PHOTOELECTRONICS
PHOTOELECTRONICS
PHOTOELECTRONICS
Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
MINISTRY OF EDUCATION AND SCIENCE OF UKRAINE<br />
Odessa I. I. Mechnikov National University<br />
<strong>PHOTOELECTRONICS</strong><br />
INTER-UNIVERSITIES SCIENTIFIC ARTICLES<br />
Founded in 1986<br />
Number 18<br />
Odessa<br />
«Astroprint»<br />
2009
UDC 621.315.592:621.383.51:537.221<br />
The results of theoretical and experimental studies in problems of optoelectronics, solar power<br />
and semiconductor material science for photoconductive materials are adduced in this collection.<br />
The prospective directions for photoelectronics are observed.<br />
The collection is introduction into the List of special editions of Ukrainian Higher Certification<br />
Comission in physics-mathematics and tehnical sciences.<br />
For lecturers, scientists, post-graduates and students.<br />
Ó çá³ðíèêó íàâåäåí³ ðåçóëüòàòè òåîðåòè÷íèõ ³ åêñïåðèìåíòàëüíèõ äîñë³äæåíü ç ïèòàíü<br />
îïòîåëåêòðîí³êè, ñîíÿ÷íî¿ åíåðãåòèêè ³ íàï³âïðîâ³äíèêîâîãî ìàòåð³àëîçíàâñòâà ôîòî -<br />
ïðîâ³äíèõ ìàòåð³àë³â. Ðîçãëÿíóòî ïåðñïåêòèâí³ íàïðÿìêè ðîçâè òêó ôîòîåëåêòðîí³êè.<br />
Çá³ðíèê âêëþ÷åíî äî Ñïèñêó ñïåö³àëüíèõ âèäàíü ÂÀÊ Óêðà¿íè ç ô³çèêî-ìàòåìàòè÷íèõ<br />
òà òåõí³÷íèõ íàóê.<br />
Äëÿ âèêëàäà÷³â, íàóêîâèõ ïðàö³âíèê³â, àñï³ðàíò³â, ñòóäåíò³â.<br />
Editorial board of “Photoelectronics”:<br />
Editor-in-Chief Smyntyna V. A.<br />
Kutalova M. I. (Odessa, Ukraine, responsible editor),<br />
Vaxman Yu. F. (Odessa, Ukraine),<br />
Litovchenko V. G. (Kiev, Ukraine),<br />
Gulyaev Yu. V. (Moscow, Russia),<br />
D’Amiko A. (Rome, Italy),<br />
Mokrickiy V. A. (Odessa, Ukraine),<br />
Neizvestny I. G. (Novosibirsk, Russia),<br />
Starodub N. F. (Kiev, Ukraine),<br />
Viculin I. M. (Odessa, Ukraine)<br />
Address of editorial board:<br />
Odessa I. I. Mechnikov National University 42, Pasteur str, Odessa, 65026, Ukraine<br />
e-mail: wadz@mail.ru, tel.: +38-0482-7266356.<br />
Information is on the site: http://www.photoelectronics.onu.edu.ua<br />
© Odessa I. I. Mechnikov National<br />
University, 2009
TABLE OF CONTENT<br />
R. V. VITER, V. A. SMYNTYNA, I. P. KONUP, YU. A. NITSUK, V. A. IVANITSA<br />
Conductivity mechanism in thin nanocryctalline tin oxide films .................................................................................................. 4<br />
G. S. FELINSKYI<br />
Noise measurements of the backward pumped distributed fiber raman amplifier .......................................................................... 9<br />
T. A. FLORKO, O. YU. KHETSELIUS, YU. V. DUBROVSKAYA, D. E. SUKHAREV<br />
Bremsstrahlung and X-ray spectra for kaonic and pionic hydrogen and nitrogen ........................................................................ 16<br />
M. V. KIRICHENKO, V. R. KOPACH, R. V. ZAITSEV, S. A. BONDARENKO<br />
Sensitivity of silicon photo-voltaic converters to the light incidence angle .................................................................................. 20<br />
L. S. MAXIMENKO, I. E. MATYASH, S. P. RUDENKO, B. K. SERDEGA, V. S. GRINEVICH, V. A. SMYNTYNA,<br />
L. N. FILEVSKAYA<br />
Spectroscopy of polarised and modulated light for nanosized tindioxide films investigation ........................................................ 24<br />
O. O. PTASHCHENKO, F. O. PTASHCHENKO, O. V. YEMETS<br />
Effect of ambient atmosphere on the surface current in silicon p-n junctions ............................................................................. 28<br />
V. A. BORSCHAK, M. I. KUTALOVA, N. P. ZATOVSKAYA, A. P. BALABAN, V. A. SMYNTYNA<br />
Dependence of space-charge region conductivity of nonideal heterojunction from photoexcitation conditions .......................... 33<br />
V. KH. KORBAN, G. P. PREPELITSA, YU. BUNYAKOVA, L. DEGTYAREVA, A. KARPENKO, S. SEREDENKO<br />
Photokinetics of the ir laser radiation effect on mixture of the CO -N -H O gases: advanced atmospheric model ....................... 36<br />
2 2 2<br />
D. À. ÊUDIY, N. P. ÊLOCHKO, G. S. KHRYPUNOV, N. À. ÊÎVTUN, K. Y. ÊRIKUN, Y. K. BELONOGOV<br />
Elaboration of cadmium sulphide film layers for economical solar cells ...................................................................................... 39<br />
I. K. DOYCHO, S. A. GEVELYUK, O. O. PTASHCHENKO, E. RYSIAKIEWICZ-PASEK, S. O. ZHUKOV<br />
Porous glasses with CdS inclusions Luminescence kinetics peculiarities ..................................................................................... 43<br />
N.V. MUDRAYA<br />
Density functional approach to atomic autoionization in an external electric field: new relativistic scheme ................................ 48<br />
V. A. SMYNTYNA, O. V. SVIRIDOVA<br />
Influence of impurities and dislocations on the value of threshold stresses and plastic deformations in silicon ............................. 52<br />
O. YU. KHETSELIUS<br />
Advanced multiconfiguration model of decay of the multipole giant resonances in the nuclei ..................................................... 57<br />
YU. F. VAKSMAN, YU. A. NITSUK, V. V. YATSUN, YU. N. PURTOV, A. S. NASIBOV, P. V. SHAPKIN<br />
Optical Properties of ZnSe:Mn Crystals .................................................................................................................................... 61<br />
A. V. GLUSHKOV<br />
Quasiparticle energy functional for finite temperatures and effective bose-condensate dynamics: theory and some<br />
illustrations ............................................................................................................................................................................... 65<br />
R. M. BALABAY, P. V. MERZLIKIN<br />
Electronic structure of heterogeneous composite: organic molecule on silicon thin film surface ................................................. 70<br />
A. A. SVINARENKO, A. V. LOBODA, N. G. SERBOV<br />
Modeling and diagnostics of interaction of the non-linear vibrational systems on the basis of temporal series<br />
(Application to semiconductor quantum generators) ................................................................................................................. 76<br />
YE. V. BRYTAVSKYI, YU. N. KARAKIS, M. I. KUTALOVA, G. G. CHEMERESYUK<br />
Effects connected with interaction of charge carriers and r-centers basic and exited states .......................................................... 84<br />
I. N. SERGA<br />
Electron internal conversion in the 125,127Ba isotopes .................................................................................................................. 88<br />
L. N. VILINSKAYA, G. M. BURLAK<br />
Sensors on the basis of aluminium metal-oxide films ................................................................................................................. 92<br />
O. O. PTASHCHENKO, F. O. PTASHCHENKO, V. V. SHUGAROVA<br />
Tunnel surface current in GaAs–AlGaAs p-n junctions, due to ammonia molecules adsorption ................................................. 95<br />
SH. D. KURMASHEV, T. M. BUGAEVA, T. I. LAVRENOVA, N. N. SADOVA<br />
Influence of the glass phase structure on the resistance of the layers in system “glass-RuO ” ....................................................... 99<br />
2<br />
A. V. IGNATENKO, A. A. SVINARENKO, G. P. PREPELITSA, T. B. PERELYGINA, V. V. BUYADZHI<br />
Optical bi-stability effect for multi-photon absorption in atomic ensembles in a strong laser field ..............................................103<br />
L. V. MYKHAYLOVSKA, A. S. MYKHAYLOVSKA<br />
Influence of the step ionization processes on the electronic temperature in thin gas-discharge tubes ..........................................106<br />
E. V. MISCHENKO<br />
Quantum measure of frequency and sensing the collisional shift and broadening of Rb hyperfine lines in medium<br />
of helium gas ...........................................................................................................................................................................112<br />
O. O. PTASHCHENKO, F. O. PTASHCHENKO, N. V. MASLEYEVA, O, V. BOGDAN<br />
Surface current in GaAs p-n junctions, passivated by sulphur atoms .........................................................................................115<br />
A. V. GLUSHKOV, YA. I. LEPIKH, A. P. FEDCHUK, A. V. LOBODA<br />
The green’s functions and density functional approach to vibrational structure in the photoelectron spectra of molecules ..........119<br />
A. V. TYURIN, A. YU. POPOV, S. A. ZHUKOV, YU. N. BERCOV<br />
Mechanism of spectral sensitizing of the emulsion containing heterophase “core –shell” microsystems ....................................128<br />
V. V. KOVALCHUK, O. V. AFANAS’EVA, O. I. LESHCHENKO, O. O. LESHCHENKO<br />
Size distributions of clusters on photoluminescence from ensembles of Si-clusters ....................................................................133<br />
I. M. VIKULIN, SH. D. KURMASHEV, P. YU. MARKOLENKO, P. P. GECHEV<br />
Radiation immunity of the planar n-p-n-transistors .................................................................................................................136<br />
K. V. AVDONIN<br />
Build-up of wave functions of the particle in the modelling periodic field ..................................................................................140<br />
V. A. ZAVADSKY, G. S. POPIK<br />
Modification of parameters IR-fotodetectors by high-energy particles ......................................................................................147<br />
Information for contributors of “Photoelectronics” articles ............................................................................................................151<br />
²íôîðìàö³ÿ äëÿ àâòîð³â çá³ðíèêà “Ôîòîåëåêòðîí³êà” ...............................................................................................................151<br />
Èíôîðìàöèÿ äëÿ àâòîðîâ ñáîðíèêà “Ôîòîýëåêòðîíèêà” .........................................................................................................152<br />
3
4<br />
UDC 621.315.592<br />
R. V. VITER 1 , V. A. SMYNTYNA 1 , I. P. KONUP 2 , YU. A. NITSUK 1 , V. A. IVANITSA 2<br />
1 Department of Experimental Physics, Odessa National University, 42, Pastera str., 65026, Odessa, Ukraine,<br />
viter_r@mail.ru; phone +38-0676639327, fax:+380-48-7233515.<br />
2 Department of Mycrobiology, Odessa National University, 2, Shampansky lane, 65000, Odessa, Ukraine,<br />
phone +38-0482-68-79-64.<br />
CONDUCTIVITY MECHANISM IN THIN NANOCRYCTALLINE<br />
TIN OXIDE FILMS<br />
Structural properties of tin oxide nanocryastalline films have been investigated by means of atomic<br />
force microscopy (AFM) and X-ray diffraction (XRD) methods. Surface morphology, roughness,<br />
crystalline size and lattice strain have been estimated. Current-voltage characteristics (I-V) have been<br />
measured at different temperatures. Temperature dependence of current has been studied. Activation<br />
energies have been evaluated and conductivity mechanism has been proposed.<br />
1. INTRODUCTION<br />
Tin oxide SnO 2 is well known as material for gas<br />
sensors [1-8]. The most important reasons of tin oxide<br />
use in sensor applications are chemical stability<br />
to different aggressive chemical pollutants and high<br />
temperature treatment [1]. Those advantages allow<br />
fabricating different types sensors, based on tin oxide<br />
to different gases [2]. Another application of tin oxide<br />
thin films is optics where they have been successfully<br />
used as transparent conducting electrodes in optical<br />
devises [2-7]. Tin oxide thin films have been successfully<br />
used for measurements in liquids to detect ammonia<br />
in water [2].<br />
It was published that tin oxide films consisting of<br />
nanoparticles showed different properties from typical<br />
polycrystalline films [3-8]. The optical characterization<br />
of the films was performed in [4]. The thickness<br />
and refractive index have been calculated. The crystalline<br />
size was estimated by means of optical methods<br />
using the absorption spectra [4]. It was observed blue<br />
shift of optical absorption spectra in comparison with<br />
polycrystalline samples [4]. The value of band gap estimated<br />
from optical absorption spectra was 0,2-0,6<br />
eV bigger, than to tin oxide single crystal (E g =3,6 eV).<br />
Electrical characterization of nanocrystalline tin<br />
oxide films has been performed in [3, 4]. No Shotky<br />
barriers have been observed and non ohmic behavior<br />
was verified [3]. However, the correct explanation of<br />
charge transfer in tin oxide tin oxide nanocrystalline<br />
films has not been performed.<br />
In this work experimental results of investigation<br />
of electrical properties are reported. Current-voltage<br />
and temperature dependence of current have been<br />
performed. Results of structural properties of the films<br />
have been reported. Activation energies were determined.<br />
Conductivity mechanism in tin oxide nanocrystalline<br />
films has been proposed.<br />
2. EXPERIMENTAL<br />
Tin oxide thin films were deposited with electrostatic<br />
spray pyrolysis technique, described in [1-3].<br />
For deposition, tin chloride (IV) ethanol solution was<br />
used [2]. Tin chloride concentration of sprayed solution<br />
and sprayed solution volume were kept constant<br />
and equaled c=0,01 mol/l and v=10 ml, correspondently.<br />
Glass substrates, with pretreatment in ethanol<br />
and ultrasonic bath, were used for films’ fabrication.<br />
Applied static voltage between capillary and glass substrate<br />
was 17 kV. After deposition, the obtained samples<br />
have been annealed at 793 K during 1 hour.<br />
I-V characterization was measured in the range<br />
of 0-200 V under different temperatures 293-393 K.<br />
Temperature dependence of current was performed at<br />
the same temperature range and with applied voltage<br />
kept constant 60 V.<br />
Atomic Force Microscopy (AFM) has been performed<br />
on the deposited SnO 2 layers in order to investigate<br />
the surface morphology of the films.<br />
XRD measurements have been performed with<br />
Philips X’Pert-MPD (CuK α , λ=0,15418 nm) difractometer<br />
to identify the nature of deposited material<br />
and determine crystalline size.<br />
Fig. 1. AFM image of tin oxide film.<br />
© R. V. Viter, V. A. Smyntyna, I. P. Konup, Yu. A. Nitsuk, V. A. Ivanitsa, 2009
3. RESULTS AND DISCUSSION<br />
The thickness of obtained films, estimated by<br />
means of profilometer Tencor P7, was 310 nm. AFM<br />
images of tin oxide nanocrystalline films are presented<br />
in figures 1, 2. The images refer to 5x5 μm 2 and 800x800<br />
nm 2 areas of tin oxide surface. The one can see that<br />
the film had polycrystalline structure with well shaped<br />
grains. Wiskers of 200-250 nm height were observed on<br />
the surface of the film. It points to high concentration<br />
of point defect on the surface of thin films [2]. Surface<br />
roughness (Rms) of the films was 26,2 nm, what seems<br />
to be suitable for sensor application.<br />
XRD data is presented in figure 3. The one can see<br />
peaks at 2θ: 26,5 , 34,5, 37,8 , 51,4, corresponding to<br />
tetragonal crystalline phase of tin oxide and one peak<br />
at 2θ=65,2, which represents orthorhombic phase of<br />
tin oxide [7,8].<br />
Crystalline size and lattice strain have been determined<br />
in figure 4, according to equation [7, 8]:<br />
() 0,9 ()<br />
β⋅cos θ ε⋅sin θ<br />
= +<br />
λ d λ<br />
Fig. 2. 2-DAFM image of surface of tin oxide films.<br />
(1)<br />
Previously [4], the crystalline size of tin oxide, deposited<br />
at the same conditions, determined from optical<br />
absorption spectra was 5,2 nm. However analysis of<br />
the AFM data showed surface agglomerates with average<br />
size of about 20 nm (fig. 2). On the other hand,<br />
crystalline size value, determined by XRD method,<br />
was compatible with optical absorption data. Similar<br />
behavior has been observed in [7], when electron<br />
microscopy images gave crystalline size of 100 nm<br />
whereas XRD analysis showed particles with 10 nm<br />
size. This phenomenon can be explained by formation<br />
of agglomerates by low size crystallites.<br />
I-V characteristics are presented in fig.5. In order<br />
to analyze charge transfer mechanism they have<br />
been plotted in different scales (fig.6, 7). At low voltages<br />
(U
I, mkA<br />
ln(J/AT 2 )<br />
6<br />
25<br />
20<br />
15<br />
10<br />
5<br />
y<br />
20 o C<br />
40 o C<br />
60 o C<br />
80 o C<br />
100 o C<br />
120 o C<br />
0<br />
0 20 40 60 80 100 120 140 160 180 200<br />
-16<br />
-17<br />
-18<br />
-19<br />
-20<br />
-21<br />
-22<br />
-23<br />
U, V<br />
Fig. 5. I-V plots of tin oxide thin films.<br />
U 1/2 , V 1/2<br />
T=293 K<br />
T=313 K<br />
T=333 K<br />
T=353 K<br />
T=373 K<br />
T=393 K<br />
2 4 6 8 10 12 14 16<br />
Fig. 6. I-V plots, rebuilt in Frenkel’s-Pool scale.<br />
With increase of applied voltage (U>50 V) measured<br />
I-V data showed Ohmic behavior. Only at T>353<br />
K nonlinear part was observed.<br />
Temperature dependence of current, measured<br />
under constant value of applied voltage U=60 V, was<br />
1<br />
plotted in ln I ~ scale and two linear parts were<br />
T<br />
found (fig.8). Activation energy values were 0,16 eV<br />
and 0,24 eV for low and high temperature regions<br />
correspondently. The activation energies E =0,16 eV<br />
1<br />
and E =0,24 eV correspond to double ionized oxygen<br />
2<br />
vacancies and defect states [8]. The one can see good<br />
correlation between energy values determined from<br />
I-V measurements and temperature dependences of<br />
current. In both cases the same surface states have<br />
been observed.<br />
ln I (mkA)<br />
3<br />
2<br />
1<br />
0<br />
-1<br />
-2<br />
-3<br />
-4<br />
ln U (V)<br />
1,5 2,0 2,5 3,0 3,5 4,0 4,5 5,0 5,5<br />
Fig. 7. I-V plots, rebuilt in double logarithm scale.<br />
CONCLUSION<br />
293 K<br />
313 K<br />
333 K<br />
353 K<br />
373 K<br />
393 K<br />
Electrical and structural properties of tin oxide<br />
nanocrystalline films have been investigated. AFM<br />
analysis showed that the obtained films had polycrystalline<br />
nature with rough surface and wiskers,<br />
what makes these films attractive for sensor applications.<br />
XRD measurements showed peaks, typical for tin<br />
oxide. Crystalline size, determined from XRD measurements,<br />
was 5,54 nm. T<br />
I-V data showed two main charge transfer mechanisms.<br />
Under applied voltages U50 V the Ohm’s mechanism dominates.<br />
The temperature dependence of current had two<br />
linear parts in Arrenius scale. The activation energies<br />
E 1 =0,16 eV and E 2 =0,24 eV concerned with oxygen<br />
vacancies and surface state defects.
ln(I), mkA<br />
-12<br />
-13<br />
-14<br />
0,0026 0,0028 0,0030 0,0032 0,0034<br />
1/T, K -1<br />
Fig. 8. Temperature dependence of current of tin oxide nanocrystalline<br />
films.<br />
UDC 621.315.592<br />
R. V. Viter, V. A. Smyntyna, I. P. Konup, Yu. A. Nitsuk, V. A. Ivanitsa<br />
CONDUCTIVITY MECHANISM IN THIN NANOCRYCTALLINE TIN OXIDE FILMS<br />
References<br />
1. Viter R., Smyntyna V., Evtushenko N., Structural properties<br />
of nanocrystalline tin dioxide films deposited by electrostatic,<br />
spray pyrolisis method // Photoelectronics. — 2005. —<br />
Vol. 15. — p.54-57<br />
2. M. Pisco, M. Consales, R. Viter, V. Smyntyna, S. Campopiano,<br />
M. Giordano, A. Cusano, A.Cutolo, Novel SnO based<br />
2<br />
optical sensor for detection of low ammonia concentrations<br />
in water at room temperatures // Intern. Sc. J. Semiconductor<br />
Physics, Quantum Electronics and Optoelectronics. —<br />
2005. — Vol. 8. — p.95-99<br />
3. A. N. Banerjee, R. Maity, S. Kundoo, and K. K. Chattopadhyay,<br />
Poole–Frenkel effect in nanocrystalline SnO :F thin<br />
2<br />
films prepared by a sol–gel dip-coating technique// phys.<br />
stat. sol. (a) -2004. — Vol. 204. — No. 5. — p. 983–989<br />
4. R.V. Viter, V.A. Smyntyna, Yu. A. Nitsuk, Optical, electrical<br />
and structural characterization of thin nanocryctalline SnO2 films for optical fiber sensors application // Proceedings of<br />
Test sensor conference 2007, Nuremberg, Germany. — May<br />
2007. — pp. 1252-1257<br />
5. Feng Gu, Shu Fen Wang, Meng kai Lu, Guang Jun Zhuo,<br />
Dong Xu and Duo Rong Yuan, Photoluminescence properties<br />
of SnO nanoparticles synthesized by sol-gel method// J.<br />
2<br />
Phys. Chem. B. — 2004. — Vol. 108. — p. 8119-8123<br />
6. Shanthi S., Subramanian C., Ramasamy P., Preperation and<br />
properties of sprayed undoped and fluorine doped tin oxide<br />
films// Materials Science and engineering, B. — 1999. —<br />
Vol. 57. — p. 127-134<br />
7. Yuji Matsui, Michio Mitsuhashi , Yoshio Goto, Early stage<br />
of tin oxide film growth in chemical vapor deposition // Surface<br />
and Coatings Technology. — 2003. — Vol.169 –170. —<br />
p. 549–552<br />
8. A.K. Mukhopadhyay, P. Mitra, A.P. Chatterjee, H.S. Maiti,<br />
Tin dioxide thin flm gas sensor// Ceramics International. —<br />
2000. — Vol. 26. — p. 123-132<br />
Abstract<br />
Structural properties of tin oxide nanocryastalline films have been investigated by means of atomic force microscopy (AFM) and<br />
X-ray diffraction (XRD) methods. Surface morphology, roughness, crystalline size and lattice strain have been estimated. Current-voltage<br />
characteristics (I-V) have been measured at different temperatures. Temperature dependence of current has been studied. Activation<br />
energies have been evaluated and conductivity mechanism has been proposed.<br />
Key words: tin oxide, nanocrystalline films, I-V characterization, XRD, AFM.<br />
ÓÄÊ 621.315.592<br />
Ð. Â. Âèòåð, Â. À. Ñìûíòûíà, È. Ï. Êîíóï, Þ. À. Íèöóê, Â. À. Èâàíèöà<br />
ÌÅÕÀÍÈÇÌ ÏÐÎÂÎÄÈÌÎÑÒÈ Â ÒÎÍÊÈÕ ÍÀÍÎÊÐÈÑÒÀËËÈ×ÅÑÊÈÕ Ï˨ÍÊÀÕ ÎÊÑÈÄÀ ÎËÎÂÀ<br />
Ðåçþìå<br />
Ñòðóêòóðíûå ñâîéñòâà íàíîêðèñòàëëè÷åñêèõ ïë¸íîê îêñèäà îëîâà áûëè èçó÷åíû ïðè ïîìîùè ìåòîäîâ àòîìíîé ñèëîâîé<br />
ìèêðîñêîïèè è äèôðàêöèè ðåíòãåíîâñêîãî èçëó÷åíèÿ. Áûëè îïðåäåëåíû ìîðôîëîãèÿ ïîâåðõíîñòè, âåëè÷èíû åå øåðîõîâàòîñòè,<br />
ðàçìåðîâ êðèñòàëëèòîâ è ìåõàíè÷åñêîãî íàïðÿæåíèÿ êðèñòàëëè÷åñêîé ðåøåòêè. Âîëüò-àìïåðíûå õàðàêòåðèñòèêè<br />
îáðàçöîâ áûëè èçó÷åíû ïðè ðàçíûõ òåìïåðàòóðàõ. Òåìïåðàòóðíàÿ çàâèñèìîñòü òåìíîâîãî òîêà áûëà èçó÷åíà. Ýíåðãèè àêòèâàöèè<br />
ïðîâîäèìîñòè áûëè îïðåäåëåíû.<br />
Êëþ÷åâûå ñëîâà: îêñèä îëîâà, âîëüò-àìïåðíûå õàðàêòåðèñòèêè, àòîìíàÿ ñèëîâàÿ ìèêðîñêîïèÿ è äèôðàêöèÿ ðåíòãåíîâñêîãî<br />
èçëó÷åíèÿ.<br />
7
8<br />
ÓÄÊ 621.315.592<br />
Ð. Â. ³òåð, Â. À. Ñìèíòèíà, ². Ï. Êîíóï, Þ. À. ͳöóê, Â. Î. ²âàíèöÿ<br />
ÌÅÕÀͲÇÌ ÏÐβÄÍÎÑÒ²  ÒÎÍÊÈÕ ÍÀÍÎÊÐÈÑÒÀ˲×ÍÈÕ Ï˲ÂÊÀÕ ÎÊÑÈÄÓ ÎËÎÂÀ<br />
Ðåçþìå<br />
Ñòðóêòóðí³ âëàñòèâîñò³ íàíîêðèñòàë³÷íèõ ïë³âîê îêñèäó îëîâà áóëî äîñë³äæåíî çà äîïîìîãîþ ìåòîä³â àòîìíî¿ ñèëîâî¿<br />
ì³êðîñêîﳿ òà äèôðàêö³¿ ðåíòãåí³âñüêîãî âèïðîì³íþâàííÿ. Áóëî âèçíà÷åíî ìîðôîëîã³ÿ ïîâåðõí³, âåëè÷èíè ¿¿ íåîäíîð³äíîñò³,<br />
ðîçì³ð³â êðèñòàë³ò³â òà ìåõàí³÷íîãî íàïðóæåííÿ êðèñòàë³÷íî¿ ãðàòêè. Âîëüò-àìïåðí³ õàðàêòåðèñòèêè çðàçê³â áóëî äîñë³äæåíî<br />
ïðè ð³çíèõ òåìïåðàòóðàõ. Òåìïåðàòóðíó çàëåæí³ñòü òåìíîâîãî ñòðóìó áóëî ïîáóäîâàíî. Åíåð㳿 àêòèâàö³¿ ïðîâ³äíîñò³<br />
áóëè âèçíà÷åí³.<br />
Êëþ÷îâ³ ñëîâà: îêñèä îëîâà, âîëüò-àìïåðí³ õàðàêòåðèñòèêè, àòîìíà ñèëîâà ì³êðîñêîï³ÿ òà äèôðàêö³ÿ ðåíòãåí³âñüêîãî<br />
âèïðîì³íþâàííÿ.
UDC 535.361:621.391.822<br />
G. S. FELINSKYI<br />
Radiophysics Faculty, Kyiv Taras Shevchenko National University,<br />
Glushkova Prospect, 2, 03127 Kyiv, Ukraine, Phone: +380-44-526-0570,<br />
Fax: +380-44-526-0531, E-mail: felinskyi@yahoo.com<br />
NOISE MEASUREMENTS OF THE BACKWARD PUMPED DISTRIBUTED<br />
FIBER RAMAN AMPLIFIER<br />
INTRODUCTION<br />
Creation of light amplifiers based on stimulated<br />
Raman scattering (SRS) in singlemode fibers is legally<br />
refer to the most powerful practical achievements [1]<br />
arisen as a result of long time fundamental researches<br />
in nonlinear optics. At first practical applications of<br />
fiber Raman amplifiers (FRA) with several pumping<br />
sources at the end of the last century [2] high-quality<br />
amplification of optical signals with a bandwidth<br />
close to limiting for silica fibers about 13 THz has been<br />
shown. Due to improvement of operational characteristics<br />
now FRA gradually supersede other types of<br />
optical amplifiers from the ultra wide band communication<br />
systems and they already became the first nonlinear<br />
optics device which has received wide practical<br />
application in long-distance optical fiber communication<br />
with terabit capacity.<br />
Such amplifiers are widely applied despite of the<br />
actual interdiction imposed by the theory. Really<br />
modern theory [3-5] predicts that the noise figure of<br />
any optical amplifier should be higher than the minimal<br />
quantum limit of 3 dB. It means that the signal to<br />
noise ratio (SNR) after the amplifier should decrease<br />
at least in 2 times. The bit error rate (BER) in turn directly<br />
depends on the signal to noise ratio. In addition<br />
the noise statistics is those that the SNR level above 6<br />
dB each decibel increasing or reducing this ratio creates<br />
reduction or accordingly increases the BER at the<br />
order on value. Hence each optical amplifier, including<br />
FRA, theoretically should appreciably increase the<br />
BER during the digital information transfer. However<br />
there is an obvious contradiction between theoretical<br />
performances about optical amplifiers noise and real<br />
practice of their application in the optical fiber systems.<br />
In particular, standard optical communication<br />
scheme of linear signal regeneration using of sequential<br />
application of optical amplifiers becomes theoretically<br />
impossible without full signal restoration from<br />
noise, in view of fast error accumulation with growth<br />
of amplifiers amount. Even more ten years ago it has<br />
been marked [5], that the physical phenomena which<br />
© G. S. Felinskyi, 2009<br />
Noise parameters of fiber Raman amplifiers are appeared in practice essentially better top theoretical<br />
limits and the amplifier noise performance is the subject of intensive investigations now. the<br />
experimental results on the amplified spontaneous emission (ASE) observation in the single mode<br />
fiber span using the backward pumped distributed FRA are presented in the report. Our measurements<br />
and quantity analysis show that physical principles of formation of the FRA optical noise with the<br />
counter pumping lead to very small gain coefficients and it almost correspond to spontaneous Raman<br />
properties in the silica fibers. Raman gain nonlinearity is the reason to reduce the real FRA noise figure<br />
much below the established 3 dB limit. Our analysis of experimental data on ASE measurements<br />
in silica fibers allows us to get the information about the nature of formation mechanism of the noise<br />
parameters in the practical FRA.<br />
result in extraordinary rare occurrence of mistakes in<br />
digital optical communication systems till now yet<br />
have not received unequivocal interpretation. Now the<br />
same remarks can be related to the noise performance<br />
of practical FRA.<br />
The experimental results of the amplified spontaneous<br />
emission (ASE) research using distributed FRA<br />
in singlrmode fiber with the counter pumping are presented<br />
in this work.<br />
FRA APPLICATION PROBLEMS AND MOTIVATION TO EX-<br />
PERIMENTS<br />
Technical specifications of optical signal amplifiers<br />
in telecommunication systems essentially influence<br />
on design strategy of modern fiber optic highways<br />
for an information exchange. The generalized quality<br />
parameter of any digital communication system is the<br />
bit error rate B er which is unequivocally determined by<br />
the signal to noise ratio Q [6]:<br />
1<br />
Ber = erfc( Q/<br />
2) , (1)<br />
2<br />
where erfc(x) is error function. Therefore to increase<br />
of signal transfer distance in the communication<br />
link one must not so much try to restore the<br />
power losses of a signal due to attenuation in a link,<br />
but mainly the maintenance of necessary value Q,<br />
for example not less than 6 dB, that corresponds<br />
B =10 er -9 .<br />
Two linear regeneration schemes of the signal in<br />
long optical fiber communication links are shown<br />
on Fig. 1. The scheme on Fig. 1à using the linear repeaters<br />
with full regeneration of a signal (Fig. 1b) is<br />
the standard application in microwave multichannel<br />
communication. Such scheme have no alternative<br />
for the microwave systems of multichannel communication<br />
because of restrictions on receiver sensitivity<br />
by own noise at its input. The cascade amplification<br />
scheme using broadband optical amplifiers (Fig. 1â) is<br />
extremely simple due to huge reduction of the equipment<br />
amount and it becomes dominating for modern<br />
high-speed optical links but this design is not theoreti-<br />
9
cally possible if the noise figure of optical amplifier, in<br />
particular FRA, considerably exceeds unit.<br />
10<br />
(a) Repeaters<br />
(Rp)<br />
Tx<br />
N×λn<br />
(b)<br />
In<br />
N×λn<br />
Demultiplexer<br />
N×λn<br />
Rp Rp Rp<br />
Repeater Block Diagram<br />
Rx 1<br />
Rx 2<br />
Rx 3<br />
...<br />
Rx N<br />
Electronic<br />
Signal<br />
Restoration<br />
Tx 1<br />
Tx 2<br />
Tx 3<br />
...<br />
Tx N<br />
Multiplexer<br />
N×λn<br />
Rx<br />
N×λn<br />
Out<br />
N×λn<br />
appears above permissible maximum and it is create<br />
the contradiction with the resulted theoretical performances<br />
about optical amplifier noise. Last circumstance<br />
practically specifies illegitimacy of direct carry<br />
the models developed for electronic amplifiers on such<br />
nonlinear photon system as FRA. As the our purpose<br />
to solve the problem about the FRA optical noise have<br />
been established, the analysis of Raman gain features<br />
in silica fibers and experimental results on amplified<br />
spontaneous emission is resulted in the present work.<br />
The noise analysis of optical Raman gain may<br />
be relieved of restrictions of traditional methods of<br />
equivalent circuits by direct experimental noise measurements<br />
of amplified spontaneous emission (ASE)<br />
in commercial FRA model with the multiwave pumping<br />
and using the theory of Raman interaction physics<br />
as nonlinear optical process.<br />
THEORETICAL BASIS OF RAMAN GAIN<br />
ANALYSIS<br />
(c) Optical amplifiers (OA)<br />
Physical source of the Raman scattering radiation<br />
is nonlinear polarization P<br />
Tx<br />
N×λn<br />
OA OA OA<br />
Rx<br />
N×λn<br />
Fig. 1. Optical signal transmission based on: a) set of recovery<br />
repeaters, for which is shown the block diagram (b) of group signal<br />
restoration equipment with wavelength division multiplexing; â)<br />
set of wideband optical amplifiers.<br />
Really amplifiers quality can be characterized by<br />
noise figure of amplifier F=Q /Q . Parameter F n in out n<br />
shows how times the input SNR Q is changed by the<br />
in<br />
amplifier due to own internal noise in relation to output<br />
SNR Q . Noise figure of the fiber amplifiers based<br />
out<br />
on inversion of population densities in laser transition<br />
is<br />
Fn = 2 nsp( G−1)/ G ≈ 2nsp<br />
, (2)<br />
where nsp = N2 / ( N2 − N1)<br />
is so called the spontaneous<br />
emission factor or population inversion factor, N1 and N are the population density of the bottom and<br />
2<br />
top levels for laser transition, accordingly, G is the gain<br />
coefficient. The equation (2) shows that at big G SNR<br />
for amplified signal is degrades by a factor 2 (or 3 äÁ)<br />
even for an ideal amplifier, for which n = 1. The prac-<br />
sp<br />
tical amplifier with n > 1 should have F ≥ 2, and the<br />
sp n<br />
minimal value of 2 (3 dB) in the modern theory [3-5]<br />
is considered a quantum limit. However the physics of<br />
processes in FRA is not deals with the population inversion<br />
of electronic energy levels. Fundamental FRA<br />
basis is the nonlinear optics and description of amplifier<br />
noise performance is required not only other theoretical<br />
approaches, but also additional experimental<br />
researches.<br />
Today FRA application practice has revealed some<br />
problems, at least, two of which have fundamental<br />
character. First, extreme simplicity of FRA technical<br />
realization is accompanied by complexity of physical<br />
processes of a nonlinear energy exchange between<br />
several pumping sources and hundreds information<br />
channels and it essentially complicates modeling and<br />
designing of devices. Second, signal quality after FRA<br />
NL induced by electric field<br />
Ep of the pumping wave. Relative power of scattered<br />
radiation on Stokes frequency ω due to the pump ac-<br />
s<br />
tion of the frequency ω is quantitatively described by<br />
p<br />
differential Raman cross section [7]:<br />
2<br />
d σ<br />
dΩdωs 3 2 NL 2<br />
ωω s pnV s < | P | > ωs<br />
=<br />
,<br />
2 2 4 p<br />
2<br />
16πε0cnp<br />
E<br />
(3)<br />
where V is the volume of scattering, ε is the dielec-<br />
0<br />
tric constant, dΩ is a solid angle element, n , n are the<br />
s p<br />
refractive indexes on frequencies ω and ω . Angular<br />
s p<br />
brackets designate the time averaging.<br />
Generally the i-th component of the induced polarization<br />
in isotropic media of fiber core irrespective<br />
from the frequency of nonlinear interaction products<br />
may be written using the third order susceptibility ten-<br />
(3)<br />
sor χ as:<br />
NL<br />
(3)<br />
P =ε<br />
i 0 ∑ χijkl<br />
EjEkEl. (4)<br />
jkl , ,<br />
If our consideration will be limited to the most<br />
general fiber types which have the weak waveguide<br />
properties it is possible to divide the electric field both<br />
the Stokes wave (s), and the pumping wave (p) into<br />
transverse part Ri 2 2<br />
(r), (where i = s, p, and r = x + y )<br />
and on the function from z: E (z). In particular the<br />
i<br />
polarization of mode on the frequency ω is arisen at<br />
s<br />
nonlinear mixing of electric fields as:<br />
ω s P<br />
i<br />
= 6 ε0 p<br />
2<br />
s (3) p p * s<br />
R R ∑ χijkl<br />
Ej( Ek) El<br />
jkl , ,<br />
(5)<br />
with complex conjugation on frequency − ω . Accord-<br />
s<br />
ingly the equation for third order susceptibility tensor<br />
(3)<br />
χ is:<br />
*<br />
(3) N 1 ∂αij ⎛∂α ⎞ kl<br />
χ ijkl = ⋅ 2 2 ∑ ⎜ ⎟ , (6)<br />
12mε0V ωv−ω + 2iωγ<br />
n ∂qn ⎝ ∂qn<br />
⎠<br />
where N is the oscillator quantity in the volume of interaction<br />
V, m is the oscillator mass, ω and ω are res-<br />
v
onant and current frequencies of molecular vibration<br />
with attenuation γ, ij k q ∂α ∂ is the differential polarizability<br />
(Raman tensor), q r is the displacement vector<br />
for molecular vibration.<br />
It is well known that the Raman active vibration q r<br />
should be not active in infrared (IR) absorption as it<br />
is not accompanied by the local dipole momentum at<br />
a molecular level in this case isotropic media with the<br />
inversion center. It means that thermal fluctuations of<br />
radiation in a fiber cannot influence directly changes<br />
phonon density of the Raman active vibrations. There<br />
is the blocking of the thermal IR noise influence on<br />
Raman scattering process in the fibers in contrast with<br />
any other material having the inversion center in its<br />
molecules.<br />
The vector q r depends on E r as a corresponding<br />
combination of the pumping and Stokes wave fields<br />
in the case of the stimulated Raman scattering which<br />
creates synchronous external force and it causes resonant<br />
behavior of the given vibration. Thus there is an<br />
amplification of power of a Stokes wave in fiber and<br />
the gain coefficient g looks as<br />
R<br />
g<br />
(3) (3)<br />
3ω<br />
Im[ χ iiii +χijji<br />
]<br />
s<br />
R =− 2<br />
ps<br />
ε0cnn<br />
p s 2Aeff<br />
, (7)<br />
ps<br />
where A eff is the effective area of the pump and signal<br />
overlapping region. The gain frequency profile g (ω) R<br />
defines the dynamics of Raman amplification. Interaction<br />
between Stokes intensity I (z, ω) and mono-<br />
s<br />
chromatic pumping I (z) in arbitrary fiber coordinate<br />
s<br />
z is described by the coupled equations [6]:<br />
dIs (, z ω )<br />
= gR( ω) Ip( z) Is( z, ω) −αs Is( z,<br />
ω ) , (8)<br />
dz<br />
dI p( z)<br />
ωp<br />
= gR( ω) Ip( z) Is( z, ω) −αpIp(<br />
z)<br />
, (9)<br />
dz ωs<br />
The classical equations (8) and (9) were almost<br />
exclusively used for the distributed FRA description.<br />
Unfortunately, analytical solutions of these equations<br />
are known only in the limited special cases even<br />
when interaction only of two waves is considered.<br />
Such special case is the approximation with no pump<br />
depletion. This approach [8] allows obtaining the analytical<br />
expressions for gain coefficients of the Stokes<br />
signal and ASE. However these equations cannot be<br />
applied to direct calculations of the amplifier noise<br />
figure as such essential nonlinear Raman features<br />
as an amplification threshold and a nonlinear mode<br />
concurrence are remained behind of the made restriction<br />
frameworks.<br />
The quantum dynamical equation for the photon<br />
number η (z) on the unit length z for Stokes photons<br />
s<br />
ω = ω – ω is [9]:<br />
s p v<br />
dη<br />
ωω<br />
s<br />
p s<br />
= C ρ( hω<br />
f ) ×<br />
dz<br />
ωv<br />
× { η ( η + 1) η −η η η −η η + ( η + 1) η } , (10)<br />
s v p s v p s v v p<br />
=∂αij ∂<br />
2<br />
k ( π<br />
2<br />
) (4<br />
2<br />
εε s p ) h<br />
where C q h V Nm v , = h 2π<br />
is Plank’s constant, ε , ε are dielectric constants for<br />
s p<br />
pumping and Stokes waves, respectively, v is phase velocity<br />
of Stokes wave.<br />
Rate equation (10) for the Stokes scattered photons<br />
consists of four terms. At the quantum analysis first two<br />
terms are often referred to as stimulated emission and<br />
stimulated absorption. Last two terms are the spontaneous<br />
absorption, and spontaneous emission, respectively.<br />
Since the phonons are assumed to be in equilibrium<br />
at temperature T, the occupation number<br />
η (ω) is the thermal equilibrium number<br />
v<br />
( ) kT<br />
ηv( ω v) = ⎡⎣exp h ωv B −1⎤⎦<br />
, where k is Boltzmann’s<br />
B<br />
constant. It should be noted that the spontaneous emission<br />
terms are proportional to n +1 for Stokes photon<br />
v<br />
and thus depends on the temperature of the fiber.<br />
The equation (10) shows that the difference in<br />
stimulated emission and absorption terms does not<br />
depend on the phonon number η and therefore is<br />
v<br />
temperature-independent. In fact, the frequency profile<br />
g (ν) may be expressed by the spontaneous Raman<br />
R<br />
cross section at zero temperature σ (ν) as [10]:<br />
0<br />
−1<br />
λ<br />
g v =σ v ⋅ , (11)<br />
R () 0 ()<br />
3<br />
s<br />
2 ps 2<br />
chAeff np<br />
where ν =ω/2πñ is the wave number, c is speed of light,<br />
and weak frequency dependence of a pumping wave<br />
refractive index n in the Stokes shifted area can be ne-<br />
p<br />
glected. The spontaneous Raman cross section σ (v) Ò<br />
at temperature Ò is related to zero Kelvin cross section<br />
σ (ν) as:<br />
0<br />
σ0 ( ν ) =σT( ν)/[ ηv( ν , T ) + 1] , (12)<br />
The spontaneous Raman spectrum and Raman<br />
gain profile for standard silica fiber is shown on<br />
Fig. 2. The essential difference between spontaneous<br />
Raman spectrum and Raman gain profile according<br />
to (11) — (12) and the data on Fig. 2 should be<br />
observed in the frequency region of Stokes shift less<br />
than 6 THz=200 cm-1 where the thermal density factor<br />
of phonon numbers essentially exceeds the unit. In<br />
more high-frequency area the thermal density factor<br />
of Stokes phonon numbers (> 200 cm-1 ) lose its frequency<br />
dependence, practically not differing from<br />
unit and consequently spontaneous Raman spectrum<br />
coincides with the Raman gain profile.<br />
Normalized Intensity,<br />
1.0<br />
0.5<br />
0<br />
Raman Scattering<br />
Spontaneous<br />
Stimulated<br />
0 200 400<br />
Raman shift, cm<br />
600<br />
-1<br />
Fig. 2. Spontaneous Raman spectrum (solid line) and Raman<br />
gain profile (doted line) in Stokes shifted frequency area from 0 to<br />
21 THz (700 cm-1 ) for standard silica fiber.<br />
Earlier we have studied [10, 11] ASE spectra in<br />
comparison to experimental data of measurements of<br />
11
the effective noise figure in the multi wavelength Raman<br />
amplifier [12]. On the preliminary simulation data<br />
the FRA noise is mainly formed in the fiber part near to<br />
the pumping source where the pump power is maximal.<br />
During the distribution on a fiber full pumping power<br />
fades, the pumping gradually starts to be exhausted and<br />
its powers not enough for effective ASE generation.<br />
Such modeling [13] allows obtaining the information<br />
on the noise properties formation in the real fiber optical<br />
amplifiers. These preliminary conclusions prove to<br />
be true our experimental ASE measurements.<br />
12<br />
EXPERIMENTAL SETUP AND<br />
MEASUREMENT METODS<br />
Experimental observations of ASE were made at<br />
the output from singlemode fibers in the backward<br />
direction to the pumping generated by an industrial<br />
sample of the pumping source [11] containing four<br />
semi-conductor laser diodes (LD). The LD’s wavelengths<br />
are 1426, 1436, 1456 and 1466 nm. Every LD<br />
had the maximal output power of 300 mW. The experimental<br />
set up is schematically shown on Fig. 3.<br />
Output pump power from each LD through the pump<br />
combiner and circulator is directed to the 50 km span<br />
of standard single mode fiber in this set up. Pumping<br />
source allows independently fix the output power of<br />
each LD in the range from 0 mW to 300 mW using<br />
digital control unit. Output ASE power from the fiber<br />
after circulator is registered by optical spectrum<br />
analyzer (OSA). Spectral resolution of OSA was set<br />
to 1 nm (~ 4 cm -1 ) for all ASE measurements. Only<br />
one LD remained active (the others LDs are switched<br />
off) at each registration of spectra with the fixed set of<br />
power levels: 100, 150, 200, 250 and 300 mW for every<br />
wavelength of the pumping LD.<br />
ASE spectral density, nW/nm<br />
ASE spectral density, nW/nm<br />
20<br />
15<br />
10<br />
λp=1426 nm<br />
5<br />
4<br />
3<br />
2<br />
1<br />
0<br />
20<br />
1450 1500 1550 1600<br />
15<br />
10<br />
5<br />
λp=1436 nm<br />
5<br />
0<br />
1450<br />
1<br />
1500 1550 1600<br />
Wavelength, nm<br />
2<br />
5<br />
4<br />
3<br />
20<br />
15<br />
10<br />
5<br />
20<br />
15<br />
10<br />
SMF 50 km<br />
4λ pump<br />
Circulator<br />
Pump combiner<br />
ASE<br />
1426 nm 1436 nm 1456 nm 1466 nm<br />
Pumping Source<br />
OSA<br />
Fig. 3. Experimental set up for ASE measurements with four<br />
LD in commercial FRA pumping source.<br />
EXPERIMENTAL RESULTS AND<br />
DISCUSSION<br />
The typical example of registered ASE spectra is<br />
shown on Fig. 4. Presented ASE measurements show, that<br />
the absolute power cross section of Stokes radiation for<br />
each of four pumping wavelengths changes from (2,75 ±<br />
0,08)⋅10 -6 at LD input pumping power of 100 mW up to<br />
(4,3 ± 0.2)⋅10 -6 for pumping power of 300 mW. The received<br />
numerical values for ASE cross section as of the<br />
order size ~10 -6 and thus, whole Stokes power is approximately<br />
-60 äÁ related to input pumping power. It<br />
correspond to quantum efficiency more likely to spontaneous<br />
Raman scattering, but it does not SRS because<br />
it quantum output should be higher on about 4–5 orders.<br />
The absolute ASE cross section was experimentally<br />
determined as relation of total power of Stokes<br />
spectrum, integrated on shifted frequencies diapason<br />
from 10 cm -1 up to 1400 cm -1 , to input pump power.<br />
5<br />
λp=1456 nm<br />
5<br />
4<br />
3<br />
2<br />
1<br />
1500 1550 1600<br />
λp=1466 nm<br />
2<br />
1<br />
3<br />
5<br />
4<br />
1500 1550 1600<br />
Wavelength, nm<br />
Fig. 4. Stokes ASE power generated with separate pumping LD.<br />
Pp, mW<br />
1- 100<br />
2- 150<br />
3- 200<br />
4- 250<br />
5- 300<br />
1650<br />
1650
The general view of the spectra submitted on fig.<br />
5a in qualitative interpretation unequivocally specifies<br />
features of a known spectrum of the spontaneous Raman<br />
light scattering in silica fibers. There is obvious<br />
similarity to a non-uniform continuum in the range of<br />
Stokes shifted frequencies from 0 cm -1 up to 900 cm -1 .<br />
The main distinction of the spontaneous Raman spectrum<br />
from spectral profile of Raman gain (see fig. 2)<br />
is shown in the raised intensity small Stokes shifted<br />
frequency components is approximately down to 200<br />
cm -1 . Distinction between the spontaneous and stimulated<br />
Raman spectra directly follows from the quantum<br />
dynamic equation (10). The terms corresponding to<br />
spontaneous Raman scattering in this equation should<br />
contains a phonon density factor of a kind η B (ω) +1<br />
[14]. The Bose factor at T=300K considerably exceeds<br />
unit in the frequency interval from 0 up to 200 cm -1 ,<br />
and it rises to infinity when frequency aspires to zero.<br />
AbsoluteASE power density, (nW/nm)<br />
(a)<br />
20<br />
15<br />
10<br />
5<br />
SMF 50 km<br />
λp=1466 nm<br />
Av. resol. ~ 4 cm -1<br />
70 cm -1<br />
5<br />
4<br />
6<br />
Frequency, (THz)<br />
12 18<br />
440 cm -1<br />
601 cm -1<br />
24<br />
Pp (mW)<br />
1 – 100<br />
2 – 150<br />
3 – 200<br />
4 – 250<br />
5 – 300<br />
Raman shift, (cm -1 3<br />
2<br />
1<br />
0<br />
0 200 400 600 800<br />
)<br />
Normalized ASE power<br />
The Raman gain profile should be formed by SRS<br />
process as zero Kelvin cross section scattering which<br />
is shown by a dotted line in the Fig. 5b. In contrast to<br />
spontaneous Raman scattering SRS does not depend<br />
on phonon density states and, accordingly, does not<br />
depend on temperature. It is the reason of distinction<br />
between the observable Raman gain spectrum and the<br />
measured spontaneous Raman spectrum and it explains<br />
the obvious tendency of ASE distribution to the<br />
structure of the Raman gain profile at pumping power<br />
is increased as one can see in Fig. 5b [15].<br />
Other feature of a spontaneous Raman scattering<br />
is by the nature the linear process and thereby it does<br />
not depend on pumping intensity. As result the spontaneous<br />
Raman cross section remains the constant for<br />
any studied material and the dashed lines in Fig. 6 correspond<br />
to the power of spontaneous Stokes radiation<br />
as function of pumping power.<br />
1.0<br />
0.8<br />
0.6<br />
0.4<br />
0.2<br />
0<br />
(b)<br />
1<br />
2<br />
3<br />
Zero Kelvin<br />
(Raman gain profile)<br />
0 200 400 600 800<br />
Raman shift, (cm -1 )<br />
Pp (mW)<br />
1 – 100<br />
2 – 200<br />
3 – 300<br />
Fig. 5. Absolute (a) and normalized (b) Stokes ASE power distributions generated by separate LD λ p =1466 nm in backward pumped<br />
FRA with terahetz bandwidth. Normalized curves (b) show the trend of ASE distribution to Raman gain profile (dotted line) when the<br />
pump power is increased [15].<br />
Power amplification of spontaneous optical noise<br />
in singlemode silica fiber as one can see in Fig. 6, is in<br />
enough small limits. Quantum efficiency of the Raman<br />
grows no more than by ~40 %, and it corresponds<br />
on\off amplification approximately on 1.9 dB when<br />
the pumping power increases in 3 times, that is from<br />
100 mW up to 300 mW.<br />
It is possible to explain such situation from the<br />
physical point of view as follows. Rather weak ASE<br />
generation in the studied pump power range is resulted<br />
from no coherence of the Stokes photons arising at not<br />
elastic scattering of pump photons on huge amount of<br />
molecular phonon vibrations with different frequencies.<br />
Therefore Stokes radiation in addition to its random<br />
phase distribution appears as distributed in a very<br />
wide frequency diapason. Both these circumstances<br />
obstruct to automatic establish the phase matching<br />
conditions necessary for coherent accumulation of<br />
Stokes radiation which would give Raman gain. In<br />
other words Raman interaction in the core of the silica<br />
based optical fiber results to “spreading” of pump<br />
power along the wide spectrum of Stokes frequencies.<br />
As the creation probability of the in phase Stokes photons<br />
with equal frequencies inversely depends on Raman<br />
radiation bandwidth as it appears available pump<br />
power insufficiently for effective Raman noise generation.<br />
In result spectral power distribution of ASE observable<br />
by us looks more likely spontaneous Raman<br />
scattering, instead of SRS.<br />
It should be noted the presented results are received<br />
with no optical signals in the studied single<br />
mode fiber piece. At signal presence its coherent power<br />
with the spectral density considerably higher than<br />
the Stokes noise density starts the concurrence for<br />
possession of the pump power during the SRS process.<br />
The expending of the pump power for 20-30 dB signal<br />
13
amplification causes the accelerated pump depletion<br />
in the propagation process along the fiber, simultaneously<br />
reducing effective length of power accumulation<br />
of Stokes noise. Therefore the ASE gain coefficients<br />
measured by us in the single mode silica fiber have the<br />
greatest possible quantities and noise power of real<br />
FRA cannot exceed the absolute values resulted on<br />
Fig. 4.<br />
Stokes ASE Power, (nW)<br />
14<br />
15<br />
10<br />
5<br />
SMF 50 km<br />
λp=1466 nm<br />
- Gain<br />
- Spontaneous scatt.<br />
- Spont. Levels<br />
0<br />
0 100 200 300<br />
Pump Power, (mW)<br />
440 cm -1<br />
70 cm -1<br />
601 cm -1<br />
Fig. 6. Stokes ASE power as function of pump power for<br />
several peaks in experimental spectra on Fig.5a (solid dots). Right<br />
hand diagram show that the Raman gain values are only decimal<br />
parts in comparison with spontaneous scattering peaks [15].<br />
The data resulted in a Fig. 6 is illustrated than the<br />
pure ASE amplification essentially depends on peak<br />
power in noise distribution which is defined by frequency<br />
position of the given maximum in the spectrum.<br />
The ASE gain coefficients appreciably grow in<br />
the spectrum points having the big intensity as a result<br />
of Raman nonlinearity. Simultaneously maximal<br />
noise amplification remains many times smaller in<br />
comparison with nonlinear amplification of the coherent<br />
signal. The output SNR after Raman amplifier<br />
becomes big in comparison with the case of FRA<br />
absence, therefore there is an improvement (even to<br />
F
UDC 535.361:621.391.822<br />
G. S. Felinskyi<br />
NOISE MEASUREMENTS OF THE BACKWARD PUMPED DISTRIBUTED FIBER RAMAN AMPLIFIER<br />
Abstract<br />
Noise parameters of fiber Raman amplifiers are appeared in practice essentially better top theoretical limits and the amplifier<br />
noise performance is the subject of intensive investigations now. the experimental results on the amplified spontaneous emission (ASE)<br />
observation in the single mode fiber span using the backward pumped distributed FRA are presented in the report. Our measurements<br />
and quantity analysis show that physical principles of formation of the FRA optical noise with the counter pumping lead to very small<br />
gain coefficients and it almost correspond to spontaneous Raman properties in the silica fibers. Raman gain nonlinearity is the reason<br />
to reduce the real FRA noise figure much below the established 3 dB limit. Our analysis of experimental data on ASE measurements in<br />
silica fibers allows us to get the information about the nature of formation mechanism of the noise parameters in the practical FRA.<br />
Key words: fiber Raman amplifiers, optical fiber communications, amplified spontaneous emission.<br />
ÓÄÊ 535.361:621.391.822<br />
Ã. Ñ. Ôåëèíñêèé<br />
ÈÇÌÅÐÅÍÈß ØÓÌÀ  ÐÀÑÏÐÅÄÅËÅÍÍÎÌ ÂÊÐ ÓÑÈËÈÒÅËÅ ÑÎ ÂÑÒÐÅ×ÍÎÉ ÍÀÊÀ×ÊÎÉ<br />
Ðåçþìå<br />
Øóìîâûå ïàðàìåòðû ðåàëüíûõ âîëîêîííûõ ÊÐ óñèëèòåëåé íà ïðàêòèêå îêàçûâàþòñÿ çíà÷èòåëüíî ëó÷øå óñòàíîâëåííûõ<br />
òåîðåòè÷åñêèõ ïðåäåëîâ, è øóìû óñèëèòåëÿ ñåé÷àñ ÿâëÿþòñÿ ïðåäìåòîì èíòåíñèâíûõ èññëåäîâàíèé.  ðàáîòå ïðåäñòàâëåíû<br />
ýêñïåðèìåíòàëüíûå ðåçóëüòàòû èññëåäîâàíèÿ óñèëåííîãî ñïîíòàííîãî èçëó÷åíèÿ (ÓÑÈ), êîòîðîå íàáëþäàëîñü â îäíîìîäîâîì<br />
âîëîêíå ïðè èñïîëüçîâàíèè ðàñïðåäåëåííîãî âîëîêîííîãî ÊÐ óñèëèòåëÿ ñî âñòðå÷íîé íàêà÷êîé. Íàøè èçìåðåíèÿ è èõ<br />
êîëè÷åñòâåííûé àíàëèç ïîêàçûâàþò, ÷òî ôèçè÷åñêèå ïðèíöèïû ôîðìèðîâàíèÿ îïòè÷åñêîãî øóìà â óñèëèòåëÿõ ñî âñòðå÷íîé<br />
íàêà÷êîé ïðèâîäÿò ê î÷åíü ìàëåíüêèì êîýôôèöèåíòàì åãî óñèëåíèÿ, à ñàì øóì ïî÷òè ñîîòâåòñòâóåò ñâîéñòâàì ñïîíòàííîãî<br />
ÊÐ â êâàðöåâûõ âîëîêíàõ. Íåëèíåéíîñòü óñèëåíèÿ ÊÐ ÿâëÿåòñÿ ãëàâíîé ïðè÷èíîé ñíèæåíèÿ ðåàëüíîãî êîýôôèöèåíòà<br />
øóìà çíà÷èòåëüíî íèæå óñòàíîâëåííîãî ïðåäåëà â 3 äÁ. Ïðåäñòàâëåííûé àíàëèç ýêñïåðèìåíòàëüíûõ äàííûõ èçìåðåíèé<br />
ÓÑÈ â êâàðöåâûõ âîëîêíàõ ïîçâîëÿåò ïîëó÷èòü èíôîðìàöèþ î ïðèðîäå ìåõàíèçìîâ ôîðìèðîâàíèÿ øóìîâûõ ïàðàìåòðîâ â<br />
ðåàëüíûõ óñèëèòåëÿõ.<br />
Êëþ÷åâûå ñëîâà: âîëîêîííûå ÊÐ-óñèëèòåëè, óñèëèòåëü ñî âñòðå÷íîé íàêà÷êîé, îïòè÷åñêèå âîëîêíà.<br />
ÓÄÊ 535.361:621.391.822<br />
Ã. Ñ. Ôåë³íñüêèé<br />
ÂÈ̲ÐÞÂÀÍÍß ØÓÌÓ Â ÐÎÇÏÎIJËÅÍÎÌÓ ÂÊРϲÄÑÈËÞÂÀײ ²Ç ÇÓÑÒв×ÍÈÌ ÍÀÊÀ×ÓÂÀÍÍßÌ<br />
Ðåçþìå<br />
Øóìîâ³ ïàðàìåòðè ðåàëüíèõ âîëîêîííèõ ÊÐ ï³äñèëþâà÷³â íà ïðàêòèö³ âèÿâëÿþòüñÿ çíà÷íî êðàùèìè çà âñòàíîâëåí³<br />
òåîðåòè÷í³ ìåæ³, à øóìè ï³äñèëþâà÷à çàðàç º ïðåäìåòîì ³íòåíñèâíèõ äîñë³äæåíü. Ó ðîáîò³ íàäàíî åêñïåðèìåíòàëüí³<br />
ðåçóëüòàòè äîñë³äæåííÿ ï³äñèëåíîãî ñïîíòàííîãî âèïðîì³íþâàííÿ (ÏÑÂ), ÿêå ñïîñòåð³ãàëîñÿ â îäíîìîäîâîìó âîëîêí³ ïðè<br />
çàñòîñóâàíí³ ðîçïîä³ëåíîãî âîëîêîííîãî ÊÐ ï³äñèëþâà÷à ³ç çóñòð³÷íèì íàêà÷óâàííÿì. Íàø³ âèì³ðþâàííÿ òà ¿õ ê³ëüê³ñíèé<br />
àíàë³ç ïîêàçóþòü, ùî ô³çè÷í³ ïðèíöèïè ôîðìóâàííÿ îïòè÷íîãî øóìó â ï³äñèëþâà÷àõ ³ç çóñòð³÷íèì íàêà÷óâàííÿì ïðèâîäÿòü<br />
äî äóæå ìàëèõ êîåô³ö³ºíò³â éîãî ï³äñèëåííÿ, à ñàì øóì ìàéæå â³äïîâ³äຠâëàñòèâîñòÿì ñïîíòàííîãî ÊÐ ó êâàðöîâèõ âîëîêíàõ.<br />
Íåë³í³éí³ñòü ï³äñèëåííÿ ÊÐ º ãîëîâíîþ ïðè÷èíîþ çíèæåííÿ ðåàëüíîãî êîåô³ö³ºíòà øóìó çíà÷íî íèæ÷å âñòàíîâëåíî¿ ìåæ³ â<br />
3 äÁ. Íàäàíî àíàë³ç åêñïåðèìåíòàëüíèõ äàíèõ âèì³ðþâàíü ÏÑ ó êâàðöîâèõ âîëîêíàõ, ÿêèé äîçâîëÿº îòðèìàòè ³íôîðìàö³þ<br />
ïðî ïðèðîäó ìåõàí³çì³â ôîðìóâàííÿ øóìîâèõ ïàðàìåòð³â ó ðåàëüíèõ ï³äñèëþâà÷àõ.<br />
Êëþ÷îâ³ ñëîâà: âîëîêîíí³ ÊÐ-ï³äñèëþâà÷³, ï³äñèëþâà÷ ç³ çóñòð³÷íèì íàêà÷óâàííÿì, îïòè÷í³ âîëîêíà.<br />
15
16<br />
UDÑ 539.192<br />
T. A. FLORKO, O. YU. KHETSELIUS, YU. V. DUBROVSKAYA, D. E. SUKHAREV<br />
Odessa National Polytechnical University, Odessa<br />
Odessa State Environmental University, Odessa<br />
I. I. Mechnikov Odessa National University, Odessa<br />
BREMSSTRAHLUNG AND X-RAY SPECTRA FOR KAONIC AND PIONIC<br />
HYDROGEN AND NITROGEN<br />
The level energies, energy shifts and transition rates are estimated for pionic and kaonic atoms of<br />
hydrogen and nitrogen on the basis of the relativistic perturbation theory with an account of nuclear<br />
and radiative effects. New data about spectra of the exotic atomic systems can be considered as a new<br />
tool for sensing the nuclear structure and creation of new X-ray sources too.<br />
1. INTRODUCTION<br />
At present time, the light hadronic (pionic, kaonic<br />
etc.) atomic systems are intensively studied and can<br />
be considered as a candidate to create the new lowenergy<br />
X-ray standards [1-12]. In the last few years<br />
transition energies in pionic [1] and kaonic atoms [2]<br />
have been measured with an unprecedented precision.<br />
Besides, an important aim is to evaluate the pion mass<br />
using high accuracy X-ray spectroscopy [1-10]. Similar<br />
endeavour are in progress with kaonic atoms. It is<br />
easily to understand that the spectroscopy of pionic<br />
and kaonic hydrogen gives unprecedented possibilities<br />
to study the strong (nuclear) interaction at low energies<br />
[5-8] by measuring the energy and natural width<br />
of the ground level with a precision of few meV. Naturally,<br />
studying the hadronic atomic systems is of a great<br />
interest for further development of atomic and nuclear<br />
theories as well as new tools for sensing the nuclear<br />
structure and fundamental interactions, including the<br />
Standard model [1-15]. The collaborators of the E570<br />
experiment [7,8] measured X-ray energy of a kaonic<br />
hydrogen atom, which is an atom consisting of a kaon<br />
(a negatively charged heavy particle) and a hydrogen<br />
nucleus (proton). The kaonic hydrogen X-rays were<br />
detected by large-area Silicon Drift Detectors, which<br />
readout system was developed by SMI (see [1,8]). It is<br />
known that the shifts and widths due to the strong interaction<br />
can be systematically understood using phenomenological<br />
optical potential models. Nevertheless,<br />
one could mention a large discrepancy between<br />
the theories and experiments on the kaonic atoms<br />
states. (for example, well known puzzle with helium<br />
2p state). A large repulsive shift (about -40 eV) has<br />
been measured by three experimental groups in the<br />
1970’s and 80’s, while a very small shift (< 1 eV) was<br />
obtained by the optical models calculated from the<br />
kaonic atom X-ray data with Z>2 (look [1]). This significant<br />
disagreement (a difference of over 5 standard<br />
deviations) between the experimental results and the<br />
theoretical calculations is known as the “kaonic helium<br />
puzzle”. A possible large shift has been predicted<br />
using the model assuming the existence of the deeply<br />
bound kaonic nuclear states. However, even using this<br />
model, the large shift of 40 eV measured in the experiments<br />
cannot be explained. A re-measurement of the<br />
shift of the kaonic helium X-rays is one of the top pri-<br />
orities in the experimental research activities. In the<br />
last papers (look, for example, [5,6] and [10] too) this<br />
problem is physically reasonably solved. In the theory<br />
of the kaonic and pionic atoms there is an important<br />
task, connected with a direct calculation of the radiative<br />
transition energies within consistent relativistic<br />
quantum mechanical and QED methods (c.f.[13-<br />
15]). The multi-configuration Dirac-Fock (MCDF)<br />
approximation is the most reliable approach for multielectron<br />
systems with a large nuclear charge; in this<br />
approach one- and two-particle relativistic effects are<br />
taken into account practically precisely. The next important<br />
step is an adequate inclusion of the radiative<br />
corrections. This topic has been a subject of intensive<br />
theoretical and experimental interest (see [13]). Nevertheless,<br />
the problem remains quite far from its final<br />
solution. It is of a great interest to study and treat these<br />
effects in the pionic and kaonic systems (for example,<br />
hydrogen, nitrogen, oxygen etc.). In this paper the hyperfine<br />
structure (HFS) level energies, energy shifts,<br />
transition rates are estimated for the pionic and kaonic<br />
atoms of hydrogen and nitrogen. New data about spectra<br />
of the hadronic systems can be considered as a new<br />
tool for sensing nuclear structure and creation of new<br />
X-ray sources. Our method is based on the relativistic<br />
perturbation theory (PT) [10,15,16,] with an accurate<br />
account of the nuclear and radiative effects. The Lamb<br />
shift polarization part is described in the Uehling-Serber<br />
approximation; the Lamb shift self-energy part is<br />
considered effectively within the advanced scheme.<br />
2. METHOD OF RELATIVISTIC<br />
PERTURBATION THEORY<br />
Let us describe the key moments of our scheme<br />
to relativistic calculation of the spectra for exotic<br />
atomic systems with an account of relativistic, correlation,<br />
nuclear, radiative effects (more details can<br />
be found for example, in ref. [15]; see also [10]). In<br />
general, the one-particle wave functions are found<br />
from solution of the Klein-Gordon equation with<br />
potential, which includes the self-consistent V 0 potential<br />
(including electric, polarization potentials of<br />
nucleus):<br />
{1/c2 [E +eV (r)] 0 0 2 + 2<br />
h∇ -m2c2 }Ψ( r)=0<br />
© T. A. Florko, O. Yu. Khetselius, Yu. V. Dubrovskaya, D. E. Sukharev, 2009
where E 0 is the total energy of the system (sum of the<br />
mass energy mc 2 and the binding energy ε 0 ). To describe<br />
the nuclear finite size effect the smooth Gaussian<br />
function of the charge distribution in a nucleus is<br />
used. With regard to normalization we have:<br />
32 2<br />
( rR) ( 4 ) exp(<br />
r)<br />
ρ = γ π −γ (1)<br />
2<br />
where γ= 4 π R , R is the effective nucleus radius.<br />
The Coulomb potential for the spherically symmetric<br />
density ρ( r ) is:<br />
nucl<br />
( ) (<br />
r<br />
∞<br />
' '2 ' ' ' '<br />
( 1 ) ( ) ( )<br />
∫ ∫ (2)<br />
V r R =− r drr ρ r R + drrρr R<br />
0<br />
It is determined by the following system of differential<br />
equations:<br />
r<br />
' 2 ' '2 ' 2<br />
( , ) = ( 1 ) ρ( , ) ≡(<br />
1 ) ( , )<br />
V nucl r R r ∫ dr r r R r y r R (3)<br />
0<br />
2 ( ) ( )<br />
y' r, R = r ρ r, R<br />
(4)<br />
52 2<br />
( rR) r ( r)<br />
ρ ' , = −8γ πexp −γ =<br />
8r<br />
=−2 γrρ ( r, R) =− ρ 2 ( r, R)<br />
(5)<br />
πr<br />
with the corresponding boundary conditions. The<br />
pion (kaon ) charge distribution is also included in the<br />
strict accordance to the scheme [2,7]. Further, one<br />
can write the Klein-Gordon type equations for oneor<br />
multi-particle system. In general, formally they fall<br />
into one-particle equations with potential, which includes<br />
the self-consistent potential, electric, polarization<br />
potentials of a nucleus. Procedure for an account<br />
of the radiative corrections is given in detail in refs.<br />
[15]. Regarding the vacuum polarization (Vac.Pol.)<br />
effect let us note that this effect is usually taken into<br />
account in the first PT order by means of the Uehling<br />
potential:<br />
∞<br />
2α<br />
U () r =− dt exp( −2rt α Z ) ×<br />
3πr<br />
∫<br />
× +<br />
1<br />
2<br />
t<br />
2<br />
1 2<br />
≡ −<br />
3π<br />
2 − α<br />
( 1 1 2 t ) C() g , (6)<br />
t r<br />
where g=r/αZ. In our calculation we usually use more<br />
exact approach. The Uehling potential, determined as<br />
a quadrature (6), is approximated with high precision<br />
by a simple analytical function. The use of new approximation<br />
of the Uehling potential [15] permits one<br />
to decrease the calculation errors for this term down to<br />
~0.5%. Besides, using such a simple function form for<br />
the Uehling potential allows its easy inclusion to the<br />
general system of differential equations. It is very important<br />
to underline that the scheme used includes automatically<br />
high-order vacuum polarization contributions,<br />
including, the well known Wichman-Kroll and<br />
Källen-Sabry ones. A scheme for estimating the selfenergy<br />
part of the Lamb shift is based on the method<br />
[22-24]. In an atomic system the radiative shift and<br />
relativistic part of the energy are, in principle, defined<br />
by one and the same physical field. It may be supposed<br />
that there exists some universal function that connects<br />
r<br />
a self-energy correction and relativistic energy. The<br />
self-energy correction for states of a hydrogen-like ion<br />
is presented by Mohr [18,19]. In ref. [22-24] this result<br />
is modified for the corresponding states of the multiparticle<br />
atomic system. Further let us note that so<br />
called relativistic recoil contribution is not calculated<br />
by us and its value is taken from refs [4,7]. The transition<br />
probabilities between the HFS sublevels are defined<br />
by the standard energy approach formula. Other<br />
details of the method used (and Superatom code) can<br />
be found in refs. [15,16].<br />
3. DATA FOR HADRONIC ATOMS AND<br />
DISCUSION<br />
We studied the X-ray spectra for the hadronic hydrogen<br />
and nitrogen. In figure 1 the experimental kaonic<br />
hydrogen X-ray energy spectra is presented [7,8].<br />
In table 1 we present the measured and theoretical Xray<br />
energies of kaonic hydrogen atom for the 2-1 transition<br />
(in keV). In figure 1 this transition is clearly identified.<br />
The notations are related to the initial (n i ) and<br />
final (n f ) quantum numbers. The calculated value of<br />
transition energy is compared with available measured<br />
(E m ) and other calculated (E c ) values [1,3,9,10].<br />
Fig. 1. The experimental kaonic hydrogen X-ray energy spectrum<br />
[7,8].<br />
Transition<br />
Calculated (E c ) and measured (E m ) kaonic H atom<br />
X-ray energies (in keV)<br />
E c ,<br />
this work<br />
E c<br />
[10]<br />
E c<br />
[3]<br />
2-1 6.420 6.65 6.48<br />
E c<br />
[9]<br />
6.480<br />
6,482<br />
E m<br />
Table 1<br />
E m<br />
[1,7] [1,8]<br />
6,44(8) 6.675(60)<br />
6,96 (9)<br />
In tables 2-4 we present the data on energy (in eV)<br />
contribution for selected levels (transitions 8k-7i and<br />
8i-7h), hyperfine transition energies and transition<br />
rates in kaonic nitrogen. The radiative effects contributions<br />
are indicated separately. For comparison, the<br />
estimating data from [3,9] are given too.<br />
In tables 5-7 we present theoretical data on energy<br />
(in eV) contribution for the selected levels (transitions<br />
5g-4f and 5f-4d), hyperfine transition energies and<br />
transition rates in the pionic nitrogen. The radiative<br />
corrections are separately indicated. For comparison,<br />
the estimating data from refs. [4,9] are given too. The<br />
detailed analysis of theoretical and separated experi-<br />
17
mental data shows that indeed there is a physically<br />
reasonable agreement between the cited data. But, obviously,<br />
there may take a place the exception too as it is<br />
shown on example of the kaonic uranium in ref. [10].<br />
Table 2<br />
Energy (in eV) contribution for the selected levels in kaonic nitrogen.<br />
The first error takes into account neglected next order radiative<br />
corrections. The second is due to the accuracy of the kaon mass<br />
(±32 ppm)<br />
Contributions 8k-7i [9]<br />
8k-7i<br />
This work<br />
8i-7h<br />
This work<br />
Coulomb 2968.4565 2968.4492 2968.5344<br />
Vac. Pol. 1.1789 1.1778 1.8758<br />
Relativistic Recoil 0.0025 0.0025 0.0025<br />
HFS Shift -0.0006 -0.0007 -0.0009<br />
Total 2969.6373 2969.6288 2970.4118<br />
Error 0.0005 0.0004 0.0004<br />
Error due to the kaon mass 0.096 0.096 0.096<br />
18<br />
Transition<br />
Table 3<br />
Hyperfine transition (8k-7i) energies in kaonic nitrogen<br />
F-<br />
F’<br />
Trans. E<br />
(eV)<br />
[3]<br />
Trans. E<br />
(eV)<br />
This work<br />
Trans. rate<br />
(s-1 )<br />
[3]<br />
Trans.<br />
rate (s -1 )<br />
This work<br />
8k-7i 8-7 2969.6365 2969.6289 1.54 × 10 13 1.51 × 10 13<br />
7-6 2969.6383 2969.6298 1.33 × 10 13 1.32 × 10 13<br />
7-7 2969.6347 2969.6264 1.31 × 10 13 1.29 × 10 13<br />
6-5 2969.6398 2969.6345 1.15 × 10 13 1.12 × 10 13<br />
6-6 2969.6367 2969.6284 0.03 × 10 13 0.02 × 10 13<br />
6-7 2969.6332 2969.6248 0.00 × 10 13 0.00 × 10 13<br />
Table 4<br />
Hyperfine transition (8i-7h) energies in kaonic nitrogen (this work)<br />
Transition F-F’ Trans. E (eV) Trans. rate (s -1 )<br />
8i — 7h 7-6 2970.4107 1.16 × 10 13<br />
6-5 2970.4135 0.99 × 10 13<br />
6-6 2970.4086 0.96 × 10 13<br />
5-4 2970.4193 0.81 × 10 13<br />
5-5 2970.4114 0.02 × 10 13<br />
5-6 2970.4073 0.00 × 10 13<br />
Table 5<br />
Energy (in eV) contribution for the selected levels in pionic nitrogen.<br />
The first error takes into account neglected next order radiative<br />
corrections. The second is due to the accuracy of the pion mass<br />
(±2.5 ppm)<br />
Contributions 5g-4f [4]<br />
5g-4f This<br />
work<br />
5f-4d This<br />
work<br />
Coulomb 4054.1180 4054.1146 4054.7152<br />
Self Energy -0.0001 -0.0002 -0.0004<br />
Vac. Pol. 1.2602 1.2599 2.9711<br />
Relativistic Recoil 0.0028 0.0028 0.0028<br />
HFS Shift -0.0008 -0.0009 -0.0030<br />
Total 4055.3801 4055.3762 4057.6857<br />
Error ±0.0011 ±0.0007 ±0.0009<br />
Error due to the pion mass ±0.010 ±0.010 ±0.010<br />
We mean the agreement between theoretical estimating<br />
data and experimental results. One should<br />
keep in mind the following important moment. In a<br />
case of good agreement between theoretical and experimental<br />
data, the corresponding levels are less sensitive<br />
to strong nuclear interaction. In the opposite<br />
case one could point to a strong-interaction effect<br />
in the exception cited (as example, some transitions<br />
in the hadronic U) [10]. In a whole, to understand<br />
further information on the low-energy kaon-nuclear<br />
(pion-nuclear) interaction, new experiments to define<br />
the shift and width of kaonic H/deuterium are now in<br />
preparation in J-Parc and in LNF (c.f. [2,6,8]). Finally,<br />
let us turn attention at the new possibilities, which<br />
are opened with X-ray, γ -lasers (raser, graser) action<br />
on hadronic system. Namely, speech is about a set of<br />
the possible new nuclear quantum-optical effects in<br />
the kaonic and pionic systems [25-27].<br />
Table 6<br />
Hyperfine transition (5g -4f) energies and transition rates in pionic<br />
nitrogen<br />
Transition<br />
F-F’<br />
Trans. E<br />
(eV)<br />
[3]<br />
Trans. E<br />
(eV)<br />
This work<br />
Trans. rate<br />
(s-1 )<br />
[3]<br />
Trans. rate<br />
(s -1 )<br />
This work<br />
5g -4f 5-4 4055.3779 4055.3744 7.13 × 10 13 7.10 × 10 13<br />
4-3 4055.3821 4055.3784 5.47 × 10 13 5.42 × 10 13<br />
4-4 4055.3762 4055.3735 5.27 × 10 13 5.23 × 10 13<br />
3-2 4055.3852 4055.3828 4.17 × 10 13 4.12 × 10 13<br />
3-3 4055.3807 4055.3769 0.36 × 10 13 0.33 × 10 13<br />
3-4 4055.3747 4055.3712 0.01 × 10 13 0.01 × 10 13<br />
Table 7<br />
Hyperfine transition (5f — 4d) energies and transition rates in<br />
pionic N (this work)<br />
Transition F-F’ Trans. E (eV) Trans. rate (s -1 )<br />
5f — 4d 4-3 4057.6821 4.52 × 10 13<br />
3-2 4057.6914 3.11 × 10 13<br />
3-3 4057.6793 2.93 × 10 13<br />
2-1 4057.6954 2.09 × 10 13<br />
2-2 4057.6892 2.21 × 10 13<br />
2-3 4057.6768 0.01 × 10 13<br />
References<br />
1. Deloff A., Fundamentals in Hadronic Atom Theory, Singapore:<br />
World Sci., 2003. — 352P.<br />
2. Hayano R.S., Hori M., Horvath D., Widman E., Antiprotonic<br />
helium and CPT invariance//Rep. Prog. Phys. — 2007. —<br />
Vol.70. — P.1995-2065.<br />
3. Trassinelli M., Indelicato P., Relativistic calculations of pionic<br />
and kaonic atoms hyperfine structure// arXiv:physics.<br />
— 2007. — 0611263v2.<br />
4. Anagnostopoulos D., Gotta D., Indelicato P., Simons L.M.,<br />
Low-energy X-ray standards from hydrogenlike pionic atoms//<br />
arXiv:physics. — 2003. — 0312090v1.<br />
5. Okada S., Beer G., Bhang H., et al, Precision measurement<br />
of the 3d→2p x-ray energy in kaonic 4 He//Phys.Lett.B. —<br />
2007. — Vol.653, N 5-6. — P. 387-391.<br />
6. Okada S., Beer G., Bhang H., et al, Precision spectroscopy of<br />
Kaonic Helium 3d → 2p X-rays //Nucl.Phys.A. — 2007. —<br />
Vol.790,N1-4. — P.663-666.<br />
7. Ito T.M., Hayano R.S., Nakamura S.N., Terada T.P., Observation<br />
of kaonic hydrogen atom x rays// Phys. Rev. C. —<br />
1998. — Vol.58. — P.2366 — 2382<br />
8. Ishiwatari T. on behalf of the SIDDHARTA Collaboration,<br />
Silicon drift detectors for the kaonic atom X-ray measurements<br />
in the SIDDHARTA experiment// Nucl. Instr. and<br />
Methods in Phys. A: Accelerators, Spectrometers, Detectors.<br />
— 2007. — Vol.581. — P.326-329.<br />
9. Santos J.P., Parente F., Boucard S., Indelicato P., Desclaux<br />
j.P., X-ray energies of circular transitions and electron scattering<br />
in kaonic atoms//Phys.Rev.A. — 2005. — Vol.71. —<br />
P.032501.<br />
10. Khetselius O.Yu., Turin A.V., Sukharev D.E., Florko T.A.,<br />
Estimating of X-ray spectra for kaonic atoms as tool for sensing<br />
the nuclear structure// Sensor Electr. and Microsyst.<br />
Techn. — 2009. — N1. — P.13-18.<br />
11. Glushkov A.V., Makarov I.T., Nikiforova E., Pravdin M.,<br />
Sleptsov I., Muon component of EAS with energies above<br />
10 17 eV// Astroparticle Physics. — 1999. — Vol.4. — P. 15-<br />
22.
12. Glushkov A.V.,Dedenko L.G.,Pravdin M.I.,Sleptsov I.E.,<br />
Spatio-temporal structure of the muon disk at E 0 ≥ 5×10 16 eV<br />
from EAS array data//JETP. — 2004. — Vol.99,N1. — P.123-<br />
132.<br />
13. Grant I.P., Relativistic Quantum Theory of Atoms and Molecules,<br />
Springer, 2007. — 650P.<br />
14. Dyall K.G., Faegri K. Jr., Introduction to relativistic quantum<br />
theory. — Oxford, 2007. — 680P.<br />
15. Glushkov A.V., Relativistic quantum theory. Quantum mechanics<br />
of atomic systems. — Odessa: Astroprint, 2008. —<br />
800P.<br />
16. Glushkov A.V., Lovett L., Florko T.A., et al, Gauge-invariant<br />
QED perturbation theory approach to studying the nuclear<br />
electric quadrupole moments, hyperfine structure constants<br />
for heavy atoms and ions// Frontiers in Quantum Systems<br />
in Chemistry and Physics (Springer). — 2008. — Vol.18. —<br />
P.505-522.<br />
17. Karshenboim S.G., Kolachevsky N.N., Ivanov V.G., Fischer<br />
M., Fendel P., Honsch T.W., 2s-Hyperfine Splitting in Light<br />
Hydrogen-like Atoms: Theory and Experiment// JETP. —<br />
2006. — Vol.102,N3. — P.367-376.<br />
18. Mohr P.J. Quantum Electrodynamics Calculations in few-<br />
Electron Systems// Phys. Scripta. — 1993. — Vol.46,N1. —<br />
P.44-52.<br />
19. Mohr P.J. Energy Levels of H-like atoms predicted by Quan-<br />
UDÑ 539.192<br />
T. A. Florko, O. Yu. Khetselius, Yu. V. Dubrovskaya, D. E. Sukharev<br />
tum Electrodyna-mics,10
INTRODUCTION<br />
The short circuit current density J SC and open<br />
circuit voltage U OC of photo-voltaic converters (PVC)<br />
are increased with intensity growth of light flux penetrating<br />
into their semiconductor base. It causes the<br />
expediency of concentrated solar radiation (CSR)<br />
used for increasing of such devices efficiency η , since<br />
η∼ JSCUOCand UOC ∼ ln( JSC/ J0)<br />
, where J 0 — diode<br />
saturation current density [1-4].<br />
One of the most favorable types of multi-junction<br />
Si-PVC, specially created for the use in CSR conditions<br />
[3,4] under the name of “photo-volt” represents<br />
a monolithic design — the set (more than 10 devices)<br />
silicon flatly-parallel diode cells with p–n junctions<br />
on single crystal oriented perpendicularly to receiving<br />
surface and connected in-series by means of metal layers<br />
between adjacent cells.<br />
The essential advantages of considered PVC type at<br />
CSR conditions in comparison with single-junction Si-<br />
PVC of planar design p–n junction which is oriented<br />
parallel to receiving surface, are: i) potential capability<br />
to much more effective conversion of CSR into electric<br />
energy and ii) generation of 10–30 times greater output<br />
voltage. The last circumstance simplifies the problem<br />
of high-voltage photoelectric systems development and<br />
provides reduction of electrical energy losses in solar<br />
batteries interconnections as well as in the efficacy of<br />
electrical energy transmission from solar batteries to<br />
the consumer. The manufacturing of “photo-volt” type<br />
PVC causes the necessity of the sufficiently expensive<br />
photolithography process use, what disappears due to<br />
the form of the receiving surface (in difference from<br />
planar design PVC [2]) made as crested or grid currentcollecting<br />
electrode with narrow and thin (~10 μm) elements<br />
divided by the gaps less than 1 mm. However,<br />
it is necessary to take into account that the significant<br />
part of CSR is taken by the PVC receiving surface under<br />
the angle α> 0 to the normal [5]. Therefore, J SC , U OC<br />
and transfer efficacy should depend upon α , as far as<br />
the irradiance E of PVC receiving surface changes with<br />
α according to the law 0 cos E E = α, where ation of such type PVC with increased efficacy for the<br />
use at CSR conditions.<br />
On the other hand, the optical location systems<br />
the Si-PVC of “photo-volt” type could be the serious<br />
alternative to the well-known semiconductor radiation<br />
sensors requiring the external source of electrical<br />
energy. Thus, in this case, the angular dependence of<br />
J SC and U OC should be more tangible as possible.<br />
In the present work, the influence of single crystal<br />
Si-PVC “photo-volt” design features on J SC and U OC<br />
dependence upon α was investigated in connection<br />
with the practical importance of two above mentioned<br />
problems. While concerning both problems simultaneously,<br />
the greatest interest is caused by the UOC ( α )<br />
dependence due to the simplicity of this parameter<br />
measurement.<br />
EXPERIMENTAL DETAILES<br />
The serial “photo-volt” type Si-PVC with the area<br />
of receiving surface about 2 cm<br />
E0= E at<br />
α= 0 [6]. The problem of the angular dependence of<br />
multi-junction Si-PVC output parameters’ minimization<br />
is one of the urgent problems with regard to cre-<br />
2 manufactured on the<br />
basis of p-type conductivity single crystal silicon wafer<br />
with resistivity about 10 Ohm∙cm were investigated.<br />
Schematic image of the samples is presented on Fig.1.<br />
Fig. 1. Principal scheme of “photovolt” type multijunction<br />
Si-PVC cross–section: 1 — metal layer of thickness t ≈ 10 μm;<br />
m<br />
2 — layer of n + -type conductivity silicon; 3 — layer of ð-type<br />
conductivity silicon; 4 — layer of p + -type conductivity silicon;<br />
5 — metal electrode.<br />
20<br />
UDC 539.2:648.75<br />
M. V. KIRICHENKO, V. R. KOPACH, R. V. ZAITSEV, S. A. BONDARENKO<br />
National Technical University “Kharkiv Polytechnical Institute”,<br />
21, Frunze Str., 61002, Kharkiv, Ukraine<br />
e-mail:kirichenko_mv@mail.ru<br />
SENSITIVITY OF SILICON PHOTO-VOLTAIC CONVERTERS<br />
TO THE LIGHT INCIDENCE ANGLE<br />
The results of output parameters dependences researches for multijunction silicon photovoltaic<br />
converters (PVC) upon solar radiation incidence angle on their receiving surface are presented. It has<br />
been shown that for improving of PVC efficiency is necessary to achieve the increased values of minority<br />
charge carriers lifetime in their base crystals as well as the optical reflection coefficient for metal/Si<br />
boundaries (interfaces) inside multijunction PVC, while for using multijunction PVC in the optical<br />
location systems the forced reduction of these values is reasonable.<br />
© M. V. Kirichenko, V. R. Kopach, R. V. Zaitsev, S. A. Bondarenko, 2009
Devices had overall dimensions 33 × 6 × 1 mm<br />
and consisted of 35 elementary diode cells by thickness<br />
150 μm each with n + -p-p + -structure which were<br />
connected in-series through the metal inter-layers<br />
with thickness about 10 μm.<br />
The determination of J SC and U OC values for investigated<br />
Si-PVC was carried out by measurement<br />
and following analytical processing of loading illuminated<br />
current versus voltage characteristics LI CVC.<br />
The measurement of LI CVC was carried out similarly<br />
to [7] under the Si-PVC receiving surface irradiation<br />
power of 5712 W/m2 , what corresponds to the degree<br />
of AM0 irradiation concentration equal to 4.2.<br />
For the light incidence angle α change detection,<br />
the investigated Si-PVC was fixed on angle measuring<br />
device, which allows the angle α variation in the range<br />
from 0° up to 90° with the accuracy of 0.01°. Measurements<br />
of LI CVC were carried out at the following values<br />
of α : from 0° up to 20° with a step 2°; from 20° up<br />
to 40° with a step 4°; from 40° up to 60° with a step 5°.<br />
Also, the LI CVC were measured precisely at angles<br />
70°, 80°, 85o and 90°. Temperature of samples at LI<br />
CVC measurements was at the level of 25 °Ñ with the<br />
help of the thermostate. The analytical processing of<br />
LI CVC data was realized similarly to [8].<br />
RESULTS AND DISCUSSION<br />
The normalized angular dependences of open cir-<br />
norm<br />
cuit voltage UOC ( α ) (curve 1) and short circuit cur-<br />
norm<br />
rent J SC ( α ) (curve 2), calculated according to the<br />
experimental values of the corresponding magnitudes<br />
norm J ( )<br />
in the following way: ( ) SC α<br />
J SC α =<br />
,<br />
J SC ( α= 0)<br />
norm U ( )<br />
( ) OC α<br />
UOC<br />
α =<br />
, are presented on the<br />
UOC<br />
( α= 0)<br />
Figure 2. Earlier [9] it was shown that in the range of<br />
α values from 40o up to the Brewster angle ϕ B (74.5o<br />
norm<br />
for silicon) trend of UOC ( α ) dependence is well de-<br />
ln ⎡f ( R,<br />
α) cos α⎤<br />
norm<br />
scribed by the ratio UOC<br />
( α) ≈ 1+<br />
⎣ ⎦<br />
,<br />
2.3(<br />
ξ2 −ξ1)<br />
where 0 ≤ f( R,<br />
α) ≤ 1 is a correcting function, taking<br />
into account the real values of reflection coefficient<br />
from the metal/Si boundaries into “photovolt” type<br />
Si-PVC. In expanded form this ratio is presented in<br />
[9], where ξ 1 < ξ 2 are absolute values of indexes in<br />
degrees of short circuit current and diode saturation<br />
current densities, accordingly. As a result of analysis<br />
norm<br />
of such UOC ( α ) dependence it has been established<br />
that, varying parameters R and Δξ = ξ2 - ξ 1 it is possible<br />
to purposefully effect on its character. So, for<br />
example, it is necessary to maximally increase param-<br />
norm<br />
eters R and Δξ for minimization of UOC ( α ) angular<br />
dependence with the purpose of “photovolt” type Si-<br />
PVC efficiency rising.<br />
Earlier, it was shown [9] that in the range of α values<br />
from 40o up to the Brewster angle B ϕ (74.5o for<br />
norm<br />
silicon), the trend of UOC ( α ) dependence is well de-<br />
ln ⎡f ( R,<br />
α) cos α⎤<br />
norm<br />
scribed by the ratio UOC<br />
( α) ≈ 1+<br />
⎣ ⎦<br />
,<br />
2.3(<br />
ξ2 −ξ1)<br />
where 0 ≤ f( R,<br />
α) ≤ 1 is a correction function taking<br />
into account the real values of reflection coefficient<br />
from the metal/Si boundaries into “photo-volt” type<br />
Si-PVC.<br />
Figure 2. Normalized values of open circuit voltage (1) and<br />
short circuit current density (2) versus light incidence angle on Si-<br />
PVC of “photovolt” type receiving surface.<br />
In expanded form this ratio is presented in [9],<br />
where 1 ξ < ξ 2 are absolute values of short circuit current<br />
and diode saturation current densities, respec-<br />
norm<br />
tively. As a result of analysis of UOC ( α ) dependence,<br />
it has been established that, varying parameters R<br />
and Δξ = ξ2 - ξ 1 , it is possible to effect on its character<br />
purposefully. For example, it is necessary to increase<br />
parameters R and Δξ maximally, for the minimization<br />
norm<br />
of UOC ( α ) angular dependence with the purpose of<br />
“photo-volt” type Si-PVC efficiency rise.<br />
norm<br />
Fig. 3. Theoretical U OC values versus α and Δξ for considered<br />
Si-PVC of “photo-volt” type at the light reflection coefficient<br />
values on metal/silicon boundaries: 1 – R = 1; 2 – R = 0.6;<br />
3 — R = 0.2.<br />
It is suggested to use the “photo-volt” type Si-<br />
PVC as sensor in the optical location systems. Obviously,<br />
for the successful solution of such a problem,<br />
the device, used in the given volume, must provide the<br />
ease of the output signal registration and, also, to have<br />
the expressed, desirably linear dependence of the registered<br />
parameter on the α angle.<br />
As follows from mentioned above, the characteristic<br />
detail of “photo-volt” type Si-PVC is the higher<br />
photo-voltage that provides simple and reliable registration<br />
of the parameter. At the same time, as it is<br />
evident from Fig.2, the concerned “photo-volt” type<br />
21
UOC α dependence<br />
on the light incidence angles on their receiving surface<br />
from 0 up to 74o . However, the results of work [9], allow<br />
to suppose that varying parameters R and Δξ will<br />
norm<br />
provide the strikingly expressed character of UOC ( α)<br />
dependence.<br />
Therefore, we carried out the numerical simula-<br />
ln ⎡f ( R,<br />
α) cos α⎤<br />
norm<br />
tion of UOC<br />
( α) ≈ 1+<br />
⎣ ⎦<br />
dependence<br />
2.3Δξ<br />
at 40î ≤ α ≤ 70î for different values of R and Δξ .<br />
Results of the simulation are presented on Fig.3, as a<br />
norm<br />
family of UOC ( αΔξ , ) surfaces for different values of<br />
the parameter R . From Fig.3 it follows evidently that<br />
varying of parameter R practically does not result in<br />
norm<br />
∂UOC the varying ( α)<br />
norm<br />
being speed of U<br />
∂α OC change<br />
norm<br />
on α , but provides the change of U OC absolute value,<br />
causing this magnitude increase with R growing.<br />
At the same time, as it is evident from Fig. 3, the<br />
norm<br />
∂UOC influence on the ( α)<br />
renders Δξ parameter,<br />
∂α<br />
being the difference of J SC and J 0 orders of values.<br />
Really, from named Fig.3, it is seen that by realization<br />
of situation, characteristic for concerned “photo-volt”<br />
type Si-PVC, when Δξ ≈ 7− 8,<br />
the value<br />
norm<br />
∂UOC ( α) → 0 as well as on the Fig. 2 at α< 74°.<br />
∂α<br />
However, at decrease of difference between J SC and<br />
J 0 , that corresponds to Δξ decrease, dependence of<br />
norm<br />
UOC ( α ) suffers substantial changes and at Δξ = 1− 2<br />
obtains practically linear character in the concerned<br />
range of α angles with sufficiently large value<br />
norm<br />
∂UOC ( α)<br />
∂α ≈ -(7.3- 14.6)∙10-3 rel.un./deg.<br />
Thus, the obtained results argue that in the case<br />
of the “photo-volt” type Si-PVC use as sensors in the<br />
optical location systems, the U OC sensitivity of such<br />
sensors to the light incidence angle on their receiving<br />
surface increase with decrease of the difference between<br />
J SC and J 0 , characterized by parameter Δξ .<br />
Value of the registered parameter U OC increases with<br />
the growth of reflection coefficient from metal/Si<br />
boundaries in “photo-volt” type Si-PVC. At the same<br />
time, it is necessary to take into account the technological<br />
difficulties of R → 1 achievement in the conditions<br />
of the Si-PVC production, and, also, that, as it is<br />
norm<br />
seen from Fig. 3, the value of U OC is less only by 5%<br />
at R = 0.6 than at R = 1.<br />
Therefore, for the use of “photo-volt” type Si-<br />
PVC as the sensor in the optical location systems,<br />
the optimum is achieved at the next combination of<br />
norm<br />
parameters influencing the UOC ( α ) dependence:<br />
Δξ = 1− 2 and R = 0.6 .<br />
At the same time, the achievement of such reflection<br />
coefficient from the metal/Si boundaries into<br />
“photo-volt” type Si-PVC offers no complications at<br />
the real conditions of Si-PVC production.<br />
It is well known [1], that the values of J SC and 0 J<br />
PVC as sensors, it is possible to achieve by a purposeful<br />
decrease of τ np , values in base crystals bulk. Since<br />
−1<br />
τ np , Nr , where N r is bulk concentration of recombination<br />
centers, then, with above mentioned purpose,<br />
the base crystals for such sensors in the process<br />
of appropriate devices manufacturing can be subject<br />
to thermal, mechanical or other types of processing<br />
directed to create in their bulk amount of recombination<br />
centers as greater as possible. It will result in substantial<br />
decrease of τnp , value. A similar effect could<br />
be achieved as well by the use of heavily doped silicon<br />
single crystal for manufacturing of concerned sensors.<br />
Such type of silicon, produced for electronic industry,<br />
has small τ np , values due to high doping level.<br />
CONCLUSION<br />
The results of experimental and theoretical research<br />
of silicon photo-converters’ sensitivity to the<br />
light incidence angle allow to make the following conclusion:<br />
1. The character of UOC ( α ) dependence for multijunction<br />
“photo-volt” type Si-PVC considerably<br />
depends on the minority charge carriers lifetime τ np ,<br />
value in the PVC base crystal, while reflection coefficient<br />
R on metal/Si boundaries of PVC effects the absolute<br />
value of U OC . It has been shown that purposeful<br />
decrease of τ np , value and increase of R value will allow<br />
to create the PVC with practically linear and easily<br />
registered UOC ( α ) dependence.<br />
2. The obtained character of UOC ( α ) dependence<br />
will allow to use the multijunction “photo-volt” type<br />
Si-PVC as sensors in the optical location systems.<br />
References<br />
1. Blakers A.W., Smeltik J., Proceedings of the 2<br />
, and consequently Δξ , substantially depend from minority<br />
charge carriers lifetime τ np , in PVC base crystals.<br />
Therefore the required value of Δξ at using such<br />
nd World Conference<br />
and Exhibition on Photovoltaic Solar Energy Conversion;<br />
July 6–10, 1998, Vienna, Austria, p. 2193.<br />
2. Verlinden P.J. High–efficiency concentrator silicon solar cells,<br />
p. 436–455. In: “Practical Handbook of Photovoltaics: Fundamentals<br />
and Applications”, edited by T. Markvart and L.<br />
Castaner; Elsevier Science Ltd., Kidlington, Oxford, 2003.<br />
3. Sater B.L., Sater N.D. High voltage silicon VMJ solar cells for<br />
up to 1000 suns intensities // Proceedings of the 29th IEEE<br />
Photovoltaic Specialists Conference, May 20 — May 24,<br />
2002. — New Orleans , USA, 2002. — P. 1019–1022.<br />
4. Guk E.G., Shuman V.B., Shwartz M.Z. Proceedings of the<br />
14th European Photovoltaic Solar Energy Conference, June<br />
30 — July 4, 1997, Barcelona, Spain, P. 154.<br />
5. Àíäðååâ Â.Ì., Ãðèëèõåñ Â.À., Ðóìÿíöåâ Â.Ä. Ôîòîýëåêòðè-<br />
÷åñêîå ïðåîáðàçîâàíèå êîíöåíòðèðîâàííîãî ñîëíå÷íîãî<br />
èçëó÷åíèÿ. — Ëåíèíãðàä: Íàóêà, 1999. — 310 ñ.<br />
6. Ëàíäñáåðã Ã.Ñ. Îïòèêà. — Ìîñêâà: Íàóêà, 1996. — 928 ñ.<br />
7. Keogh W., Cuevas A. Simple flashlamp I–V testing of solar<br />
cells // Proceedings of the 26th IEEE Photovoltaic Specialists<br />
Conference, Anaheim, CA, September 30 — October 3. —<br />
1997. — P. 199–202.<br />
8. Kerschaver E., Einhaus R., Szlufcik J., Nijs J.,Mertens R.<br />
Simple and fast extraction technique for the parameters in<br />
the double exponential model for I–V characteristic of solar<br />
cells // Proceedings of the 14th European Photovoltaic Solar<br />
Energy Conference, June 30 — July 4, 1997. — Barcelona,<br />
Spain, 1997. — P. 2438–2441.<br />
9. Kopach V.R., Kirichenko M.V., Shramko S.V. et al. New approach<br />
to the efficiency increase problem for multijunction<br />
silicon photovoltaic converters with vertical diode cells //<br />
Functional Materials. — 2008. — Vol. — No. 2. — P. 253-<br />
258.<br />
norm<br />
Si-PVC has weakly expressed ( )<br />
22
UDC 539.2:648.75.<br />
M. V. Kirichenko, V. R. Kopach, R. V. Zaitsev, S. A. Bondarenko<br />
SENSITIVITY OF SILICON PHOTO-VOLTAIC CONVERTERS TO THE LIGHT INCIDENCE ANGLE<br />
Àbstract<br />
The results of output parameters dependences researches for multijunction silicon photovoltaic converters (PVC) upon solar radiation<br />
incidence angle on their receiving surface are presented. It has been shown that for improving of PVC efficiency is necessary<br />
to achieve the increased values of minority charge carriers lifetime in their base crystals as well as the optical reflection coefficient for<br />
metal/Si boundaries (interfaces) inside multijunction PVC, while for using multijunction PVC in the optical location systems the forced<br />
reduction of these values is reasonable.<br />
Key words: sensitivity, photovoltaic converters, receiving surface.<br />
ÓÄÊ 539.2:648.75<br />
Ì. Â. Êèðè÷åíêî, Â. Ð. Êîïà÷, Ð. Â. Çàéöåâ, Ñ. À. Áîíäàðåíêî<br />
×ÓÂÑÒÂÈÒÅËÜÍÎÑÒÜ ÊÐÅÌÍÈÅÂÛÕ ÔÎÒÎÝËÅÊÒÐÈ×ÅÑÊÈÕ ÏÐÅÎÁÐÀÇÎÂÀÒÅËÅÉ Ê ÓÃËÓ ÏÀÄÅÍÈß<br />
ÑÂÅÒÀ ÍÀ ÈÕ ÏÐÈÅÌÍÓÞ ÏÎÂÅÐÕÍÎÑÒÜ<br />
Ðåçþìå<br />
Ïðèâåäåíû ðåçóëüòàòû èññëåäîâàíèé çàâèñèìîñòåé âûõîäíûõ ïàðàìåòðîâ ìíîãîïåðåõîäíûõ êðåìíèåâûõ ôîòîýëåêòðè-<br />
÷åñêèõ ïðåîáðàçîâàòåëåé (ÔÝÏ) îò óãëà ïàäåíèÿ ñîëíå÷íîãî èçëó÷åíèÿ íà èõ ïðèåìíóþ ïîâåðõíîñòü. Ïîêàçàíî, ÷òî äëÿ óâåëè÷åíèÿ<br />
ÊÏÄ ÔÝÏ íåîáõîäèìî îáåñïå÷èòü ïîâûøåíèå çíà÷åíèé âåëè÷èí âðåìåíè æèçíè íåîñíîâíûõ íîñèòåëåé çàðÿäà â<br />
áàçîâûõ êðèñòàëëàõ è êîýôôèöèåíòà îïòè÷åñêîãî îòðàæåíèÿ îò ãðàíèö ìåòàëë/Si âíóòðè ìíîãîïåðåõîäíûõ ÔÝÏ, â òî âðåìÿ<br />
êàê ïðè èñïîëüçîâàíèè ìíîãîïåðåõîäíûõ ÔÝÏ â ñèñòåìàõ îïòè÷åñêîé ëîêàöèè öåëåñîîáðàçíûì ÿâëÿåòñÿ ïðèíóäèòåëüíîå<br />
ñíèæåíèå ýòèõ âåëè÷èí.<br />
Êëþ÷åâûå ñëîâà: ôîòîýëåêòðè÷åñêèå ïðåîáðàçîâàòåëè, ÷óâñòèòåëüíîñòü, ïðè¸ìíàÿ ïîâåðõíîñòü.<br />
ÓÄÊ 539.2:648.75<br />
Ì. Â. ʳð³÷åíêî, Â. Ð. Êîïà÷, Ð. Â. Çàéöåâ, Ñ. Î. Áîíäàðåíêî<br />
×ÓÒËȲÑÒÜ ÊÐÅÌͲªÂÈÕ ÔÎÒÎÅËÅÊÒÐÈ×ÍÈÕ ÏÅÐÅÒÂÎÐÞÂÀײ ÄÎ ÊÓÒÀ ÏÀIJÍÍß Ñ²ÒËÀ ÍÀ ¯Õ<br />
ÏÐÈÉÌÀËÜÍÓ ÏÎÂÅÐÕÍÞ<br />
Ðåçþìå<br />
Íàâåäåíî ðåçóëüòàòè äîñë³äæåíü çàëåæíîñòåé âèõ³äíèõ ïàðàìåòð³â áàãàòîïåðåõ³äíèõ êðåìí³ºâèõ ôîòîåëåêòðè÷íèõ<br />
ïåðåòâîðþâà÷³â (ÔÅÏ) â³ä êóòà ïàä³ííÿ ñîíÿ÷íîãî âèïðîì³íþâàííÿ íà ¿õ ïðèéìàëüíó ïîâåðõíþ. Ïîêàçàíî, ùî äëÿ çá³ëüøåííÿ<br />
ÊÊÄ ÔÅÏ íåîáõ³äíî çàáåçïå÷èòè ï³äâèùåí³ çíà÷åííÿ ÷àñó æèòòÿ íåîñíîâíèõ íîñ³¿â çàðÿäó â áàçîâèõ êðèñòàëàõ òà<br />
êîåô³ö³ºíòà îïòè÷íîãî â³äáèòòÿ â³ä ãðàíèöü ìåòàë/Si âñåðåäèí³ áàãàòîïåðåõ³äíèõ ÔÅÏ, ó òîé ÷àñ, ÿê ïðè âèêîðèñòàíí³ áàãàòîïåðåõ³äíèõ<br />
ÔÅÏ ó ñèñòåìàõ îïòè÷íî¿ ëîêàö³¿ âèçíà÷åííÿ íàïðÿìó ðîçïîâñþäæåííÿ âèïðîì³íþâàííÿ äîö³ëüíèì º ïðèìóñîâå<br />
çíèæåííÿ öèõ âåëè÷èí.<br />
Êëþ÷îâ³ ñëîâà: ôîòîåëåêòðè÷í³ ïåðåòâîðþâà÷³, ïðèéìàëüíà ïîâåðõíÿ, ÷óòëèâ³ñòü.<br />
23
24<br />
UDC 535.5<br />
L. S. MAXIMENKO 1 , I. E. MATYASH 1 , S. P. RUDENKO 1 , B. K. SERDEGA 1 ,<br />
V. S. GRINEVICH 2 , V. A. SMYNTYNA 2 , L. N. FILEVSKAYA 2<br />
1 Lashkarev Institute of Physics of Semiconductors, National Academy of Sciences of Ukraine,Kiev-28, Prospect Nauki,45<br />
2 Odessa I.I. Mechnikov National University, Odessa, Ukraine, Odessa, 65082, Dvoryanskaya str.2 grinevich@onu.edu.ua<br />
SPECTROSCOPY OF POLARISED AND MODULATED LIGHT<br />
FOR NANOSIZED TINDIOXIDE FILMS INVESTIGATION<br />
The peculiarities of Surface Plasmons Resonance (SPR) in nanosized tin dioxide films, deposited<br />
on a prism of total interior reflection, were experimentally investigated using methods of the polarized<br />
and modulated radiation of light (PM). It was found that the layers, obtained by special technology<br />
using polymer materials as structuring additives are the combination of polycrystalline nanosized<br />
grains with air pores. This result has confirmed the supposition about the considerable porosity of the<br />
obtained layers. The obtained results confirm the considerable (PM) method’s sensitivity for the aims<br />
of material’s optical parameters detecting.<br />
INTRODUCTION<br />
In the modern gas analyses there is a natural transition<br />
to the thin films’ adsorptive sensitive elements<br />
with a perfectly developed physical surface based on<br />
oxide nanodimensional materials. One of such a material<br />
is tin dioxide which has perfect sensitivity to<br />
a composition of a environmental atmosphere and<br />
chemical resistivity to a poisoning media. Transparent<br />
Tin dioxide films with nano sized grains may be<br />
applied as optical detectors of environmental compositions.<br />
It has become possible due to the noticed<br />
optical property of such films to answer the presence<br />
of different chemical compounds in the environment<br />
(both gaseous and liquid).<br />
The later circumstance defined the urgency of the<br />
detailed researches of such films’ optical properties.<br />
Among possible diagnostic methods the surface<br />
plasmons resonance phenomenon (SPR) is a unique<br />
one because it is a basis of the most sensitive methods<br />
applied for the registration of media dielectric<br />
functions changes. The application of polarized and<br />
modulated light for detecting of SPR in oxide materials<br />
with nano grains is becoming proved and effective.<br />
The peculiarities of the SPR in the nanosized Tin<br />
dioxide films deposited on the surface of total internal<br />
reflection prism were investigated using the technique<br />
of modulation of polarized light radiation (PM). The<br />
research supposed the presence of electrons’ plasma in<br />
the obtained layers, which is indirectly confirmed by a<br />
property of a considerable electrons’ degeneration in<br />
the films.<br />
TECHNOLOGY OF SAMPLES<br />
PREPARATION AND THE INVESTIGATIONS<br />
TECHNIQUE<br />
Transparent nanodimensional Tin dioxide films,<br />
obtained using polymer materials as structuring additives<br />
and Tin containing precursor of SnO 2 were used as<br />
samples. The samples’ preparation methods described<br />
in [1] had several stages: preparation of tin containing<br />
organic filler, preparation of the polymer material so-<br />
lution, and introduction of tin containing compound<br />
into it. The resultant gel was deposited on a glass substrate<br />
and annealed in a muffle furnace. Temperature<br />
(500 0 C) and the annealing time (2 hours) were chosen<br />
as necessary parameters for both polymer and tin dioxide<br />
precursor decomposition. Thin tin dioxide layers<br />
with well developed nanostructure and considerable<br />
porosity were formed after the complete removal of decay<br />
products by means of annealing both of polymer<br />
and tin dioxide containing precursor and also after the<br />
complete oxidation of the films up to tin dioxide.<br />
The nanostructure of the films was determined at<br />
the AFM [2] investigations and the typical surface of<br />
such films (AFM image) is shown on Fig. 1.<br />
Fig. 1. The films investigated typical surface AFM image.<br />
In the presented work the films were obtained from<br />
solutions having 0,03% of PVA as a structural additive<br />
and 1% of Tin containing SnO 2 precursor.<br />
The total internal reflection on the border film-external<br />
environment was used as a detecting phenomenon.<br />
The theoretical evidence and the experimental<br />
© L. S. Maximenko, I. E. Matyash, S. P. Rudenko, B. K. Serdega, V. S. Grinevich, V. A. Smyntyna, L. N. Filevskaya, 2009
technique of the total internal reflection measurement<br />
are described in details in [3].<br />
Fig. 2. The experimental equipment optical scheme for angles<br />
characteristics of a polarization difference measurements<br />
using modulation of polarization: LGN- Helium-Neon<br />
Laser, PEM-photo elastic modulator of polarization, FP- the<br />
phase plate, p,s linear polarizations, azimuths of them are parallel<br />
and perpendicular to the plane of incidence, G-Glann s<br />
prism, FD-photo detector, ôcr the total internal reflection critical<br />
angle, N0, N1,N2- refraction indices of glass, films, air correspondingly.<br />
The general optical scheme of the experimental<br />
equipment for measuring of both total internal reflection<br />
characteristics and polarization parameters difference<br />
is given at Fig.2. Hellium-Neon laser, LGN-113<br />
with the fixed wave lengths 0,63 mkm and 1,15 mkm<br />
is used as a source of linearly polarized radiation, together<br />
with monochromator MDR-4 with a halogen<br />
lamp KGM-150 at the input and the polarizer on the<br />
out let. A modulator of polarization –REM was used<br />
as a dynamic phase plate functioning in two regimes.<br />
In the both cases, by rotating the modulator round the<br />
optical axe of the measuring device, it’s position was<br />
chosen so, that the out let polarized radiation azimuth<br />
was parallel and perpendicular in turn to the plane of<br />
accident (p and s polarization, correspondingly ).<br />
After the interaction with the half cylinder and the<br />
surface of resonance sensitive Tin dioxide film, the radiation<br />
was directed to photo detector, FD (Silicium<br />
or Germanium photodiode) which after absorption of<br />
radiation generates a signal comprising an alternative<br />
component. It is proportional to the reflection indices<br />
difference of p and s polarization of the detected<br />
radiation.<br />
Thus, the equipment with the modulation of polarized<br />
light permits to obtain not only the angle dependant<br />
reflection indices R and R correspondingly,<br />
s p<br />
but also their difference, ΔR. ΔR factor is not a result<br />
of mathematical act, but is the physical value independently<br />
and directly measured.<br />
The investigated films were deposited on glass<br />
substrates which permited to supplement the glass segment<br />
to a half cylinder, using the contact of a substrate<br />
with a segment by immersion liquid (glycerin).<br />
THE MEASUREMENTS RESULTS AND<br />
THEIR DISCUSSION<br />
All the three characteristics (R s , R p and ΔR) for<br />
one of the samples are given at Fig.3 in relative, but<br />
comparable units. Wave length of the scanning radiation,<br />
λ =630 nm.<br />
ΔR, R , R , arb.un. ���. ���. ��. ��.<br />
S P<br />
5,0<br />
4,5<br />
4,0<br />
3,5<br />
3,0<br />
2,5<br />
2,0<br />
1,5<br />
1,0<br />
0,5<br />
0,0<br />
-0,5<br />
30 40 50 60 70<br />
����, ����, ���. ���.<br />
Angle, grad<br />
ΔR<br />
R S<br />
R P<br />
Fig. 3. Experimental dependencies for reflection indices Rs<br />
and Rp and their difference ΔR of the investigated samples.<br />
�R, ΔR, abs.un. ���. ��.<br />
0,00<br />
-0,01<br />
-0,02<br />
-0,03<br />
-0,04<br />
-0,05<br />
400 500 600 700 800 900 1000<br />
����� �����, ��<br />
�, nm<br />
experimental<br />
�����������<br />
theoretical ������<br />
Fig. 4. The polarization difference vs wave length at a fixed<br />
angle, which is more than critical.<br />
The most interesting is the angles’ range where<br />
their values are more than critical ones where an inequality<br />
of R and R testifies about the broken total<br />
s p<br />
internal reflection, which is possible when an absorption<br />
has place in the sample. As soon as, the difference<br />
between R and R is registered on the background of<br />
s p<br />
considerable signals, then the big error is inevitable.<br />
This not favourable situation may be eliminated in a<br />
case when a polarization difference, which is measured<br />
relatively zero, may be amplified and, hence, fixed<br />
for sure. Thus, the polarization difference, ΔR=R - s<br />
R certainly demonstrates the break of total internal<br />
p<br />
reflection which change its sign at an angle equal to<br />
the critical one. It was a definite interest to retrace the<br />
differential difference dependence vs wave length at a<br />
fixed incidence angle which is more than critical. The<br />
experimental results at the incidence angle of 460 are<br />
shown at Fig.4 by the continues line in comparison<br />
with the theoretically calculated ones on the basis of<br />
Frehnels equations — by the broken line. The adjustment<br />
parameters were refraction and absorption indices<br />
for the Tin dioxide film.<br />
As it is seen at Fig.4 the coincidence of experi-<br />
25
mental and theoretical results is more than satisfactory<br />
in the wave length range 500-1000 nm. It is appeared,<br />
that n and k are linearly and uniquely dependent on the<br />
wave length: n=1, 28+0,000005 * λ; k= 0,223-λ/4700,<br />
λ in nm. Changes of n or k on more than 5 * 10 -3 gives<br />
the mismatch of experimental and theoretical data.<br />
Changes of k move a theoretical curve higher or<br />
lower the experimental one, but variation of n changes<br />
the theoretical curve’ slope.<br />
The reflective index values obtained in our work<br />
are in good agreement with other authors’ data [4]/<br />
That is, the reflective index value obtained as a result<br />
of our samples investigation is within the indices<br />
values interval corresponding to indices of pure Tin<br />
dioxide (n=1,56) and for air (n=1,003).The pores existence<br />
was expected in the investigated layers which is<br />
resulted from polymers application at the films’ production.<br />
Pores within the said technology are formed<br />
at polymer’s decay during the annealing procedure.<br />
Thus, the above discussed results confirm the considerable<br />
sensitiveness of the method for our material’s<br />
optical parameters detecting.<br />
26<br />
CONCLUSION<br />
The new technique for the Surface Plasmons<br />
Resonans parameters measuring by means of modulation<br />
of polarized light was applied to nanosized Tin<br />
dioxide layers in order to obtain their optical indices.<br />
UDC 535.5<br />
The system’s — porous layer of SnO 2 reflection index<br />
showed its satisfactory agreement with other authors’<br />
data. The obtained reflection index values characterize<br />
the object as a system of nanosize Tin dioxide with<br />
air pores. This result confirms the supposition of such<br />
pores presence in the films.<br />
The polarized light modulation method has high<br />
sensitivity for optical parameters detecting at Plasmons<br />
Resonans in the investigated layers and, hence,<br />
gives good perspectives in gaseous environment detecting.<br />
Such a technique of PM in SPR of the layers<br />
is, besides all, is a perfect confirmation of electrons’<br />
plasma presence in them.<br />
References<br />
1. L.N. Filevskaya, V.A. Smyntyna, V.S. Grinevich. Morphology<br />
of nanostructured SnO 2 films prepared with polymers<br />
employment // Photoelectronics. Inter-universities scientific<br />
articles, 2006, ¹ 15, p.p.11-13.<br />
2. V.A. Smyntyna, L.N. Filevskaya, V.S. Grinevich, Morphology<br />
and optical properties of SnO 2 nanofilms // Semiconductor<br />
physics, quantum electronics & optoelectronics, Volume 11,<br />
¹ 2, 2008, p.p.163-171.<br />
3. Ë.È. Áåðåæèíñêèé, Ë.Ñ. Ìàêñèìåíêî, È.Å. Ìàòÿø,<br />
Ñ.Ï. Ðóäåíêî, Á.Ê. Ñåðäåãà. Ïîëÿðèçàöèîííî-ìîäóëÿöèîííàÿ<br />
ñïåêòðîñêîïèÿ ïîâåðõíîñòíîãî ïëàçìîííîãî<br />
ðåçîíàíñà // Îïòèêà è ñïåêòðîñêîïèÿ. 2008. — Ò.105. —<br />
¹2. — Ñ. 281-289.<br />
4. M.Anastasescu, M.Gartner, S.Mihaiu, C.Anastasescu,<br />
M.Purica, E.Manea, M.Zaharescu. Optical and structural<br />
properties of SnO 2 — based sol-gel thin films // International<br />
Semiconductor Conference, 2006, Volume 1, Issue, Sinaia,<br />
Romania, 27-29 Sept. 2006, Page(s):163 — 166.<br />
L. S. Maximenko, I. E. Matyash, S. P. Rudenko, B. K. Serdega, V. S. Grinevich, V. A. Smyntyna, L. N. Filevskaya<br />
SPECTROSCOPY OF POLARISED AND MODULATED LIGHT FOR NANOSIZED TINDIOXIDE FILMS<br />
INVESTIGATION<br />
Abstract<br />
The peculiarities of the Surface Plasmons Resonance (SPR) in nanosized tin dioxide films, deposited on the prism of total interior<br />
reflection, were experimentally investigated using methods of the polarized and modulated radiation of light (PM). It was found that the<br />
layers, obtained by special technology using polymer materials as structuring additives are the combination of polycrystalline nanosized<br />
grains with air pores. This result has confirmed the supposition about the considerable porosity of the obtained layers. The obtained<br />
results confirm the considerable PM method’s sensitivity for the aims of material’s optical parameters detecting.<br />
Key words: polarizing modulation, thin film, tin dioxide.<br />
ÓÄÊ 535.5<br />
Ë. Ñ. Ìàêñèìåíêî, È. Å. Ìàòÿø, Ñ. Ï. Ðóäåíêî, Á. Ê. Ñåðäåãà, Â. Ñ. Ãðèíåâè÷, Â. À. Ñìûíòûíà, Ë. Í. Ôèëåâñêàÿ<br />
ÏÎËßÐÈÇÀÖÈÎÍÍÎ-ÌÎÄÓËßÖÈÎÍÍÀß ÑÏÅÊÒÐÎÑÊÎÏÈß ÍÀÍÎÐÀÇÌÅÐÍÛÕ ÏËÅÍÎÊ ÄÂÓÎÊÈÑÈ<br />
ÎËÎÂÀ<br />
Ðåçþìå<br />
Ýêñïåðèìåíòàëüíî ñ ïðèìåíåíèåì ìåòîäèêè, îñíîâàííîé íà ïîëÿðèçàöèîííîé ìîäóëÿöèè (ÏÌ) èçëó÷åíèÿ, èññëåäîâàíû<br />
îñîáåííîñòè ïîâåðõíîñòíîãî ïëàçìîííîãî ðåçîíàíñà (ÏÏÐ) â íàíîðàçìåðíûõ ïëåíêàõ äèîêñèäà îëîâà, íàíåñåííûõ<br />
íà ïîâåðõíîñòü ïðèçìû ïîëíîãî âíóòðåííåãî îòðàæåíèÿ. Èññëåäóåìûå ñëîè ÿâëÿþòñÿ ñî÷åòàíèåì ïîëèêðèñòàëëè÷åñêèõ<br />
íàíîðàçìåðíûõ çåðåí ñ âîçäóøíûìè ïîðàìè, ÷òî ïîäòâåðäèëî ïåðâîíà÷àëüíîå ïðåäïîëîæåíèå î çíà÷èòåëüíîé ïîðèñòîñòè<br />
ïîëó÷åííûõ ñ èñïîëüçîâàíèåì ïîëèìåðîâ ïëåíîê äâóîêèñè îëîâà. Ïîëó÷åííûå ðåçóëüòàòû ñâèäåòåëüñòâóþò î çíà÷èòåëüíîé<br />
÷óâñòâèòåëüíîñòè ìåòîäà ÏÌ â îïðåäåëåíèè îïòè÷åñêèõ ïàðàìåòðîâ ìàòåðèàëà.<br />
Êëþ÷åâûå ñëîâà: ïîëÿðèçàöèîííàÿ ìîäóëÿöèÿ, òîíêèå ïëåíêè, äâóîêèñü îëîâà.
ÓÄÊ 535.5<br />
Ë. Ñ. Ìàêñèìåíêî, ². ª. Ìàòÿø, Ñ. Ï. Ðóäåíêî, Á. Ê. Ñåðäåãà, Â. Ñ. Ãð³íåâè÷, Â. À. Ñìèíòèíà, Ë. Ì. Ô³ëåâñüêà<br />
ÏÎËßÐÈÇÀÖ²ÉÍÎ-ÌÎÄÓËßÖ²ÉÍÀ ÑÏÅÊÒÐÎÑÊÎÏ²ß ÍÀÍÎÐÎÇ̲ÐÍÈÕ Ï˲ÂÎÊ ÄÂÎÎÊÈÑÓ ÎËÎÂÀ<br />
Ðåçþìå<br />
Åêñïåðèìåíòàëüíî ³ç çàñòîñóâàííÿì ìåòîäèêè, çàñíîâàíî¿ íà ïîëÿðèçàö³éí³é ìîäóëÿö³¿ (ÏÌ) âèïðîì³íþâàííÿ, äîñë³äæåí³<br />
îñîáëèâîñò³ ïîâåðõíåâîãî ïëàçìîííîãî ðåçîíàíñó (ÏÏÐ) ó íàíîðîçì³ðíèõ ïë³âêàõ ä³îêñèäó îëîâà, íàíåñåíèõ íà ïîâåðõíþ<br />
ïðèçìè ïîâíîãî âíóòð³øíüîãî â³äáèòòÿ. Äîñë³äæóâàí³ øàðè º ñïîëó÷åííÿì ïîë³êðèñòàë³÷íèõ íàíîðîçì³ðíèõ çåðåí<br />
ç ïîâ³òðÿíèìè ïîðàìè, ùî ï³äòâåðäèëî ïåðâ³ñíå ïðèïóùåííÿ ïðî çíà÷íó ïîðèñò³ñòü îòðèìàíèõ ç âèêîðèñòàííÿì ïîë³ìåð³â<br />
ïë³âîê äâîîêèñó îëîâà. Îòðèìàí³ ðåçóëüòàòè ñâ³ä÷àòü ïðî çíà÷íó ÷óòëèâ³ñòü ìåòîäó ÏÌ ó âèçíà÷åíí³ îïòè÷íèõ ïàðàìåòð³â<br />
ìàòåð³àëó.<br />
Êëþ÷îâ³ ñëîâà: ïîëÿðèçàö³éíà ìîäóëÿö³ÿ, òîíê³ ïë³âêè, äâîîêèñ îëîâà.<br />
27
28<br />
UDC 621.315.592<br />
O. O. PTASHCHENKO 1 , F. O. PTASHCHENKO 2 , O. V. YEMETS 1<br />
1 I. I. Mechnikov National University of Odessa, Dvoryanska St., 2, Odessa, 65026, Ukraine<br />
2 Odessa National Maritume Academy, Odessa, Didrikhsona St., 8, Odessa, 65029, Ukraine<br />
EFFECT OF AMBIENT ATMOSPHERE ON THE SURFACE CURRENT<br />
IN SILICON P-N JUNCTIONS<br />
1. INTRODUCTION<br />
Gas sensors on p-n junctions [1, 2] have some<br />
advantages in comparison with these, based on oxide<br />
polycrystalline films [3] and Shottky diodes [4]. P-n<br />
junctions on wide-band semiconductors have high<br />
potential barriers for current carriers, which results in<br />
low background currents, high sensitivity and selectivity<br />
to the gas components [5, 6].<br />
The sensitivity of p-n sensors to donor gases as<br />
ammonia is due to forming of a surface conducting<br />
channel in the electric field induced by the ammonia<br />
ions adsorbed on the surface of the natural oxide layer<br />
[1, 2]. This mechanism is valid only for adsorbed molecules<br />
which are ionized on the semiconductor surface.<br />
And it causes the gas selectivity of these sensors.<br />
The surface conducting channel is produced in these<br />
sensors under condition<br />
m<br />
Ns > Ns<br />
, (1)<br />
m<br />
where N s and Ns are the surface densities of adsorbed<br />
molecules (ions) and surface electron states in<br />
the semiconductor, respectively. This determines the<br />
threshold gas concentration for these sensors.<br />
Characteristics of p-n junctions in silicon as ammonia<br />
sensors were studied in previous works [7, 8].<br />
It was shown that ammonia sensitivity of these structures<br />
is due to enhancing of the surface recombination,<br />
caused by NH molecules adsorption. The<br />
3<br />
difference in the sensitivity mechanism can lead to<br />
differences in selectivity and other characteristics of<br />
sensors.<br />
The purpose of this work is a comparative study of<br />
the influence of ammonia, water and ethylene vapors<br />
on stationary surface currents in silicon p-n junctions,<br />
as well as on their kinetics.<br />
2. EXPERIMENT<br />
The influence of ammonia, water and ethylene vapors on I-V characteristics of forward and reverse<br />
currents, as well as on the kinetics of the surface current in silicon p-n structures was studied.<br />
All these vapors enhance both the forward and reverse currents. The gas sensitivity of p-n structures at<br />
forward biases is due to enhanced surface recombination, while at reverse biases a surface conductive<br />
channel shorts the p-n junction. The sensitivity to ammonia is much higher than to other vapors. It is<br />
explained as a result of donor properties of NH 3 molecules. The response time of silicon p-n junctions<br />
as gas sensors at room temperature is below 60 s.<br />
I-V measurements were carried out on silicon pn<br />
junctions with the structure described in previous<br />
works [7, 8]. The effect of vapors over water solutions<br />
of several NH 3 concentrations, over distilled water<br />
and over liquid ethylene was studied on stationary I-V<br />
characteristics, as well as on kinetics of surface current<br />
in p-n junctions.<br />
I-V characteristic of the forward current in a typical<br />
p-n structure is presented as curve 1 in Fig. 1. Over<br />
the current range between 10 nA and 1mA the I–V<br />
curve can be described with the expression<br />
IV ( ) = I0exp( qV/ nkT)<br />
, (2)<br />
where I is a constant; q is the electron charge; V de-<br />
0<br />
notes bias voltage; k is the Boltzmann constant; T is<br />
temperature; n ≈ 1.1 is the ideality constant. Some deviation<br />
from the value n=1 can be ascribed to recombination<br />
on deep levels in p-n junction and (or) at the<br />
surface [9]. Curves 2, 3, 4 in Fig. 1 were obtained in<br />
air with vapors of water, ethylene, and ammonia, respectively.<br />
The partial pressures of water, ethylene and<br />
ammonia vapors were of 2000Pa, 5000Pa, and 50 Pa,<br />
accordingly. A comparison between curves 1, 2, and<br />
3 in Fig. 1 shows that the sensitivity of p-n structures<br />
to ammonia vapors is the highest and to water — the<br />
lowest.<br />
10 -5<br />
10 -6<br />
10 -7<br />
10 -8<br />
I,A<br />
2<br />
0 0,1 0,2 0,3 0,4<br />
V, Volts<br />
Fig. 1. Forward branches of I-V characteristics of a p-n structure<br />
in air (1) and in vapors of water (2), ethylene (3) and ammonia<br />
(4).<br />
© O. O. Ptashchenko, F. O. Ptashchenko, O. V. Yemets, 2009<br />
3<br />
4<br />
1
Fig. 2 represents I–V characteristics of the reverse<br />
current in a p-n junction. Curve 1 was measured in air,<br />
and curves 2–4 were obtained in air with vapors of water,<br />
ethylene, and ammonia, respectively. It is evident<br />
from Fig. 2 that the studied vapors strongly enhance<br />
the reverse current in silicon p-n structures.<br />
10 -5<br />
10 -6<br />
10 -7<br />
10 -8<br />
I,A<br />
4<br />
0 1 2 3 4 5<br />
V, Volts<br />
Fig. 2. Reverse branches of I-V characteristics of a p-n structure<br />
in air (1) and in vapors of water (2), ethylene (3) and ammonia<br />
(4).<br />
Curves 1, 2 and 3 in Fig. 3 depict I–V characteristics<br />
of the additional current in a p-n structure, due<br />
to adsorption of water, ethylene, and NH molecules,<br />
3<br />
accordingly. It is seen that, at a high enough reverse<br />
voltage, the additional reverse currents are higher than<br />
the corresponding forward currents.<br />
2<br />
0<br />
-2<br />
-4<br />
-6<br />
I, �A<br />
1<br />
1A<br />
2<br />
3<br />
-8<br />
-5 -4 -3 -2 -1 0<br />
V, Volts<br />
Fig. 3. I-V characteristics of the additional currents in a p-n<br />
structure in vapors of water (1, 1A), ethylene (2) and ammonia (3).<br />
Ordinates of curve 1A are multiplied by 10.<br />
The (absolute, current-) sensitivity of a gas sensor<br />
can be defined as<br />
S =ΔI Δ P , (3)<br />
I<br />
3<br />
2<br />
1<br />
where ΔI is the change in the current (at a fixed voltage),<br />
which is due to a change ΔP in the corresponding<br />
gas partial pressure [10]. And the relative sensitivity is<br />
S =ΔI ( I Δ P)<br />
, (4)<br />
R<br />
0<br />
where I denotes the current in the pure air at the same<br />
0<br />
bias voltage.<br />
The sensitivities of studied p-n junctions to water,<br />
ethylene (C H OH) and ammonia vapors at a forward<br />
2 5<br />
bias voltage of 0.25 V and a reverse bias voltage of 3 V<br />
are presented in Tab. 1.<br />
Gas sensitivities of p-n structures<br />
Table 1<br />
H 2 O C 2 H 5 OH NH 3<br />
S I (0.25 V), nA/kPa 6 11 11000<br />
S I (–3 V), nA/kPa 70 800 20000<br />
S R (0.25 V), 1/kPa 0.1 0.23 200<br />
S R (-3 V), 1/kPa 3 50 900<br />
It is seen in Tab. 1 that the studied p-n structures<br />
can be used, practically, as ammonia selective sensors.<br />
In an ammonia-free atmosphere these structures are<br />
sensors of water and ethylene. The reverse bias is preferable<br />
for the sensors.<br />
Fig. 4 illustrates the kinetics of forward (a) and reverse<br />
(b) currents in a p-n structure after let in- and<br />
out of ammonia vapor with a partial pressure of 50 Pa.<br />
Similar curves were measured for water- and ethylene<br />
vapors. The response time t r for current rise was estimated<br />
as the duration of the current increase to 90%<br />
of its stationary value after letting in the corresponding<br />
vapor into the container with the sample. And the decay<br />
time t d was obtained in a similar way, for the current<br />
decrease from the stationary value to 10% of it. These<br />
procedures were carried out in regimes of forward and<br />
reverse bias. The resulting response- and decay times<br />
are presented in Tab. 2.<br />
I, ��<br />
0,70<br />
0,68<br />
0,66<br />
0,64<br />
0<br />
I, �� ��<br />
3,1<br />
3,0<br />
2,9<br />
100 200 300<br />
t, s<br />
2,8<br />
0 100 200 300<br />
t, s<br />
Fig. 4. Kinetics of forward (a) and reverse (b) currents<br />
in a p-n structure after let in- and out of ammonia<br />
vapor with a partial pressure of 50 Pa.<br />
a<br />
b<br />
29
30<br />
Rise- and decay times of p-n gas sensors<br />
H 2 O C 2 H 5 OH NH 3<br />
t r (0.25 V), s 30–35 10–16 25–40<br />
t d (0.25 V), s 10–15 5–8 20–30<br />
t r (-3 V), s 35–55 50–55 25–30<br />
t d (-3 V), s 25–30 10–12 8–9<br />
Table 2<br />
The data in Tab. 2 show that the rise time of the<br />
signal for all the studied vapors is longer than the decay<br />
time in both regimes. And the response time of the<br />
p-n structures as gas sensors does not exceed 55 s.<br />
3. DISCUSSION<br />
The mechanism of the ammonia sensitivity of the<br />
forward current in silicon p-n structures was discussed<br />
in previous works [7, 8]. Adsorbed and subsequently<br />
ionized molecules of a donor gas form the electric<br />
field which bends down c- and v- bands in the crystal<br />
at the surface. Under a high enough gas partial pressure,<br />
a surface channel with electron conductivity is<br />
formed. This situation is realized in p-n structures on<br />
wide-band semiconductors at low enough biases [1,<br />
2, 5, 6]. With an increased bias voltage, electrons and<br />
holes are injected into the channel, and a regime of<br />
double injection is realized which results in a superlinear<br />
rise of the current [6]. The double injection leads,<br />
at high enough injection current, to destruction of the<br />
channel.<br />
In the case of silicon p-n junctions, the destruction<br />
of the channel by injected charge currents occurs<br />
at relatively low forward bias voltages of ~0.1V, and the<br />
linear section of I–V curve practically is not realized<br />
[6]. Our experiments confirm this conclusion. Curve<br />
4 in Fig. 1 has a large section that corresponds to formula<br />
(1) with ideality coefficient n=2.6. Such value<br />
of n suggests that the excess current, due to ammonia<br />
molecules adsorption, can be ascribed to the phononassisted<br />
tunnel recombination at deep surface states<br />
[9].<br />
I–V characteristics 2 and 3 in Fig. 1, measured<br />
in water- and ethylene vapors, respectively, have pronounced<br />
linear sections in a semi-logarithmic plot,<br />
with ideality coefficients of 1.13 and 2.1, which argues<br />
that the increase of the forward current in these vapors<br />
is due to enhanced surface recombination. Thus, the<br />
mechanism of the sensitivity of silicon p-n junctions<br />
to ammonia-, ethylene- and ammonia vapors is the<br />
same.<br />
I–V characteristics of the excess reverse current,<br />
due to water- and ethylene molecules adsorption,<br />
which are plotted as curves 1and 2 in Fig. 3, have large<br />
linear sections. This means that the surface conductive<br />
channel, formed as a result of water- and ethylene<br />
molecules adsorption, is not destroyed at a reverse<br />
bias. And this channel is responsible for gas sensitivity<br />
of silicon p-n structures at reverse biases.<br />
Curve 3 in Fig. 3, as I–V characteristic of the excess<br />
reverse current in ammonia vapors, is superlinear.<br />
This can be tentatively ascribed to injection processes<br />
or (and) strong-field effects.<br />
An interesting result of our study is that the sensitivity<br />
of silicon p-n structures as gas sensors is higher<br />
at reverse bias than at forward bias. Gas sensitivity of<br />
the forward current was observed only at low bias voltages<br />
V
studied vapors are of the same order of magnitude,<br />
while the sensitivities differ by three orders. For the<br />
electronic mechanism is also characteristic inequality<br />
τ > τ , which was observed for all studied gases at for-<br />
r d<br />
ward and reverse biases, as seen in Tab. 3. Tentatively,<br />
the response time of the studied p-n gas sensors is due<br />
to the recharging of surface states, as a result of the adsorption<br />
of molecules from the ambient atmosphere.<br />
4. CONCLUSIONS<br />
Forward and reverse currents in silicon p-n junctions<br />
are sensitive to ammonia, ethylene and water vapors<br />
in the ambient air. The sensitivity to NH 3 vapor<br />
is by orders of magnitude higher than to other studied<br />
vapors. Therefore silicon p-n junctions can be used<br />
as selective ammonia vapor sensors. Selectivity of the<br />
sensor is due to donor properties of NH 3 molecules.<br />
The sensitivity of silicon p-n structures to the mentioned<br />
vapors at forward biases is caused by enhancing<br />
of surface recombination, as a result of band bending<br />
in p-region, due to electric field of adsorbed ions.<br />
The gas sensitivity of studied p-n structures at reverse<br />
biases is due to forming of a surface conductive<br />
channel which shorts the p-n junction.<br />
The forward bias voltage of the sensor is limited by<br />
exponential rise of bulk injection current and screening<br />
of the electric field, induced by adsorbed ions, by<br />
injected electrons and holes. Therefore the reverse<br />
bias is preferable for the sensors and provides higher<br />
gas sensitivity, than the forward bias.<br />
The response time of silicon p-n sensors is below<br />
UDC 621.315.592<br />
O. O. Ptashchenko, F. O. Ptashchenko, O. V. Yemets<br />
60 s at room temperature for all the studied vapors at<br />
forward and reverse biases. This time can be ascribed<br />
to recharging of slow surface centers.<br />
References<br />
1. Ïòàùåíêî Î. Î., Àðòåìåíêî Î. Ñ., Ïòàùåíêî Ô. Î.<br />
Âïëèâ ãàçîâîãî ñåðåäîâèùà íà ïîâåðõíåâèé ñòðóì â pn<br />
ãåòåðîñòðóêòóðàõ íà îñíîâ³ GaAs–AlGaAs // Ô³çèêà ³<br />
õ³ì³ÿ òâåðäîãî ò³ëà . — 2001. — Ò. 2, ¹ 3. — Ñ. 481 — 485.<br />
2. Ïòàùåíêî Î. Î., Àðòåìåíêî Î. Ñ., Ïòàùåíêî Ô. Î.<br />
Âïëèâ ïàð³â àì³àêó íà ïîâåðõíåâèé ñòðóì â p-n ïåððåõîäàõ<br />
íà îñíîâ³ íàï³âïðîâ³äíèê³â À3Â5 // Æóðíàë ô³çè÷íèõ<br />
äîñë³äæåíü. — 2003. — Ò. 7, ¹4. — Ñ. 419 — 425.<br />
3. Áóãàéîâà M. E., Koâaëü Â. M., Ëàçàðåíêî B. ². òà ³í. Ãàçîâ³<br />
ñåíñîðè íà îñíîâ³ îêñèäó öèíêó (îãëÿä) // Ñåíñîðíà<br />
åëåêòðîí³êà ³ ì³êðîñèñòåìí³ òåõíîëî㳿. — 2005. — ¹3. —<br />
Ñ. 34 — 42.<br />
4. Áàëþáà Â. È., Ãðèöûê Â. Þ., Äàâûäîâà Ò. À. è äð. Ñåíñîðû<br />
àììèàêà íà îñíîâå äèîäîâ Pd-n-Si // Ôèçèêà è õèìèÿ<br />
ïîëóïðîâîäíèêîâ. — 2005. — Ò. 39, ¹ 2. Ñ. 285 — 288.<br />
5. Ptashchenko O. O., Artemenko O. S., Dmytruk M. L. et al.<br />
Effect of ammonia vapors on the surface morphology and<br />
surface current in p-n junctions on GaP // Photoelectronics.<br />
— 2005. — No. 14. — P. 97 — 100.<br />
6. Ptashchenko F. O. Effect of ammonia vapors on surface currents<br />
in InGaN p-n junctions // Photoelectronics. — 2007. —<br />
No. 17. — P. 113 — 116.<br />
7. Ïòàùåíêî Ô. Î. Âïëèâ ïàð³â àì³àêó íà ïîâåðõõíåâèé<br />
ñòðóì ó êðåìí³ºâèõ p-n ïåðåõîäàõ // ³ñíèê ÎÍÓ, ñåð.<br />
Ô³çèêà. — 2006. — Ò. 11, ¹. 7. — Ñ. 116 — 119.<br />
8. Ptashchenko O. O., Ptashchenko F. O., Yemets O. V. Effect<br />
of ammonia vapors on the surface current in silicon p-n junctions<br />
// Photoelectronics. — 2006. — No. 16. — P. 89 — 93.<br />
9. Ðtashchenko A. A., Ptashchenko F. A. Tunnel surface recombination<br />
in p-n junctions // Photoelectronics. — 2000. —<br />
¹ 10. — P. 69 — 71.<br />
10. Âàøïàíîâ Þ. À., Ñìûíòûíà Â. À. Àäñîðáöèîííàÿ ÷óâñòâèòåëüíîñòü<br />
ïîëóïðîâîäíèêîâ. — Îäåññà: Àñòðîïðèíò,<br />
2005. — 216 p.<br />
EFFECT OF AMBIENT ATMOSPHERE ON THE SURFACE CURRENT IN SILICON P-N JUNCTIONS<br />
Abstract<br />
The influence of ammonia, water and ethylene vapors on I-V characteristics of forward and reverse currents, as well as on the kinetics<br />
of the surface current in silicon p-n structures was studied. All these vapors enhance both the forward and reverse currents. The gas<br />
sensitivity of p-n structures at forward biases is due to enhanced surface recombination, while at reverse biases a surface conductive channel<br />
shorts the p-n junction. The sensitivity to ammonia is much higher than to other vapors. It is explained as a result of donor properties<br />
of NH 3 molecules. The response time of silicon p-n junctions as gas sensors at room temperature is below 60s.<br />
Key words: ambient atmosphere, surface current, silicon.<br />
ÓÄÊ 621.315.592<br />
À. À. Ïòàùåíêî, Ô. À. Ïòàùåíêî, Å. Â. Åìåö<br />
ÂËÈßÍÈÅ ÎÊÐÓÆÀÞÙÅÉ ÀÒÌÎÑÔÅÐÛ ÍÀ ÏÎÂÅÐÕÍÎÑÒÍÛÉ ÒÎÊ Â ÊÐÅÌÍÈÅÂÛÕ P-N ÏÅÐÅÕÎÄÀÕ<br />
Ðåçþìå<br />
Èññëåäîâàíî âëèÿíèå ïàðîâ àììèàêà, âîäû è ýòèëåíà íà ÂÀÕ ïðÿìîãî è îáðàòíîãî òîêîâ, à òàêæå íà êèíåòèêó ïîâåðõíîñòíîãî<br />
òîêà â êðåìíèåâûõ p-n ñòðóêòóðàõ. Âñå óêàçàííûå ïàðû ïîâûøàþò è ïðÿìîé, è îáðàòíûé òîêè. Ãàçîâàÿ ÷óâñòâèòåëüíîñòü<br />
p-n ñòðóêòóð ïðè ïðÿìîì ñìåùåíèè îáóñëîâëåíà ðîñòîì èíòåíñèâíîñòè ïîâåðõíîñòíîé ðåêîìáèíàöèè, à ïðè<br />
îáðàòíîì ñìåùåíèè ïðîâîäÿùèé êàíàë çàêîðà÷èâàåò p-n ïåðåõîä. ×óâñòâèòåëüíîñòü ê àììèàêó çíà÷èòåëüíî âûøå, ÷åì ê<br />
äðóãèì ïàðàì. Ýòî îáúÿñíÿåòñÿ äîíîðíûìè ñâîéñòâàìè ìîëåêóë NH 3 . Âðåìÿ ñðàáàòûâàíèÿ êðåìíèåâûõ p-n ïåðåõîäîâ êàê<br />
ãàçîâûõ ñåíñîðîâ ïðè êîìíàòíîé òåìïåðàòóðå íå ïðåâûøàåò 60 ñ.<br />
Êëþ÷åâûå ñëîâà: ïîâåðõíîñòíûé òîê, îêðóæàþùàÿ àòìîñôåðà, êðåìíèé.<br />
31
32<br />
ÓÄÊ 621.315.592<br />
Î. Î. Ïòàùåíêî, Ô. Î. Ïòàùåíêî, Î. Â. ªìåöü<br />
ÂÏËÈ ÍÀÂÊÎËÈØÍÜί ÀÒÌÎÑÔÅÐÈ ÍÀ ÏÎÂÅÐÕÍÅÂÈÉ ÑÒÐÓÌ Ó ÊÐÅÌͲªÂÈÕ P-N ÏÅÐÅÕÎÄÀÕ<br />
Ðåçþìå<br />
Äîñë³äæåíî âïëèâ ïàð³â àì³àêó, âîäè ³ åòèëåíó íà ÂÀÕ ïðÿìîãî ³ çâîðîòíîãî ñòðóì³â, à òàêîæ íà ê³íåòèêó ïîâåðõíåâîãî<br />
ñòðóìó â êðåìí³ºâèõ p-n ñòðóêòóðàõ. Âñ³ óêàçàí³ ïàðè ï³äâèùóþòü ³ ïðÿìèé, ³ çâîðîòíèé ñòðóìè. Ãàçîâà ÷óòëèâ³ñòü p-n<br />
ñòðóêòóð ïðè ïðÿìîìó çì³ùåíí³ îáóìîâëåíà çðîñòàííÿì ³íòåíñèâíîñò³ ïîâåðõíåâî¿ ðåêîìá³íàö³¿, à ïðè çâîðîòíîìó çì³ùåíí³<br />
ïðîâ³äíèé êàíàë çàêîðî÷óº p-n ïåðåõ³ä. ×óòëèâ³ñòü äî àì³àêó çíà÷íî âèùà, í³æ äî ³íøèõ ïàð³â. Öå ïîÿñíþºòüñÿ äîíîðíèìè<br />
âëàñòèâîñòÿìè ìîëåêóë NH 3 . ×àñ ñïðàöþâàííÿ êðåìí³ºâèõ p-n ïåðåõîä³â ÿê ãàçîâèõ ñåíñîð³â ïðè ê³ìíàòí³é òåìïåðàòóð³ íå<br />
ïåðåâèùóº 60 ñ.<br />
Êëþ÷îâ³ ñëîâà: ïîâåðõíåâèé ñòðóì, íàâêîëèøíÿ àòìîñôåðà, êðåìí³é.
UDC 621.315.592<br />
V. A. BORSCHAK, M. I. KUTALOVA, N. P. ZATOVSKAYA, A. P. BALABAN, V. A. SMYNTYNA<br />
Odessa I. I. Mechnikov National University, RL-3,<br />
2, Dvorianskaya str., Odessa 65082, Ukraine. Tel. +38 048 — 7266356.<br />
Fax +38 048 — 7266356, E-mail: wadz@mail.ru<br />
DEPENDENCE OF SPACE-CHARGE REGION CONDUCTIVITY<br />
OF NONIDEAL HETEROJUNCTION FROM PHOTOEXCITATION<br />
CONDITIONS<br />
It is shown that nonideal CdS-Cu 2 S heterojunction illumination results in essential space-charge<br />
region width reduction and change of a potential barrier form. It is established that this change is the<br />
most expressed near the heteroboundary and occurs even at very small intensity of stimulating light.<br />
It is connected with the capture of nonequilibrium charge on local centers, presented in space-charge<br />
region. Such change of the form of a potential barrier results in essential change of tunnel hopping<br />
conductivity of a spatial charge nonideal heterojunction.<br />
Investigation of the current transport mechanism<br />
in heterojunction, used as optical and x-ray images<br />
sensors inevitably takes in the account the light influence<br />
on their tunnel-jumping conductivity. Impact of<br />
light on nonideal heterojunction essentially influences<br />
the parameters of its space charge region (SCR) [1],<br />
and hence on tunnel-jumping conductivity of SCR,<br />
and heterojunction as a whole. As the sensor generated<br />
signal strongly depends on heterojunction conductivity,<br />
the question about light influence on SCR and so<br />
on conductivity is represented as rather significant.<br />
Experimental investigations of light influence on<br />
conductivity CdS-Cu 2 S heterojunction are described<br />
in [2]. It is established, that with the increase of excitation<br />
intensity with white or only short-wave (λ ≈<br />
520nm) light heterojunction conductivity is essentially<br />
increased both on direct and alternating current even<br />
in a short-circuited condition, i.e. at constant barrier<br />
height. At the same time, photocapacity growth is observed<br />
even at light illumination essentially smaller than<br />
the solar. Ratio Ñ ph /C d for some elements achieved 10<br />
and more units that testifies the barrier width reduction.<br />
Such phenomenon can result in essential growth<br />
tunnel-jumping conductivity in SCR .<br />
In [3] the current transport in heterojunction<br />
without illumination was considered, but it was not<br />
taken into account the SCR parameters change under<br />
light influence. Therefore, the offered model cannot<br />
be applied directly to the sensor work description. We<br />
shall consider how it is possible quantitatively to take<br />
into account the influence of light on jumping current<br />
transport in nonideal heterojunction.<br />
For definition SCR heterojunction conductivity it<br />
is necessary to set the function Fermi E F (x) level position<br />
in each point x [3]. Conductivity G σ (x) of SCR<br />
part is calculated from 0 up to x as the solution of the<br />
integral equation. However, the solution of this equation<br />
is also determined by the form of a potential barrier<br />
φ (x). For dark heterojunction φ (x) depends only<br />
on submitted bias U and shows the known square-law<br />
formula. At heterojunction illumination generated<br />
in wide band CdS nonbasic carriers (holes) are captured<br />
in SCR on the traps, presented there. We shall<br />
assume, that holes are captured by the centers with a<br />
© V. A. Borschak, M. I. Kutalova, N. P. Zatovskaya, A. P. Balaban, V. A. Smyntyna, 2009<br />
single energy level, which concentration is equal N t .<br />
Then, apparently from fig. 1, because of band bending<br />
in SCR, the energy distance from level Fermi up<br />
to a level of holes traps E Ft (determining their filling<br />
degree) essentially depends on coordinate x. It results<br />
in non-uniform filling, and concentration of the captured<br />
charge (changeable along an x axis) will be determined<br />
by expression:<br />
⎡ϕ0⎛ω−x⎞⎤ px ( ) =Δp0exp⎢ ⎜ ⎟<br />
kT ω<br />
⎥<br />
⎣ ⎝ ⎠⎦<br />
(1)<br />
here Δp — the concentration of photoexcited<br />
0<br />
holes in CdS quasineutral region. We shall note, that<br />
in formula (1) capture holes centers concentration is<br />
absent, because limits of considered here models the<br />
capture centers number essentially exceed the number<br />
of nonequilibrium holes, i.e. N > > p (x) or E > > kT.<br />
t Ft<br />
Thus, the dependence of a potential barrier φ (õ) can<br />
be defined from Poisson equation:<br />
2<br />
d ϕ(<br />
x) e<br />
=− N 2<br />
D + p x<br />
dx εε0<br />
[ ( )]<br />
(2)<br />
where N D — concentration of ionized sallow donors,<br />
not compensated completely in CdS, determining<br />
barrier width in darkness, when ð (õ) ≡ 0. Captured<br />
charge dependence from coordinate õ results in a significant<br />
deviation of a potential shape of a barrier φ (x)<br />
from the square-law form, characteristic for constant<br />
distribution of the charge creating a built-in field. The<br />
potential barrier of illuminated heterojunction will be<br />
described by expression:<br />
ϕ − ⎡ ⎛ Δ ⎞<br />
ϕ = + + − +<br />
( x)<br />
0 eU<br />
1<br />
Δp ⎢<br />
α<br />
1+ ( e −1−α) ⎢⎣ αN<br />
D<br />
2<br />
x<br />
2<br />
ω<br />
2 p<br />
⎜<br />
⎝αND x<br />
2⎟<br />
⎠ω<br />
+<br />
x<br />
2Δp<br />
⎛ −α<br />
α ⎞⎤<br />
ω ee 1 F<br />
2 ⎜ − −α ⎟⎥+Δ<br />
0<br />
ND<br />
(3)<br />
α ⎝ ⎠⎦⎥<br />
where α = (φ −ΕU)/kT. The parameter Δp, ,ncluded in<br />
0<br />
(3), can be determined knowing the character of distribution<br />
of the captured nonequilibrium charge set by<br />
the formula (3), and average value of a nonequilibrium<br />
33
charge p` , captured on traps in SCR region, which can<br />
t<br />
be determined from photocapacity value (C ). Having<br />
ph<br />
measured heterojunction dark capacity and its photocapacity<br />
at various stimulating light intensities under<br />
the formula of the flat condenser<br />
C=εε S/ω (4)<br />
0<br />
it is possible easily to determine appropriate to each<br />
of these capacities barrier region width ω (Ñ , C ), so<br />
d ph<br />
also values N (C ) and p (C , C ). It is obvious, that<br />
D T d d ph<br />
the size of average captured in SCR nonequilibrium<br />
charge is connected with ð (õ) by the ratio:<br />
ω<br />
1<br />
p = p( x) dx<br />
ω ∫ . (5)<br />
34<br />
t<br />
0<br />
Fig. 1. The zone diagram CdS-Cu S heterojunction with hole<br />
2<br />
traps in CdS<br />
It is possible to determine the appropriate value of<br />
parameter Δp, included in (3), for each value of stimulating<br />
light intensity by means of calculating p (C , C )<br />
t d ph<br />
from (5) with the account (1) for each value of photocapacity<br />
Ñ . ph<br />
UDC 621.315.592<br />
V. A. Borschak, M. I. Kutalova, N. P. Zatovskaya, A. P. Balaban, V. A. Smyntyna<br />
From the equation (3) it is seen, that at absence<br />
of photoexcitation of wide band material, i.e. at Δð =<br />
0, the expression (3) transforms into the square-law<br />
form. However, already at small values Δð there is an<br />
essential deviation φ (õ) from a square-law dependence<br />
especially near heterojunction where the captured<br />
nonequilibrium charge is maximal and φ (õ) gets<br />
the character close to exponential. At values Δð appropriate<br />
to large capacities, change φ (õ) form also is<br />
the most essential in frontier area where E F (x) has the<br />
maximal value and which, hence, has minimal tunnel-jumping<br />
conductivity. Thus, function φ (õ), and<br />
also E F (x) essentially depend on intensity of illumination<br />
and consequently from photocapacity which is<br />
easily measured experimentally. As E F (x) determines<br />
the values of parameters R’(E F ), N(E F ), W(E F ), which<br />
determines tunnel-jumping conductivity mechanism<br />
in SCR, the SCR conductivity G σ essentially depends<br />
on a type of function φ (õ) and the stimulating light intensity.<br />
It means, that illumination influences not only<br />
barrier width reduction on current-transport, but also<br />
the change of its form.<br />
References<br />
1. D.V.Vassilevski, V.A.Borschak, M.S.Vinogradov. “Influence<br />
of tunnel effects on the kinetics of the photocapacitance in<br />
nonideal heterojunctins” Solid-State Electronics, 1994,<br />
Vol.37, No.9, p.1680-1682.<br />
2. Smyntyna, V.A. Borschak, M.I. Kutalova, N.P. Zatovskaya,<br />
A.P. Balaban. Investigation in temperature and frequency<br />
dependences for conductivity in barrier region of nonideal<br />
heterojunction. — Photoelectronics. — 2005. — ¹14. —<br />
P. 5-7.<br />
3. Smyntyna V. A., Borschak V. A., Kutalova M. I., Zatovskaya<br />
N. P., Balaban A. P. External bias influence on the transmission<br />
processes in nonideal heterojunction // Photoelectronics.<br />
— 2008. — ¹17. — P. 23-26.<br />
4. Â.À.Áîðùàê, Ä.Ë.Âàñèëåâñêèé Òîêîïåðåíîñ ïî ëîêàëèçîâàííûì<br />
ñîñòîÿíèÿì â íåèäåàëüíûõ ãåòåðîñòðóêòóðàõ.<br />
Çàâèñèìîñòü ïðîâîäèìîñòè îáëàñòè ïðîñòðàíñòâåííîãî<br />
çàðÿäà íåèäåàëüíîãî ãåòåðîïåðåõîäà îò óñëîâèé ôîòîâîçáóæäåíèÿ.<br />
Ïîëóïðîâîäíèêîâàÿ òåõíèêà. — 1999–<br />
Âûï. 17. — ñ. 24-29.<br />
DEPENDENCE OF SPACE-CHARGE REGION CONDUCTIVITY OF NONIDEAL HETEROJUNCTION FROM<br />
PHOTOEXCITATION CONDITIONS<br />
Abstract<br />
It is shown that nonideal CdS-Cu 2 S heterojunction illumination results in essential space-charge region width reduction and<br />
change of a potential barrier form. It is established that this change is the most expressed near the heteroboundary and occurs even at<br />
very small intensity of stimulating light. It is connected with the capture of nonequilibrium charge on local centers, presented in spacecharge<br />
region. Such change of the form of a potential barrier results in essential change of tunnel hopping conductivity of a spatial charge<br />
nonideal heterojunction.<br />
Key word: space — nonideal heterojunction, photoexcitation conditions.
ÓÄÊ 621.315.592<br />
Â. À. Áîðùàê, Ì. È. Êóòàëîâà, Í. Ï. Çàòîâñêàÿ, À. Ï. Áàëàáàí, Â. À. Ñìûíòûíà<br />
ÇÀÂÈÑÈÌÎÑÒÜ ÏÐÎÂÎÄÈÌÎÑÒÈ ÎÁËÀÑÒÈ ÏÐÎÑÒÐÀÍÑÒÂÅÍÍÎÃÎ ÇÀÐßÄÀ ÍÅÈÄÅÀËÜÍÎÃÎ<br />
ÃÅÒÅÐÎÏÅÐÅÕÎÄÀ ÎÒ ÓÑËÎÂÈÉ ÔÎÒÎÂÎÇÁÓÆÄÅÍÈß<br />
Ðåôåðàò<br />
Ïîêàçàíî, ÷òî îñâåùåíèå íåèäåàëüíîãî ãåòåðîïåðåõîäà CdS-Cu 2 S ïðèâîäèò ê ñóùåñòâåííîìó ñîêðàùåíèþ øèðèíû<br />
îáëàñòè ïðîñòðàíñòâåííîãî çàðÿäà è èçìåíåíèþ ôîðìû ïîòåíöèàëüíîãî áàðüåðà. Óñòàíîâëåíî, ÷òî âáëèçè ãåòåðîãðàíèöû<br />
ýòî èçìåíåíèå ìàêñèìàëüíî âûðàæåíî è ïðîèñõîäèò äàæå ïðè î÷åíü íåáîëüøèõ èíòåíñèâíîñòÿõ âîçáóæäàþùåãî ñâåòà. Ýòî<br />
ñâÿçàíî ñ çàõâàòîì íåðàâíîâåñíîãî çàðÿäà íà ïðèñóòñòâóþùèå â îáëàñòè ïðîñòðàíñòâåííîãî çàðÿäà ëîêàëüíûå öåíòðû. Òàêîå<br />
èçìåíåíèå ôîðìû ïîòåíöèàëüíîãî áàðüåðà ïðèâîäèò ê ñóùåñòâåííîìó èçìåíåíèþ òóííåëüíî-ïðûæêîâîé ïðîâîäèìîñòè îáëàñòè<br />
ïðîñòðàíñòâåííîãî çàðÿäà íåèäåàëüíîãî ãåòåðîïåðåõîäà.<br />
Êëþ÷åâûå ñëîâà: ïðîñòðàíñòâåííûé çàðÿä, íåèäåàëüíûé ãåòåðîïåðåõîä, óñëîâèÿ ôîòîâîçáóæäåíèÿ.<br />
ÓÄÊ 621.315.592<br />
Â. À. Áîðùàê, Ì. ². Êóòàëîâà, Í. Ï. Çàòîâñüêà, À. Ï. Áàëàáàí, Â. À. Ñìèíòèíà<br />
ÇÀËÅÆͲÑÒÜ ÏÐβÄÍÎÑÒ² ÎÁËÀÑÒ² ÏÐÎÑÒÎÐÎÂÎÃÎ ÇÀÐßÄÓ ÍŲÄÅÀËÜÍÎÃÎ ÃÅÒÅÐÎÏÅÐÅÕÎÄÓ<br />
Â²Ä ÓÌΠÔÎÒÎÇÁÓÄÆÅÍÍß<br />
Ðåçþìå<br />
Äîâåäåíî, ùî çàñâ³òëåííÿ íå³äåàëüíîãî ãåòåðîïåðåõîäó CdS-Cu 2 S âèêëèêຠâàæëèâå ñêîðî÷åííÿ øèðèíè îáëàñò³ ïðîñòîðîâîãî<br />
çàðÿäà òà çì³íè ôîðìè ïîòåíö³éíîãî áàð’ºðó.<br />
Âñòàíîâëåíî, ùî ïîáëèçó ãåòåðîìåæ³ öÿ çì³íà — ìàêñèìàëüíà ³ â³äáóâàºòüñÿ ïðè äóæå íåâåëèêèõ ³íòåíñèâíîñòÿõ çáóäæóþ÷îãî<br />
ñâ³òëà. Öå ïîâ’ÿçàíî ç çàõîïëåííÿì íåð³âíîâàæíîãî çàðÿäó íà ïðèñóòí³ ëîêàëüí³ öåíòðè. Òàê³ çì³íè ôîðìè ïîòåíö³éíîãî<br />
áàð’ºðó âèêëèêàþòü ñóòòºâ³ çì³íè òóíåëüíî –ñòðèáêîâî¿ ïðîâ³äíîñò³ îáëàñò³ ïðîñòîðîâîãî çàðÿäó íå³äåàëüíîãî<br />
ãåòåðîïåðåõîäó.<br />
Êëþ÷îâ³ ñëîâà: ïðîñòîðîâèé çàðÿä, íå³äàëüíèé ãåòðîïåðåõ³ä, óìîâè ôîòîçáóäæåííÿ.<br />
35
36<br />
UDÑ 539.142, 539.184<br />
V. KH. KORBAN, G. P. PREPELITSA, YU. BUNYAKOVA, L. DEGTYAREVA, A. KARPENKO,<br />
S. SEREDENKO<br />
Odessa National Polytechnical University, Odessa<br />
Odessa State Environmental University, Odessa<br />
PHOTOKINETICS OF THE IR LASER RADIATION EFFECT ON MIXTURE<br />
OF THE CO 2 -N 2 -H 2 O GASES: ADVANCED ATMOSPHERIC MODEL<br />
A kinetics of energy exchange in the mixture of the atmosphere CO 2 -N 2 -H 2 O gases under passing<br />
the powerful CO 2 laser radiation pulses within the three-mode model of kinetical processes is studied.<br />
More accurate data for the absorption coefficient are presented.<br />
1. INTRODUCTION<br />
At present time the environmental physics has<br />
a great progress, provided by implementation of the<br />
modern quantum electronics and laser physics methods<br />
and technologies in order to study unusual features<br />
of the “laser radiation- substance (gases, solids<br />
etc.) interaction. A special interest attracts a problem<br />
of interaction of the powerful laser radiation with an<br />
aerosol ensemble and search of new non-linear optical<br />
effects. The latter is directly related with problems of<br />
modern aerosol laser physics (c.f.[1-13]). One could<br />
remind that there is a redistribution of molecules on<br />
the energy levels of internal degree of freedom in the<br />
resonant absorption of IR laser radiation by the atmospheric<br />
molecular gases. As a result of quite complicated<br />
processes one could define an essential changing<br />
of the gases absorption coefficient due to the saturation<br />
of absorption [1].<br />
One interesting effect else to be mentioned is an<br />
effect of the kinetic cooling of environment (mixture<br />
of gases), as it was at first predicted in ref. [2,5]. Usually<br />
the effect of kinetical cooling (CO 2 ) in a process of<br />
absorption of the laser pulse energy by molecular gas is<br />
considered for the middle latitude atmosphere and for<br />
special form of a laser pulse. Besides, the approximate<br />
values for constants of collisional deactivation and<br />
resonant transfer in reaction CO 2 -N 2 are usually used.<br />
In series of papers (see, for example, [11-13], computational<br />
modelling of the energy and heat exchange<br />
kinetics in the mixture of the CO 2 -N 2 -H 2 0 atmospheric<br />
gases interacting with IR laser radiation has been<br />
carried out within the general three-mode kinetical<br />
model. It is obvious that using more precise values for<br />
all model constants and generally speaking the more<br />
advanced atmospheric model parameters may lead to<br />
quantitative changing in the temporary dependence of<br />
the resonant absorption coefficient by CO 2 .<br />
Let us remind that the creation and accumulation<br />
of the excited molecules of nitrogen owing to the resonant<br />
transfer of excitation from the molecules CO 2<br />
results in the change of environment polarizability.<br />
Perturbing the complex conductivity of environment,<br />
all these effects are able to transform significantly the<br />
impulse energetics of IR lasers in an atmosphere and<br />
significantly change realization of different non-linear<br />
laser-aerosol effects.<br />
The aim of this paper is to present more accurate<br />
data for kinetics of energy and heat exchange in the<br />
mixture CO 2 -N 2 -H 2 0 gases in atmosphere under passing<br />
the powerful CO 2 laser radiation pulses on the basis<br />
of using the more advanced atmospheric model and<br />
more precise values for all kinetical model constants.<br />
2. ADVANCED THREE-MODE KINETICAL<br />
MODEL FOR THE “LASER PULSE —<br />
MEDIUM” INTERACTION<br />
As usually, we start from the modified three-mode<br />
model of kinetic processes (see, for example, [1,11-<br />
13] in order to take into consideration the energy exchange<br />
and relaxation processes in the ÑÎ 2 . — N 2 —<br />
H 2 O mixture interacting with a laser radiation. As in<br />
ref. [11-13] we consider a kinetics of three levels: 10°0,<br />
00°1 (ÑÎ 2 ) and v = 1 (N 2 ). Availability of atmospheric<br />
constituents O 2 and H 2 O is allowed for the definition<br />
of the rate of vibrating-transitional relaxation of N 2 .<br />
The system of balance equations for relative populations<br />
is written in a standard form as follows:<br />
dx1<br />
0<br />
=−βω+ ( 2 gP10 ) x1 +βω x2 + 2 β gP10 ) x1<br />
,<br />
dt<br />
dx2<br />
0<br />
=ωx1−( ω+ Q+ P20) x2 + Qx3 + P20x2 , (1)<br />
dt<br />
dx 3<br />
0<br />
=δQx2 −( δ Q + P30) x3 + P30x3 .<br />
dt<br />
Here the following notations are used:<br />
x = N / N 1 100 CO , 2<br />
x = N / N 2 001 CO , (2)<br />
2<br />
x = N 3 N / N<br />
2 CO2<br />
δ ,<br />
where N N are the level populations 10°0, 00°1<br />
100, 001<br />
(ÑÎ ); 2 2 N N is the level population v = 1 (N ); N 2 CO2<br />
is the concentration of CO molecules; Δ is the ratio<br />
2<br />
of the common concentrations of ÑÎ and N in the<br />
2 2<br />
atmosphere (Δ = 3.85⋅10-4 0 0<br />
); x 1 , x 2 and 0<br />
x 3 are the<br />
equilibrium relative values of populations under gas<br />
temperature T:<br />
© V. Kh. Korban, G. P. Prepelitsa, Yu. Bunyakova, L. Degtyareva, A. Karpenko, S. Seredenko, 2009
( )<br />
x = exp − E T , (3)<br />
0<br />
1 1<br />
( )<br />
x = x = exp E T ;<br />
0 0<br />
2 3 2<br />
The values E and E in (1) are the energies (K)<br />
1 2<br />
of levels 10°0, 00°1 (consider the energy of quantum<br />
N equal to E ); P , P and P are the probabilities<br />
2 2 10 20 30<br />
(s-1 ) of the collisional deactivation of levels 10°0, 00°1<br />
(ÑÎ ) and v = 1 (N ), Q is the probability (s 2 2 -1 ) of resonant<br />
transfer in the reaction ÑÎ → N ,ω is the prob-<br />
2 2<br />
ability (s-1 ) of ÑÎ light excitation, g = 3 is the statisti-<br />
2<br />
cal weight of level 02°0, β = (1+g) -1 = 1/4.<br />
As usually, the solution of the differential equations<br />
system (1) allows defining a coefficient of absorption<br />
of the radiation by the CO molecules according to the<br />
2<br />
formula:<br />
α =σ( x − x ) N . (4)<br />
CO2 1 2 CO2<br />
The σ in (4) is dependent upon the thermodynamical<br />
medium parameters as follows [2]:<br />
1<br />
2<br />
P ⎛ T ⎞<br />
σ=σ0 ⎜ ⎟<br />
P0 T0<br />
, (5)<br />
⎝ ⎠<br />
Here T and p are the air temperature and pressure,<br />
σ is the cross-section of resonant absorption under<br />
0<br />
T = T , p = p .<br />
0 0<br />
One could remind that the absorption coefficient<br />
for carbon dioxide and water vapour is dependent<br />
upon the thermodynamical parameters of aerosol atmosphere.<br />
In particular, for radiation of CO -laser the<br />
2<br />
coefficient of absorption by atmosphere defined as<br />
α =α +α<br />
g<br />
CO2 H2O is equal in conditions, which are typical for summer<br />
mid-latitudes, α g (H=0) = 2.4∙10 6 ñm -1 , from which<br />
0.8∙10 6 ñm -1 accounts for CO 2 and the rest — for water<br />
vapour (data are from ref. [2]) . On the large heights the<br />
sharp decrease of air moisture occurs and absorption<br />
coefficient is mainly defined by the carbon dioxide.<br />
The changing population of the low level 10°0<br />
(ÑÎ 2 ), population of the level 00°1, the vibratingtransitional<br />
relaxation (VT-relaxation) and the inter<br />
modal vibrating-vibrating relaxation (VV’-relaxation)<br />
processes define the physics of resonant absorption<br />
processes. Moreover, the above indicated processes<br />
result in a redistribution of the energy between the<br />
vibrating and transitional freedom of the molecules.<br />
According to ref.[1], the threshold value, which corresponds<br />
to the decrease of absorption coefficient in<br />
two times, for the strength of saturation of absorption<br />
in vibrating-rotary conversion give I sat = (2 ÷ 5) 10 5 W<br />
cm -2 for atmospheric CO 2 . In this case the pulse duration<br />
t i must satisfy the condition t R
the gas under powerful CO 2 laser radiation passing in atmosphere//<br />
J.Techn.Phys. — 1974. — Vol.14. — P.1063-1069.<br />
6. Ambrosov S.V., Prepelitsa G.P. Spectroscopy of Atmosphere<br />
gases atoms and molecules: Nonlinear Spectroscopic Effects//<br />
Proc.32 nd EPS Conf. EGAS. — Vilnius. — 2000. — P.30.<br />
7. Parkinson S., Young P. Uncertainty and sensitivity in global<br />
carbon cycle modeling // Climate Research. — 1998. —<br />
Vol. 9. — No. 3. — P. 157-174.<br />
8. Stephens B.B., Keeling R.F., Heimann M., Six K.D., Murnane<br />
R., Caldeira K. Testing global ocean carbon cycle models<br />
using measurements of atmospheric O 2 and CO 2 concentration<br />
// Global Biogeochemical Cycles. — 1998. — Vol. 12. —<br />
No. 2. — P. 213-230.<br />
9. Glushkov A.V., Ambrosov S.V., Malinovskaya S.V. l, Spectroscopy<br />
of carbon dioxide: Oscillator strengths and energies<br />
of transitions in spectra of ÑÎ 2 // Optics and Spectroscopy. —<br />
1999. — Ò.80,N1. — Ð.60-65.<br />
10. Glushkov A.V., Malinovskaya S.V.,Shpinareva I.M., Kozlovskaya<br />
V.P., Gura V.I., Quantum stochastic modelling<br />
energy transfer and effect of rotational and V-T relaxation<br />
on multiphoton excitation and dissociation for CF 3 Br molecules//<br />
Int. Journ.Quant.Chem. — 2005, — Vol.104,N(5). —<br />
P.512-520.<br />
11. Trenberth K.E., Stepaniak D.P., Caron J.M. Interannual<br />
variations in the atmospheric heat budget // J. Geophys.<br />
Res. — 2002. — Vol. 107. — P. 4-1 — 4-15.<br />
12. Turin A.V., Prepelitsa G.P., Kozlovskaya V.P., Kinetics of<br />
energy and heat exchange in mixture CO 2 -N 2 -H 2 0 of atmospheric<br />
gases interacting with IR laser radiation: Precise<br />
3-mode kinetical model// Phys. Aerodisp. Syst. — 2003. —<br />
N40. — P.123-128.<br />
38<br />
UDÑ 539.142, 539.184<br />
V. Kh. Korban, G. P. Prepelitsa, Yu. Bunyakova, L. Degtyareva, A. Karpenko, S. Seredenko<br />
13. Prepelitsa G.P., Shpinareva I.M., Bunyakova Yu.Ya., Photokinetics<br />
of interaction and energy exchange for ir laser radiation<br />
with mixture CO 2 -N 2 -H 2 0 of atmospheric gases//Photokinetics.<br />
— 2006. — Vol.16. — P.139-141.<br />
14. Wang, C, ENSO, climate variability, and the Walker and<br />
Hadley circulations. The Hadley Circulation: Present, Past,<br />
and Future (Eds H. F. Diaz and R. S. Bradley). Springer<br />
(2004).<br />
15. Boer G.J., Sargent N.E. Vertically integrated budgets of<br />
mass and energy for the globe // J. Atmos. Sci. — 1995. —<br />
Vol. 42. — P. 1592-1613.<br />
16. Kistler R., Kalnay E., Collins W., Saha S., White G., Woollen<br />
J., Chelliah M., Ebisuzaki W., Kanamitsu M., Kousky V.,<br />
van den Dool H., Jenne R., Fiorino M. The NCEP-NCAR<br />
50-year reanalysis: monthly means CD-ROM and documentation<br />
// Bull. Amer. Meteor. Soc. — 2001. — Vol. 82. —<br />
P. 247-267.<br />
17. Fyfe J.C., Boer G.J., Flato G.M. Predictable winter climate<br />
in the North Atlantic sector during the 1997–1999 ENSO<br />
cycle // Geophysical Research Letters. — 1999. — Vol. 26. —<br />
No. 21. — P. 1601-1604.<br />
18. Plattner G-K., Joos F., Stocker T.F., Marchal O. Feedback<br />
mechanisms and sensitivities of ocean carbon uptake under<br />
global warming // Tellus. — 2001. — Vol. 53B. — No. 5. —<br />
P. 564-592.<br />
19. Jin X., Shi G. A simulation of CO 2 uptake in a three dimensional<br />
ocean carbon cycle model // Acta Meteorologica Sinica.<br />
— 2001. — Vol. 15. — No. 1. — P. 29-39.<br />
20. Rivkin B.B., Legendre L. Biogenic carbon cycling in the<br />
upper ocean: effect of microbial respiration // Science. —<br />
2001. — Vol. 291. — P. 2398-2400.<br />
PHOTOKINETICS OF THE IR LASER RADIATION EFFECT ON MIXTURE OF THE CO 2 -N 2 -H 2 O GASES: ADVANCED<br />
ATMOSPHERIC MODEL<br />
Abstract<br />
A kinetics of energy exchange in the mixture of the atmosphere CO 2 -N 2 -H 2 O gases under passing the powerful CO 2 laser radiation<br />
pulses within the three-mode model of kinetical processes is studied. More accurate data for the absorption coefficient are presented.<br />
Key words: photokinetics, laser field, mixture of gases, atmospheric model.<br />
ÓÄÊ 539.142, 539.184<br />
Â. Õ. Êîðáàí, Ã. Ï.Ïðåïåëèöà, Þ. Áóíÿêîâà, Ë. Äåãòÿðåâà, À. Êàðïåíêî, Ñ. Ñåðåäåíêî<br />
ÔÎÒÎÊÈÍÅÒÈÊÀ ÂÇÀÈÌÎÄÅÉÑÒÂÈß ÈÊ ËÀÇÅÐÍÎÃÎ ÈÇËÓ×ÅÍÈß ÑÎ ÑÌÅÑÜÞ CO 2 -N 2 -H 2 O ÃÀÇÎÂ:<br />
ÓÒÎ×ÍÅÍÍÀß ÀÒÌÎÑÔÅÐÍÀß ÌÎÄÅËÜ<br />
Ðåçþìå<br />
Ðàññìîòðåíà ôîòîêèíåòèêà ýíåðãîîáìåíà â ñìåñè CO 2 -N 2 -H 2 0 àòìîñôåðíûõ ãàçîâ ïðè ïðîõîæäåíèè ÷åðåç àòìîñôåðó<br />
ìîùíîãî èçëó÷åíèÿ CO 2 ëàçåðà â ðàìêàõ óòî÷íåííîé 3-ìîäîâîé ìîäåëè êèíåòè÷åñêèõ ïðîöåññîâ. Ïîëó÷åíû áîëåå òî÷íûå<br />
çíà÷åíèÿ êîýôôèöèåíòà ïîãëîùåíèÿ.<br />
Êëþ÷åâûå ñëîâà: ôîòîêèíåòèêà, ëàçåðíîå ïîëå, ñìåñü ãàçîâ, àòìîñôåðíàÿ ìîäåëü.<br />
ÓÄÊ 539.142, 539.184<br />
Â. Õ. Êîðáàí, Ã. Ï.Ïðåïåëèöà, Þ. Áóíÿêîâà Ë. Äåãòÿðåâà, À. Êàðïåíêî, Ñ. Ñåðåäåíêî<br />
ÔÎÒÎʲÍÅÒÈÊÀ ÂÇÀªÌÎIJ¯ ²× ËÀÇÅÐÍÎÃÎ ÂÈÏÐÎ̲ÍÞÂÀÍÍß ²Ç ÑÓ̲ØÅÉ CO 2 -N 2 -H 2 O ÃÀDzÂ:<br />
ÓÄÎÑÊÎÍÀËÅÍÀ ÀÒÌÎÑÔÅÐÍÀ ÌÎÄÅËÜ<br />
Ðåçþìå<br />
Ðîçãëÿíóòî ôîòîê³íåòèêó åíåðãîîáì³íó ó ñóì³øó CO 2 -N 2 -H 2 O àòìîñôåðíèõ ãàç³â ïðè ïðîõîäæåíí³ ñêð³çü àòìîñôåðó ì³öíîãî<br />
âèïðîì³íþâàííÿ CO 2 ëàçåðà ó ìåæàõ óòî÷íåíî¿ 3-ìîäîâî¿ ìîäåë³ ê³íåòè÷íèõ ïðîöåñ³â. Îòðèìàí³ á³ëüø òî÷í³ îö³íêè äëÿ<br />
êîåô³ö³ºíòà ïîãëèíàííÿ.<br />
Êëþ÷îâ³ ñëîâà: ôîòîê³íåòèêà, ëàçåðíå ïîëå, ñóì³ø ãàç³â, àòìîñôåðíà ìîäåëü.
UDÑ 621.383:537.221<br />
D. À. ÊUDIY, N. P. ÊLOCHKO, G. S. KHRYPUNOV, N. À. ÊÎVTUN, K. Y. ÊRIKUN,<br />
Y. K. BELONOGOV 1<br />
National technical university “Kharkiv polytechnic institute”,<br />
Kharkiv, Frunze Street 21, 61002,<br />
tel. (057) 731-56-91, kudiy@ukr.net, klochko_np@mail.ru, khrip@ukr.net, orcsin@gmail.com<br />
1 Voronezh State technical university, Voronezh, Moscow Avenue 14.<br />
belonogov@phys.vorstu.ru<br />
ELABORATION OF CADMIUM SULPHIDE FILM LAYERS<br />
FOR ECONOMICAL SOLAR CELLS<br />
The structural and optical properties of CdS films received by the liquid-phase chemical deposition<br />
method are investigated. The analysis of surface film was conducted by scanning electron microscopy<br />
method. The structural parameters are determined by the X-Ray difractogram method while the<br />
definition of dispersion coherent areas and microdeformations were defined by analytical processing<br />
of X-Ray difractogram. The X-ray electron-probe microanalysis was fulfilled as well.<br />
The mathematical processing of CdS layers transmittion<br />
spectra was carried out. The structural and<br />
optical properties of the investigated CdS films are<br />
defined by the CdS layer thickness and its annealing<br />
mode.<br />
The film solar cells (SC) developed on the base of<br />
heterojunction CdTe/CdS are the most perspective<br />
for large-scale applications [1]. The cadmium telluride<br />
base layers provide the maximal theoretical photovoltaic<br />
conversion efficiency in ground conditions<br />
for single-junction semiconductive SC — more than<br />
29 % caused by optimal cadmium telluride band gap<br />
(1,46 eV) [2]. Cadmium sulphide is the most suitable<br />
hetero-partner for the formation of an effective separating<br />
barrier to cadmium telluride according to the<br />
theoretical diagram of the heterojunction p-CdTe/n-<br />
CdS giving the height of a potential barrier equal to<br />
1,02 eV [3]. The last fact allows to divide effectively the<br />
non-equilibrium charge carriers concentration generated<br />
under illumination and to increase the theoretical<br />
open-circuit voltage limit up to 1V.<br />
Moreover, the interface of CdTe and CdS have<br />
the minimal lattice mismatching as compared to other<br />
combinations of A II B VI semiconductors [4], what<br />
causes the possibility to form high-quality separating<br />
junction characterized by low density of saturation diode<br />
current and by high shunt resistance.<br />
The present-day effective thin film SC on the base<br />
of CdS/CdTe have been developed by high-temperature<br />
expensive vacuum manufacturing methods [5]<br />
what limits the opportunities of the cost reduction of<br />
such devices. At the same time, the application of thin<br />
semiconductor film interfacing layers, deposited by<br />
economical and high technology processes, allows the<br />
creation of competitive film solar cells in comparison<br />
with traditional electric power sources. It causes an<br />
actuality of development of economical low-temperature<br />
chemical technologies for the deposition of such<br />
heterojunctions’components.<br />
Traditionally, the film solar cells on the base of<br />
CdS/CdTe are created on glass substrates covered by<br />
sublayer of transparent conducting oxide, penetratable<br />
to solar radiation absorbed in the cadmium tel-<br />
© D. À. Êudiy, N. P. Êlochko, G. S. Khrypunov, N. À. Êîvtun, K. Y. Êrikun, Y. K. Belonogov, 2009<br />
luride base layer. The typical design of CdS/CdTe film<br />
SC is shown in Fig. 1.<br />
Fig. 1. FTO/CdS/CdTe device structure scheme.<br />
A maximal theoretical short current density of<br />
CdTe solar cell equals to the 30.8 mÀ/cm2 if solar radiation<br />
power is about 100 mW/cm2 [6].<br />
However, approximately 7.1 mÀ/cm2 of this short<br />
current density is an electricical part caused by generation<br />
of non-equilibrium charge carriers as a result<br />
of absorption of photons with energy exceeding the<br />
CdS band gap.<br />
Thus, the recombination of non-equilibrium<br />
charge carriers in the CdS layer could influence negatively<br />
on the CS photocurrent by means of essential<br />
reduce of the contribution of the charge carriers generated<br />
by photons in a spectral range from 350 up to<br />
520 nm to the SC photovoltage. So, the short-wave<br />
edge of the CdS/CdTe CS spectra is caused not only<br />
by an absorption edge of transparent conducting electrode<br />
prepared from FTO (fluoride doped tin oxide),<br />
but also by absorption edge of the CdS.<br />
The transparency of the cadmium sulphide layer<br />
in a 520-850 nm range also influences significantly the<br />
photocurrent, because it determines the flux density<br />
of the photons arriving to the cadmium telluride layer<br />
(note that the long wavelength edge of this spectral<br />
range is caused by absorption edge of CdTe).<br />
39
Thus, the decrease of cadmium sulphide layer<br />
thickness in the case of equal other parameters, has<br />
to enlarge the photocurrent. Therefore, traditionally<br />
the only requirement to cadmium sulphide film used<br />
in a solar cell design was the high transparency in the<br />
visible range.<br />
We consider, however, that the elaboration of cadmium<br />
sulphide deposition technology must be fulfilled<br />
by means of optimization of CdS layer when it becomes<br />
a member of a multilayer composition of the SC device<br />
structure. We have shown in [7], that the optimal thickness<br />
of CdS films deposited by vacuum evaporation is<br />
0.35 μm and the reduction of CdS thickness results in<br />
shunting of the heterojunction as a sequence of direct<br />
contact of the FTO frontal electrode with the CdTe<br />
base layer through the cadmium sulphide pinholes.<br />
So, the high transparency of cadmium sulphide<br />
must go along with the formation of non-porous CdS<br />
layer. As well as for the creation of sharp heterojunction<br />
during high-temperature manufacture of the device<br />
structure, it is important to minimize the diffusion interaction<br />
between cadmium sulphide and cadmium telluride,<br />
so traditionally before CdTe deposition, the CdS<br />
layers were annealed at 400 î Ñ during 25 minutes in air.<br />
Thus, it is necessary to provide the high transparency<br />
and an absence of the pinholes not only in as-prepared<br />
cadmium sulphide but in air-annealed one as well.<br />
In this work, we investigate the structure and optical<br />
properties of the cadmium sulphide films, prepared<br />
by chemical bath deposition. Plates of borosilicate<br />
glass Ê8, covered with FTO were used as substrates<br />
for CdS films. Cadmium sulphide chemical bath deposition<br />
was carried out in aqueous chloride solution,<br />
consisted of cadmium chloride CdÑl 2 0,011M, thiourea<br />
(NH 2 ) 2 CS 0,014M and ammonium hydroxide<br />
NH 4 OH 3,0M at 75 î Ñ. The solution was magnetically<br />
stirred in a vessel contained FTO covered glass substrate<br />
for 10 minutes, and then the films were washed<br />
out by distilled water and dried out on air.<br />
The investigations of the films by scanning electron<br />
microscopy method (SEM) on REM — 1M have<br />
shown, that in the case of approximately 0.1 μm thick<br />
as-deposited CdS films measured by profilometer Deflak,<br />
the pinholes have not appeared as a result of air<br />
annealing, but a compression of the film by means of<br />
diminution f surface relief was observed (Fig. 2 a, b).<br />
X-ray diffraction (XRD) patterns of the CdS/FTO<br />
heterostructures, before and after air annealing, were<br />
taken with a use of DRON — 4M diffractometer using<br />
CuK α radiation and have testified only FTO reflections<br />
for as-deposited films because of the amorphous<br />
nature of CdS. However, the air-annealed compositions<br />
glass/FTO/CdS demonstrated the appearance<br />
of (100) reflection of cadmium sulphide of hexagonal<br />
modification (Fig. 3 a, b).<br />
The results of the X-ray electron-probe microanalysis<br />
fulfilled by means of JSM-840 with the system<br />
of the power dispersion analysis LINK-860 testify<br />
the ratio of cadmium to sulfur atomic concentrations<br />
in as-deposited CdS films equal to 1.1. The annealing<br />
results in depletion of the volatile component sulfur,<br />
and, as a sequence, this ratio grown up to 1.2.<br />
The analysis of FTO/CdS transmission and reflection<br />
spectra recorded by means of spectrophotometer<br />
SF-46 (Fig. 4 a,b), has shown the direct allowed optical<br />
40<br />
transitions in cadmium sulphide films. Before the annealing<br />
an optical band gap was 2.27 eV. The air-annealing<br />
results in increase of the band gap up to 2.35 eV.<br />
Fig. 2. SEM pattern of chemical bath deposited CdS film on<br />
FTO/glass substrate as prepared (a) and after air annealing (b)<br />
Fig. 3. XRD patterns of glass/FTO/CdS before (à) and after<br />
air annealing (b)<br />
In the wave lengths ranges 400-550 nm and 550-<br />
850 nm, the increase of FTO/CdS transparency from<br />
Ò 400-550 = 47,14 % up to Ò 400-550 = 50,57 % and from<br />
Ò 550-850 = 75,29 % up to Ò 550-850 = 78,62 % were observed.
Fig. 4. The transmittance (a) and reflectance (b) of FTO/CdS<br />
before (1) and after the procedure.<br />
UDÑ 621.383:537.221<br />
CONCLUSIONS<br />
D. À. Êudiy, N. P. Êlochko, G. S. Khrypunov, N. À. Êîvtun, K. Y. Êrikun, Y. K. Belonogov<br />
1. Thus, the cadmium sulphide films were deposited<br />
by chemical bath deposition method. These<br />
films are considered suitable for the use in the design<br />
of highly effective solar cells on the base of cadmium<br />
telluride, because after the air annealing of FTO/CdS<br />
composition at 400 î Ñ for 25 minutes in accordance<br />
with technology of the device heterostructure formation<br />
this composition demonstrated a favourable combination<br />
of optical and structural parameters.<br />
2. After the air annealing the pinholes in cadmium<br />
sulphide layers were absent, the CdS films were polycrystalline.<br />
Owing to the air-annealing the transparency<br />
of FTO/CdS layers in the range of wave lengths 400-<br />
550 nm was Ò 400-550 = 50.57 %, and in a range of wave<br />
lengths 550-850 nm it achieved Ò 550-850 . = 78, 62 %.<br />
The work is supported by the grant of Russia ¹08-<br />
08-99071-r_îfi<br />
References<br />
1. Goetzberger A., Luther J., Willere G. Solar cells: past, present,<br />
future // Solar Energy Material & Solar Cells. — 2002. —<br />
Vol. 74, ¹1-4. — P.1-11.<br />
2. Hamakawa Y. Solar PV energy conversion and 21st century’s<br />
civilization // Solar Energy Materials & Solar Cells.—<br />
2002. — Vol. 74, ¹ 1- 4. — P. 13-23.<br />
3. Andersson B.A. Materials Availability for large-scale thinfilm<br />
photovoltaic // Progress of Photovoltaic: Research and<br />
Applications. — 2000. — Vol. 8, ¹1. — P. 61-76.<br />
4. Bube R. Properties of Semiconductors Materials. Photovoltaic<br />
Materials. — USA: Imperial College Press, 2000. — vol.1,<br />
¹6. — P. 69-72.<br />
5. 16.5% — Efficiency CdS/CdTe polycrystalline thin-film solar<br />
cells / Wu X., Keame J.C., Dhere R.G., De Hart C., Duda<br />
A., Gessert T.A., Asher S., Levi D.H., Sheldon P. // Proceeding<br />
17th European Photovoltaic Solar Energy Conference. —<br />
Munich (Germany). — 2001. — P. 995- 999.<br />
6. Device performance characterization and junction mechanisms<br />
in CdS/CdTe solar cells / Oman D.M., Dugan K.M.,<br />
Killian J.K., Cekala C.S., Ferikides C.S., Morel D.L. // Solar<br />
Energy Materials Solar Cells. — 2000. — Vol. 58, ¹3. —<br />
P. 361-373.<br />
7. Õðèïóíîâ Ã.Ñ., Áîéêî Á.Ò., Êîïà÷ Ã.²., Ìåð³óö À.Â.,<br />
Êóä³é Ä.À., Íîâèêîâ Â.Î. Îïòèì³çàö³ÿ ôîòîåëåêòðè÷íèõ<br />
ïðîöåñ³â ó ïë³âêîâèõ ñîíÿ÷íèõ åëåìåíòàõ íà îñíîâ³<br />
CdTe // Íàóêîâèé â³ñíèê ×åðí³âåöüêîãî óí³âåðñèòåòó. —<br />
2005. — Â.237. — Ô³çèêà. Åëåêòðîí³êà. — Ñ.80-85.<br />
ELABORATION OF CADMIUM SULPHIDE FILM LAYERS FOR ECONOMICAL SOLAR CELLS<br />
Abstract<br />
The structural and optical properties CdS films, which received by the liquid-phase chemical deposition method, are investigated.<br />
The analysis of surface film has conducted by scanning electron microscopy method. The structural parameters are determined by the<br />
X-Ray difractogram method, which the definition of dispersion cogerent areas and microdeformations were defined by analytical processing<br />
X-Ray difractogram. The X-ray electron-probe microanalysis has fulfilled.<br />
The mathematical processing of CdS layers transmittion specters are carried out. The structural and optical properties investigated<br />
CdS films are defined by the thickness and annealing modes CdS layer.<br />
Key words: liquid-phase chemical deposition method, X-Ray difractogram method, dispersion cogerent area, microdeformations,<br />
scanning electron microscopy, X-ray electron-probe microanalysis.<br />
41
42<br />
ÓÄÊ 621.383:537.221<br />
Ä. À. Êóäèé, Í. Ï. Êëî÷êî, Ã. Ñ. Õðèïóíîâ, Í. À. Êîâòóí, K. Þ. Êðèêóí, Å. K. Áåëîíîãîâ<br />
ÐÀÇÐÀÁÎÒÊÀ ÏËÅÍÎ×ÍÛÕ ÑËÎÅ ÑÓËÜÔÈÄÀ ÊÀÄÌÈß ÄËß ÝÊÎÍÎÌÈ×ÍÛÕ ÑÎËÍÅ×ÍÛÕ ÝËÅÌÅÍÒÎÂ<br />
Ðåçþìå<br />
Èññëåäîâàíî ñòðóêòóðó è îïòè÷åñêèå ñâîéñòâà ïëåíîê ñóëüôèäà êàäìèÿ, ïîëó÷åííûõ ìåòîäîì æèäêîôàçíîãî õèìè-<br />
÷åñêîãî îñàæäåíèÿ. Àíàëèç ïîâåðõíîñòè ïëåíêè ïðîâåäåí ñ ïîìîùüþ ðàñòðîâîé ýëåêòðîííîé ìèêðîñêîïèè. Ñòðóêòóðíûå<br />
ïàðàìåòðû îïðåäåëåíû ðåíòãåí-äèôðàêòîìåòðè÷åñêèì ìåòîäîì, â êîòîðîì îáëàñòè êîãåðåíòíîãî ðàññåèâàíèÿ (î.ê.ð.) è<br />
ìèêðîäåôîðìàöèè îïðåäåëÿëèñü ïóòåì àíàëèòè÷åñêîé îáðàáîòêè îòäåëüíûõ ðåíòãåíäèôðàêòîãðàìì. Âûïîëíåí ðåíòãåíîñïåêòðàëüíûé<br />
ýëåêòðîííî-çîíäîâûé ìèêðîàíàëèç.<br />
Ïðîâåäåíà ìàòåìàòè÷åñêàÿ îáðàáîòêà ñïåêòðîâ ïðîïóñêàíèÿ ñëîåâ ñóëüôèäà êàäìèÿ. Ñòðóêòóðíûå è îïòè÷åñêèå ñâîéñòâà<br />
èññëåäîâàííûõ ïëåíîê CdS îïðåäåëÿþòñÿ òîëùèíîé è ðåæèìàìè îòæèãà ñëîåâ CdS.<br />
Êëþ÷åâûå ñëîâà: Ìåòîä æèäêîôàçíîãî õèìè÷åñêîãî îñàæäåíèÿ, ðåíòãåíäèôðàêòîìåòðè÷åñêèé ìåòîä, îáëàñòü êîãåðåíòíîãî<br />
ðàññåèâàíèÿ, ìèêðîäåôîðìàöèè, ðàñòðîâàÿ ýëåêòðîííàÿ ìèêðîñêîïèÿ, ðåíòãåíîñïåêòðàëüíûé ýëåêòðîííî-çîíäîâûé<br />
ìèêðîàíàëèç.<br />
ÓÄÊ 621.383:537.221<br />
Ä. À. Êóä³é, Í. Ï. Êëî÷êî, Ã. Ñ. Õðèïóíîâ, Í. À. Êîâòóí, K. Þ. Êðèêóí, ª. K. Áºëîíîãîâ<br />
ÐÎÇÐÎÁÊÀ Ï˲ÂÊÎÂÈÕ ØÀв ÑÓËÜÔ²ÄÓ ÊÀÄÌ²Þ ÄËß ÅÊÎÍÎ̲×ÍÈÕ ÑÎÍß×ÍÈÕ ÅËÅÌÅÍÒ²Â<br />
Ðåçþìå<br />
Äîñë³äæåíî ñòðóêòóðó ³ îïòè÷í³ âëàñòèâîñò³ ïë³âîê ñóëüô³äó êàäì³þ, îòðèìàíèõ ìåòîäîì ð³äèííîôàçîâîãî õ³ì³÷íîãî<br />
îñàäæåííÿ. Àíàë³ç ïîâåðõí³ ïë³âêè ïðîâåäåíèé çà äîïîìîãîþ ðàñòðîâî¿ åëåêòðîííî¿ ì³êðîñêîﳿ. Ñòðóêòóðí³ ïàðàìåòðè âèçíà÷åí³<br />
ðåíòãåí-äèôðàêòîìåòðè÷íèì ìåòîäîì, â ÿêîìó îáëàñò³ êîãåðåíòíîãî ðîçñ³ÿííÿ (î.ê.ð.) òà ì³êðîäåôîðìàö³¿ âèçíà÷àëèñü<br />
øëÿõîì àíàë³òè÷íî¿ îáðîáêè îêðåìèõ ðåíòãåíäèôðàêòîãðàì. Âèêîíàíèé ðåíòãåíîñïåêòðàëüíèé åëåêòðîííî-çîíäîâèé<br />
ì³êðîàíàë³ç.<br />
Ïðîâåäåíà ìàòåìàòè÷íà îáðîáêà ñïåêòð³â ïðîïóñêàííÿ øàð³â ñóëüô³äó êàäì³þ. Ñòðóêòóðí³ òà îïòè÷í³ âëàñòèâîñò³ äîñë³äæåíèõ<br />
ïë³âîê CdS âèçíà÷àþòüñÿ òîâùèíîþ òà ðåæèìàìè â³äïàëó øàð³â CdS.<br />
Êëþ÷îâ³ ñëîâà: Ìåòîä ð³äèííîôàçíîãî õ³ì³÷íîãî îñàäæåííÿ, ðåíòãåí-äèôðàêòîìåòðè÷åñêèé ìåòîä, îáëàñòü êîãåðåíòíîãî<br />
ðîçñ³ÿííÿ, ì³êðîäåôîðìàö³¿, ðàñòðîâà åëåêòðîííà ì³êðîñêîï³ÿ, ðåíòãåíîñïåêòðàëüíèé åëåêòðîííî-çîíäîâèé ì³êðîàíàë³ç.
UDC 621.315.592<br />
I. K. DOYCHO 1 , S. A. GEVELYUK 1 , O. O. PTASHCHENKO 1 , E. RYSIAKIEWICZ-PASEK 2 , S. O. ZHUKOV 1<br />
1 I. I. Mechnikov Odesa National University, Dvoryanska St., 2, Odesa, 65082, Ukraine<br />
2 Institute of Physics, Wroc³aw University of Technology, W.Wyspianskiego 27, 50-370 Wroc³aw, Poland<br />
POROUS GLASSES WITH CDS INCLUSIONS LUMINESCENCE KINETICS<br />
PECULIARITIES<br />
The kinetics of the luminescence of porous glasses with CdS inclusions was studied at liquid nitrogen<br />
temperature. It is shown that the luminescence decay curves after excitation switch-off are different<br />
for different types of glasses. For CdS in matrices with small enough pores, a short-time “flash”<br />
of the luminescence was observed after the excitation turn-off. Additional doping of the samples with<br />
Na 2 S enhanced the “flash”. The flash intensity and duration depended on the exciting photons energy.<br />
The effect is explained by thermo-optical excitation of electrons.<br />
1. INTRODUCTION<br />
Nanocrystals of semiconductors and, in particular<br />
III–V materials, in porous glass are prospective<br />
for optical and electronic applications. The complexity<br />
of porous glass morphology and a rich luminescent<br />
spectrum of bulk CdS [1-4] allows to expect<br />
the appearance of various point defects which could<br />
be considered as shallow traps and, in other cases, are<br />
the centers of luminescence. While studying the luminescence,<br />
it is necessary to take into account both the<br />
contribution of charge carrier recombination as well<br />
as the changes in optical properties of the porous matrix<br />
caused by the excitation. Some information about<br />
the mechanism of the luminescence excitation, as well<br />
as on the nature of radiative recombination, could be<br />
obtained by an analysis of the luminescence rise and<br />
decay curves [1, 5].<br />
The present paper investigates photo-luminescence<br />
stationary and transient characteristics of CdS<br />
nano-formations in a porous glass matrix at the temperature<br />
of 77 K. A tentative model is used in order to<br />
explain some peculiarities of the luminescence decay<br />
curves observed.<br />
2. EXPERIMENTAL PROCEDURE<br />
Porous silicate glasses obtained from the two-phase<br />
sodium-boron-silica glasses by chemical etching of the<br />
unstable sodium-borate phase in hydrochloric acid [6]<br />
were used in our experiments as a model matrix medium<br />
for forming the semiconductor nano-particles.<br />
An initial two-phase glass was annealed at 760 K or<br />
930 K for 100 and 150 hours, respectively, in order<br />
to enhance the phase separation. After a subsequent<br />
etching of the glass annealed at 760 K, we obtained<br />
mesoporous glass with the mean pore radius about 15<br />
nm (glass of À type). A similar treatment of the glass<br />
annealed at 930 K gives the mean pore radius of 75 nm<br />
(glass of C type).<br />
The etching solution used for etching the sodiumborate<br />
phase off, interacted also with the silicate skeleton,<br />
which led to the formation of the secondary silica<br />
gel inside the pores.<br />
It is possible to remove the secondary silica gel<br />
from pores almost completely by the subsequent treatment<br />
in KOH solution. However, it leads to an excessive<br />
etching of the pores. After such additional treatment<br />
of A-glass we obtained a glass with gel-free pores<br />
of 23 nm radius. It is referred in this paper as B-glass.<br />
An additional KOH solution etching of C-glass<br />
produced gel-free pores with the radius of 160 nm. It<br />
is referred as D-glass.<br />
CdS were impregnated into the glasses containing<br />
silica gel (A) and also into B- and D-glasses which<br />
were practically free of silica gel. CdS nanoclusters in<br />
the porous matrices were obtained by chemical deposition<br />
using the technique described in [3, 7].<br />
For the increase of the of sulfur ions content (playing<br />
the role of a luminescence activator), some of the<br />
prepared samples were saturated with Na 2 S in water<br />
solution [8].<br />
Luminescence was excited by the monochromatized<br />
light of a 1 kW Xenon lamp over the wavelength<br />
range of 400-700 nm. The luminescence spectra, as<br />
well as the excitation spectra, were measured in stationary<br />
and transient regimes. Luminescence rise-<br />
and decay curves were analyzed. All the measurements<br />
were performed at the liquid nitrogen temperature (77<br />
K) using a standard set-up.<br />
Some of specimens were gamma-irradiated using<br />
a Co 60 facility with the power of 9.8 rad/s. This power<br />
was used to induce some modification of the material<br />
recombination properties [9], called a small dose effect.<br />
3. EXPERIMENTAL RESULTS<br />
© I. K. Doycho, S. A. Gevelyuk, O. O. Ptashchenko, E. Rysiakiewicz-pasek, S. O. Zhukov, 2009<br />
Curves 1 and 2 at Fig. 1 present the photoluminescence<br />
spectra of CdS clusters in an A-matrix before<br />
and after Na 2 S saturation, respectively. It is seen that<br />
the main red luminescence band is strongly broadened.<br />
The short-well shoulder shows that it is not<br />
elementary one. Na 2 S saturation enhances the luminescence<br />
intensity and shifts the spectrum to shorter<br />
wavelengths.<br />
Fig. 2 illustrates the luminescence spectrum evolution<br />
after switch-off the excitation of A-sample with<br />
Na 2 S excess. It is seen that the luminescence spectrum<br />
changes in time. The phosphorescence band at 725<br />
43
nm has much higher decay time than the short-wavelength<br />
part of the spectrum. It suggests that the photon<br />
generation of 725 nm band and the short-wavelength<br />
radiative recombination are caused by the differently<br />
located centers.<br />
44<br />
�, arb.un.<br />
5<br />
4<br />
3<br />
2<br />
1<br />
2<br />
0<br />
0.4 0.5 0.6 0.7 0.8<br />
�, �m 0.9<br />
Fig. 1. Photoluminescence spectra of CdS clus-ters in A-matrix:<br />
1 — initial; 2 — after Na 2 S saturation<br />
�, arb.un.<br />
6<br />
5 1<br />
4<br />
3<br />
2<br />
1<br />
0<br />
0.55 0.6 0.65 0.7 0.75<br />
�, �m<br />
Fig. 2. Phosphorescence spectra of A-sample with Na S ex- 2<br />
cess at different times after switching off the excitation: 1 — 0.2 s;<br />
2 — 0.8 s; 3 — 1.2 s; 4 — 2.0 s.<br />
Curves 1, 2, and 3 at Fig. 3 show the luminescence<br />
rise after switching on the excitation in A-, B-, and<br />
D-samples. The rise curves for all of the glass types<br />
practically coincide and are exponential:<br />
Φ ( t) =Φ [1−exp( −t/ τ )] , (1)<br />
0<br />
where Φ 0 is a constant; τ r denotes the rise time. For all<br />
the curves at Fig. 3 τ r =1.8 s.<br />
2<br />
3<br />
4<br />
1<br />
r<br />
�, arb.un.<br />
1.0<br />
0.8<br />
0.6<br />
0.4<br />
0.2<br />
0<br />
3<br />
1<br />
2<br />
0 3 6<br />
Fig. 3. CdS nanoclusters photoluminescence rise curves for<br />
different types of porous glass matrix after 500 nm excitation: 1 —<br />
A; 2 — B; 3 — D.<br />
Curve 1 at Fig. 4 represents the phosphorescence<br />
decay for CdS in pores of B-glass (in pores with small<br />
diameter). The curve has a shoulder at the beginning<br />
of the decay. After γ-irradiation with a dose of 104 rad,<br />
the shoulder disappears (low dose effect), what is illustrated<br />
by curve 3, and the decay curve is practically<br />
identical to curve 2 (for CdS in wide pores).<br />
1.0<br />
0.8<br />
0.6<br />
0.4<br />
0.2<br />
0<br />
�, arb.un.<br />
3<br />
2<br />
1<br />
t, s<br />
0 1 2 3 4<br />
t, s<br />
Fig. 4. CdS phosphorescence decay curves in B- (1 and 3) and<br />
D-(2) matrices. 3 – after γ-irradiation with a dose of 10 4 rad.<br />
The decay curve 2 at Fig. 4, measured on CdS in<br />
D-matrix, is exponential:<br />
Φ(t) = Φ exp(-t/τ ), (2)<br />
o d<br />
where τ is the decay time. For curve 2 τ =1.78 s. The<br />
d d<br />
rise- and decay times are very high:<br />
τ , τ >> τ , (3)<br />
r d n<br />
where τ is the life time for electrons (τ ~1 ms). The<br />
n n<br />
strong inequality (3) suggests that CdS microcrystals<br />
9
in pores have very high concentration of shallow traps,<br />
which are in equilibrium with c-band.<br />
Our measurements show that the phosphorescence<br />
kinetics of CdS clusters in A-glass usually exhibits a<br />
flash-up after the excitation turn-off, as demonstrated<br />
at Fig 5. Curve 1, measured after excitation with photons<br />
of λ = 550 nm (hν = 2.25 eV), has a pronounced<br />
shoulder. Excitation at λ = 450 nm (hν = 2.76 eV)<br />
leads to a flash, as illustrated with curve 2 at Fig. 5.<br />
A treatment of the sample in Na 2 S enhances the flash,<br />
as shown by curves 3, 4. A comparison of curves 3 and<br />
4 at Fig. 5, obtained after excitation with photons of<br />
hν = 2.25 eV and hν = 2.76 eV, correspondingly, demonstrates<br />
that the flash becomes more detectable with<br />
increasing photon energy.<br />
�, arb.un.<br />
1.2<br />
1.0<br />
0.8<br />
0.6<br />
0.4<br />
0.2<br />
0<br />
0<br />
2<br />
1<br />
3<br />
4<br />
1 2 3 4<br />
t, s<br />
Fig. 5. Decay kinetics of CdS nanoclusters phosphorescence<br />
in A — matrices: 1, 2 – initial; 3, 4 – after Na 2 S-treatment.<br />
Curves 1, 3 were measured after excitation with λ=550 nm; 2,<br />
4 – after illumination with λ=450 nm.<br />
The intensity of the flash is different for various<br />
phosphorescence spectral bands. The most intensive<br />
flash is observed in the band at 725 nm for CdS clusters<br />
in A-matrix with the excess of Na 2 S after the 450<br />
nm excitation. This band is dominant in the stationary<br />
photoluminescence spectrum of these samples,<br />
as shown by curve 2 at Fig. 1, and has a higher decay<br />
time, as shown at Fig. 2.<br />
4. MODEL AND DISCUSSION<br />
The most interesting result of our experiment is<br />
the effect of the mean pore size on the photoluminescence<br />
and phosphorescence spectra and the kinetics.<br />
We can analyze this effect by the use of a recombination<br />
scheme represented at Fig. 6.<br />
The electron-hole pair generation of intensity G<br />
is provided by photons absorption. Additional generation<br />
G t has photo-thermal nature. The observed very<br />
long rise- and decay times suggest the presence of a<br />
high concentration of shallow traps t, being in thermal<br />
equilibrium with c-band. Radiative recombination<br />
occurs at deep r-centers. Non-radiative recombination<br />
takes place at deep s-centers (“fast recombination<br />
centers”).The corresponding differential equations set<br />
is as follows:<br />
dn<br />
= G+ Gt − Ctpn t + CtNctnt −Cspsn− Crprn ; (4)<br />
dt<br />
dpr<br />
= Cnp ′ r r − Crpn r ; (5)<br />
dt<br />
dps<br />
= Cnp ′ s s − Cspn s ; (6)<br />
dt<br />
dnt<br />
= Cpn t t − CN t ct,<br />
(7)<br />
dt<br />
where n, p are the free carriers concentrations; C , C , C t r s<br />
denote the electron capture coefficients for t-, r- and<br />
s- centers, respectively; C′ , C′ are the hole capture<br />
r s<br />
coefficients for r- and s- centers, correspondingly; n , t<br />
n , n are the concentrations of electrons on t-, r- and<br />
r s<br />
s- centers, respectively; p , p , p are the concentrations<br />
t r s<br />
of empty t-, r- and s- centers, respectively;<br />
N ≡ N exp( − E / kT)<br />
, (8)<br />
G<br />
Gt<br />
ct c ct<br />
t<br />
Fig. 6. Recombination scheme for photo-luminescence in<br />
CdS clusters<br />
N is the effective state density in c-band; E is the<br />
c ct<br />
depth of t-level; kT is the Boltzmann factor.<br />
The neutrality equation can be written as<br />
n+ nt 0<br />
= pr − pr + ps,<br />
(9)<br />
0<br />
where p r is determined by the concentration of electrons<br />
on deep t traps. In (9) the strong inequality<br />
d<br />
n>>p<br />
is taken into account as well.<br />
(10)<br />
For s-centers, as fast non-radiative recombination<br />
centers in CdS, the inequalities<br />
C>>C ; p p (13)<br />
takes place.<br />
td<br />
s<br />
c<br />
r<br />
v<br />
45
The observed exponential rise- and decay curves,<br />
depicted at Figs.3 and 4, suggest that differential equations<br />
(4) — (7) for our case are linearized. It means<br />
that in (12)<br />
pr = const . (14)<br />
And this could occur if we have in the equality (9)<br />
0<br />
pr ≅ pr >> n+ nτ, ps<br />
. (15)<br />
Moreover, it means that the concentration n of t<br />
electrons, captured on shallow t-centers, is proportional<br />
to free electrons concentration n. It could occur<br />
in the case of<br />
n gr /(1 − gr) G . (27)<br />
Our measurements reveal the following conditions<br />
for the flash-effect in CdS nanocrystals: a) the pores<br />
in the matrix must be small enough. The effect occurs<br />
only in A- and B-matrices; b) the flash is observed only<br />
in 725 nm phosphorescence band; c) Na S treatment<br />
2<br />
enhances the effect; d) excitation with photons of hν =<br />
2.75 eV gives more pronounced effect than of 2.25eV;<br />
e) γ-irradiation suppresses the effect.<br />
The first condition suggests that the flash-effect<br />
is characteristic for small enough CdS nanocrystals.<br />
This observation is in agreement with results [5] and<br />
the fact that lowering the nanocrystals’ size makes<br />
slower the CdS phosphorescence decay [3, 4, 7].<br />
The influence of Na S treatment on the flash inten-<br />
2<br />
sity argues that point defects, containing sulfur atoms,<br />
are responsible for this effect. Inhibition of the effect<br />
by γ-radiation can be explained by destroying the lowsize<br />
nanoclusters as a result of γ-quanta absorption [9].<br />
The photoluminescence and phosphorescence of CdS<br />
in D-glasses is less sensitive to radiation, which is in<br />
accordance with previously reported results [9]. The<br />
flash-effect disappeared after a long time storage (over<br />
half a year; A-specimens were exposed to the air). It is<br />
. also suppressed by low temperature annealing of the<br />
As seen at Figs. 4 and 5, CdS clusters under certain samples in air. This phenomenon could be attributed<br />
conditions exhibit a phosphorescence flash. The flash- to slow oxidation of excess sulfur during the storage in<br />
effect could be explained by using the model presented the open air.<br />
above. The role of photothermal electron generation<br />
and presence of two kinds of recombination centers<br />
are essential for this explanation. From equation (4),<br />
in a stationary case, we obtain for the stationary elec-<br />
5. CONCLUSIONS<br />
tron concentration<br />
1. The phosphorescence of CdS nanoclusters in<br />
st 0 st<br />
n = ( G+ Gt)/( Crpr + Csps )<br />
and for the radiative recombination intensity<br />
(20) porous glass exhibits a flash after excitation turn-off.<br />
This effect occurs only in glass with sufficiently small<br />
pores and can be ascribed to recombination in small<br />
st<br />
R<br />
r<br />
= gr( G+ Gt)<br />
, (21)<br />
enough CdS clusters. The flash is observed only in the<br />
spectral band of hν = 1.71 eV which corresponds to<br />
where<br />
the recombination of free electrons at deep centers.<br />
gr 0 0 st<br />
≡ Crpr /( Crpr + Csps ) . (22)<br />
2. The recombination model, that includes two<br />
kinds of recombination centers and centers respon-<br />
After turn off excitation G the intensity of the photothermal<br />
excitation<br />
Gt ≠ 0 , (23)<br />
because the concentration of electrons on the corresponding<br />
centers exceeds its equilibrium value. Moreover,<br />
the fast recombination centers (s-centers) for a<br />
short time are filled by electrons, so<br />
sible for the photo-thermal electron generation, could<br />
explain the flash-effect.<br />
3. The introduction of the excess sulphur ions enhances<br />
the effect. It suggests that the centers, responsible<br />
for flash, contain sulphur atoms.<br />
4. The flash-effect is suppressed by γ-irradiation<br />
as well as by low-temperature annealing in air. This<br />
instability could be ascribed to small CdS clusters destruction.<br />
The γ-stability of CdS nanoclusters is en-<br />
Ps → 0 , (24) hanced with the increase of their size.<br />
and one could obtain for the electrons concentration References<br />
n = G C p<br />
(25)<br />
1. Guyot-Sionnest P., Halas N.J., Mattoussi H., Wang Z.L.,<br />
Woggon U. Quantum Confined Semiconductor Nanostruc-<br />
46<br />
0<br />
t /( r r)
tures // MRS Symposium Proceedings. — 2004. — V. 789.<br />
429 p.<br />
2. Smyntyna V.A., Skobeeva V.M., Malushin N.V., Pomogailo<br />
A.D. Influence of matrix on photoluminesence of CdS<br />
nanocrystals // Photoelectronics. — 2006. — V. 15. P. 38-42.<br />
3. Gevelyuk S.A., Doycho I.K., Mak V.T., Zhukov S.A. Photoluminescence<br />
and structural properties of nano-size CdS inclusions<br />
in porous glasses // Photoelectronics — 2007. — V. 16.<br />
P. 75-79.<br />
4. Rysiakiewicz-Pasek E., Polañska J., Gevelyuk S.A., Doycho<br />
I.K., Mak V.T., Zhukov S.A. The photoluminescent<br />
properties of CdS clusters of different size in porous glasses //<br />
Optica Applicata. — 2008. — V. 38, ¹ 1. P. 93-100.<br />
5. Gaponenko S.V. Optical Properties of Semiconductor Nanocrystals<br />
// Cambridge University Press: 1998. — 300 p.<br />
UDC 621.315.592<br />
I. K. Doycho, S. A. Gevelyuk, O. O. Ptashchenko, E. Rysiakiewicz-Pasek, S. O. Zhukov<br />
6. Rysiakiewicz-Pasek E., Gevelyuk S.A., Doycho I.K., VorobjovaV.A.<br />
Application of Porous Glass in Ophtalmic Prosthetic<br />
Repair // Journal of Porous Materials. — 2004. — V. 11, ¹ 1.<br />
P. 21-29.<br />
7. Rysiakiewicz-Pasek E., Zalewska M., Polañska J. Optical<br />
Properties of CdS-doped Porous Glasses // Optical Materials.<br />
— 2008. — V. 30, ¹ 5. P. 777-779.<br />
8. Fu Z., Zhou S., Shi J., Zhang S. Effects of Precursors on<br />
the Crystal Structure and Photoluminescence of CdS Nanocrystalline//<br />
Material Research Bulletin. — 2005. — V. 40. P.<br />
1591-1598.<br />
9. Doycho I.K., Gevelyuk S.A., Kovalenko M.P., Prokopovich<br />
L.P., Rysiakiewicz-Pasek E. Small doses γ-irradiation effect<br />
on the photoluminescence properties of porous glasses //<br />
Optica Applicata. — 2003. — V. 33, ¹ 1. P. 55-60.<br />
POROUS GLASSES WITH CDS INCLUSIONS LUMINESCENCE KINETICS PECULIARITIES<br />
Abstract<br />
The kinetics of the luminescence of porous glass with CdS inclusions was studied at liquid nitrogen temperature. It is shown that<br />
the luminescence decay curves after excitation switching-off are different for different types of glasses. For CdS in matrices with small<br />
enough pores, a short-time “flash” of the luminescence was observed after the excitation turn off. Additional doping of the samples<br />
with Na 2 S enhanced the “flash”. The flash intensity and duration depended on the exciting photons energy. The effect is interpreted by<br />
thermo-optical excitation of electrons.<br />
Key words: doped porous glasses, cadmium sulphide, luminescence kinetics.<br />
ÓÄÊ 621.315.592<br />
È. Ê. Äîé÷î, Ñ. À. Ãåâåëþê, À. À. Ïòàùåíêî, Å. Ðûñàêåâè÷-Ïàñåê, Ñ. À. Æóêîâ<br />
ÎÑÎÁÅÍÍÎÑÒÈ ÊÈÍÅÒÈÊÈ ËÞÌÈÍÅÑÖÅÍÖÈÈ ÏÎÐÈÑÒÎÃÎ ÑÒÅÊËÀ Ñ ÂÊËÞ×ÅÍÈßÌÈ CDS<br />
Ðåçþìå<br />
Èññëåäîâàíà êèíåòèêà ôîòîëþìèíåñöåíöèè ïîðèñòîãî ñòåêëà ñ âêëþ÷åíèÿìè CdS ïðè òåìïåðàòóðå æèäêîãî àçîòà. Ïîêàçàíî,<br />
÷òî êðèâûå ñïàäà ôîòîëþìèíåñöåíöèè ïîñëå âûêëþ÷åíèÿ âîçáóæäåíèÿ ðàçëè÷íû äëÿ ðàçëè÷íûõ òèïîâ ñò¸êîë. Äëÿ<br />
CdS â ìàòðèöàõ ñ äîñòàòî÷íî ìàëûìè ïîðàìè íàáëþäàëàñü êðàòêîâðåìåííàÿ âñïûøêà ëþìèíåñöåíöèè ïîñëå âûêëþ÷åíèÿ<br />
âîçáóæäåíèÿ. Äîïîëíèòåëüíîå íàñûùåíèå îáðàçöîâ Na 2 S óñèëèâàëî óêàçàííóþ âñïûøêó. Èíòåíñèâíîñòü è äëèòåëüíîñòü<br />
âñïûøêè çàâèñåëà îò ýíåðãèè âîçáóæäàþùèõ ôîòîíîâ. Ýôôåêò îáúÿñí¸í òåðìîîïòè÷åñêèì âîçáóæäåíèåì ýëåêòðîíîâ.<br />
Êëþ÷åâûå ñëîâà: êèíåòèêà ëþìèíåñöåíöèè, ïîðèñòîå ñòåêëî, ñóëüôèä êàäìèÿ.<br />
ÓÄÊ 621.315.592<br />
². Ê. Äîé÷î, Ñ. À. Ãåâåëþê, Î. Î. Ïòàùåíêî, Å. Ðèøÿêåâè÷-Ïàñåê, Ñ. Î. Æóêîâ<br />
ÎÑÎÁËÈÂÎÑÒ² ʲÍÅÒÈÊÈ ËÞ̲ÍÅÑÖÅÍÖ²¯ ØÏÀÐÈÑÒÎÃÎ ÑÊËÀ Ç ÂÊÐÀÏËÅÍÍßÌÈ CDS<br />
Ðåçþìå<br />
Äîñë³äæåíî ê³íåòèêó ëþì³íåñöåíö³¿ øïàðèñòîãî ñêëà ³ç âêðàïëåííÿìè CdS ïðè òåìïåðàòóð³ ð³äêîãî àçîòó. Ïðîäåìîíñòðîâàíî,<br />
ùî êðèâ³ ñïàäó ëþì³íåñöåíö³¿ ï³ñëÿ âèìêíåííÿ çáóäæåííÿ íåîäíàêîâ³ äëÿ ð³çíèõ òèï³â ñêëà. Äëÿ CdS â ìàòðèöÿõ<br />
ç äîñòàòíüî ìàëèìè ïîðàìè ñïîñòåð³ãàâñÿ êîðîòêî÷àñíèé ñïàëàõ ëþì³íåñöåíö³¿ ï³ñëÿ âèìêíåííÿ çáóäæåííÿ. Äîäàòêîâå íàñè÷åííÿ<br />
çðàçê³â Na 2 S ï³äñèëþâàëî çàçíà÷åíèé ñïàëàõ. ²íòåíñèâí³ñòü ³ òðèâàë³ñòü ñïàëàõó çàëåæàëà â³ä åíåð㳿 çáóäæóþ÷èõ<br />
ôîòîí³â. Åôåêò ïîÿñíåíî òåðìîîïòè÷íèì çáóäæåííÿì åëåêòðîí³â.<br />
Êëþ÷îâ³ ñëîâà: ê³íåòèêà ëþì³íåñöåö³¿, øïàðèñòå ñêëî, ñóëüô³ä êàäì³þ.<br />
47
48<br />
UDÑ 539.186<br />
N.V. MUDRAYA<br />
Odessa National Polytechnical University, Odessa<br />
DENSITY FUNCTIONAL APPROACH TO ATOMIC AUTOIONIZATION<br />
IN AN EXTERNAL ELECTRIC FIELD: NEW RELATIVISTIC SCHEME<br />
Within the S-matrix Gell-Mann and Low formalism and the relativistic perturbation theory we<br />
present a new relativistic density functional theory scheme to description of the atomic autoionization<br />
in an external dc electric and laser field.<br />
INTRODUCTION<br />
The photo- and autoionization phenomena in<br />
atomic and molecular systems, solids, semiconductors<br />
etc attracts a great interest because of the very<br />
perspective applications in the quantum electronics,<br />
laser physics, technical physics and creation of the<br />
new types of devices in the opto- and molecular electronics<br />
[1-6]. In the last years extensive experimental<br />
and theoretical studies of photo-, auto- and multiple<br />
ionization in strong laser fields have revealed a number<br />
of unexpected effects and features. Moreover,<br />
some phenomena should can be described as multiphoton<br />
ones and independent electron processes<br />
(see e.g. [1]). This was concluded on the basis of a<br />
detailed analysis of experimental data together with<br />
precise calculations of single photo-ionization rates<br />
[2, 3]. Several experiments (see e.g.[4]) then revealed<br />
a pronounced “knee” structure in the yield vs peak<br />
laser intensity, typically plotted with logarithmic axis<br />
because of the wide range of values covered. Further<br />
ingenious experiments that resolved the joint momentum<br />
distributions of the outgoing electrons then<br />
showed that they often leave the atom with the same<br />
moments. This has triggered a number of theoretical<br />
studies of this process, including S-matrix calculations<br />
for the full cross sections [3], and investigations<br />
of simplified classical and quantum models. Among<br />
the models are so-called aligned-electron models,<br />
in which electrons move in a one-dimensional (1D)<br />
regularized Coulomb potential, or quasi three-dimensional<br />
(3D) ones with the centre of mass of the<br />
electrons confined to move along the field polarization<br />
axis. At the same time, an exact solution of the<br />
time-dependent Schrödinger equation even for two<br />
electrons in a laser field remains a formidable task.. It<br />
is well known that the autoionization phenomena in<br />
the heavy atomic systems should be considered exclusively<br />
within the relativistic formalism. The suitable<br />
basis is the time-dependent Dirac equation, which<br />
is remained by the very complicated problem to be<br />
solved. Surely, if one consider a dc electric field and<br />
its effect on the autoionization process in the atomic<br />
system, the standard atomic relativistic approaches<br />
can be used as the zeroth approximation. Besides, one<br />
must take into account a group of the known complicated<br />
correlation effects. Here we mean, for example,<br />
the relaxation processes due to Coulomb interaction<br />
between moving away electron (electrons in a case of<br />
the laser induces multiple ionization) and resulting in<br />
the electron distribution in the vacancy field have no<br />
time to be over prior to the transition. It is known that<br />
a consistent theory of the atomic autoionization is to<br />
take into account correctly a definite number of the<br />
correlation (polarization, relaxation) effects, including<br />
the energy dependence of the vacancy mass operator,<br />
the continuum pressure, spreading of the initial<br />
state over a set of configurations etc. [1-5]. It should<br />
be reminded that hitherto these effects are not described<br />
adequately in the modern theoretical scheme.<br />
As example, let us remind such wide-spread methods<br />
as the : Dirac-Fock, relativistic Hartree-Fock methods,<br />
random phase approximation (RPA) and RPA<br />
with exchange, different model and pseudo potential<br />
schemes, density-functional formalism and its relativistic<br />
generalization etc. (see e.g. [3-10]).<br />
In this paper we present a new relativistic density<br />
functional theory scheme to description of the atomic<br />
autoionization in an external dc electric and laser field<br />
within the S-matrix Gell-Mann and Low formalism<br />
and the relativistic perturbation theory. New scheme<br />
has to be applied to studying the autoionization phenomena<br />
characteristics in the atomic and molecular<br />
(obviously heavy) systems, and quasi-molecules and<br />
solids. The novel elements consist in an implementation<br />
of the relativistic Dirac-Kohn-Sham density<br />
functional theoretical scheme to the S-matrix Gell-<br />
Mann and Low formalism and using the optimized<br />
electron wave functions basis’s of the relativistic perturbation<br />
theory in order to describe the fundamental<br />
atomic characteristic of autoionization in an external<br />
field [3,10].<br />
AN OPTIMIZED RELATIVISTIC DENSITY<br />
FUNCTIONAL APPROACH TO AN ATOMIC<br />
AUTOIONIZATION<br />
As usually, we will describe the multielectron system<br />
(atom, molecule etc.) by the Dirac relativistic<br />
Hamiltonian (the atomic units are used) (see e.g.[3]):<br />
∑ i ∑ ( i j)<br />
(1)<br />
H = h(r ) + V rr<br />
i i> j<br />
Here h(r) is one-particle Dirac Hamiltonian for<br />
electron in a field of a nucleus (nuclei), V is potential<br />
of the inter-electron interaction. In order to take into<br />
account the retardion effect and magnetic interaction<br />
© N.V. Mudraya, 2009
in the lowest order on parameter α 2 (the fine structure<br />
constant) one could write :<br />
( 1−αα<br />
i j)<br />
( i j) ( ij ij)<br />
V r r = exp iω r ⋅ (2)<br />
rij<br />
where ω is the transition frequency; α ,α are the Di-<br />
ij i j<br />
rac matrices. The Dirac equation potential includes<br />
the electric potential of a nucleus and exchange-correlation<br />
potentials. The standard KS exchange potential<br />
is as follows [7]:<br />
V () r (1/ )[3 ()] r<br />
KS<br />
X<br />
2 1/3<br />
=− π πρ (3)<br />
In the local density approximation the relativistic<br />
potential is as follows [7-9]:<br />
δEX[ ρ(<br />
r)]<br />
VX[ ρ ( r), r]<br />
=<br />
(4)<br />
δρ()<br />
r<br />
where EX[ ρ ( r)]<br />
is the exchange energy of the multielectron<br />
system corresponding to the homogeneous<br />
density ρ () r , which is obtained from a Hamiltonian<br />
having a transverse vector potential describing the<br />
photons. In this theory the exchange potential is [10]:<br />
2 1/2<br />
KS 3 [ β+ ( β + 1) ] 1<br />
VX[ ρ ( r), r] = VX ( r)<br />
⋅{ ln − } (5)<br />
2 1/2<br />
2 ββ ( + 1) 2<br />
where<br />
2 1/3<br />
β= 3 πρ ( r)] / c.<br />
The corresponding correlation functional is [3]:<br />
1/3<br />
VC[ ρ ( r), r] = −0.0333⋅b⋅ ln[1 + 18.3768 ⋅ρ ( r)<br />
] , (6)<br />
where b is the optimization parameter (look details in<br />
ref. [3,11]). Earlier it has been shown [3,9,10] that an<br />
adequate description of the atomic characteristics requires<br />
using the optimized basis’s of wave functions.<br />
Within the frame of QED PT approach [3] to the<br />
atomic autoionization effect, the corresponding transition<br />
probability is knowingly defined by the square<br />
of an electron interaction matrix element having the<br />
form (see details also in ref. [12]):<br />
( )( )( )( )<br />
ω<br />
V1234 = ⎡⎣ j1j2j3j4⎡⎣ ×<br />
μ ⎛ jj 1 3<br />
× ∑ ( − 1) ⎜<br />
λμ m −m λ ⎞<br />
⎟×<br />
ReQλ ( 1234)<br />
;<br />
μ<br />
= + . (7)<br />
⎝ 1 3 ⎠<br />
Qλ Qul<br />
Qλ Br<br />
Qλ<br />
Qul<br />
Br<br />
The terms Qλ and Qλ correspond to subdivision<br />
of the potential into Coulomb part cos|ω|r /r 12 12<br />
and Breat one, cos|ω|r α α /r . The real part of the<br />
12 1 2 12<br />
electron interaction matrix element is determined using<br />
expansion in terms of Bessel functions:<br />
12 1 2<br />
∑() J 1 ( r< ) J 1 ( r> ) Pλ(<br />
cosrr<br />
1 2)<br />
. (8)<br />
λ+ −λ−<br />
λ= 0<br />
cos ω r12<br />
π<br />
= ×<br />
r 2 rr<br />
× λ ω ω<br />
2 2<br />
where J is the 1st order Bessel function, (λ)=2λ+1.<br />
Qul<br />
The Coulomb part Qλ is expressed in terms of radial<br />
integrals R λ , angular coefficients S λ [3]:<br />
1 2<br />
ReQ<br />
Qul<br />
λ<br />
{ Rl( ) Sλ( ) R ( %<br />
λ<br />
% ) S ( %<br />
λ<br />
% )<br />
( 1243) ( 1243) ( 1243) ( 1243 )} .<br />
1<br />
= Re<br />
Z<br />
1243 1243 + 1243 1243<br />
+ R %% S %% + R %%%% S %%%% (9)<br />
λ λ λ λ<br />
As a result, the Auger decay probability is expressed<br />
in terms of ReQ (1243) matrix elements:<br />
λ<br />
Re R 1243 =<br />
∫∫<br />
=<br />
λ ( )<br />
() () () ()<br />
() 1 () 1<br />
( ) ( )<br />
= drr r f r f r f r f r ×<br />
2 2<br />
11 2 1 1 3 1 2 2 4 2<br />
× Zλ r< Zλ r<br />
. (10)<br />
><br />
where f is the large component of radial part of single<br />
electron state Dirac function and function Z is defined<br />
as follows [3]:<br />
λ+ 1<br />
2<br />
1 ( 13 )<br />
() ⎡ 2 ⎤ J αω r<br />
λ+<br />
1 2<br />
Zλ<br />
= ⎢ ⎥<br />
.<br />
13 Z λ<br />
⎢⎣ ω α ⎥⎦ r Γ( λ+ 3<br />
2)<br />
The angular coefficient is defined by standard way<br />
[7]. The other items in (9) include small components<br />
of the Dirac functions; the sign “∼” means that in (9)<br />
the large radial component f is to be changed by the<br />
i<br />
small g one and the moment l is to be changed by<br />
i i<br />
l% i = li<br />
−1for<br />
Dirac number æ > 0 and l +1 for æ
50<br />
2<br />
∑ ( ) ( )<br />
()( )<br />
∑∑ Qλ αkγβ Qλ βγkα , (14)<br />
λ λ j βγ≤ f k> f<br />
α<br />
The formula (14) defines the total autoionization<br />
level width. The partial items of the ∑∑ sum an-<br />
βγ k<br />
swer to contributions of α-1→(βγ) -1K channels resulting<br />
in formation of two new vacancies βγ and one free<br />
electron k: ω =ω +ω –ω . The final expression for<br />
k α β α<br />
the width in the representation of jj-coupling scheme<br />
of single-electron moments has the form:<br />
o o<br />
1 1<br />
o o<br />
2 2 ∑<br />
jl k k<br />
o o<br />
1 1<br />
o o<br />
2 2 o<br />
2<br />
(15)<br />
Γ (2 jl,2 jl; J) = 2 | Γ(2<br />
jl,2 jl;1 l, kjl)|<br />
Here the summation is made over all possible<br />
decay channels. The formulas for the autoionization<br />
probability include the radial integrals R α (αkγβ),<br />
where one of the functions describes electron in the<br />
continuum state. When calculating this integral, the<br />
correct normalization of the function Ψ k is a problem.<br />
Naturally, the correctly normalized function should<br />
have the following asymptotic at r→0<br />
1<br />
⎧ −2<br />
−<br />
2<br />
⎡ ( Z) ⎤<br />
f ⎫<br />
sin(<br />
kr ) ,<br />
− 1 ⎪ ω+ α +δ<br />
2 ⎣ ⎦<br />
⎬→( πω)<br />
⎨<br />
(16)<br />
1<br />
g<br />
−2<br />
−<br />
⎭ ⎪⎡ 2<br />
ω−( α Z) ⎤ cos(<br />
kr +δ)<br />
.<br />
⎩⎣<br />
⎦<br />
When integrating the master system, the function<br />
is calculated simultaneously:<br />
()<br />
N r<br />
{ ( ) ( ) } 1 −<br />
2 −2 2 2<br />
2<br />
⎡<br />
k f ⎡<br />
k k Z ⎤ −<br />
g ⎡ k k Z ⎤ ⎤<br />
⎣ ⎦ ⎣ ⎦<br />
= πω ω + α + ω + α<br />
⎣ ⎦ .<br />
It can be shown (see [3] that at r → ∞, N(r)→N , k<br />
where N is the normalization of functions f , g of<br />
k k k<br />
continuous spectrum satisfying the condition (17). In<br />
this relation, the procedure is equivalent to the same<br />
procedure in a case of the Auger decay probability determination<br />
[12].<br />
UDÑ 539.186<br />
N. V. Mudraya<br />
=<br />
CONCLUSION<br />
Therefore, we describe a new relativistic density<br />
functional theory scheme to description of the atomic<br />
autoionization in an external dc electric and laser<br />
field, starting from the S-matrix Gell-Mann and Low<br />
formalism and the relativistic perturbation theory. The<br />
important new element is in the using the generalized<br />
Dirac-Kohn-Sham procedure for generation of the<br />
wave functions basis’s, which is based on the condition<br />
that the calibration-non-invariant contribution of the<br />
second order polarization diagrams to the imaginary<br />
part of the multi-electron system energy is minimized<br />
already at the first non-disappearing approximation<br />
of the relativistic perturbation theory [11]. Besides,<br />
we use the correct relativistic exchange-correlation<br />
functionals that hitherto has not done in any paper.<br />
Application of the new scheme to studying the autoionization<br />
phenomena in the heavy atomic systems is<br />
now in progress.<br />
References<br />
1. Aglitsky E.V., Ñàôðîíîâà Ó.è. Ñïåêòðîñêîïèÿ àòîìíûõ<br />
ñèñòåì. Ýíåðãîàòîèçäàò.: Ìîñêâà, 1999.<br />
2. Êàëåêøîâ Â.ô., Êóõàðåíêî Þ. A., Ôðàéäðàéõîâ Ñ.À.<br />
Ñïåêòðîñêîïèÿ è ýëåêòðîííàÿ äèôðàêöèÿ ïîâåðõíîñòåé.<br />
Íàóêà: Ìîñêâà, 1999.<br />
3. Ãëóøêîâ À.â., Ðåëÿòèâèñòñêàÿ êâàíòîâàÿ òåîðèÿ. Êâàíòîâàÿ<br />
ìåõàíèêà àòîìíîé ñèñòåìû.Îäåññà: Àñòðîïðèíò.<br />
2008. — Ñ 900.<br />
4. Amusia M.Ya. Atomic photoeffect. Acad.Press: N. — Y.,<br />
1998.<br />
5. Ëåòîõîâ V. Íåëèíåéíûå ôîòîïðîöåññû â àòîìàõ è ìîëåêóëàõ.<br />
Íàóêà: Ìîñêâà, 1999.<br />
6. Ivanova E.P., Ivanov L.N., Aglitsky E.V., Modern Trends<br />
in Spectroscopy of Multicharged Ions// Physics Rep. —<br />
1991. — Vol.166,N6. — P.315-390.<br />
7. Kohn W., Sham L.J. Quantum density oscillations in an<br />
inhomogeneous electron gas//Phys. Rev. A. — 1995. —<br />
Vol.137,N6. — P.1697–1706.<br />
8. Hohenberg P., Kohn W., Inhomogeneous electron gas //Phys.<br />
Rev.B. — 1999. — Vol.136,N2. — P.864-875.<br />
9. Gross E.G., Kohn W. Exchange-correlation functionals in<br />
density functional theory. — N-Y: Plenum, 2005. — 380P.<br />
10. Wilson S., The Fundamentals of Electron Density, Density<br />
Matrix and Density Functional Theory in Atoms, Molecules<br />
and the Solid State, Series: Progress in Theoretical Chemistry<br />
and Physics, Eds. Gidopoulos N.I. and Wilson S. — Amsterdam:<br />
Springer, 2004. — Vol.14. — 244P.<br />
11. Glushkov A.V., Ivanov L.N. Radiation Decay of Atomic<br />
States: atomic residue and gauge non-invariant contributions<br />
// Phys. Lett.A. — 1997. — Vol.170,N1. — P.33-37<br />
12. Ambrosov S.V., Glushkov A.V., Nikola L.V., Sensing the<br />
Auger spectra for solids: New quantum approach// Sensor<br />
Electr. and Microsyst. Techn. — 2006. — N3. — P.46-50.<br />
DENSITY FUNCTIONAL APPROACH TO ATOMIC AUTOIONIZATION IN AN EXTERNAL ELECTRIC FIELD: NEW<br />
RELATIVISTIC SCHEME<br />
Abstract<br />
Within the S-matrix Gell-Mann and Low formalism and the relativistic perturbation theory we present a new relativistic density<br />
functional theory scheme to description of the atomic autoionization in an external dc electric and laser field.<br />
Key words: autoionization, density functional theory, electric field.
ÓÄÊ 539.186<br />
Í. Â. Ìóäðàÿ<br />
ÌÅÒÎÄ ÔÓÍÊÖÈÎÍÀËÀ ÏËÎÒÍÎÑÒÈ Â ÎÏÈÑÀÍÈÈ ÀÒÎÌÍÎÉ ÀÂÒÎÈÎÍÈÇÀÖÈÈ ÂÎ ÂÍÅØÍÅÌ<br />
ÝËÅÊÒÐÈ×ÅÑÊÐÎÌ ÏÎËÅ: ÍÎÂÀß ÐÅËßÒÈÂÈÑÒÑÊÀß ÑÕÅÌÀ<br />
Ðåçþìå<br />
 ðàìêàõ S-ìàòðè÷íîãî ôîðìàëèçìà Ãåëë-Ìàíà è Ëîó è ðåëÿòèâèñòñêîé òåîðèè âîçìóùåíèé èçëîæåíà íîâàÿ ðåëÿòèâèñòñêàÿ<br />
ñõåìà ìåòîäà ôóíêöèîíàëà ïëîòíîñòè äëÿ îïèñàíèÿ õàðàêòåðèñòèê àòîìíîé àâòîèîíèçàöèè âî âíåøíåì ýëåêòðè-<br />
÷åñêîì è ëàçåðíîì ïîëå.<br />
Êëþ÷åâûå ñëîâà: àâòîèîíèçàöèÿ, òåîðèÿ ôóíêöèîíàëà ïëîòíîñòè, ýëåêòðè÷åñêîå ïîëå.<br />
ÓÄÊ 539.186<br />
Í. Â. Ìóäðà<br />
ÌÅÒÎÄ ÔÓÍÊÖ²ÎÍÀËÓ ÃÓÑÒÈÍÈ Â ÎÏÈÑÀÍͲ ÀÒÎÌÍί ÀÂÒβÎͲÇÀÖ²¯ Ó ÇÎÂͲØÍÜÎÌÓ<br />
ÅËÅÊÒÐÈ×ÍÎÌÓ ÏÎ˲: ÍÎÂÀ ÐÅËßÒȲÑÒÑÜÊÀ ÑÕÅÌÀ<br />
Ðåçþìå<br />
 ìåæàõ S- ìàòðè÷íîãî ôîðìàë³çìó Ãåëë-Ìàíà òà Ëîó ³ ðåëÿòèâ³ñòñüêî¿ òåî𳿠çáóðåíü âèêëàäåíà íîâà ðåëÿòèâ³ñòñüêà<br />
ñõåìà ìåòîäà ôóíêö³îíàëó ãóñòèíè äëÿ îïèñó õàðàêòåðèñòèê àòîìíî¿ àâòî³îí³çàö³¿ ó çîâí³øíüîìó åëåêòðè÷íîìó òà ëàçåðíîìó<br />
ïîë³.<br />
Êëþ÷îâ³ ñëîâà: àâòî³îí³çàö³ÿ, òåîð³ÿ ôóíêö³îíàëó ãóñòèíè, åëåêòðè÷íå ïîëå.<br />
51
52<br />
UDC 537.311.33:622.382.33<br />
V. A. SMYNTYNA, O. V. SVIRIDOVA<br />
I. I. Mechnikov National University of Odessa, Dvoryanskaya Str., 2, Odessa, 65026, Ukraine, e-mail: sviridova_olya@mail.ru<br />
INFLUENCE OF IMPURITIES AND DISLOCATIONS ON THE VALUE<br />
OF THRESHOLD STRESSES AND PLASTIC DEFORMATIONS IN SILICON<br />
The dependence of a plastic flow stress and deformation values on the presence of clear and<br />
precipitated by impurity initial structural defects in epitaxial p — silicon without foreign impurity and<br />
in epitaxial p — silicon with oxygen impurity is investigated. It is established, that, boron doping of<br />
silicon is the reason of threshold stress reduction in comparison with threshold stress for clear from<br />
defects silicon and leads to reduction of its hardness. Presence of oxygen atoms, precipitating dislocations<br />
in plates, stimulates the increase of threshold stress.<br />
1. INTRODUCTION<br />
Structural and impurity defects, their distribution<br />
in initial semiconductor plates at technological processing,<br />
can make decisive influence on the process of<br />
new defects’ generation that influences on degradation<br />
properties and on percentage yield of devices. In<br />
spite of the fact that it has been studied many years,<br />
the problem of defects in silicon remains actual till<br />
now. First of all, it is connected with the increase of<br />
electronic microcircuits integration, and with transition<br />
from micro technologies to nano technologies.<br />
Many works are devoted to studying recombination<br />
active defects arising in the course of silicon crystals<br />
cultivation [1]; defects in silicon nanowires [2]; problems<br />
of iron gettering in silicon by means of oxygen<br />
precipitates [3, 4]; optical attenuation on silicon divacancies<br />
[5]; controllable cultivation of dislocations<br />
[6]; redistributions of dislocations in silicon [7] and<br />
other defect properties.<br />
It is informed [8] that silicon crystals, containing<br />
appreciable quantity of oxygen atoms, are the best basis<br />
for integrated microcircuits creation, than clearer<br />
crystals. Despite of considerable amount of works on<br />
this problem [9, 10], a number of questions connected<br />
with impurity influence on the stress of the beginning<br />
of a plastic flow remain unsolved.<br />
The purpose of given work is the establishment of<br />
laws of stresses and relative deformations changes under<br />
the influence of structural and impurity variations<br />
of silicon plates.<br />
2. OBJECTS AND METHODS OF RESEARCH<br />
We studied epitaxial boron-doped silicon plates of<br />
grade BDS 10 (111) with the diameter of 60 mm and<br />
thickness of 405 microns.<br />
Following methods and equipment were used for<br />
studying defects on a silicon surface:<br />
– a method of selective chemical etching by Sirtle<br />
[11];<br />
– scanning electronic microscopy of a surface<br />
(SEMS), by means of scanning electronic microscope-analyzer<br />
“Cam Scan” — 4D with a system of<br />
the energetic dispersive analyzer “Link — 860” (with<br />
the usage of “Zaf” program, mass sensitivity of the de-<br />
vice is 0,01 %, beam diameter ranges from 5∙10 -9 to<br />
1∙10 -6 ) [12];<br />
– optical methods of researches with the usage of<br />
metallographic microscope “ÌÌÐ — 2Д;<br />
– Ozhe electronic spectroscopy (OES), by means<br />
of spectrometer LAS-3000, manufactured by “Riber”<br />
(with spatial resolution of 3 microns and energetic<br />
permission of analyzer of 0,3 %).<br />
Selective chemical etching was applied to samples<br />
before studying of their defects with the usage of metallographic<br />
microscope “ÌÌÐ — 2Д and of electronic<br />
microscope “Cam Scan”. For etching of plates with<br />
(111) -oriented surface plane Sirtle etchant was used.<br />
Its chemical compound is as following: 50 g of CrO 3<br />
+100 ml of H 2 O + 100 ml of HF (46 %). Etching time<br />
was from 2 till 15 minutes, etching speed was about<br />
2 — 3 microns/minute. Preliminary processing of<br />
plates in Caro and hydrogen-ammonia compositions<br />
[13] was made before selective etching. It allowed us to<br />
raise revealing properties of selective etchant. The revealed<br />
defects looked like dislocational etching poles,<br />
lines of dislocations or dislocational grids.<br />
3. RESEARCH AND CALCULATION<br />
PECULIARITIES<br />
It is established, that deformations of a rigid body<br />
arise under the influence of mechanical stresses. Stress<br />
τ dependence on deformation ε is presented on the<br />
graph of (fig. 1) [14].<br />
Crystal deformation occurs not only under the influence<br />
of external mechanical stresses. Crystal doping,<br />
presence of uncontrollable oxygen, carbon, hydrogen<br />
impurities and impurities of other elements in<br />
the course of cultivation always leads to the change of<br />
lattice constant, and, hence, to existence of areas with<br />
changed mechanical potential [15]. Elastic stresses in<br />
silicon lattice, caused by implantation of atoms of another<br />
size, are described by Poisson formula [11]:<br />
1+υ<br />
τ=ω⋅2μ⋅ ⋅C<br />
, (1)<br />
1−υ<br />
with ω — Vegard constant;<br />
υ= - S / S — Poisson constant (tab. 1); (2)<br />
12 11<br />
© V. A. Smyntyna, O. V. Sviridova, 2009
μ= 1/ S44<br />
— shear modulus (tab. 1); (3)<br />
С — impurity concentration;<br />
S — elastic compliance coefficient [14].<br />
mn<br />
Fig. 1. Typical dependence of τ= f () ε for covalent crystals:<br />
τ UFL , τ LFL , ε UFL , ε LFL — shear stresses τ and the deformations<br />
е , corresponding upper fluidity limit (UFL) and lower<br />
fluidity limit (LFL).<br />
In the given work the area of plastic flows (from<br />
( τUFL , ε UFL ) to ( τLFL, ε LFL ) ) on the curve of fig. 1 was<br />
investigated. Namely, we studied the value of residual<br />
stresses and deformations, arising in a crystal after<br />
cancellation of deformation and formation of structural<br />
defects. We were not interested in mechanical<br />
deformation of a crystal (stretching, compression,<br />
blow, bend, cave-in and other), but in initial (internal)<br />
deformation which possesses the crystal before exposure<br />
to external deformation. This initial deformation<br />
of a crystal is formed in the course of growth and subsequent<br />
doping. Stresses and deformations brought<br />
into a crystal by defects, formed in the course of its<br />
growth, were investigated in this work. Useful (doping)<br />
impurity as, for example, boron in p-silicon and<br />
phosphorus in n-silicon, as well as any other impurity,<br />
refers to the category of point defects.<br />
Table 1<br />
Results of required calculations of critical stresses and<br />
deformations, and also intermediate values for clear silicon.<br />
Lattice constant 0 , A a & [18] 5,431<br />
Vegard constant for boron ω boron [17]<br />
3<br />
2,8 10 −<br />
⋅<br />
Vegard constant for oxygen ω oxygen [17]<br />
4<br />
110 −<br />
⋅<br />
Elastic compliance<br />
2<br />
m<br />
coefficient,<br />
N [14]<br />
S11<br />
12<br />
2,14 10 −<br />
⋅<br />
S11<br />
12<br />
7,68 10 −<br />
S44<br />
⋅<br />
12<br />
12,6 10 −<br />
⋅<br />
Shear modulus м, 2<br />
N<br />
m (3)<br />
Poisson constant х (2)<br />
13<br />
7,94⋅10 0,28<br />
Atom concentration С, -3<br />
cm [18]<br />
22<br />
510 ⋅<br />
N<br />
Critical stress value τ UFL , [14] 2<br />
m<br />
8<br />
10<br />
Critical deformation value ε UFL (8)<br />
3<br />
1, 3 10 −<br />
⋅<br />
For silicon plates, containing dislocations with<br />
residual stresses around dislocation cores, the top of<br />
fluidity limit achievement comes at smaller stresses,<br />
than for crystals without dislocations. Studying a defect<br />
picture on a surface of silicon plates, it is possible<br />
to define value of these residual stresses and deformations,<br />
and, also, concentration of point defects.<br />
After selective chemical etching, two types of<br />
defect distribution picture in boron-doped epitaxial<br />
silicon plates were found out by means of SEMS.<br />
Thereupon the investigated plates were divided into 2<br />
groups. By methods of x-ray and OES analyses it was<br />
established, that plates of the second group contain<br />
oxygen impurity, and plates of the first group do not<br />
contain oxygen atoms. A typical representative of the<br />
first group of plates is the plate ¹ 1 (fig. 2) with big<br />
period of dislocational grid, consisting of 60 0 dislocations.<br />
A typical representative of the second group of<br />
plates is the plate ¹ 2 (fig. 3) with small period of dislocational<br />
grid, highly precipitated by oxygen atoms.<br />
SEMS analysis with a system of the energetic dispersive<br />
analyzer “Link — 860” showed that oxygen atoms<br />
are placed not only lengthways of dislocational grid,<br />
but also in its lattice sites (spheres on fig. 3).<br />
Dislocational grids considered to be repeating<br />
linear defects, density of which is expressed through<br />
a number of dislocational lines, crossing a surface of<br />
unit area, perpendicular to dislocational lines [16].<br />
After definition of dislocations’ amount n from<br />
the pictures of plates’ surface (tab. 2), we calculated<br />
surface density of dislocations, using formula<br />
N<br />
surf .<br />
n ⋅image<br />
increase<br />
= . (4)<br />
image square<br />
Fig. 2. Image of boron precipitated dislocational grid in psilicon,<br />
received after selective chemical etching by Sirtle (depth<br />
of analysis x = 5 mkm, image increase is 2300 times, (111) — orientated<br />
surface plane).<br />
Value of surface density of dislocations is represented<br />
in table 2.<br />
53
The surface density of dislocations is connected<br />
with value of relative deformation е (tab. 2) and with<br />
silicon lattice constant a by a well-known ratio<br />
0<br />
[11]:<br />
2<br />
ε= a0 ⋅ Nsurf<br />
. . (5)<br />
Relative deformation, described by Vegard law,<br />
arises at the process of impurity diffusion [17]:<br />
ε=ω⋅ С , (6)<br />
with ω — Vegard constant, С — impurity concentration<br />
in relative dimensionless units (tab. 2).<br />
In the case of several kinds of impurity resultant<br />
relative deformation is defined as the sum of contributions:<br />
ε= ∑ εi<br />
. (7)<br />
i<br />
Shear stress τ UFL and deformation еUFL in the isotropic<br />
structure are connected by ratio [14]<br />
with μ – shear modulus (tab. 1).<br />
54<br />
τ<br />
UFL<br />
ε UFL = , (8)<br />
μ<br />
Fig. 3. Image of oxygen precipitated dislocational grid in psilicon,<br />
received after selective chemical etching by Sirtle (depth<br />
of analysis x = 5 mkm, image increase is 2300 times, (111) — orientated<br />
surface plane).<br />
In the process of dislocation multiplication internal<br />
stresses become comparable with external and the<br />
real stresses effecting dislocation, will differ from the<br />
external. It is considered [14], that the real stresses effecting<br />
dislocation are equal to<br />
τ=τ eff . +τ inn.<br />
, (9)<br />
with τ eff . — external stresses, effecting a crystal,<br />
τ inn.<br />
— internal stresses.<br />
Long-range (internal) mechanical stress of uniformly<br />
distributed dislocations can be calculated, con-<br />
sidering additive character of internal stresses of each<br />
dislocation, according to expression<br />
1/2<br />
τ inn. =α⋅ b ⋅μ⋅Nsurf.<br />
, (10)<br />
Here<br />
α= 1 −υ/(2 π ) , (11)<br />
with х — Poisson constant (2), N surf . — dislocation<br />
density, b — Burgers’ vector magnitude.<br />
Table 2<br />
Results of stresses and relative deformations calculations, and also<br />
intermediate values for the investigated silicon plates<br />
Plates<br />
¹ 1<br />
Plates<br />
¹ 2<br />
Surface orientation plane (111) (111)<br />
Amount n of dislocations on figure,<br />
averaged for all plates of one type<br />
17 260<br />
Value of relative deformations in the<br />
area of dislocational grids е (5)<br />
6<br />
110 −<br />
⋅<br />
6<br />
210 −<br />
⋅<br />
Surface density of dislocations<br />
-2<br />
. , Nsurf cm (4)<br />
Stress value in the area of dislocational<br />
357 1320<br />
grids , 2<br />
N<br />
τ<br />
m<br />
(13)<br />
6<br />
9,3 ⋅10<br />
7<br />
1, 8 ⋅10<br />
Concentration of boron impurity<br />
С boron ,% (6)<br />
0,04 0,04<br />
Concentration of boron impurity<br />
-3<br />
Сboron, cm [18]<br />
19<br />
210 ⋅<br />
19<br />
210 ⋅<br />
Concentration of oxygen impurity<br />
С oxygen ,% (6)<br />
0,01<br />
Concentration of oxygen impurity,<br />
-3<br />
Сoxygen , cm [18]<br />
18<br />
510 ⋅<br />
As it is seen from fig. 2 and fig. 3, observable dislocational<br />
grids consist of 600 dislocations. The Burgers’<br />
vector magnitude of 600 dislocations equals to the lattice<br />
constant 0 a = b . Introducing the expressions of<br />
е and б , given by Eqs. (5) and (11) in Eq. (10) one<br />
gets:<br />
1−υ<br />
τ inn.<br />
= ⋅ε⋅μ . (12)<br />
2π<br />
As the surface picture was taken without any external<br />
stresses, for τ eff . one gives: τ eff . = 0 . Real stresses,<br />
effecting dislocation are equal to internal stress (tab.<br />
2) of uniformly distributed dislocations:<br />
τ=τ inn.<br />
(13)<br />
On the other hand, apparently from Eq. (10), the<br />
stress, effecting dislocation for lack of external stresses,<br />
applied to a crystal, is in direct ratio to a square<br />
root from dislocation density.<br />
4. RESULTS AND DISCUSSION<br />
Data, used in presented work and received from<br />
the analysis of pictures of a surface and as a result of<br />
calculations, averaged for each type of plates, are given<br />
in tables 1 and 2. Steams of stress values, received as a<br />
result of calculations, and deformations are tabulated<br />
for the analysis in table 3.
Table 3<br />
Values of residual stresses and deformations for plates ¹1 and<br />
plates ¹ 2, and also threshold values of stresses and deformations<br />
for clear from defects silicon.<br />
Plates ¹ 1<br />
Plates ¹ 2<br />
Threshold values for clear<br />
from defects silicon<br />
ε , 2<br />
N<br />
τ<br />
m<br />
6<br />
9,3 ⋅10<br />
7<br />
1, 8 ⋅10<br />
6<br />
110 −<br />
⋅<br />
6<br />
210 −<br />
⋅<br />
3<br />
1, 3 10 −<br />
⋅<br />
8<br />
10<br />
As it follows from table 3, residual stresses and deformations<br />
for a case of boron-doped silicon (plates<br />
¹ 1) on the curve of fig. 1 are placed more to the left<br />
of threshold stresses and deformations for clear from<br />
defects silicon that is they correspond to areas of elastic<br />
stresses and deformations. It means that in a result<br />
of boron precipitation of dislocations the new phase of<br />
Si-B was formed. Its threshold of plasticity is more to<br />
the left of a threshold of plasticity of clear silicon and,<br />
taking into account the formation of dislocations,<br />
more to the left of residual stresses and deformations<br />
for a plate ¹ 1. Thus, the received values of residual<br />
stresses and deformations for a plate ¹ 1 correspond<br />
not to silicon, but to Si-B. As the threshold of a plastic<br />
flow for this phase is less, than for silicon it is possible<br />
to draw a conclusion that in a result of boron doping<br />
hardness of silicon plates reduces.<br />
In a case of oxygen presence in boron-doped silicon<br />
plates, (plates ¹ 2) residual stresses and deformations<br />
are more to the left of threshold stresses and<br />
deformations for clear from defects silicon, but more<br />
to the right of residual stresses and deformations for a<br />
case of plates ¹ 1.<br />
It is experimentally proved, that value of ф UFL and,<br />
hence, of е UFL for germanium monotonously increases<br />
with the increase of oxygen concentration [14]. It<br />
is informed; that similar results were found out for<br />
silicon, only the mechanism of strengthening of these<br />
crystals in the presence of oxygen is not investigated<br />
up to the end [9].<br />
5. CONCLUSION<br />
The accessory of residual stresses and deformations’<br />
values of plates ¹ 1 to areas of elastic stresses<br />
and deformations for clear from defects silicon is<br />
caused by the fact, that boron atoms have smaller covalent<br />
radius (0,08 nanometers) [19] in comparison<br />
with covalent radius of silicon atoms (0,1175 nanometers)<br />
[19]. As a result the period of silicon crystal lattice<br />
increases when boron is placed in lattice sites. It is<br />
connected with the fact, that placing of boron atoms<br />
in silicon crystal lattice sites leads to lattice compression<br />
in the doped area, and, hence, to corresponding<br />
stretching of silicon crystal lattice [20]. As a result mechanical<br />
stresses leading to existence of stretching area<br />
appear in silicon crystal [21]. Therefore at the application<br />
of mechanical stresses, smaller, than threshold<br />
stresses for clear from defects silicon, the plastic flow<br />
is observed in plates ¹ 1. As residual stresses and deformations<br />
for plates ¹ 1 are less than threshold values<br />
for clear from defects silicon, it is obvious (fig. 1),<br />
that threshold values of stresses and deformations for<br />
p — silicon are placed in the area, more to the left of<br />
residual stresses and deformations.<br />
In the case of oxygen atoms in lattice sites of silicon<br />
crystal, because of larger covalent radius of oxygen<br />
atoms in comparison with covalent radius of silicon<br />
atoms, compression areas (caused by lattice stretching<br />
in the field of oxygen atom placement [22]) will be<br />
formed in silicon crystal. Interstitials (both boron and<br />
oxygen) in silicon crystal lattice form areas of compression<br />
[22]. Strength reduction of plates ¹ 1 justifies<br />
to primary placement of boron in silicon lattice<br />
sites. Position of residual stresses and deformations for<br />
plates ¹ 2 to the right of residual stresses and deformations<br />
for plates ¹ 1 (increase of deformation value)<br />
is caused by oxygen presence in plates ¹ 2. Concentration<br />
of oxygen atoms equal to 0,01 % in plates ¹<br />
2, is enough for compensation of hardness reduction<br />
effect, called by boron presence in the plates [23]. The<br />
presented results justify that controllable introduction<br />
of oxygen impurity can be used for increase of mechanical<br />
hardness of silicon crystals.<br />
References<br />
1. S. Diez, S. Rein, T. Roth, S. W. Glunz, Cobalt related defect<br />
levels in silicon analyzed by temperature- and injection-dependent<br />
lifetime spectroscopy, // J. Appl. Phys. 101, 033710<br />
(2007).<br />
2. R.P. Wang, Defects in silicon nanowires, // Appl. Phys. Lett.<br />
88, 142104 (2006).<br />
3. A. Haarahiltunen, H. Väinölä, O. Anttila, M. Yli-Koski and<br />
J. Sinkkonen, Experimental and theoretical study of heterogeneous<br />
iron precipitation in silicon, // J. Appl. Phys. 101,<br />
043507 (2007).<br />
4. A. Haarahiltunen, H. Väinölä, O. Anttila, E. Saarnilehto, M.<br />
Yli-Koski, J. Storgårds, and J. Sinkkonen, Modeling of heterogeneous<br />
precipitation of iron in silicon, // Appl. Phys. Lett.<br />
87, 151908 (2005).<br />
5. P. J. Foster, J. K. Doylend, P. Mascher, A. P. Knights, P. G.<br />
Coleman, Optical attenuation in defect-engineered silicon rib<br />
waveguides, // J. Appl. Phys. 99, 073101 (2006).<br />
6. M. Kammler, D. Chidambarrao, K. W. Schwarz, C. T. Black,<br />
and F. M. Ross, Controlled nucleation of dislocations by a<br />
spatially localized stress field, // Appl. Phys. Lett. 87, 133116<br />
(2005).<br />
7. A. M. Orlov, A. A. Solov`ev, A. A. Skvortsov, I. O. Yavtushenko,<br />
Redistribution of dislocations in silicon near stress concentrators,<br />
// Physics of the Solid State 47, pp. 2049 — 2054<br />
(2005).<br />
8. Á. Â. Ïåòóõîâ, Âëèÿíèå äèíàìè÷åñêîãî ñòàðåíèÿ äèñëîêàöèé<br />
íà äåôîðìàöèîííîå ïîâåäåíèå ïðèìåñíûõ ïîëóïðîâîäíèêîâ<br />
// ÔÒÏ 36 (2), ñòð. 129 — 133 (2002).<br />
9. Â. Ì. Áàáè÷, À. È. Áëåöêàí, Å. Ô. Âåíãåð, Êèñëîðîä â<br />
ìîíîêðèñòàëëàõ êðåìíèÿ. Èíòåðïðåñ ËÒÄ, Ê. (1997).<br />
240 ñ.<br />
10. Joao Justo, // Defects and Diffusion Forum. 200 — 202, pp.<br />
97 (2002).<br />
11. Ê. Ðåéâè, Äåôåêòû è ïðèìåñè â ïîëóïðîâîäíèêîâîì<br />
êðåìíèè. Ìèð, Ì. (1994). 471 ñ.<br />
12. Æ. Ôèëèáåð, Æ. Ôîíòåí, Ý. Âèêàðèî, Æ. Ýíîê, Ô. Ìîðèñ,<br />
Æ. Ðþñò, Ô. Áàñòåíåð, Ð. Òèêñüå è äð., Â ñá.: Ìèêðîàíàëèç<br />
è ðàñòðîâàÿ ýëåêòðîííàÿ ìèêðîñêîïèÿ / Ïîä ðåä.<br />
Ô.Ìîðèñ, Ë.Ìåíè, Ð. Òèêñüå. Ìåòàëëóðãèÿ, Ì. (1995).<br />
498 c.<br />
13. Î. À. Êóëèíè÷, À. À. Ëèñîâñêàÿ, Í. Í. Ñàäîâà, Î ïîâûøåíèè<br />
âûÿâëÿþùåé ñïîñîáíîñòè èçáèðàòåëüíîãî òðàâëåíèÿ<br />
ìîíîêðèñòàëëîâ êðåìíèÿ // ÓÔÆ 35 (11), ñòð.<br />
1691 — 1694 (1990).<br />
14. Þ. À. Êîíöåâîé, Þ. Ì. Ëèòâèíîâ, Ý. À. Ôàòòàõîâ, Ïëàñòè÷íîñòü<br />
è ïðî÷íîñòü ïîëóïðîâîäíèêîâûõ ìàòåðèàëîâ è<br />
ñòðóêòóð. Ðàäèî è ñâÿçü, Ì. (1992). 240 ñ.<br />
15. Â. À. Ñìûíòûíà, Î. À. Êóëèíè÷, Ì. À. Ãëàóáåðìàí, Ý. Ò. Ðîãîâñêàÿ,<br />
Ã. Ã. ×åìåðåñþê, È. Ð. ßöóíñêèé, Î. Â. Ñâèðèäîâà,<br />
Âëèÿíèå äåôåêòîâ íà ðàñïðåäåëåíèå êîíöåíòðàöèè ëå-<br />
55
ãèðóþùåé ïðèìåñè ïî ïëàñòèíå ìîíîêðèñòàëëè÷åñêîãî<br />
êðåìíèÿ // Ñåíñîð, 21 (3), ñòð. 19 — 23 (2006).<br />
16. Ã. Ìàòàðå, Ýëåêòðîíèêà äåôåêòîâ â ïîëóïðîâîäíèêàõ.<br />
Ìèð, Ì. (1974). 464 ñ.<br />
17. Ë. Í. Àëåêñàíäðîâ, Ïåðåõîäíûå îáëàñòè ýïèòàêñèàëüíûõ<br />
ïëåíîê â ïîëóïðîâîäíèêàõ. Íàóêà, Íîâîñèáèðñê. (1978).<br />
272 ñ.<br />
18. Ï. È. Áàðàíñêèé, Â. Ï. Êëî÷êîâ, Ì. Â. Ïîòûêåâè÷, Ïîëóïðîâîäíèêîâàÿ<br />
ýëåêòðîíèêà. Ñïðàâî÷íèê. Íàóêîâà<br />
äóìêà, Êèåâ. (1995). 704 ñ.<br />
19. Â. È. Ïëåáàíîâè÷, À. È. Áåëîóñ, À. Ð. ×åëÿäèíñêèé, Â. Á. Îäæàåâ,<br />
Ñîçäàíèå áåçäèñëîêàöèîííûõ èîííî-ëåãèðîâàííûõ<br />
ñëîåâ êðåìíèÿ // ÔÒÒ 50 (8), ñòð. 1378 — 1382 (2008).<br />
20. È. Í. Ñìèðíîâ, Èçìåíåíèÿ ïåðèîäà êðèñòàëëè÷åñêîé ðå-<br />
56<br />
UDC 537.311.33:622.382.33<br />
V. A. Smyntyna, O. V. Sviridova<br />
øåòêè êðåìíèÿ, âûçûâàåìûå äèôôóçèåé áîðà, ìûøüÿêà<br />
è ñóðüìû // Äîêëàäû Àêàäåìèè íàóê ÑÑÑÐ. Òåõíè÷åñêàÿ<br />
ôèçèêà 221 (2), ñòð. 332 — 334 (1975).<br />
21. È. Í. Ñìèðíîâ, È. È. Ïåòðîâ, Ò. Ô. Ãîðÿ÷åâà, Èññëåäîâàíèå<br />
äèôôóçèè áîðà â êðåìíèé ðåíòãåíîâñêèìè äèôðàêöèîííûìè<br />
ìåòîäàìè // Ýëåêòðîííàÿ òåõíèêà. Ñåðèÿ 2<br />
(Ïîëóïðîâîäíèêîâûå ïðèáîðû) 97 (5), ñòð. 11 — 18<br />
(1995).<br />
22. È. Í. Ñìèðíîâ, Äåôîðìàöèÿ êðèñòàëëè÷åñêîé ðåøåòêè<br />
êðåìíèÿ, âûçûâàåìàÿ áîìáàðäèðîâêîé èîíàìè áîðà è<br />
êèñëîðîäà // Äîêëàäû Àêàäåìèè íàóê ÑÑÑÐ. Ôèçè÷åñêàÿ<br />
õèìèÿ 225 (3), ñòð. 621 — 623 (1995).<br />
23. Î. Â. Ñâ³ð³äîâà, Çì³öíþþ÷èé âïëèâ äîì³øêè êèñíþ íà<br />
êðèñòàëè êðåìí³þ // ÅÂÐÈÊÀ — 2008, Â22 (2008).<br />
INFLUENCE OF IMPURITIES AND DISLOCATIONS ON THE VALUE OF THRESHOLD STRESSES AND PLASTIC<br />
DEFORMATIONS IN SILICON<br />
Abstract<br />
The dependence of a plastic flow stress and deformation values on the presence of clear and precipitated by impurity initial structural<br />
defects in epitaxial p — silicon without foreign impurity and in epitaxial p — silicon with oxygen impurity is investigated. It is<br />
established, that, boron doping of silicon is the reason of threshold stress reduction in comparison with threshold stress for clear from<br />
defects silicon and leads to reduction of its hardness. Presence of oxygen, precipitating dislocations in plates, stimulates the increase of<br />
threshold stress.<br />
Key words: dislocations, threshold stresses, plastic deformations.<br />
ÓÄÊ 537.311.33:622.382.33<br />
Â. À. Ñìûíòûíà, Î. Â. Ñâèðèäîâà<br />
ÂËÈßÍÈÅ ÏÐÈÌÅÑÅÉ È ÄÈÑËÎÊÀÖÈÉ ÍÀ ÂÅËÈ×ÈÍÓ ÏÎÐÎÃÎÂÛÕ ÍÀÏÐßÆÅÍÈÉ È ÏËÀÑÒÈ×ÅÑÊÈÕ<br />
ÄÅÔÎÐÌÀÖÈÉ Â ÊÐÅÌÍÈÈ<br />
Ðåçþìå<br />
Èññëåäîâàíà çàâèñèìîñòü âåëè÷èíû íàïðÿæåíèé è äåôîðìàöèé íà÷àëà ïëàñòè÷åñêîãî òå÷åíèÿ îò ïðèñóòñòâèÿ ÷èñòûõ<br />
è ïðåöèïèòèðîâàííûõ ïðèìåñÿìè èñõîäíûõ ñòðóêòóðíûõ äåôåêòîâ â ýïèòàêñèàëüíîì p- êðåìíèè áåç ñòîðîííèõ ïðèìåñåé<br />
è â ýïèòàêñèàëüíîì p- êðåìíèè ñ ïðèìåñüþ êèñëîðîäà. Óñòàíîâëåíî, ÷òî, ëåãèðîâàíèå êðåìíèÿ áîðîì ÿâëÿåòñÿ ïðè÷èíîé<br />
óìåíüøåíèÿ ïîðîãîâûõ íàïðÿæåíèé ïî ñðàâíåíèþ ñ ïîðîãîâûìè íàïðÿæåíèÿìè äëÿ ÷èñòîãî îò äåôåêòîâ êðåìíèÿ è ïðèâîäèò<br />
ê óìåíüøåíèþ åãî ïðî÷íîñòè. Ïðèñóòñòâèå â ïëàñòèíàõ êèñëîðîäà, ïðåöèïèòèðóþùåãî äèñëîêàöèè, ñòèìóëèðóåò âîçðàñòàíèå<br />
ïîðîãîâûõ íàïðÿæåíèé.<br />
Êëþ÷åâûå ñëîâà: äèñëîêàöèè, ïîðîãîâîå íàïðÿæåíèå, ïëàñòè÷åñêèå äåôîðìàöèè.<br />
ÓÄÊ 537.311.33:622.382.33<br />
Â. À. Ñìèíòèíà, Î. Â. Ñâ³ð³äîâà<br />
ÂÏËÈ ÄÎ̲ØÎÊ ² ÄÈÑËÎÊÀÖ²É ÍÀ ÂÅËÈ×ÈÍÓ ÏÎÐÎÃÎÂί ÍÀÏÐÓÃÈ ² ÏËÀÑÒ×Íί ÄÅÔÎÐÌÀÖ²¯<br />
 ÊÐÅÌͲ¯<br />
Ðåçþìå<br />
Äîñë³äæåíî çàëåæí³ñòü âåëè÷èíè íàïðóãè ³ äåôîðìàö³¿ ïî÷àòêó ïëàñòè÷íî¿ òå÷³¿ â³ä íàÿâíîñò³ ÷èñòèõ ³ ïðåöèï³òîâàíèõ<br />
äîì³øêàìè ïî÷àòêîâèõ ñòðóêòóðíèõ äåôåêò³â â åï³òàêñ³àëüíîìó p- êðåìí³¿ áåç ñòîðîíí³õ äîì³øîê ³ â åï³òàêñ³àëüíîìó<br />
p- êðåìí³¿ ç äîì³øêîþ êèñíþ. Âñòàíîâëåíî, ùî, ëåãóâàííÿ êðåìí³þ áîðîì º ïðè÷èíîþ çìåíøåííÿ ïîðîãîâî¿ íàïðóãè â<br />
ïîð³âíÿíí³ ³ç ïîðîãîâîþ íàïðóãîþ äëÿ ÷èñòîãî â³ä äîì³øîê êðåìí³þ ³ ïðèçâîäèòü äî çíèæåííÿ éîãî ì³öíîñò³. Ïðèñóòí³ñòü â<br />
ïëàñòèíàõ êèñíþ, ïðåöèï³òóþ÷îãî äèñëîêàö³¿, ñòèìóëþº çðîñòàííÿ ïîðîãîâî¿ íàïðóãè.<br />
Êëþ÷îâ³ ñëîâà: äèñëîêàö³¿, ïîðîãîâà íàïðóãà, ïëàñòè÷í³ äåôîðìàö³¿.
UDÑ 530.145; 539.1;539.18<br />
O. YU. KHETSELIUS<br />
I. I. Mechnikov Odessa National University<br />
ADVANCED MULTICONFIGURATION MODEL OF DECAY<br />
OF THE MULTIPOLE GIANT RESONANCES IN THE NUCLEI<br />
1. INTRODUCTION<br />
As it is well known, the multipole giant resonances<br />
are the highly excited states of nuclei, which are interpretive<br />
as the collective coherent vibrations with a<br />
participance of large number of nucleons [1-8]. Experimentally,<br />
the multipole giant resonances are manifested<br />
as the wide maximums in the dependence of<br />
cross-section of the nuclear reactions on the incident<br />
particle energy r in the spectrum of incident particles.<br />
A classification of the multipole giant resonances as<br />
the states of collective type is usually fulfilled on the<br />
quantum numbers of vibration excitations: entire angle<br />
momentum (J) and parity π (J π ). The multipole<br />
giant resonances are observed in the spectra of majority<br />
of nuclei and situated ,as a rule, in the continuous<br />
spectrum of excitations in a nucleus (with width of order<br />
of several MeV). Two main theoretical approaches<br />
to a description of the multipole giant resonances are<br />
usually used [1-5]. In the phenomenological theories<br />
it is supposed that the strong collectivization of states<br />
allows to apply the hydrodynamic models to the description<br />
of vibrations of the nuclear form and volume.<br />
The microscopic theory is in fact based on the shell<br />
model of a nucleus. It is well known different versions<br />
of the quasiparticle-phonon model of a nucleus, designed<br />
for describing littlie-quasiparticle components<br />
of the wave functions for low, intermediate and high<br />
excitation energies (see [2,3,6]). In the simple interpretation<br />
an excitation of the multipole giant resonances<br />
is the result of transition of the nucleons from<br />
one closed shell to another one, i.e. the multipole giant<br />
resonances is the result of the coherent summation<br />
of many particle-hole (p-h) transitions with the necessary<br />
corresponding momentum and parity.<br />
As a rule, the multipole giant resonances are situated<br />
under the excitation energies, which exceed the<br />
thresholds of emission of the particles from a nucleus.<br />
Studying the multipole giant resonances decay channels<br />
allows to reveal the mechanisms of its forming, connection<br />
with other excitations etc. The interaction of a<br />
nucleus with external field with forming the multipole<br />
giant resonances occurs during several stages. There is<br />
a production of the p-h excitation which is corresponding<br />
to the 1p-1h states over the Fermi surface (the first<br />
stage). Then the excited pair interacts with nuclear<br />
nucleons with the creating another 1p-1h excited state<br />
or two p-h pairs ( 2p-2h state; second stage). Then the<br />
© O. Yu. Khetselius, 2009<br />
It is presented an advanced generalized multiconfiguration approach to describe a decay of highexcited<br />
states (the multipole giant resonances), which is based on the mutual using the shell models<br />
(with extended basis) and microscopic model of pre-equilibrium decay with statistical account for<br />
complex configurations 2p2h, 3p3h etc. The new model is applied to an analysis of the reaction (μ - n)<br />
on the nucleus 40 Ca.<br />
3p-3h and more complicated states are created till the<br />
statistical equilibrium takes a place. The full width of<br />
the multipole giant resonances is provided by the direct<br />
decay to continuum (Γ ↑ ) and decay of the 1p-1h<br />
configurations on more complicated multi-particle<br />
(Γ ↓ ) ones. The mixing with complex configurations<br />
leads to the loss of the coherence and creating states<br />
of the compound nucleus. It’s known that an account<br />
of complex configurations has significant meaning for<br />
adequate explanation of the widths, structure and decay<br />
properties of the multipole giant resonances (see<br />
[2-5]). Here we present generalized multiconfiguration<br />
model to describe a decay of high-excited states, which<br />
is based on the mutual using the shell models (with<br />
limited basis) and microscopic Zhivopistsev-Slivnov<br />
model [5] of the pre-equilibrium decay with statistical<br />
account for complex 2p2h, 3p3h configurations etc.<br />
The model is applied to analysis of reaction (μ - n) on<br />
the nucleus 40 Ca. The comparison with experimental<br />
and other theoretical data is presented.<br />
2. GENERALIZED MULTICONFIGURATION<br />
MODEL OF THE MULTIPOLE GIANT<br />
RESONANCES DECAY<br />
The multipole giant resonances are treated on<br />
the basis of the multiparticle shell model. Process of<br />
creation of the collective state (of the multipole giant<br />
resonance) and an emission process of nucleons are<br />
described by the diagram in fig.1.<br />
Here V is effective Hamiltonian of interaction,<br />
μ<br />
resulted in capture of muon by nucleus with transformation<br />
of proton to neutron and emission by antineutrino.<br />
Isobaric analogs of isospin and spin-isospin<br />
resonances of finite nucleus are excited. The diagrams<br />
n<br />
for photonuclear reactions look to be analogous; Γ 22<br />
%<br />
is the full vertex part (full amplitude of interaction,<br />
which transfers the interacting p-h pair to the finite<br />
n<br />
npnh state. The full vertex Γ Γ 22<br />
% is defined by the<br />
system of equations within quantum Green function<br />
modified approach [3,5].<br />
All possible configurations are divided on two<br />
groups: i). group of complicated configurations “n ”, 1<br />
which must be considered within shell model with<br />
account for residual interaction; ii). statistical group<br />
“n ” of complex configurations with large state den-<br />
2<br />
sity p(n,E)>> and strong overlapping the states<br />
57
G n >>D n-1 >D n (D n is an averaged distance between<br />
states with 2n exciton; G n is an averaged width). Matrix<br />
elements of bond are small and characterized<br />
by a little dispersion. To take into account a<br />
collectivity of separated complex configurations for<br />
input state a diagonalization of residual interaction on<br />
the increased basis (ph,ph+phonon, ph+2 phonon) is<br />
used. All complex configurations are considered within<br />
the pre-equilibrium decay model by Feschbach-<br />
Zhivopistsev et al [5] with additional account of “n 1 ”<br />
group configurations. The input wave functions of the<br />
multipole giant resonances for nuclei with closed or<br />
almost closed shells are found from diagonalization of<br />
residual interaction on the effective 1p1h basis [9-15].<br />
58<br />
μ<br />
V<br />
~<br />
ν<br />
n<br />
p<br />
ϕ<br />
~<br />
Γ<br />
22n<br />
Fig. 1. Diagram of process for production of the collective<br />
state (multipole giant resonances) and emission of nucleons (or<br />
more complex particles).<br />
Statistical multistep negative muon capture<br />
through scalar intermediate states of compound nucleus<br />
is important. Intensities of nucleon spectra can<br />
be written by standard way [1,2]. In particular, an intensity<br />
of nucleonic spectra is defined as follows:<br />
dI<br />
( Eμ, l, εf, Jπ<br />
) =<br />
dε<br />
f<br />
↑<br />
(, , ) n 1 ↓<br />
Γn l εf Jπ ⎡ − Γk( J π)<br />
⎤<br />
= ⋅⎢ ⎥⋅Λμ(<br />
Eμ, Jπ)<br />
n= 1, Γn( Jπ) ⎢⎣ k = 1 Γk( Jπ)<br />
⎥⎦<br />
Δ n=<br />
1<br />
where<br />
∑ ∏ (1)<br />
•<br />
•<br />
•<br />
•<br />
Γ (, l ε , Jπ ) = 2 π⋅ρ( , ε ) ρ ( , ,<br />
↓<br />
2<br />
Γk( Jπ ) = 2 π⋅<br />
k+<br />
1<br />
( b)<br />
>ρ ( Nk+ 1,<br />
Jπ , Eμ)<br />
Eμ =ε f + UB + BN<br />
Here l is the orbital moment of the emission nucleon,<br />
ε is its energy; B is the bond energy of nucleon<br />
f N<br />
in the compound nucleus; Λμ( Eμ, Jπ<br />
) is probability of<br />
μ-capture with excitation of the state φ (E , Jπ) with<br />
in μ<br />
energy E , spin J and parity π. In oppositeness to stan-<br />
μ<br />
dard theories [2,5], we take into account an interference<br />
between contributions of separated “dangerous”<br />
configurations. From the other side, above indicated<br />
features of the statistical group of configurations arte<br />
not fulfilled for the “dangerous” configurations. However,<br />
the value n ( 1)<br />
n<br />
↓<br />
Γ for some dangerous configura-<br />
2<br />
tion is weakly dependent upon the energy. Indeed,<br />
configuration n 1 is the superposition of the large number<br />
of configurations, i.e. [2]:<br />
2<br />
| < n1 | In1, n+<br />
1 | n+<br />
1 > |<br />
↓<br />
Γ n ( n1<br />
) = ∑<br />
2 2<br />
n+ 1 ( Eμ − En+<br />
1) +Γn+<br />
1/4<br />
Generally, the expressions for the n-step contribution<br />
to the emission spectrum are modified as follows:<br />
dI Γ (, l ε, Jπ) Γ ( Jπ)<br />
E l ε Jπ<br />
= ⋅ +<br />
dε Γ ( Jπ) Γ ( Jπ)<br />
( μ , , f , ) (<br />
↑<br />
n2 n−1, n2<br />
f n2n−1 ∑<br />
{ n1 }<br />
↑<br />
n1 l<br />
n1 f<br />
J<br />
J<br />
↓<br />
n−1, n J<br />
1<br />
n−1J n−<br />
2<br />
∏<br />
k = 1<br />
↓<br />
k<br />
k<br />
J<br />
J<br />
+<br />
Γ (, ε<br />
Γ (<br />
, π) Γ<br />
⋅<br />
π) Γ<br />
( π) ⎡<br />
⋅ ⎢<br />
( π) ⎢⎣ Γ (<br />
Γ (<br />
π)<br />
⎤<br />
⎥×<br />
π)<br />
⎥⎦<br />
×Λ% ( E , Jπ)<br />
where<br />
∑<br />
μ μ<br />
Γ = Γ +Γ<br />
↓ ↓ ↓<br />
n−1 { n1<br />
}<br />
n−1, n1 n−1, n2<br />
| | Γ<br />
Γ =<br />
↓<br />
Nn−1 Nn−1, Nn ( n1<br />
)<br />
Nn 2<br />
n1<br />
n−1, n1<br />
2 2<br />
( Eμ− En<br />
) +Γ /4<br />
1 n1<br />
(2)<br />
Supposing the input state is isolated, in formalism<br />
of the input ph-states one could write as follows:<br />
Γ1( Jπ) Λμ( ϕin ( E , ))<br />
i i Jπ<br />
Λ %<br />
μ =<br />
2 2<br />
( Eμ− Ei)<br />
+Γn<br />
/4<br />
1<br />
where<br />
∑<br />
Γ ( ϕ ) =Γ +Γ + Γ<br />
↑ ↓ ↓<br />
1 in 1 1, n2 { n1<br />
}<br />
1, n1<br />
The other technical details of the presented approach<br />
can be found in refs. [2,4,5, 15,16,18].<br />
3. RESULTS AND CONCLUSION<br />
The wave functions of the input state {φ in } in the<br />
reaction 40 Ca (μ - n) are calculated within the shell<br />
model [12,15,18]. As one could wait for that a col-<br />
lectivity of initial input state leads to significant de-<br />
↓ creasing Γ . The separation into groups n1 and n is<br />
i<br />
2<br />
naturally accounted for the 2p2h configuration space<br />
[2,18] and the contribution of configurations “ph +<br />
phonon” and weakly correlated 2p2h states is revealed<br />
[4,5,16]. A probability of transition to the “dangerous”<br />
configurations 2p2h is defined by the value of<br />
matrix element:<br />
2<br />
| |<br />
and additionally by density ρ(2p2h,Jπ,E) for statistical<br />
group n . The contribution of weakly correlated 2p2h<br />
2<br />
configurations is defined by the following expression:<br />
↓<br />
2<br />
Γ 2p2h= 2 π⋅ | >ρ 2p2h The residual interaction has been chosen in the<br />
form of Soper forces (see [5]):<br />
V=g (1-α+α⋅σ σ )⋅Δ(r -r ),<br />
0 1 2 1 2<br />
3 where g /(4πr )=-3 MeV, α=0,135.The phonons have<br />
0 o<br />
been considered in the collective model and calculation<br />
parameters in the collective model and generalized
andom phase approximation are chosen according to<br />
ref.[4,5]. The phonons contribution is distributed as<br />
follows: 2 + (E=3,9 MeV; β=0,075)~ 42%, 3 - (E=3,736<br />
MeV; β=0,345)~8%, 5 - (E=4,491 MeV; β=0,216)~3%<br />
etc. with growth of the phonon moment.<br />
Our theoretical results are compared with experimental<br />
data and other calculation results [2] in fig.2,3.<br />
In the range of 5-13MeV the experiment gives the intensity<br />
~10% from the equilibrium one. As it has been<br />
shown earlier (c.f.[4,5], the 1 - , 2 - states do not give the<br />
significant contribution. However, these states exhaust<br />
~80% of the intensity of the μ - -capture. This fact is<br />
completely corresponding to results [16] and independently<br />
to the data from ref. [5] and ref. [19].<br />
The analysis shows also that only an accurate mu-<br />
tual account for 0 ± ,1 +<br />
, 2+ ,3 + and more high multipoles<br />
(plus more less correct microscopic calculation of<br />
↓<br />
Γin ( Jπ , E)<br />
, the input and 2p2h states, separation of<br />
the 2p2h space n configurations n and n etc.) allows<br />
1 2<br />
to fill the range of high and middle part of the spectrum.<br />
Preliminary estimates show that an agreement<br />
between theoretical and experimental data is more<br />
improved in this case, especially, in the high energy<br />
part of the spectrum.<br />
1,0<br />
0,1<br />
0,0<br />
0 3 6 9 12, MeV<br />
Fig. 2. The comparison of the calculated spectra (curve 1)<br />
with experimental data (curve 3) [20] and theoretical data from<br />
the Zhivopistsev-Slivnov model (curve 2) [5].<br />
0,1<br />
0,0<br />
0 3 6 9 12, MeV<br />
± ±<br />
Fig. 3. The mutual account of the 0 ± , 1 + , 2 + , 3 ± , 4 - , 5 - — multipoles:<br />
the curve 1 — the present paper; the curve 1 is corresponding<br />
to the pre-equilibrium and direct part of the spectrum and the<br />
curve 3 is corresponding to the equilibrium part (see text).<br />
References<br />
1. Bohr O., Mottelsson B., Structure of atomic nucleus. — N-Y.:<br />
Plenum, 1995. — 450P.<br />
2. Ñîëîâü¸â Â.Ã. Òåîðèÿ àòîìíîãî ÿäðà. Êâàçè÷àñòèöû è<br />
ôîíîíû. — Ìîñêâà:, Ýíåðãîèçäàò 1999. — 300P.<br />
3. Izenberg I., Grainer B., Models of nuclei. Collective and onebody<br />
phenomena. — N-Y. :Plenum Press, 2005. — 360P.<br />
4. Paar N., Vretenar D., Ring P., Neutrino-nuclei reactions with<br />
relativistic quasiparticle RPA// J. Phys. G. Nucl. and Particle<br />
Phys. — 2008. — Vol.35. — P.014058.<br />
5. Zhivopistsev F.A., Slivnov A.M., multiconfiguration model<br />
of decay of the multipole giant resonances // Izv.AN Ser.<br />
Phys. — 1994. — Vol.48. — P.821-825.<br />
6. Serot B.D., Walecka J.D., Advances in Nuclear Physics: The<br />
Relativistic Nuclear Many Body Problem. — N. — Y.: Plenum<br />
Press, 1999. — Vol.16.<br />
7. Tsoneva N., Lenske H., Low energy dipole excitations in<br />
nuclei at the N=50,82 and Z=50 shell closures as signatures<br />
for a neutron skin// J. Phys. G. Nucl. and Particle Phys. —<br />
2008. — Vol.35. — P.014047.<br />
8. Glushkov A.V., Malinovskaya S.V., Cooperative laser-nuclear<br />
processes: border lines effects// In: New projects and<br />
new lines of research in nuclear physics. Eds. G.Fazio and<br />
F.Hanappe, Singapore : World Scientific. — 2003. — P.241-<br />
250.<br />
9. Nagasawa T., Haga A., Nakano M., Hyperfine splitting of<br />
hydrogenlike atoms based on relativistic mean field theory//<br />
Phys.Rev.C. — 2004. — Vol.69. — P.034322.<br />
10. Benczer-Koller N., The role of magnetic moments in the determination<br />
of nuclear wave functions of short-lived excited<br />
states// J.Phys.CS. — 2005. — Vol.20. — P.51-58.<br />
11. Tomaselli M., Schneider S.M., Kankeleit E., Kuhl T., Ground<br />
state magnetization of 209 Bi in a dynamic-correlation model//<br />
Phys.Rev.C. — 1999. — Vol.51, N6. — P.2989-2997.<br />
12. Dikmen E., Novoselsky A., Vallieres M., Shell model calculation<br />
of low-lying states of 110 Sb// J.Phys.G.: Nucl.Part.<br />
Phys. — 2007. — Vol.34. — P.529-535.<br />
13. Stoitsov M., Cescato M.L., Ring P., Sharma M.M., Nuclear<br />
breathing mode in the relativistic mean-field theory//J. Phys.<br />
G: Nucl. Part. Phys. — 1994. — Vol.20. — P.L149-L156.<br />
14. Khetselius O.Yu., Hyperfine structure of atomic spectra. —<br />
Odessa: Astroprint, 2008. — 210P.<br />
15. Khetselius O.Yu., Relativistic Calculating the Hyperfine<br />
Structure Parameters for Heavy-Elements and Laser Detecting<br />
the Isotopes and Nuclear Reaction Products//Physica<br />
Scripta. — 2009. — Vol.62. — P.71-76.<br />
16. Khetselius O.Yu. et al, Generalized multiconfiguration model<br />
of decay of the multipole giant resonances applied to analysis<br />
of reaction (μ - n) on the nucleus 40 Ca//Trans. of SLAC<br />
(MENU, Stanford). — 2008. — Vol.1. — P.186-192.<br />
17. Khetselius O.Yu., Relativistic Calculating the Spectral Lines<br />
Hyperfine Structure Parameters for Heavy Ions // Spectral<br />
Line Shapes (AIP). — 2008. — Vol. 15. — P.363-365.<br />
18. Khetselius O.Yu., Turin A.V., Sukharev D.E., Florko T.A.,<br />
Estimating of X-ray spectra for kaonic atoms as tool for sensing<br />
the nuclear structure// Sensor Electr. and Microsyst.<br />
Techn. — 2009. — N1. — P.P.11-16.<br />
19. Glushkov A.V., Malinovskaya S.V., Quantum theory of the<br />
cooperative muon-nuclear processes: Discharge of metastable<br />
nuclei during negative muon capture// Recent Advances<br />
in Theory of Phys. and Chem. Systems (Springer). — 2006. —<br />
Vol.15. — P.301-328.<br />
20. Waitkovskaya I. et al, Analysis of reaction (μ - n) on the nucleus<br />
40 Ca//Nucl.Phys. — 1999. — Vol.15. — P.2154-2158.<br />
59
60<br />
UDÑ 530.145; 539.1;539.18<br />
O. Yu. Khetselius<br />
ADVANCED MULTICONFIGURATION MODEL OF DECAY OF THE MULTIPOLE GIANT RESONANCES IN THE<br />
NUCLEI<br />
Abstract<br />
It is presented an advanced generalized multiconfiguration approach to describe a decay of high-excited states (the multipole giant<br />
resonances), which is based on the mutual using the shell models (with extended basis) and microscopic model of pre-equilibrium decay<br />
with statistical account for complex configurations 2p2h, 3p3h etc. The new model is applied to an analysis of the reaction (μ - n) on the<br />
nucleus 40 Ca.<br />
Key words: multipole giant resonances, generalized multiconfiguration model, reaction (μ - n) on the nucleus 40 Ca.<br />
ÓÄÊ 530.145; 539.1;539.18<br />
O. Þ. Õåöåëèóñ<br />
ÓÑÎÂÅÐØÅÍÑÒÂÎÂÀÍÍÀß ÌÍÎÃÎÊÎÍÔÈÃÓÐÀÖÈÎÍÍÀß ÌÎÄÅËÜ ÐÀÑÏÀÄÀ ÌÓËÜÒÈÏÎËÜÍÛÕ<br />
ÃÈÃÀÍÒÑÊÈÕ ÐÅÇÎÍÀÍÑΠ ßÄÐÀÕ<br />
Ðåçþìå<br />
Ðàçðàáîòàí óñîâåðøåíñòâîâàííûé îáîáùåííûé ìíîãîêîíôèãóðàöèîííûé ïîäõîä äëÿ îïèñàíèÿ ðàñïàäà âûñîêî âîçáóæäåííûõ<br />
ñîñòîÿíèé (ìóëüòèïîëüíûå ãèãàíòñêèå ðåçîíàíñû) ÿäåð, êîòîðûé áàçèðóåòñÿ íà îäíîâðåìåííîì èñïîëüçîâàíèè<br />
îáîëî÷å÷íîé ìîäåëè (ñ ðàñøèðåííûì áàçèñîì) è ìèêðîñêîïè÷åñêîé ìîäåëè ïðåäðàâíîâåñíîãî ðàñïàäà ñî ñòàòèñòè÷åñêèì<br />
ó÷åòîì ñëîæíûõ êîíôèãóðàöèé òèïà 2p2h, 3p3h è äðóãèõ. Íîâûé ïîäõîä èñïîëüçîâàí äëÿ àíàëèçà ðåàêöèè (μ - n) íà ÿäðå<br />
40 Ca.<br />
Êëþ÷åâûå ñëîâà: ìóëüòèïîëüíûå ãèãàíòñêèå ðåçîíàíñû, îáîáùåííàÿ ìíîãî-êîíôèãóðàöèîííàÿ ìîäåëü, ðåàêöèÿ (μ - n)<br />
íà ÿäðå 40 Ca.<br />
ÓÄÊ 530.145; 539.1;539.18<br />
O. Þ. Õåöåë³óñ<br />
ÓÄÎÑÊÎÍÀËÅÍÀ ÁÀÃÀÒÎÊÎÍÔ²ÃÓÐÀÖ²ÉÍÀ ÌÎÄÅËÜ ÐÎÇÏÀÄÓ ÌÓËÜÒÈÏÎËÜÍÈÕ Ã²ÃÀÍÒÑÜÊÈÕ<br />
ÐÅÇÎÍÀÍѲ  ßÄÐÀÕ<br />
Ðåçþìå<br />
Ðîçâèíóòî óäîñêîíàëåíèé óçàãàëüíåíèé áàãàòî êîíô³ãóðàö³éíèé ï³äõ³ä äëÿ îïèñó ðîçïàäó âèñîêî çáóäæåíèõ ñòàí³â<br />
(ìóëüòèïîëüí³ ã³ãàíòñüê³ ðåçîíàíñè) ÿäåð, ÿêà áàçóºòüñÿ íà îäíî÷àñíîìó âèêîðèñòàíí³ îáîëîíêîâî¿ ìîäåë³ (ç ðîçøèðåíèì<br />
áàçèñîì) òà ì³êðîñêîï³÷íî¿ ìîäåë³ ïðåäðàâíîâ³ñíîãî ðîçïàäó ³ç ñòàòèñòè÷íèì óðàõóâàííÿì ñêëàäíèõ êîíô³ãóðàö³é òèïó 2p2h,<br />
3p3h òà ³íøèõ. Íîâèé ï³äõ³ä âèêîðèñòàíî äëÿ àíàë³çó ðåàêö³¿ (μ - n) íà ÿäð³ 40 Ca.<br />
Êëþ÷îâ³ ñëîâà: ìóëüòèïîëüí³ ã³ãàíòñüê³ ðåçîíàíñè, óçàãàëüíåíà áàãàòîêîíô³ãóðàö³éíà ìîäåëü, ðåàêö³ÿ (μ - n) íà ÿäð³ 40 Ca.
UDÑ 621.315.592<br />
YU. F. VAKSMAN 1 , YU. A. NITSUK 1 , V. V. YATSUN 1 , YU. N. PURTOV 1 , A. S. NASIBOV 2 , P. V. SHAPKIN 2<br />
1 I. I. Mechnikov National University, 65026 Odessa, Ukraine<br />
2 P. N. Lebedev Physical Institute, Russian Academy of Sciences,<br />
117924 Moscow, Russia<br />
OPTICAL PROPERTIES OF ZnSe:Mn CRYSTALS<br />
ZnSe single crystals with diffusion doping of Mn have been investigated. Absorption, luminescence<br />
and photoconductivity of ZnSe:Mn crystals have been studied and analyzed in the visible region<br />
of the spectrum. Concentration of Mn impurity was estimated from absorption edge. The electron<br />
transition scheme in ZnSe:Mn was proposed.<br />
1. INTRODUCTION<br />
Semiconductor compounds A 2 B 6 with dopants<br />
of transition metals are described by internal transitions<br />
in 3d states — absorption and luminescence.<br />
Investigation of internal transitions luminescence on<br />
these states and luminescent centers formed by Mn<br />
is interesting because ZnS:Mn and ZnSå:Mn crystals<br />
are rather good phosphors [1]. In this work, diffusion<br />
doping of ZnSe single crystals with Mn is described.<br />
The optical absorption, luminescence and photoconductivity<br />
of ZnSe:Mn crystals have been investigated<br />
and analyzed in the visible range of the spectrum.<br />
Concentration of Mn impurity was estimated from<br />
absorption edge shift.<br />
The purpose of this study is to develop the procedure<br />
of diffusion Mn doping of the ZnSe crystals,<br />
to identify the optical absorption, luminescence and<br />
photoconductivity spectra of obtained samples.<br />
2. EXPERIMENTAL<br />
The samples for study were prepared via diffusion<br />
Mn doping of pure ZnSe single crystals. Undoped<br />
crystals were obtained by the method of free growth on<br />
a single-crystal ZnSe substrate with the growth plane<br />
(111) or (100). This method was described in detail,<br />
and the main characteristics of the ZnSe crystals were<br />
obtained in [2, 3]. The selection of temperature profiles<br />
and design of the growth chamber excluded the<br />
possibility of a contact between the crystal and chamber<br />
walls. The dislocation density in the crystals obtained<br />
was no higher than 10 4 cm –2 .<br />
Initially, the crystal doping was provided by impurity<br />
diffusion towards crystal bulk from evaporated<br />
surface layer of metallic Mn in He+Ar atmosphere.<br />
Then, the crystals have been annealed at 1173-1223Ê.<br />
Diffusion process time was 5 hours. However, this<br />
method didn’t form crystals with high concentration<br />
of Mn, thus Mn atoms diffused into the crystal bulk<br />
resulted from high diffusion coefficient of Mn. As result,<br />
ZnSe:Mn crystals with low Mn concentration<br />
(10 16 cm -3 ) were obtained.<br />
The method described in [4,5] was used for highly<br />
doped ZnSe:Mn crystals obtaining. Metal powderlike<br />
Mn was used as the source of impurity. To prevent<br />
crystal etching, Mn powder was mixed with ZnSe<br />
© Yu. F. Vaksman, Yu. A. Nitsuk, V. V. Yatsun, Yu. N. Purtov, A. S. Nasibov, P. V. Shapkin, 2009<br />
powder in 1:1 ratio. Diffusion process was performed<br />
in He+Ar atmosphere in the temperature range from<br />
1173 to 1223 K. The diffusion process was 5h long.<br />
The spectra of optical density were measured using<br />
an MDR-6 monochromator with diffraction grating<br />
1200 grooves/mm in the visible region. The light<br />
intensity was registered by photomultiplier FEU-100.<br />
The optical density spectra were measured at 77 and<br />
293 K.<br />
Photoluminescence spectra were measured by<br />
ISP-51 quartz prism spectrograph. Photoluminescence<br />
excitation was provided by super luminescent<br />
diode EDEV-3LA1 Edison Opto Corporation with<br />
λ max =400 nm.<br />
Indium contacts were deposited on the surface<br />
of crystals for photoconductivity measurements. The<br />
contacts were formed by firing in vacuum at 600 K.<br />
Monochromator MUM-2 was used for photoconductivity<br />
spectra measurements. Halogen lamp was used<br />
for excitation of the spectra.<br />
3. OPTICAL ABSORPTION OF ZnSe:Mn IN<br />
THE VISIBLE REGION SPECTRUM<br />
The optical density (D * ) spectra of ZnSe:Mn crystals,<br />
obtained at different annealing temperatures,<br />
are presented in fig.1. The spectra of undoped ZnSe<br />
crystals characterized by absorption edge at 2.82 eV<br />
(fig 1, curve 1) at Ò =77K. The second linear area was<br />
located at 2.76 eV, associated with an unresilient exciton-exciton<br />
interaction [3]. No features was observed<br />
at the energies lower than 2.6 eV.<br />
Mn doping led to absorption edge shift (fig.1,<br />
curves 2-4). The shift value increased with annealing<br />
temperature raise. The change of the band gap (meV)<br />
as the function of impurity concentration was discussed<br />
in [3]:<br />
13 13<br />
5 ⎛3⎞ eN<br />
Δ Eg<br />
= −210 ⋅ ⎜ ⎟<br />
⎝π⎠ 4πε0εs<br />
, (1)<br />
where å-electron charge, N-impurity concentration<br />
(cm -3 ), ε s =8.66 — ZnSe dielectric constant. As result,<br />
Mn-dopant concentrations have been calculated. The<br />
obtained values are presented in Table. Maximum of<br />
Mn concentration was observed (6∙10 19 cm -3 ) for the<br />
samples annealed at 1223Ê.<br />
61
(D * ) 2<br />
300<br />
250<br />
200<br />
150<br />
100<br />
50<br />
0<br />
62<br />
2.65 2.7 2.75 2.8 2.85 2.9<br />
3<br />
2<br />
1<br />
E, eV<br />
Fig. 1. The optical-density spectra of ZnSe (1) and ZnSe:Mn<br />
(2,3) crystals doped with Mn at temperatures of (2) 1173 and (3)<br />
1223K. T =77 K.<br />
m<br />
In the visible region ZnSe:Mn optical-density<br />
spectra have several absorption lines, which intensity<br />
increases with Mn concentration enhance (fig.2).<br />
Three absorption lines at 2.31, 2.47 and 2.67 eV can<br />
be separated<br />
Table<br />
The change of the band gap (meV) in the ZnSe:Mn crystals<br />
Crystal type 77 Ê 300 Ê ΔÅ g , meV N, cm -3<br />
ZnSe, undoped 2.82 2.68 --- ---<br />
ZnSe:Mn,<br />
doped at 1173K<br />
2.78 2.64 40 2∙1018 ZnSe:Mn,<br />
doped at 1223K<br />
2.69 2.55 130 6∙1019 Absorption measurements at 77-300Ê showed that<br />
lines at 2.31, and 2.47 eV didn’t change their positions<br />
with temperature raise. Line at 2.67 eV at 300 K is located<br />
at the conductivity band because the band gap of<br />
ZnSe:Mn is 2.55-2.64 eV at this temperature. The one<br />
can suppose that intracenter transitions are the origins<br />
of those lines. According to [6], absorption line at 2.31<br />
eV is due to transition from the ground state 6 À 1 (G) to<br />
excited state 4 Ò 1 (G) of Mn 2+ ion. The line at 2.47 eV<br />
is due to 6 À 1 (G)→ 4 T 2 (G) transitions and the line at<br />
2.67 eV is due to 6 À 1 (G)→ 4 Å 1 (G) intracenter transitions.<br />
3. ZnSe:Mn PHOTOLUMINESCENCE<br />
SPECTRA<br />
Photoluminescence measurements have been performed<br />
at 77-600 K. At 77K ZnSe:Mn crystals spectra<br />
had two narrow lines at 2.12 and 2.31 eV. The intensity<br />
of the lines increased with Mn concentration increase<br />
(fig.3, curves 1,2). Lines positions didn’t change with<br />
the temperature increase that evidences about intracenter<br />
nature of this lines.<br />
D *<br />
3.3<br />
3.2<br />
3.1<br />
3<br />
2.9<br />
2.8<br />
2.7<br />
2.6<br />
2.5<br />
3<br />
2<br />
1<br />
2.4<br />
1.9 2 2.1 2.2 2.3 2.4 2.5 2.6 2.7<br />
E, eV<br />
Fig. 2. The optical-density spectra of ZnSe (1) and ZnSe:Mn<br />
(2,3) crystals in the visible region of the spectrum at 77 K. Curve<br />
2 corresponds to the sample annealed at 1173 K and curve 3 corresponds<br />
to annealing at 1223 K.<br />
I, arb. un.<br />
1<br />
0.9<br />
0.8<br />
0.7<br />
0.6<br />
0.5<br />
0.4<br />
0.3<br />
0.2<br />
0.1<br />
0<br />
1.8 1.9 2 2.1 2.2 2.3 2.4<br />
3<br />
2<br />
1<br />
E, eV<br />
Fig. 3. The photoluminescence spectra of ZnSe:Mn crystals at<br />
77(1,2) and 400 K (3). Curve 1 corresponds to the sample annealed<br />
at 1173 K and curves 2,3 correspond to annealing at 1223 K.<br />
The temperature dependence of the luminescence<br />
at 300-600 K showed that line at 2.31 eV disappeared
at 400 K (fig.3, curve 3) and line at 2.12 eV disappeared<br />
at 600 K. Luminescence lines half-width increases<br />
with the temperature increase:<br />
⎛2kT ⎞<br />
E1/2 = E0⎜<br />
⎟ , (2)<br />
⎝ hΩ<br />
⎠<br />
The equation (2) is obtained from the model of<br />
configuration coordinates, where Å is the lines half-<br />
0<br />
width at 0 K.<br />
1/2<br />
4. ZnSe:Mn PHOTOCONDUCTIVITY<br />
SPECTRA<br />
It is established, that ZnSe:Mn crystals had photosensitivity.<br />
ZnSe:Mn crystals photoconductivity spectra<br />
at different temperatures are shown in fig.4. The<br />
one can see, one line is observed at 2.78 eV under 77 K<br />
(fig.4). This line present in spectra of undoped ZnSe<br />
crystals and could be associated with intraband transitions.<br />
Low-energy photoconductivity part increases<br />
with the temperature increase (fig4, curves 2-4).<br />
Ip.c., arb. un.<br />
1<br />
0.9<br />
0.8<br />
0.7<br />
0.6<br />
0.5<br />
0.4<br />
0.3<br />
0.2<br />
0.1<br />
4<br />
0<br />
1.8 1.9 2 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8<br />
3<br />
2 1<br />
�, eV<br />
Fig. 4. The photoconductivity spectra of ZnSe:Mn crystals at<br />
77(1), 293 (2), 323 (3) and 403Ê (4).<br />
The permanent lines at 2.31 and 2.47 eV appeared<br />
in spectra at temperatures over 293 K. These lines positions<br />
are identical to absorption lines. The intensity<br />
of photoconductivity lines was changed with the temperature.<br />
At room temperatures high energy lines were<br />
dominated whereas at 403 K 2.31 eV lines intensity<br />
becomes maximal.<br />
Electron transitions scheme of ZnSe:Mn based on<br />
optical properties investigations, is shown in fig.5.<br />
As it is mentioned above, absorption lines at 2.31,<br />
2.47 and 2.67 eV are the result of transitions from the<br />
ground state 6À (G) to Mn excited states (fig.5, transi-<br />
1<br />
tions 1-3). According to [1], the ground state of Mn<br />
ion is located 0.1 eV higher than valence band.<br />
Photoluminescence lines at 2.12 and 2.31 eV<br />
are resulted by transitions from excited states to the<br />
ground state of Mn ion (fig.5, transitions 4,5).<br />
Presented scheme allows to explain photoconductivity,<br />
which is due two stage process. First, optical<br />
transitions 2 and 3 take place and then thermal electron<br />
transition to conductance band starts (transitions<br />
6 and 7). The absence of low energy photoconductivity<br />
up to 300 K, can be explained by the impossibility<br />
of thermal transitions of electrons from 4 E 1 (G) to conductance<br />
band. It is worth to say that similar results<br />
have been obtained by us before for ZnSe:Cr [8].<br />
Eg<br />
Mn 2+<br />
6<br />
1<br />
7<br />
2<br />
3<br />
4 5<br />
6 �1(G)<br />
Fig. 5. The electron transition scheme in ZnSe:Mn crystals.<br />
5. CONCLUSIONS<br />
Ec<br />
EV<br />
The studies carried out allow us to conclude the<br />
following.<br />
1. A procedure of diffusion Mn doping of the ZnSe<br />
crystals has been developed. Maximal Mn concentration,<br />
estimated from the absorption edge, was 6∙1019 cm-3 .<br />
2. The nature of ZnSe:Mn crystals absorption<br />
lines in the visible region of the spectrum have been<br />
identified.<br />
3. The identity of absorption, photoluminescence<br />
and photoconductivity lines in ZnSe:Mn was<br />
shown.<br />
4. The electron transition scheme in the ZnSe:Mn<br />
crystals was proposed.<br />
References<br />
1. Àãåêÿí Â.Ô. Âíóòðèöåíòðîâûå ïåðåõîäû èîíîâ ãðóïïû<br />
æåëåçà â ïîëóïðîâîäíèêîâûõ ìàòðèöàõ òèïà À 2 Â 6 //<br />
ÔÒÒ. — 2002. — Ò. 44, ¹. 11. — Ñ. 1921-1939.<br />
2. Korostelin Yu.V., Kozlovsky V.I., Nasibov A.S., Shapkin P.V.<br />
Vapour growth and doping of ZnSe single crystals // J.Cryst.<br />
Growth. — 1999. — V. 197. — P.449-454.<br />
3. Âàêñìàí Þ.Ô., Íèöóê Þ.À., Ïóðòîâ Þ.Í., Øàïêèí<br />
Ï.Â. Ñîáñòâåííûå è ïðèìåñíûå äåôåêòû â ìîíîêðèñòàëëàõ<br />
ZnSe:In, ïîëó÷åííûõ ìåòîäîì ñâîáîäíîãî ðîñòà<br />
// ÔÒÏ. — 2001. — Ò.35, ¹8. — ñ. 920-926.<br />
4. Âàêñìàí Þ.Ô., Ïàâëîâ Â.Â., Íèöóê Þ.À., Ïóðòîâ Þ.Í.,<br />
Íàñèáîâ À.Ñ. Øàïêèí Ï.Â. Îïòè÷åñêîå ïîãëîùåíèå è<br />
äèôôóçèÿ õðîìà â ìîíîêðèñòàëëàõ ZnSe // ÔÒÏ. —<br />
2005. — Ò. 39, ¹4. — Ñ. 401-404.<br />
5. Âàêñìàí Þ.Ô., Ïàâëîâ Â.Â., Íèöóê Þ.À., Ïóðòîâ Þ.Í.,<br />
Íàñèáîâ À.Ñ., Øàïêèí Ï.Â. Ïîëó÷åíèå è îïòè÷åñêèå<br />
ñâîéñòâà ìîíîêðèñòàëëîâ ZnSe, ëåãèðîâàííûõ êîáàëüòîì<br />
//ÔÒÏ. — 2006. — Ò.40, ¹.7. — Ñ. 815-818.<br />
6. Õìåëåíêî Î.Â., Îìåëü÷åíêî Ñ.À Âëèÿíèå ïëàñòè÷åñêîé<br />
63
64<br />
äåôîðìàöèè íà çàðÿäîâîå ñîñòîÿíèå èîíîâ Mn 2+ â êðèñòàëëàõ<br />
ZnSe // Âåñòíèê Äíåïðîïåòðîâñêîãî óíèâåðñèòåòà.<br />
— 2008. — ¹15.<br />
7. Áóëàíûé Ì.Ô., Êîâàëåíêî À.Â., Ïîëåæàåâ Á.À. Ýëåêòðîëþìèíåñöåíòíûå<br />
èñòî÷íèêè ñâåòà íà îñíîâå ìîíîêðèñòàëëîâ<br />
ZnSe:Mn ñ îïòèìàëüíûìè ÿðêîñòíûìè<br />
UDÑ 621.315.592<br />
Yu. F. Vaksman, Yu. A. Nitsuk, V. V. Yatsun, Yu. N. Purtov, A. S. Nasibov, P. V. Shapkin<br />
OPTICAL PROPERTIES OF ZnSe:Mn CRYSTALS<br />
õàðàêòåðèñòèêàìè // ÆÒÔ. — 2003. — Ò.73, ¹.2. —<br />
Ñ. 133-135.<br />
8. Âàêñìàí Þ.Ô., Ïàâëîâ Â.Â., Íèöóê Þ.À., Ïóðòîâ Þ.Í.<br />
Îïòè÷åñêèå ñâîéñòâà êðèñòàëëîâ ZnSe, ëåãèðîâàííûõ<br />
ïåðåõîäíûìè ýëåìåíòàìè // ³ñíèê Îäåñüêîãî íàö. óíòó.<br />
— 2006. — Ò.11, âèï. 7. — Ô³çèêà. — Ñ. 47-53.<br />
Abstract<br />
ZnSe single crystals with diffusion doping of Mn have been investigated. Absorption, luminescence and photoconductivity of ZnSe:<br />
Mn crystals have been studied and analyzed in the visible region of the spectrum. Concentration of Mn impurity was estimated from<br />
absorption edge. The electron transition scheme in ZnSe:Mn was proposed.<br />
Key words: diffusion doping, optical-density, photoluminescence, photoconductivity, intracenter transition.<br />
ÓÄÊ 621.315.592<br />
Þ. Ô. Âàêñìàí, Þ. À. Íèöóê, Â. Â. ßöóí, Þ. Í. Ïóðòîâ, À. Ñ. Íàñèáîâ, Ï. Â. Øàïêèí<br />
ÎÏÒÈ×ÅÑÊÈÅ ÑÂÎÉÑÒÂÀ ÊÐÈÑÒÀËËΠZnSe:Mn<br />
Ðåçþìå<br />
Èññëåäîâàíû ìîíîêðèñòàëëû ZnSe:Mn, ïîëó÷åííûå ìåòîäîì äèôôóçèîííîãî ëåãèðîâàíèÿ. Èññëåäîâàíû ñïåêòðû îïòè÷åñêîé<br />
ïëîòíîñòè, ôîòîëþìèíåñöåíöèè è ôîòîïðîâîäèìîñòè â âèäèìîé îáëàñòè. Ïî âåëè÷èíå ñìåùåíèÿ êðàÿ ïîãëîùåíèÿ<br />
îïðåäåëåíà êîíöåíòðàöèÿ ìàðãàíöà â èññëåäóåìûõ êðèñòàëëàõ. Ïîñòðîåíà ñõåìà îïòè÷åñêèõ ïåðåõîäîâ â êðèñòàëëàõ<br />
ZnSe:Mn.<br />
Êëþ÷åâûå ñëîâà: äèôôóçèîííîå ëåãèðîâàíèå, îïòè÷åñêàÿ ïëîðòíîñòü, ôîòîëþìèíåñöåíöèÿ, ôîòîïðîâîäèìîñòü, âíóòðèöåíòðîâûå<br />
ïåðåõîäû.<br />
ÓÄÊ 621.315.592<br />
Þ. Ô. Âàêñìàí, Þ. À. ͳöóê, Â. Â. ßöóí, Þ. Ì. Ïóðòîâ, Î. Ñ. Íàñèáîâ, Ï. Â. Øàïê³í<br />
ÎÏÒÈ×Ͳ ÂËÀÑÒÈÂÎÑÒ² ÊÐÈÑÒÀ˲ ZnSe:Mn<br />
Ðåçþìå<br />
Äîñë³äæåíî ìîíîêðèñòàëè ZnSe:Mn, îòðèìàí³ ìåòîäîì äèôóç³éíîãî ëåãóâàííÿ. Ïðîâåäåí³ äîñë³äæåííÿ ñïåêòð³â îïòè÷íî¿<br />
ãóñòèíè, ôîòîëþì³íåñöåíö³¿ òà ôîòîïðîâ³äíîñò³ â âèäèì³é îáëàñò³. Ïî çì³ùåííþ êðàþ ïîãëèíàííÿ âèçíà÷åíî êîíöåíòðàö³¿<br />
ìàðãàíöþ â äîñë³äæóâàíèõ êðèñòàëàõ. Ïîáóäîâàíà ñõåìà îïòè÷íèõ ïåðåõîä³â â êðèñòàëàõ ZnSe:Mn.<br />
Êëþ÷îâ³ ñëîâà: äèôóç³éíå ëåãóâàííÿ, îïòè÷íà ãóñòèíà, ôîòîëþì³íåñöåíö³ÿ, ôîòîïðîâ³äí³ñòü, âíóòð³öåíòðîâ³ ïåðåõîäè.
UDÑ 539.19+539.182<br />
A. V. GLUSHKOV<br />
Odessa State University, Odessa<br />
QUASIPARTICLE ENERGY FUNCTIONAL FOR FINITE TEMPERATURES<br />
AND EFFECTIVE BOSE-CONDENSATE DYNAMICS: THEORY AND SOME<br />
ILLUSTRATIONS<br />
INTRODUCTION<br />
At present time a density functional theory became<br />
by a powerful tool in studying the electron structure of<br />
different materials, including atomic and molecular<br />
systems, solids, semiconductors etc. [1-16]. A construction<br />
of the correct energy functionals of a density<br />
for multi-body systems represents very actual and<br />
important problem of the modern theory of semiconductors<br />
and solids, thermodynamics, statistical physics<br />
(including a theory of non-equilibrium thermodynamical<br />
processes), quantum mechanics and others.<br />
In last time a development of formalism of the energy<br />
density functional has been considered in many<br />
papers (see [1–7]). Its application is indeed based on<br />
the two known theorems by Hohenbreg-Kohn (τ = 0,<br />
where τ is a temperature) and Mermin (τ ≠ 0) [1,2].<br />
According to these theorems, an energy and thermodynamical<br />
potential of the multi-body system are<br />
universal density functionals. Though these theorems<br />
predict an existence of such a density functional, however<br />
its practical realization is connected with a number<br />
of the significant difficulties (see [1-3,8-17]). The<br />
problem is complicated under consideration of the<br />
non-stationary tasks (the known theorem by Runge-<br />
Gross about 1-1 mapping between time-dependent<br />
densities and the external potentials [2]).<br />
Let us remind some important results of the density<br />
functional theory. It should be mentioned a constructive<br />
approach to delivering optimal representations<br />
for an exact density functional [1,2,8-16], which has<br />
been used for generalization of the Hohenberg-Kohn<br />
theorem in order to get an effective density functional<br />
for large molecules. As alternative version one could<br />
consider a quasiparticle density functional formalism<br />
by Beznosjuk-Kryachko (see [1-3,8]). The latter develops<br />
a quasiparticle conception of Kohn-Sham and<br />
the Levi-Valone method [2,3]. In fact it has been done<br />
an attempt practically to realize an idea of the Hohenberg-Kohn<br />
theorem. More advanced analogous approach<br />
with account of the multi-particle correlations<br />
is developed in ref. [8,17,18]. It has been shown a fundamental<br />
feature of the Weizsacker universal density<br />
functional, which describes an energy of the effective<br />
condensate of interacting bosons. In ref. [14] (see also<br />
[8,19-21]) it has been firstly developed a QED theory<br />
© A. V. Glushkov, 2009<br />
It is considered a theory of the quasiparticle energy functional under non-zeroth temperatures τ<br />
and some its applications. A thermodynamical potential for multielectron system in external stationary<br />
field for given τ is defined by dynamics of effective Bose-condensate in atoms of the physical space<br />
of electrons. Structure of this space is defined by the cell system of surfaces of zeroth flux for entropy<br />
pulse under availability of the zeroth current of the Bose-condensate density.<br />
of a density functional formalism and constructed an<br />
optimized one-quasiparticle representation in a theory<br />
of multi-electron systems. The lowest order multibody<br />
effects, in particular, the gauge dependent radiative<br />
contribution for the certain class of the photon<br />
propagators calibration are treated in QED formulation<br />
and new density functional integral-differential<br />
equations are derived. The minimal value of the gauge<br />
dependent radiative contribution is considered to be<br />
the typical representative of the multi-electron correlation<br />
effects, whose minimization is a reasonable<br />
criteria in the searching for the optimal QED perturbation<br />
theory one-electron basis.<br />
New several schemes for application of the density<br />
functional theory are based on using so called hybrid<br />
functionals [9-13,15], which improve the description<br />
of the optical absorption spectra and related properties<br />
for solids, semiconductors etc. However, the principal<br />
problems are remained. For example, one could<br />
remind the limitations stemming from (semi) local<br />
density approximation (LDA) functionals [2,15] (<br />
A thermochemistry: up to 1 eV error; The structural<br />
properties: 23% error; The elastic constants: 10% error;<br />
great problems for the strongly correlated systems<br />
(transition metal oxides); The Van der Waals bonding<br />
missing; Band gap problem; Problem of description of<br />
the electronic excitations, etc). There is a little more<br />
situation in the Perdew-Burke-Ernzenhof (PBE0) and<br />
Heyd-Scuseria- Ernzerhof (HSE0) hybrid schemes.<br />
In particular, in the last scheme the lattice constants<br />
and bulk modules are clearly improved for insulators<br />
and semiconductors in comparison with the LDA and<br />
even PBE0 schemes. The band gaps are excellent for<br />
wide range of semiconductors, except for very large<br />
gap systems. The transition metals are problematic,<br />
at least in terms of the bulk modules. At last, the atomization<br />
energies are not improved compared to the<br />
PBE scheme. Sufficiently full review of the modern<br />
state of art for the density functional theories and their<br />
applications in theory of semiconductors is presented<br />
in the recent report [15] (see also [9-13]).<br />
Further, it is natural to note that a density functional<br />
theory for the zeroth temperatures τ = 0 is developed<br />
in a majority of the papers. At the same time,<br />
similar theories for τ ≠ 0 have a whole number of the<br />
significant problems [1-4]. It is self-understood that<br />
65
their applications are not widely known. In ref. [8] it is<br />
considered a theory of the quasiparticle energy functional<br />
under non-zeroth temperatures τ and shown<br />
that a thermodynamical potential for multielectron<br />
system in external stationary field for given τ is defined<br />
by dynamics of effective Bose-condensate in atoms of<br />
the physical space of electrons. Below we will consider<br />
an advanced version of this theory and give some its<br />
illustrations regarding the molecular structure mapping,<br />
a theory of semiconductors in a laser field etc.<br />
66<br />
2. A QUASIPARTICLE DENSITY<br />
FUNCTIONAL THEORY<br />
A formalism of the energy density functional theory<br />
for non-zeroth temperatures, developed below (see<br />
also [8]), is based on the two fundamental results [1].<br />
In a large canonical ensemble under given temperature<br />
a density distribution ï( r ) directly defines a value<br />
of V( r ) — μ (here μ is a chemical potential). For given<br />
V( r ) and μ it is existed a functional of ï( r ):<br />
r r r 3 r<br />
Ω V −μ ⎡⎣n′ () r ⎤⎦<br />
= ∫(<br />
V() r −μ ) n′ () r d r+ F( n′ () r ) , (1)<br />
which reaches an absolute minimum, when a density<br />
ï′( r ) is a regular density ï( r ) ~ V( r ). A value<br />
of Ω is in minimum equal to a thermodynamical<br />
V-μ<br />
potential; F is an universal density functional, which<br />
represents an internal energy of a system and is dependent<br />
upon τ. A distribution ï( r ) is searched as a<br />
solution of the fundamental equation (naturally under<br />
corresponding boundary conditions) of the following<br />
type:<br />
N r<br />
δF⎡ ⎣n () r ⎤<br />
⎦ r<br />
V N r + () r =μ.<br />
(2)<br />
δn<br />
() r<br />
An universal functional of the electrons (fermions)<br />
is described for a given temperature τ as follows [8]:<br />
r ⎛ 1 ⎞<br />
FF⎡⎣n′ () r ⎤⎦ = Spρ ′ ⎜T + U − ln ρ ′ ⎟=<br />
⎝ β ⎠<br />
r r r<br />
= U ⎡⎣n′ () r ⎤⎦+Δ S⎡⎣n′ () r ⎤⎦+ G⎡⎣n′ () r ⎤⎦,<br />
r r r<br />
G⎡⎣n′ () r ⎤⎦ = T ⎡⎣n′ () r ⎤⎦+τS⎡⎣n′ () r ⎤⎦,<br />
(3)<br />
where β = (k τ) B -1 , ρ′ is a large canonical operator of<br />
a matrice of density; G is a free Helmholtz energy of<br />
the non-interacting electrons, S is an entropy, ΔS is<br />
an effective energy of correlation; TU , are the operators<br />
of the kinetical energy and energy of the Coulomb<br />
interaction:<br />
2 N e 1<br />
U = ∑ r r ,<br />
2 i, j= 1(<br />
ri − rj)<br />
N<br />
N — represented density nF() r is created by a set of<br />
N N N N<br />
the fermionic wave functions { nF ←Ψ F }; F0F n ⎡ ⎤ Ψ ⎣ ⎦<br />
realizes a minimal mathematical expectation of inter-<br />
N<br />
nal energy of the electrons for the fixed density nF() r .<br />
Let us turn attention on the following important circumstance:<br />
ρ′ in the equation (3) is an operator of a<br />
density of the interacting particles system and it does<br />
not reduced to anti-symmetrized multiplying the den-<br />
sity matrices. That’s why an entropy in (3) differs from<br />
an ideal gas entropy. There is an effective energy of<br />
correlation ΔS besides the Coulomb energy U in the<br />
cited formulae.<br />
Generalizing above noted results, we define the<br />
universal energy density functionals of the effective interacting<br />
and non-interacting bosons, every of which<br />
has a mass and a charge of electrons (in the second<br />
case only a mass) [4]:<br />
N N N N<br />
F ⎡ b ⎣n ⎤ b ⎦ = G⎡ ⎣n ⎤ b ⎦+ U ⎡<br />
⎣n ⎤ b ⎦+ΔS⎡ ⎣n ⎤ b ⎦ , (4)<br />
N N<br />
F ⎡ b0⎣n ⎤ b ⎦ = G⎡ ⎣n ⎤ b ⎦ , (5)<br />
Here N — represented density of bosons N<br />
n b is cre-<br />
N N<br />
ated by a set of N — boson wave functions b b n Ψ → ;<br />
N N<br />
b 0 b n ⎡ ⎤ Ψ ⎣ ⎦ and N N<br />
b 0’<br />
b n ⎡ ⎤ Ψ ⎣ ⎦ are the wave functions of the<br />
interacting and non-interacting bosons correspondingly,<br />
which realize a minimal mathematical expectation<br />
of the internal energy of bosons for the fixed density<br />
N<br />
n b ( r ). The conditions of the N — representation<br />
for a matrice of density of the first order for fermions<br />
and bosons are reduced to the expansion conditions<br />
as follows:<br />
Φ 1 r 2<br />
F ()<br />
N r ηα Ψα r ηα<br />
1<br />
nF() r = N∑<br />
,0≤<br />
≤ , (6)<br />
β( εα−μ) β( εα−μ) α 1+ e 1+<br />
e N<br />
b 1 r 2<br />
b ()<br />
N r ηα Ψα r ηα<br />
nb() r = N∑<br />
,0≤ ≤1,<br />
(7)<br />
βε ( α−μ) βε ( α−μ)<br />
α 1+ e 1+<br />
e<br />
Naturally, a set of the fermionic densities<br />
is contained in a set of the bosonic densities<br />
N r N r<br />
{ nF () r } ⊂{<br />
nb () r } in the ideal gas approximation.<br />
Further a class of the N — represented fermionic<br />
densities for system of N — electrons is considered.<br />
A thermodynamical potential can be written as follows:<br />
NF N N r N r<br />
Ω ( ) = G ⎡ b n ⎤ i + V( ⎡n ⎤ i , r iV ) ni () r dω<br />
−μ ⎣ ⎦ ∫ ⎣ ⎦<br />
. (8)<br />
ω<br />
A potential of the external effective field of the<br />
interacting bosons system is defined by the following<br />
auxiliary universal density functionals:<br />
N r r N r<br />
V( ⎡<br />
⎣n ⎤<br />
⎦, r) = V() r −μ+ VH( ⎡<br />
⎣n ⎤<br />
⎦,<br />
r)<br />
, (9)<br />
∫<br />
ω<br />
r r<br />
( , ) ()<br />
N N N N<br />
V ⎡ H ⎣n ⎤<br />
⎦ r n r dω= H ⎡ b ⎣n ⎤<br />
⎦+ H ⎡ bF ⎣n ⎤<br />
⎦,<br />
(10)<br />
N N N<br />
H ⎡ b ⎣n ⎤<br />
⎦ = F ⎡ b ⎣n ⎤<br />
⎦−G ⎡ b ⎣n ⎤<br />
⎦ ,<br />
N N N<br />
H ⎡ bF ⎣n ⎤<br />
⎦ = F ⎡ F ⎣n ⎤<br />
⎦−F ⎡ b ⎣n ⎤<br />
⎦ ,<br />
N<br />
where H ⎡ b ⎣n ⎤<br />
⎦ is an energy of correlation of the N<br />
N<br />
interacting bosons, H ⎡ bF ⎣n ⎤<br />
⎦ is a correction to an energy<br />
of correlation of the bosons, accounting for the<br />
Pauli principle.<br />
The wave function of the inhomogeneous Bosecondensate<br />
is as follows:<br />
N<br />
N r r i D N r<br />
( ⎡n ⎤; r, , r ) exp S ( ⎡n ⎤,<br />
r)<br />
Ψ bђ0H⎣ ⎦i1 K N =<br />
⎧<br />
⎨∑ ⎩ i=<br />
1 h<br />
b ⎣ ⎦<br />
⎫<br />
⎬.<br />
⎭<br />
Here:
N r<br />
D N r ih ⎛n () r ⎞ i<br />
Sbђ( ⎡<br />
⎣n ⎤<br />
⎦, ir)<br />
= S⎜<br />
⎟<br />
2 ⎜ N ⎟<br />
⎝ ⎠<br />
is a generalized action function, which is proportional<br />
to the specific entropy of “i” boson, where it is right<br />
the following definition S(n) = -lnn.<br />
Starting from D<br />
S bђ , it is easy to define a functional<br />
of a pulse of the entropy for “i“ boson as follows:<br />
r N r r D N r<br />
PEђ( ⎣<br />
⎡n ⎦<br />
⎤, ri i ) = Im { ∇rSb<br />
( ⎡<br />
⎣n ⎤<br />
⎦,<br />
r)<br />
} ,<br />
which is directly connected with the Heisenberg principle<br />
of indefiniteness.<br />
A mathematical expectation of operator of the ki-<br />
N N<br />
netical energy on the wave functions b 0’<br />
n ⎡ ⎤ Ψ ⎣ ⎦ is defined<br />
as follows:<br />
r r r<br />
T n = P n r n r mdω<br />
=<br />
( ) ()<br />
M<br />
N 2 N N<br />
⎡ b ⎣<br />
⎤<br />
⎦ ∑∫ ⎡ E ⎣<br />
⎤<br />
⎦,<br />
/2 j<br />
j=<br />
1<br />
r r r<br />
= h m ∇ n r n r dω<br />
M<br />
2<br />
N N<br />
∑∫ ( /8 ) ⎡<br />
⎣ () ⎤<br />
⎦<br />
/ ()<br />
j=<br />
1<br />
r j<br />
N<br />
( ⎡<br />
⎣<br />
⎤<br />
⎦ ) () ∑ { rђ b(<br />
⎡<br />
⎣<br />
⎤<br />
⎦j<br />
) }<br />
r<br />
i=<br />
1<br />
(11)<br />
when the following condition r of the zeroth flux for a<br />
N r<br />
pulse of the entropy PE( ⎡<br />
⎣n ⎤<br />
⎦r)<br />
:<br />
r N r r r<br />
P ( ⎡ E ⎣n ⎤<br />
⎦r)<br />
⋅ lЈ() r = 0, j = 1, K , M . (12)<br />
j<br />
r∈Ј<br />
j<br />
is fulfilled Mon<br />
the boundaries {£} of M atoms (the relationship<br />
∑ ω j =ω is fulfilled for volumes ω ).A math-<br />
j<br />
ematical expectation j= 1 of the Bose-condensate density<br />
current is as follows:<br />
N N N D N<br />
j n , r = n r Re ∇ S n , r = 0<br />
r r r r<br />
.<br />
The conditions (12), i.e. the conditions of the zeroth<br />
flux for a pulse of entropy of distribution for the<br />
Bose-condensate density define in fact the surfaces,<br />
the cell system of which defines an atomic structure<br />
of the physical space of electrons under availability of<br />
zeroth current of a density. Taking into account of the<br />
equations (2), (8)–(12) it is clear that the electrons<br />
system states in an external scalar field V( r ) fro given<br />
temperature τ are defined by the iterative solution of<br />
the following system of the differential equations for<br />
each r ∈ ω [4]: j<br />
r 2<br />
2 2 N r<br />
h r N r h ⎡∇r r n () r ⎤<br />
∇ r<br />
n () r + ⎢ N r ⎥ −<br />
4m 8m⎢⎣<br />
n () r ⎥⎦<br />
N<br />
δS⎡ ⎣n ⎤<br />
⎦ r<br />
−τ + V() r = μ,<br />
N r<br />
%<br />
(13)<br />
δn<br />
() r<br />
N r<br />
δV ( ⎡ H n ⎤,<br />
r)<br />
N N<br />
V% r r r ⎣ ⎦ r<br />
() r = V() r + VH( ⎡<br />
⎣n ⎤<br />
⎦, r) +<br />
n N r () r , (14)<br />
δn<br />
() r<br />
r N r r<br />
∇r r n () r ⋅ l r<br />
J 0<br />
j r∈J = . (15)<br />
j<br />
Here the functional Ò [n b N ] is the well known Weizsacker<br />
correction [1] that is a fundamental part of the<br />
universal density functional for particles of quantum<br />
statistics. The obtained system of the functional<br />
equations (13)–(15) is transferred to the analogous<br />
Beznosjuk system under τ → 0. Further if the multibody<br />
correlations corrections (an effect of energy<br />
dependence of the interparticle interaction, the continuum<br />
pressure etc.) are taken into account in an energy<br />
of interaction between particles, then the system<br />
(13)–(15) coincides with the corresponding equations<br />
system from ref. [8].<br />
Thus, from all consideration it is followed that a<br />
thermodynamical potential for multielectron system<br />
in an external local scalar stationary field V( r r ) for a<br />
given temperature τ is defined by a dynamics of the effective<br />
N-particle Bose-condensate in semi-spaces —<br />
atoms of the physical space of the electrons system.<br />
An atomic structure of the physical space of electrons<br />
is defined by the cell system of surfaces of zeroth flux<br />
for entropy pulse (condition (12)) under availability<br />
of the zeroth current of the Bose-condensate density.<br />
Probably, the same results can be received within the<br />
Lagrange density functional theory [17,18], which is<br />
based on the Green’s functions method. In fact speech<br />
is about using equations for the Matzubarian Green’s<br />
functions.<br />
3. SOME ILLUSTRATIONS OF THEORY AND<br />
CONCLUSION<br />
It is obvious that there is a whole set of the different<br />
applications of the developed formalism. As<br />
an obvious perspective application of the presented<br />
theory else it should be mentioned a quantum geometry<br />
and hadrodynamics [21-23]. As a natural application<br />
of the theory one could indicate a description<br />
of the Bader lodges [3,24,25]. Really, the conditions<br />
(15) are equivalent to the Bader conditions for topological<br />
breaking of the physical space of electrons on<br />
atoms. So, all structural constructions, developed in<br />
this theory, are true in the presented density functional<br />
formalism under non-zeroth temperature.<br />
Such constructions can be manifested in an excitation<br />
dynamics of the semiconductors by a short<br />
(femto-second diapason) laser pulse within the description<br />
by non-equilibrium Green’s function too<br />
[26,27]. It is interesting to note that all theoretical results<br />
(fundamental final formula) of the Bonits method<br />
are reproduced in two new approaches: i). the Ullrich-Erhard-Gross<br />
version of the density functional<br />
approach to multi-electron systems in a laser field [28]<br />
and ii).S-matrix Gell-Mann and Low QED approach<br />
to atomic systems in a laser field, developed in refs.<br />
[29-31]. However the authors of the last cited papers<br />
did not carry out the calculation for semiconductors<br />
in a laser filed.<br />
The next illustration of the theory application<br />
may be as follows. The well known phenomenon of<br />
the Benar’s cells convection in a thermodynamics (a<br />
property of structurization by a net of the hexahedrons<br />
for the non-perturbed atmosphere in a position of the<br />
blocking high pressure crest) can be in the known essence<br />
indicated into correspondence to above mentioned<br />
topological breaking of the physical space of<br />
electrons on atoms [32-37]. The region of a crest is<br />
covered by a grid of the hexahedrons as some fractal of<br />
the density tightening for a barical surface to minimal<br />
one when decreasing of the air density in the clouds<br />
region is treated as the tightening hedrons of a fractal<br />
due to the movement to equilibrium on density. On<br />
67
the mathematical language an operation of covering<br />
the Benar convection region by a net of the hexahedrons<br />
is fulfilled by a conform transformation of the<br />
indicated region on series of multi-angles by means of<br />
the Christoffel-Schwarz integral [32].<br />
The above cited topological breaking the physical<br />
space of electrons on atoms can be taken into accordance<br />
recently discovered [38] (see also [39]) phenomenon<br />
of the cells structurization in the hypothetical<br />
π-electron nano-organic superconducting analog<br />
system , which imitates the “human brain” on the<br />
basis of the computer simulation with using effective<br />
polarization interaction potentials in the π-electron<br />
system of organic supermolecules.<br />
At present time a great attention is turned to a development<br />
of the advanced methods of the numerical<br />
statistical Monte-Carlo modeling for many-body systems<br />
[40] (see also [41]). The known success is reached<br />
in the Monte-Carlo modeling the Bose-systems, where<br />
a great number of the useful results is received. At the<br />
same time, this approach for the Fermi-systems deals<br />
with the well known problems (excluding the electron<br />
gas). This circumstance is connected with the fact that<br />
the trial function must be anty-symmetrized one and<br />
a calculation of the determinants in the trial function<br />
is very slow and complicated procedure. Using the<br />
above constructed boson density functional for multielectron<br />
systems allows in principle to overcome this<br />
problem in the significant manner. Naturally, here it<br />
is required a correct definition of the correction Í in bF<br />
equation (10), which takes into account a difference<br />
between the Bose and Fermi-systems.<br />
Acknowledgement. Author would like to thank<br />
Prof. W. Kohn, L.Sham, E. Gross, I. Kaplan, C. Roothaan,<br />
Yu.Lozovik , A.Theophilou, S. Wilson, for useful<br />
discussions. Besides, author would like to thank<br />
Prof. M.Bonits for providing results of the Green’s<br />
functions method calculation for a laser field excitation<br />
dynamics of a semiconductor.<br />
References<br />
1. Kohn W., Sham L.J. Quantum density oscillations in an<br />
inhomogeneous electron gas//Phys. Rev. A. — 1995. —<br />
Vol.137,N6. — P.1697–1706.<br />
2. Gross E.G., Kohn W. Exchange-correlation functionals in<br />
density functional theory. — N-Y: Plenum, 2005. — 380P.;<br />
Appel H., Gross E.K.U., Time-dependent natural orbitals<br />
and occupation numbers//Phys. Rev. Lett. — 2009. —<br />
Vol.102. — P.024331.<br />
3. The Fundamentals of Electron Density, Density Matrix and<br />
Density Functional Theory in Atoms, Molecules and the<br />
Solid State, Series: Progress in Theoretical Chemistry and<br />
Physics, Eds. Gidopoulos N.I. and Wilson S. — Amsterdam:<br />
Springer, 2004. — Vol.14, 244P.<br />
4. Glushkov A.V., Relativistic quantum theory. Quantum mechanics<br />
of atomic systems. — Odessa: Astroprint, 2008. —<br />
800P.<br />
5. Durand Ph., Malrieu J.P. Ab initio methods of Quantum<br />
Chemistry. — N.Y.:Acad., 2007. — 560P.<br />
6. Zhao Y., Truhlar D.G., Density Functionals with Broad Applicability<br />
in Chemistry//Accounts of Chem. Res. — 2008. —<br />
Vol.41. — P.157-167.<br />
7. Êîëîìèåö Â.Ì.Ïðèáëèæåíèå ìåñòíîé ïëîòíîñòè â àòîìíîì<br />
ÿäðå. — Êèåâ. — Íàóêîâà äóìêà., 1999. — Ñ 164.<br />
8. Glushkov A.V., Relativistic and correlation effects in spectra<br />
of atomic systems.. — Odessa: Astroprint, 2006. — 400P.<br />
9. Kaghazchi P., Simeone F.C., Soliman K.A., Kibler L.A.,<br />
Jacob T., Bridging the gap between nanoparticles and single<br />
crystal surfaces//Faraday Discuss. — 2009. — Vol.140. —<br />
P.69-80.<br />
68<br />
10. Gehrke R., Gruene P., Fielicke A., Meijer G., Reuter K.,<br />
Nature of Ar bonding to small Co + clusters and its effect<br />
n<br />
on the structure determination by far-infrared absorption<br />
spectroscopy//J. Chem. Phys. — 2009. — Vol.130. —<br />
P.034306.<br />
11. Zheng R., Wei W., Shi Q., Density functional theory study on<br />
sum-frequency vibrational spectroscopy of arabinose chiral<br />
solutions// J. Phys. Chem. A. — 2009. — Vol.113. — P.157-<br />
164.<br />
12. Rinke P., Janotti A., Scheffler M., Van de Walle C.G., Defect<br />
formation energies without the band-gap problem:<br />
Combining density-functional theory and the GW approach<br />
for the silicon self-interstitial//Phys. Rev. Lett. — 2009. —<br />
Vol.102. — P.026402.<br />
13. Krishna V., Time-dependent density-functional theory for<br />
nonadiabatic electronic dynamics// Phys. Rev. Lett. —<br />
2009. — Vol.102. — P.053002.<br />
14. Glushkov A.V., Ivanov L.N. Radiation decay of atomic states:<br />
atomic residue and gauge non-invariant contributions //<br />
Phys. Lett.A. — 1999. — Vol.170. — P.33-38.<br />
15. Paier J., Marsman M., Kresse G., Hummer K., Absorption<br />
spectra from TDDFT: do hybrid functionals account for excitonic<br />
effects. — Vienna: Univ. Vienna, 2008. — 54P.<br />
16. Alonso J. L., Andrade X., Echenique P., Falceto F., Prada-<br />
Gracia D., Rubio A., Efficient formalism for large-scale ab<br />
initio molecular dynamics based on time-dependent density<br />
functional theory //Phys.Rev.Lett. — 2008. — Vol.101. —<br />
P.96403.<br />
17. Glushkov A.V., Ambrosov S.V., Loboda A.V. et al, QED calculation<br />
of heavy multicharged ions with account for correlation,<br />
radiative and nuclear effects// Recent Advances in<br />
Theory of Phys. and Chem. Systems 2006. — Vol.15. — P.285-<br />
300.<br />
18. Glushkov A.V., Malinovskaya S.V., Lovett L. et al, Green’s<br />
function method in quantum chemistry: New algorithm for<br />
the Dirac equation with complex energy and Fermi-model<br />
nuclear potential//Int. Journ. of Quantum Chemistry. —<br />
2009. — Vol.109. — N10. — P.1331-1345.<br />
19. Glushkov A.V., Khetselius O.Yu., Malinovskaya S.V., Optics<br />
and spectroscopy of cooperative laser-electron nuclear<br />
processes in atomic and molecular systems — new trend<br />
in quantum optics// Europ.Phys.Journ. ST. — 2008. —<br />
Vol.160,N1. — P.195-204.<br />
20. Glushkov A.V., Khetselius O.Yu., Malinovskaya S.V., Spectroscopy<br />
of cooperative laser-electron nuclear effects in multiatomic<br />
molecules// Molec. Physics . — 2008. — Vol.106. —<br />
N910. — P.1257-1260.<br />
21. Glushkov A.V., Energy Approach to Resonance states of compound<br />
super-heavy nucleus and EPPP in heavy nucleus collisions//<br />
Low Energy Antiproton Physics 2005. — Vol.796. —<br />
P.206-210.<br />
22. Schmid N., Engel E., Dreizler R.M. Density functional approach<br />
to quantum hadrodynamics// Phys.Rev.C. — 2005. —<br />
Vol.52. — P.164-169.<br />
23. Glushkov A.V., Quantization of quasistationary states for<br />
multi-particle Dirac-Kohn-Sham equation in collision problem:<br />
New approach//Proc. of International conference on<br />
Geometry-08. — Odessa 2008. — P.168.<br />
24. Wilson S., Handbook on Molecular Physics and Quantum<br />
Chemistry. — Chichester: Wiley. — 2007. — 700P.<br />
25. Bader R.F., Tal Y., Andersen S.G., et al, Molecular charge<br />
distributions: Method of breaking on the lodgies// Israel<br />
J.Chem. — 1990. — V.19. — P.8-13.<br />
26. Bonits M., Excitation of a semiconductor by a short laser<br />
pulse. Description using non-equilibrium Green’s functions.:<br />
Univ. of Kiel, 2008. — 64P.<br />
27. Introduction to Computational Methods in Many-Body<br />
Physics, Eds. Bonitz M. Semkat D. — Princeton: Rinton<br />
Press, 2006. — 650P.<br />
28. Ullrich C., Erhard S., Gross E. Density Functional Approach<br />
to Atoms in Strong Laser Pulses//Superintense Laser Atoms<br />
Physics. — N-Y.: Kluwer,2006. — P.1-18.<br />
29. Glushkov A.V., Atom in electromagnetic field. — Kiel:,<br />
2005. — 400P.<br />
30. Glushkov A.V. et al, QED approach to atoms in a laser field:<br />
Multi-photon resonances and above threshold ionization//<br />
Frontiers in Quantum Systems in Chemistry and Physics. Series:<br />
Progress in Theoretical Chemistry and Physics 2008. —<br />
Vol.18. — P.541-558.
31. Glushkov A.V., Loboda A.V., Gurnitskaya E.P., Svinarenko<br />
A. A., QED theory of radiation emission and absorption<br />
lines for atoms and atomic ensembles in a strong laser field//<br />
Physica Scripta. — 2009. — Vol.134. — P.305001.<br />
32. Glushkov A.V., Super low-frequency planetary solitons. Entropy<br />
approach and hydrodynamical pre-calculating atmosphere<br />
processes in 4-D space// Preprint OSU, 2001, N1. —<br />
8p.<br />
33. Glushkov A.V., Khokhlov V.N., Tsenenko I.A. Atmospheric<br />
teleconnection patterns and eddy kinetic energy content:<br />
wavelet analysis// Nonlinear Processes in Geophysics. —<br />
2004. — V.11,N3. — P.285-293.<br />
34. Khokhlov V.N., Glushkov A.V., Loboda N.S., Bunyakova<br />
Yu. Ya., Short-range forecast of atmospheric pollutants<br />
using non-linear prediction method// Atmospheric Environment<br />
2008. — Vol.42. — P. 7284–7292..<br />
35. Ðóçîâ Â.ä., Ãëóøêîâ À.â., Âàñ÷åíêî Â.í., Àñòðîôèçè÷åñêàÿ<br />
ìîäåëü Çåìíîãî ãëîáàëüíîãî êëèìàòà Êèåâ.Íàóêîâà<br />
äóìêà 2005. — ñ 250.<br />
36. Rusov V.D., Glushkov A.V., Vaschenko V.N., Mavrodiev S.,<br />
Vachev B., Galactic cosmic rays -cloud effect and bifurcation<br />
UDÑ 539.19+539.182<br />
A. V. Glushkov<br />
model of Earth global climate// Bound Vol. of Observatorie<br />
Montagne de Moussalla. — 2007. — Vol.12. — P.80-90.<br />
37. Rusov V.N., Glushkov A.V., Loboda A., , On possible genesis<br />
of fractal dimensions in the turbulent pulsations of cosmic<br />
plasma — galactic-origin rays — turbulent pulsation in<br />
planetary atmosphere system// Advances in Space Research<br />
2008. — Vol.42,N9. — P.1614-1627.<br />
38. Glushkov A.V., Lovett L., Is it real to create an artificial superconductive<br />
nano-organic analog of the human brain//<br />
Preprint UK National Acad.of Sciences N HB-3, London,<br />
2007. — 15P.<br />
39. Glushkov A.V., New form of effective functional for account<br />
of the polarization effects in treating the π-electron states<br />
for organic molecules// Journ. Struct. Chem. — 1998. —<br />
Vol.34,N5. — P.12-19.<br />
40. Ceperley D.M., Monte-Carlo method in statistical physics.<br />
— N. — Y.: Plenum, 2000. — 650P.<br />
41. Glushkov A.V., , Monte-Carlo quantum chemistry of biogene<br />
amines. Laser and neutron capture effects// Theory and Applications<br />
of Computational Chemistry 2009. — Vol.1102. —<br />
P.131-150.<br />
QUASIPARTICLE ENERGY FUNCTIONAL FOR FINITE TEMPERATURES AND EFFECTIVE BOSE-CONDENSATE<br />
DYNAMICS: THEORY AND SOME ILLUSTRATIONS<br />
Abstract<br />
It is considered a theory of the quasiparticle energy functional under non-zeroth temperatures τ and some its applications. A thermodynamical<br />
potential for multielectron system in external stationary field for given τ is defined by dynamics of effective Bose-condensate<br />
in atoms of physical space of electrons. Structure of this space is defined by the cell system of surfaces of zeroth flux for entropy pulse<br />
under availability of the zeroth current of the bose-condensate density.<br />
Key words: density functional, effective Bose-condensate, atoms of physical space of electrons.<br />
ÓÄÊ 539.19+539.182<br />
À. Â. Ãëóøêîâ<br />
ÊÂÀÇÈ×ÀÑÒÈ×ÍÛÉ ÝÍÅÐÃÅÒÈ×ÅÑÊÈÉ ÔÓÍÊÖÈÎÍÀË ÏÐÈ ÊÎÍÅ×ÍÛÕ ÒÅÌÏÅÐÀÒÓÐÀÕ È ÄÈÍÀÌÈÊÀ<br />
ÝÔÔÅÊÒÈÂÍÎÃÎ ÁÎÇÅ-ÊÎÍÄÅÍÑÀÒÀ: ÒÅÎÐÈß È ÍÅÊÎÒÎÐÛÅ ÈËËÞÑÒÐÀÖÈÈ<br />
Ðåçþìå<br />
Ðàññìîòðåíû òåîðèÿ êâàçè÷àñòè÷íîãî ýíåðãåòè÷åñêîãî ôóíêöèîíàëà ïðè íåíóëåâîé òåìïåðàòóðå τ è å¸ íåêîòîðûå<br />
èëëþñòðàöèè. Òåðìîäèíàìè÷åñêèé ïîòåíöèàë ìíîãî-ýëåêòðîííîé ñèñòåìû âî âíåøíåì ñòàöèîíàðíîì ïîëå äëÿ äàííîé τ<br />
îïðåäåëÿåòñÿ äèíàìèêîé ýôôåêòèâíîãî ìíîãî÷àñòè÷íîãî áîçå-êîíäåíñàòà â àòîìàõ ôèçè÷åñêîãî ïðîñòðàíñòâà ýëåêòðîíîâ.<br />
Ñòðóêòóðà èñêîìîãî ïðîñòðàíñòâà îïðåäåëÿåòñÿ ÿ÷åèñòîé ñèñòåìîé ïîâåðõíîñòåé íóëåâîãî ïîòîêà èìïóëüñà ýíòðîïèè ïðè<br />
íàëè÷èè íóëåâîãî òîêà ïëîòíîñòè áîçå-êîíäåíñàòà.<br />
Êëþ÷åâûå ñëîâà: ôóíêöèîíàë ïëîòíîñòè, ýôôåêòèâíûé áîçå-êîíäåíñàò, àòîìû ôèçè÷åñêîãî ïðîñòðàíñòâà ýëåêòðîíîâ.<br />
ÓÄÊ 539.19+539.182<br />
Î. Â. Ãëóøêîâ<br />
ÊÂÀDz×ÀÑÒÈÍÊÎÂÈÉ ÅÍÅÐÃÅÒÈ×ÍÈÉ ÔÓÍÊÖ²ÎÍÀË ÏÐÈ ÑʲÍ×ÅÍÈÕ ÒÅÌÏÅÐÀÒÓÐÀÕ ² ÄÈÍÀ̲ÊÀ<br />
ÅÔÅÊÒÈÂÍÎÃÎ ÁÎÇÅ-ÊÎÍÄÅÍÑÀÒÓ: ÒÅÎÐ²ß ² ÄÅßʲ ²ËÞÑÒÐÀÖ²¯<br />
Ðåçþìå<br />
Ðîçãëÿíóòî òåîð³þ êâàç³-÷àñòèíêîâîãî åíåðãåòè÷íîãî ôóíêö³îíàëó ïðè íåíóëüîâ³é òåìïåðàòóð³ τ òà äåÿê³ ³¿ ³ëþñòðàö³¿.<br />
Òåðìîäèíàì³÷íèé ïîòåíö³àë áàãàòîåëåêòðîííî¿ ñèñòåìè ó çîâí³øíüîìó ñòàö³îíàðíîìó ïîë³ äëÿ äàíî¿ τ âèçíà÷àºòüñÿ<br />
äèíàì³êîþ åôåêòèâíîãî áàãàòî÷àñòèíêîâîãî áîçå-êîíäåíñàòó â àòîìàõ ô³çè÷íîãî ïðîñòîðó åëåêòðîí³â. Ñòðóêòóðà òàêîãî<br />
ïðîñòîðó âèçíà÷àºòüñÿ êîì³ðêîâîþ ñèñòåìîþ ïîâåðõîíü íóëüîâîãî ïîòîêó ³ìïóëüñó åíòðîﳿ ïðè íàÿâíîñò³ íóëüîâîãî òîêó<br />
ãóñòèíè áîçå-êîíäåíñàòó.<br />
Êëþ÷îâ³ ñëîâà: ôóíêö³îíàë ãóñòèíè, åôåêòèâíèé áîçå-êîíäåíñàò, àòîìè ô³çè÷íîãî ïðîñòîðó åëåêòðîí³â.<br />
69
70<br />
UDC 74.78.<br />
R. M. BALABAY, P. V. MERZLIKIN<br />
Krivoi Rog State Pedagogical University, Department of Physics, Krivoi Rog, Ukraine,<br />
phone:(0564)715721, e-mail: oks_pol@cabletv.dp.ua.<br />
ELECTRONIC STRUCTURE OF HETEROGENEOUS COMPOSITE: ORGANIC<br />
MOLECULE ON SILICON THIN FILM SURFACE<br />
Atomic composites such as molecule on silicon film substrate are interesting for molecular electronics<br />
and other applications. Therefore, the better understanding of mechanisms of interactions<br />
inside such systems is important. The space distribution of the valence electron density was calculated<br />
using Car-Parrinello molecular dynamic and ab initio norm-conserving pseudopotential. CH 4 N 2 O<br />
and CONH 5 molecules on silicon thin film (100) surface were examined.<br />
INTRODUCTION<br />
Nanoscience performs fundamental investigations<br />
of properties of nanomaterials and phenomena<br />
which appear in nanomaterials. Within the list of nanoobjects,<br />
the objects with lowered dimensionality are<br />
stood out. They are: quasi-2D electron gas (quantum<br />
wells), quasi-1D electron gas (quantum wires) and<br />
quasi-0D electron gas (quantum dots) [1]. Quantum<br />
wires are not observed spontaneously in nature and<br />
must be produced in a laboratory. Quantum wires<br />
could be created with the use of the semiconductor<br />
heterostructures [2] or through other way (for example<br />
with use of heterogeneous composition: organic molecule<br />
on silicon film surface).<br />
Atomic composites such as molecule on silicon<br />
film substrate are of great interest for a wide range<br />
of technological applications. For the latter, organic<br />
molecules are used to modify the electronic properties<br />
of metal and semiconductor surfaces.<br />
During the last few years there has been noticed<br />
the growth of interest from molecular electronics side<br />
stimulated largely by the experimental realization of<br />
molecular wires and systems where a single organic<br />
molecule or a few molecules carry an electric current<br />
between a pair of metal contacts [3, 4]. In some cases,<br />
such systems exhibit switching behavior and/or negative<br />
differential resistance [5, 6] phenomena that may<br />
be exploited in future molecular electronic devices.<br />
Hybride molecular/semiconductor nano-electronic<br />
devices are another intriguing possibility and fundamental<br />
research that may ultimately lead to their creation<br />
is also being pursued at the present time [7, 8,<br />
9, 10].<br />
SCOPE OF THE WORK<br />
The task of our research is the exploration of the<br />
possibility of forming of quantum wires in heterogeneous<br />
composites: organic molecule on silicon film<br />
and the topology of 1D electron gas channel created.<br />
The better understanding of mechanisms of interactions<br />
inside such systems is important for practical applications.<br />
One of the fundamental problems of semiconductor<br />
nanotechnology is the electron distribution<br />
and the electron move type in complicated geometry<br />
structures. It is very difficult to realize the experimental<br />
solution of this problem. Therefore, the ab initio<br />
calculation methods which not require experimental<br />
data are applicable the best.<br />
One of such methods is the electron density functional<br />
theory. The central place in this theory is occupied<br />
by self-consistent charge density which completely<br />
determines the system’s ground state. Besides<br />
presenting the information about the total energy,<br />
forces and stresses, the charge density is interesting itself<br />
and could often provide the illuminating insights<br />
which help to understand the physical properties of<br />
the solid. It opens a “window” for viewing the chemical<br />
bonds, it allows one to judge the nature of interatomic<br />
forces as well as to find the plausible interstitial<br />
positions. Therefore, the numerical ab initio calculation<br />
is required for detection particulars of forming<br />
of quantum wires in heterogeneous composition: organic<br />
molecule on silicon film. For this purpose, the<br />
Car-Parrinello molecular dynamic [11] and ab initio<br />
norm-conserving pseudopotential [12] realized within<br />
special software [13] were used.<br />
METHOD OF CALCULATION<br />
The Car-Parrinello method (CP method) is based<br />
on density-functional theory (DFT) and the Born-<br />
Oppenheimer (OB) adiabatic approximation. For<br />
conventional DFT electronic structure calculations,<br />
the Kohn-Sham (KS) equations are solved self-consistently.<br />
In the BO approximation, wave function<br />
are r considered<br />
r r as the functions of ionic positions<br />
{}:{ Rl ψi(<br />
r; Rl)}<br />
, but in the CP method, the { i ( ) } r<br />
ψ<br />
are treated as classical dynamical variables independent<br />
on { Rl} r<br />
. They are postulated to evolve by Newton’s<br />
equations of motion so that “dynamical simulated<br />
annealing” could be performed to search a global<br />
minimum of the electronic configuration.<br />
The Lagrangian in the CP method is introduced as<br />
r r<br />
r 2 r 1<br />
r<br />
2<br />
L{ ψψ , &, R, R&, } =μ∑∫ψ & i( r) dr + ∑MlR&<br />
l −<br />
i 2 l<br />
r<br />
, (1)<br />
r * r r<br />
E[{ ψ ( r)},{ R}] + ε ψ ( r) ψ ( r) dr−δ<br />
∑∑ ∫<br />
( )<br />
l l ij i j ij<br />
i j<br />
© R. M. Balabay, P. V. Merzlikin, 2009
where { ψ i}<br />
are the single-electron orbitals, and the<br />
electronic density pr ( )<br />
r is assumed to be given by<br />
r<br />
pr ( ) =<br />
r 2<br />
ψ ( r)<br />
. (2)<br />
∑<br />
i<br />
The first term of Eq. (1) is a fictitious classical mechanical<br />
kinetic energy of { ψ i}<br />
. The second term is<br />
an ionic kinetic energy, and E is the total energy (the<br />
sum of the electronic energy and the ion-ion Coulomb<br />
interaction energy). Lagrangian multipliers ε ij are introduced<br />
to satisfy the orthonormality constraints on<br />
{ ψ i}<br />
. The details of the electronic energy are described<br />
in a lot of literature in the field.<br />
From Eq. (2), equations of motion for i ψ and l Rr<br />
are derived as<br />
r δE<br />
r<br />
μψ i( r) = − ( )<br />
* r + ∑ε<br />
ijψj r (3)<br />
δψi<br />
( r ) j<br />
r<br />
MR&& l l =−∇lE<br />
(4)<br />
If an electronic structure reaches the state where<br />
no force acts on ψ i , that is, the left-hand side of Eq.<br />
(3) is equal to zero and this equation is identical to<br />
the KS equation and ψ i becomes the eigenstate of<br />
the KS equation. To attain this, the kinetic energy of<br />
{ ψ i}<br />
is gradually reduced until { ψ i}<br />
are frozen. This<br />
procedure is called “dynamical simulated annealing.”<br />
If one also relaxes the ions with Eq. (4), the minimization<br />
with respect to electronic and ionic configurations<br />
could be executed simultaneously.<br />
It is also possible to evolve { R i}<br />
and { ψ i}<br />
without<br />
reducing the kinetic energy. If { ψ i}<br />
are kept close<br />
to eigenstates during the time evolution, ionic trajectories<br />
generated by Eq. (4) are physically meaningful.<br />
When the CP method is applied to study the<br />
dynamical evolution of a system consisting of ions<br />
and electrons, it is called the “ab initio molecular dynamics.”<br />
THE CALCULATION RESULTS AND<br />
DISCUSSION<br />
Worked calculations allow to obtain the information<br />
on the details of electronic construction of some<br />
organic molecules (Fig.1 — Fig.2) and on the charges’<br />
redistribution in heterogeneous system: adsorbed organic<br />
molecule on silicon thin film (100) surface with<br />
4 Å thickness (Fig.3 — Fig.4).<br />
The density space partial distributions of valence<br />
electrons (taking into account only valence electrons)<br />
is determined by the theory of pseudopotential and<br />
by the fact that only the valence electrons undergo<br />
catastrophic changes under interactions for different<br />
iso-values, allow to determine the levels of hierarchy<br />
of connection between atoms in molecule CH 4 N 2 O<br />
(Fig.1), CONH 5 (Fig.2).<br />
As seen from Fig. 1 flat CH 4 N 2 O molecule has<br />
three weakly connected parts: CO, NH 2 , NH2, which<br />
contain most of electronic density of molecule. An<br />
absence of distinctly expressed spherical contour in<br />
distribution points leads to the absence of ionic and<br />
the presence of covalence polar (strong) and van der<br />
Waal’s (weak) types of the bounds.<br />
i<br />
Fig. 1. Density space partial distributions of valence electrons<br />
in CH4N2O molecule for iso-values: (a) 0.8-0.9 from maximal<br />
value, (b) 0.7-0.8 from maximal value, (c) 0.6-0.7 from maximal<br />
value, (d) 0.5-0.6 from maximal value, (e) cross section in molecule<br />
plane.<br />
71
Fig. 2. Density space partial distributions of valence electrons in CONH 5 molecule for iso-values: (a) 0.9-1.0 from maximal value<br />
(b) 0.8-0.9 from maximal value, (c) 0.7-0.8 from maximal value, (d) 0.6-0.7 from maximal value, (e) 0.5-0.6 from maximal value, (f)<br />
0.4-0.5 from maximal value, (g) 0.3-0.4 from maximal value(look-out over the two inversely situated molecules in primitive cell), (h)<br />
cross section of molecule in vicinity of O atom.<br />
A review of the density partial distributions of<br />
valence electrons in CONH 5 molecule for different<br />
iso-values allows to divide the two fragments in the<br />
molecule: ÑH 3 and NOH 2 . A part of electronic density<br />
inside these fragments is more than one in area<br />
between them what could be seen in cross section of<br />
molecule in the vicinity of O atom (Fig. 2(h)).<br />
For creation of the heterogeneous composite, the<br />
molecules are situated on distance of 1 Å from the film<br />
surface and 10.86 Å from each other. The film thickness<br />
is 4 Å. On both surfaces of film, the two molecules<br />
are situated with inversion symmetry in respect<br />
to atomic system’s center of symmetry (Fig. 3(a)).<br />
The calculation algorithm means the use of periodic<br />
space lattice with atomic basis (it reflects the features<br />
of the investigating system) which certainly ought to<br />
have the inverse symmetry. The atomic basis of the<br />
primitive cubic unit cell of the superlattice which rep-<br />
72<br />
resents thin silicon film with two (100) surfaces with<br />
CH 4 N 2 O or CONH 5 molecule on each side of film<br />
and consisted of 48 atoms (32 of them are Si atoms<br />
and 16 atoms of two molecules). In such a way, there is<br />
the space-periodic system of molecules on silicon film<br />
surface. Maps of density of valence electrons, shown<br />
on Fig.3-4, demonstrate the interaction between electrons<br />
of thin silicon film and the molecules considered<br />
above. The analysis of these distributions allows to<br />
segregate the two types of electronic system realignment:<br />
one concerned with reaction of molecules with<br />
surface and other concerned with molecules’ influence<br />
of the formation of the new quantum objects in<br />
silicon film: i.e. the electronic wires. More distinct<br />
space realignment of electrons with quantum wires<br />
(with thickness about 1 Å) formation is observed under<br />
flat molecule CH 4 N 2 O more clear than under the<br />
non- flat CONH 5 .
Fig. 3. Density space partial distributions of valence electrons in heterogeneous composition: two inversely situated symmetric molecules<br />
CH 4 N 2 O on (100) silicon surface for iso-values: (a) 0.9-1.0 from maximal value, (b) (110) cross section of primitive cell.<br />
Fig. 4. Density space partial distributions of valence electrons in heterogeneous composition: two inversely situated symmetric<br />
molecules CONH 5 on (100) silicon surface for iso-values: (a) 0.8-0.9 from maximal value, (b) 0.7-0.8 from maximal value, (c) 0.6-0.7<br />
from maximal value, (d) 0.5-0.6 from maximal value, (e) 0.4-0.5 from maximal value, (f) 0.3-0.4 from maximal value, (g) (110) cross<br />
section of primitive cell.<br />
73
74<br />
CONCLUSIONS<br />
1. The analysis of valence electrons’ density allows<br />
to affirm the creation of space-periodic arrays of<br />
quantum wires in silicon film obtained by periodic deposition<br />
flat organic molecules onto film surface, for<br />
instance the CH N O molecule.<br />
4 2<br />
2. Precise deposition of the flat molecules onto<br />
the surface is quite realistic considering the modern<br />
technologies and these composites will play the role<br />
of the gate under which the conductive 1D dimension<br />
channel is formed.<br />
3. The atomic heterogeneous composite could be<br />
used to develop the next generation of computing devices<br />
with switching mechanisms of signal amplification.<br />
References<br />
1. Øèê À. ß., Áàêàåâà Ë. Ã., Ìóñèõèí Ñ. Ô., Ðûêîâ Ñ. À.<br />
Ôèçèêà íèçêî ðàçìåðíûõ ñèñòåì. — Ñ. — Ïá: Íàóêà,<br />
2001. — 160ñ.<br />
2. Óàéòñàéäå Äæ., Ýéãëåð Ä., Àíäåðñ Ð. è äð. Íàíîòåõíîëîãèè<br />
â áëèæàéøåì äåñÿòèëåòèè/Ïîä ðåä. Ðîêî Ì. Ê., Óèëüÿìñà<br />
Ð. Ñ., Àëèâèñàòîñà Ï. Ïðîãíîç íàïðàâëåíèé. —<br />
Ì: Ìèð. — 2002. — 293ñ.<br />
3. L. A. Bumm, J. J. Arnold, M. T. Cygan, T. D. Dunbar,<br />
T. P. Burgin, L. Jones II, D. L. Allara, J. M. Tour, and<br />
P. S. Weiss. Are Single Molecular Wires Conducting // Science.<br />
— 1996. — Vol. 271. — pp. 1705 — 1707.<br />
4. M. A. Reed, C. Zhou, C. J. Muller, T. P. Burgin, and J. M.<br />
Tour. Conductance of a Molecular Junction // Science. —<br />
1997. — Vol. 278. — pp. 252 — 254.<br />
UDC 74.78.73.21<br />
R. M. Balabay , P. V. Merzlikin<br />
5. J. Chen, M. A. Reed, A. M. Rawlett and J. M. Tour. Large<br />
On-Off Ratios and Negative Differential Resistance in a Molecular<br />
Electronic Device // Science. — 1999. — Vol. 286. —<br />
pp. 1550 — 1552.<br />
6. C. P. Collier, G. Mattersteig, E. W. Wong, Y. Luo, K. Beverly,<br />
J. Sampaio, F. M. Raymo, J. F. Stoddart and J. R. Heath. A<br />
[2]Catenane-Based Solid State Electronically Reconfigurable<br />
Switch // Science. — 2000. — Vol. 289. — pp. 1172 — 1175.<br />
7. R. A. Wolkow. Controlled Molecular Aadsorption on Silicon:<br />
Laying a Foundation for Molecular Devices // Annual Review<br />
of Physical Chemistry. – 1999. — Vol. 50. — pp. 413 — 441.<br />
8. G. P. Lopinski, D.D.M. Wayner, R.A. Wolkow. Self-directed<br />
growth of molecular nanostructures on silicon // Nature. —<br />
2000. — Vol. 406. — pp. 48 — 51.<br />
9. J. — H. Cho, D. — H. Oh, and L. Kleinman. One-dimensional<br />
molecular wire on hydrogenated Si(001) // Phys.<br />
Rev. B. — 2002. — Vol. 65. — 310.<br />
10. W. A. Hofer, A. J. Fisher, G. P. Lopinski, R. A. Wolkow.<br />
Electronic structure and STM images of self-assembled styrene<br />
lines on a Si(100) surface // Chemical physics letters —<br />
2002. — Vol. 365. — pp. 129-134.<br />
11. Marx D., Hutter J. Ab initio molecular dynamics: theory and<br />
implementation”, published in “Modern methods and algorithms<br />
of quantum chemistry”, J. Grotendorst (Ed.), John<br />
von Neuman Institute for computing, Julich, // NIC Series,<br />
v. 1, ISBN 3-00-005618-1, 2000, pp. 301-449<br />
12. Hartwigsen C., Goedecker S., Hutter J.. Relativistic separable<br />
dual-space Gaussian pseudopotentils from H to Rn.//<br />
Phys. Rev. B, v. 58, N. 7, 1998, pp. 3641-3662<br />
13. Áàëàáàé Ð.Ì., Ãðèùåíêî Í.Â. Ïðîãðàììíîå îáåñïå÷åíèå<br />
äëÿ ðàñ÷åòîâ ñ íà÷àëà òâåðäîòåëüíûõ ñòðóêòóð //<br />
Ìàòåð³àëè ̳æíàðîäíî¿ íàóêîâî-ïðàêòè÷íî¿ êîíôåðåíö³¿<br />
“Ïðîáëåìè åëåêòðîííî¿ ïðîìèñëîâîñò³ ó Ïåðåõ³äíèé<br />
ïåðèîä”. — Çá³ðíèê íàóêîâèõ ïðàöü Ñõ³äíîóêðà¿íñüêîãî<br />
äåðæàâíîãî óí³âåðñèòåòó. — Ëóãàíñüê. — 1998. —<br />
ñ.124-128.<br />
ELECTRONIC STRUCTURE OF HETEROGENEOUS COMPOSITE: ORGANIC MOLECULE ON SILICON THIN FILM<br />
SURFACE<br />
Abstract<br />
Atomic composites such as molecule on silicon film substrate are interesting for molecular electronics and other applications.<br />
Therefore the better understanding of mechanisms of interactions inside such systems is important. Space distribution of the valence<br />
electron density was calculated using Car-Parrinello molecular dynamic and ab initio norm-conserving pseudopotential. CH 4 N 2 O and<br />
CONH 5 molecules on silicon thin film (100) surface were examined.<br />
Key words: heterogeneous composite, electronic structure, film surface.<br />
ÓÄÊ 74.78 73.21<br />
Ð. Ì. Áàëàáàé, Ï. Â. Ìåðçëèêèí<br />
ÝËÅÊÒÐÎÍÍÀß ÑÒÐÓÊÒÓÐÀ ÃÅÒÅÐÎÃÅÍÍÎÉ ÊÎÌÏÎÇÈÖÈÈ: ÎÐÃÀÍÈ×ÅÑÊÈÅ ÌÎËÅÊÓËÛ ÍÀ<br />
ÏÎÂÅÐÕÍÎÑÒÈ ÒÎÍÊÎÉ ÏËÅÍÊÈ ÊÐÅÌÍÈß<br />
Ðåçþìå<br />
Èçó÷àåòñÿ âîçìîæíîñòü ôîðìèðîâàíèÿ êâàíòîâûõ íèòåé â ãåòåðîãåííîé êîìïîçèöèè: îðãàíè÷åñêèå ìîëåêóëû íà ïîâåðõíîñòè<br />
òîíêîé ïëåíêè êðåìíèÿ.<br />
Èíôîðìàöèþ î ôèçè÷åñêèõ ñâîéñòâàõ òàêèõ ñèñòåì ìîæíî èçâëå÷ü èç ðàñïðåäåëåíèÿ ýëåêòðîííîé ïëîòíîñòè. Äëÿ ðàñ-<br />
÷åòà òàêîãî ðàñïðåäåëåíèÿ èñïîëüçîâàëàñü êâàíòîâî-ìåõàíè÷åñêàÿ ìîëåêóëÿðíàÿ äèíàìèêà Êàð-Ïàðèíåëëî (Car-Parrinello).<br />
Ðàñ÷åò ïðîâîäèëñÿ äëÿ ìîëåêóë CH 4 N 2 O è CONH 5 íà ïîâåðõíîñòè Si(100).<br />
Àíàëèç ðàñïðåäåëåíèé ýëåêòðîííîé ïëîòíîñòè ïîçâîëÿåò ãîâîðèòü î ïîëó÷åíèè ïðîñòðàíñòâåííî ïåðèîäè÷åñêèõ ìàññèâîâ<br />
êâàíòîâûõ íèòåé â êðåìíèåâîé ïëåíêå ïðè ïåðèîäè÷åñêîì íàíåñåíèè íà ïîâåðõíîñòü ïëåíêè ïëîñêèõ îðãàíè÷åñêèõ<br />
ìîëåêóë, íàïðèìåð CH 4 N 2 O.<br />
Êëþ÷åâûå ñëîâà: ãåòåðîãåííàÿ êîìïîçèöèèÿ, îðãàíè÷åñêèå ìîëåêóëû ïîâåðõíîñòü ïëåíêè.
ÓÄÊ 74.78 73.21<br />
Ð. Ì. Áàëàáàé, Ï. Â. Ìåðçëèê³í<br />
ÅËÅÊÒÐÎÍÍÀ ÑÒÐÓÊÒÓÐÀ ÃÅÒÅÐÎÃÅÍÍί ÊÎÌÏÎÇÈÖ²¯: ÎÐÃÀͲ×Ͳ ÌÎËÅÊÓËÈ ÍÀ ÏÎÂÅÐÕͲ ÒÎÍÊί<br />
Ï˲ÂÊÈ ÊÐÅÌͲÞ<br />
Ðåçþìå<br />
Âèâ÷àºòüñÿ ìîæëèâ³ñòü ôîðìóâàííÿ êâàíòîâèõ äðîò³â â ãåòåðîãåíí³é êîìïîçèö³¿: îðãàí³÷í³ ìîëåêóëè íà ïîâåðõí³ òîíêî¿<br />
ïë³âêè êðåìí³þ.<br />
²íôîðìàö³þ ïðî ô³çè÷í³ âëàñòèâîñò³ òàêèõ ñèñòåì ìîæíà îòðèìàòè ç ðîçïîä³ëó åëåêòðîííî¿ ãóñòèíè. Äëÿ ðîçðàõóíêó<br />
òàêîãî ðîçïîä³ëó âèêîðèñòîâóâàëàñü êâàíòîâî-ìåõàí³÷íà ìîëåêóëÿðíà äèíàì³êà Êàð-Ïàð³íåëëî (Car-Parrinello). Ðîçðàõóíîê<br />
ïðîâîäèâñÿ äëÿ ìîëåêóë CH 4 N 2 O òà CONH 5 íà ïîâåðõí³ Si(100).<br />
Àíàë³ç ðîçïîä³ë³â åëåêòðîííî¿ ãóñòèíè äîçâîëÿº ãîâîðèòè ïðî îòðèìàííÿ ïðîñòîðîâî ïåð³îäè÷íèõ ìàñèâ³â êâàíòîâèõ íèòîê<br />
â êðåìí³ºâ³é ïë³âö³ ïðè ïåð³îäè÷íîìó íàíåñåíí³ íà ïîâåðõíþ ïë³âêè ïëîñêèõ îðãàí³÷íèõ ìîëåêóë, íàïðèêëàä CH 4 N 2 O.<br />
Êëþ÷îâ³ ñëîâà: ãåòåðîãåííà êîìïîçèö³ÿ, îðãàí³÷í³ ìîëåêóëè, ïîâåðõíÿ ïë³âêè.<br />
75
Many physical and radiotechnical systems —<br />
multielement semiconductors and gas lasers, different<br />
radiotechnical devices, etc can be considered<br />
in the first approximation as set of autogenerators,<br />
coupled by different way (c.f.[1,2]). Scheme of two<br />
autogenerators (semiconductor quantum generators<br />
(1), coupled by means optical waveguide (2), is<br />
presented been in figure 1. An important feature of<br />
these systems is connected with possibility of realizing<br />
so called synchronic (sinphase) regimes of auto<br />
oscillations, when relative phases of oscillations of<br />
different elements are fixed. Another important feature<br />
is realizing the stochastic regime of oscillations<br />
and chaos elements. In refs.[1-4] it has been numerically<br />
studied a regular and chaotic dynamics of<br />
the system of the Van-der-Poll autogenerators with<br />
account of finiteness of the signals propagation time<br />
between them and also with special kind of interoscillators<br />
interaction forces. Chaos theory establishes<br />
that apparently complex irregular behaviour<br />
could be the outcome of a simple deterministic system<br />
with a few dominant nonlinear interdependent<br />
variables.<br />
Fig. 1. The grid of autogenerators (semiconductor quantum<br />
generators (SQG), coupled by means of the general resonator:<br />
1 — SQG, 2 — resonator (dielectric plate)<br />
The past decade has witnessed a large number of<br />
studies employing the ideas gained from the science of<br />
chaos to characterize, model, and predict the dynamics<br />
of various geophysical phenomena (c.f.[1-20]).<br />
The outcomes of such studies are very encouraging, as<br />
they not only revealed that the dynamics of the apparently<br />
irregular phenomena could be understood from<br />
a chaotic deterministic point of view but also reported<br />
76<br />
UDÑ 539.124 : 541.47<br />
A. A. SVINARENKO, A. V. LOBODA, N. G. SERBOV<br />
Odessa State Environmental University<br />
MODELING AND DIAGNOSTICS OF INTERACTION OF THE NON-<br />
LINEAR VIBRATIONAL SYSTEMS ON THE BASIS OF TEMPORAL SERIES<br />
(APPLICATION TO SEMICONDUCTOR QUANTUM GENERATORS)<br />
It is studied an employing a variety of techniques for characterizing the dynamics of the coupled<br />
semiconductor quantum generators and identifying the presence of chaotic dynamics in this time (frequency)<br />
series. The statistical techniques used are the autocorrelation function and the Fourier power<br />
spectrum, whereas the mutual information approach, correlation integral analysis, false nearest neighbour<br />
algorithm, Lyapunov exponents analysis, and surrogate data method are used for comprehensive<br />
characterization.<br />
very good predictions using such an approach for different<br />
systems.<br />
The present study attempts to employ a variety<br />
of techniques for characterizing the dynamics of the<br />
coupled semiconductor quantum generators (autogenerators)<br />
More specifically, we attempt to identify the<br />
possible presence of chaotic dynamics in this time (frequency)<br />
series. The techniques employed range from<br />
standard statistical techniques that can provide general<br />
indications regarding the dynamics of the phenomenon<br />
to specific ones that can provide comprehensive characterization<br />
of the dynamics. The statistical techniques<br />
used (c.f.[17-20]) are the autocorrelation function and<br />
the Fourier power spectrum, whereas the mutual information<br />
approach, the correlation integral analysis, the<br />
false nearest neighbour algorithm, the Lyapunov exponents<br />
analysis, and the surrogate data method are employed<br />
for comprehensive characterization.<br />
2. INVESTIGATION OF CHAOS IN THE<br />
VIBRATION DYNAMICS<br />
In ref.[2-4,19] it has been studied a regular and<br />
chaotic dynamics of the system of the Van-der-Poll<br />
autogenerators with account of finiteness of the signals<br />
propagation time between them and also with special<br />
kind of inter-oscillators interaction forces. The cases<br />
of little and large non-linearity in the system were<br />
considered. Phase diagram for system of two coupled<br />
autogenerators, interacting with retardion, has been<br />
obtained and the regions, where single-frequency sinphase<br />
oscillation regime (1), multi-frequency sinphase<br />
one (2), chaotic one (3) are realized, were found.. In<br />
general the same situation takes a place in a case of the<br />
grid of autogenerators (semiconductor quantum generators,<br />
coupled by means of the general resonator. In<br />
ref. [3,4,19] we have carried out an analysis of oscillations<br />
in system (Fig.2) for a grid of the coupled (N>2)<br />
semiconductor quantum generators (autogenerators).<br />
This case is more complicated in comparison with a<br />
case of the system of two coupled autogenerators.<br />
Figure 1 shows the vibration dynamics time series<br />
for a grid of autogenerators for the time interval<br />
t=nτ (τ=25[2π/w 0 ]). As it can be seen in Fig. 1, the<br />
© A. A. Svinarenko, A. V. Loboda, N. G. Serbov, 2009
systems exhibits significant variations without any apparent<br />
cyclicity. It is clear that a visual inspection of<br />
the (irregular) amplitude level series does not provide<br />
any clues regarding its dynamical behaviour, whether<br />
chaotic or stochastic. To detect some regularity (or irregularity)<br />
in the time series, the Fourier power spectrum<br />
is often analyzed.<br />
For a purely random process, the power spectrum<br />
oscillates randomly about a constant value, indicating<br />
that no frequency explains any more of the variance of<br />
the sequence than any other frequency. For a periodic<br />
or quasi-periodic sequence, only peaks at certain frequencies<br />
exist; measurement noise adds a continuous<br />
floor to the spectrum. Chaotic signals may also have<br />
sharp spectral lines but even in the absence of noise<br />
there will be continuous part (broadband) of the spectrum.<br />
The broad power spectrum falling as a power of<br />
frequency is a first indication of chaotic behaviour,<br />
though it alone does not characterize chaos [5,15,18].<br />
From this point of view, the corresponding series analyzed<br />
in this study is presumably chaotic [8] However,<br />
more well-defined conclusion on the dynamics of the<br />
time series can be made after the data will be treated<br />
by methods of chaos theory.<br />
Let us consider scalar measurements s(n) = s(t 0 +<br />
nΔt) = s(n), where t 0 is the start time, Δt is the time<br />
step, and is n the number of the measurements. In a<br />
general case, s(n) is any time series, particularly the<br />
amplitude level.<br />
3<br />
2<br />
1<br />
0<br />
-1<br />
-2<br />
-3<br />
62 64 66 68 70 72 74 76 78 80 82 84 86 88 90 92 94 96 98 100 102 104<br />
Fig. 2. The vibration dynamics time series for a grid of autogenerators<br />
Since processes resulting in the chaotic behaviour<br />
are fundamentally multivariate, it is necessary to reconstruct<br />
phase space using as well as possible information<br />
contained in the s(n). Such a reconstruction<br />
results in a certain set of d-dimensional vectors y(n)<br />
replacing the scalar measurements. Packard et al. [7]<br />
introduced the method of using time-delay coordinates<br />
to reconstruct the phase space of an observed dynamical<br />
system. The main idea is that the direct use of<br />
the lagged variables s(n + τ), where τ is some integer to<br />
be determined, results in a coordinate system in which<br />
the structure of orbits in phase space can be captured.<br />
Then using a collection of time lags to create a vector<br />
in d dimensions,<br />
y(n) = [s(n), s(n + τ), s(n + 2τ), …,<br />
s(n + (d−1)τ)], (1)<br />
the required coordinates are provided. In a nonlinear<br />
system, the s(n + jτ) are some unknown nonlinear<br />
combination of the actual physical variables that comprise<br />
the source of the measurements. The dimension<br />
d is also called the embedding dimension, d E . The example<br />
of the Lorenz attractor given by Abarbanel et<br />
al. [5,6] is a good choice to illustrate the efficiency of<br />
the method.<br />
3.1. CHOOSING TIME LAG<br />
The statement of Mañé [13] and Takens [12] that<br />
any time lag will be acceptable is not terribly useful for<br />
extracting physics from data. If τ is chosen too small,<br />
then the coordinates s(n + jτ) and s(n + (j + 1)τ) are<br />
so close to each other in numerical value that they<br />
cannot be distinguished from each other. Similarly, if τ<br />
is too large, then s(n + jτ) and s(n + (j + 1)τ) are completely<br />
independent of each other in a statistical sense.<br />
Also, if τ is too small or too large, then the correlation<br />
dimension of attractor can be under- or overestimated<br />
respectively [8,18]. It is therefore necessary to choose<br />
some intermediate (and more appropriate) position<br />
between above cases.<br />
First approach is to compute the linear autocorrelation<br />
function<br />
N 1<br />
∑[(<br />
sm+δ) −s][( sm) −s]<br />
N m=<br />
1<br />
CL<br />
( δ ) =<br />
, (2)<br />
N 1<br />
2<br />
∑[(<br />
sm) − s]<br />
N m=<br />
1<br />
N 1<br />
where s = ∑ s( m)<br />
, and to look for that time lag<br />
N m=<br />
1<br />
where C (Δ) first passes through zero (see [18]). This<br />
L<br />
gives a good hint of choice for τ at that s(n + jτ) and<br />
s(n + (j + 1)τ) are linearly independent. However, a<br />
linear independence of two variables does not mean<br />
that these variables are nonlinearly independent since<br />
a nonlinear relationship can differs from linear one.<br />
It is therefore preferably to utilize approach with a<br />
nonlinear concept of independence, e.g. the average<br />
mutual information.<br />
Briefly, the concept of mutual information can<br />
be described as follows. Let there are two systems, A<br />
and B, with measurements a and b . The amount one<br />
i k<br />
learns in bits about a measurement of a from a mea-<br />
i<br />
surement of b is given by the arguments of informa-<br />
k<br />
tion theory [9] as<br />
⎛ PAB ( ai, b ) ⎞ k<br />
IAB ( ai, bk)<br />
= log2⎜<br />
⎟,<br />
(3)<br />
⎝PA( ai) PB( bk)<br />
⎠<br />
where the probability of observing a out of the set of<br />
all A is P (a ), and the probability of finding b in a<br />
A i<br />
measurement B is P (b ), and the joint probability of<br />
B i<br />
the measurement of a and b is P (a , b ). The mutual<br />
AB i k<br />
information I of two measurements a and b is sym-<br />
i k<br />
metric and non-negative, and equals to zero if only<br />
the systems are independent. The average mutual information<br />
between any value a from system A and b i k<br />
from B is the average over all possible measurements<br />
of I (a , b ),<br />
AB i k<br />
I () τ =∑ P ( a , b ) I ( a , b ) . (4)<br />
AB AB i k AB i k<br />
ai, bk<br />
To place this definition to a context of observations<br />
from a certain physical system, let us think of the<br />
77
sets of measurements s(n) as the A and of the measurements<br />
a time lag τ later, s(n + τ), as B set. The average<br />
mutual information between observations at n and<br />
n + τ is then<br />
78<br />
I () τ =∑ P ( a , b ) I ( a , b ) . (5)<br />
AB AB i k AB i k<br />
ai, bk<br />
Now we have to decide what property of I(τ) we<br />
should select, in order to establish which among the<br />
various values of τ we should use in making the data<br />
vectors y(n). In ref. [11] it has been suggested, as a prescription,<br />
that it is necessary to choose that τ where<br />
the first minimum of I(τ) occurs. Figure 3a presents<br />
the variations of the autocorrelation coefficient for the<br />
amplitude level.<br />
Fig. 3 (a) Autocorrelation function and (b) average mutual<br />
information<br />
As it can be seen, the autocorrelation function exhibits<br />
some kind of exponential decay up to a lag time<br />
of about 100 time units (sec). Such an exponential<br />
decay can be an indication of the presence of chaotic<br />
dynamics in the process of the level variations. On the<br />
other hand, the autocorrelation coefficient failed to<br />
achieve zero, i.e. the autocorrelation function analysis<br />
not provides us with any value of τ. Such an analysis<br />
can be certainly extended to values exceeding 1000,<br />
but it is known [15] that an attractor cannot be ad-<br />
equately reconstructed for very large values of τ. Figure<br />
3b shows the variation of the mutual information<br />
function against the lag time. The mutual information<br />
function exhibits an initial rapid decay (up to a lag<br />
time of about 10) followed more slow decrease before<br />
attaining near-saturation at the first minimum. Thus,<br />
we can use in following investigations the value of τ<br />
equals to 40 that is obtained by using the average mutual<br />
information analysis.<br />
Let us also note that the autocorrelation function<br />
and average mutual information ca be to some extent<br />
considered as analogues of the linear redundancy and<br />
general redundancy, respectively, which was applied<br />
in the test for nonlinearity. If a time series under consideration<br />
have an n-dimensional Gaussian distribution,<br />
these statistics are theoretically equivalent as it is<br />
shown by Paluš (see [15]). The general redundancies<br />
detect all dependences in the time series, while the linear<br />
redundancies are sensitive only to linear structures.<br />
Although we do not perform the test for nonlinearity<br />
of Paluš in full, the simple comparison of the curves<br />
in Figs. 3a and 3b shows that most of features observed<br />
in the autocorrelation function values are missing in<br />
the average mutual information. In other words, the<br />
nature of curves in Figs. 3a and 3b is substantially<br />
different. From this fact, a possible nonlinear nature<br />
of process resulting in the vibrations amplitude level<br />
variations can be concluded.<br />
3.2. CHOOSING EMBEDDING DIMENSION<br />
The goal of the embedding dimension determination<br />
is to reconstruct a Euclidean space R d large<br />
enough so that the set of points d A can be unfolded<br />
without ambiguity. In accordance with the embedding<br />
theorem, the embedding dimension, d E , must<br />
be greater, or at least equal, than a dimension of attractor,<br />
d A , i.e. d E > d A . In other words, we can choose<br />
a fortiori large dimension d E , e.g. 10 or 15, since the<br />
previous analysis provides us prospects that the dynamics<br />
of our system is probably chaotic. However,<br />
two problems arise with working in dimensions larger<br />
than really required by the data and time-delay embedding<br />
[5,6,18] . First, many of computations for<br />
extracting interesting properties from the data require<br />
searches and other operations in R d whose computational<br />
cost rises exponentially with d. Second, but<br />
more significant from the physical point of view, in<br />
the presence of noise or other high dimensional contamination<br />
of the observations, the extra dimensions<br />
are not populated by dynamics, already captured by<br />
a smaller dimension, but entirely by the contaminating<br />
signal. In too large an embedding space one is<br />
unnecessarily spending time working around aspects<br />
of a bad representation of the observations which are<br />
solely filled with noise. It is therefore necessary to determine<br />
the dimension d A .<br />
There are several standard approaches to reconstruct<br />
the attractor dimension (see, e.g., [5,6,15]), but<br />
let us consider in this study two methods only. The correlation<br />
integral analysis is one of the widely used techniques<br />
to investigate the signatures of chaos in a time<br />
series. The analysis uses the correlation integral, C(r),<br />
to distinguish between chaotic and stochastic systems.
To compute the correlation integral, the algorithm of<br />
Grassberger and Procaccia [10] is the most commonly<br />
used approach. According to this algorithm, the correlation<br />
integral is computed as<br />
2<br />
Cr () = lim H( r−||<br />
i − j || )<br />
N →∞ Nn ( −1) ∑ y y , (6)<br />
i, j<br />
(1 ≤< i j≤N) where H is the Heaviside step function with H(u) = 1<br />
for u > 0 and H(u) = 0 for u ≤ 0, r is the radius of<br />
sphere centered on y or y , and N is the number of data<br />
i j<br />
measurements. If the time series is characterized by an<br />
attractor, then the correlation integral C(r) is related to<br />
the radius r given by<br />
log Cr ( )<br />
d = lim , (7)<br />
r→0<br />
log r<br />
N→∞<br />
where d is correlation exponent that can be determined<br />
as the slop of line in the coordinates log C(r)<br />
versus log r by a least-squares fit of a straight line over<br />
a certain range of r, called the scaling region. If the<br />
correlation exponent attains saturation with an increase<br />
in the embedding dimension, then the system is<br />
generally considered to exhibit chaotic dynamics. The<br />
saturation value of the correlation exponent is defined<br />
as the correlation dimension (d ) of the attractor. The<br />
2<br />
nearest integer above the saturation value provides the<br />
minimum or optimum embedding dimension for reconstructing<br />
the phase-space or the number of variables<br />
necessary to model the dynamics of the system.<br />
On the other hand, if the correlation exponent increases<br />
without bound with increase in the embedding<br />
dimension, the system under investigation is generally<br />
considered stochastic. In this study, the correlation<br />
functions and the exponents was computed for the<br />
hourly amplitude level. Figure 4 shows the correlation<br />
dimension results, i.e. the relationship between<br />
the correlation exponent and embedding dimension<br />
values.<br />
Fig. 4. Relationship between correlation exponent and embedding<br />
dimension for vibrations amplitude level data for original<br />
time series (line 1), mean values of surrogate data sets (line 2),<br />
and one surrogate realization (line 3). Error bars indicate minimal<br />
values of correlation exponent among all realizations of surrogate<br />
data.<br />
As it can be seen, the correlation exponent value<br />
increases with embedding dimension up to a certain<br />
value, and then saturates beyond that value.<br />
The saturation of the correlation exponent beyond a<br />
certain embedding dimension is an indication of the<br />
existence of deterministic dynamics. The saturation<br />
value of the correlation exponent, i.e. correlation dimension<br />
of attractor, for the amplitude level series<br />
is about 3.5 and occurs at the embedding dimension<br />
value of 6. The low, non-integer correlation dimension<br />
value indicates the existence of low-dimensional<br />
chaos in the vibrations dynamics of the autogenerators.<br />
.<br />
The nearest integer above the correlation dimension<br />
value can be considered equal to the minimum<br />
dimension of the phase-space essential to embed the<br />
attractor. The value of the embedding dimension at<br />
which the saturation of the correlation dimension occurs<br />
is considered to provide the upper bound on the<br />
dimension of the phase-space sufficient to describe the<br />
motion of the attractor. Furthermore, the dimension<br />
of the embedding phase-space is equal to the number<br />
of variables present in the evolution of the system dynamics.<br />
Therefore, the results from the present study<br />
indicate that to model the dynamics of process resulting<br />
in the amplitude level variations the minimum<br />
number of variables essential is equal to 4 and the<br />
number of variables sufficient is equal to 6. Therefore,<br />
the amplitude level attractor should be embedded at<br />
least in a four-dimensional phase-space. The results<br />
indicate also that the upper bound on the dimension<br />
of the phase-space sufficient to describe the motion of<br />
the attractor, and hence the number of variables sufficient<br />
to model the dynamics of process resulting in<br />
the level variations is equal to 6.<br />
There are certain important limitations in the use<br />
of the correlation integral analysis in the search for<br />
chaos. For instance, the selection of inappropriate<br />
values for the parameters involved in the method may<br />
result in an underestimation (or overestimation) of the<br />
attractor dimension [8]. Consequently, finite and low<br />
correlation dimensions could be observed even for a<br />
stochastic process [18]. To verify the results obtained<br />
by the correlation integral analysis, we use surrogate<br />
data method.<br />
The method of surrogate data [16] is an approach<br />
that makes use of the substitute data generated in accordance<br />
to the probabilistic structure underlying<br />
the original data. This means that the surrogate data<br />
possess some of the properties, such as the mean,<br />
the standard deviation, the cumulative distribution<br />
function, the power spectrum, etc., but are otherwise<br />
postulated as random, generated according to a specific<br />
null hypothesis. Here, the null hypothesis consists<br />
of a candidate linear process, and the goal is to<br />
reject the hypothesis that the original data have come<br />
from a linear stochastic process. One reasonable statistics<br />
suggested by Theiler et al. [16] is obtained as<br />
follows.<br />
Let Q orig denote the statistic computed for the original<br />
time series and Q si for the ith surrogate series generated<br />
under the null hypothesis. Let μ s and σ s denote,<br />
respectively, the mean and standard deviation of the<br />
distribution of Q s . Then the measure of significance S<br />
is given by<br />
79
| Qorig<br />
−μs<br />
|<br />
S =<br />
. (8)<br />
σs<br />
An S value of ∼2 cannot be considered very significant,<br />
whereas an S value of ∼10 is highly significant<br />
[16]. The details on the null hypothesis and surrogate<br />
data generation are described in ref. [18].<br />
To detect nonlinearity in the amplitude level data,<br />
the one hundred realizations of surrogate data sets were<br />
generated according to a null hypothesis in accordance<br />
to the probabilistic structure underlying the original<br />
data. The correlation integrals and the correlation exponents,<br />
for embedding dimension values from 1 to 20,<br />
were computed for each of the surrogate data sets using<br />
the Grassberger-Procaccia algorithm as explained earlier.<br />
Figure 4 shows the relationship between the correlation<br />
exponent values and the embedding dimension<br />
values for the original data set and mean values of the<br />
surrogate data sets as well as for one surrogate realization.<br />
It is interesting to note that the similar picture<br />
takes a place in the principally other geophysical system<br />
[17], that confirms the fundamental idea about genesis<br />
of the fractal dimensions and chaotic features in physically<br />
different systems. One can be stressed, however<br />
the similar features are manifested in such complicated<br />
system as “cosmic plasma — galactic-origin rays — turbulent<br />
pulsation in planetary atmosphere system” [20].<br />
As it can be seen from Fig. 4, a significant difference<br />
in the estimates of the correlation exponents, between<br />
the original and surrogate data sets, is observed. In the<br />
case of the original data, a saturation of the correlation<br />
exponent is observed after a certain embedding<br />
dimension value (i.e., 6), whereas the correlation exponents<br />
computed for the surrogate data sets continue<br />
increasing with the increasing embedding dimension.<br />
The significance values (S) of the correlation exponent<br />
are computed for each embedding dimension and are<br />
shown in Fig. 5. The significance values lie mostly in<br />
the range between 10 and 50. The high significance<br />
values of the statistic indicate that the null hypothesis<br />
(the data arise from a linear stochastic process) can be<br />
rejected and hence the original data might have come<br />
from a nonlinear process.<br />
Fig. 5. Relationship between significance values of correlation<br />
dimension and embedding dimension<br />
80<br />
Let us consider another method for determining<br />
d that comes from asking the basic question addressed<br />
E<br />
in the embedding theorem: when has one eliminated<br />
false crossing of the orbit with itself which arose by<br />
virtue of having projected the attractor into a too low<br />
dimensional space? In other words, when points in<br />
dimension d are neighbours of one other? By examining<br />
this question in dimension one, then dimension<br />
two, etc. until there are no incorrect or false neighbours<br />
remaining, one should be able to establish, from<br />
geometrical consideration alone, a value for the necessary<br />
embedding dimension. Such an approach was<br />
described by Kennel et al. [6]. In dimension d each<br />
vector<br />
y(k) = [s(k), s(k + τ), s(k + 2τ), …,<br />
s(k + (d−1)τ)] (9)<br />
has a nearest neighbour yNN (k) with nearness in the<br />
sense of some distance function. The Euclidean distance<br />
in dimension d between y(k) and yNN (k) we call<br />
R (k): d<br />
2 NN 2 NN<br />
2<br />
Rd( k) = [ sk ( ) − s ( k)] + [ sk ( +τ) − s ( k+τ<br />
)] +<br />
(10)<br />
NN<br />
2<br />
... + [ sk ( +τ( d−1)) − s ( k+τ( d−1))]<br />
.<br />
R (k) is presumably small when one has a lot a<br />
d<br />
data, and for a dataset with N measurements, this distance<br />
is of order 1/N1/d . In dimension d + 1 this nearest-neighbour<br />
distance is changed due to the (d + 1)st<br />
coordinates s(k + dτ) and sNN (k + dτ) to<br />
2 2 NN<br />
2<br />
Rd+ 1 ( k) = Rd( k) + [ s( k+ dτ) − s ( k+ dτ<br />
)] . (11)<br />
We can define some threshold size R to decide<br />
T<br />
when neighbours are false. Then if<br />
NN<br />
| sk ( + dτ) − s ( k+ dτ<br />
)|<br />
> RT<br />
, (12)<br />
R ( k)<br />
d<br />
the nearest neighbours at time point k are declared<br />
false. Kennel et al.[6] showed that for values in the<br />
range 10 ≤ R T ≤ 50 the number of false neighbours<br />
identified by this criterion is constant. In practice,<br />
the percentage of false nearest neighbours is determined<br />
for each dimension d. A value at which the<br />
percentage is almost equal to zero can be considered<br />
as the embedding dimension. Figure 6 displays the<br />
percentage of false nearest neighbours that was determined<br />
for the amplitude level series, for phase-spaces<br />
reconstructed with embedding dimensions from 1 to<br />
20. As it can be seen, the percentage of false neighbours<br />
drops to almost zero at 4 or 5. This indicates<br />
that a four or five-dimensional phase-space is necessary<br />
to represent the dynamics (or unfold the attractor)<br />
of the amplitude level series. From the other<br />
hand, the mean percentage of false nearest neighbours<br />
computed for the surrogate data sets decreases<br />
steadily but at 20 is about 35%. Such a result seems<br />
to be in close agreement with that was obtained from<br />
the correlation integral analysis, providing further<br />
support to the observation made earlier regarding the<br />
presence of low-dimensional chaotic dynamics in<br />
the amplitude level variations.
Fig. 6. Embedding dimension estimation using false nearest<br />
neighbour method for amplitude level data for original time series<br />
(line 1), mean values of surrogate data sets (line 2), and one surrogate<br />
realization (line 3). Error bars indicate minimal percentage of<br />
false nearest neighbour among all realizations of surrogate data<br />
3.3. LYAPUNOV EXPONENTS<br />
Lyapunov exponents are the dynamical invariants<br />
of the nonlinear system. They are very useful when<br />
physics of process is considered. Using the spectrum<br />
of Lyapunov exponents, the average predictability of<br />
nonlinear system can be estimated. In a general case,<br />
the orbits of chaotic attractors are unpredictable, but<br />
there is the limited predictability of chaotic physical<br />
system, which is defined by the global and local<br />
Lyapunov exponents. A concept of Lyapunov exponents<br />
existed long before the establishment of chaos<br />
theory, and was developed to characterize the stability<br />
of absolute value of the eigenvalues of the linearized<br />
dynamics averaged over the attractor. A negative<br />
exponent indicates a local average rate of contraction<br />
while a positive value indicates a local average rate of<br />
expansion. In the chaos theory, the spectrum of Lyapunov<br />
exponents is considered a measure of the effect<br />
of perturbing the initial conditions of a dynamical<br />
system. Note that both positive and negative Lyapunov<br />
exponents can coexist in a dissipative system, which is<br />
then chaotic.<br />
Since the Lyapunov exponents are defined as asymptotic<br />
average rates, they are independent of the<br />
initial conditions, and therefore they do comprise an<br />
invariant measure of attractor. In fact, if one manages<br />
to derive the whole spectrum of Lyapunov exponents,<br />
other invariants of the system, i.e. Kolmogorov entropy<br />
and attractor’s dimension can be found. The<br />
Kolmogorov entropy, K, measures the average rate at<br />
which information about the state is lost with time.<br />
An estimate of this measure is the sum of the positive<br />
Lyapunov exponents. The inverse of the Kolmogorov<br />
entropy is equal to the average predictability. The estimate<br />
of the dimension of the attractor is provided by<br />
the Kaplan and Yorke conjecture (see [15,18]):<br />
j<br />
∑ λα<br />
α= 1<br />
dL= j+<br />
| λ<br />
,<br />
|<br />
(13)<br />
j<br />
j+<br />
1<br />
j+ 1<br />
where j is such that ∑ λ α > 0 and ∑ λ α < 0 , and the<br />
α= 1<br />
α= 1<br />
Lyapunov exponents λ are taken in descending order.<br />
α<br />
There are several approaches to computing the Lyapunov<br />
exponents (see, e.g., [5,6]); in this paper, we use<br />
one [18] which computes the whole spectrum and is<br />
based on the Jacobin matrix of the system function<br />
[14]. To calculate the spectrum of Lyapunov exponents<br />
from the amplitude level data, we use the time<br />
delay τ = 40 and embed the data in the four-dimensional<br />
space. Such a choice of the input parameters is<br />
the result of the previous calculations. Table 2 summarizes<br />
the results of the Lyapunov exponent analysis.<br />
For the time series under consideration, there exist two<br />
positive exponents (indicating expansion along two<br />
directions) and two negative ones (indicating contraction<br />
along remaining directions). The Kaplan-Yorke<br />
dimension is equal to 3.33; this value is very close to<br />
the correlation dimension which was defined by the<br />
Grassberger-Procaccia algorithm. The estimations<br />
of the Kolmogorov entropy and average predictability<br />
can show a limit, up to which the amplitude level<br />
data can be on average predicted. Surely, the important<br />
moment is a check of the statistical significance<br />
of results.<br />
Table 2<br />
Results of Lyapunov exponents analysis for amplitude level: λ 1 −λ 4<br />
are the Lyapunov exponents in descending order, d L is the Kaplan-<br />
Yorke attractor dimension, K is the Kolmogorov entropy, and P is<br />
the average predictability<br />
λ 1 λ 2 λ 3 λ 4 d L K P<br />
0.0082 0.0017 −0.0047 −0.0167 3.33 0.0094 124.3<br />
It is worth to remind that results of state-space<br />
reconstruction are highly sensitive to the length of<br />
data set (i.e. it must be sufficiently large) as well as to<br />
the time lag and embedding dimension determined.<br />
Indeed, there are limitations on the applicability of<br />
chaos theory for observed (finite) time series arising<br />
from the basic assumptions that the time series must<br />
be infinite. A finite and small data set may probably<br />
results in an underestimation of the actual dimension<br />
of the process. Nevertheless, we check the robustness<br />
of our results with respect to the size of time series by<br />
dividing the data in 2 sets of 4500 points and in 4 sets<br />
of 2250 points and using the methods described above<br />
for these subsets. The main assumption is that the results<br />
obtained for the subsets are close to the results<br />
obtained for the whole time series. We received the dimensions<br />
determined for various time lags. It has been<br />
found the value τ = 40, which is the time lag provided<br />
by the mutual information approach, and the correlation<br />
dimension is 3.33 due to the saturation at the embedding<br />
dimension 6 (see Fig. 4). Thus, the statistical<br />
convergence tests, described here, together with surrogate<br />
data approach that, was above applied, provide<br />
the satisfactory significance of our data regarding the<br />
state-space reconstruction.<br />
81
82<br />
4. CONCLUSIONS AND DISCUSSION<br />
This paper investigated the existence of chaotic<br />
behaviour in the non-linear vibrational systems on the<br />
basis of temporal series, in particular, for a grid of the<br />
semiconductor autogenerators. The mutual information<br />
approach, the correlation integral analysis, the<br />
false nearest neighbour algorithm, the Lyapunov exponent’s<br />
analysis, and the surrogate data method were<br />
used in the analysis. The mutual information approach<br />
provided a time lag which is needed to reconstruct<br />
phase space. Such an approach allowed concluding<br />
the possible nonlinear nature of process resulting in<br />
the amplitude level variations. The correlation dimension<br />
method provided a low fractal-dimensional<br />
attractor thus suggesting a possibility of the existence<br />
of chaotic behaviour. The method of surrogate data,<br />
for detecting nonlinearity, provided significant differences<br />
in the correlation exponents between the original<br />
data series and the surrogate data sets. This finding<br />
indicates that the null hypothesis (linear stochastic<br />
process) can be rejected. The main conclusion is that<br />
the system exhibits a nonlinear behaviour and possibly<br />
low-dimensional chaos. Thus, a short-term prediction<br />
based on nonlinear dynamics is possible. The Lyapunov<br />
exponents analysis supported this conclusion. It<br />
can be noted that the nonleading exponents are notoriously<br />
difficult to estimate from time series data.<br />
Moreover, the interpretation of inverse Lyapunov exponents<br />
as predictability times can results in ambiguous<br />
conclusions.<br />
Though a large number of studies employed the<br />
ideas gained from the science of chaos, there have also<br />
been widespread criticisms on the application of chaos<br />
theory. The fact that observational time series are almost<br />
always finite and are inherently contaminated by<br />
noise, such as errors arising from measurements, necessitate<br />
addressing the above issues in the application<br />
of chaos theory. The basis for the criticisms of studies<br />
investigating and reporting existence of chaos in the<br />
amplitude variations is our strong belief that they are<br />
influenced by a large number of variables and, therefore,<br />
are stochastic. On the other hand, the outcomes<br />
of the present study provide support to the claims that<br />
the (seemingly) highly irregular processes could be the<br />
result of simple deterministic systems with a few degrees<br />
of freedom. Therefore, the hypothesis of chaos<br />
in is reasonable and can provide an alternative approach<br />
for characterizing and modelling the dynamics<br />
of processes resulting in the vibrations processes in<br />
the complicated system. The future investigations can<br />
be realized and include , for example, the nonlinear<br />
prediction method or artificial neural network approach.<br />
In any case the presented analysis has a great<br />
importance for correct treating dynamics of different<br />
semiconductors and quantum electronics devices,<br />
nanoelectronics and corresponding nano-devices, including<br />
the molecular electronics.<br />
References<br />
1. Âåäåíîâ À.À., Ýæîâ À.À., Ëåâ÷åíêî Ý.Á, Íåëèíåéíûå<br />
ñèñòåìû ñ ïàìÿòüþ è ôóíêöèÿìè íåéðîíà, Â: Íåëèíåéíûå<br />
âîëíû. Ñòðóêòóðà è ðàçäâîåíèÿ Ý.Ä Ãàïîíîâ-Ãðåõîâ<br />
À.Â.,Ðàáèíîâè÷ Ì.È.Íàóêà, Ìîñêâà 1997,Ñ .53-69,<br />
2. Serbov N.G., Svinarenko A.A., Wavelet and multifractal analysis<br />
of oscillations in system of couled autogenerators in chaotic<br />
regime// Photoelectronics. — 2006 . — N15. — P.27-30.<br />
3. Serbov N.G., Svinarenko A.A., Wavelet and multifractal<br />
analysis of oscillations in a grid of couled autogenerators//<br />
Photoelectronics. — 2007. — N16. — P.53-56.<br />
4. Svinarenko A.A., Serbov N.G., Chernyakova Yu.G., Wavelet<br />
and multifractal analysis of oscillations in a grid of couled autogenerators:<br />
a case of strong non-linearity// Photoelectronics.<br />
— 2008. — N17. — P.42-45.<br />
5. Abarbanel H., Brown R., Sidorowich J., Tsimring L., The<br />
analysis of observed chaotic data in physical systems//Rev<br />
Modern Phys. — 1993. — Vol.65. — P.1331–1392.<br />
6. Kennel M., Brown R., Abarbanel H., Determining embedding<br />
dimension for phase-space reconstruction using a geometrical<br />
construction//Phys Rev A. — 1998. — Vol.45. —<br />
P.3403–3411.<br />
7. Packard N., Crutchfield J., Farmer J., Shaw R., Geometry<br />
from a time series//Phys Rev Lett. — 1998. — Vol.45. —<br />
P.712–716.<br />
8. Havstad J., Ehlers C., Attractor dimension of nonstationary<br />
dynamical systems from small data sets//Phys Rev A. —<br />
1999. — Vol.39. — P.845–853.<br />
9. Gallager R.G., Information theory and reliable communication,<br />
Wiley, New York. — 1996.<br />
10. Grassberger P., Procaccia I., Measuring the strangeness of<br />
strange attractors//Physica D. — 1999 Vol.9. — P.189–208.<br />
11. Fraser A., Swinney H., Independent coordinates for strange<br />
attractors from mutual information// Phys Rev A. — 1996. —<br />
Vol.33. — P.1134–1140.<br />
12. Takens F Detecting strange attractors in turbulence. In: Rand<br />
DA, Young LS (eds) Dynamical systems and turbulence, Warwick<br />
1999. (Lecture notes in mathematics No 898). Springer,<br />
Berlin Heidelberg New York, pp 366–381<br />
13. Mañé R On the dimensions of the compact invariant sets of<br />
certain non-linear maps. In: Rand DA, Young LS Dynamical<br />
systems and turbulence, Warwick 1990. (Lecture notes in<br />
mathematics No 898). Springer, Berlin Heidelberg N. — Y.,<br />
p. 230–242<br />
14. Sano M, Sawada Y (1985) Measurement of the Lyapunov<br />
spectrum from a chaotic time series//Phys Rev.Lett. —<br />
1995. — Vol.55. — P.1082–1085<br />
15. Sivakumar B., Chaos theory in geophysics: past, present and<br />
future//Chaos, Solitons & Fractals. — 2004. — Vol.19. —<br />
P.441–462.<br />
16. Theiler J., Eubank S., Longtin A., Galdrikian B., Farmer J.,<br />
Testing for nonlinearity in time series: The method of surrogate<br />
data// Physica D. — 1998. — Vol.58. — P.77–94.<br />
17. Glushkov A.V., Khokhlov V.N., Tsenenko I.A. Atmospheric<br />
teleconnection patterns: wavelet analysis// Nonlin. Proc.in<br />
Geophys. — 2004. — V.11,N3. — P.285-293.<br />
18. Khokhlov V.N., Glushkov A.V., Loboda N.S., Serbov N.G.,<br />
Zhurbenko K., Signatures of low-dimensional chaos in hourly<br />
water level measurements at coastal site// Stoch. Environ.<br />
Res. Risk Assess. . — 2008. — Vol.22. — P.777-788.<br />
19. Bunyakova Yu.Ya., Glushkov A.V.,Fedchuk A.P., Serbov<br />
N.G., Svinarenko A.A., Tsenenko I.A., Sensing non-linear<br />
chaotic features in dynamics of system of couled autogenerators:<br />
standard multifractal analysis// Sensor Electr. and Microsyst.<br />
Techn. — 2007. — N1. — P.14-17.<br />
20. Rusov V.D., Glushkov A.V., Loboda A., et al, On possible<br />
genesis of fractal dimensions in the turbulent pulsations of<br />
cosmic plasma — galactic-origin rays — turbulent pulsation<br />
in planetary atmosphere system// Advances in Space Research.<br />
— 2008. — Vol.42,N9. — P.1614-1617.
UDÑ 539.124 : 541.47<br />
A. A. Svinarenko, A. V. Loboda, N. G. Serbov<br />
MODELING AND DIAGNOSTICS OF INTERACTION OF THE NON-LINEAR VIBRATIONAL SYSTEMS ON THE BASIS<br />
OF TEMPORAL SERIES (APPLICATION TO SEMICONDUCTOR QUANTUM GENERATORS)<br />
Abstract<br />
It is studied an employing a variety of techniques for characterizing the dynamics of the coupled semiconductor quantum generators<br />
and identifying the presence of chaotic dynamics in this time (frequency) series. The statistical techniques used are the autocorrelation<br />
function and the Fourier power spectrum, whereas the mutual information approach, correlation integral analysis, false nearest<br />
neighbour algorithm, Lyapunov exponents analysis, and surrogate data method are used for comprehensive characterization.<br />
Key words: non-linear vibrational systems, quantum generators, statistical techniques.<br />
ÓÄÊ 539.124 : 541.47<br />
À. À. Ñâèíàðåíêî, À. Â. Ëîáîäà, Í. Ã. Ñåðáîâ<br />
ÌÎÄÅËÈÐÎÂÀÍÈÅ È ÄÈÀÃÍÎÑÒÈÊÀ ÂÇÀÈÌÎÄÅÉÑÒÂÈÉ Ó ÍÅËÈÍÅÉÍÛÕ ÊÎËÅÁÀÒÅËÜÍÛÕ ÑÈÑÒÅÌÀÕ<br />
ÍÀ ÎÑÍÎÂÅ ÀÍÀËÈÇÀ ÂÐÅÌÅÍÍÛÕ ÐßÄΠ(ÏÐÈÌÅÍÅÍÈÅ Ê ÏÎËÓÏÐÎÂÎÄÍÈÊÎÂÛÌ ÊÂÀÍÒÎÂÛÌ<br />
ÃÅÍÅÐÀÒÎÐÀÌ)<br />
Ðåçþìå<br />
Íîâûå ñòàòèñòè÷åñêèå ìåòîäèêè èñïîëüçîâàíû äëÿ èçó÷åíèÿ äèíàìèêè ïîëóïðîâîäíèêîâûõ êâàíòîâûõ ãåíåðàòîðîâ è<br />
èäåíòèôèêàöèè õàîñà â ñîîòâåòñòâóþùèõ âðåìåííûõ ðÿäàõ. Äëÿ âûÿâëåíèÿ õàðàêòåðíûõ îñîáåííîñòåé äèíàìèêè ñèñòåì<br />
ïðèìåíåíû ìåòîä àâòîêîððåëÿöèîííûõ ôóíêöèé, Ôóðüå — ñïåêòðû, ìåòîäèêà âçàèìíîé èíôîðìàöèè, ìåòîä êîððåëÿöèîííûõ<br />
èíòåãðàëîâ, àëãîðèòì “ëîæíûõ áëèæàéøèõ ñîñåäåé”, àíàëèç íà îñíîâå ýêñïîíåíò Ëÿïóíîâà, ìåòîä ñóððîãàòíûõ äàííûõ.<br />
Êëþ÷åâûå ñëîâà: íåëèíåéíûå êîëåáàòåëüíûå ñèñòåìû, êâàíòîâûå ãåíåðàòîðû, ñòàòèñòè÷åñêèå ìåòîäû àíàëèçà.<br />
ÓÄÊ 539.124 : 541.47<br />
À. À. Ñâèíàðåíêî, À. Â. Ëîáîäà, Ì. Ã. Ñåðáîâ<br />
ÌÎÄÅËÞÂÀÍÍß ÒÀ IJÀÃÍÎÑÒÈÊÀ ÂÇÀªÌÎÄ²É Ó ÍÅ˲ͲÉÍÈÕ ÊÎËÈÂÀËÜÍÈÕ ÑÈÑÒÅÌÀÕ ÍÀ ÎÑÍβ<br />
ÀÍÀ˲ÇÓ ×ÀÑÎÂÈÕ ÐßIJ (ÇÀÑÒÎÑÓÂÀÍÍß ÄÎ ÍÀϲÂÏÐβÄÍÈÊÎÂÈÕ ÊÂÀÍÒÎÂÈÕ ÃÅÍÅÐÀÒÎв )<br />
Ðåçþìå<br />
Çàñòîñîâàí³ íîâ³ ñòàòèñòè÷í³ ìåòîäèêè äëÿ ç’ÿñóâàííÿ äèíàì³êè íàï³âïðîâ³äíèêîâèõ êâàíòîâèõ ãåíåðàòîð³â òà ³äåíòèô³êàö³¿<br />
õàîñó ó ÷àñîâèõ ðÿäàõ. Äëÿ âèÿâëåííÿ õàðàêòåðíèõ îñîáëèâîñòåé ñèñòåìè âèêîðèñòàí³ ìåòîä àâòîêîðåëÿö³éíèõ ôóíêö³é,<br />
Ôóð’º — ñïåêòðîñêîï³ÿ, ìåòîäèêà âçàºìíî¿ ³íôîðìàö³¿, ìåòîä êîðåëÿö³éíèõ ³íòåãðàë³â, àëãîðèòì “õèáíèõ íàéáëèæ÷èõ<br />
ñóñ³ä³â”, àíàë³ç íà ï³äñòàâ³ åêñïîíåíò Ëÿïóíîâà, ìåòîä ñóðîãàòíèõ äàíèõ.<br />
Êëþ÷îâ³ ñëîâà: íåë³í³éí³ êîëèâàëüí³ ñèñòåìè, êâàíòîâ³ ãåíåðàòîðè, ñòàòèñòè÷í³ ìåòîäè àíàë³çó.<br />
83
Semiconductor crystals of cadmium sulphide with shifted to excited R’ one [2]. And the occupation of<br />
S- and R-centers were applied in the studies. When these levels with holes is determined by the corre-<br />
samples treated by visible light and intensive IR-illusponding maxima. For the same reason the first maximination,<br />
their relaxation characteristics, lux-current mum (shortwave one) is more sensible to changes in<br />
dependencies, photocurrent spectral distribution and each light intensities.<br />
curve for quenching coefficient distribution corresponded<br />
to Bube-Rose model [1,2]. The dependence<br />
The lower intensity of quenching light at Lв = const ,<br />
the lower Q becomes. And at lower intensities of in-<br />
of quenching value on wavelength had two maxima trinsic excitation this dependence shows evidently. At<br />
within the range 1000-1400 μm, that certified the the same time the value of quenching coefficient in-<br />
presence of R-centers excited states [2].<br />
We investigated the change in quenching value uncreases<br />
with decrease of L в excitation at unchanging<br />
intensity of quenching light. The increase was higher if<br />
der various intensities of applied light fluxes. All the<br />
measurements were carried out under the stationary<br />
conditions. The relaxation maintained in each point<br />
the applied intensities L were insignificant.<br />
g<br />
Experimental particularities of Q (L , L ) men-<br />
g B<br />
tioned above confirm the validity of formula (1) in our<br />
(up to 20 minutes) to avoid the processes described in case. There are some limits imposed during its deriva-<br />
[3, 4].<br />
tion, and one was the following: all the holes knocked<br />
For Q(L L ) there is only one expression in lit-<br />
g B<br />
erature [5] that requires to measure the variables such<br />
out of R — R’ levels by light remain in valence band<br />
and decrease capturing to S — centers. But this sup-<br />
as free carriers and complicated considerably the calposition is not valid. Cross-section of holes capture by<br />
culations, made them low exact and practically unac- S- and R — centers are equal. The hole being newly<br />
ceptable.<br />
photoexcited locates dimensionally near R’ — cen-<br />
We used the dependence of IR-quenching value ter and most probably will be captured by it [7]. The<br />
on intensities of applied light flows, being given earlier similar, probably multiple, oscillations does not show<br />
in [6]:<br />
on registered external parameters and lead to useless<br />
⎡⎛ τp ⎞ L ′ ′ gαβ τp<br />
⎤<br />
Q = 1− + ⋅100<br />
0<br />
⎢⎜ ⎟ ⎥ 0<br />
⎢⎣⎝ τnв⎠ nL<br />
αβτ ⎥⎦<br />
(1)<br />
absorption of IR-light photons. Obviously this process<br />
can mask the dependence on intensity of IR-light.<br />
As critical levels of illumination both by exciting<br />
and quenching light we will observe such light fluxes<br />
where Q - IR-quenching coefficient; τn, τ p — lifetimes<br />
for nonequilibrium electrons and holes; L в ,<br />
L — the value for incident quanta of exiting and<br />
g<br />
quenching light; α , α ′ — the part of photons absorbed<br />
by our sample; β , β ′ — quantum yields under exciting<br />
and quenching light treatment.<br />
Expression (1) shows the dependence of optical<br />
quenching value on intensities of exciting L в and<br />
quenching L light. It should be noted that the men-<br />
g<br />
tioned ratio is valid for low intensities of quenching<br />
light and high levels of photoexcitation, when the<br />
changes in recombination centers occupation can not<br />
be taken into consideration.<br />
The studies showed that under various intensity<br />
combinations the shortwave maximum occurred lower<br />
than long-wave one. This is explained by thermal<br />
supply of captured carriers. At the expense of photon<br />
absorption, the part of holes from the basic R-level<br />
when in spectral allocation of quenching coefficient<br />
not only the mentioned numerical changes is seen at<br />
spectral distribution of quenching coefficient, but the<br />
qualitative changes become.<br />
As it was noted previously, with increase of intrinsic<br />
excitation intensity and decrease of IR-light flux<br />
the value of quenching decreases in accordance to formula<br />
(1). The conditions, when shortwave maximum<br />
disappeared completely in curve of spectral distribution<br />
Q (λ) but long-wave maximum still presented,<br />
were created (Fig. 1a). Formula (1) can not be applied<br />
for such a case because tolerance limits made at its<br />
derivation were broken.<br />
The processes took place in the case could be explained<br />
as follows. The lower the value L the lower<br />
g<br />
the number of holes is knocked out of R-centers by<br />
IR-photons. Respectively, the lower number of holes<br />
enters the centers of quick recombination and the<br />
84<br />
UDC 621.315.592<br />
YE. V. BRYTAVSKYI, YU. N. KARAKIS, M. I. KUTALOVA, G. G. CHEMERESYUK<br />
Odesa I. I. Mechnikov National University, E-mail: wadz@mail.ru, t 8048-7266356<br />
EFFECTS CONNECTED WITH INTERACTION OF CHARGE CARRIERS<br />
AND R-CENTERS BASIC AND EXITED STATES<br />
The critical modes of illumination for samples with sensitization centers by exciting and quenching<br />
light were investigated. And the conditions when the spectral distribution of infrared quenching<br />
coefficient changed qualitatively have been founded. Disappearance of quenching shortwave maximum<br />
within the range of 1000 μm connected with photoexcitation of holes from R-centers under the<br />
high intrinsic conductivity. The anomalous shape of quenching curve without long-wave maximum in<br />
the range of 1400 μm was obtained. The observed dependence is explained by the decrease of quantum<br />
yield for infrared illumination.<br />
© Ye. V. Brytavskyi, Yu. N. Karakis, M. I. Kutalova, G. G. Chemeresyuk, 2009
losses of main carriers (electrons) become lower. The<br />
value of coefficient Q estimates (through current)<br />
namely this relative decrease. The higher the intensity<br />
of intrinsic light and, respectively, the initial concentration<br />
of free electrons, the quickly their decrease by<br />
recombination becomes negligibly small. In the first<br />
place, the shortwave maximum Q (λ) disappears from<br />
the curve, because it is connected with holes release<br />
from basic state of R-centers, and charge concentration<br />
there is lower at the expense of thermal pumping<br />
to R’- states [2].<br />
Q ,% (�)<br />
20 20<br />
10 10<br />
20<br />
10<br />
800 1200<br />
λ , ��,<br />
1600<br />
Q ,% (�)<br />
��, λ,<br />
��<br />
800 1200 1600<br />
Figure 1. Qualitative changes in quenching. (à) — L â ” L g ;<br />
L g → 0; (á) — L â ↑, L g<br />
At maximum flux of IR-light and maximum level<br />
of intrinsic excitation we observed the disappearance<br />
of long-wave maximum for photocurrent quenching<br />
(1300-1400 μm), whereas shortwave maximum Q (λ)<br />
within the range 950-1000 μm still remained (Figure<br />
1b). This phenomenon was not described in literature<br />
previously.<br />
The maximum levels of exposure were determined<br />
by the possibilities of experimental equipment. Within<br />
the maximum of sample photosensitivity (520-530<br />
μm) the intensity of monochromatic light provided<br />
the illumination of order 5-6 lx.<br />
In infrared part of spectrum we ran into the obstacle<br />
of equal raise in illumination within spectral regions<br />
of both maxima. The known procedures to control<br />
light flux by the width of monochromator output<br />
slit or by application of neutral light filter gave nonproportional<br />
values, because these regions were located<br />
far from each other (up to 400 μm). And we took<br />
up the procedure to vary the tube filament at rather<br />
narrow output slit of monochromator.<br />
The assumptions of this procedure consist in the<br />
following: with increase of filament according to Wein<br />
law the spectrum of source emittance shifts slightly<br />
to shortwave part. As the result, illumination in near<br />
shortwave part of IR-spectrum increases somewhat<br />
quickly, than in long-wave part. In this case the greater<br />
influence gives the mechanisms to form Q (λ) maxima.<br />
Shortwave maximum of quenching always locates<br />
lower than long-wave one because of redistribution in<br />
captured holes concentration and it should disappear<br />
first at non-optimized ratio for exposure intensities.<br />
The anomalous shape of Q (λ) curve (Fig. 1b) is<br />
explained as follows. In accordance with formula (1),<br />
the value of coefficient Q does not depend on intensity<br />
L g itself but on product βL g ‘ which includes the value<br />
of quantum yield. The authors [8] noted, that at some<br />
ratios of light flux intensities the value of quantum<br />
yield can quickly decrease for infrared illumination in<br />
the samples with R-centers. The magnitudes of order<br />
β’= 0,026÷0,072 [7] were registered experimentally.<br />
At such low values namely the decrease of β’ can be<br />
decisive factor even under the considerably high magnitudes<br />
in numerator (1).<br />
The mechanism to explain the shape of Figure 1b<br />
dependence is suggested as follows. Under illumination<br />
by light with wavelength corresponded to activation<br />
energy of R’-centers (Figure 2), the number of<br />
free sites there increases. And decrease of thermally<br />
excited holes from R-level must raise. In its turn this<br />
leads to increase of free sites on these levels. As the<br />
result, the recurrent captures of holes to R’-centers<br />
increase, and quantum yield for IR-illumination becomes<br />
lower.<br />
f<br />
e +<br />
e –<br />
I<br />
S<br />
Figure. 2 Diagram for transitions of electrons and holes under<br />
considerable intensity of exciting light and infrared illumination.<br />
We note that the described effect can be achieved<br />
for each specified temperature only within very narrow<br />
range of existing R-center concentration and applied<br />
intensities of intrinsic light and infrared illumination.<br />
The studies are carried out under the condition when<br />
only the intensity of intrinsic light changes in relation<br />
to the other three registered parameters. If the intrinsic<br />
excitation is considerably high, there is the great<br />
number of free holes in V-band. The additional charge<br />
knocked out from R-centers by IR-illumination does<br />
not significantly change their concentration, and in<br />
the end, the current flow. Besides, R-centers become<br />
strongly occupied by holes (probably, even p r ≈ N r ;<br />
p r′ ≈ N r′ [1]). Respectively small changes caused by<br />
II<br />
kT<br />
R<br />
Ec<br />
Ev<br />
R�<br />
h�<br />
85
IR-photons are unable to influence the existed ratio<br />
of charge concentration on these centres. And the free<br />
places created there will be occupied immediately by<br />
holes from valence band.<br />
If intensity of intrinsic light is sufficiently high,<br />
concentration of localized vacancies will be low too.<br />
In this case, there are a lot of sites on R-centres not<br />
occupied by holes before IR-light switched on. And<br />
IR-excitation can not change their number and the<br />
balance of capture — emptying processes considerably.<br />
The studies carried out correspond to the movement<br />
along AB line of sketch figure 3.<br />
Area 1 in Figure 3 shows the intensities when<br />
the standard Rose mechanism is carried out [1]. The<br />
families of Q (λ) curves were measured under such<br />
conditions and formula (1) was obtained for such<br />
light fluxes. During its derivation the authors made<br />
simplifications required the conditions L g ↑ > L â ↑<br />
[6], when collection of quanta of intrinsic and infrared<br />
light is incident on the sample, and the value of<br />
quenching light L g is higher. And formula (1) can not<br />
be applied in area 2 in Figure 3 because of the abovementioned<br />
cause.<br />
�<br />
86<br />
Lg<br />
2<br />
1<br />
Figure 3. The shape of possible relationships between the intensities<br />
of exciting and quenching light: 1 — the area for photocurrent<br />
IR-quenching effect; 2 — the area for intensities without<br />
quenching; 3 — the area with anomalous quenching effect.<br />
The quenching effect can not carry out because of<br />
the following three reasons:<br />
I. Firstly, under low intrinsic excitation (L → â<br />
0) the number of free nonequilibrium carrier pairs is<br />
found smaller than the value, that can provide their<br />
recombination only through S-centres (see Fig. 2);<br />
3<br />
�<br />
�<br />
L�<br />
II. Secondly, insignificant activation of holes from<br />
R-centres (L →0) L concealed almost completely<br />
â g<br />
by dissipation processes, captures to traps etc. These<br />
traps do not practically reach S-centres;<br />
III. Small numbers of additional holes that however<br />
reach fast-recombination centres lead to small<br />
decrease of main carriers — electrons and practically<br />
do not influence on photocurrent change.<br />
Let’s make the observation when the area 3 of Figure<br />
3 can be reached along the line CB. This means<br />
that the level of intrinsic excitation is registered at the<br />
highest position and infrared flux increases gradually.<br />
At low L magnitudes the quenching does not occur<br />
g<br />
because of the third reason for area 2. At middle Lg magnitudes the quenching can be observed but it is<br />
insignificant. This corresponds to range condition of<br />
area 1 in Figure 3, showed by curve in Figure 1a. For<br />
the higher intensities the mechanism of anomalous<br />
quenching described above becomes valid.<br />
At that time the quenching maxima of Q (λ) dependence<br />
conduct differently. Shortwave maximum<br />
within the range 1000 μm can increase slightly at the<br />
expense of complete emptying in basic state of R-centers.<br />
Long-wave maximum (area of 1400 μm) can not<br />
appear even under these conditions because of the<br />
small magnitudes for quantum yield. The holes photoexcited<br />
from R’-states remain in R-centers areas at<br />
the expense of repeated captures and do not contribute<br />
to recombination on S-centers.<br />
References:<br />
1. À.Ðîóç Îñíîâû òåîðèè ôîòîïðîâîäèìîñòè Ìîñêâà,<br />
Ìèð,1998-Ñ.192<br />
2. Ð.Áüþá Ôîòîïðîâîäèìîñòü òâ¸ðäûõ òåë, Ìîñêâà, 2002ã.,Ñ<br />
558.<br />
3. ÊàðàêèñÞ.Í., Êóòàëîâà Ì.È.,Çàòîâñêàÿ Í.Ï. Îñîáåííîñòè<br />
ðåëàêñàöèè ôîòîòîêà â êðèñòàëëàõ ñóëüôèäà êàäìèÿ<br />
Ïåðâàÿ Óêðàèíñêàÿ êîíôåðåíöèÿ ïî ôèçèêå ïîëóïðîâîäíèêîâ,<br />
Îäåññà,Òåçèñû äîêëàäîâ, Ò 2, 2002ã.,<br />
4. Karakis Yu.N., Borschak V.F., Zotov V.V., Kutalova M.I. Relaxation<br />
characteristics of cadmium sulphide crystals with IRquenching.<br />
— Photoelectronics, vol. 11, 2002. — P. 51-55.<br />
5. Ñåðäþê Â.Â., ×åìåðåñþê Ã.Ã., Òåðåê Ì. Ôîòîýëåêòðè÷åñêèå<br />
ïðîöåññû â ïîëóïðîâîäíèêàõ , Êèåâ ,1992 — Ñ151.<br />
6. Novikova M.A., Karakis Yu.N., Kutalova M.I. Particularities<br />
of current transfer in crystals with two types of recombination<br />
centres. — Photoelectronics, vol. 14, 2005. — P. 58-61.<br />
7. Dragoev A.A., Karakis Yu.N., Kutalova M.I. Peculiarities in<br />
photoexcitation of carriers from deep traps. — Photoelectronics,<br />
vol. 15, 2006. — P. 54-56.<br />
8. Dragoev A.A., Zatovskaya N.P., Karakis Yu.N., Kutalova<br />
M. I. Sensors of infrared illumination controlled by electric<br />
field. — 2nd International Scientific and Technical Conference<br />
Sensor Elec –tronics and Microsystems Technology,<br />
Book of abstracts , P. 115.
UDC 621.315.592<br />
Ye. V. Brytavskyi, Yu. N. Karakis, M. I. Kutalova, G. G. Chemeresyuk<br />
EFFECTS CONNECTED WITH INTERACTION OF CHARGE CARRIERS AND R-CENTERS BASIC AND EXITED<br />
STATES<br />
Abstract<br />
The critical modes of illumination for samples with sensitization centers by exciting and quenching light were investigated. And the<br />
conditions when the spectral distribution of infrared quenching coefficient changed qualitatively have been founded. Disappearance of<br />
quenching shortwave maximum within the range of 1000 μm connected with photoexcitation of holes from R-centers under the high<br />
intrinsic conductivity. The anomalous shape of quenching curve without long-wave maximum in the range of 1400 μm was obtained. The<br />
observed dependence is explained by the decrease of quantum yield for infrared illumination.<br />
Key words: quantum yield, spectral distribution, infrared illumination.<br />
ÓÄÊ 621.315.592<br />
Å. Â. Áðèòàâñêèé, Þ. Í. Êàðàêèñ, Ì. È. Êóòàëîâà, Ã. Ã. ×åìåðåñþê<br />
ÝÔÔÅÊÒÛ, ÑÂßÇÀÍÍÛÅ ÑÎ ÂÇÀÈÌÎÄÅÉÑÒÂÈÅÌ ÍÎÑÈÒÅËÅÉ ÇÀÐßÄÀ Ñ ÎÑÍÎÂÍÛÌ<br />
È ÂÎÇÁÓÆĨÍÍÛÌ ÑÎÑÒÎßÍÈÅÌ R-ÖÅÍÒÐÎÂ<br />
Ðåçþìå<br />
Èññëåäîâàíû êðèòè÷åñêèå ðåæèìû îñâåùåíèÿ âîçáóæäàþùèì è ãàñÿùèì ñâåòîì îáðàçöîâ ñ öåíòðàìè î÷óâñòâëåíèÿ.<br />
Îðåäåëåíû óñëîâèÿ, ïðè êîòîðûõ ñïåêòðàëüíîå ðàñïðåäåëåíèå êîýôôèöèåíòà èíôðàêðàñíîãî ãàøåíèÿ ïðåòåðïåâàåò êà÷åñòâåííûå<br />
èçìåíåíèÿ. Èñ÷åçíîâåíèå êîðîòêîâîëíîâîãî ìàêñèìóìà ãàøåíèÿ â îáëàñòè 1000íì ñâÿçàíî ñ ôîòîâîçáóæäåíèåì<br />
äûðîê ñ R — öåíòðîâ â óñëîâèÿõ áîëüøîé ñîáñòâåííîé ïðîâîäèìîñòè. Îïðåäåë¸í àíîìàëüíûé âèä êðèâîé ãàøåíèÿ áåç äëèííîâîëíîâîãî<br />
ìàêñèìóìà â îáëàñòè 1400íì. Íàáëþäàåìàÿ çàâèñèìîñòü îáúÿñíÿåòñÿ óìåíüøåíèåì êâàíòîâîãî âûõîäà äëÿ èíôðàêðàñíîãî<br />
èçëó÷åíèÿ.<br />
Êëþ÷åâûå ñëîâà: íîñèòåëè çàðÿäà, èíôðàêðàñíîe èçëó÷åíèå, êâàíòîâûé âûõîä.<br />
ÓÄÊ 621.315. 592<br />
Å. Â. Áðèòàâñüêèé, Þ. Í. Êàðàê³ñ, Ì. ². Êóòàëîâà, Ã. Ã. ×åìåðåñþê<br />
ÅÔÅÊÒÈ, ÏΠ, ßÇÀͲ Dz ÂÇÀªÌÎIJªÞ ÍÎѲ¯Â ÇÀÐßÄÓ Ç ÎÑÍÎÂÍÈÌ ² ÇÁÓÄÆÅÍÈÌ ÑÒÀÍÎÌ R-ÖÅÍÒвÂ.<br />
Ðåçþìå<br />
Äîñë³äæåí³ êðèòè÷í³ óìîâè çàñâ³òëåííÿ çáóäæóþ÷èì ³ ãàñíó÷èì ñâ³òëîì çðàçê³â ç öåíòðàìè ÷óòëèâîñò³. Çíàéäåí³ óìîâè,<br />
ïðè ÿêèõ ñïåêòðàëüíèé ðîçïîä³ë êîåô³ö³åíòà ³íôðà÷åðâîíîãî ãàñ³ííÿ â³ä÷óâຠÿê³ñí³ çì³íè. Çíèêíåííÿ êîðîòêîõâèëüîâîãî<br />
ìàêñèìóìà ãàñ³ííÿ â îáëàñò³ 1000íì ïîâ , ÿçàíî ç ôîòîçáóäæåííÿì ä³ðîê ç R — öåíòð³â çà óìîâ á³ëüøî¿ âëàñíî¿ ïðîâ³äíîñò³.<br />
Ðîçðàõîâàíî àíîìàëüíèé âèãëÿä êðèâî¿ ãàñ³ííÿ áåç äîâãîõâèëüîâîãî ìàêñèìóìà â îáëàñò³ 1400íì. Ñïîñòåð³ãàºìà çàëåæí³ñòü<br />
ïîÿñíþºòüñÿ çìåíüøåííÿì êâàíòîâîãî âèõîäó äëÿ ³íôðà÷åðâîíîãî âèïðîì³íþâàííÿ.<br />
Êëþ÷îâ³ ñëîâà: íîñ³¿ çàðÿäó, ³íôðà÷åðâîíå âèïðîì³íþâàííÿ, êâàíòîâèé âèõ³ä.<br />
87
88<br />
UDÑ 539.184<br />
I. N. SERGA<br />
Odessa National Polytechnical University, Odessa<br />
ELECTRON INTERNAL CONVERSION IN THE 125,127 Ba ISOTOPES<br />
The work is devoted to a theoretical studying of the electron internal conversion phenomenon in<br />
the 125,127 Ba nuclides. The relativistic Dirac-Fock (DF) method (modified Dirac code) is used in order<br />
to estimate the electron conversion coefficients in the 125,127 Ba nuclides.<br />
INTRODUCTION<br />
Hitherto a problem of the internal conversion<br />
studying attracts a great interest, which is provided<br />
by significant difficulties in theoretical and experimental<br />
definition of the corresponding electron internal<br />
conversion coefficients despite on the known<br />
progress in development of the experimental methodologies<br />
and technique [1-20]. From physical; point<br />
of view, an internal conversion is a radioactive decay<br />
process where an excited nucleus interacts with an<br />
electron in one of the lower atomic orbitals, causing<br />
the electron to be emitted from the atom. Thus, in an<br />
internal conversion process, a high-energy electron is<br />
emitted from the radioactive atom, but without beta<br />
decay taking place. Decay spectroscopy using on-line<br />
mass separators [2] has the advantage of allowing the<br />
study of level properties in the low energy region, including<br />
band head information, since gamma transitions<br />
between low-spin states can be measured more<br />
intensively under lower background conditions than<br />
can those with in-beam spectroscopy measurements.<br />
Below we consider spectra of the barium isotopes and<br />
turn attention on definition of the corresponding internal<br />
conversion electron coefficients. The neutronde@cient<br />
nuclides of 125,127 Ba have been studied by<br />
means of in-beam spectroscopy, and the level structures<br />
for high-spin states were interpreted within the<br />
framework of the IBFM model [2]. Decay studies of<br />
these nuclides have rarely been reported. Therefore,<br />
the last 1996, 1999 evaluations were mainly based<br />
on in-beam studies. The E1 transitions between parity<br />
doublets are characterized by a two to four orders<br />
of magnitude enhancement compared to those of<br />
more normal cases. The 127–130 Ba isotopes were studied<br />
by in-beam conversion electron measurements<br />
by Cottle-Glasmacher-Johnson-Kemper (1993) and<br />
Cottle-Glasmacher-Kemper (1992) ( see full review<br />
in ref. [2]) to investigate for parity doublets, and no<br />
evidence of them was observed [13-16]. These studies,<br />
however, focused only on the conversion electron<br />
measurements; the transition probabilities were<br />
not measured. It is necessary to measure not only the<br />
conversion electrons, but also the half-lives of the excited<br />
states in order to explain this feature. The aim<br />
of this investigation is to study theoretically the level<br />
properties of 125,127 Ba in the low energy region, focusing<br />
on consideration of the E1 transitions, and their<br />
electron conversion characteristics using the relativistic<br />
DF method (modified Dirac code) (see refs. [5-<br />
7]). Below we present the key details of this approach<br />
and then consider the low energy levels spectrum and<br />
internal conversion schemes in nuclides of 125,127 Ba.<br />
2. THE RELATIVISTIC DIRAC-FOCK<br />
METHOD<br />
Standard approach to calculating the electron<br />
conversion characteristics is based on the usual nonrelativistic<br />
Hartree-Fock (HF) or Hartree-Fock-Slater<br />
(HFS) approach with account of the finite nuclear<br />
size [17]. More comprehensive calculation must be<br />
based on the relativistic approaches, in particular, the<br />
well known DF method (see e.g. [5-7]). Though these<br />
methods are the most wide-spread calculation methods,<br />
but, as a rule, the corresponding orbital basis’s are<br />
not optimized. It often lead to the quite large errors<br />
in calculating the fundamental atomic characteristics.<br />
Besides, some problems are connected with correct<br />
definition of the nuclear size effects, QED corrections<br />
etc. In ref. [17] calculation was based on relativistic<br />
HFS wave functions with the coefficient C = 1 for the<br />
exchange term and on the assumption that the potential<br />
is the same for all electrons, including the emitted<br />
one. Experimental binding energies were used. Finite<br />
nuclear size is taken into account, but the penetration<br />
terms are not tabulated. In our work we use ab initio<br />
relativistic DF (modified Dirac code) approach [5-7]<br />
to calculating wave functions basis’s and electron conversion<br />
coefficients for Ba isotopes with account of the<br />
nuclear effects.<br />
To define a nuclear potential it is usually used the<br />
Fermi model for a charge distribution ρ () r :<br />
с( r) = с0 /{1 + exp[( r− c) / a)]}<br />
(1)<br />
where the parameter a=0.523 fm, the parameter ñ is<br />
chosen by such a way that it is true the following condition<br />
for average-squared radius:<br />
1/2 =(0.836⋅A1/3 +0.5700)fm.<br />
Further let us present the formulas for the finite<br />
size nuclear potential and its derivatives on the nuclear<br />
radius. If the point-like nucleus has the central potential<br />
W( R), then a transition to the finite size nuclear<br />
potential is realized by exchanging W(r) by the potential<br />
[8]:<br />
r<br />
2 2<br />
( ) = () ∫ ρ ( ) + ∫ ()( ρ ) . (2)<br />
W r R W r dr r r R dr r W r r R<br />
0<br />
∞<br />
r<br />
© I. N. Serga, 2009
We assume it as some zeroth approximation. The<br />
nuclear potential for spherically symmetric density<br />
ρ rR is:<br />
( )<br />
nucl ( ) (<br />
r<br />
∞<br />
' '2 ' ' ' '<br />
( 1 ) ( ) ( )<br />
∫ ∫ (3)<br />
V r R =− r drr ρ r R + drrρr R<br />
0<br />
It is defined by the following system of differential<br />
equations [7]:<br />
r<br />
'<br />
nucl<br />
2 ' '2 ' 2<br />
0<br />
2<br />
( , ) = ( 1 ) ρ( , ) ≡(<br />
1 ) ( , )<br />
V r R r ∫ drr r R r y r R (4)<br />
0<br />
( ) = ρ ( )<br />
y' r, R r r, R<br />
с '( r) = ( с / a)exp[( r− c) / a]{1+ exp[( r− c) / a)]}<br />
with corresponding boundary conditions. The master<br />
system of equations includes the equations for density<br />
distribution function too. The normalisation of electron<br />
radial functions f and g provides the behaviour of these<br />
i i<br />
functions for large values of radial valuable as follows:<br />
g (r)→r i -1 [(E+1)/E] 1/2 sin(pr +Δ ), (5)<br />
i<br />
f (r)→r i -1 (i/|i|) [(E-1)/E] 1/2 cos (pr+Δ ). i<br />
The DF equations for N-electron system are written<br />
and contain the potential: V(r)=V(r|nlj)+V +V(r|R),<br />
ex<br />
which includes the nuclear, electric and mean-filed potentials.<br />
The general analysis shows that the DF method<br />
allow getting the results, which are more precise in<br />
comparison with analogous HF or HFS data. From the<br />
other side, it is well known in the modern atomic physics<br />
that above cited mean-filed methods are needed to<br />
be improved by means using obligatorily optimization<br />
of the relativistic orbital basis’s (see details in ref. [7]).<br />
3. LOW ENERGY LEVEL SPECTRUM AND<br />
INTERNAL CONVERSION IN 125,127 BA<br />
NUCLIDES<br />
In ref. [2] the half-lives and electron conversion<br />
coefficients were measured for the first time through<br />
the decays of 125,127 La, by means of a delayed coincidence<br />
technique and a cooled Si (Li) detector, respectively.<br />
In this work the radioactivities of 125,127 La were<br />
produced with the heavy-ion-induced fusion evaporation<br />
reactions of nat Mo( 32 S, pxn) with a 160-MeV 32 S<br />
beam (~50 pnA) from a tandem accelerator (MP20)<br />
of JAERI [2]. Internal conversion electrons were measured<br />
with a cooled Si (Li) detector (500mm 2 x 3 mm;<br />
2.5-keV full width at half maximum at 976-keV electrons<br />
of 207 Bi). Simultaneously, gamma rays were measured<br />
with a 20% HPGe detector with 180° geometry<br />
in order to obtain the peak intensity ratios between the<br />
electrons and g rays. Some conversion coefficients of<br />
transitions below 100 keV were determined by means<br />
of x-gamma or γ-γ coincidence measurements with<br />
the LEPS and the Ge detector with 180° geometry.<br />
In addition, gamma-ray intensity measurements were<br />
performed with the HPGe detector and the LEPS with<br />
source-todetector distances of 10 and 5 cm, respectively.<br />
The full energy peak efficiencies for the detectors<br />
were determined using standard sources of 56 Co,<br />
133 Ba, 152 Eu, and 241 Am. Further, the half-life of 127m Ba<br />
r<br />
2<br />
was deduced to be 1.93(7) s from the decay curves of<br />
the 24.0-, 56.3-, and 80.2-keV gamma rays by spectrum<br />
multiscaling measurements with an A=127 beam<br />
[2], which mainly consisted of the metallic state of<br />
barium. The contribution of 127 La decay to the 56.3keV<br />
gamma ray was corrected using the other gammaray<br />
intensities of 127 La. It should be also noted that the<br />
relative intensities of the three gamma-rays associated<br />
with the decay of 127m Ba were obtained on experiment<br />
[2] much more precisely than were those of Liang et<br />
al. It is important to note that above 100 keV, the a K s<br />
were determined in experiment [2] by taking the peak<br />
intensity ratios of the electrons to the simultaneously<br />
measured gamma rays. These values were normalized<br />
by using the pure E2 (2 + →0 + ) 230-keV transition in<br />
124 Ba. The experimental values and the assigned multipolarities<br />
are shown in Fig. 1 and are also listed in.<br />
Table 1,2 together with our theoretical data and calculation<br />
values by Rossel et al. [17, 18]. In 125 Ba nuclide<br />
the a K values of the 43.7- and 67.2-keV transitions<br />
were deduced from the x-gamma coincidence method<br />
with the 281.9- and 521.6-keV g rays, respectively. In<br />
this method, the coincidence events with β + -particles<br />
were also taken into account to distinguish the K x rays<br />
from the originating EC decay and the internal conversion<br />
process. Assuming the multipolarity of the<br />
56.3-keV transition to M1 or E2, the a T value of the<br />
25.1-keV transition was deduced to be 6.9–20.2. The<br />
calculated value by Rossel et al. [17] supports that the<br />
25.1-keV transition is M1. Similarly, the a T value of the<br />
24.0-keV transition was deduced to be (0.6–1.6)⋅10 3<br />
from the gamma-ray relative intensity. It is interesting<br />
to compare here our theoretical estimates and data<br />
by by Rossel et al [17], which are 1.1⋅10 3 and 8.5⋅10 4<br />
for M2 and E3, respectively, the 24.0-keV transition<br />
can be considered mainly an M2 transition. The other<br />
a K values of the 79.4-, 114.3-, 128.7-, 134.3-, 220.4-,<br />
243.0-, 253.3-,269.6-, 285.6-, and 318.7-keV γ transitions<br />
associated with the decay of 127 La are deduced<br />
from the electron internal conversion measurements.<br />
E i t l K i ffi i t f t iti i 125 B (<br />
Fig. 1. Experimental K conversion coefficients of transitions<br />
in 125 Ba (a) and 127 Ba (b).<br />
89
Table 1<br />
The theoretical and experimental internal conversion coefficients<br />
of transitions in 125 Ba.<br />
E γ<br />
(keV)<br />
43.7<br />
98.7<br />
134.0<br />
168.5<br />
193.5<br />
216.3<br />
237.3<br />
281.9<br />
90<br />
Assign.<br />
Multip.<br />
M1/E2<br />
M1/E2<br />
M1/E2<br />
M1/E2<br />
M1/E2<br />
M1/E2<br />
E2<br />
E2<br />
Exp.<br />
[2]<br />
8.9(5)<br />
1.0(4)<br />
0.36(14)<br />
0.21(8)<br />
0.12(4)<br />
0.13(4)<br />
0.056(17)<br />
0.046(14)<br />
Theory<br />
[17]<br />
E2<br />
7.71<br />
1.28<br />
0.486<br />
0.236<br />
0.149<br />
0.104<br />
0.0770<br />
0.0446<br />
Theory<br />
[17]<br />
M1<br />
9.37<br />
0.895<br />
0.378<br />
0.200<br />
0.137<br />
0.101<br />
0.0791<br />
0.0510<br />
Our<br />
Theory<br />
E2<br />
7.5039<br />
1.1183<br />
0.4951<br />
0.2396<br />
0.1504<br />
0.1307<br />
0.0782<br />
0.0459<br />
Our<br />
Theory<br />
M1<br />
9.0562<br />
0.9340<br />
0.3824<br />
0.2113<br />
0.1480<br />
0.1298<br />
0.0801<br />
0.0522<br />
The analysis of theoretical data shows that at first<br />
as DF theory as HFS [17] theory provide in principle<br />
physically reasonable description of the electron internal<br />
conversion phenomenon and its characteristics.<br />
It is obvious that using DF approach gives more high<br />
accuracy for majority of transitions. At the same time<br />
more accurate account for the correlation, QED, nuclear<br />
effects is needed to get more comprehensive description<br />
of the phenomenon. One of the obvious ways<br />
to reach it is in using more sophisticated theoretical<br />
schemes, in particular, optimized QED perturbation<br />
theory with the optimized DF zeroth approximation<br />
(see e.g. [5,7]).<br />
Table 2<br />
The theoretical and experimental internal conversion coefficients<br />
of transitions in 127Ba E γ<br />
(keV)<br />
25.1<br />
56.3<br />
79.4<br />
114.3<br />
128.7<br />
220.4<br />
243.0<br />
253.3<br />
269.6<br />
285.6<br />
318.7<br />
Assign.<br />
Multip.<br />
M1<br />
M1/E2<br />
M1/E2<br />
M1/E2<br />
M1/E2<br />
M1/E2<br />
M1/E2<br />
M1/E2<br />
E2<br />
M1/E2<br />
E2<br />
Experiment<br />
[2]<br />
6.9–20.2<br />
5.0(4)<br />
2.0(1)<br />
1.0(5)<br />
1.2(5)<br />
0.097(30)<br />
0.078(25)<br />
0.087(30)<br />
0.062(28)<br />
0.047(15)<br />
0.035(10)<br />
Theory<br />
[17]<br />
E2<br />
537<br />
5.50<br />
1.26<br />
0.803<br />
0.553<br />
0.0975<br />
0.0713<br />
0.0625<br />
0.0513<br />
0.0428<br />
0.0305<br />
Theory<br />
[17]<br />
M1<br />
8.46<br />
4.53<br />
0.666<br />
0.591<br />
0.423<br />
0.0963<br />
0.0742<br />
0.0665<br />
0.0564<br />
0.0484<br />
0.0364<br />
Our<br />
Theory<br />
E2<br />
108<br />
5.7220<br />
1.5812<br />
0.9320<br />
0.7791<br />
0.0988<br />
0.0746<br />
0.0693<br />
0.0568<br />
0.0447<br />
0.0323<br />
Our<br />
Theory<br />
M1<br />
7.0421<br />
4.7010<br />
1.0292<br />
0.7062<br />
0.6851<br />
0.0973<br />
0.0771<br />
0.0735<br />
0.0621<br />
0.0502<br />
0.0388<br />
References<br />
1. Dragoun O., Rysavy M., Dragoun O., Spalek A., Internal<br />
conversion coefficients for superheavy elements// J. Phys. G:<br />
Nucl. Part. Phys. — 2000. — Vol.26. — P.1461-1466.<br />
2. Shibata M., Iimura H., Asai M., Osa A., Kawade K., Ichikawa<br />
S., Oshima M., Sekine T., Shinohara N., Half-life and<br />
UDÑ 539.184<br />
I.N. Serga<br />
ELECTRON INTERNAL CONVERSION IN THE 125,127 BA ISOTOPES<br />
internal conversion electron measurements in low-lying levels<br />
of 125,127 Ba //Phys.Rev -2002. — Vol.C65. — P.024305.<br />
3. Karpeshin F.F., Trzhaskovskaya M. B., Gangrskii Yu. P. ,<br />
Resonance Internal Conversion in Hydrogen-Like Ions //<br />
JETP. — 2004. — Vol.99, N2. — P.286-289.<br />
4. Kopytin I. V., Karelin K. N., and Nekipelov A. A. Exact Inclusion<br />
of the Coulomb Field in the Photobeta Decay of a<br />
Nucleus and Problem of Bypassed Elements// Phys. Atom.<br />
Nucl. — 2004. — Vol. 67, N8. — P.1429-1441.<br />
5. Grant I. P., Relativistic quantum theory of atoms and molecules.<br />
— N. — Y.:Springer, 2007. — 286P.<br />
6. “Dirac, a relativistic ab initio electronic structure program,<br />
Release DIRAC04.0 (2004)”, written by H. J. Aa. Jensen,<br />
T. Saue, and L. Visscher with contributions from V. Bakken,<br />
E. Eliav, T. Enevoldsen, T. Fleig, O. Fossgaard, T. Helgaker,<br />
J. Laerdahl, C. V. Larsen, P. Norman, J. Olsen, M. Pernpointner,<br />
J. K. Pedersen, K. Ruud, P. Salek, J. N. P. van Stralen,<br />
J. Thyssen, O. Visser, and T. Winther,2004, Ð 39.<br />
7. Glushkov A.V., Relativistic quantum theory. Quantum mechanics<br />
of atomic systems. — Odessa: Astroprint, 2008. —<br />
900P.<br />
8. Glushkov A.V., Rusov V.D., Ambrosov S.V., Loboda A.V., Resonance<br />
states of compound super-heavy nucleus and EPPP<br />
in heavy nucleus collisions // New Projects and New Lines of<br />
research in Nuclear phys. Eds. Fazio G.,Hanappe F. — Singapore<br />
: World Sci., 2003. — P.142-154.<br />
9. Coursol N., Gorozhankin V.M., Yakushev E.A., Briançon C.,<br />
Vylov A., Analysis of internal conversion coefficients // Applied<br />
Radiation and Isotopes. — 2000. — Vol.52, N3. — P.557-<br />
567.<br />
10. Nica N., Hardy J.C., Iacob V.E., Rockwell W.E., Trzhaskovskaya<br />
M.B., Internal conversion coefficients for rheavy elements,<br />
Phys. Rev. C-2007. — Vol.75. — P.024308.<br />
11. Raman, S., Nestor Jr., C.W., Ichihara, A. and Trzhaskovskaya,<br />
M.B., How good are the internal conversion coefficients<br />
now// Phys. Rev. — 2002. — Vol.C66. — P.044312.<br />
12. Band I.M., Trzhaskovskaya M.B., Nestor Jr., C.W., Tikkanen<br />
P.O. and Raman S., Dirac-Fock internal conversion<br />
coefficients// Atomic Data and Nucl Dat Tables. — 2002. —<br />
Vol.81. — P.1-334.<br />
13. Kibedi T., Burrows T., Trzhaskovskaya M.B., Davidson P.,<br />
Nestor C., Evaluation of theore-tical conversion coefficients<br />
using BrIcc // Nucl. Instr. and Meth. — 2008. — Vol.A589. —<br />
P.202-229<br />
14. Hofmann, C.R., Soff, G., Total and differential conversion<br />
coefficients for the internal pair creation in extended nuclei//<br />
Atomic Data and Nucl Dat Tables. — 1996. — Vol.63. —<br />
P.189-273.<br />
15. Coursol N., Gorozhankin V. M. , Yakushev E. A. , Briançon C.<br />
and Vylov V., Analysis of internal conversion coefficients//Applied<br />
Rad. and Isotopes. — 2000. — Vol.52,N3. — P.557-567.<br />
16. Safronova U.I., Mancini R., Atomic data for dielectronic satellite<br />
lines and dielectronic recombination into Ne 5+ // Atomic<br />
Data and Nucl. Data Tabl. — 2009. — Vol.95. — P.54-95.<br />
17. Rossel F., Fries H.M., Alder K., Pauli H.C., Internal conversion<br />
coefficients for all atomic shells //Atomic Data Nucl.<br />
Tables. — 1998. — Vol.21. — P.91-289.<br />
18. Firestone R.B., Shirley V.S., Table of Isotopes, 8th ed. —<br />
N.-Y.: Wiley, 2006.<br />
Abstract<br />
The work is devoted to a theoretical studying of the electron internal conversion phenomenon in the 125,127 Ba nuclides. The relativistic<br />
Dirac-Fock (modified Dirac code) method is used in order to estimate the electron conversion coefficients in the 125,127 Ba nuclides.<br />
Key words: electron internal conversion, relativistic theory, barium isotopes.
ÓÄÊ 539.184<br />
È. Í. Ñåðãà<br />
ÂÍÓÒÐÅÍÍßß ÊÎÍÂÅÐÑÈß ÝËÅÊÒÐÎÍÎÂ Â ÈÇÎÒÎÏÀÕ 125,127 BA<br />
Ðåçþìå<br />
Ðàáîòà ïîñâÿùåíà òåîðåòè÷åñêîìó èçó÷åíèþ ýôôåêòà âíóòðåííåé êîíâåðñèè ýëåêòðîíîâ â èçîòîïàõ 125,127 Ba. Íà îñíîâå<br />
îïòèìèçèðîâàííîãî ðåëÿòèâèñòñêîãî ìåòîäà Äèðàêà-Ôîêà (Dirac êîä) âûïîëíåíà îöåíêà êîýôôèöèåíòîâ âíóòðåííåé êîíâåðñèè<br />
ýëåêòðîíîâ â íóêëèäàõ 125,127 Ba.<br />
Êëþ÷åâûå ñëîâà: âíóòðåííÿÿ êîíâåðñèÿ ýëåêòðîíîâ, ðåëÿòèâèñòñêàÿ òåîðèÿ, èçîòîïû áàðèÿ.<br />
ÓÄÊ 539.184<br />
². Ì. Ѻðãà<br />
ÂÍÓÒвØÍß ÊÎÍÂÅÐÑ²ß ÅËÅÊÒÐÎͲ  ²ÇÎÒÎÏÀÕ 125,127 BA<br />
Ðåçþìå<br />
Ðîáîòà ïðèñâÿ÷åíà òåîðåòè÷íîìó âèâ÷åííþ åôåêòó âíóòð³øíüî¿ êîíâåðñ³¿ åëåêòðîí³â â ³çîòîïàõ 125,127 Ba. Íà îñíîâ³ îïòèì³çîâàíîãî<br />
ðåëÿòèâ³ñòñüêîãî ìåòîäó ijðàêà-Ôîêà (Dirac êîä) âèêîíàíà îö³íêà êîåô³ö³ºíò³â âíóòð³øíüî¿ êîíâåðñ³¿ åëåêòðîí³â<br />
ó íóêë³äàõ 125,127 Ba.<br />
Êëþ÷îâ³ ñëîâà: âíóòð³øíÿ êîíâåðñ³ÿ åëåêòðîí³â, ðåëÿòèâ³ñòñüêà òåîð³ÿ, ³çîòîïè áàð³ÿ.<br />
91
Previously, the influence of hydrogen peroxide<br />
concentration, dissolved in water solutions upon intensity<br />
of a luminescence of aluminium metal-oxide<br />
films, dipped in these solutions, has been investigated<br />
[1]. Because of the fact that aluminum oxide films are<br />
structures with high catalytic activity surface, it is possible<br />
the using these films as environment composition<br />
sensors.<br />
In [2] the presence of a series of thermostimulated<br />
luminescence (TSL) maxima, thermostimulated exoelectronic<br />
emission (TSEE) is shown and also the occurrence<br />
of potential difference between electrodes of<br />
Al-Al 2 O 3 -SnO 2 structure was observed for aluminum<br />
oxide films, which adsorbed only water or water vapours<br />
[3]. It indicates the sensitivity of investigated<br />
structures to a level of humidity, and, obviously, is<br />
caused by surface metal oxide states. In this connection,<br />
it was interesting to continue studying of influence<br />
of chemical processing upon aluminum oxide<br />
films on intensity of an arising luminescence in water<br />
solutions of various compounds. With this purpose it<br />
is important to study TSL spectra. Besides, it was of<br />
interest also to find out the influence of other substances,<br />
dissolved in water solutions, on intensity of<br />
aluminum oxide films luminescence dipped into these<br />
solutions.<br />
In this work, aluminum oxide films, formed on<br />
aluminum foil of the technical cleanliness by an electrochemical<br />
method in a water solution of sorrel acids<br />
[3] were investigated. TSL curves have been measured<br />
on the experimental setup, consisting of the lightproof<br />
box, IR-absorbing filter, photoelectronic multiplier,<br />
low current measuring device, copper-constantan<br />
thermocouple and recorder. Heating speed was maintained<br />
constant and was 0,3 Ê/sec.<br />
It is known [2], that in the temperature interval<br />
from room up to 600 K some maxima are observed<br />
in TSL spectra of aluminum oxide films. They correspond<br />
to various forms adsorbate and adsorbent binding.<br />
We have established, that as result of processing<br />
aluminum oxide in water solutions of some inorganic<br />
compounds, additional ÒSL bands appear and peaks,<br />
which existed earlier, quench. Chemical processing<br />
was carried out by boiling in Na 2 SO 4 , NaCl, NaI 0,1<br />
N water solutions.<br />
In fig. 1 ÒSL curves of Al 2 O 3 films are presented<br />
in the temperature interval from 450 up to 630 K,<br />
processed in: 1 — water, 2 — NaCl, 3 — Na 2 SO 4 ,<br />
4 — NaI. Only the centers, responsible for TSL in this<br />
temperature interval, are not deactivated by sodium<br />
92<br />
UDC 535.37<br />
L. N. VILINSKAYA, G. M. BURLAK<br />
The Odessa State Academy of Building and Architecture<br />
4, Didrichson str., Odessa, 65082 Ukraine, ph. 8 (048) 7-206-743<br />
SENSORS ON THE BASIS OF ALUMINIUM METAL-OXIDE FILMS<br />
Aluminum oxide films sensitivity to water solutions of inorganic compounds, the ammonia dissolved<br />
in water and in a gas atmosphere, and also to water vapours is established. The possibility of<br />
sensors’ fabrication on base of metal oxide films is shown.<br />
ions which were contained in all compounds we used.<br />
From fig. 1 it is seen, that the maximum at 560 K is<br />
observed in all samples. Therefore, obviously, is connected<br />
with dissociative water adsorption which also<br />
was available in all cases.<br />
B, rel.un.<br />
10<br />
8<br />
6<br />
4<br />
2<br />
0<br />
460 480 500 520 540 560 580 600 620 640<br />
T, K<br />
Fig. 1. TSL curves of Al 2 O 3 films processed in: 1 — water,<br />
2 — NaCl, 3 — Na 2 SO 4 , 4 — NaI<br />
The observed catalytic activity increase develops<br />
in the fact that after processing in sodium chloride<br />
the effective lightsum storage of the sample takes not<br />
hours but minutes. It can be explained by the fact that<br />
chlorine ions in a small amount increase catalytic<br />
activity of aluminum oxide surface. Chlorine forms<br />
strong bonds with valence and coordinated unsaturated<br />
aluminum atoms. While samples are in the air<br />
the dissociative adsorption of water and its vapour appears<br />
with creation of charged dopant OH - complexes<br />
which carry out a role of adsorptive nature traps. This<br />
process occurs quickly enough. As the energy distance<br />
between the levels created by hydroxyl and chlorine<br />
ions on a surface is very small (approximately 0,06 eV)<br />
[3], a part of carriers transfers to chlorine levels even<br />
at room temperatures and then are thermally released<br />
to the conduction band. As to others used anions, they<br />
do not change surface catalytic activity. Therefore,<br />
enough quantity charged hydroxyl groups has no time<br />
to be created.<br />
The influence of ammonia concentration dissolved<br />
in water solutions (in particular, sea water) on<br />
the intensity of aluminum oxide films luminescence,<br />
dipped in these solutions has been studied. The luminescence<br />
intensity was weak. Thus, the alternating<br />
1<br />
2<br />
3<br />
4<br />
© L. N. Vilinskaya, G. M. Burlak, 2009
voltage of about 1,3 V was applied providing electrolyte<br />
ions to the semiconductor surface but insufficient<br />
for electroluminescence excitation. Preliminary,<br />
the intensity aluminum oxide films luminescence in<br />
NaCl 3 % solution has been determined. After that,<br />
concentration of ammonia in this electrolyte was created.<br />
Luminescence intensity changes were measured.<br />
The addition of the few doses of ammonia increased<br />
aluminum oxide films emission intensity (Fig.2). The<br />
sensors sensitivity depends on technology metal oxide<br />
films preparation and temperature. The sensors were<br />
investigated in a temperature interval 278-354 K.<br />
The threshold of detecting is maximal in temperature<br />
interval 288-302 K. These results allow us to create<br />
sensors for ammonia concentration measurements<br />
in water solutions, operating at temperature range of<br />
278-353 K.<br />
B, rel.un.<br />
10<br />
8<br />
6<br />
4<br />
2<br />
0<br />
10 -9<br />
10 -7<br />
N, mol/l<br />
Fig. 2. Dependence aluminum metal-oxide films luminescence<br />
intensity on concentration of ammonia dissolved in water<br />
UDC 535.37<br />
L. N. Vilinskaya, G. M. Burlak<br />
SENSORS ON THE BASIS OF ALUMINIUM METAL-OXIDE FILMS<br />
10 -5<br />
10 -3<br />
The influence water vapour and ammonia concentration<br />
in a gas atmosphere on value EMF generated<br />
by structures Al-Al O -SnO was studied. Such struc-<br />
2 3 2<br />
tures were also created by electrochemical oxidizing of<br />
aluminum foil in sorrel acid water solution with next<br />
deposition of SnO layer on aluminum oxide film by<br />
2<br />
SnCl pyrolysis method. It is established, that the ap-<br />
4<br />
pearance of EMF occurs in following way. Because<br />
porosity of SnO layers, a penetration of adsorbed<br />
2<br />
molecules of water or ammonia from a gas atmosphere<br />
into micro pores of oxide film is possible. The aluminum<br />
oxide surface is catalytically active and consequently<br />
adsorbed molecules dissociation takes place.<br />
Dissociation products (positive and negative charged<br />
ions and electrons) get space separation along the film<br />
thickness because of the difference in their diffusion<br />
coefficients. The space separation continues until an<br />
internal electric field will arise and balance the diffusion<br />
flow of dissociation products. As a result, potential<br />
difference between aluminum and SnO electrode<br />
2<br />
arises. The value of this potential difference depends<br />
on ammonia concentration in a gas atmosphere. Detection<br />
threshold of these sensors is 0,1 g/m3 .<br />
It can be concluded, that aluminum oxide films<br />
are able to be used as sensors for ammonia detection in<br />
a gas atmosphere, in water and some inorganic compound<br />
solutions.<br />
References<br />
1. Áóðëàê Ã.Ì., Âèëèíñêàÿ Ë.Í. Âëèÿíèå ïåðåêèñè âîäîðîäà<br />
íà èíòåíñèâíîñòü ëþìèíåñöåíöèè ïîëóïðîâîäíèêîâûõ<br />
ïëåíîê. //Ñåíñîðíàÿ ýëåêòðîíèêà è ìèêðîñèñòåìíûå<br />
òåõíîëîãèè.. ¹4. 2008. Ñ.53-55.<br />
2. Ìèõî Â.Â., Ðîáóë Þ.Â., Òèìîôååâà Å.Þ., Áàëàáàí<br />
À.Ï.//Ôîòîýëåêòðîíèêà. — 2000. — Âûï.9. — Ñ.35-37.<br />
3. Ìèõî Â.Â., Ñåìåíþê Ë.Í. Ãåíåðàöiÿ ÝÐÑ ó ïëiâêàõ<br />
îêñèäó àëþìiíiþ //ÓÔÆ. 1995. — Ò.40. — ¹11-12. —<br />
Ñ.1209-1211.<br />
Abstract<br />
Aluminum oxide films sensitivity to water vapors, to water solutions of inorganic compounds, the ammonia dissolved in water and<br />
a gas atmosphere is established. The possibility of sensors fabrication on a basis of metal oxide films is shown.<br />
Key words: sensors, sensitivity, aluminum oxide.<br />
ÓÄÊ 535.37<br />
Ë. Í. Âèëèíñêàÿ, Ã. Ì. Áóðëàê<br />
ÑÅÍÑÎÐÛ ÍÀ ÎÑÍÎÂÅ ÏËÅÍÎÊ ÎÊÑÈÄÀ ÀËÞÌÈÍÈß<br />
Ðåçþìå<br />
Óñòàíîâëåíà ÷óâñòâèòåëüíîñòü îêñèäíûõ ïëåíîê àëþìèíèÿ ê âîäíûì ðàñòâîðàì íåîðãàíè÷åñêèõ ñîåäèíåíèé, àììèàêà,<br />
ðàñòâîðåííîãî â âîäå è â ãàçîâîé àòìîñôåðå, à òàêæå ê ïàðàì âîäû. Ïîêàçàíà âîçìîæíîñòü ñîçäàíèÿ ñåíñîðîâ íà îñíîâå<br />
îêñèäíûõ ïëåíîê.<br />
Êëþ÷åâûå ñëîâà: ÷óâñòâèòåëüíîñòü, ñåíñîðû, îêñèä àëþìèíèÿ.<br />
93
94<br />
ÓÄÊ 535.37<br />
Ë. Í. ³ë³íñüêà, Ã. Ì. Áóðëàê<br />
ÑÅÍÑÎÐÈ ÍÀ ÎÑÍβ Ï˲ÂÎÊ ÎÊÑÈÄÓ ÀËÞ̲ͲÞ<br />
Ðåçþìå<br />
Âñòàíîâëåíî ÷óòëèâ³ñòü îêñèäíèõ ïë³âîê àëþì³í³þ äî âîäÿíèõ ðîç÷èí³â íåîðãàí³÷íèõ ñïîëóê, àì³àêó, ðîç÷èíåíîãî ó âîä³<br />
òà ó ãàçîâ³é àòìîñôåð³, à òàêîæ äî âîäÿíî¿ ïàðè. Ïîêàçàíî ìîæëèâ³ñòü ñòâîðåííÿ ñåíñîð³â íà îñíîâ³ îêñèäíèõ ïë³âîê.<br />
Êëþ÷îâ³ ñëîâà: ÷óòëèâ³ñòü, ñåíñîðè, îêñèäí³ ïë³âêè àëþì³í³þ.
UDC 621.315.592<br />
O. O. PTASHCHENKO 1 , F. O. PTASHCHENKO 2 , V. V. SHUGAROVA 1<br />
1 I. I. Mechnikov National University of Odessa, Dvoryanska St., 2, Odessa, 65026, Ukraine<br />
2 Odessa National Maritime Academy, Odessa, Didrikhsona St., 8, Odessa, 65029, Ukraine<br />
TUNNEL SURFACE CURRENT IN GaAs–AlGaAs P-N JUNCTIONS, DUE<br />
TO AMMONIA MOLECULES ADSORPTION<br />
1. INTRODUCTION<br />
P-n junctions as gas-sensitive devices [1, 2] have<br />
some advantages in comparison with structures, based<br />
on oxide polycrystalline films [3, 4] and Shottky diodes<br />
[5, 6]. P-n junctions have high potential barriers<br />
for current carriers, which results in low background<br />
currents. Sensors on p-n junctions [1, 2] have crystal<br />
structure, high sensitivity at room temperature<br />
In previous papers the gas sensitivity of p-n structures<br />
on GaAs and GaAs–AlGaAs [1, 2], GaP [7],<br />
InGaN [8], and Si [9, 10] was investigated. . It was<br />
shown that the gas sensitivity of all these p-n junctions<br />
is due to forming of a surface conducting channel in<br />
the electric field induced by the ammonia ions adsorbed<br />
on the surface of the natural oxide layer. The<br />
gas-sensitivity of the forward current in p-n junctions<br />
is limited by strong rise of the injection current with<br />
the bias voltage. The voltage-limit for the reverse current<br />
is substantially higher. Therefore, under some<br />
conditions, the gas-sensitivity of the reverse current<br />
can be higher than the forward one.<br />
The aim of this work is a study of the influence<br />
of ammonia vapors on the forward and reverse currents<br />
in a GaAs–AlGaAs p-n heterostructure with a<br />
degenerated p + -GaAs layer. It is shown that, at high<br />
enough concentration of ammonia in the ambient<br />
atmosphere, a surface conducting channel with degenerated<br />
electrons is formed and tunnel forward and<br />
reverse currents are observed.<br />
2. EXPERIMENT<br />
The influence of ammonia vapors on I-V characteristics of forward and reverse currents and kinetics<br />
of surface currents in GaAs-AlGaAs p-n junctions with degenerated p + region was studied. It<br />
is shown that ammonia molecules adsorption, under sufficiently high NH 3 partial pressure, forms in<br />
p-AlGaAs a surface conducting channel with degenerated electrons. P-n junctions with degenerated<br />
p + region have higher gas sensitivity at reverse bias than at forward bias. This effect is explained by<br />
tunnel injection of electrons into the conducting channel from the degenerated p + region at a reverse<br />
bias. The rise time of the surface current in an ammonia vapor atmosphere of ~20s is due to filling up<br />
deep electron traps.<br />
The measurements were carried out on p+-<br />
GaAs(Zn) — p-Ga 0.45 Al 0.55 (Ge) — p-Ga 0.65 Al 0.35 (Ge) —<br />
n- Ga 0.45 Al 0.55 (Ge) structures with a degenerated p +<br />
layer.<br />
I-V characteristics of the forward and reverse currents<br />
were measured in air with various concentrations<br />
of ammonia vapors. The current kinetics at the change<br />
of the ambient atmosphere was observed.<br />
Fig.1 represents I–V characteristics of the forward<br />
current in a p-n structure in air (1) and in air with am-<br />
© O. O. Ptashchenko, F. O. Ptashchenko, V. V. Shugarova, 2009<br />
monia vapors of various partial pressures. The forward<br />
current increases with enhanced NH 3 concentration.<br />
At an ammonia pressure P>1000Pa a pronounced<br />
peak in the I-V curve is observed witch can be ascribed<br />
to electron tunneling between the c–band in the surface<br />
conducting channel and the v– band in the degenerated<br />
p + region. It means also that electrons in the<br />
channel are degenerated.<br />
10 -6<br />
10 -7<br />
10 -8<br />
I, A<br />
0<br />
5<br />
6<br />
4<br />
3<br />
0.5 1.0<br />
V, Volts<br />
1.5<br />
Fig. 1. I–V characteristics of the forward current in a p-n<br />
structure in air (1) and in ammonia vapors of a pressure: 2 — 50 Pà;<br />
3 — 100 Pà; 4 — 200 Pà; 5 — 1000 Pà; 6 — 4000 Pà<br />
Curves 1– 4 in Fig. 2 delineate the I–V curves of<br />
he forward and reverse currents in a p–n junction,<br />
placed in air with various ammonia partial pressures. It<br />
is seen that the reverse current is greater than forward<br />
one at the same ammonia pressure. It is characteristic<br />
for tunnel currents in tunnel- and inverted diodes.<br />
2<br />
1<br />
95
96<br />
3. DISCUSSION<br />
Curves 1 and 2 in Fig. 2 were obtained at ammonia<br />
pressures of 20 Pa and 100 Pa, respectively. As seen<br />
from the I–V curves sections, corresponding to V>0,<br />
the tunnel forward current under these pressures is<br />
small. And the reverse currents are remarkably higher<br />
than forward ones. It is characteristic of inverted diodes<br />
and argues that tunneling of electrons at reverse<br />
biases occurs from the degenerated p + region.<br />
I, μA<br />
0.8<br />
0.4<br />
0<br />
-0.4<br />
1<br />
2<br />
3<br />
-0.8<br />
-6 -4 -2 0<br />
V, Volts<br />
2<br />
4<br />
Fig. 2. I–V characteristics of a p-n structure in ammonia<br />
vapors of a pressure: 1 — 20 Pà; 2 — 100 Pà; 3 — 1000 Pà; 4 —<br />
4000 Pà<br />
Fig. 3 shows a simplified band diagram of the p-n<br />
junction with degenerated p + region. The Fermi level<br />
F is located in the v-band in p + 0<br />
0 0,2 0,4 0,6<br />
Vr, Volts<br />
region. The occupied<br />
states area is dashed. However the Fermi level lies by<br />
ΔE below c-band in n-region. In our case n-region<br />
Fig. 4. Low-bias sections of I–V characteristics of the reverse<br />
current in a p-n structure in ammonia vapors of a pressure: 1 —<br />
20 Pà; 2 — 50 Pà; 3 — 100 Pà; 4 — 4000 Pà<br />
corresponds to the conducting surface channel. De- The dependence of the cutoff voltage on the ampending<br />
on the ammonia partial pressure, ΔE change monia partial pressure P is shown in Fig. 5. This depen-<br />
can be estimated from the expression<br />
dence is logarithmic, however, the proportionality coef-<br />
ns = Nc exp( −Δ E/ kT)<br />
, (1)<br />
ficient is nkT p / q with n =1.4 — 1.5 instead kT / q .<br />
p<br />
The discrepancy between the model prediction<br />
where n is the electrons concentration in the channel<br />
s<br />
at the surface; N is the effective density of states in c-<br />
c<br />
band; k is the Boltzmann constant; T is temperature.<br />
It is evident from Fig. 3, that the strong rise of the current<br />
with reverse bias voltage V must begin at<br />
r<br />
Vr= V0 = Δ E/ q,<br />
(2)<br />
where q is the electron charge. Therefore the cutoff reverse<br />
bias voltage V must logarithmic depend on the<br />
0<br />
electrons concentration in the conducting channel,<br />
formed by the electric field of ammonia ions:<br />
and the experimental data can be ascribed to a variety<br />
of factors. The most important of them are: a) dependence<br />
of the effective channel width on the surface<br />
electrons concentration n ; b) non-linearity of the de-<br />
s<br />
pendence n (P), caused by deep traps in the channel.<br />
s<br />
A strong influence of trapping processes on the<br />
surface current in studied p-n structures is evident<br />
from a comparison between the rise- and decay curves<br />
of the surface current after let in- and off ammonia<br />
vapor in the container with the p-n structure.<br />
Fig. 6 illustrates the kinetic of the surface current<br />
in a p-n structure after let in and removal of ammo-<br />
V0 = kT / qln( Nc / ns)<br />
. (3)<br />
Fig. 4 represents low-bias sections of I–V characnia<br />
vapor from the container with the sample. The rise<br />
curve is exponential, i. e.<br />
teristics of the reverse current in a p-n structure, situated<br />
in ammonia vapors of various partial pressures.<br />
The cutoff voltage was estimated as the intersection of<br />
linearly extrapolated I-V curve with the abscise.<br />
It ( ) = Ist[1 −exp( −t/ τ r )] , (4)<br />
where I is the stationary value of I; the rise time τ =19.5 s.<br />
st r<br />
The decay curve is not exponential, with the 90% decay<br />
F<br />
�E<br />
Fig. 3. Band diagram of an inverted diode at thermal equilibrium<br />
I, nA<br />
20<br />
15<br />
10<br />
5<br />
1<br />
2<br />
3<br />
4<br />
C<br />
V
time τ 90 =4s. This time is comparable with the time of<br />
changing the atmosphere in the container. Therefore the<br />
true value of the surface current decay time τ d
98<br />
surface current in p-n junctions on GaP // Photoelectronics.<br />
— 2005. — No. 14. — P. 97 — 100.<br />
8. Ptashchenko F. O. Effect of ammonia vapors on surface currents<br />
in InGaN p-n junctions // Photoelectronics. — 2007. —<br />
No. 17. — P. 113 — 116.<br />
9. Ïòàùåíêî Ô. Î. Âïëèâ ïàð³â àì³àêó íà ïîâåðõõíåâèé<br />
ñòðóì ó êðåìí³ºâèõ p-n ïåðåõîäàõ // ³ñíèê ÎÍÓ, ñåð.<br />
Ô³çèêà. — 2006. — Ò. 11, ¹. 7. — Ñ. 116 — 119.<br />
UDC 621.315.592<br />
O. O. Ptashchenko, F. O. Ptashchenko, V. V. Shugarova<br />
10. Ptashchenko O. O., Ptashchenko F. O., Yemets O. V. Effect<br />
of ammonia vapors on the surface current in silicon p-n<br />
junctions // Photoelectronics. — 2006. — No. 16. — P. 89 —<br />
93.<br />
11. Ptashchenko O. O., Ptashchenko F. O., Masleyeva N. V. et<br />
al. Effect of sulfur atoms on the surface current in GaAs p-n<br />
junctions // Photoelectronics. — 2007. — No. 17. — P. 36 —<br />
39.<br />
TUNNEL SURFACE CURRENT IN GaAs–AlGaAs P-N JUNCTIONS, DUE TO AMMONIA MOLECULES ADSORPTION<br />
Abstract<br />
The influence of ammonia vapors on I-V characteristics of forward and reverse currents and kinetics of surface currents in GaAs-<br />
AlGaAs p-n junctions with degenerated p + region was studied. It is shown that ammonia molecules adsorption, under sufficiently high<br />
NH 3 partial pressure, forms in p-AlGaAs a surface conducting channel with degenerated electrons. P-n junctions with degenerated p +<br />
region have higher gas sensitivity at reverse bias than at forward bias. This effect is explained by tunnel injection of electrons into the conducting<br />
channel from the degenerated p + region at a reverse bias. The rise time of the surface current in an ammonia vapor atmosphere<br />
of ~20 s is due to filling up deep electron traps.<br />
Key words: tunnel surfeace current, adsorbption, junctions.<br />
ÓÄÊ 621.315.592<br />
À. À. Ïòàùåíêî, Ô. À. Ïòàùåíêî, Â. Â. Øóãàðîâà<br />
ÒÓÍÍÅËÜÍÛÉ ÏÎÂÅÐÕÍÎÑÒÍÛÉ ÒÎÊ Â P-N ÏÅÐÅÕÎÄÀÕ ÍÀ ÎÑÍÎÂÅ GaAs–AlGaAs, ÎÁÓÑËÎÂËÅÍÍÛÉ<br />
ÀÄÑÎÐÁÖÈÅÉ ÌÎËÅÊÓË ÀÌÌÈÀÊÀ<br />
Ðåçþìå<br />
Èññëåäîâàíî âëèÿíèå ïàðîâ àììèàêà íà ÂÀÕ ïðÿìîãî è îáðàòíîãî òîêîâ è íà êèíåòèêó ïîâåðõíîñòíûõ òîêîâ â p-n<br />
ïåðåõîäàõ íà îñíîâå GaAs-AlGaAs ñ âûðîæäåííîé p + îáëàñòüþ. Ïîêàçàíî, ÷òî àäñîðáöèÿ ìîëåêóë àììèàêà ïðè äîñòàòî÷íî<br />
âûñîêîì ïàðöèàëüíîì äàâëåíèè NH 3 ñîçäàåò â p-AlGaAs ïîâåðõíîñòíûé ïðîâîäÿùèé êàíàë ñ âûðîæäåííûìè ýëåêòðîíàìè.<br />
P-n ïåðåõîäû ñ âûðîæäåííîé p + îáëàñòüþ èìåþò áîëåå âûñîêóþ ãàçîâóþ ÷óâñòâèòåëüíîñòü ïðè îáðàòíîì ñìåùåíèè, ÷åì<br />
ïðè ïðÿìîì ñìåùåíèè. Ýòîò ýôôåêò îáúÿñíÿåòñÿ òóííåëüíîé èíæåêöèåé ýëåêòðîíîâ â ïðîâîäÿùèé êàíàë èç âûðîæäåííîé<br />
p + îáëàñòè ïðè îáðàòíîì ñìåùåíèè. Âðåìÿ íàðàñòàíèÿ ïîâåðõíîñòíîãî òîêà ~20 ñ â ïàðàõ àììèàêà ñâÿçàíî ñ çàïîëíåíèåì<br />
ãëóáîêèõ ýëåêòðîííûõ ëîâóøåê.<br />
Êëþ÷åâûå ñëîâà: òóííåëüíûé ïîâåðõíîñòíûé òîê, àäñîðáöèÿ, ïåðåõîä.<br />
ÓÄÊ 621.315.592<br />
Î. Î. Ïòàùåíêî, Ô. Î. Ïòàùåíêî, Â. Â. Øóãàðîâà<br />
ÒÓÍÅËÜÍÈÉ ÏÎÂÅÐÕÍÅÂÈÉ ÑÒÐÓÌ Â P-N ÏÅÐÅÕÎÄÀÕ ÍÀ ÎÑÍβ GaAs–AlGaAs, ÎÁÓÌÎÂËÅÍÈÉ<br />
ÀÄÑÎÐÁÖ²ªÞ ÌÎËÅÊÓË À̲ÀÊÓ<br />
Ðåçþìå<br />
Äîñë³äæåíî âïëèâ ïàð³â àì³àêó íà ÂÀÕ ïðÿìîãî ³ çâîðîòíîãî ñòðóì³â òà íà ê³íåòèêó ïîâåðõíåâèõ ñòðóì³â ó p-n ïåðåõîäàõ<br />
íà îñíîâ³ GaAs-AlGaAs ç âèðîäæåíîþ p + îáëàñòþ. Ïîêàçàíî, ùî àäñîðáö³ÿ ìîëåêóë àì³àêó ïðè äîñòàòíüî âèñîêîìó ïàðö³àëüíîìó<br />
òèñêó NH 3 ñòâîðþº â p-AlGaAs ïîâåðõíåâèé ïðîâ³äíèé êàíàë ç âèðîäæåíèìè åëåêòðîíàìè. P-n ïåðåõîäè ç âèðîäæåíîþ<br />
p + îáëàñòþ ìàþòü âèùó ãàçîâó ÷óòëèâ³ñòü ïðè çâîðîòíîìó çì³ùåíí³, í³æ ïðè ïðÿìîìó çì³ùåíí³. Öåé åôåêò ïîÿñíþºòüñÿ<br />
òóíåëüíîþ ³íæåêö³ºþ åëåêòðîí³â â ïðîâ³äíèé êàíàë ³ç âèðîäæåíî¿ p + îáëàñò³ ïðè çâîðîòíîìó çì³ùåíí³. ×àñ íàðîñòàííÿ ïîâåðõíåâîãî<br />
ñòðóìó ~20 ñ â ïàðàõ àì³àêó ïîâ’ÿçàíèé ³ç çàïîâíåííÿì ãëèáîêèõ åëåêòðîííèõ ïàñòîê.<br />
Êëþ÷îâ³ ñëîâà: òóíåëüíèé ïîâåðõíåâèé ñòðóì, àäñîðáö³ÿ, ïåðåõ³ä.
UDC 544.187.2; 621.315.59<br />
SH. D. KURMASHEV, T. M. BUGAEVA, T. I. LAVRENOVA, N. N. SADOVA<br />
Odessa I.I.Mechnicov National University<br />
Odessa, 65026, Ukraine, e-mail: kurm@mail.css.od.ua. Tel. (0482) — 746-66-58.<br />
INFLUENCE OF THE GLASS PHASE STRUCTURE ON THE RESISTANCE<br />
OF THE LAYERS IN SYSTEM “GLASS-RuO 2 ”<br />
Influence of quantitative and qualitive (phase and granule-metric) composition of initial components<br />
on the electro physical properties of thick films on the base of the systems “glass — clusters<br />
RuO 2 , Bi 2 Ru 2 O 7 ” was investigated. At the fixed values of functional material content m f (RuO 2 ) and<br />
glass phase m g , the increase of mass of crystalline phase m cr leads to decrease of conductivity, therefore<br />
to the increase of resistance of resistive layer.<br />
1. INTRODUCTION<br />
Development of submicron and nanotechnology<br />
in electronics involves not only active elements (lasers,<br />
photodetectors and etc) but also those elements,<br />
which are recognized as passive. They also include the<br />
resistors of the integrated circuits. As it is generally<br />
known, the resistance compositions on the basis of<br />
dioxide of ruthenium (RuO 2 ) have good electro physical<br />
properties [1]. Thick-film structures on the basis<br />
of the systems “glass-RuO 2 , Bi 2 Ru 2 O 7 ” used as the<br />
conductive resistive elements of the hybrid integrated<br />
circuits, are little affected by high temperatures, as dioxide<br />
of ruthenium does not dissolve in a glass matrix.<br />
It allows to increase annealing temperature of resistance<br />
pastes up to 1000 î Ñ. But the problems, related<br />
to structural-phases transitions resulted from external<br />
factors, which influence electro physical properties<br />
of thick films on the base of RuO 2, Bi 2 Ru 2 O 7 , are still<br />
unsolved. There is no one statifactory model of conductivity<br />
mechanisms in the separate components<br />
of ceramic layers and also information about contribution<br />
of micro- and nano-geometry defects to the<br />
conductivity mechanisms of thick resistance films.<br />
Composition materials belong to multi phase heterosystems,<br />
which include components with different<br />
physical and chemical properties.<br />
The study of properties of initial components,<br />
organic and inorganic compositions, conductive and<br />
dielectric phases, morphology and particle size distribution<br />
become the primary goals in production of<br />
thick-film elements.<br />
In the present work , the influence of quantitative<br />
and qualitive composition (phase and granule-metric)<br />
of initial components on the electro physical parameters<br />
of thick films on the base of the systems “glass —<br />
clusters RuO 2 , Bi 2 Ru 2 O 7 ” was investigated.<br />
2. EXPERIMENTAL<br />
The main components of the composition systems<br />
of thick film elements are:<br />
– small-dispersion powders of functional material<br />
(metals, oxides of metals), which provide formation of<br />
conducting paths;<br />
© Sh. D. Kurmashev, T. M. Bugaeva, T. I. Lavrenova, N. N. Sadova, 2009<br />
– special glass frit carrying out the role of permanent<br />
binder.<br />
Functional materials (conducting phase) are<br />
brought into paste as ultrafine particles with the maximal<br />
size lower than 5 μm. The glasses are used as permanent<br />
binder. On one side, the glass frit provides the<br />
adherence of metal-enamel elements. On the other<br />
side it creates “hard framework”, fixing position of<br />
conducting particles inside the structure.<br />
The shape and dispersion of particles of conducting<br />
phase (RuO 2 ) strongly depends on the ruthenium<br />
dioxide powder fabrication method. Usually, it is obtained<br />
by hydrochloric ruthenium decomposition. At<br />
a temperature a 300-400 î Ñ ruthenium dioxide forms<br />
as ultrafine spherical particles. The optimal size of<br />
annealed particles is (0,05 ÷ 0,1) μm. Maximal size<br />
of the ruthenium dioxide particles shouldn’t exceed<br />
(0,2 ÷ 0,3) parts of thickness of annealing layer. Character<br />
of conductivity of resistive layers concerns by<br />
potential barrier height of dielectric layer between<br />
conducting particles. If the dielectric layer between<br />
conducting particles is less than 100 Å, the tunnel<br />
current is the basic mechanism of conductivity. If the<br />
layer is more than 100 Å, tunnel effect is improbable<br />
and charge carriers with energy higher than the height<br />
of barrier can go over it. Thermal emission becomes<br />
the basic mechanism of conductivity.<br />
The influence of permanent binder (glasses of different<br />
types) seems to be not so strong in chemical<br />
interaction with a conducting phase, but it increases<br />
through moistening and dissolution of its particles.<br />
Moistening of functional material by glass and chemical<br />
activity of glass have more important meaning. If<br />
glass forms a thick continuous layer round every conductive<br />
particle, the contact between particles is violated.<br />
Consequently, it is needed, that glass moistened<br />
particles not fully, but rather enough, that particles<br />
were fixed in a matrix.<br />
Electro physical properties of capacitance-resistance<br />
elements largely depend on ratio of conductor<br />
phase and permanent binder concentrations. Dependence<br />
of electro physical properties of composition<br />
structures on the basis of “glass-RuO 2 ” from ratio of<br />
conducting phase RuÎ 2 and glass concentrations, sizes<br />
of particles of glass and temperature of annealing<br />
was investigated. The films were made from powders<br />
of lead-boron-silicate glass of marked ¹ 279, 2005<br />
99
(PbÎ, S³Î 2 , B 2 O 3 , Al 2 O 3 ) with the fixed sizes of particles<br />
(0.5; 1; 3 and 5 μm) and functional material RuÎ 2<br />
with the sizes of particles (0.05 ÷ 0.1) μm.<br />
The system of analysis of images “QUANTIM-<br />
ET — 720” and raster electronic microscope were used<br />
for researches. Investigations of glasses phase composition<br />
were carried by x-ray technique on DRON-2<br />
with silicon grating monochromator (voltage 16 kV,<br />
intensity of current 2 mA).<br />
Dependence of resistance of the resistors from<br />
concentration ratio of conductor phase and glass at<br />
the fixed temperature of annealing (870 î Ñ) has been<br />
obtained (fig. 1).<br />
Fig. 1. Dependence of surface resistivity of thick layers from<br />
concentration ratio of conductor phase and glass. Size of glass particles,<br />
μm: 1 — 0.5; 2 — 1; 3 — 3; 4 — 5<br />
The samples with low content of ruthenium dioxide<br />
had the most influence from glass-frit particle size<br />
on resistance of resistors. Resistance of films increased<br />
with the enhance of glass portion. The highest rate of<br />
resistance raise took place for glass-frit with the particles<br />
size of 0,5 μm. With increase of RuO 2 concentration,<br />
the resistance approaches to the saturated value<br />
and does not depend on the glass-frit particles size.<br />
Dependence of resistance from the particle sizes<br />
for high resistivity films is affected by coalescence<br />
of the particles and the influence of the components<br />
dispersion on the geometrical sizes of. With glass particles<br />
size decreasing, the current chainlets length<br />
increases and their cross section area decreases. It is<br />
observed mixed type of conductivity in the systems<br />
“RuÎ 2 — glass”. The charge transport processes take<br />
place in the conducting phase and glass-frit. In layers<br />
with high resistivity, the main contribution in conductivity<br />
is performed by glass-frit. Therefore, the state of<br />
this phase acts important role in the process of charge<br />
transport.<br />
If the influence of high-quality composition of<br />
conducting (functional) material is usually investigated<br />
in details, usually, the glass is considered an<br />
amorphous homogeneous environment. However, in<br />
the process of investigations it is set, that glasses can<br />
be crystallized as result of heat treatment. The x-ray<br />
radiation can stimulate crystallization [2]. In addition,<br />
100<br />
particles of functional material can become the nucleation<br />
centers of crystallization of glass matrix. The<br />
presence of the crystallites in a matrix can cause local<br />
changes of glass melt point and coating of functional<br />
material particles by glass at heat treatment of film. It<br />
leads to change of the parameters of the formed layer.<br />
The phase composition of glasses has been determined<br />
by XRD method. At first, the initial content of<br />
the particles has been investigated. It was found XRD<br />
peak at θ=15,50 (d = 3,35 Å) on the XRD diagram of<br />
glasses ¹2005 in the initial state. The XRD diagram<br />
of glass ¹279 did not have any peaks, related to crystalline<br />
phase.<br />
For determination of influence of heat treatment<br />
on the phase state of glass, it was performed the annealing<br />
procedure at 8700Ñ during 10 minutes. At the<br />
same time, the samples were separated on two groups<br />
to take in account influence of x-ray radiation. The<br />
first group was annealed with subsequent XRD analysis.<br />
The second one was processed by the consequence<br />
XRD-annealing-XRD.<br />
After annealing, XRD peak at θ=15,50 of ¹2005<br />
samples (first group) increased in comparison to the initial<br />
state. Weak peaks appeared at θ=12,10 (d = 4,27 о<br />
А )<br />
and θ=29,50° (d = 1,82 о<br />
А ) in XRD curve of the second<br />
group after x-ray irradiation and annealing, that<br />
testifies to the increase of concentration of crystalline<br />
phase. Identification of the observed peaks within<br />
card index of ASTM showed that the peaks belonged<br />
to α-SiO (quartz) modification. Comparison of the<br />
2<br />
XRD diagrams of glasses ¹2005 before and after heat<br />
treatment showed that X-ray reflections from crystalline<br />
phase increased as a result of increase of volume of<br />
crystalline phase in these glasses after heat treatment.<br />
Glasses ¹279 of first and second groups remained<br />
amorphous.<br />
The obtained results allowed to make some conclusions<br />
about the influence of heat treatment on<br />
phase composition of glasses:<br />
– glass ¹279 had strong amorphous structure,<br />
the crystalline phase nucleation disappeared by time<br />
and never appear after the heat treatment;<br />
– glass ¹2005 already contained the crystalline<br />
phase in the initial state, the volume of crystalline<br />
phase increased with time in all stages of heat treatment.<br />
Thus, it is found that heat treatment caused formation<br />
of new phases, and also rebuilding of energetic<br />
zones of the system. As the crystalline phase found<br />
in glass α-SiO had the temperature of melting over<br />
2<br />
15000Ñ, and coalescence of resistance layers took<br />
place usually at 870 0Ñ. The crystalline phase of glass<br />
didn’t melt and there were local structural deviations<br />
in the volume of matrix, that influenced the formation<br />
of conducting chainlets. In addition, the random<br />
breaks of conducting chainlets made disorder of structure<br />
of chainlets of conductivity in the volume of the<br />
film. It is set that presence of crystalline phase α-SiO2 in glass increases resistance of the sample (∼ 10%).<br />
Properties of glass phase play an important role in<br />
conductivity of films. In fiq.2, the dependence of surface<br />
resistivity of thick films from the percent concentration<br />
of crystalline phase m in glass-frit is presented.<br />
cr<br />
It is seen that with the increase of m the resistance of<br />
cr,<br />
films increased.
Fig. 2. Dependence of surface resistivity of films from percent<br />
concentration of crystalline phase m cr in glass-frit. Table of<br />
contents of glass-frit in the system “glass — clusters RuO 2 ” m g , %:<br />
1 — 40; 2 — 50; 3 — 60<br />
3. DISCUSSION<br />
Electrical properties of the films on the base of the<br />
systems “glass — clusters RuO 2 , Bi 2 Ru 2 O 7 ” are defined<br />
by the mechanisms of conductivity, which strongly depend<br />
on physical properties of initial components and<br />
microstructure of layers. The structure of films depends<br />
on the technological factors of their fabrication<br />
and properties of initial components. Phase composition<br />
of glass-frit is also an important factor.<br />
In [2] it was obtained mixed character of conductivity<br />
as combination of processes which flow in conducting<br />
phase and glass-phase. The conducting phase<br />
had metallic conductivity. Charge transport through<br />
the thin layers of glass-phase, surrounding the conducting<br />
phase, takes place by means of tunneling effect<br />
in a low energetic zone, which appears after doping of<br />
glass with ions, diffunding from the conducting phase.<br />
Through research of resistance layers on basis RuO 2 ,<br />
it was determined traces of presence of cristobalite in<br />
the layers– one of crystalline modifications SiO 2 , which<br />
decreased conductivity of resistors. According to that, it<br />
is interesting to develop a model which explains influencing<br />
of crystalline phase α-SiO 2 in amorphous glass<br />
matrix on conductivity of thick resistance films, based<br />
on the system “glass — clusters RuO 2 ”.<br />
A thick-film element can be presented as aggregate<br />
of conducting chainlets from one electrode to other,<br />
which consist of conducting particles of functional<br />
material. Conducting ³-chain appears with probability<br />
p ³ , which includes probability of that all elements of<br />
chainlet are conductors (p 1 ) and probability of continuity<br />
of chainlet in a glass matrix from one electrode to<br />
other (p 2 ). It means that p ³ = p 1 ⋅ p 2 . In this model probability<br />
p 1 of conductivities of all elements of chainlet is<br />
proportional to mass part of functional material m f in<br />
bulk of dry powder in paste m p : p 1 = k 1 ∙m f / m p . Here k 1<br />
is coefficient of proportion. Probability of formation<br />
of continuous chainlet from one electrode to other is<br />
proportional to mass of functional material m f and inversely<br />
proportional to mass of glass-frit m g : p 2 = k 2 ∙<br />
m f /m g . Here, k 2 is coefficient of proportion. Mass of<br />
powder m p = m f + m g .<br />
In the case of presence of crystalline phase in<br />
glass it is necessary to take into account another factor.<br />
The presence of crystalline phase SiO in fusible<br />
2<br />
glass is equivalently adding unfire-polished particles<br />
in the glass, beacause temperature of melting of any<br />
of modifications SiO is considerably higher than the<br />
2<br />
temperature of sintering of layer. Under sintering, the<br />
local regions of the “not-melts” are formed in the glass,<br />
which prevent the process of forming of structure in the<br />
limited areas. The “not-melts” restrict the distribution<br />
of liquid glass and formation of homogeneous sintered<br />
structure. The particles of functional material in these<br />
regions do not form good contacts because of lack of<br />
pressing forces which arise up at sintering of glass. Thus,<br />
the non sintered and non conductive contacts appear,<br />
what is equal to the breaks of leading chainlets.<br />
For calculation of possibility of formation of nonconducting<br />
contacts, it is necessary to put p (the prob-<br />
3<br />
ability of that all contacts are conducting between the<br />
elements of chainlet) in equation for p , that is equal<br />
³<br />
to p = p ⋅ p p . With the increase of mass of crystalline<br />
³ 1 2 3<br />
phase m in glass-frit and functional material powder<br />
cr<br />
mass m , the probability p is decreased: p = 1–k ∙m /m .<br />
p 3 3 3 cr p<br />
Here k is coefficient of proportion. From the resulted<br />
3<br />
equations for p , p and p , it is not difficult to get p .<br />
1 2 3 ³<br />
As a thick-film resistor in this case is presented<br />
as system which consists of aggregate of conducting<br />
N<br />
chainlets, its conductivity equals σ= ∑ σipi<br />
, where<br />
i=<br />
1<br />
σ is conductivity of i –th chainlet, N is the amount<br />
i<br />
of chainlets.<br />
If all chainlets are formed in identical terms, it is<br />
possible to assume that p =p. Then<br />
³<br />
2<br />
N mf(<br />
m 3 ) N<br />
f + mg −k<br />
mcr<br />
∑ i 1 2 2 ∑ i .<br />
σ= p σ = k k<br />
σ<br />
i= 1 mm g p<br />
i=<br />
1<br />
It is obviously seen from this expression, that increase<br />
of maintenance of crystalline constituent m in cr<br />
a glass phase m results to increase of resistance of thick<br />
g<br />
film, what was confirmed experimentally (fig.2).<br />
CONCLUSIONS<br />
Most influence of sizes of particles of glass-frit on<br />
resistance of resistors at the fixed temperature of annealing<br />
takes place for samples with low content of<br />
ruthenium dioxide. Resistance of layers increases with<br />
the increase of content of glass.<br />
At the fixed values of functional material content<br />
m (RuO ) and glass phase m , the increase of mass of<br />
f 2 g<br />
crystalline phase m leads to decrease of conductiv-<br />
cr<br />
ity, therefore to the increase of resistance of resistive<br />
layer.<br />
References<br />
1. Ïàíîâ Ë. È. Îïûò ñîâåðøåíñòâîâàíèÿ òîëñòîïëåíî÷íîé<br />
òåõíîëîãèè / Ë. È. Ïàíîâ, Ð. Ã. Ñèäîðåö // Òåõíîëîãèÿ è<br />
êîíñòðóèðîâàíèå â ýëåêòðîííîé àïïàðàòóðå. — 2002. —<br />
¹1. — Ñ. 43-46.<br />
2. Êóðìàøåâ Ø. Ä. Âëèÿíèå êðèñòàëëè÷åñêîé ôàçû SiO 2<br />
íà ýëåêòðîôèçè÷åñêèå êîìïîíåíòû êîìïîçèöèîííûõ<br />
ñòðóêòóð íà áàçå ñòåêëî-Bi 2 Ru 2 O 7 / Ø. Ä. Êóðìàøåâ,<br />
Ò. È. Ëàâðåíîâà, Ò. Í. Áóãàåâà // Òåçèñû äîêë. XXII<br />
íàó÷í. êîíô. ñòðàí ÑÍà “Äèñïåðñíûå ñèñòåìû”. —<br />
Îäåññà, 2006. — Ñ. 215.<br />
101
102<br />
UDC 544.187.2; 621.315.59<br />
Sh. D. Kurmashev, T. M. Bugaeva, T. I. Lavrenova, N. N. Sadova<br />
INFLUENCE OF THE GLASS PHASE STRUCTURE ON THE RESISTANCE OF THE LAYERS IN SYSTEM<br />
“GLASS-RuO 2 ”<br />
Influence of quantitative and qualitive (phase and granule-metric) composition of initial components on the electro physical properties<br />
of thick films on the base of the systems “glass — clusters RuO 2 , Bi 2 Ru 2 O 7 ” was investigated. At the fixed values of functional<br />
material content m f (RuO 2 ) and glass phase m g , the increase of mass of crystalline phase m cr leads to decrease of conductivity, therefore to<br />
the increase of resistance of resistive layer.<br />
Key words: sructure, phase, resistance, layer.<br />
ÓÄÊ 544.187.2; 621.315.59<br />
ÂËÈßÍÈÅ ÑÒÐÓÊÒÓÐÛ ÑÒÅÊËßÍÍÎÉ ÔÀÇÛ ÍÀ ÑÎÏÐÎÒÈÂËÅÍÈÅ ÐÅÇÈÑÒÈÂÍÛÕ ÏËÅÍÎÊ Â ÑÈÑÒÅÌÅ<br />
“ÑÒÅÊËÎ-RuO 2 ”<br />
Ø. Ä. Êóðìàøåâ, Ò. Í. Áóãàåâà, Ò. È. Ëàâðåíîâà, Í. Í. Ñàäîâà<br />
Èññëåäîâàíî âëèÿíèå êîëè÷åñòâåííîãî è êà÷åñòâåííîãî (ôàçîâûé è ãðàíóëîìåòðè÷åñêèé) ñîñòàâà èñõîäíûõ êîìïîíåíòîâ<br />
íà ýëåêòðîôèçè÷åñêèå ïàðàìåòðû òîëñòûõ ïëåíîê íà áàçå ñèñòåì “ñòåêëî — êëàñòåðû RuO 2 , Bi 2 Ru 2 O 7 ”. Ïðè çàäàííûõ<br />
âåëè÷èíàõ ñîäåðæàíèÿ ôóíêöèîíàëüíîãî ìàòåðèàëà (RuO 2 ) è ñòåêëÿííîé ôàçû óâåëè÷åíèå ìàññû êðèñòàëëè÷åñêîé ôàçû<br />
ïðèâîäèò ê óâåëè÷åíèþ ñîïðîòèâëåíèÿ ðåçèñòèâíîé ïëåíêè.<br />
Êëþ÷åâûå ñëîâà: ñòðóêòóðà, ôàçà, ñîïðîòèâëåíèå, ïë¸íêà.<br />
ÓÄÊ 544..187.2; 621.315.59<br />
ÂÏËÈ ÑÒÐÓÊÒÓÐÈ ÑÊËßÍί ÔÀÇÈ ÍÀ ÎϲРÐÅÇÈÑÒÈÂÍÈÕ Ï˲ÂÎÊ Â ÑÈÑÒÅ̲ “ÑÊËÎ-RuO 2 ”<br />
Ø. Ä. Êóðìàøåâ, Ò. Í. Áóãàåâà, Ò. È. Ëàâðåíîâà, Í. Í. Ñàäîâà<br />
Äîñë³äæåíî âïëèâ ê³ëüê³ñíîãî òà ÿê³ñíîãî (ôàçîâèé ³ ãðàíóëîìåòðè÷íèé) ñêëàäó âèõ³äíèõ êîìïîíåíò³â íà åëåêòðîô³çè÷í³<br />
ïàðàìåòðè òîâñòèõ ïë³âîê íà áàç³ ñèñòåì “ñêëî — êëàñòåðè RuO 2 , Bi 2 Ru 2 O 7 ”. Ïðè çàäàíèõ âåëè÷èíàõ âì³ñòó ôóíêö³îíàëüíîãî<br />
ìàòåð³àëó (RuO 2 ) ³ ñêëÿíî¿ ôàçè çá³ëüøåííÿ ìàñè êðèñòàë³÷íî¿ ôàçè ïðèçâîäèòü äî çìåíøåííÿ âåëè÷èíè ïðîâ³äíîñò³,<br />
òîáòî äî çá³ëüøåííÿ îïîðó ðåçèñòèâíî¿ ïë³âêè.<br />
Êëþ÷îâ³ ñëîâà: ñòðóêòóðà, ôàçà, îï³ð, ïë³âêà.
UDÑ 539.19+539.182<br />
A. V. IGNATENKO, A. A. SVINARENKO, G. P. PREPELITSA, T. B. PERELYGINA, V. V. BUYADZHI<br />
Odessa National Polytechnical University<br />
OPTICAL BI-STABILITY EFFECT FOR MULTI-PHOTON ABSORPTION<br />
IN ATOMIC ENSEMBLES IN A STRONG LASER FIELD<br />
Within the density matrices formalism it is considered the multi-photon absorption in the atomic<br />
ensemble of the two-level atoms, which interacts with resonant laser filed. The hysteresis dependence<br />
of output amplitude on the input electromagnetic wave under 3-photon absorption (bi-stability effect)<br />
is found in the caesium vapour (the pressure 0.1 Torr, particle density 3.5⋅10 15 cm -3 ).<br />
INTRODUCTION<br />
The interaction of atomic ensembles s with the<br />
external alternating fields, in particular, laser fields,<br />
has been the subject of intensive experimental and<br />
theoretical investigation in the quantum optics and<br />
electronics [1-15]. The appearance of the powerful<br />
laser sources allowing to obtain the radiation field amplitude<br />
of the order of atomic field in the wide range<br />
of wavelengths results to systematic investigations of<br />
the nonlinear interaction of radiation with atomic ensembles.<br />
A whole number of interesting non-linear<br />
optical phenomena may take a place, in particular,<br />
multi=photon absorption and emission, multi-photon<br />
excitation and ionization, at last different be-stability<br />
and hysteresis phenomena. Another important topic<br />
is a problem of governing and control of non-linear<br />
processes in a stochastic, multi-mode laser field [4,5].<br />
The principal aim of quantum coherent control is to<br />
steer an atomic ensemble towards a desired final state<br />
through interaction with light while simultaneously<br />
inhibiting paths leading to undesirable outcomes. This<br />
type of quantum interference is inherent in non-linear<br />
multi-photon processes. Controlling mechanisms<br />
have been proposed and demonstrated for atomic,<br />
molecular and solid-state systems [1-3]. In ref. [12]<br />
the effect of pulse-shaping on transient populations<br />
of the excited Rb atoms ensembles is tested. At present<br />
time, a progress is achieved in the description of<br />
the processes of interaction atoms with the harmonic<br />
emission field [1,2]. But in the realistic laser field the<br />
according processes are in significant degree differ<br />
from ones in the harmonic field. The latest theoretical<br />
works claim a qualitative study of the phenomenon<br />
though in some simple cases it is possible a quite acceptable<br />
quantitative description [1-3,15].<br />
Among existed approaches one could mention the<br />
Green function method, the density-matrix formalism,<br />
time-dependent density functional formalism,<br />
direct numerical solution of the quantum-mechanical<br />
equations, multi-body multi-photon approach, the<br />
time-independent Floquet formalism, S-matrix Gell-<br />
Mann and Low formalism etc. [1-13]. The effects of<br />
the different laser line shape on the intensity and spectrum<br />
of resonance fluorescence from a two-level atom<br />
are studied in many papers (c.f. [1-5]). Nevertheless,<br />
in a whole one can note that a problem of correct description<br />
of the non-linear atomic dynamics in a sto-<br />
© A. V. Ignatenko, A. A. Svinarenko, G. P. Prepelitsa, T. B. Perelygina, V. V. Buyadzhi, 2009<br />
chastic, multi-mode laser field is quite far from the<br />
final solution. It requires developing the advanced approaches<br />
to description of multi-photon dynamics of<br />
atomic ensembles in a strong laser field and adequate<br />
treating different non-linear optical and photon-correlation<br />
effects. In this paper within the density matrices<br />
formalism it has been considered the multi-photon<br />
absorption in the atomic ensemble of the two-level<br />
atoms, which interacts with resonant laser filed. The<br />
bi-stability effect for three-photon absorption in the<br />
Cs vapors is found.<br />
Figure 1. Calculated hysteresis dependence of the<br />
output field upon the input electromagnetic wave amplitude<br />
for the 3-photon absorption in the Cs vapor<br />
2. OPTICAL BI-STABILITY EFFECT FOR<br />
MULTI-PHOTON ABSORPTION<br />
Below we consider the optical passing bi-stability<br />
effect for multiphoton absorption in the atomic<br />
ensemble and calculate the hysteresis dependence of<br />
output amplitude on the input electromagnetic wave<br />
under multi-photon absorption. We use the formalism<br />
of the density matrices [3,4].<br />
Let us suppose that the ensemble of N two-level<br />
atoms interacts with a resonant laser radiation field.<br />
The sum of frequencies of the external light fields is<br />
ω+ω s =ω o (where ω o is the quantum transition frequency<br />
in a two-level system). Besides, there are the<br />
103
“exchange forces”, which may have the different nature.<br />
In particular they are connected with exchange<br />
through the radiation field, dipole-dipole electrostatic<br />
interaction etc.<br />
Let us suppose that an electromagnetic field (including<br />
two waves) acts on the system:<br />
104<br />
− −<br />
Er ( , t) = E( r, t) + E( r, t)<br />
+ ê.ñ. =<br />
j L j s j<br />
− −<br />
EсL exp( −ω i t+ ikLrj) + Eso( rj)exp( −ω i st)<br />
+ c.<br />
. (1)<br />
Here ω+ω =ω -Δ, Δ is the detuning from the two-<br />
s o<br />
photon resonance, r is a radius-vector of the “j” atom.<br />
j<br />
The resulting field for frequency ω is defined likely (1)<br />
s<br />
as follows:<br />
− −<br />
Es ( rj, t) = Eso( rj)exp( −iω st)<br />
+<br />
) )<br />
+ ∂ ∂<br />
∑<br />
2<br />
(1/ c ) (1/ rij )<br />
2<br />
[[ Psi ( t') nij ] nij ] /<br />
2<br />
t .<br />
ii ( ≠ j)<br />
Here t’=t-r ij /c=t-τ ij , σ i are the Pauli matrices and<br />
(2)<br />
+<br />
Psi () t =−rELσi−()exp( t −ω i t−ikLrj) - polarization, (3)<br />
−1<br />
r = h [( d<br />
)<br />
e ) d /( ω +ω ) + d ( d<br />
)<br />
e ) /( ω +ω ) ;<br />
∑<br />
α<br />
bα L αa bα bα αa L bα s<br />
The equations for atomic variables, which describe<br />
the processes of the two-photon amplifying and<br />
absorption, have a standard form [3-5]:<br />
∂σ j+ / ∂t−i( ω 0 + i/ T2)<br />
σ j+<br />
=<br />
+ +<br />
=−(/ i h)<br />
rEs ( rj,) t EL( rj,) t σj3,<br />
o<br />
∂σ j3 / ∂ t+ ( σ j3 − σ j3)<br />
T1<br />
=<br />
+ +<br />
= (2 i/ h)<br />
xrE ( r , t) E ( r , t) σ −cc<br />
. .]<br />
s j L j j−<br />
(4)<br />
where σ j3 = σ jaa - σ jbb is a difference of populations in<br />
the “j” atom. Then the full populations difference is<br />
defined as follows:<br />
∑<br />
j<br />
j3(), j3<br />
()/ . (5)<br />
D = σ t σ ≈ D t N<br />
Transition to slowly changing density matrice:<br />
ρ j− =σj−exp[ i( ω+ωs) t− ikLrj] (6)<br />
allows to get the following system of equations for<br />
density matrice:<br />
∂ρ / ∂ t+ i( Δ − iT ) ρ =<br />
= ( iD / h N) d E ( r ) + ( GD /2 N) C ρ ( t),<br />
(7)<br />
j+ −1<br />
2 j+<br />
* +<br />
s j ∑<br />
j<br />
ij j+<br />
∂D/ ∂ t+ ( D− D )/ T ) =<br />
= ρ − −<br />
0 1<br />
* −<br />
(2 iкс / h)<br />
∑(<br />
d Esо( rj)<br />
j<br />
+ . .)<br />
∑∑<br />
−G C [ ρ ( t) ρ ( t) +ρ ( t) ρ ( t)]<br />
i j<br />
ij i− i+ i+ j−<br />
(8)<br />
where G- is a constant of the “k”-photon decay of<br />
atom and Ñ ij =sin k s r ij /( k s r ij ).<br />
Let us introduce the notations for normalizing<br />
amplitudes of the falling and passing fields:<br />
y (4 TT / ) d E so<br />
+<br />
= h<br />
x = (4 TT / h ) [ d ε + d E ], (9)<br />
2 1/2 *<br />
1 2 ,<br />
2 1/2 * + * +<br />
1 2<br />
so so<br />
+<br />
( q) * +<br />
∑ d Es ( rj) j<br />
q( rj)<br />
ε = Ψ<br />
We use the eigen-functions Ψ(q) and eigen-values<br />
of the q matrice of the inter-atomic interaction<br />
Ñ ij . [4]. The system of the differential equations in fact<br />
describes the bi-stable behavior of an ensemble of the<br />
two-level atoms system under the multi-photon absorption<br />
for different possible geometries of the radiated<br />
medium (different Frenel numbers) [5].<br />
3. THE MULTI-PHOTON ABSORPTION IN<br />
THE CESIUM ENSEMBLE<br />
Let us consider a process of the multi-photon absorption<br />
for three-photon absorption in the Cs vapors<br />
(the cesium ensemble). This process is characterized<br />
by the bi-stability effect The corresponding Cs level<br />
energies are presented in table 1 (from ref. [6]). In figure<br />
1 we present the calculated hysteresis dependence<br />
of output field upon the input electromagnetic wave<br />
amplitude for the three-photon absorption in the Cs<br />
vapor (this is corresponding to transition 6F-7D). The<br />
following parameters have been used: particle concentration<br />
n=3.5⋅1015 cm-3 , pressure 0,1 Torr, time<br />
T =5⋅10 2 -10 s and time T =130ns (the time of spontane-<br />
1<br />
ous transition 6F-7D). Really, the time of realizing the<br />
stationary regime is less than Ò . 1<br />
Table 1<br />
Energies of the cesium terms<br />
Term of the ground<br />
state<br />
Cs (6s-2S ) 1/2<br />
31435.5<br />
Term of the excited<br />
state<br />
6p( 2P ) 1/2<br />
6p( 2P ) 3/2<br />
7d( 2D ) 5/2<br />
6f( 2F ) 7/2<br />
Energy of the level,<br />
cm -1<br />
11256<br />
11813<br />
26134<br />
28120<br />
It is obvious, a great interest attracts an experimental<br />
study of the cited effect, that is a non-trivial<br />
task because of the increased energetic of laser fields<br />
and other factors. (Fig.1).<br />
References<br />
1. Batani D., Joachain C J, Matter in super-intense laser fields,<br />
AIP Serie, N. — Y., 2006. — 650P.<br />
2. Ullrich C.A., Erhard S., Gross E.K.U., Superintense Laser<br />
Atoms Physics, New York : Acad.Press., 2006. — 580P.<br />
3. Delone N.B., Interaction of laser radiation with substance, :<br />
Nauka, 1999. — 278P.<br />
4.`Àêóëèí Â.Ì., Êàðëîâ Í.Á. Èíòåíñèâíûå ðåçîíàíñíûå<br />
âçàèìîäåéñòâèÿ â êâàíòîâîé ýëåêòðîíèêå. Ìîñêâà.<br />
Íàóêà.1997 Ñ312 ..<br />
5.`Andreev A.V., Emeliyanov V., Ilyinsky Yu., Cooperative phenomena<br />
in optics, Nauka, 1998. — 230P.<br />
6.`Ãëóøêîâ À.Â. Àòîì â ýëåêòðîìàãíèòíîì ïîëå.Êèåâ,<br />
ÊÍÒ, 2005. — Ñ 400.<br />
7. Mercouris T., Nikolaides C.A. Solution of the many-electron<br />
many-photon problem for strong fields: Application to Li -<br />
in one and two-colour laser fields//Phys.Rev.A. — 2003. —<br />
Vol.67. — P.063403-1-063403-12.<br />
8. Mocken G.R., Keitel C.H., Bound atomic dynamics in the<br />
MeV regime// J.Phys. B: At. Mol. Opt. Phys. — 2004. —<br />
Vol.37. — P.L.275-283.<br />
9. Kamta G.L., Starace A.F. Elucidating the mechanisms of<br />
double ionization using intense half-cycle, single-cycle and<br />
double-half-cycle pulses// Phys.Rev.A. — 2003. — Vol.68. —<br />
P.043413-1-043413-11.<br />
10. Luc-Koenig E., Lyras A., Lecomte J. — M., Aymar M. Eigenchannel<br />
R-matrix study of two-photon processes including<br />
above-threshold ionization in magnesium// J.Phys. B: At.<br />
Mol. Opt. Phys. — 1997. — Vol.30. — P.5213-5232.<br />
11. Becker A., Faisal F.H.M., S-matrix analysis of coincident
measurement of two-electron energy distribution for double<br />
ionization of He in an intense laser field//Phys.rev.Lett. —<br />
2002. — Vol.89,N18. — P.193003-1-193003-4.<br />
12. Courade E., Anderlini M., Ciampini D. Etal, Two-photon<br />
ionization of cold rubidiun atoms with near resonant intermediate<br />
state//J.Phys.B. At.Mol.Opt.Phys. — 2004. —<br />
Vol.37. — P.967-979.<br />
13. Glushkov A.V., Khetselius O.Yu., Loboda A.V., Svinarenko<br />
A.A., QED approach to atoms in a laser field: Multi-photon<br />
resonances and above threshold ionization//Frontiers in<br />
UDÑ 539.19+539.182<br />
A. V. Ignatenko, A. A. Svinarenko, G. P. Prepelitsa, T. B. Perelygina, V. V. Buyadzhi<br />
Quantum Systems in Chemistry and Physics (Springer). —<br />
2008. — Vol.18. — P.541-578.<br />
14. Singer K. , Reetz-Lamour M., Amthor T., et al, Suppression<br />
of excitation and spectral broadening induced by interactions<br />
in a cold gas of Rydberg atoms//Phys. Rev.Lett. — 2004. —<br />
Vol.93. — P.163001.<br />
15. Ignatenko A.V., Probabilities of the radiative transitions between<br />
stark sublevels in spectrum of atom in an dc electric<br />
field: new approach//Photoelectronics. — 2007. — N16. —<br />
P.71-74.<br />
OPTICAL BI-STABILITY EFFECT FOR MULTI-PHOTON ABSORPTION IN ATOMIC ENSEMBLES IN A STRONG<br />
LASER FIELD<br />
Abstract<br />
Within the density matrices formalism it is considered the multi-photon absorption in the atomic ensemble of the two-level atoms,<br />
which interacts with resonant laser filed. The hysteresis dependence of output amplitude on the input electromagnetic wave under 3photon<br />
absorption (bi-stability effect) is found in the caesium vapour (the pressure 0.1 Torr, particle density 3.5⋅10 15 cm -3 ).<br />
Key words: atomic ensembles, laser field, multi-photon processes, bi-stability effect.<br />
ÓÄÊ 539.19+539.182<br />
À. Â. Èãíàòåíêî, À. À. Ñâèíàðåíêî, Ã. Ï. Ïðåïåëèöà, T. Á. Ïåðåëûãèíà, Â. Â. Áóÿäæè<br />
ÝÔÔÅÊÒ ÎÏÒÈ×ÅÑÊÎÉ ÁÈ-ÑÒÀÁÈËÜÍÎÑÒÈ ÏÐÈ ÌÍÎÃÎÔÎÒÎÍÍÎÌ ÏÎÃËÎÙÅÍÈÈ ÄËß ÀÒÎÌÍÛÕ<br />
ÀÍÑÀÌÁËÅÉ Â ÑÈËÜÍÎÌ ÏÎËÅ ËÀÇÅÐÍÎÃÎ ÈÇËÓ×ÅÍÈß<br />
Ðåçþìå<br />
 ðàìêàõ ôîðìàëèçìà ìàòðèö ïëîòíîñòè ðàññìîòðåíî ìíîãîôîòîííîå ïîãëîùåíèå â àòîìíûõ àíñàìáëÿõ äâóõóðîâíåâûõ<br />
àòîìîâ, êîòîðûå âçàèìîäåéñòâóþò ñ ïîëåì ðåçîíàíñíîãî ëàçåðíîãî èçëó÷åíèÿ. Îïðåäåëåíà ãèñòåðåçèñíàÿ çàâèñèìîñòü<br />
âûõîäíîãî ïîëÿ îò àìïëèòóäû ïàäàþùåé âîëíû ïðè òðåõôîòîííîì ïîãëîùåíèè (ýôôåêò áè-ñòàáèëüíîñòè) â ïàðàõ öåçèÿ<br />
(ïëîòíîñòü ÷àñòèö 3,5⋅10 15 ñì -3 , äàâëåíèå 0.1 Toðð.).<br />
Êëþ÷åâûå ñëîâà: àòîìíûå àíñàìáëè, ëàçåðíîå ïîëå, ìíîãîôîòîííûå ïðîöåññû, ýôôåêò áè-ñòàáèëüíîñòè.<br />
ÓÄÊ 539.19+539.182<br />
Ã. Â. ²ãíàòåíêî, À. À. Ñâèíàðåíêî, Ã. Ï. Ïðåïåëèöà, T. Á. Ïåðåëèã³íà, Â. Â. Áóÿäæ³<br />
ÅÔÅÊÒ ÎÏÒÈ×Íί Á²-ÑÒÀÁ²ËÜÍÎÑÒ² ÏÐÈ ÁÀÃÀÒÎÔÎÒÎÍÍÎÌÓ ÏÎÃËÈÍÀÍͲ ÄËß ÀÒÎÌÍÈÕ<br />
ÀÍÑÀÌÁË²Â Ó ÑÈËÜÍÎÌÓ ÏÎ˲ ËÀÇÅÐÍÎÃÎ ÂÈÏÐÎ̲ÍÞÂÀÍÍß<br />
Ðåçþìå<br />
 ìåæàõ ôîðìàë³çìó ìàòðèöü ãóñòèíè ðîçãëÿíóòî áàãàòîôîòîííå ïîãëèíàííÿ â àòîìíèõ àíñàìáëÿõ äâîð³âíåâèõ àòîì³â,<br />
ùî âçàºìîä³þòü ç ïîëåì ðåçîíàíñíîãî ëàçåðíîãî âèïðîì³íþâàííÿ. Âèçíà÷åíà ã³ñòåðåçèñíà çàëåæí³ñòü âèõ³äíîãî ïîëÿ â³ä<br />
àìïë³òóäè ïàäàþ÷î¿ õâèë³ ïðè òðüîõôîòîííîìó ïîãëèíàíí³ (åôåêò á³-ñòàá³ëüíîñò³) ó ïàðàõ öåç³ÿ (ãóñòèíà ÷àñòèíîê 3,5⋅10 15<br />
ñì -3 , òèñê 0,1 Toðð.).<br />
Êëþ÷îâ³ ñëîâà: àòîìí³ àíñàìáë³, ëàçåðíå ïîëå, áàãàòîôîòîíí³ ïðîöåñè, åôåêò á³-ñòàá³ëüíîñò³.<br />
105
106<br />
UDC 633.9<br />
L. V. MYKHAYLOVSKA 1 , A. S. MYKHAYLOVSKA 2<br />
1 I.I,Mechnikov Odesa National University, 65082, Odesa, Ukraine<br />
2 Ruhr-Universität Bochum, 44780, Bochum, Germany<br />
Å-mail: lidam@onu.edu.ua<br />
INFLUENCE OF THE STEP IONIZATION PROCESSES ON THE<br />
ELECTRONIC TEMPERATURE IN THIN GAS-DISCHARGE TUBES<br />
Taking into account the processes of the direct and step ionization and on the basis of the closed<br />
system of the balance equations we have performed theoretical analysis of electronic temperature dependence<br />
on the external parameters such as the value of the discharge current, pressure of the working<br />
gas and the diameter of the discharge capillary. The conditions were found which indicate the<br />
decisive role of the step ionization in charged particles creation.<br />
INTRODUCTION<br />
The study of the low-temperature plasma of the<br />
positive column in thin tubes is interesting not only<br />
as of the active medium of the wave-guiding lasers<br />
but also for the possible creation of the small-size energy<br />
sources for the gas laser emitting devices of low<br />
pressure. The out-come characteristics of the devices<br />
which use the gaseous discharge depend on the inner<br />
parameter of the positive column (PC) which is characteristic<br />
for the discharge. A significant number of<br />
theoretical and experimental papers [1-5] is devoted to<br />
the studies of the inner parameters in the active media<br />
of gaseous lasers. In these papers the main attention is<br />
paid to the influence of the discharge current I r the ,<br />
gas pressure p and the inner radius of the discharge<br />
capillary R 0 on the longitudinal electric field strength<br />
E z , the electronic concentration N e , as well as on the<br />
electronic temperature T e [3-6] assuming their Maxwell-type<br />
distribution over the velocity values. The inner<br />
parameters of the discharge processes are defined<br />
by the given external parameters.<br />
The condition required for the stationary gaseous<br />
discharge existence is the temporal stability of the<br />
electronic concentration. The balance of the charged<br />
particles in the plasma of low-pressure PC is formed<br />
by ionization processes in the gaseous volume and the<br />
following deficiency of the charged particles due to the<br />
drift of electrons and ions towards the wall of the tube.<br />
It should be noted that the probability of the charged<br />
particles creation as well as the processes concerning<br />
the decrease of the charged particles, depend on the<br />
electronic temperature.<br />
SCOPE OF THE WORK<br />
In the present paper, basing on the closed system<br />
of equations obtained in [7], for the evaluation<br />
of the inner parameters of PC of the discharge tube,<br />
the stable value of electronic temperature with the simultaneous<br />
account of the direct and step ionization.<br />
Usually, either the ionization of non-excited atoms<br />
(direct ionization), or the ionization of the excited atoms<br />
(step ionization) is taken into account [4, 6]. The<br />
simultaneous account of the direct and step ionization<br />
allows to evaluate and compare the input of these two<br />
processes into atoms’ ionization. The analysis is conducted<br />
with the use of two main approximations — the<br />
diffuse regime of the discharge and the Maxwell distribution<br />
of electrons over the velocity values.<br />
Let us write down the simplified (interpolated)<br />
equations for direct and step ionizations with the simultaneous<br />
account of the balance between the number<br />
of charged particles e N and metastable atoms m N<br />
in the discharge controlled by the diffusion<br />
ν 0iNe +ν miN e=ν adNe. (1)<br />
( )<br />
ν ⋅ N = ν +ν ⋅ N . (2)<br />
0m e md mj m<br />
Here<br />
tion, mi<br />
0i k0iN0 kmi Nm<br />
ν = – the frequency of direct ioniza-<br />
ν = — the frequency of step ionization<br />
from the metastable level, ν ad — the frequency of the<br />
diffuse departures of the electrons towards the tube<br />
walls, ν 0m = k0mN0 — the frequency of the metastable<br />
level’s excitation as a result of the atoms, being in the<br />
basic state, coincidence with the free electrons, ν md —<br />
the frequency of diffuse departure of metastable atoms<br />
to the walls of the tube, ν mj = kmj Ne—<br />
the frequency of<br />
the metastable atoms extinguishing due to their coincidence<br />
with free electrons, 0 , k i k mi — the constants<br />
of the direct and step ionization velocity, correspondingly,<br />
k0m and kmj = kmi + km0—<br />
the constants of the<br />
excitation and extinguishing velocities of the metastable<br />
state by electron impact, correspondingly. The<br />
main role in the process of states’ destruction is played<br />
by the metatstable state ionization with the constant<br />
k mi , as well as the transition of the excited atom to the<br />
k .<br />
ground state with the constant m0<br />
MODEL AND DISCUSSION<br />
It is known that the frequency of the diffuse departures<br />
in the cylindrical PC is being expressed by<br />
2<br />
d D ν = Λ , where D — the corresponding diffusion<br />
coefficient, 0 2.405 R Λ= — diffusion length,<br />
R 0 — the discharge tube capillary radius. For the<br />
charged particles, the ambipolar diffusion coefficient<br />
is: Dad =μ i kTee , where μ i — the ions’ mobility. According<br />
to [2], the ions’ mobility with the account of<br />
© L. V. Mykhaylovska, A. S. Mykhaylovska, 2009
the resonant recharging, is proportional to the ion’s<br />
velocity and inversely proportional to the gas pressure.<br />
That’s why, the ambipolar diffusion coefficient is:<br />
Dad = Da0⋅ TeT p,<br />
ànd the corresponding frequency<br />
equals to ( )( ) 2<br />
ν ad = Da0 TeT p 2.405 R0<br />
.<br />
For metastable atoms, the diffusion coefficient<br />
is proportional to the atoms’ velocity and inversely<br />
proportional to the density of gas. Thus, one obtains<br />
Dmd = Dm0⋅ T T p and then the corresponding of<br />
diffusive departures is ( )( ) 2<br />
ν md = Dm0 T T p 2.405 R0<br />
. The values of the constants Da0, D m0<br />
are defined by<br />
the type of gas and the system of units chosen [7].<br />
In the written above simplified balance equations<br />
the particles’ space distribution over the area of<br />
the capillary cross-section is not considered. It is assumed<br />
that 0 , N Nmcorrespond to the values of the<br />
particles’ concentration in the ground and metastable<br />
states at the center of the tube and N e — the concentration<br />
of the electrons at the center of the tube.<br />
It should be noted that according to the Mendeleev-<br />
Clapeiron law, the density of atoms N0 is mutually<br />
connected with gas temperature through the relation<br />
N0 = p kT = N00⋅ p T (ð –gas pressure in the discharge<br />
tube, T — gas temperature, and the constant N 00 value<br />
is determined by the system of units choice).<br />
According to [2], while the energy spectrum of<br />
electrons is of the Maxwellian type, the frequency of<br />
the direct ionization of the atom from the ground state<br />
and that for the atom transfer from the ground to the<br />
excited one the following expression could be used:<br />
v0i = veCi( eU0i + 2kTe) exp(<br />
−eU0i kTe) ⋅ N0.<br />
(3)<br />
Here ve = 8kTe<br />
π m — average thermal velocity<br />
of the electron, C i – the constant characteristic for<br />
the given process, U 0i — ionization potential for the<br />
atom transfer to the excited state.<br />
The frequency of the step ionization is determined<br />
by Thomson formula and could be written as:<br />
Cmi ⋅ve<br />
ν mi = ×<br />
kTe ⋅eU<br />
mi<br />
⎛ ⎞<br />
∞<br />
⎜ eU mi eU mi exp( −tdt<br />
) ⎟<br />
× ⎜exp( − ) − ⋅ Nm<br />
kTe kT ∫ ⎟⋅<br />
,<br />
⎜ e eU t<br />
mi ⎟<br />
⎝ kTe<br />
⎠<br />
(4)<br />
ãäå mi C — the constant for given process, earlier, for ad<br />
U mi — the<br />
excited atom ionization potential.<br />
It is seen that one could get the following expressions<br />
for the density of metastable atoms and electronic<br />
density from the balance equations (1-2) as a<br />
functions of electronic temperature, pressure and the<br />
temperature of the working gas and of the radius of the<br />
discharge tube:<br />
( νad −ν0i) νmd<br />
Ne( Te, p, T, R0)<br />
= , (5a)<br />
ν0mkmi −( νad −ν0i)<br />
kmj<br />
νad −ν0iν0mNe Nm( Te, p, T, R0)<br />
= = . (5b)<br />
kmi ν md + kmj Ne<br />
It is seen, also, that for Nm ≥ 0 and Ne ≥ 0 only<br />
if νad −ν0i≥ 0 and ν0mkmi −( νad −ν 0i) kmj<br />
> 0 . From<br />
these equations, with the use of the definitions made<br />
D and N 0 , the following boundary value<br />
could be proposed for the combination pR 0 :<br />
2<br />
Da0⋅Te ⋅T T ⎛ pR0 ⎞ Da0⋅Te ⋅T<br />
T<br />
< ⎜<br />
00 ( 0 0 ) 2.405<br />
⎟ ≤<br />
. (6)<br />
N ⋅ k N i + k mkmi kmj<br />
⎝ ⎠<br />
00 ⋅ k0i<br />
The two boundary equations for the electronic<br />
temperature follow the simplified balance equations<br />
(1) and (2). So, when the excited atoms are absent in<br />
the plasma of gaseous discharge, and only the non-excited<br />
atoms are being ionized the Schottky condition<br />
could be obtained from (1) or (5b) for the evaluation<br />
of the electronic temperature:<br />
ν ad ( Te) =ν 0i<br />
( Te)<br />
(7à)<br />
what means the equality of the frequencies of the ionization<br />
and of the diffuse departures of electrons to the<br />
walls of the tube. Inserting the expressions for the frequencies<br />
into this equality, one could obtain the following<br />
equation for the T e under the given pressure<br />
p and the gas temperature T in the tube of the radius<br />
R 0<br />
2<br />
TeN00 ⎛ pR0⎞<br />
1<br />
=<br />
k0i( Te) D<br />
⎜<br />
a0<br />
2.405<br />
⎟ . (7b)<br />
⎝ ⎠ T T<br />
During the increase of the electrons’ density caused<br />
by the increase of the discharge current the number of<br />
the metastable atoms increases as well and the mechanism<br />
of the step ionization starts to act. In the limit of<br />
Ne →∞ one could obtain from (5à) or (5b) the other<br />
following equation for the evaluation of e T<br />
kmj ( Te)<br />
ν 0m( Te) = ⎡ ad( Te) 0i(<br />
Te)<br />
kmi ( T<br />
⎣ν −ν ⎤⎦<br />
, (8à)<br />
e )<br />
which is the analog of Schottky condition with the account<br />
of the processes of creation and destruction of<br />
the metastable atoms. Using the expressions for the<br />
frequencies, we obtain the following equation for e T<br />
Tk e mj( Te)<br />
=<br />
k0i( Te) kmj( Te) + k0m( Te) kmi( Te)<br />
2<br />
N00 ⎛ pR0⎞<br />
1<br />
=<br />
D<br />
⎜<br />
a0<br />
2.405<br />
⎟<br />
. (8b)<br />
⎝ ⎠ T T<br />
CALCULATION RESULTS<br />
At Fig.1 are presented the solutions of the equations<br />
(7b) (curves 1) and of the equation (8b) (curves<br />
2), as well as the difference between these solutions<br />
(curves 3) in dependence of the product pR 0 . It is seen<br />
that the solutions of these equations behave almost<br />
uniformly in the chosen range of pR 0 . The difference<br />
between these solutions decreases with the increase of<br />
pR 0 . The increase of the working gas temperature T<br />
causes the somewhat increase of the corresponding<br />
electronic temperature along with the non-significant<br />
increase difference between solutions. The character<br />
of the dependencies is not changing.<br />
107
Te, eV<br />
15<br />
10<br />
5<br />
108<br />
1<br />
2<br />
3<br />
0<br />
0 0.2 0.4 0.6 0.8 1 1.2<br />
pRo, mmHg*cm<br />
4<br />
Te, eV<br />
3<br />
2<br />
1<br />
1 1.2 1.4 1.6 1.8<br />
pRo, mmHg*cm<br />
2<br />
Fig.1. The dependencies of the equations (7) and (8) solutions<br />
on the product pR 0 for two regions of pR0 changes. Curves<br />
1 — solutions of equation (7), curves 2 – solutions of equation (8),<br />
curves 3 – the difference of these solutions<br />
In two limiting cases which correspond to Ne → 0<br />
and Ne →∞, the equations (7) and (8) for the electronic<br />
temperatures do not include the electronic concentration.<br />
From other part, in general case, the electronic<br />
density depends on the total discharge current, which<br />
R0<br />
could be written in the form Ir = 2πeμeEz∫<br />
rNe() r dr ,<br />
0<br />
where e e( , ) pT<br />
μ =μ - the electronic mobility, z E<br />
- the longitudinal field strength, which stabilizes in<br />
the discharge. In the PC of the discharge, the electronic<br />
distribution over the section area is Besselian<br />
Ne( r) = Ne0⋅ J0(2.405 r R0)<br />
. Then, under the given<br />
discharge current, it is possible to obtain the following<br />
N<br />
expression for e0<br />
( , , , , )<br />
1<br />
2<br />
3<br />
Ir<br />
2<br />
⎛2.405 ⎞<br />
πμ e eEz ⎝ R0<br />
⎠<br />
Ne0 Te p R0 T Ir<br />
= ⎜ ⎟ . (9)<br />
5<br />
This value of electronic density should coincide<br />
with the value obtained from the balance equations<br />
(5à). That’s why it is necessary to solve the following<br />
equation taking into account the both the direct and<br />
step ionization<br />
( 0 )<br />
( )<br />
2<br />
νad −ν i νmd I ⎛ r 2.405 ⎞<br />
= ⎜ ⎟<br />
ν0mkmi − νad −ν0i kmj 5πeμeEz<br />
⎝ R0<br />
⎠<br />
. (10)<br />
The strength of the longitudinal electric field held<br />
in the discharge Ez = Ez( Te, p, R0, T)<br />
is determined by<br />
the balance equation for the electronic energy with the<br />
account of the elastic energy losses during the impact<br />
energy exchanges between electrons and atoms of the<br />
working gas as well as of the non-elastic losses on the<br />
excitation and ionization of the atoms. Besides, the<br />
energy balance in thin gas-discharge tubes includes<br />
the energy transferred by the charged particles to the<br />
tube walls [4]. In present paper the longitudinal field<br />
strength Ez is calculated through the following equation<br />
obtained in [7]:<br />
∑<br />
eE = mν<br />
⋅( ν ⋅Δε +ν ⋅Δε + ν ⋅Δε ) . (11)<br />
2 2<br />
z m c g w w i i<br />
i<br />
Here ν m — the effective electron collision frequency,<br />
ν c — elastic collisions frequency for electronneutral<br />
atom pairs, which in general case depends<br />
on the electronic temperature and atoms’ density,<br />
Δε g = (2 mM) ⋅ (3kTe 2) — energy losses during elastic<br />
collisions of electrons with neutral atoms, νw− electron collisions frequency with discharge tube walls<br />
which equals to the frequency of electronic diffuse<br />
departures to the walls ν w =ν ad , Δεw − energy losses<br />
of the charged particles on the tube walls, ν i — frequency<br />
of non-elastic collisions of the electrons with<br />
atoms which excites or ionizes the atom, Δε i — electron<br />
energy losses given to the excitation or ionization<br />
of the atoms.<br />
We would like to limit ourselves to the case of account<br />
only the electron energy losses on the tube walls<br />
when the power contains only the recombination<br />
energy for positive ions and electrons eU and of ki-<br />
i<br />
netic energy they possessed before the recombination<br />
Δε w = eU0i + Wkp + 2kTe,<br />
where U 0i is the ionization<br />
potential of the atom in the recombination state, 2kT e<br />
and W kp -are the average kinetic energies of the electron<br />
and ion during their approach the wall.<br />
Taking into account the processes of direct and<br />
step ionization, one could rewrite the last term in<br />
brackets in formula (11) for the power density in the<br />
form: ∑ ν⋅Δε= i i e( k0iN0Ui + k0mN0Um + kmiNmUmi) .<br />
Using the i expression (6) for the metastable<br />
atoms’ density N m , it is evident that:<br />
∑ ν⋅Δε= i i e⎡⎣( k0i + k0m) N0Um +νadUmi⎤⎦.<br />
i<br />
The losses on the direct iomization and atoms’<br />
excitation into metastable state without the step<br />
ionization process are described by the expression:<br />
∑ ν⋅Δε= i i e( k0iUi + k0mUm) N0.<br />
i It is possible to make it clear that:<br />
k + k N U +ν U ≥ k U + k U N , if<br />
( ) ( )<br />
0i 0m 0 m ad mi 0i i 0m m 0<br />
νad ≥ν 0i<br />
.<br />
The results of the numeric solution of the equation<br />
(10) for the discharge in He are presented at<br />
Fig.2 — 4. The partial dependencies on the gas<br />
pressure and capillary radius are presented as they<br />
in contrary to the solutions of equations (7) and
(8), from the equation (10) the solution’s dependence<br />
on the product pR 0 , and Te≠f( pR0)<br />
doesn’t<br />
follow. We have used the following constants:<br />
−17<br />
Ci<br />
= 0.12⋅10 cm2 −17<br />
/eV, Cm<br />
= 0.45⋅10 cm2 /eV,<br />
−14<br />
Cmi<br />
= 6.5⋅10 cm2 /eV, eU 0i = 24.6 eV, eU 0m = 20 eV,<br />
eU mi = 4.6 eV. Besides, it is assumed that the gas temperature<br />
equals to that of the outer walls’ one. For<br />
this latter parameter we have used the experimentally<br />
0 0<br />
0<br />
measured value of TC= 20 C+ 7 ⋅ Ir,<br />
where 20 C is<br />
the temperature of the ambient, I r — the discharge<br />
current in mA.<br />
The numeric calculations have shown that the<br />
electronic temperature in the real situation with the<br />
account of both the direct and step ionization takes<br />
the value which doesn’t coincide with any limit value<br />
determined by equations (7) and (8). At the figures<br />
presented, it is seen that the electronic temperature’s<br />
value obtained as the solution of the equation (10), is<br />
smaller than the temperature obtained from the equation<br />
(7), but the greater than that obtained from equation<br />
(8). Only at small pressures and not significant<br />
discharge currents and in thin tubes, the real electronic<br />
temperature is almost the same as the electronic<br />
temperature obtained from Schottky’s condition (7)<br />
(see, please, the curves at Fig.2à). With the increase of<br />
the pressure, the step ionization plays the greater role<br />
and the solution of the equation (10), even for thin<br />
tubes, differs slightly from the solution of the modified<br />
Schottky equation (8) (see, please, the curves at<br />
Figs.2, 3).<br />
In the present paper, in the framework of the assumptions<br />
taken, the electronic temperature depends<br />
only on the temperature of the ambient and is proportional<br />
to the discharge current value. The proportionality<br />
quotient is a constant determined from the<br />
experiment. Really, according to the results of [7],<br />
the temperature of the working gas is the function<br />
of the discharge inner parameters such as: the electronic<br />
temperature, longitudinal electric field strength<br />
and electrons’ concentration. As it is seen from Fig.4<br />
(curves 1), with the increase of discharge current the<br />
metastable atoms’ concentration increases as well and<br />
the electronic temperature decreases as a result. But,<br />
afterwards, under the following growth of the discharge<br />
current and of the gas temperature, the gas density decreases<br />
due to the termal gas expansion to the ballast<br />
volume. This effect is accompanied by the increase of<br />
the electronic temperature. The placement of the discharge<br />
tube into the thermostat causes the decrease of<br />
the electronic temperature which could be explained<br />
by the growing role of the step ionization due to the<br />
gas pressure increase (see, please, Fig.4, curves 2).<br />
Thus, under the taken assumptions, the working gas<br />
temperature T remains stable and the solutions of the<br />
equations (7b) and (8b) do not dependent on the discharge<br />
current as it is seen at Fig.4 (horizontal lines).<br />
Fig. 2. The results of the numeric solution of the equation<br />
(10). The solid lines represent the dependencies of the electronic<br />
temperature during the gas pressure changes. Punctured and<br />
pin-stripe ones correspond to the solutions of the equations (7)<br />
and (8). The discharge current Ir= 10mA<br />
, gas temperature<br />
0<br />
T = 20 C+ 7 ⋅ Ir,<br />
à) discharge capillary radius R0= 0.05cm,<br />
b) discharge capillary radius R0= 0.5cm<br />
Fig. 3. The results of the numeric solution of the equation<br />
(10). The solid line represents the dependence of the electronic<br />
on the capillary radius. The punctured and pin-stripe ones correspond<br />
to the solution of the equations (7) è (8). Discharge current<br />
Ir= 10mA<br />
, gas pressure p = 10Torr<br />
, gas temperature<br />
0<br />
T = 20 C+ 7 ⋅<br />
Ir<br />
b<br />
�<br />
109
Fig. 4. The results of the numeric solution of the equation<br />
(10). The solid lines – dependencies of the electronic temperature<br />
on the discharge current. Gas pressure p = 10Torr<br />
, capillary<br />
0<br />
radius R0= 0.05cm,<br />
1 – gas temperature T = 20 C+ 7 ⋅ Ir,<br />
0<br />
2 – gas temperature T = 20 C .<br />
110<br />
CONCLUSIONS<br />
As a result, we would like to make the following<br />
conclusions:<br />
1. The performed theoretical analysis has shown<br />
that it is impossible to calculate the electronic temperature<br />
in the positive column of the gaseous discharge<br />
using only the Schottky equation (7) which takes into<br />
account the processes of direct ionization (limiting<br />
case of small discharge currents Ir → 0 ) or the equation<br />
(8) which operates only the processes of step ion-<br />
UDC 633.9<br />
L.V. Mikhaylovskaya, A.S. Mykhaylovska<br />
ization (limiting case of extremely high discharge currents<br />
Ir →∞). In the intermediate<br />
2. In the intermediate case of the medium values<br />
of discharge currents, the electronic temperature is determined<br />
by the equation (10) obtained in the present<br />
paper, which accounts both direct and step ionization.<br />
3. As a result of this equation solution, the dependencies<br />
of electronic temperature on gas pressure, discharge<br />
current value as well as the capillary radius value.<br />
4. It is shown that in the real range of discharge<br />
parameters changes such as gas pressure, discharge<br />
current and capillary radius, for the physical processes<br />
in positive column plasma adequate description it is<br />
necessary to take into account all possible ways of atoms<br />
excitation and ionization.<br />
REFERENCES<br />
1. Ýíöèêëîïåäèÿ íèçêîòåìïåðàòóðíîé ïëàçìû. Ò.ÕI-4. Ãàçîâûå<br />
è ïëàçìåííûå ëàçåðû. (Ïîä ðåä. Â.Å.Ôîðòîâà) Ì.:<br />
Ôèçìàòëèò. 2005 ã.396 ñòð.<br />
2. Ðàéçåð Þ.Ï. Ôèçèêà ãàçîâîãî ðàçðÿäà. — Ì.: Íàóêà,<br />
1997. — 592ñ.<br />
3. Ñìèðíîâ Á.Ì. Ôèçèêà ñëàáîèîíèçîâàííîãî ãàçà (â çàäà-<br />
÷àõ ñ ðåøåíèÿìè). — Ì.: Íàóêà, 1999. — 423ñ.<br />
4. Ìèëåíèí Â.Ì., Òèìîôååâ Í.À. Ïëàçìà ãàçîðàçðÿäíûõ<br />
èñòî÷íèêîâ ñâåòà íèçêîãî äàâëåíèÿ. — Ë.:Ýíåðãîèçäàò,<br />
1999. — 240 ñ.<br />
5. Ãðàíîâñêèé Â.Ë. Ýëåêòðè÷åñêèé òîê â ãàçå– Ì.: Íàóêà,<br />
1994. — 544ñ.<br />
6. Å. A. Áîãäàíîâ, À. À. Êóäðÿâöåâ, Ë. Ä. Öåíäèí, Ð. Ð. Àðñëàíáåêîâ,<br />
Â. È. Êîëîáîâ, Â. Â. Êóäðÿâöåâ. Âëèÿíèå ìåòàñòàáèëüíûõ<br />
àòîìîâ è íåëîêàëüíîãî ýëåêòðîííîãî ðàñïðåäåëåíèÿ<br />
íà õàðàêòåðèñòèêè ïîëîæèòåëüíîãî ñòîëáà â<br />
àðãîíå.//ÆÒÔ. — 2004. — Ò.74, âûï.6. — Ñ.35-42.<br />
7. Mikhailovskaya L.V. Energy balance and gas temperature in<br />
plasma of positive column in gas discharge narrow tubes. //<br />
Proc. SPIE. — 1999. — V. 3686. — P. 62-69.<br />
INFLUENCE OF THE STEP IONIZATION PROCESSES ON THE ELECTRONIC TEMPERATURE IN THIN<br />
GAS-DISCHARGE TUBES<br />
Abstract<br />
The theoretical analysis of dependence of the steady-stated electron temperature on the external parameters of discharge, such,<br />
as a value of discharge current, pressure of working gas and size of radius of discharge capillary, has been done taking into account the<br />
simultaneous processes of direct and stepped ionization. This analysis has been done on the basis of solution of the closed system of the<br />
balance equations. Conditions at which a determining role in formation of the charged particles plays stepped ionization are found.<br />
Key word: step ionization, processes, gas discharge tubes.<br />
ÓÄÊ 633.9<br />
Ë. Â. Ìèõàéëîâñêàÿ, À. Ñ. Ìèõàéëîâñêàÿ<br />
ÂËÈßÍÈÅ ÏÐÎÖÅÑÑΠÑÒÓÏÅÍ×ÀÒÎÉ ÈÎÍÈÇÀÖÈÈ ÍÀ ÝËÅÊÒÐÎÍÍÓÞ ÒÅÌÏÅÐÀÒÓÐÓ Â ÓÇÊÈÕ<br />
ÃÀÇÎÐÀÇÐßÄÍÛÕ ÒÐÓÁÊÀÕ<br />
Ðåçþìå<br />
Ñ ó÷åòîì ïðîöåññîâ ïðÿìîé è ñòóïåí÷àòîé èîíèçàöèè íà îñíîâå ðåøåíèÿ çàìêíóòîé ñèñòåìû áàëàíñíûõ óðàâíåíèé<br />
ïðîâåäåí òåîðåòè÷åñêèé àíàëèç çàâèñèìîñòè óñòàíîâèâøåéñÿ òåìïåðàòóðû ýëåêòðîíîâ îò âíåøíèõ ïàðàìåòðîâ ðàçðÿäà, òàêèõ,<br />
êàê âåëè÷èíà òîêà ðàçðÿäà, äàâëåíèå ðàáî÷åãî ãàçà è ðàçìåðà ðàäèóñà ðàçðÿäíîãî êàïèëëÿðà. Íàéäåíû óñëîâèÿ, ïðè<br />
êîòîðûõ îïðåäåëÿþùóþ ðîëü â îáðàçîâàíèè çàðÿæåííûõ ÷àñòèö èãðàåò ñòóïåí÷àòàÿ èîíèçàöèÿ.<br />
Êëþ÷åâûå ñëîâà: ñòóïåí÷àòàÿ èîíèçàöèÿ, ïðîöåññ, ãàçîðàçðÿäíàÿ òðóáêà.
ÓÄÊ 633.9<br />
Ë. Â. Ìèõàéëîâñüêà, À. Ñ. Ìèõàéëîâñüêà<br />
ÂÏËÈ ÏÐÎÖÅѲ ÑÒÓϲÍ×ÀÑÒί ²ÎͲÇÀÖ²¯ ÍÀ ÅËÅÊÒÐÎÍÍÓ ÒÅÌÏÅÐÀÒÓÐÓ Â ÂÓÇÜÊÈÕ<br />
ÃÀÇÎÐÎÇÐßÄÍÈÕ ÒÐÓÁÊÀÕ<br />
Ðåçþìå<br />
Ç óðàõóâàííÿì ïðîöåñ³â ïðÿìî¿ òà ñòóï³í÷àñòî¿ ³îí³çàö³¿ íà ï³äñòàâ³ ð³øåííÿ çàìêíóòî¿ ñèñòåìè áàëàíñíèõ ð³âíÿíü ïðîâåäåíî<br />
òåîðåòè÷íèé àíàë³ç çàëåæíîñò³ óñòàëåíî¿ òåìïåðàòóðè åëåêòðîí³â â³ä çîâí³øí³õ ïàðàìåòð³â ðîçðÿäó, òàêèõ, ÿê âåëè÷èíà<br />
ðîçðÿäíîãî ñòðóìó, òèñê ðîáî÷îãî ãàçó òà ðîçì³ð ðàä³óñà ðîçðÿäíîãî êàï³ëÿðó. Çíàéäåí³ óìîâè, ïðè ÿêèõ âèçíà÷àëüíó ðîëü â<br />
óòâîðåíí³ çàðÿäæåíèõ ÷àñòèíîê ãðຠñòóï³í÷àñòà ³îí³çàö³ÿ.<br />
Êëþ÷îâ³ ñëîâà: ñòóï³í÷àñòà ³îí³çàö³ÿ, ïðîöåñ, ãàçîðàçðÿäíà òðóáêà.<br />
111
112<br />
UDÑ 539.184<br />
E. V. MISCHENKO<br />
Odessa National Polytechnical University, Odessa<br />
QUANTUM MEASURE OF FREQUENCY AND SENSING THE COLLISIONAL<br />
SHIFT AND BROADENING OF Rb HYPERFINE LINES IN MEDIUM<br />
OF HELIUM GAS<br />
New theoretical data regarding the collisional shift and broadening of the hyperfine spectral lines<br />
for Rb in the atmosphere of the buffer inert He gas within relativistic perturbation theory formalism<br />
are presented.<br />
1. INTRODUCTION<br />
Studying the collisional shifts and broadening of<br />
the hyperfine structure lines for heavy elements (alkali,<br />
alkali-earth, lanthanides, actinides and others) in<br />
an atmosphere of inert gases is one of the important<br />
and actual topics of the modern atomic and molecular<br />
optics, quantum and photo-electronics etc. [1-8].<br />
It should be mentioned also that the heavy atoms are<br />
interesting from the point of view of studying a role of<br />
weak interactions in atomic optics and heavy-elements<br />
chemistry. Besides, calculation of the hyperfine structure<br />
line shift and broadening allows one to check the<br />
quality of wave functions and study a contribution of<br />
relativistic effects in two-center (multi-center) atomic<br />
systems . From the applied science point of view, the<br />
mentioned physical effects form a basis for creating<br />
an atomic quantum measure of frequency [2,8]. For<br />
a long time the corresponding phenomenon for thallium<br />
atom attracted a special attention because of<br />
possibility to create the thallium quantum frequency<br />
measure. Alexandrov and co-workers [8] have realized<br />
the optical pumping of the thallium atoms on the line<br />
of 21GHz, which corresponds to transition between<br />
the components of hyperfine structure for the ground<br />
state, and have measured the collisional shift of this<br />
line due to buffer (bath) gas. Naturally, the inert buffer<br />
gases (He, Ar etc.) were used. The detailed non-relativistic<br />
theory of the collisional shift and broadening of<br />
the hyperfine structure lines for simple elements (light<br />
alkali elements etc.) has been developed by many authors<br />
(see discussions in refs. [1-8]). However, consideration<br />
of heavy elements faces serious difficulties<br />
related with account for the relativistic and correlation<br />
corrections. It is very curious that until now a consistent,<br />
accurate quantum mechanical approach for<br />
calculating main characteristics of the collisional processes<br />
was not developed though many different simplified<br />
models have been proposed (see, for example<br />
[6-13]). The most widespread approach is based on the<br />
calculation of the corresponding collision cross-section,<br />
in particular, in a case of the van der Waals interaction<br />
between colliding particles [2]. However, such<br />
an approach does not factually define any difference<br />
between the Penning process and resonant collisional<br />
one and gives often non-correct results for cross-sections.<br />
More consistent method requires data on the<br />
process probability (cross-section, collision strength)<br />
G(R) as a function of internuclear distance. It should<br />
be noted that these data for many physically interesting<br />
tasks are practically absent at present time. In ref.<br />
[6,13] a new relativistic optimized approach, based on<br />
the gauge-invariant perturbation theory with using the<br />
optimized wave functions basis’s, is developed in order<br />
applied to calculating the inter atomic potentials,<br />
hyper fine structure collision shift for heavy atoms in<br />
an atmosphere of inert gases. In this paper we present<br />
some new theoretical data regarding the collision shift<br />
and broadening the hyperfine spectral lines for Rb<br />
atom in an atmosphere of the inert gas He.<br />
2. THE SPECTRAL BROADENING OF<br />
THE ATOMIC HYPERFINE LINES IN THE<br />
BUFFER GAS<br />
To calculate the collision shift of hyperfine structure<br />
spectral lines one could use the following expression<br />
known from kinetic theory of spectral line form<br />
(see, for example, [2,5]):<br />
( ) ( ( ) ) 2<br />
0<br />
∞<br />
D 4πw<br />
fp= = w R exp U R kT R dR<br />
p kT ∫ δ −<br />
(1)<br />
0<br />
where U(R) is the effective potential of the interatomic<br />
interaction, which has a central symmetry in a case of<br />
the systems A—B (in our case, for example, B=He;<br />
A=Rb); T is a temperature, w is a frequency of the<br />
0<br />
hyperfine structure transition in the isolated active<br />
atom; Δω(R)=Dw(R)/w is the relative local shift of<br />
0<br />
the hyperfine structure lines, which is arisen due to the<br />
disposition of the active atoms (say, atom of Rb and<br />
helium He) on a distance R. To calculate an effective<br />
potential of the interatomic interaction we use a method<br />
of the exchange perturbation theory (the modified<br />
version ÅL-ÍÀV) [1,5]. Within exactness to second<br />
order terms on potential of Coulomb interaction of<br />
the valent electrons and atomic cores one can write:<br />
S0 C ⎛ 6 2 1 ⎞<br />
δω ( R)<br />
= + Ω 1+ Ω2 − ,<br />
6 ⎜ + ⎟ (2)<br />
1− S0 R ⎜Ea Ea + E ⎟<br />
⎝ B ⎠<br />
Here Ñ is the van der Waals constant for interac-<br />
6<br />
tion À-Â (e.g., a pair of Rb-Íå); I, Å are the ioniza-<br />
1a,b<br />
tion potential and excitation energy on the first level<br />
for atoms A, B correspondingly; S is the overlapping<br />
0<br />
© E. V. Mischenko, 2009
integral; The value of E α ,b can be simply defined as<br />
follows:<br />
( )<br />
Eα , b = Ia, b+ E1<br />
a, b 2,<br />
The values Ω , Ω in the expression (2) are the<br />
1 2<br />
non-exchange and exchange non-perturbation sums<br />
of the first order correspondingly, which are defined<br />
as follows:<br />
( 1 )<br />
() 1 () 1<br />
2 〈Φ H Φ 〉 V<br />
Ω= 1<br />
N −S ρ E −E<br />
( 1 )<br />
' ' '<br />
'<br />
∑<br />
0 CT k k 0<br />
(3)<br />
0 0 k 0 k<br />
() 1 () 1<br />
2 〈Φ H Φ 〉 U<br />
'<br />
Ω 2=<br />
∑<br />
N −S ρ E −E<br />
' ' '<br />
0 CT k k 0<br />
0 0 k 0 k<br />
() 1 () 1 / () 1 () 1<br />
с =
method and determination of Cs atomic properties// Phys.<br />
Rev. A. — 2005. — Vol.71. — P.032509.<br />
15. Ivanov L.N.,Ivanova E.P. Extrapolation of atomic ion energies<br />
by model potential method: Na-like spectra // Atom.<br />
Data Nucl Tabl. — 1999-Vol.24,N2. — p.95-101.<br />
16. Bekov G.I., Vidolova-Angelova E., Ivanov L.N., Letokhov<br />
V.S., Mishin V.I., Laser spectroscopy of narrow two-timely<br />
excited autoionization states for ytterbium atom// JETP. —<br />
1991. — Vol.80,N3. — P.866-878.<br />
17. Basar Gu., Basar Go., Acar G., Ozturk I.K., Kroger S., Hyperfine<br />
structure investigations of MnI: Experimental and<br />
theoretical studies of the hyperfine structure in the even configurations//<br />
Phys.Scripta. — 2003. — Vol.67. — P.476-484.<br />
18. Dorofeev D.L., Zon B.A., Kretinin I.Y., Chernov V.E.,<br />
Method of the Quantum Defect Green’s Function for Calculation<br />
of Dynamic Atomic Polarizabilities// Optics and<br />
Spectroscopy. — 2005. — Vol. 99, N4. — P.540-548.<br />
19. Mischenko E.V., Transition energies and oscillator strengths<br />
in helium within equation of motion approach with density<br />
functional method for effective account of correlation’s//<br />
Photoelectronics. — 2006 . — N15. — P.58-60.<br />
114<br />
UDÑ 539.184<br />
E. V. Mischenko<br />
20. The Fundamentals of Electron Density, Density Matrix and<br />
Density Functional Theory in Atoms, Molecules and the<br />
Solid State. Series: Progress in Theoretical Chemistry and<br />
Physics. Eds. Gidopoulos N.I. and Wilson S. — Springer,<br />
2004. — Vol.14. — 244P.<br />
21. Kolachevsky N.N., Precise laser spectroscopy of cold atoms<br />
and search of drift of the fine structure constant//Physics Uspekhi.<br />
— 2008. — Vol.178. — P.1225-1230.<br />
22. Courade E., Anderlini M., Ciampini D. Etal, Two-photon<br />
ionization of cold rubidiun atoms with near resonant intermediate<br />
state//J.Phys.B. At.Mol.Opt.Phys. — 2004. —<br />
Vol.37. — P.967-979.<br />
23. Singer K., Reetz-Lamour M., Amthor T., Marcassa L.G.,<br />
Weidemuller M., Spectral broadening and suppression of<br />
excitation induced by ultralong-range interactions in a cold<br />
gas of Rydberg atoms //Phys. Rev. Lett. — 2004. — Vol.93. —<br />
P.163001.<br />
24. Loboda A.V., Mischenko E.V. et al, Spectral Broadening of<br />
excitation induced by ultralong-range interaction in a cold gas<br />
of Rydberg atoms// Spectral Line Shapes (AIP). — 2008. —<br />
Vol.15. — P.260-264.<br />
QUANTUM MEASURE OF FREQUENCY AND SENSING THE COLLISIONAL SHIFT AND BROADENING OF RB<br />
HYPERFINE LINES IN MEDIUM OF HELIUM GAS<br />
Abstract<br />
New theoretical data regarding the collisional shift and broadening of the hyperfine spectral lines for Rb in the atmosphere of the<br />
buffer inert He gas within relativistic perturbation theory formalism are presented.<br />
Key words: atoms of Rb, spectral shift, inert gas He.<br />
ÓÄÊ 539.184<br />
E. Â. Ìèùåíêî<br />
ÊÂÀÍÒÎÂÀß ÌÅÐÀ ×ÀÑÒÎÒÛ È ÄÅÒÅÊÒÈÐÎÂÀÍÈÅ ÑÒÎËÊÍÎÂÈÒÅËÜÍÎÃÎ ÑÄÂÈÃÀ È ÓØÈÐÅÍÈß ËÈÍÈÉ<br />
ÑÂÅÐÕÒÎÍÊÎÉ ÑÒÐÓÊÒÓÐÛ RB Â ÀÒÌÎÑÔÅÐÅ ÃÅËÈß È ÕÎËÎÄÍÎÌ ÃÀÇÅ ÀÒÎÌÎÂ RB<br />
Ðåçþìå<br />
Ïðèâåäåíû íîâûå òåîðåòè÷åñêèå äàííûå ïî ñòîëêíîâèòåëüíîìó ñäâèãó è óøèðåíèþ ñïåêòðàëüíûõ ëèíèé ñâåðõòîíêîé<br />
ñòðóêòóðû àòîìîâ Rb â àòìîñôåðå èíåðòíîãî ãàçà Íå, ïîëó÷åííûå â ðàìêàõ ôîðìàëèçìà ðåëÿòèâèñòñêîé òåîðèè âîçìóùåíèé.<br />
Êëþ÷åâûå ñëîâà: àòîìû Rb, ñïåêòðàëüíûé ñäâèã, èíåðòíûé ãàç Íå.<br />
ÓÄÊ 539.184<br />
Î. Â. ̳ùåíêî<br />
ÊÂÀÍÒÎÂÀ ̲ÐÀ ×ÀÑÒÎÒÈ ² ÄÅÒÅÊÒÓÂÀÍÍß ÇÑÓÂÓ ÒÀ ÓØÈÐÅÍÍß Ë²Í²É ÍÀÄÒÎÍÊί ÑÒÐÓÊÒÓÐÈ RB ÇÀ<br />
ÐÀÕÓÍÎÊ Ç²ÒÊÍÅÍÜ Â ÀÒÌÎÑÔÅв ÃÅ˲ß<br />
Ðåçþìå<br />
Íàâåäåí³ íîâ³ òåîðåòè÷í³ äàí³ ïî çñóâó òà óøèðåííþ ñïåêòðàëüíèõ ë³í³é ïîíàäòîíêî¿ ñòðóêòóðè àòîì³â Rb çà ðàõóíîê<br />
ç³òêíåíü â àòìîñôåð³ ³íåðòíîãî ãàçó Íå, îòðèìàí³ ó ìåæàõ ôîðìàë³çìó ðåëÿòèâ³ñòñüêî¿ òåî𳿠çáóðåíü.<br />
Êëþ÷îâ³ ñëîâà: àòîìè Rb , ñïåêòðàëüíèé çñóâ, ³íåðòíèé ãàç Íå.
UDC 621.315.592<br />
O. O. PTASHCHENKO 1 , F. O. PTASHCHENKO 2 , N. V. MASLEYEVA 1 , O. V. BOGDAN 1<br />
1 I. I. Mechnikov National University of Odessa, Dvoryanska St., 2, Odessa, 65026, Ukraine<br />
2 Odessa National Maritime Academy, Odessa, Didrikhsona St., 8, Odessa, 65029, Ukraine<br />
SURFACE CURRENT IN GaAs P-N JUNCTIONS, PASSIVATED BY<br />
SULPHUR ATOMS<br />
1. INTRODUCTION<br />
P-n junctions on wide-band III–V semiconductors<br />
can be used as gas sensors [1, 2]. Such sensors<br />
have low background currents, are sensitive at room<br />
temperature, have a response time of 100 s [3]. The gas<br />
sensitivity of these sensors is due to forming of a surface<br />
conducting channel in the electric field induced<br />
by the ions adsorbed on the surface of the natural oxide<br />
layer [1, 2].<br />
The threshold gas partial pressure of a sensor on<br />
p-n junction depends on the surface states density in<br />
the semiconductor [3]. The results of calculations [3]<br />
predict rise of the sensitivity to low concentrations of<br />
a donor gas when the surface states density in the p-n<br />
junction is diminished.<br />
The surface states density in GaAs can be lowered<br />
by sulphur atoms deposition from some solutions [4].<br />
The sulphur-passivation reduces the excess forward<br />
current and reverse current in GaAs p-n junctions,<br />
enhances the photosensitivity in the spectral region<br />
of strong absorption, substantially increases the sensitivity<br />
to ammonia vapors [5]. However the stability of<br />
the characteristics of sulphur-passivated p-n junctions<br />
was not investigated.<br />
The aim of this work is a study of the influence of<br />
the storage in a neutral gas on the surface currents in<br />
sulphur-passivated GaAs p-n junctions. Effect of the<br />
storage in helium atmosphere at room temperature on<br />
I-V characteristics of forward and reverse currents in<br />
sulphur-passivated GaAs p-n structures was studied.<br />
2. EXPERIMENT<br />
Influence of the storage (low-temperature annealing) of sulphur-passivated GaAs p-n structures<br />
in a neutral (helium) atmosphere at room temperature on I-V characteristics of forward and reverse<br />
currents was studied. The storage strongly reduces the excess forward current and the reverse current<br />
in p-n junctions. The ideality coefficient of I-V characteristics decreases with the storage. This effect<br />
has two stages. It is showed that all these phenomena can be explained by lowering of the surface<br />
recombination centers density and reduction of the electrically active centers concentration in the<br />
surface depletion layer.<br />
I-V characteristics were measured on GaAs pn<br />
junctions with the structure described in previous<br />
works [1, 2]. The sulphur atoms deposition (passivation)<br />
was carried out by a treatment of different durations<br />
in 30% water solutions of Na 2 S . H 2 O [5].<br />
I-V characteristics of the forward current in a typical<br />
p-n structure are presented in Fig. 1.<br />
Curve 1 was measured before the treatment. Over<br />
the current range between 1μA and 1mA the I–V curve<br />
can be described with the expression<br />
IV ( ) = I0exp( qV/ nkT t ) , (1)<br />
where I is a constant; q is the electron charge; V de-<br />
0<br />
notes bias voltage; k is the Boltzmann constant; T is<br />
temperature; nt ≈ 2 is the ideality constant. Such I-V<br />
curves can be ascribed to recombination on deep levels<br />
in p-n junction and (or) at the surface [6]. And the<br />
corresponding current is known as a recombination<br />
current.<br />
10 -5<br />
© O. O. Ptashchenko, F. O. Ptashchenko, N. V. Masleyeva, O. V. Bogdan, 2009<br />
10 -6<br />
10 -7<br />
10 -8<br />
I,A<br />
– 1<br />
– 2<br />
– 3<br />
–4<br />
– 5<br />
– 6<br />
– 7<br />
0 0,2 0,4 0,6 0,8<br />
V, Volts<br />
Fig. 1. I–V characteristics of the forward current of a p-n<br />
structure: initial (1) and after S-treatment and subsequent storage<br />
in helium: 2 – 2.4 . 103 s; 3 – 7.2 . 103 s; 4 – 1.7 . 105 s; 5 – 5.2 . 105 s;<br />
6 – 2.6 . 106 s; 7 – 7.8 . 106 s<br />
At lower biases curve 1 has a section of an excess<br />
current, which has an ideality constant n >2 and cor-<br />
t<br />
responds to the phonon-assisted tunnel recombination<br />
at deep centers [6]. This recombination is located<br />
at the p-n junction non-homogeneities, which cause<br />
local increase of the electric field [6].<br />
Curves 2 to 7 in Fig. 1 were obtained after passivation<br />
of 40 s duration. Curve 2, measured after subsequent<br />
40 min storage, exhibits an increase of the<br />
115
excess current. And the further storage leads to a substantial<br />
decrease of the excess current, as illustrated by<br />
curves 3 to 7. After the storage during one month the<br />
excess current disappears, and the I–V characteristic<br />
corresponds to (1) over the current range from 10-8 A<br />
≈ 2 .<br />
to 10 -3 A with ideality constant n t<br />
116<br />
10 -5<br />
10 -6<br />
10 -7<br />
10 -8<br />
I, �<br />
10 3<br />
2<br />
1<br />
10 4<br />
10 5<br />
3<br />
3A<br />
10<br />
t, s<br />
6 10 7<br />
Fig. 2. Effect of the storage duration on the forward current at<br />
V=0.4 V (1) and at V=0.7 V (2) and on the reverse current at V=-4<br />
V (3, 3A). Curve 3A is shifted down by 0.7<br />
Curve 1 in Fig. 2 represents the dependence of the<br />
excess current (at the voltage of 0.4 V) on the storage<br />
duration. Curve 2 illustrates such dependence for<br />
the recombination current (at the voltage of 0.7 V). A<br />
comparison between curves 1 and 2 shows that the influence<br />
of passivation and the subsequent storage on<br />
the recombination current is much weaker than on the<br />
excess current.<br />
The effect of the passivation and the subsequent<br />
storage on the I–V characteristic of reverse current in<br />
the same p-n junction is illustrated by Fig. 3. Curve<br />
1 was obtained before passivation. Curve 2, measured<br />
after subsequent storage during 40 min, exhibits an<br />
increase of the reverse current. And further storage<br />
decreases the reverse current, as is seen from curves 3<br />
to 6. Curve 3 in Fig. 2 presents the dependence of the<br />
reverse current, measured at a voltage of –4 V, on the<br />
storage duration. Curve 3A is obtained from curve 3 by<br />
a shift down by 0.7. This curve practically coincides<br />
with curve 1. It means that the time-dependences of<br />
the excess forward current and the reverse current are<br />
identical.<br />
3. DISCUSSION<br />
The presented experimental results show that<br />
storage in helium atmosphere at room temperature<br />
substantial reduces recombination and excess forward<br />
currents, as well as reverse current in sulphur-passivated<br />
GaAs p-n junctions. The storage time dependence<br />
of the excess forward current and the reverse current<br />
are identical. It suggests that decrease of these currents<br />
has the same nature. It is known [6] that the excess<br />
current in GaAs p-n structures is localized in nonhomogeneities<br />
of the p-n unction. A strong influence<br />
of the passivation and the subsequent storage on the<br />
excess current suggests that these non-homogeneities<br />
are placed on the surface of our structures. And the<br />
same can be concluded about the localization of the<br />
reverse current in our samples.<br />
10 -5<br />
10 -6<br />
10 -7<br />
10 -8<br />
I,A<br />
1<br />
2<br />
3<br />
0 1 2 3 4 5<br />
V, Volts<br />
Fig. 3. I–V characteristics of the reverse current of a p-n<br />
structure: initial (1) and after S-treatment and subsequent storage<br />
in helium: 2 – 2.4 . 103 s; 3 – 1.7 . 105 s; 4 – 1.2 . 106 s; 5 – 2.6 . 106 s;<br />
6 – 7.8 . 106 s<br />
Curve 2 in Fig. 2 illustrates the time dependence of<br />
the recombination current during the storage. The surface<br />
component of this current can be expressed [6] as<br />
1 1 1<br />
s p s s<br />
2 2<br />
ns<br />
2<br />
ps<br />
I = ql n(0)( C N / 2) ( D + D ) . (2)<br />
Here l is the perimeter of the p-n junction; n(0) is<br />
p<br />
the surface electron concentration at the point, where<br />
the surface recombination centers are half-occupied;<br />
N and C are the density of surface recombination<br />
s s<br />
centers and their electron capture coefficient, correspondingly;<br />
D and D are the diffusion coefficients<br />
ns ps<br />
for electrons and holes at the surface, respectively.<br />
Sulphur-passivation of GaAs reduces the density<br />
of surface recombination centers [4]. Therefore the<br />
decrease in the recombination current, illustrated by<br />
curve 2 in Fig. 2, can be ascribed to the lowering of the<br />
density N of surface recombination centers in (2).<br />
s<br />
An analysis of curve 2 in Fig. 2 with the help of relation<br />
(2) gives the time-dependence N /N (0), where<br />
s s<br />
N (0) is the concentration of surface recombination<br />
s<br />
centers before storage. This dependence is represented<br />
by curve 1 in Fig. 4.<br />
Some information about the concentration N of<br />
electrically active centers at the surface of a p-n structure<br />
gives an analysis of ideality coefficient n of I–V<br />
t<br />
characteristics [6]. This coefficient is related to N by<br />
expression [6]<br />
5<br />
4
where<br />
N = N (1 − 2 / n ) , (3)<br />
t t<br />
N m kT q<br />
ε is the permittivity; m denotes the tunnel effective<br />
t<br />
mass of charge carriers.<br />
1,2<br />
0,8<br />
0,4<br />
0<br />
Ns / Ns0, N/Nt<br />
10 3<br />
2<br />
3<br />
2 2<br />
t = 12 ε t(<br />
) /( h ) , (4)<br />
10 4<br />
10 5<br />
3A<br />
1<br />
10<br />
t, s<br />
6 10 7<br />
Fig. 4. Effect of the storage duration on the concentrations:<br />
1 – N s /N s (0); 2 – N/N t in surface non-homogeneities; 3, 3A – average<br />
of N/N t in surface. Ordinates of curve 3A are normalized at<br />
t=7.2 . 10 3 s<br />
Curves 1 and 2 in Fig. 5 present the time-dependences<br />
of the ideality coefficients at low and high injection<br />
levels, respectively. An analysis of curve 1 by<br />
using formula (3) gives the change of electrically active<br />
centers concentration N in the surface depletion<br />
layer of non-homogeneities, which are responsible for<br />
the excess current. N(t) dependence during the storage<br />
for these centers is depicted as curve 2 in Fig. 4. The<br />
dependence N(t) for the homogeneous region of the<br />
surface, obtained from analogous analysis of curve 2 in<br />
Fig. 5, is represented by curve 3 in Fig. 4.<br />
A comparison between curves 2 and 3 in Fig. 4<br />
shows that the storage much stronger reduces the concentration<br />
of electrically active centers (non-compensated<br />
acceptors) in surface non-homogeneities, than<br />
in average at the surface of p-n structure. After a storage<br />
for one month curve 2 in Fig. 4 trends to curve 2.<br />
It means that the non-homogeneities at the surface,<br />
which are responsible for the excess current, disappear.<br />
The main defects that increase the excess current<br />
in p-n junctions on III–V semiconductors are dislocations<br />
[6]. And sulphur atoms are donors in GaAs.<br />
It permits to conclude that during the storage S atoms<br />
at the surface diffuse to dislocations and compensate<br />
their charge.<br />
As mentioned, curves 1 and 3 in Fig. 4 represent<br />
the time-dependences of the surface recombination<br />
centers concentration N s (normalized to initial its<br />
value N s (0)) and the electrically active centers at the<br />
surface N (normalized to the quantity N t , defined<br />
by Eq. (4)). Curve 3A was obtained from curve 3 by<br />
normalization to its value at t=2 h. The courses of<br />
curves 1 and 3A at t>10 4 s are similar. It means that<br />
the density of surface recombination centers and the<br />
concentration of electrically active centers vary identically<br />
during the storage. However, at the beginning,<br />
at t
The forward current at high injection level (at<br />
I>1μA) also decreases with the storage, due to lowering<br />
of both the density of surface recombination centers<br />
and the concentration of electrically active centers<br />
in the surface depletion layer.<br />
All these effects can be ascribed to compensation<br />
of acceptors (perhaps, related to dislocations)<br />
by sulphur atoms in the surface depletion layer and<br />
to destroying of surface states, acting as recombination<br />
centers.<br />
References<br />
1. Ïòàùåíêî Î. Î., Àðòåìåíêî Î. Ñ., Ïòàùåíêî Ô. Î.<br />
Âïëèâ ãàçîâîãî ñåðåäîâèùà íà ïîâåðõíåâèé ñòðóì â pn<br />
ãåòåðîñòðóêòóðàõ íà îñíîâ³ GaAs–AlGaAs // Ô³çèêà ³<br />
õ³ì³ÿ òâåðäîãî ò³ëà . — 2001. — Ò. 2, ¹ 3. — Ñ. 481 — 485.<br />
2. Ptashchenko O. O., Artemenko O. S., Dmytruk M. L. et al.<br />
Effect of ammonia vapors on the surface morphology and<br />
118<br />
UDC 621.315.592<br />
O. O. Ptashchenko, F. O. Ptashchenko, N. V. Masleyeva, O. V. Bogdan<br />
SURFACE CURRENT IN GaAs P-N JUNCTIONS, PASSIVATED BY SULPHUR ATOMS<br />
surface current in p-n junctions on GaP // Photoelectronics. —<br />
2005. — No. 14. — P. 97 — 100.<br />
3. Ïòàùåíêî À. À., Ïòàùåíêî Ô. À. P-n — ïåðåõîäû íà îñíîâå<br />
GaAs è äðóãèõ ïîëóïðîâîäíèêîâ À III B V êàê ãàçîâûå<br />
ñåíñîðû // Äåâÿòàÿ êîíôåðåíöèÿ “Àðñåíèä ãàëëèÿ è ïîëóïðîâîäíèêîâûå<br />
ñîåäèíåíèÿ ãðóïïû III-IV”: Ìàòåðèàëû<br />
êîíôåðåíöèè/ — Òîìñê: Òîìñêèé ãîñóíèâåðñèòåò,<br />
2006. — Ñ. 496 — 499.<br />
4. Äìèòðóê Í. Ë., Áîðêîâñêàÿ Î. Þ., Ìàìîíîâà Ë. Â. Ñóëüôèäíàÿ<br />
ïàññèâàöèÿ òåêñòóðèðîâàííîé ãðàíèöû ðàçäåëà<br />
ïîâåðõíîñòíî-áàðüåðíîãî ôîòîïðåîáðàçîâàòåëÿ íà îñíîâå<br />
àðñåíèäà ãàëëèÿ // Æóðíàë òåõíè÷åñêîé ôèçèêè. —<br />
1999. — Ò. 69, ¹6. — Ñ. 132 — 134.<br />
5. Ptashchenko O. O., Ptashchenko F. O., Masleyeva N. V. et<br />
al. Effect of sulfur atoms on the surface current in GaAs p-n<br />
junctions. // Photoelectronics. — 2007. — No 17. — P. 36 —<br />
39.<br />
6. Ptashchenko A. A., Ptashchenko F. A. Tunnel surface recombination<br />
in optoelectronic device modelling // Proc.<br />
SPIE. — 1997. — V. 3182. — P. 145 — 149.<br />
Abstract<br />
Influence of the storage (low-temperature annealing) of sulphur-passivated GaAs p-n structures in a neutral (helium) atmosphere<br />
at room temperature on I-V characteristics of forward and reverse currents was studied. The storage strongly reduces the excess forward<br />
current and the reverse current in p-n junctions. The ideality coefficient of I-V characteristics decreases with the storage. This effect has<br />
two stages. It is showed that all these phenomena can be explained by lowering of the surface recombination centers density and reduction<br />
of the electrically active centers concentration in the surface depletion layer.<br />
Key words: influence, surface current, P — N junctions.<br />
ÓÄÊ 621.315.592<br />
À. À. Ïòàùåíêî, Ô. À. Ïòàùåíêî, Í. Â. Ìàñëååâà, Î. Â. Áîãäàí<br />
ÏÎÂÅÐÕÍÎÑÒÍÛÉ ÒÎÊ Â P-N ÏÅÐÅÕÎÄÀÕ ÍÀ ÎÑÍÎÂÅ GaAs, ÏÀÑÑÈÂÈÐÎÂÀÍÍÛÕ ÀÒÎÌÀÌÈ ÑÅÐÛ<br />
Ðåçþìå<br />
Èññëåäîâàíî âëèÿíèå õðàíåíèÿ (íèçêîòåìïåðàòóðíîãî îòæèãà) ïàññèâèðîâàííûõ àòîìàìè ñåðû p-n ïåðåõîäîâ íà îñíîâå<br />
GaAs â íåéòðàëüíîé àòìîñôåðå (â ãåëèè) ïðè êîìíàòíîé òåìïåðàòóðå íà ÂÀÕ ïðÿìîãî è îáðàòíîãî òîêîâ. Ïðè õðàíåíèè<br />
ñóùåñòâåííî óìåíüøàþòñÿ ïðÿìîé èçáûòî÷íûé òîê è îáðàòíûé òîê â p-n ïåðåõîäàõ. Êîýôôèöèåíò èäåàëüíîñòè ÂÀÕ óìåíüøàåòñÿ<br />
â ïðîöåññå õðàíåíèÿ. Ýòîò ïðîöåññ äâóõñòàäèéíûé. Ïîêàçàíî, ÷òî âñå ýòè ÿâëåíèÿ ìîæíî îáúÿñíèòü óìåíüøåíèåì<br />
ïëîòíîñòè ïîâåðõíîñòíûõ öåíòðîâ ðåêîìáèíàöèè è óìåíüøåíèåì êîíöåíòðàöèè ýëåêòðè÷åñêè àêòèâíûõ öåíòðîâ â ïîâåðõíîñòíîì<br />
îáåäíåííîì ñëîå.<br />
Êëþ÷åâûå ñëîâà: ïîâåðõíîñòíûé òîê, P-N — ïåðåõîä, èññëåäîâàíèÿ.<br />
ÓÄÊ 621.315.592<br />
Î. Î. Ïòàùåíêî, Ô. Î. Ïòàùåíêî, Í. Â. Ìàñ뺺âà, Î. Â. Áîãäàí<br />
ÏÎÂÅÐÕÍÅÂÈÉ ÑÒÐÓÌ Ó P-N ÏÅÐÅÕÎÄÀÕ ÍÀ ÎÑÍβ GaAs, ÏÀÑÈÂÎÂÀÍÈÕ ÀÒÎÌÀÌÈ Ñ²ÐÊÈ<br />
Ðåçþìå<br />
Äîñë³äæåíî âïëèâ çáåð³ãàííÿ (íèçüêîòåìïåðàòóðíîãî â³äïàëó) ïàñèâîâàíèõ àòîìàìè ñ³ðêè p-n ïåðåõîä³â íà îñíîâ³ GaAs<br />
ó íåéòðàëüí³é àòìîñôåð³ (â ãå볿) ïðè ê³ìíàòí³é òåìïåðàòóð³ íà ÂÀÕ ïðÿìîãî ³ çâîðîòíîãî ñòðóì³â. Ïðè çáåð³ãàíí³ çíà÷íî<br />
çìåíøóþòüñÿ íàäëèøêîâèé ïðÿìèé ñòðóì òà çâîðîòíèé ñòðóì ó p-n ïåðåõîäàõ. Êîåô³ö³ºíò ³äåàëüíîñò³ ÂÀÕ çìåíøóºòüñÿ â<br />
ïðîöåñ³ çáåð³ãàííÿ. Äàíèé ïðîöåñ º äâîñòàä³éíèé. Ïîêàçàíî, ùî âñ³ ö³ ÿâèùà ìîæíà ïîÿñíèòè çìåíøåííÿì ù³ëüíîñò³ ïîâåðõíåâèõ<br />
öåíòð³â ðåêîìá³íàö³¿ òà çìåíøåííÿì êîíöåíòðàö³¿ åëåêòðè÷íî àêòèâíèõ öåíòð³â ó ïîâåðõíåâîìó çá³äíåíîìó øàð³.<br />
Êëþ÷îâ³ ñëîâà: ïîâåðõíåâèé ñòðóì, Ð-N — ïåðåõ³ä, äîñë³äæåííÿ.
UDÑ 539.19+539.182<br />
A. V. GLUSHKOV, YA. I. LEPIKH, A. P. FEDCHUK, A. V. LOBODA<br />
I. I. Mechnikov Odessa National University, Odessa<br />
Odessa National Polytechnical University<br />
THE GREEN’S FUNCTIONS AND DENSITY FUNCTIONAL APPROACH<br />
TO VIBRATIONAL STRUCTURE IN THE PHOTOELECTRON SPECTRA<br />
OF MOLECULES<br />
We present the basis’s of the new combined theoretical approach to vibrational structure in photoelectron<br />
spectra of molecules. The approach is based on the Green’s function method and quasiparticle<br />
density functional theory (DFT). The density of states, which describe the vibrational structure in<br />
photoelectron spectra, is defined with the use of combined DFT-Green’s-functions approach and is<br />
well approximated by using only the first order coupling constants in the one-particle approximation.<br />
Using the DFT theory leads to significant simplification of the calculation.<br />
I. INTRODUCTION<br />
A number of phenomena, provided by interaction<br />
of electrons with vibrations of the atomic nuclei<br />
in molecules or solids is manifested in the molecular<br />
photoelectron spectra. Indeed, the physics of the<br />
interaction of electrons with vibrations of the atomic<br />
nuclei in molecules or solids is more richer (c.f.[1-<br />
16]). One could mention here a great field of the resonant<br />
collisions of electrons with molecules, which are<br />
one of the most efficient pathways for the transfer of<br />
energy from electronic to nuclear motion. While the<br />
corresponding theory has been refined over the years<br />
with sophisticated and elaborate non-local treatments<br />
of the reaction dynamics, such studies have<br />
for the most part treated the nuclear dynamics in one<br />
dimension. This situation has resulted from the fact<br />
that, as the field of electron-molecule scattering developed,<br />
both experimentally and theoretically, the<br />
phenomena of vibrational excitation and dissociative<br />
attachment were first understood for diatomics,<br />
and it seemed natural to extend that understanding<br />
to polyatomic molecules using one-dimensional or<br />
single-mode models of the nuclear motion. However<br />
a series of experimental measurements of these phenomena<br />
in small polyatomic molecules have proven<br />
to be uninterpretable in terms of atomic motion with<br />
a single degree of freedom. Primary among these are<br />
potential-energy surfaces which describe the behavior<br />
of the electronic energy with respect to the locations<br />
of the nuclei, subject to the underlying Born-Oppenheimer<br />
or clamped nuclei approximation. From the<br />
ground- and excited-state wave functions one could<br />
in principle obtain all properties that arise from a solution<br />
to the vibrational Schrödinger equation that<br />
gives the frequencies, and, with the derivatives of the<br />
dipole moment, the infrared intensities [1-6]. Electronic<br />
excited states are also accessible along with<br />
electronic and photoelectron spectra.<br />
As it is often takes a place, the old multi-body<br />
quantum theoretical approaches, which have been<br />
primarily developed in a theory of superfluity and superconductivity,<br />
and generally speaking in a theory<br />
of solids, became by the powerful tools for develop-<br />
© A. V. Glushkov, Ya. I. Lepikh, A. P. Fedchuk, A. V. Loboda, 2009<br />
ing new conceptions in a theory of molecules [7-11].<br />
Many of them offers a synthesis of cluster expansions,<br />
Brueckner’s summation of ladder diagrams, the summation<br />
of ring diagrams Gell-Mann and an infinite-order<br />
generalization of many-body perturbation<br />
theory (MBPT). Using the quantum-field methods in<br />
molecular theory allowed to obtain a very powerful approach<br />
for correlation in many-electron systems.<br />
The Green’s method is very well known in a<br />
quantum theory of field, quantum theory of solids<br />
(c.f.[7,8]). Naturally, an attractive idea was to use<br />
it in the molecular theory. Returning to problem of<br />
description of the vibrational structure in photoelectron<br />
spectra of molecules, it is easily understand that<br />
this approach has great perspective (c.f.[12,13]). One<br />
could note that the experimental photoelectron (PE)<br />
spectra usually show a pronounced vibrational structure.<br />
Usually the electronic Green’s function is defined<br />
for fixed position of the nuclei. As result, only<br />
vertical ionization potentials (V.I.P.’s) can be calculated<br />
[14]. The cited method, however, requires as<br />
input data the geometries, frequencies, and potential<br />
functions of the initial and final states. Since in most<br />
cases at least a part of these data are unavailable, the<br />
calculations have been carried out with the objective<br />
of determining the missing data by comparison<br />
with experiment. Naturally, the Franck-Condon<br />
factors are functions of the derivatives of the difference<br />
between the potential curves of the initial and<br />
final states with respect to the normal coordinates.<br />
To avoid the difficulty and to gain additional information<br />
about the ionization process, the Green’s<br />
functions approach has been extended to include the<br />
vibrational effects in the photoelectron spectra. Nevertheless,<br />
there are well known great difficulties of<br />
the correct interpretation of the photoelectron spectra<br />
for any molecules.<br />
Further let us remember that for larger molecules<br />
and solids, far more approximate but more easily<br />
applied methods such as density-functional theory<br />
(DFT) [15-17] or from the wave-function world the<br />
simplest correlated model MBPT are preferred [3].<br />
Indeed, in the last decades DFT theory became by a<br />
great, quickly developing field of the modern quan-<br />
119
tum computational chemistry of molecules, solids.<br />
Naturally, this approach does not allow to reach a<br />
spectroscopic accuracy in description of the different<br />
molecular properties, nevertheless, the key idea is very<br />
attractive and can be used in new combined theoretical<br />
approaches.<br />
Here we present the basis’s of the new combined<br />
theoretical approach to vibrational structure in photoelectron<br />
spectra of molecules. The approach is<br />
based on the Green’s function method (Cederbaum-<br />
Domske version) [12,13] and Fermi-liquid DFT formalism<br />
[11,14] (see also [18-23,25]). The density of<br />
states, which describe the vibrational structure in molecular<br />
photoelectron spectra, is calculated with the<br />
help of combined DFT-Green’s-functions approach.<br />
In addition to exact solution of one-bode problem<br />
different approaches to calculate reorganization and<br />
many-body effects are presented. The density of states<br />
is well approximated by using only the first order coupling<br />
constants in the one-particle approximation.<br />
It is important that the calculational procedure is<br />
significantly simplified with using the quasiparticle<br />
DFT formalism. Thus quite simple method becomes<br />
a powerful tool in interpreting the vibrational structure<br />
of photoelectron spectra for different molecular<br />
systems.<br />
2. The Hamiltonian of the system. The density of<br />
states in one-body and many-body solution<br />
The quantity which contains the information<br />
about the ionization potentials and the molecular vibrational<br />
structure due to quick ionization is the density<br />
of occupied states [12]:<br />
120<br />
−1<br />
ihº t t<br />
k( ) (1/2 h)<br />
0 a (0) k(<br />
)<br />
k<br />
0<br />
N º = π ∫ dte 〈ψ a t ψ 〉 , (1)<br />
where Ψ〉 0 is the exact ground state wavefunction of<br />
the reference molecule and ak() t is an electron destruction<br />
operator, both in the Heisenberg picture.<br />
For particle attachment the quantity of interest is the<br />
density of unoccupied states:<br />
1<br />
ihº t<br />
t<br />
k ( ) (1/ 2 h)<br />
0 a k(t)a (0) k 0<br />
N dte −<br />
º = π ∫ 〈ψ ψ 〉 (2)<br />
Usually in order to calculate the value (1) states<br />
for photon absorption one should express the Hamiltonian<br />
of the molecule in the second quantization formalism<br />
(see [12,13]). The corresponding Hamiltonian<br />
is as follows:<br />
H = TΕ( ∂/ ∂ x) + TΝ( ∂/ ∂ X) + U( x, X)<br />
, (3)<br />
where TΕ is the kinetic energy operator for the electrons,<br />
TΝ is the kinetic energy operator for the nuclei,<br />
and U represents the interaction<br />
U( x, X) = UΕΕ ( x) + UΝΝ ( X) + UΕΝ ( x, X)<br />
, (4)<br />
where x denotes electron coordinates, X denotes nuclear<br />
coordinates, UΕΕ represents the Coulomb interaction<br />
between electrons, etc. Introducing the field<br />
operator:<br />
Ψ( R, θ , x) = ∑ φi( x, R, θ) ai( R,<br />
θ)<br />
(5)<br />
i<br />
where the ô are Hartree-Fock (HF) one–particle<br />
i<br />
functions and the a are destruction operators for a HF<br />
i<br />
particle in the “i” state, them the Hamiltonian in a occupation<br />
number representation is given as<br />
H = H ( R, θ ) + U ( R, θ ) + T ( ∂/ ∂ R)<br />
, (6)<br />
ΕΝ 0 ΝΝ 0 Ν<br />
H<br />
t 1<br />
= ∑º ( Raa ) + ∑V<br />
i<br />
2<br />
t t<br />
( Raaaa )<br />
[ V<br />
t<br />
( R)] a a ,<br />
EN i i i ijkl i j l k<br />
−∑ ∑ (7)<br />
ikkj i j<br />
ij k∈ f<br />
Vijkl 2<br />
ij e r<br />
'<br />
r<br />
−1<br />
kl<br />
=〈 − 〉<br />
The i ( ) R ∈ are the one-particle HF energies and f<br />
denotes the set of orbitals occupied in the HF ground<br />
state. As usually in the adiabatic approximation one<br />
could write the eigenfunctions to H as products<br />
xR , , θ〉 0 E × R〉<br />
N ,and further expand i ( ) R ∈ , Vijkl ( R ) ,<br />
and UNN ( R, θ ) about R leaving the operators a 0 i and<br />
a unchanged [13]:<br />
t<br />
i<br />
H= ∑<br />
i<br />
t 1<br />
t t<br />
º i ( R0) aa i i + ∑ Vijkl ( R0) aaaa i j l k −<br />
2<br />
− [ V ( R ) − V ( R )] a a +<br />
∑∑<br />
ij k∈ f<br />
ikjk 0 ikkj 0<br />
t<br />
i j<br />
⎛∂∈ ⎞<br />
M<br />
1<br />
+ ∑ [ ∑ ⎜ ⎟ ( R− Rso<br />
) +<br />
i s= 1 ⎝ ∂Rs<br />
⎠0<br />
2<br />
⎛ ∂ ∈ ⎞ i<br />
M 1<br />
+ ∑ ∑ ⎜ ⎟ ( Rs − Rso)<br />
×<br />
2 ⎝∂R ∂R<br />
⎠<br />
i s, s'=<br />
1 s s'0<br />
t<br />
⎛ ∂ ⎞<br />
× ( Rs' − Rs'0) ] aiai + ... + UNN( R0, θ 0)<br />
+ ... + TN<br />
⎜ ⎟<br />
, (8)<br />
⎝∂R⎠ where M is the number of normal coordinates.<br />
Choosing R as the equilibrium geometry on the<br />
0<br />
HF level and introducing dimensionless normal coordinates<br />
Q one can write the following Hamiltonian<br />
s<br />
(the subscript 0 stands for R ) [13]:<br />
0<br />
H = H + H + H + H ,<br />
(1) (2)<br />
E N EN EN<br />
1<br />
H = º ( R) aa + V ( R) aaaa −<br />
∑ ∑<br />
t t t<br />
E<br />
i<br />
i 0 i i<br />
2<br />
ijkl 0 i j l k<br />
t<br />
−∑∑ [ Vikjk ( R0) −Vikkj<br />
( R0)] aiaj, i, j k∈f M<br />
t 1<br />
HN = h ∑ ω s( bsbs + ),<br />
s=<br />
1 2<br />
⎛ ∂є<br />
⎞<br />
M<br />
(1) −1/<br />
2<br />
i<br />
t t<br />
HEN = 2 ∑ ⎜ ⎟ ( bs+ bs)[ aiai− ni]<br />
+<br />
s= 1 ⎝∂Qs⎠0 i<br />
M<br />
s, s'=<br />
1<br />
2<br />
⎛ ∂ є ⎞ i<br />
⎜ ⎟<br />
⎝∂Qs∂Qs'⎠0 bs t<br />
bs bs' t<br />
bs' t<br />
aiai ni<br />
1<br />
+ ∑ ∑<br />
( + )( + )[ − ],<br />
4<br />
⎛∂V⎞ H b b<br />
M<br />
(2) −3/2<br />
EN = 2 ∑∑<br />
s= 1<br />
ijkl<br />
⎜ ⎟ (<br />
⎝ ∂Qs<br />
⎠0<br />
s +<br />
t<br />
s ) ×<br />
t t<br />
vaaa 1 i j k<br />
t t<br />
vaaaa 2 l k i j<br />
t t<br />
vaaaa<br />
3 j k l i<br />
×δ [ +δ + 2 δ ]<br />
1 ⎛ ∂ V ⎞<br />
+ ∑ ∑<br />
( + ) ×<br />
8<br />
M 2<br />
ss , '= 1<br />
ijkl<br />
⎜ ⎟<br />
⎜ QsQ ⎟ '' ⎝<br />
∂ ∂ s ⎠0<br />
bs t<br />
bs<br />
t t t<br />
× ( bs' + bs'[ δ v1aiajak +<br />
2 l k<br />
t<br />
i<br />
t<br />
j 2 3<br />
t<br />
j k l<br />
t<br />
i],<br />
+δ vaaaa + δvaaaa<br />
(9)
ni = 1 i ∈ f , δσ f = 1 ( ijkl)<br />
∈σ,<br />
= 0, i ∉ f , = 0 ( ijkl)<br />
∉σ ,<br />
where the index set v means that at least 1 k<br />
or i φ and j φ are unoccupied, v2 the orbitals is unoccupied, and v that 3 k<br />
φ and l φ<br />
that at most one of<br />
φ and j φ or l φ<br />
and φ j are unoccupied. Besides, here for simplicity all<br />
terms leading to anharmonicities are neglected. The<br />
t<br />
ωs are the HF frequencies and the b s and b s are destruction<br />
and creation operators for vibrational quanta<br />
defined by<br />
t<br />
Q = (1/ 2)( b + b ),<br />
s s s<br />
f<br />
/ (1/ 2)( )<br />
t<br />
∂ ∂ Qs = bs − bs<br />
(10)<br />
The first term H E describes the electronic motion<br />
for nuclei fixed at the HF ground state geometry. The<br />
second term H N describes the motion of the nuclei in<br />
the harmonic HF potential (the extension to anhar-<br />
(1)<br />
monic terms can easily be done). H EN represents the<br />
coupling of the HF particles with the nuclear motion.<br />
The coupling constants are the normal coordinate<br />
derivatives of the HF one-particle energies. The first<br />
(1)<br />
sum in the expression for H EN is responsible for the<br />
geometry shifts and the second one for the charge of<br />
frequencies due to electrons. There is also a modification<br />
of the interaction between electrons through the<br />
(2)<br />
coupling to the nuclear motion. The term H EN , which<br />
describes this modification, is due to its nature less im-<br />
(1)<br />
portant than H EN .<br />
The exact solution of the one-body HF problem<br />
has been given in ref.[12]. Correspondingly in the<br />
one-particle picture the density of occupied states is<br />
given by<br />
1<br />
1<br />
0 1<br />
iє є( t )<br />
i H<br />
k<br />
0<br />
t<br />
Nk( ) dte 0 e 0 ,<br />
2<br />
−<br />
∞<br />
−<br />
−<br />
±<br />
= 〈 〉<br />
π −∞<br />
∫<br />
×〈ψ0 T<br />
t<br />
ak() t ak(0)<br />
ψ 0〉<br />
(12)<br />
where T is Wick’s time ordering operator and the function<br />
Nє k ( ) then follows from the relation<br />
π Nk( º ) = aIm Gkk( º −aiη) , (13)<br />
a =− signº<br />
k ,<br />
where η is a positive infinitesimal. Choosing the unperturbed<br />
Hamiltonian H 0 to be<br />
t<br />
H0= ∑ º i aiai + HN<br />
(14)<br />
one finds for the corresponding Green’s functions<br />
0<br />
Gkk ( º ) =δkk ' /( º −º k −aiη) (15)<br />
The Dyson equation<br />
0 0<br />
Gkk ' = Gkk ' + ∑ GkkΣ kk ''Gk'' k ' (16)<br />
k ''<br />
h %<br />
h<br />
º (11)<br />
h<br />
with<br />
relates the Green’s functions to the free ones introducing<br />
a new function kk '' ( ) є Σ called the (proper) self-energy<br />
part. In in order to calculate Σ kk ' , a well-known<br />
diagrammatic method is used. The sum of Feynman<br />
diagrams leading to the self-energy part is shown in<br />
Fig. 1. All notations are standard.<br />
The one-body problem treated above results in the<br />
exact solution of the Dyson equation with the selfenergy<br />
part given by the infinite number of diagrams<br />
shown in the first row of Fig. 1 and the corresponding<br />
Green’s function is as follows [12-14]:<br />
g<br />
0<br />
M<br />
∑ s<br />
t<br />
s s<br />
M<br />
∑<br />
k<br />
s s<br />
t<br />
s<br />
s= 1 s=<br />
1<br />
M<br />
k t t<br />
∑ γ ss '( bs+ bs)( bs' + bs')<br />
ss , '= 1<br />
H% = h ω b b + g ( b + b ) +<br />
1 ⎛ ∂є ⎞<br />
1⎛<br />
∂ є ⎞<br />
.<br />
2<br />
i i i<br />
i<br />
s =± ⎜ ⎟ , γ ss'=±<br />
⎜ ⎟<br />
2 ⎝∂Qs ⎠ 4 Q 0 ⎝∂ s∂Qs'⎠0 In a diagrammatic method in order to obtain the<br />
function N k ( º ) one should calculate the Green’s<br />
function G kk ' ( º ) first:<br />
kk '<br />
Fig. 1. The sum of diagrams contributing to the self-energy part<br />
−1<br />
ih<br />
h ∫−∞<br />
{ }<br />
−1º<br />
t<br />
G ( º ) =− i dte ×<br />
∞<br />
( m )<br />
OB<br />
−1<br />
Gkk′ () t =±δkk′ iexp⎡ ⎣−in εk Δε t⎤<br />
⎦×<br />
∑<br />
n<br />
2<br />
n$ k Uk 0 exp(<br />
ink kt)<br />
× ± ⋅ω<br />
(17)<br />
121
The corresponding Dyson-like equation is as follows:<br />
OB OB<br />
Gkk′ () ∈ = Gkk′ () ∈ + ∑Gkk () ∈Ô kk′′ Gk′′<br />
k′<br />
() ∈ (18)<br />
122<br />
kk′<br />
The expression for a sum of the first 2 diagrams<br />
appearing in Fig. 2 are written by a standard way:<br />
Ф<br />
kk′<br />
where<br />
()<br />
( − )<br />
Vklij Vklij Vk′ lijUniiUnj jUnll ∈ = ∑ ∑<br />
+<br />
∈+ E −E −E<br />
+<br />
i, j∈F ni, nj, nl l i j<br />
l∉F ( Vklij −Vklij<br />
) Vk′ lijUniiUnj jUnll ∑ ∑ (19)<br />
∈+ E −E −E<br />
i, j∈F ni, nj, nl l i j<br />
l∉F 2<br />
U $<br />
ni i 0<br />
i = n Ui<br />
and E $<br />
i =∈i mΔ∈i m hni<br />
⋅ωi<br />
The direct method for calculation of N (∈) as the<br />
k<br />
imaginary part of the corresponding Green’s function<br />
implicitly includes the determination of the V. I. P. s of<br />
the reference molecule and then of Nk() ∈ . The zeros<br />
of the functions<br />
op () ∈ =∈− ∈ +Σ() ∈<br />
D ⎡ k ⎣<br />
⎤<br />
⎦ , (20)<br />
k<br />
op<br />
where ( ∈ +Σ) denotes the kth eigenvalue of the di-<br />
k<br />
agonal matrix of the one-particle energies added to the<br />
matrix of the self-energy part, are the negative V. I. P. ‘s<br />
for a given geometry. Further it is easily to write [12]:<br />
( VIP) ( F)<br />
.. . k k k<br />
=− ∈ + ,<br />
1<br />
Fk =Σkk( −( V.. I P.<br />
) ) ≈ Σkk( ∈k)<br />
. (21)<br />
k<br />
1 −∂Σkk ( ∈k ) / ∂∈<br />
N 1<br />
Expanding the ionic energy Ek − about the equilibrium<br />
geometry of the reference molecule in a power<br />
series of the normal coordinates of this molecule:<br />
where<br />
( F )<br />
⎛∂∈ + ⎞<br />
E = E − Q −<br />
M<br />
N−1 N−1<br />
k k<br />
k k () 0 ∑ ⎜ ⎟ s<br />
s= 1 ⎝ ∂Qs<br />
⎠0<br />
q<br />
k<br />
=<br />
∑<br />
Fig. 2. Perturbation expansion of Ô kk<br />
2<br />
( V V )<br />
where kk Ф ′ , is equal to Σ kk′ , less the diagrams of the<br />
first row in Fig.1. The perturbation expansion of Ô is<br />
OB<br />
shown in Fig. 2 where iG kk′<br />
, is symbolized by a double<br />
solid line.<br />
N ( k Fk E0)<br />
2 ⎛ ⎞<br />
M 1 ∂ ∈ + −<br />
− ∑ ⎜ ⎟ QQ s s′<br />
(22)<br />
2! ss , ′= 1⎜<br />
∂Qs∂Q ⎟<br />
s′<br />
⎝ ⎠0<br />
leads to a set of linear equations in the unknown normal<br />
coordinate shifts ΔQ , S<br />
2<br />
( k Fk) ( k Fk)<br />
∑<br />
⎛∂∈ + ⎞ ⎛∂∈ + ⎞<br />
− ⎜ ⎟ = ⎜ ⎟ δQs′<br />
, (23)<br />
∂Q ⎜ ∂Q ∂Q<br />
⎟<br />
⎝ ⎠ ⎝ ⎠<br />
s 0<br />
s′≠ s s s′<br />
0<br />
( F )<br />
⎡ 2<br />
⎛∂ ∈ k + ⎞ ⎤<br />
k<br />
⎢⎜ 2 ⎟ s⎥ s<br />
⎢<br />
⎜ Q ⎟<br />
⎝ ∂ 3 ⎠0⎥<br />
+ −hω δ Q , s = 1... M,<br />
⎣ ⎦<br />
where ω s are frequencies of the reference molecule.<br />
The new coupling constants are then:<br />
( ) ( )<br />
g =± 1/ 2 ⎡⎣∂∈ + F / ∂Q<br />
⎤⎦<br />
(24)<br />
1 k k l 0<br />
⎛1⎞ 2<br />
γ ll′ = ± ⎜ ⎟⎣<br />
⎡∂ ( ∈ k + Fk) / ∂Ql/ ∂Ql′<br />
⎤<br />
4<br />
⎦0<br />
⎝ ⎠<br />
Thee coupling constants l g and ll y ′ are calculated<br />
by the well-known perturbation expansion of the selfenergy<br />
part using the Hamiltonian H of Eq. (6). In<br />
EN<br />
second order one obtains:<br />
( − )<br />
V V V<br />
∈ = +<br />
() 2<br />
ksij ksji ksij<br />
∑ kk () ∑<br />
i, j ∈+∈s −∈i −∈j<br />
s∉F ( − )<br />
V V V<br />
+<br />
∈+∈ −∈ −∈<br />
and the coupling constant g , can be written as<br />
l<br />
g<br />
ksij − ksji ⎡∂∈s ∂∈ ∂∈ i j ⎤<br />
2 ⎢ − − ⎥<br />
( VIP .. . )<br />
∂Q s i j<br />
l ∂Ql ∂Ql<br />
⎣<br />
⎡− +∈ −∈ −∈ ⎤ ⎣ ⎦<br />
k<br />
⎦<br />
l<br />
∑<br />
ksij ksji ksij<br />
(25)<br />
i, j<br />
s∉F s i j<br />
k( ) ∑ kk ( ) k<br />
( ) ∑ ( )<br />
1 ∂∈ 1 + q ∂/ ∂∈ ⎡− V. I. P.<br />
⎤<br />
k<br />
≈±<br />
⎣ ⎦<br />
, (26)<br />
2 ∂Ql 1 − ∂/ ∂∈ ⎡ kk ⎣− VIP . . . ⎤<br />
k ⎦<br />
2<br />
( Vksij −Vksji<br />
)<br />
∂∈<br />
∑ ∂ ⎡ ⎤<br />
k<br />
2<br />
Ql ⎣− ( VIP .. . ) +∈ k s −∈i −∈j⎦<br />
(27)
It is suitable to use further the pole strength of the<br />
corresponding Green’s function<br />
[ ] 1 −<br />
VIP<br />
⎧ ∂<br />
⎫<br />
ρ k = ⎨1 − ∑ kk −( . . .) k ⎬ ;1≥ρk ≥0,<br />
(28)<br />
⎩ ∂∈<br />
⎭<br />
and then<br />
( )<br />
0<br />
0 −1/2<br />
gl ≈ gl ⎡⎣ρ k + qk<br />
ρk −1<br />
⎤⎦<br />
, gl =± 2 ∂∈k / ∂ Ql<br />
(29)<br />
Below we firstly give the DFT definition of the pole<br />
strength corresponding to V. I. P.’s and confirm earlier<br />
data [13,14]: p ≈0,8-0,95. The closeness of p to 1 in<br />
k k<br />
fact means that a role of the multi-body correlation<br />
0<br />
effects is small ( gl ≈ gl<br />
). The above presented results<br />
can be usefully treated in the terms of the correlation<br />
and reorganization effects. Usually it is introduced the<br />
following expression for an I.P.:<br />
( IP . . )<br />
k<br />
( Vkikj −Vkijk<br />
)<br />
∑<br />
=−∈k − −<br />
∈ −∈<br />
j∉, i∈F j i<br />
( Vkijl −Vkilj<br />
)<br />
1<br />
− −δ<br />
2 ∈ +∈ −∈ −∈<br />
∑ ( 1 ik ) (30)<br />
i∈F k i j l<br />
jl , ∉F<br />
2<br />
( Vkjpq −Vkjqp<br />
)<br />
1<br />
− ∑ ( 1−δkp )( 1−δkq<br />
)<br />
2 pq , ∈F∈<br />
k +∈i −∈p −∈q<br />
j∉F The first correction term is due to reorganization,<br />
the remaining correction terms are due to correlation<br />
effects. Then the coupling constant g , can be written as<br />
l<br />
gl ⎧<br />
0 ⎪<br />
gl<br />
⎨1<br />
⎪⎩<br />
2<br />
( Vkkkj<br />
)<br />
∑<br />
2<br />
j∉F( ∈j −∈k)<br />
2<br />
( Vkijl −Vkilj<br />
)<br />
∑<br />
1 2<br />
( ∈ k +∈i −∈j −∈l)<br />
≈ + −<br />
⎡<br />
1<br />
− ⎢<br />
( −δ ki ) +<br />
2 ⎢i∈F<br />
⎢⎣ jl , ∉F<br />
2<br />
⎫<br />
( Vkjpq −V ⎤<br />
kjqp )<br />
⎪<br />
+ 2 ( 1 kq )( 1 kp ) ⎥<br />
∑ −δ −δ ⎬ (31)<br />
pq , ∈F(<br />
∈<br />
⎥ ) j, ∉F<br />
k +∈i −∈p −∈q ⎥⎦⎭<br />
⎪<br />
The second coupling constant can be written<br />
γ = γ<br />
⎛ g<br />
⎜<br />
⎝<br />
⎞ 1<br />
⎟+ ⎠<br />
2g<br />
∂ ⎛ g<br />
⎜<br />
Q ⎝<br />
⎞<br />
⎟<br />
⎠<br />
γ , is defined analogously<br />
0<br />
ll<br />
0 l 0<br />
l<br />
ll ll 0<br />
gl 4<br />
l<br />
∂ l<br />
0<br />
gl<br />
0<br />
g l .<br />
(32)<br />
3. QUASIPARTICLE DENSITY FUNCTIONAL<br />
THEORY<br />
The quasiparticle Fermi-liquid version of the DFT<br />
theory has been presented in ref. [11,14] (see also [18-<br />
23,25]), starting from the problem of searching for the<br />
optimal one-electron representation [2-5]. One of the<br />
simplified recipes represents the Kohn-Sham DFT<br />
theory [15]. Earlier new QED DFT version, based on<br />
the formally exact QED perturbation theory (energy<br />
approach), has been developed and a new approach<br />
to construction of the optimized one-quasiparticle<br />
representation has been proposed (look details in ref.<br />
[11]). The energy approach uses the adiabatic Gell-<br />
Mann and Low formula for the energy shift ΔE with<br />
electrodynamic scattering matrice. In a modern theory<br />
of molecules there is a number of tasks, where an<br />
accurate account for the complex exchange-correlation<br />
effects, including the continuum pressure, energy<br />
dependence of a mass operator etc., is critically important.<br />
It includes also the calculation of the vibration<br />
structure for the molecular systems. In this case it<br />
can be very useful the quasiparticle DFT [11,14].<br />
In order to get the master equations and construct<br />
an optimal basis of the one-particle wave functions<br />
ϕ λ one could use the Green’s function method. Let<br />
us define the one-particle Hamiltonian for functions<br />
ϕ λ so that the Greens’ function pole part in the ( λ ϕ<br />
) representation is diagonal on λ . Starting equation<br />
is the Dyson equation for multi-electron (for example<br />
atom or molecule):<br />
∑<br />
∑<br />
2 /<br />
( p /2 Zα / rα) G( x, x , )<br />
ε− + ⋅ ε −<br />
/ / /<br />
∫ dx ( x, x , ) ( x x ) (33)<br />
− ε =δ −<br />
where x = (,) r s are the spatial and spin variables, ∑<br />
is the mass operator; Z , as usually, a charge of a nucleus<br />
(nuclei) “ α ”, G is the Green’s function. In the<br />
/<br />
representation of auxiliary functions ϕ λ the equation<br />
(67) has the following form:<br />
p Z<br />
( ε⋅δ −[ − + ( , , ε )] ) =δ<br />
λλ1 2<br />
2<br />
α<br />
rα<br />
/<br />
xx λλ1 G /<br />
λλ<br />
/<br />
λλ<br />
∑ ∑ (34)<br />
where λ 1 is an index of summation. It is natural to<br />
choose ϕ λ so that the following expression will be diagonal:<br />
∑ ∑ (35)<br />
2 /<br />
[ p /2 − Zα / rα + ( x, x , ε )] λλ = E ( )<br />
1 λ ε ⋅δλλ1<br />
α<br />
Then the Green’s function is diagonal on λ :<br />
G / = G ⋅δ /, G = 1/[ ε−E ( ε )] (36)<br />
λλ λ λλ λ λ<br />
/<br />
and the functions ϕ λ , which diagonalizes G , satisfy to<br />
equation as follows: :<br />
∑<br />
2<br />
( p /2 −<br />
/<br />
Zα / rα) ϕλ( x,<br />
ε ) +<br />
/<br />
∫ ∑(<br />
xx , , )<br />
α<br />
/<br />
λ( x1, ) dx1 Eλ( )<br />
/<br />
λ(<br />
x,<br />
) (37)<br />
+ ε ϕ ε − ε ϕ ε<br />
One could introduce the mixed representation for<br />
a mass operator as follows:<br />
∑ ∑<br />
∫ 1 1 1 (38)<br />
( x, p, ε ) = ( x, x , ε)exp[ i( r−r) p] dr<br />
Then equation (37) with account for of the expression<br />
(38) can be written as follows:<br />
2 /<br />
[ p /2 − Z / r + ( x, p, ε)] ϕ ( x,<br />
ε ) =<br />
∑ ∑<br />
α<br />
α α λ<br />
/<br />
= Eλ() ε ϕλ(,) x ε<br />
(39)<br />
It can be shown that an operator p = iv in (33) acts<br />
on functions which are on the right of ∑ ( x, p, ε)<br />
. So,<br />
in order to find the one-particle energies, defined by<br />
the pole part of the Green’s function G, it is sufficient<br />
/<br />
to know the functions ϕ λ under ε=ε λ . The Greens’<br />
function pole part is as follows:<br />
λ<br />
G / = a δ / /( ε−ε + iγ<br />
) (40)<br />
λλ λλ<br />
λ λ<br />
123
where<br />
124<br />
λ<br />
a = 1/(1 − ∂E / δε)| ,( ∂E/ ∂ε )| = ( ∂E/ ∂ε )|<br />
λ ε=ε /<br />
λ ε=ελ λλ<br />
∑ ∑ (41)<br />
ε = ε = − + ε<br />
λ Eλ() 2<br />
{ p /2 Zα / rα (, x p,)}|<br />
λλ<br />
α<br />
/<br />
λ( x) /<br />
λ( x,<br />
λ)<br />
The functions ϕ<br />
following equation:<br />
=ϕ ε are satisfying to<br />
∑ ∑ (42)<br />
2<br />
[ p /2 − Zα / rα + ( x, p, ελ)] ϕ λ =ελϕλ( x)<br />
α<br />
Introducing an expansion for self-energy part ∑<br />
2 2<br />
into set on degrees x, ε−εF, p − pF<br />
(here ε F and F p<br />
are the Fermi energy and pulse correspondingly):<br />
∑ ∑<br />
( x, p, ε ) = ( x)<br />
+<br />
+∂∑ ∂ − + ∂∑ ∂εε−ε +<br />
then equation (42) is rewritten as follows:<br />
2<br />
[ p /2 − Z / r + ( x)<br />
+<br />
0<br />
(<br />
2 2<br />
/ p )( p<br />
2<br />
pF)<br />
( / )( F)<br />
...<br />
∑ ∑<br />
∑ ∑ (43)<br />
α α 0<br />
( /<br />
2<br />
) ]<br />
α<br />
λ( ) (1 / ) λ λ(<br />
)<br />
+ p ∂ ∂p p Φ x = −∂ ∂ε ε Φ x<br />
The functions Φ λ in (77) are orthogonal with a<br />
-1 1<br />
weight ρ = a [1 / ]<br />
k<br />
− = −∂∑ ∂ε . Now one can introduce<br />
1/2<br />
the wave functions of the quasiparticles a −<br />
ϕ λ = Φ λ ,<br />
which are, as usually, orthogonal with weight 1. For<br />
complete definition of { ϕ λ}<br />
it should be determined<br />
2<br />
the values 0 , / , / p ∑ ∂∑ ∂ ∂∑ ∂ε.<br />
Naturally, the equations<br />
(43) can be obtained on the basis of the variational<br />
principle, if we start from a Lagrangian of a system<br />
L q (density functional). It should be defined as a<br />
functional of the following quasiparticle densities:<br />
∑<br />
0 () r nλ |<br />
2<br />
λ()|,<br />
r<br />
1()<br />
r<br />
λ<br />
∑ nλ |<br />
λ<br />
2<br />
λ()|,<br />
r<br />
2 () r ∑ nλ[<br />
λ<br />
*<br />
λ λ<br />
*<br />
λ λ].<br />
ν = Φ<br />
ν = ∇Φ<br />
ν = Φ Φ −Φ Φ<br />
(44)<br />
The densities 0 ν and ν 1 are similar to the HF electron<br />
density ρ ( ρ=ν⋅ a ) and kinetical energy density<br />
correspondingly; the density ν 2 has no an analog in the<br />
HF or standard Kohn-Sham theory and appears as result<br />
of account for the energy dependence of the mass<br />
operator. A Lagrangian L q can be written as a sum of a<br />
free Lagrangian and Lagrangian of interaction:<br />
0 int<br />
Lq = Lq + Lq<br />
,<br />
0<br />
where a free Lagrangian L q has a standard form:<br />
0 *<br />
Lq = ∫ dr∑nλΦλ( i∂/ ∂t−εp) Φλ,<br />
(45)<br />
λ<br />
And an interaction Lagrangian is defined in the<br />
form, which is characteristic for a standard Kohn-<br />
Sham DFT theory (as a sum of the Coulomb and exchange-correlation<br />
terms), however, it takes into account<br />
for the energy dependence of a mass operator:<br />
1<br />
L L F r r r r drdr<br />
2<br />
int<br />
q = K −<br />
2 ik , = 0<br />
βik ( 1, 2) νi( 1) νk(<br />
2) 1 2<br />
∑ ∫ (46)<br />
where β ik are some constants (look below), F is an effective<br />
potential of the exchange-correlation interaction. Let<br />
us explain here the essence of the introduced constants.<br />
Indeed, in some degree they have the same essence as the<br />
similar constants in the well-known Landau Fermi-liquid<br />
theory and Migdal finite Fermi-systems theory. The<br />
Coulomb interaction part LK looks as follows:<br />
1<br />
LK=− [1 − 2( r1)] ν0( r1)[1−<br />
2 ∫ ∑<br />
− ν −<br />
∑ 2( r2)] 0( r2)/| r1 r2 | dr1dr (47)<br />
2<br />
∑ ∑ . Regarding the exchange-cor-<br />
where 2 =∂ / ∂ε<br />
relation potential F, it should be noted the there are<br />
many possible approximations (directly in the density<br />
–functional theory and its modern generalizations).<br />
One of the suitable forms for this potential is the Ivanov-Ivanova<br />
potential (look details in ref. [11]):<br />
∫<br />
F( r , r ) = X ( drρ ( r)/ r −rr−r− 1 2<br />
(0)1/3<br />
c<br />
1 2<br />
' (0)1/3 / /<br />
−( ∫ dr ρc( r )/ r1 −r⋅ //<br />
dr<br />
(0)1/ 3 //<br />
c r<br />
//<br />
r r2 dr<br />
(0)1/ 3<br />
c r<br />
∫ ∫ (48)<br />
⋅ ρ ( )/ − )/ ρ ( )<br />
where X is the numerical coefficient. It has been<br />
obtained on the basis of calculating the Rayleigh-<br />
Schrödinger perturbation theory Feynman diagrams<br />
of the second and higher order (so called polarization<br />
diagrams) in the Thomas-Fermi approximation [26].<br />
The relativistic generalization of the potential (48) is<br />
obtained in ref.[11].<br />
In the local density approximation in the density<br />
functional the potential F can be expressed through<br />
the exchange-correlation pseudo-potential V XC as follows<br />
[15,]:<br />
Frr ( 1, 2 ) =δVXC / δν0⋅δ( r1− r2).<br />
(49)<br />
Further, one can get the following expressions for<br />
int<br />
i q / 1 L ∑ =−δ δν :<br />
ex<br />
∑ 0 = (1 − ∑ e) VK<br />
+ ∑ 0 +<br />
1 2 2 2<br />
+ β00δ VXC / δν ⋅ν 0 +β00δVXC / δν0⋅ν 0 +<br />
2<br />
2 2<br />
+β01δVXC / δν0⋅ν 1 +β01δ VXC<br />
/ δν0⋅ν0ν 1 +<br />
2 2<br />
+β δ V / δν ⋅ν ν + β δV / δν ⋅ν (50)<br />
02 XC 0 0 2 02 XC 0 2<br />
∑<br />
1 =β01δVXC / δν0⋅ν 0 +<br />
+β δV / δν ⋅ν + β δV / δν ⋅ν ;<br />
12 XC 0 2 11 XC 0 1<br />
∑<br />
2 =β02δVXC / δν0⋅ν 0 +<br />
+β12δVXC / δν0 ⋅ν 1 + β22δVXC / δν0 ⋅ν 2;<br />
ex<br />
Here V K is the Coulomb term (look above), ∑ 0 is the<br />
exchange term. Using the known canonical relation-<br />
* *<br />
ship Hq =ΦλδLq / δΦ λ +ΦλδLq / δΦλ − Lq<br />
after some<br />
transformations one can receive the expression for the<br />
quasiparticle Hamiltonian, which is corresponding to<br />
a Lagrangian L q :<br />
0 int 0 1<br />
2<br />
Hq = Hq + Hq = Hq − LK + β00δVXC / δν0⋅ν 0 +<br />
2<br />
1<br />
2<br />
+β01δVXC / δν0⋅ν0⋅ν 1 + β11δVXC / δν0⋅ν1− 2<br />
1<br />
2<br />
− β22δVXC / δν0⋅ν 2<br />
(51)<br />
2
In further applications as potential V XC it is suitable<br />
to use the exchange-correlation pseudo-potential<br />
which contains the correlation (Gunnarsson-Lundqvist)<br />
potential and exchanger Kohn-Sham one<br />
[11]:<br />
1/3<br />
VXC ( r) = f( θ) VX( r) −0,0333⋅ ln[1+ 18,376 ⋅ρ ( r)]<br />
(52)<br />
where<br />
2 1/3<br />
VX=−(1/ π)[3 π ⋅ρ ( r)]<br />
2 1/3<br />
is the Kohn-Sham exchange potential, θ= [3 πρ ] / c ,<br />
and function f ( θ ) is some function. Using the above<br />
written formula, one can simply define the values (46),<br />
(50).<br />
Further let us give the corresponding comments<br />
regarding the constants β . First of all, it is obvi-<br />
ik<br />
ous that the terms with constants β01, β11, β12, β 22 give<br />
omitted contribution to the energy functional (at<br />
least in the zeroth approximation in comparison<br />
with others), so they can be equal to zero. The value<br />
for a constant β 00 in some degree is dependent upon<br />
the definition of the potential V XC . If as V XC it is use<br />
one of the correct exchange-correlation potentials<br />
from the standard density functional theory, then<br />
without losing a community of statement, the constant<br />
β 00 can be equal to 1. The constant β 02 can<br />
be in principle calculated by analytical way, but it is<br />
very useful to remember its connection with a spectroscopic<br />
factor Fsp of the molecular system (it is<br />
usually defined from the ionization cross-sections)<br />
[11,14]:<br />
⎧ ∂<br />
⎫<br />
Fsp = ⎨1 − ∑ kk[ −(<br />
V. I. P.)<br />
k ] ⎬ (53)<br />
⎩ ∂∈<br />
⎭<br />
The term ∂∑ / ∂ε is defined above. It is easily to<br />
understand the this definition is in fact corresponding<br />
to the pole strength of the corresponding Green’s<br />
function. It is interesting to discuss the possible<br />
analogous universality of β and the constants in the<br />
ik<br />
well-known Landau Fermi-liquid theory and Migdal<br />
finite Fermi-systems theory. Indeed, as we know<br />
now, the entire universality of the constants in the<br />
last theories is absent, though a range of its changing<br />
is quite little. Without a detailed explanation, we note<br />
here that the corresponding constants in our theory<br />
possess the same universality as ones in the Landau<br />
Fermi-liquid theory and Migdal finite Fermi-systems<br />
theory. More detailed explanation requires a careful<br />
check. Further it is obvious that omitting the energy<br />
dependence of the mass operator (i.e. supposing<br />
β 02 = 0 ) the quasiparticle density functional theory<br />
can be resulted in the standard Kohn-Sham theory.<br />
In this essence presented approach to definition of<br />
the functions basis { Φ λ}<br />
of Hamiltonian H q can be<br />
treated as an improved in comparison with similar<br />
basise’s of other one-particle representations (HF,<br />
Hatree-Fock-Slater, Kohn-Sham etc.). Naturally,<br />
this advancement can be manifested during studying<br />
those properties of the multi-electron systems, when<br />
an accurate account for the complex exchange-correlation<br />
effects, including a continuum pressure, energy<br />
dependence of a mass operator etc., is critically<br />
important.<br />
4. APPLICATION OF THE COMBINED<br />
METHOD TO DIATOMICS<br />
PHOTOELECTRON SPECTRA<br />
As an object of studying we choose the diatomic<br />
molecule of N for application of the combined<br />
2<br />
Green’s function method and quasiparticle DFT approach.<br />
The nitrogen molecule has been naturally discussed<br />
in many papers. The valence V. I. P. ‘s of N2 have been calculated [1,13,14,24] by the method of<br />
Green’s functions and therefore the pole strengths pk are known and the mean values q can be estimated. It<br />
k<br />
should be reminded that the N molecule is the clas-<br />
2<br />
sical example where the known Koopmans’ theorem<br />
even fails in reproducing the sequence of the V. I. P. ‘s<br />
in the PE spectrum. From the HF calculation of Cade<br />
et al.[24] one finds that including reorganization the<br />
V. I. P. ‘s assigned by g σ and σu improve while for π V.<br />
I. P. the good agreement between the Koopmans value<br />
and the experimental one is lost, leading to the same<br />
sequence as given by Koopmans’ theorem. Earlier<br />
[13,14] it has been shown that the nitrogen spectra can<br />
be in principle reproduced by applying a one-particle<br />
theory with account of the correlation and reorganization<br />
effects. The above-mentioned Green’s functions<br />
calculation which takes account of reorganization and<br />
correlation effects leads to the experimental sequence<br />
of V. I. P.’s. In Table 1 the experimental V. I. P. ‘s (a),<br />
the one-particle HF energies (b), the V. I. P. ‘s calculated<br />
by Koopmans’ theorem plus the contribution<br />
of reorganization (c), the V. I. P. ‘s calculated with<br />
Green’s functions method (d), the combined Green<br />
functions and DFT approach (e) and the corresponding<br />
pole strengths (d,e) are listed.<br />
Table 1<br />
The experimental and calculated V. I. P.’s (in eV) of N . R is the<br />
2 k<br />
contribution of reorganization; p stands for pole strength<br />
k<br />
Orbital<br />
Exptla V.I.P. , s - b<br />
∈k<br />
(<br />
−∈ +<br />
+ R<br />
k<br />
k<br />
c<br />
)<br />
Calc d<br />
V.I.P. , s<br />
d<br />
ρk<br />
Calc e<br />
V.I.P. , s<br />
e<br />
ρk<br />
3 σg 15,60 17,36 16,01 15,50 0,91 15,52 0,913<br />
1 πu 16,98 17,10 15,67 16,83 0,94 16,85 0,942<br />
2 σu 18,78 20,92 19,93 18,59 0,87 18,63 0,885<br />
Analysis shows that the data, obtained within the<br />
standard Green functions approach and combined<br />
method are very much close. Taking into account a<br />
simplification of the method scheme within the DFT<br />
approach, the standard Green’s function theory ( in<br />
particular the Cederbaum-Domske theory [12]) looks<br />
more attractive else. As it is known, of the three bands<br />
in the experimental low-energy spectrum of the N 2<br />
molecule ( Fig. 3), only the lπ u band exhibits a strong<br />
vibrational structure.<br />
When the change of frequency due to ionization is<br />
small, the density of states can be well approximated<br />
using only one parameter g:<br />
∞ n<br />
−s<br />
S<br />
() ∈ = ∑ δ( ∈−∈ +Δ∈ + ⋅h ω)<br />
,<br />
N e n<br />
k k k<br />
n=<br />
0 n!<br />
S g<br />
( ) 2 −<br />
2<br />
= ω<br />
h (54)<br />
125
Fig. 3. Experimental and calculated PE spectra of N 2 . The<br />
uppermost spectrum is calculated with S 0 and Eq. (54). The middle<br />
spectrum is calculated with values of S from (29) (see text and<br />
[12,14]).<br />
In case the frequencies change considerably, the<br />
intensity distribution of the most intensive lines can<br />
analogously be well approximated by an effective parameter<br />
S. In fig.1 the experimental and calculated<br />
photoelectron spectra for the N 2 .molecule are presented.<br />
The uppermost spectrum is calculated with<br />
S 0 (i.e. the constant S calculated with g 0 ) and Eq.<br />
(54). The middle spectrum is calculated with values<br />
of S from Eq. (29). It is important to note that<br />
the original Green’s functions and combined Green<br />
functions +DFT approach coincide in the scale of<br />
the figure. In a whole the agreement between the<br />
calculated spectrum (the corrected g ) and the experimental<br />
one is improved. Regarding the inclusion<br />
of the anharmonicites it should be mentioned<br />
that a theory can be generalized by means a standard<br />
normal coordinate expansion of the Hamiltonian to<br />
third and higher orders.<br />
126<br />
5. CONCLUSIONS<br />
So, firstly we present a new combined theoretical<br />
approach to vibrational structure in molecular photoelectron<br />
spectra, which is based on the Green’s function<br />
method and DFT approach. The density of states,<br />
which describe the vibrational structure in molecular<br />
photoelectron spectra, is calculated with the help of<br />
combined DFT-Green’s-functions approach. It is<br />
important that the calculational procedure is significantly<br />
simplified with using the quasiparticle DFT formalism.<br />
In result, we believe that quite simple theory<br />
become a powerful tool in interpreting the vibrational<br />
structure of the molecular photoelectron spectra.<br />
References<br />
1. Turner D., Baker C., Baker A., Brunrile C., Molecular Photoelectron<br />
Spectroscopy, New York: Wiley. — 2005. — 540P.<br />
2. Kaplan I.G., Theory of intermolecular interactions, World<br />
Sci.: Singapore, 2005. — 350P.<br />
3. Wilson S., Handbook on Molecular Physics and Quantum<br />
Chemistry. — Chichester: Wiley. — 2007. — 700P.<br />
4. Crawford T., Schaefer H.F., III, in Reviews of Computational<br />
Chemistry, eds. K. B. Lipkowitz and D.B. Boyd, New York:<br />
Wiley, 2000. — Vol.14. — P.33–136.<br />
5. Nayak M., Chaudhuri R., Das.B., Ab initio calculation<br />
of electron-nucleus scalar-pseudo-scalar interaction constant<br />
W s in heavy polar molecules//Phys.Rev.A. — 2007. —<br />
Vol.75. — P.022510.<br />
6. Bartlett R.J., Musia³ M.N., Coupled-cluster theory in quantum<br />
chemistry//Rev. Mod. Phys. — 2007. — Vol.79. — P.291-<br />
328.<br />
7. Grant I. P., Relativistic quantum theory of atoms and molecules.<br />
— N. — Y.:Springer, 2007. — 286P.<br />
8. Abrikosov A., Gorkov L., Dzyaloshinskii E., Quantum Field<br />
Theoretical Methods in Statistical Physics ,Oxford: Pergamon,<br />
1995. — 400P.<br />
9. Dorofeev D., Zon B.A., Kretinin I., Chernov V.E., Method<br />
of quantum defect Green’s function for calculation of dynamic<br />
atomic polarizabilities// Optics and Spectr. — 2005. —<br />
Vol. 99. — P.540-548.<br />
10. Zheng R., Wei W., Shi Q., Density functional theory study on<br />
sum-frequency vibrational spectroscopy of arabinose chiral<br />
solutions// J. Phys. Chem. A. — 2009. — Vol.113. — P.157-<br />
164.<br />
11. Glushkov A.V., Relativistic quantum theory. Quantum mechanics<br />
of atomic systems. — Odessa: Astroprint, 2008. — Ð<br />
800.<br />
12. Köppel H., Domcke W., Cederbaum L.S., Green’s function<br />
method in quantum chemistry// Adv. Chem. Phys. —<br />
1999. — Vol.57. — P.59-132.<br />
13. Cederbaum L.S., Domcke W., On vibrational structure of<br />
photoelectron spectra by the Green’s functions method//<br />
J.Chem. Phys. — 1994. — Vol.60. — P.2878-2896.<br />
14. Glushkov A.V., New approach to theoretical definition of<br />
ionization potentials for molecules on the basis of Green’s<br />
function method//Journ.of Phys.Chem. — 1999-Vol.66. —<br />
P.2671-2677.<br />
15. Kohn W., Sham L.J. Quantum density oscillations in an<br />
inhomogeneous electron gas//Phys. Rev. A. — 1999. —<br />
Vol.137,N6. — P.1697–1706.<br />
16. Gross E.G., Kohn W. Exchange-correlation functionals in<br />
density functional theory. — N-Y: Plenum, 2005. — 380P.;<br />
17. The Fundamentals of Electron Density, Density Matrix and<br />
Density Functional Theory in Atoms, Molecules and the<br />
Solid State, Series: Progress in Theoretical Chemistry and<br />
Physics, Eds. Gidopoulos N.I. and Wilson S. — Amsterdam:<br />
Springer, 2004. — Vol.14, 244P.<br />
18. Glushkov A.V., Relativistic and correlation effects in spectra<br />
of atomic systems.. — Odessa: Astroprint, 2006. — Ð 400.<br />
19. Glushkov A.V., Khetselius O.Yu., Malinovskaya S.V., Optics<br />
and spectroscopy of cooperative laser-electron nuclear
processes in atomic and molecular systems — new trend<br />
in quantum optics// Europ.Phys.Journ. ST. — 2008. —<br />
Vol.160,N1. — P.195-204.<br />
20. Glushkov A.V., Khetselius O.Yu., Malinovskaya S.V., Spectroscopy<br />
of cooperative laser-electron nuclear effects in multiatomic<br />
molecules// Molec. Physics . — 2008. — Vol.106. —<br />
N9&10. — P.1257-1260.<br />
21. Glushkov A.V., Malinovskaya S.V., et al, Quantum stochastic<br />
modeling energy transfer and effect of rotational, V-T<br />
relaxation on multi-photon excitation and dissociation for<br />
molecules// Int. Journ. of Quantum Chemistry. — 2005. —<br />
Vol.105-P.512-516.<br />
22. Glushkov A.V., Malinovskaya S.V., Lovett L. et al, Green’s<br />
function method in quantum chemistry: New algorithm for<br />
the Dirac equation with complex energy and Fermi-model<br />
nuclear potential// Int. Journ. of Quantum Chemistry. —<br />
2009. — Vol.109. — N10. — P.1331-1345.<br />
UDÑ 539.19+539.182<br />
A. P. Fedchuk, A. V. Glushkov, Ya. I. Lepikh, A. V. Loboda<br />
23. Glushkov A.V., Florko T.A., Lovett L. et al, Gauge-invariant<br />
QED perturbation theory approach to nuclear electric quadrupole<br />
moments, hyperfine structure constants for heavy atoms<br />
and ions// Frontiers in Quantum Systems in Chemistry<br />
and Physics . — 2008. — Vol.18. — P.505-522.<br />
24. Appel H., Gross E.K.U., Time-dependent natural orbitals and<br />
occupation numbers//Phys. Rev. Lett. — 2009. — Vol.102. —<br />
P.024331.<br />
25. Glushkov A.V., Lepikh Ya.I., Fedchuk A.P., Loboda A.V.,<br />
Lovett L., Electrodynamical and quantum — chemical approaches<br />
to modelling the electrochemical and catalytic processes<br />
on metals, metal alloys and semiconductors//Int. Journ.<br />
of Quantum Chemistry. — 2009. — Vol.111, N4, to be printed.<br />
26. Ivanova E.P., Ivanov L.N., Glushkov A.V., Kramida A.E.,<br />
High-order corrections in the relativistic perturbation theory<br />
with model zeroth approximation//Physica Scripta. —<br />
1999. — V.32. — P.512-524.<br />
THE GREEN’S FUNCTIONS AND DENSITY FUNCTIONAL APPROACH TO VIBRATIONAL STRUCTURE IN THE<br />
PHOTOELECTRON SPECTRA OF MOLECULES<br />
Abstract<br />
We present the basis’s of the new combined theoretical approach to vibrational structure in photoelectron spectra of molecules. The<br />
approach is based on the Green’s function method and density functional theory (DFT). The density of states, which describe the vibrational<br />
structure in photoelectron spectra, is defined with the use of combined DFT-Green’s-functions approach and is well approximated<br />
by using only first order coupling constants in the one-particle approximation. It is important that the calculational procedure is<br />
significantly simplified with using the DFT.<br />
Key words: photoelectron spectra, Green’s functions method, density functional theory.<br />
ÓÄÊ 539.19+539.182<br />
A. Â. Ãëóøêîâ, ß. È. Ëåïèõ, A. Ï. Ôåä÷óê, A. Â. Ëîáîäà<br />
ÌÅÒÎÄ ÔÓÍÊÖÈÉ ÃÐÈÍÀ È ÔÓÍÊÖÈÎÍÀËÀ ÏËÎÒÍÎÑÒÈ Â ÎÏÐÅÄÅËÅÍÈÈ ÊÎËÅÁÀÒÅËÜÍÎÉ<br />
ÑÒÐÓÊÒÓÐÛ ÔÎÒÎÝËÅÊÒÐÎÍÍÛÕ ÑÏÅÊÒÐÎÂ ÌÎËÅÊÓË<br />
Ðåçþìå<br />
Èçëîæåíû îñíîâû íîâîãî êîìáèíèðîâàííîãî òåîðåòè÷åñêîãî ìåòîäà îïèñàíèÿ êîëåáàòåëüíîé ñòðóêòóðû äëÿ ôîòîýëåêòðîííûõ<br />
ñïåêòðîâ ìîëåêóë, êîòîðûé áàçèðóåòñÿ íà ìåòîäå ôóíêöèé Ãðèíà è òåîðèè ôóíêöèîíàëà ïëîòíîñòè (ÒÔÏ). Ïëîòíîñòü<br />
ñîñòîÿíèé, îïèñûâàþùàÿ êîëåáàòåëüíóþ ñòðóêòóðó ñïåêòðà, îïðåäåëÿåòñÿ ñ èñïîëüçîâàíèåì ìåòîäà ôóíêöèé Ãðèíà<br />
è êâàçè÷àñòè÷íîé ÒÔÏ è ïðèåìëåìî àïïðîêñèìèðóåòñÿ ñ èñïîëüçîâàíèåì êîíñòàíò ñâÿçè òîëüêî ïåðâîãî ïîðÿäêà óæå â<br />
îäíî÷àñòè÷íîì ïðèáëèæåíèè.<br />
Êëþ÷åâûå ñëîâà: ôoòoýëåêòðîííûé ñïåêòð, ìåòîä ôóíêöèé Ãðèíà, òåîðèÿ ôóíêöèîíàëà ïëîòíîñòè.<br />
ÓÄÊ 539.19+539.182<br />
Î. Â. Ãëóøêîâ, ß. ². Ëåï³õ, Î. Ï. Ôåä÷óê, A. Â. Ëîáîäà<br />
ÌÅÒÎÄ ÔÓÍÊÖ²É ÃвÍÀ ² ÔÓÍÊÖ²ÎÍÀËÓ ÃÓÑÒÈÍÈ Ó ÂÈÇÍÀ×ÅÍͲ ²ÁÐÀÖ²ÉÍί ÑÒÐÓÊÒÓÐÈ<br />
ÔÎÒÎÅËÅÊÒÐÎÍÍÈÕ ÑÏÅÊÒв ÌÎËÅÊÓË<br />
Ðåçþìå<br />
Âèêëàäåí³ îñíîâè íîâîãî êîìá³íîâàíîãî òåîðåòè÷íîãî ìåòîäó îïèñó â³áðàö³éíî¿ ñòðóêòóðè äëÿ ôîòîåëåêòðîííèõ ñïåêòð³â<br />
ìîëåêóë, ÿêèé áàçóºòüñÿ íà ìåòîä³ ôóíêö³é Ãð³íà ³ òåî𳿠ôóíêö³îíàëó ãóñòèíè (ÒÔÃ). Ãóñòèíà ñòàí³â, ùî îïèñóº â³áðàö³éíó<br />
ñòðóêòóðó ñïåêòðà, âèçíà÷àºòüñÿ ç âèêîðèñòàííÿì ìåòîäó ôóíêö³é Ãð³íà òà êâàç³÷àñòèíêîâî¿ ÒÔà ³ ïðèéíÿòíî àïðîêñèìóºòüñÿ<br />
ç âèêîðèñòàííÿì ñòàëèõ çâ’ÿçêó ò³ëüêè ïåðøîãî ïîðÿäêó íàâ³òü ó îäíî÷àñòèíêîâîìó íàáëèæåíí³.<br />
Êëþ÷îâ³ ñëîâà: ôoòoåëåêòðîííèé ñïåêòð, ìåòîä ôóíêö³é Ãð³íà, òåîð³ÿ ôóíêö³îíàëà ãóñòèíè.<br />
127
128<br />
UDC 621.315.592<br />
A. V. TYURIN, A. YU. POPOV, S. A. ZHUKOV, YU. N. BERCOV<br />
Scientific Research Institute of Physics, I.I. Mechnikov National Odesa University,<br />
27 Paster str., 27, Ukraine, 65082 E-mail: zhukov@onu.edu.ua, bercov@gmail.com<br />
MECHANISM OF SPECTRAL SENSITIZING OF THE EMULSION<br />
CONTAINING HETEROPHASE “CORE –SHELL” MICROSYSTEMS<br />
The new approach to the process of spectral sensitizing of emulsions created on base of heterophase<br />
microcrystals of “non- photosensitive core — photosensitive silver-haloid shell” structure is<br />
offered. The distinctive feature of the given system is a possibility of sensitizer dye introduction on the<br />
“core-shell” border. Considering such spatial separation of dye adsorbed on a core by the shell of silver<br />
halide, the mechanism of sensitization, which provides the expansion of emulsion spectral sensitivity<br />
area, is offered.<br />
INTRODUCTION<br />
Holographic emulsions based on silver halide till<br />
now are the most sensitive among all recording media<br />
and it attracts the permanent interest to them. These<br />
emulsions are suitable for recording of the reflecting<br />
and transmitting transparent volume holograms possessing<br />
both high diffraction efficiency, and high angular<br />
and spectral selectivity [1,2]. Nevertheless, in<br />
this sphere till now there are problems demanding the<br />
solution, one of which is expansion of emulsion spectral<br />
sensitivity range. In the given work the new approach<br />
to spectral sensitization of holographic emulsions<br />
is considered, which allows to expand a range<br />
of their spectral sensitivity considerably and to obtain<br />
single-layer emulsions, suitable for color and infrared<br />
holography.<br />
PROBLEM STATEMENT<br />
In some cases, for example for recording of color<br />
holograms it is necessary to expand emulsion spectral<br />
sensitivity to all visible range. Moreover, for the solution<br />
of some technical problems it is necessary to have<br />
emulsion, which is sensitive to infrared (IR) area of<br />
spectrum. The latter problem is solved by introduction<br />
of the dyes absorbing light in near infrared area of a<br />
spectrum with generation of non-equilibrium charge<br />
carriers. However, used dyes are unstable even during<br />
storage at low temperatures, and they decompose to<br />
components, therefore such emulsions are insensitive<br />
and tend to fogging and degradation of the latent image.<br />
Besides in usual AgHal emulsions, especially at<br />
exceeded concentration of dye, the IR sensitivity decreases<br />
because of so-called dyes self-desensitizing<br />
phenomenon of the second kind [3]. In this case, illumination<br />
quanta interaction with dye molecules excited<br />
by light with non- excited ones happens, and as<br />
the result, the luminescence of the nearest dye appears<br />
instead of generation of the free charge carrier. The<br />
solving of the problem could be achieved through spatial<br />
division of dyes interacting this way.<br />
The similar situation arises also while using two or<br />
more dyes absorbing light in different areas of a spec-<br />
trum. They also can interact with each other and desensitize<br />
the emulsion. For solution of this problem,<br />
the multilayered emulsions are used, each layer of<br />
which is sensitized by its dye, but such solution in the<br />
case of holographic emulsion is not optimal.<br />
To solve both problems described above, it is possible<br />
to apply the model of emulsions based on heterophase<br />
micro-crystals of “non-photosensitive core —<br />
photosensitive silver-halide shell” structure [4].<br />
Due to special features of hetero-phase microcrystals<br />
structure the dyes — sensitizers could be absorbed<br />
not only on an external surface of silver-halide<br />
shell (as it is done traditionally), but also on the internal<br />
surface on the border core — shell surface [5],<br />
what allows to solve a problem of spatial separation of<br />
various dyes, and to increase their storage stability as<br />
well.<br />
The goal of the given work is to study the features<br />
of the described above spectral sensitizing of photosensitive<br />
hetero-phase micro-crystals emulsions (including<br />
the holographic ones), and to create the working<br />
model of the electron-hole processes occurring in<br />
the emulsion when the latent image is created.<br />
MATERIALS AND METHODS<br />
General scheme of synthesis process of emulsion<br />
with hetero-phase micro-crystals “core CaF 2 — AgBr<br />
shell” is described in detail in [5]. Depending on purpose<br />
of emulsion the average size of microcrystals in<br />
them is:<br />
~0.35 microns — emulsion intended for the photographic<br />
purposes (“photographic” emulsion); ~0.05<br />
microns — emulsion intended for record of holograms<br />
(“holographic” emulsion).<br />
For spectral sensitization of such emulsions we<br />
used three various dyes:<br />
I — natrium salt of 3,3 3,3’-di-γ-sulfopropyl-1,1’diethyl-5,5’-dicarboethoximid-carbocyaninebetaine<br />
II — pyridine salt of 3,3’-di-γ-sulfopropyl-9-ethyl-4,5,4’,5’-dibenzothiacarbocyaninebetaine<br />
III — 1,1 ‘-diethyl-quini-2,2 ‘-cyaniniodide.<br />
Infusion of such dyes in emulsion occurs as in molecular<br />
(Ì) and aggregated phases (J and H aggregates<br />
[6]). Wavelengths of absorption maxima of molecular<br />
© A. V. Tyurin, A. Yu. Popov, S. A. Zhukov, Yu. N. Bercov, 2009
M and aggregated J and H phases of dyes I-III adsorbed<br />
on a surface of micro-crystals in halogen-silver<br />
emulsion are specified in the Table.<br />
Table<br />
Dye Í-band (nm) M-band(nm) J-band (nm)<br />
I 508 519 -<br />
II - 630 680-690<br />
III - 545 578<br />
It has to be marked that the long-wave part of absorption<br />
band of J-aggregates of dyes often extends up<br />
to 900 nm.<br />
Spectral sensitization of hetero-phase emulsions<br />
using marked dyes was carried out stage by stage. Therefore,<br />
for experimental research three types of sensitized<br />
emulsion were prepared (symbolically a,b,c):<br />
à) emulsion with hetero-phase micro-crystals “<br />
CaF 2 core — AgBr shell ”, prepared under the procedure<br />
described in work [5], was sensitized with dyes II<br />
or III in concentration 10 -4 mole of dye for mole CaF 2 ,<br />
by traditional way (control emulsion);<br />
b) to suspension containing homogeneous calcium<br />
fluoride particles of given size we added 10 -4 mole<br />
of dye II or III for mole on ìîëü CaF 2 . These conditions<br />
provided practically full adsorption of dye on a<br />
surface of core CaF 2 . Then an AgBr shell was grown<br />
up on core CaF 2 with adsorbed dye II or III under the<br />
procedure [5];<br />
c) “b” emulsion but additionally sensitized with<br />
the dye I in concentration 10 -4 mole of dye for mole<br />
CaF 2 ;<br />
For developing of obtained emulsions surface and<br />
deep developers [7] were used. Such ways of developing<br />
allow us to observe position of the centers of the<br />
latent image all over volume of halogen-silver shell<br />
of composite system. Surface developer –only those<br />
hetero-phase micro-crystals in which centers of the<br />
latent image are on external surface of AgBr shell can<br />
be developed. Deep developer provides partial dissolution<br />
of AgBr shell top layers. So thus we can develop<br />
those hetero-phase micro-crystals in which centers of<br />
the latent image are in volume of AgBr shell up to the<br />
surface of non-photosensitive core.<br />
RESULTS<br />
On Fig. 1 results of the spectral sensitometric tests<br />
are presented, which are concerned to three mentioned<br />
above types (“a”, “b”, “c”) of “photographic” emulsions,<br />
containing dye II. Concentration of dye was<br />
high enough for the existence in emulsion not only a<br />
molecular phases of dye, but also its J-aggregates. Exposure<br />
and processing of layers of the specified types<br />
of emulsions was carried out in spectral-sensitometer<br />
ISP-3 with the help of a standard technique [8].<br />
As it follows from the data submitted on Fig. 1.A,<br />
for exposed emulsion “a” optical density after processing<br />
with the help of both surface and deep developers<br />
(curves 1, 1’) in comparison with non-exposed emulsion<br />
“a” (curve 3) are practically equal, both in the<br />
absorption area of molecular dye II (λ max = 630 nm)<br />
and in the absorption area of the J-aggregate of dye II<br />
(λ max = 690 nm).<br />
° — developed in surface developer<br />
• — developed in deep developer<br />
Fig. 1. A. Spectral distribution of optical density (D) of exposed<br />
layers of “photographic” emulsions “à” (1,1’) and “b” (2,2 ‘)<br />
containing the dye II after development in surface (1,2) and deep<br />
(1’, 2’) developers. Bar line (3) marks the fog optical density of<br />
non-exposed and developed samples of emulsion “à” and “b” both<br />
in surface and deep developers. B. Spectral distribution of optical<br />
density (D-D ) of exposed layers of “photographic” emulsion “c”<br />
0<br />
after developing in surface (1) and deep (1 ‘) developers, D — fog<br />
0<br />
optical density of non-exposed and developed both in surface and<br />
deep developers of emulsion samples “c”.<br />
For the exposed emulsion “b” developed with surface<br />
developer (curve 2) in comparison with non-exposed<br />
emulsion “b” (curve 3), on wavelengths λ > 600<br />
nm, the reduction of optical density (enlightenment)<br />
occurs and main part of this enlightenment falls on<br />
the absorption area of J-aggregate of dye II (λ max = 690<br />
nm). Application of deep developer in a case of exposed<br />
emulsion “a” (curve 2’), in comparison with non-exposed<br />
emulsion after processing with deep developer<br />
(curve 3), shows an increase of optical density in the absorption<br />
band of both molecular dye II (band M, λ max =<br />
630 nm) and in the absorption area of J-aggregate of<br />
dye II (band J, λ max = 690 nm). However, if for emulsion<br />
“a” this increase occurs almost equally both in a band<br />
M and in band J, in case of emulsion “b” an increase of<br />
absorption in band J occurs in the much greater degree<br />
than in band M. If we compare between each other exposed<br />
emulsions “a” and “b” with sensitized dye II, in a<br />
case of emulsion “b”, the expansion of a spectral range<br />
of a photosensitivity in long-wave area (see curves 1 and<br />
2’ in long-wave area) is observed.<br />
For “photographic” emulsion “c” Fig. 1 B, after<br />
processing in surface developer (curve 1) the increase<br />
of the optical density in the absorption field of a molecular<br />
band of dye I (λ max = 519 nm) is observed as well<br />
as small increase of the optical density in the absorption<br />
field of dye II both in molecular (λ max = 630 nm)<br />
and in J-aggregated (λ max = 690 nm) phases. Application<br />
of deep developer results in reduction of absorption<br />
in M band of dye I and essential increase of optical<br />
density in absorption band of J-aggregates of dye II<br />
(λ max = 690 nm) (curve 1’).<br />
The same spectral sensitometric tests were carried<br />
out for “holographic” emulsions (types “à” and “b”).<br />
Experimental data of such researches are presented at<br />
Fig. 2.<br />
129
Fig. 2 A. Spectral distribution of optical density (D-D ) of ex-<br />
0<br />
posed layers of “holographic” emulsions “à” (1) and “b” (1’) with<br />
dye II after development in deep developer. B. Spectral distribution<br />
of optical density (D-D ) of exposed layers of “holographic”<br />
0<br />
emulsion “à” (1) and “b” (1’) with dye III after development in<br />
deep developer<br />
D is the optical density of a fog of non-exposed<br />
0<br />
and developed in deep developer emulsion samples<br />
“a” and “b”.<br />
From Fig. 2 one could see, that the sensitization<br />
both with dye II and dye III of “holographic” emulsion<br />
of “b” type in comparison with emulsions of “a”<br />
130<br />
type sensitized by the same dyes results in expansion of<br />
emulsion sensitivity spectral area in long-wave part of a<br />
spectrum. If for dye II this displacement is insignificant,<br />
for dye III it is essential. It should be mentioned that for<br />
“photographic” emulsion of “b” type the expansion of<br />
the spectral sensitivity in long-wave region in comparison<br />
with to emulsion of “a” type is also observed.<br />
Observed spectral sensitivity displacement both<br />
for “photographic” and “holographic” emulsions occurs<br />
up to 900 nm and it is natural to assume that it is<br />
caused by absorption of J-aggregates of dye. The reason<br />
of this effect (in case of emulsion of type “b”) is<br />
the greater efficacy of generated in J-aggregates nonequilibrium<br />
carriers of charge use for latent image<br />
creating. It could occur due to the spatial division of<br />
molecular dye and J-aggregates of dye by AgBr cover<br />
because molecular dye is basically adsorbed on external<br />
surface of a cover and J-aggregate of dye — on<br />
internal surface. Spatial division of various phases of<br />
dye eliminates interaction at photo-excitation of the<br />
J-aggregate, which leads only to a recombination luminescence<br />
of molecular dye, instead of charge free<br />
carriers generation in silver halogenide [9].<br />
Increase of CaF 2 dye concentration adsorbed on<br />
core at spatial division of the molecular and aggregated<br />
dye by AgBr cover is accompanied, besides the<br />
expansion of spectral sensitivity area, by decrease of<br />
fog level on the cover external surface (see Fig. 3).<br />
Fig. 3 Spectral sensitograms of emulsions developed by deep developer containing dye II in different concentrations (a mole dye<br />
/mole AgBr): À-2∙10 -5 ; B — 4∙10 -5 ; C — 6∙10 -5<br />
Seen fogging decrease on an outer surface of AgBr<br />
shell results from reaction [10]<br />
0 + Ag + p → Agm → Agm-1<br />
m<br />
0Ag + 0 + → Ag + Agi . (1)<br />
m-1<br />
Occurrence of the holes on an outer surface of a<br />
shell is caused by the following reason. Adsorption of<br />
J-aggregate of dye II on an interior surface of a shell is<br />
accompanied by reaction<br />
+ 0 0 + Ag + J → Agn + J (2).<br />
n<br />
As it has been established by us in [11], the main<br />
level of the J-aggregate of dye II lays below the level of<br />
a valence band top of AgBr, therefore the hole is localized<br />
on J + transfers in valence band AgBr<br />
J + → J0 + p (3),<br />
Then the hole in valence band migrates to an outer<br />
surface of AgBr shell and provides the reaction (1).<br />
To prove our assumption, we carried out the lowtemperature<br />
research of the luminescence of these<br />
emulsions. On Fig. 4, one could see the spectra of<br />
low-temperature (Ò = 77 Ê) luminescence and excitation<br />
of a luminescence not only in “holographic”<br />
emulsions of type “b” but also as an intermediate stage<br />
of its preparation: — before we cover CaF core with<br />
2
adsorbed on them dye II with a AgBr shell. These luminescent<br />
studies were performed under such conditions:<br />
time of sample excitation and time of its luminescence<br />
registration are equal, and they equal to 10 -4<br />
s and dark interval is 1.1∙10 -3 s at the modulation frequency<br />
of 400 Hz.<br />
As it follows from the mentioned luminescent<br />
data, the phosphorescence of molecular dye II (λ max =<br />
800 nm) (curve 1) is excited by light not from the absorption<br />
area of molecular dye (λ max = 630 nm), but<br />
from absorption area of J-aggregate of dye II (λ max =<br />
630 nm) (curve 1’). This fact proves the presence of<br />
the interaction between excited dyes, which result in<br />
emulsion desensitization. After AgBr shell is created,<br />
the phosphorescence of molecular dye II under excitation<br />
of the J-aggregate of dye disappears, and there<br />
emerges a luminescence with λ max = 700 nm (curve 2).<br />
Excitation of this luminescence (curve 2’) is both<br />
caused by absorption AgBr of a shell (λ max = 430-450<br />
nm) and absorption of molecular dye (λ max = 630 nm)<br />
as the excitation of the luminescence from absorption<br />
area of the J-aggregate of dye II is absent.<br />
Fig. 4. Spectra of low-temperature Ò 77 K luminescence (À)<br />
and excitation of luminescence (B) of “holographic” emulsion of<br />
“b” type at the intermediate stage of its preparation: curves 1,1’ —<br />
when we do not create AgBr shell on CaF core with adsorbed on<br />
2<br />
them dye II; curves 2,2 ’ — “holographic” emulsion of type “b”<br />
with dye II<br />
Spectra of a luminescence (À) were obtained at<br />
the excitation with λ = 690 nm — curve 1 and with<br />
λ = 450, 630 nm — curve 2; Spectra of excitation (B)<br />
were given for a luminescence at λ = 800 nm — curve<br />
1’ and on λ = 700 nm — curve 2’.<br />
As after AgBr shell creation, the luminescence of<br />
molecular dye at excitation of the J-aggregate of dye<br />
has disappeared, hence, it testifies that self- desensitizing<br />
action of dye caused by their interaction is eliminated<br />
and thus, it provides the significant displacement<br />
of spectral sensitivity of emulsion of type “b” in<br />
the long-wave area.<br />
DISCUSSION<br />
Earlier in [12], it was shown the basic opportunity<br />
of covering of the dye adsorbed on non-silver CaF 2<br />
core by AgBr shell without specification of a aggregate<br />
phases of dye and its role in the process of an spectral<br />
sensitization. The comparison of the luminescent and<br />
sensitometric results, proves the existence of interaction<br />
between the molecular and J-aggregated dye on<br />
core CaF 2 . For a case of hetero-phase micro-crystal,<br />
we assumed that the shell divides interaction among<br />
the molecular and aggregated phases of dye that results<br />
in observable expansion of area of spectral sensitivity of<br />
various composite of the system.<br />
It should also be noted that for the case of holographic<br />
emulsions with hetero-phase micro-crystals<br />
spatial separation of the molecular and aggregated<br />
dye by shell contributes to the effective separation of<br />
photo-excited non-equilibrium charge carriers. Separation<br />
of non-equilibrium charge carriers, in the case<br />
of composite system, could be illustrated by occurrence<br />
of a enlightenment in spectral area λ > 600 nm<br />
for emulsion “b” , when we develop it in the surface<br />
developer (Fig.1, curve 2). Observed enlightenment<br />
testifies that the quantity of the centers of the latent<br />
image located on AgBr shell surface of emulsion “b”<br />
micro-crystals after the exposure and development in<br />
surface developer, decreases in comparison with nonexposed<br />
emulsion. The greatest reduction takes place<br />
when illumination is made in an absorption band of<br />
J-aggregates of dye. As to our opinion, such reduction<br />
could proceed under the schemes, offered in reactions<br />
(1) — (3). As far as we have already determined, Jaggregates<br />
of dye are located basically on an internal<br />
surface of AgBr shell adjoining micro-crystals’ core,<br />
when absorbing the light by J-aggregates, the generated<br />
holes migrate through a AgBr shell to its surface<br />
and there they result in destruction (reduction, neutralization)<br />
of the centers of the latent image, i.e. in<br />
“enlightenment”. The AgBr shell presence provides<br />
translation of free holes of the J-aggregate of dye,<br />
adsorbed on core CaF 2 through all thickness of the<br />
shell, leads to reduction of a fogging and causes the<br />
substantial increase of the diffraction efficacy of such<br />
emulsions. It could be used for the creation of “direct<br />
positive” images.<br />
The ability of hetero-phase micro-crystals halogen<br />
silver shell to separate interacting phases of the<br />
dye can be used for spatial separation of different sensitizer<br />
dyes, which application earlier was complicated<br />
because of their interaction resulting in emulsion desensitization.<br />
Such dyes’ separation allows to replace<br />
the multilayered emulsions with single-layered emulsions,<br />
which spectral sensitivity is determined by the<br />
dyes, which are located on different surfaces of a AgBr<br />
shell without desensitization effect.<br />
CONCLUSIONS<br />
As a result of the studies of the newly modified<br />
emulsions and discussion of the e3xperimentral results<br />
we can state the follows:<br />
1. The spectral sensitization of composite system<br />
“non-photosensitive core — halogen silver shell” allows<br />
adsorbing the sensitizer dye not only on external,<br />
but also on an internal surface of halogen silver shell.<br />
2. It is determined, that halogen silver shell effectively<br />
separates the aggregated and interacting molecular<br />
dye phases (aggregated dye on non-photosensitive<br />
core and molecular dye on a surface of halogen silver<br />
shell), that eliminates effect of dye self-desensitization<br />
and results in expansion of area of emulsions with hetero-phase<br />
micro-crystals spectral sensitization.<br />
131
3. Dyes’ spatial separation allows to replace the<br />
multilayer emulsions with the single layer one with the<br />
spectral sensitivity determined by dyes at different surfaces<br />
of AgBr shell without desensitization effect.<br />
References<br />
1. Á.È. Øàïèðî Òåîðåòè÷åñêèå íà÷àëà ôîòîãðàôè÷åñêîãî<br />
ïðîöåññà. Ì.: Ýäèòîðèàë ÓÐÑÑ. 2000, 209 ñ.<br />
2. Ïàòåíò 27837 Óêðà¿íè Ñïîñ³á âèãîòîâëåííÿ ãàëîãåíñð³áíî¿<br />
ôîòîãðàô³÷íî¿ åìóëüñ³¿. Â.Ì. Á³ëîóñ, Ä.Ã. ͳæíåð,<br />
Â.Ï. ×óðàøîâ.<br />
3. Â.Ì. Áåëîóñ, Ë.È. Ìàí÷åíêî, À.Þ. Ïîïîâ, À.Â. Òþðèí,<br />
Â.Ï. ×óðàøîâ, Þ.Á. Øóãàéëî Ôîòîãðàôè÷åñêàÿ ýìóëüñèÿ<br />
ñ ãåòåðîôàçíûìè ìèêðîêðèñòàëëàìè — íîâàÿ ñðåäà<br />
äëÿ çàïèñè ãëóáîêèõ òðåõìåðíûõ ïðîïóñêàþùèõ ãîëîãðàìì//<br />
Îïòèêà è ñïåêòðîñêîïèÿ. 1999, ò.86, ¹2, ñ. 344-<br />
348.<br />
4. À.Þ. Ïîïîâ, Ë.È. Ìàí÷åíêî, À.Â. Òþðèí, Þ.Á. Øóãàéëî<br />
Ðåãèñòðàöèÿ è âîñïðîèçâåäåíèå ñâåòîâûõ ïó÷êîâ<br />
ñ òîïîëîãè÷åñêèìè äåôåêòàìè // Ñáîðíèê “Ôîòîýëåêòðîíèêà”<br />
âûï.9, Îäåññà 2000, “Àñòðîïðèíò”, c.126-129.<br />
5. D.G. Nizhner, V.M. Belous, V.P. Churashov Preparation and<br />
Properties of Photographic Emulsions with Heterophase Microcrystals<br />
Comprising Nonsilver Cores and Silver Halide<br />
Shells// Imaging Science and Technology. 1999, v. 39, N 1,<br />
p. 56-66.<br />
132<br />
UDC 621.315.592<br />
Tyurin A. V., Popov A. Yu., S. A. Zhukov, Yu.N.Bercov<br />
6. F. Dietz Zum stand der Theorie der spektralen Sensibilisierung<br />
//J. für Signalsufzeichnungsmaterialien. 1998, bd. 6, N 4,<br />
s. 245-266; N5, s. 341-361.<br />
7. Ê.Â. ×èáèñîâ Ôîòîãðàôè÷åñêîå ïðîÿâëåíèå. Ìîñêâà.<br />
1999. “Íàóêà”.Ñ 208<br />
8. Þ.Í. Ãîðîõîâñêèé Ñïåêòðàëüíûå èññëåäîâàíèÿ ôîòîãðàôè÷åñêîãî<br />
ïðîöåññà. Ìîñêâà. 1960. Èçä. ôèç. — ìàò.<br />
ëèòåðàòóðû. Ñ 390<br />
9. À.Â. Òþðèí, Â.Ï. ×óðàøîâ, Ñ.À. Æóêîâ, Ë.È. Ìàí÷åíêî,<br />
Ò.Ô. Ëåâèöêàÿ, Î.È. Ñâèðèäîâà Âçàèìîäåéñòâèå ìîëåêóëÿðíûõ<br />
è ïîëèìîëåêóëÿðíûõ ôîðì êðàñèòåëÿ// Îïòèêà<br />
è ñïåêòðîñêîïèÿ. 2008, ò. 104, ¹ 1, ñ. 97-103.<br />
10. À.Â. Òþðèí, Â.Ï. ×óðàøîâ, Ñ.À. Æóêîâ, Î.Â. Ïàâëîâà<br />
Ìåõàíèçì àíòèñòîêñîâîé ëþìèíåñöåíöèè ãàëîãåíñåðåáðÿíîé<br />
ýìóëüñèè ñåíñèáèëèçèðîâàííîé êðàñèòåëåì//<br />
Îïòèêà è ñïåêòðîñêîïèÿ. 2008, ò. 104, ¹ 2, ñ.237-244.<br />
10. James,T.H. The Theory of the Photographic Process. // 1997,<br />
New York: Editorial Macmillan, 456 pp.<br />
11. Â.Ì. Áåëîóñ, Ä.Ã. Íèæíåð, Í.À. Îðëîâñêàÿ, Ï.Ã. Õåðñîíñêàÿ<br />
Çàðàùèâàíèå àäñîðáèðîâàííîãî êðàñèòåëÿ îáîëî÷êîé<br />
ãàëîãåíèäà ñåðåáðà â ãåòåðîôàçíûõ ìèêðîêðèñòàëëàõ//<br />
Æóðí. íàó÷í. è ïðèêë. ôîòî- è êèíåìàòîãðàôèè.<br />
1999, ò. 34, ¹ 4, ñ. 299-301.<br />
12. Belous V.M., Nizhner D.G., Orlovskaya N.A., Khersonskaya<br />
P.G. Growth of the Adsorbed Dye by Halogen-Silver Core in<br />
Hetero-Phase Micro-crystals//Journal of Scientific and Applied<br />
Photo-and Cinematography 1999. v. 34. ¹ 4. p. 299<br />
MECHANISM OF SPECTRAL SENSITISATION OF EMULSION CONTAINING HETEROPHASE “CORE –SHELL”<br />
MICROSYSTEMS<br />
The new approach to the process of spectral sensitization of emulsions created on a base of hetero-phase micro-crystals of “non-<br />
photosensitive core — photosensitive silver-halide shell” structure is offered. Distinctive feature of the given systems is a possibility of<br />
sensitizer dye introduction to “core-shell” border. Considering such spatial separation of dye adsorbed on a core by the shell of silver<br />
halide, the mechanism of sensitization which provides the expansion of emulsion spectral sensitivity area is offered.<br />
Key words: sensitisation, heterophase microsystems, emulsion.<br />
ÓÄÊ 621.315.592<br />
À. Â. Òþðèí, À. Þ. Ïîïîâ, Ñ. À. Æóêîâ, Þ. Í. Áåðêîâ<br />
ÌÅÕÀÍÈÇÌ ÑÏÅÊÒÐÀËÜÍÎÉ ÑÅÍÑÈÁÈËÈÇÀÖÈÈ ÝÌÓËÜÑÈÈ, ÑÎÄÅÐÆÀÙÅÉ ÃÅÒÅÐÎÔÀÇÍÛÅ<br />
ÌÈÊÐÎÑÈÑÒÅÌÛ “ßÄÐÎ — ÎÁÎËÎ×ÊÀ”<br />
Ïðåäëîæåí íîâûé ïîäõîä ê ïðîöåññó ñïåêòðàëüíîé ñåíñèáèëèçàöèè ýìóëüñèé, ñîçäàííûõ íà îñíîâå ãåòåðîôàçíûõ ìèêðîêðèñòàëëîâ<br />
ñîñòàâà “íåñâåòî÷óâñòâèòåëüíîå ÿäðî — ñâåòî÷óâñòâèòåëüíàÿ ñåðåáðÿíî-ãàëîèäíàÿ îáîëî÷êà”. Îòëè÷èòåëüíîé<br />
òåõíîëîãè÷åñêîé îñîáåííîñòüþ äàííîãî ïðîöåññà ÿâëÿåòñÿ ââåäåíèå êðàñèòåëÿ ñåíñèáèëèçàòîðà íà ãðàíèöó ðàçäåëà “ÿäðîîáîëî÷êà”.<br />
Ñ ó÷åòîì ýòîé îñîáåííîñòè — ïðîñòðàíñòâåííîãî îòäåëåíèÿ îáîëî÷êîé ãàëîãåíèäà ñåðåáðà êðàñèòåëÿ àäñîðáèðîâàííîãî<br />
íà ÿäðå — ïðåäëîæåí ìåõàíèçì ñåíñèáèëèçàöèè, îáåñïå÷èâàþùèé ðàñøèðåíèå îáëàñòè ñïåêòðàëüíîé ÷óâñòâèòåëüíîñòè<br />
ýìóëüñèè.<br />
Êëþ÷åâûå ñëîâà: ñåíñèáèëèçàöèÿ ýìóëüñèè, ãåòåðîôàçíûå ìèêðîñèñòåìû.<br />
ÓÄÊ 621.315.592<br />
Î. Â. Òþðèí, À. Þ. Ïîïîâ, Ñ. Î. Æóêîâ, Þ. Ì. Áåðêîâ<br />
ÌÅÕÀͲÇÌ ÑÏÅÊÒÐÀËÜÍί ÑÅÍÑÈÁ²Ë²ÇÀÖ²¯ ÅÌÓËÜѲ¯, ßÊÀ Â̲ÙÓª ÃÅÒÅÐÎÔÀÇͲ ̲ÊÐÎÑÈÑÒÅÌÈ<br />
“ßÄÐÎ — ÎÁÎËÎÍÊÀ”<br />
Çàïðîïîíîâàíî íîâèé ï³äõ³ä äî ïðîöåñó ñïåêòðàëüíî¿ ñåíñèá³ë³çàö³¿ åìóëüñ³é, ñòâîðåíèõ íà îñíîâ³ ãåòåðîôàçíèõ ì³êðîêðèñòàë³â<br />
ñêëàäó “íåñâ³òëî÷óòëèâå ÿäðî — ñâ³òëî÷óòëèâà ñð³áíî-ãàëî¿äíà îáîëîíêà”. Òåõíîëîã³÷íîþ îñîáëèâ³ñòþ äàíîãî<br />
ïðîöåñó º ââåäåííÿ áàðâíèêà ñåíñèá³ë³çàòîðà íà ãðàíèöþ ðîçïîä³ëó “ÿäðî — îáîëîíêà”. Ç óðàõóâàííÿì ö³º¿ îñîáëèâîñò³<br />
— ïðîñòîðîâîãî ðîçïîä³ëó îáîëîíêîþ ãàëîãåí³äà ñð³áëà áàðâíèêà àäñîðáîâàíîãî íà ÿäð³ — çàïðîïîíîâàíèé ìåõàí³çì ñåíñèá³ë³çàö³¿,<br />
ùî çàáåçïå÷óº ðîçøèðåííÿ îáëàñò³ ñïåêòðàëüíî¿ ÷óòëèâîñò³ åìóëüñ³¿.<br />
Êëþ÷îâ³ ñëîâà: ñåíñèá³ë³çàö³ÿ åìóëüñ³¿, ãåòåðîôàçí³ ì³êðîñèñòåìè.
UDC 530.145+678.9<br />
V. V. KOVALCHUK, O. V. AFANAS’EVA, O. I. LESHCHENKO, O. O. LESHCHENKO<br />
Odessa State Institute of the Measuring Technique<br />
SIZE DISTRIBUTIONS OF CLUSTERS ON PHOTOLUMINESCENCE FROM<br />
ENSEMBLES OF SI-CLUSTERS<br />
The quantum confinement model to obtain the photoluminescence (PL) spectra from ensembles<br />
of Si-clusters was used. The size of Si-clusters in ensembles are considered with a Gaussian distribution.<br />
The work demonstrates that a PL peak spectrum in the red-shift region is in agreement with<br />
experiments; a red-shift in the PL peak can be determine due to the size distribution and thus used to<br />
obtain the physically reasonable information about binding energy.<br />
The observations of visible photoluminescence<br />
(PL) in silicon clusters (Si-clusters) suggest that the<br />
Si clusters may become a material for optical applications<br />
if their electronic and optical properties are well<br />
understood [1-9]. The quantum confinement model<br />
[10] has been used to determine the origin and mechanism<br />
of visible PL in Si-clusters.<br />
To explain the effect of the size distributions on<br />
the PL spectra from the Si nanocluster ensembles is<br />
an alternative.<br />
The aim of this work is to report a simple theoretical<br />
framework with a minimal set of parameters to<br />
explain the PL spectra of Si-cluster ensembles.<br />
Our work is based on the quantum confinement<br />
model [10], which has been generally used to determine<br />
the novel electronic and optical properties of<br />
semiconductor clusters. We sugested that disorder<br />
plays a key role and model Si-cluster sizes by a Gaussian<br />
size distribution.<br />
From the experiments [11], it appears reasonable<br />
to assume Si-clusters with a Gaussian distribution of<br />
diameter d centered around a mean d Q ,<br />
( ) 2<br />
d − d<br />
C<br />
0<br />
Рd exp<br />
2<br />
2<br />
2<br />
N ⎛ ⎞<br />
= ⎜− ⎟ (1)<br />
πσ ⎜ σ ⎟<br />
⎝ ⎠<br />
The number of carriers (N ) in a cluster of diam-<br />
C<br />
eter d participating in the PL process is proportional<br />
to d3 : N =ad c 3 , where a is a constant.<br />
For a Si-cluster sample consisting of varying diameters<br />
the probability of carriers participating in the<br />
PL process is given by a product of the above two expressions,<br />
( ) 2<br />
d − d<br />
1 ⎛ ⎞<br />
3<br />
0<br />
Pcd = bd exp⎜−<br />
⎟ (2)<br />
2<br />
2<br />
2πσ<br />
⎜ 2σ<br />
⎟<br />
⎝ ⎠<br />
where b is a suitable normalization constant.<br />
In the quantum confinement model the PL process<br />
is attributed to an energy shift of the carriers<br />
(electrons and holes) and is proportional to l/d2 . The<br />
PL energy hω is given by<br />
h E E c d<br />
where E is the bulk silicon gap, E the exciton binding<br />
b<br />
energy, and c an appropriately dimensioned constant.<br />
The energy shift due to confinement ΔE is<br />
2<br />
ω= g − b + , (3)<br />
© V. V. Kovalchuk, O. V. Afanas’eva, O. I. Leshchenko, O. O. Leshchenko, 2009<br />
( g b)<br />
Δ E = hω− E − E , (4)<br />
2<br />
Δ E0= c d , (5)<br />
where we have also defined a shift 0 E Δ related to the<br />
mean nanocluster diameter d . 0<br />
The PL line shape is approximately Gaussian if σ<br />
is small. For limited σ the factor (ΔE) -1/2 in the exponential<br />
outweighs the polynomial dependence in the<br />
prefactor, resulting in an asymmetric curve with the<br />
shoulder on the short-wavelength (high-energy) side,<br />
which is in agreement with the results of experiments<br />
[11].<br />
In the following, we shall present calculations<br />
based upon our expressions and compare them with<br />
experimentally reported PL sjpectra. We have selected<br />
a representative set associated with the natural theoretical<br />
framework [3-5,10]. In the comparison with the<br />
experimental data, we have faithfully transformed λ to<br />
the energy hω (eV)= 12.4/ λ, where λ is in nm. Recall<br />
that hω=Δ E+ ( Eg − Eb)<br />
from the Gaussian shape. We<br />
compare the theoretical spectra with the one experimentally<br />
obtained at low-power density by Zhang et<br />
al. [11]. We note that the asymmetric line shape, with<br />
the shoulder on the short-wavelength (high-energy)<br />
side, is in agreement with the experimental spectra.<br />
It can be found obviously that the line shapes of PL<br />
spectra deviate from the Gaussian shape. The theoretical<br />
spectra are obtained with a mean diameter d = 0<br />
55 A and σ = 6 A and with a mean diameter d = 45 A<br />
0<br />
and σ = 5 A. Our results show the positions of the PL<br />
peaks if we neglect the effect of the size distribution of<br />
Si-clusters and use the mean energy upshift ΔE in the<br />
0<br />
calculations. A further noteworthy feature is an energy<br />
downshift in the PL peak due to the distribution of<br />
cluster sizes. This enables us to predict the physically<br />
reasonable exciton binding energy E = E + ΔE -hω b g p p<br />
other than E = E + ΔE -hω . For example, we can<br />
b g 0 p<br />
obtain the exciton binding energy of the Si-cluster<br />
with mean diameter 55 A, E = 0.10 eV, in which the<br />
b<br />
peak of the PL is hω =1.48 eV [11], ΔE is obtained E p p t<br />
from bulk band gap of silicon [7,8].<br />
By using the quantum confinement model in neglecting<br />
size distributions of Si -clusters, the exciton<br />
binding energy E , based on experiment [11], is higher<br />
b<br />
than that of the theoretical calculations [15]. In general,<br />
the quantum confinement model only uses the<br />
average size of Si-clusters to obtain the confinement<br />
133
energy. In neglecting the size distributions of Si nanoclusters,<br />
the confinement energy is overestimated. The<br />
exciton binding energy ( E b < 0.10 eV) is more reasonable<br />
when excluding the overestimated part of the<br />
confinement energy from the size distributions of Si<br />
nanoclusters. Thus, the exciton binding energy (E b <<br />
0.10 eV) is small and physically more reasonable.<br />
Our calculation shows that the experimental PL<br />
spectra have a peak at 1.48 eV and a peak at 1.59 eV<br />
in agreement with the quantum conmement model in<br />
combination with the Gaussian size distributions of<br />
Si-clusters.<br />
Indeed, the quantum confinement model, in<br />
combination with the Gaussian distribution of Sicluster<br />
sizes, very well explains the blue shift of the PL<br />
spectra from the Si-cluster-based materials, but still<br />
slightly overestimates the confinement energy. The<br />
widths of the theoretical curves are slightly narrower<br />
compared with the experimental ones. This reflects a<br />
tempera ture effect similar to statistical distribution of<br />
the shapes of the Si-clusters.<br />
We would like to stress, that a whole new class of<br />
materials called clathrates can be generated by altering<br />
the bond angles from their ideal tetrahedral value<br />
in the diamond structure [4,5]. Importantly, the band<br />
gaps of these structures are substantially larger than<br />
that of Si in the diamond structure and well into the<br />
visible region. It is worth noting that the high pressure<br />
phases of bulk Si such as β -Sn, on the other hand, are<br />
metallic with no gap.<br />
When the temperature effect is considered, the<br />
PL spectra as a function of the energy will involve the<br />
temperature distribution function of the electronic<br />
and hole states in the Si nanocluster ensembles. The<br />
statistical average effect results in a broadening of the<br />
PL spectra. The temperature effect results in a thermal<br />
broadening ~ 50 meV at room temperature, which is<br />
not enough to explain the reported broadening of the<br />
PL spectra.<br />
Next, we discuss the shape effect of the Si-clusters<br />
in the ensembles on the PL spectra. Clusters with the<br />
same volume but different shapes can result in different<br />
confinement energies. For example, Si-clusters in<br />
a simple ometry, like spheres, cubes and cylinders, etc.<br />
have different confinement energies [4]. The carriers<br />
with different confinement energies participate in the<br />
PL process and cause the broadening of the PL spectra.<br />
Therefore, the shape effect may also partially contribute<br />
to the broadening of the PL spectra; but the<br />
broadening of the PL spectra (for example, ~ 50 meV<br />
in sphere and cylinder shape Si-clusters) can explain<br />
the experimental results (Fig.1)<br />
On the other hand, the effect of the electron-phonon<br />
interaction on the PL features of the ensembles<br />
of Si nanoclusters as the natural line-width broadening.<br />
In the excitonic state the lattice vibration occurs<br />
because an electron has been transferred from a bonding-like<br />
(valence) to an antibonding-like (conduction)<br />
state, which tends to weaken the bonds [5].<br />
The local amplitude of the lattice vibration is directly<br />
connected to the exciton density. The atomic<br />
displacements are obtained by minimization on the<br />
total energy using PDFT-scheme, for examle [4]. The<br />
variation of the exiton energy gap with respect to the<br />
134<br />
lattice vibration arising from the electron-phonon<br />
interaction is the deformation potentials. It seems as<br />
if the natural line-width broadening is caused by the<br />
electron-phonon interaction. In fact, the hypothesis<br />
that the electron-phonon interaction is responsible for<br />
the broadening of the PL spectra is not plausible.<br />
Fig. 1. Model (up) [4] and STM images [2] acquired from the<br />
same region of the cluster deposited Si(111) 7x7 surface (down).<br />
The image represents an empty state image with bias of 1,5 V<br />
The electron-phonon interaction gives a typical<br />
broadening ( ~ kT < 25 meV), which is too small to<br />
explain the reported full width at half maximum (~<br />
200-300 meV) of the PL spectra of me ensembles of<br />
Si-clusters.<br />
An explanation of this broadening based on the<br />
electron-phonon interaction would involve unphysical<br />
considerations.<br />
Therefore, to obtain more accurate results for the<br />
electron-phonon interaction, it would be necessary,<br />
for instance, to make a very careful theoretical calculation<br />
[4] on the Si-clusters.<br />
In the present work, our aim is mainly to, obtain<br />
the size of Si-clusters from the PL spectra. We only<br />
want to give a method of explaining the PL spectra<br />
from the ensembles of Si-clusters by neglecting me<br />
electron-phonon interaction, the temperature effect<br />
and the shape distributions of Si-clusters.<br />
The problem will be worth further studying in experimental<br />
observations and theoretical calculations.<br />
However, the novel features (width broadening) of me<br />
PL-spectra may be partially due to ubese effects; the<br />
size distributions of the Si-clusters in the ensembles<br />
would influence physical phenomena other than the<br />
PL spectra.<br />
In summary, the method of the quantum confinement<br />
model are used to describe the PL spectra from<br />
an ensembles of Si nanoclusters. The diameters d of<br />
the Si -clusters in the ensembles are considered with a<br />
Gaussian distribution.<br />
This our work demonstrates a PL spectrum with a<br />
line-shape asymmetry on the wavelength (or energy)<br />
scale, in agreement with experiments; a downshift in<br />
the PL peak due to the size distribution.
Cluster approach allows formulating in a new fashion<br />
material sciences concept, it is essential to expand<br />
its possibilities for the decision of modern problems of<br />
nanoelectronics.<br />
We would like to express our deep thanks to Prof.<br />
Dr. M. Drozdov for the helpful comments of the calculation<br />
results.<br />
References<br />
1. Ho M.S, Hwang I.S., Tsong T.T. Direct Observation of Electromigration<br />
of Si Magic Clusters on Si() Surfaces // Phys.<br />
Rev.Lett. — V.84, N25. — 2000. — P.5792-5795<br />
2. M. O. Watanabe, T. Miyazaki, T. Kanayama Deposition of<br />
Hydrogenated Si Clusters on Si(111) — (7 x 7) Surfaces //<br />
Phys.Rev.Lett. — 1998. — v.81,N24. — P. 5362- 5365<br />
3. Drozdov V.A., Êîvalchuk V.V. Electronic processes in nanostructures<br />
with silicon subphase // J.of Phys.Studies. —<br />
2003. — v.4, ¹ 7. — p.393-401<br />
UDC 530.145+678.9<br />
V. V. Kovalchuk, O. V. Afanas’eva, O. I. Leshchenko, O. O. Leshchenko<br />
4. Kovalchuk V.V. Cluster modification semiconductor’s heterostructures.<br />
— : Hi-Tech Press, 2007. — 309 p.<br />
5. Jarrold M.F. Nanosurface chemistry on size-selected silicon<br />
clusters // Science. — 1999. — v. 252. — Ð. 1085-1092<br />
6. Êîvalchuk V.V., Drozdov V.A., Moiseev L.M., Moiseeva V.O.<br />
Optical spectra of polyhedral clusters: influence of the matrix<br />
surroundings // Photoelectronics. — 2005. — ¹ 14. — p.12-18<br />
7. Delerue C., Allan G., Lannoo M. Optical band gap of Si<br />
nanoclusters// J. Lum. — 1999. — v.80. — p.65-73<br />
8. Pavesi L., Dal Negro L., Mazzoleni C., Franzo G., Priolo<br />
F. Optical gain in silicon nanocrystals // Nature. — 2000. —<br />
v.408. — ð.440-444<br />
9. Kamenutsu Y., Suzuki K., Kondo M. Luminescence properties<br />
of a cubic silicon cluster octasilacubane // Phys.Rev. —<br />
1999. — B 51. — P.10666–10669<br />
10. Drozdov V.A., Êîvalchuk V.V., Ìîiseev L.Ì., Osipenko<br />
O.V.Quantum Confinement and Optical Properties of Clusters<br />
// Photoelectronics, 2007. — No 16. — Ð.3-6<br />
11. Zhang R.M., Kozbas P.D. Gaussian distribution of Silicon clusters//<br />
Appl.Phys.Lett.,2004. — 165. — P. 2684-2689<br />
SIZE DISTRIBUTIONS OF CLUSTERS ON PHOTOLUMINESCENCE FROM ENSEMBLES OF SI-CLUSTERS<br />
Abstract<br />
The quantum confinement model to obtain the photoluminescence (PL) spectra from ensembles of Si-clusters was used. The size<br />
of Si-clusters in ensembles are considered with a Gaussian distribution. The work demonstrates that a PL peak spectrum in the red-shift<br />
region is in agreement with experiments; a red-shift in the PL peak can be determine due to the size distribution and thus used to obtain<br />
the physically reasonable information about binding energy.<br />
Key words: clusters, photoluminescence, distributions.<br />
ÓÄÊ 530.145+678.9<br />
Â. Â. Koâaëü÷óê, O. Â. Àôàíàñüåâà, À. Î. Ëåùåíêî, Î. È. Ëåùåíêî<br />
ÐÀÑÏÐÅÄÅËÅÍÈÅ ÏÎ ÐÀÇÌÅÐÀÌ ÊËÀÑÒÅÐÎÂ ÏÐÈ ÔÎÒÎËÞÌÈÍÅÑÖÅÍÖÈÈ, ÎÁÓÑËÎÂËÅÍÍÎÉ<br />
ÑÓÙÅÑÒÂÎÂÀÍÈÅÌ ÀÍÑÀÌÁËß SI-ÊËÀÑÒÅÐÎÂ<br />
Ðåçþìå<br />
Èñïîëüçîâàíà êâàíòîâàÿ ìîäåëü êîíôàéìåíòà ñ öåëüþ ïîëó÷åíèÿ ñïåêòðîâ ôîòîëþìèíåñöåíöèè (ÔË), âûçâàííîé íàëè÷èåì<br />
àíñàìáëÿ Si-êëàñòåðîâ. Ðàçáðîñ ðàçìåðîâ Si-êëàñòåðîâ â àíñàìáëå àïðîêñèìèðîâàí â ðàìêàõ ðàñïðåäåëåíèÿ Ãàóññà. Â<br />
ðàáîòå ïîêàçàíî, ÷òî ñïåêòðû ÔË èìåþò àñèììåòðèþ ïèêà â îáëàñòü êðàñíîãî ñìåùåíèÿ, ÷òî ñîãëàñóåòñÿ ñ ýêñïåðèìåíòîì;<br />
êðàñíûé ñäâèã ïèêà ñïåêòðà ÔË îïðåäåëÿåòñÿ ðàñïðåäåëåíèåì ïî ðàçìåðàì êëàñòåðîâ, ÷òî ïîçâîëÿåò ïîëó÷èòü ôèçè÷åñêè<br />
ðàçóìíóþ èíôîðìàöèþ îá ýíåðãèè ñâÿçè.<br />
Êëþ÷åâûå ñëîâà: êëàñòåðû, ôîòîëþìèíåñöåíöèÿ, ðàñïðåäåëåíèå.<br />
ÓÄÊ 530.145+678.9<br />
Â. Â. Koâaëü÷óê, Î. Â. Àôàíàñüºâà, Î. ². Ëåùåíêî, Î. Î. Ëåùåíêî<br />
ÐÎÇÏÎÄ²Ë ÏÎ ÐÎÇ̲ÐÀÕ ÊËÀÑÒÅв ÏÐÈ ÔÎÒÎËÞ̲ÍÅÑÖÅÍÖ²¯, ÎÁÓÌÎÂËÅÍί ²ÑÍÓÂÀÍÍßÌ<br />
ÀÍÑÀÌÁËÞ SI-ÊËÀÑÒÅвÂ<br />
Ðåçþìå<br />
Âèêîðèñòàíà êâàíòîâà ìîäåëü êîíôàéìåíòà ç ìåòîþ îäåðæàííÿ ñïåêòð³â ôîòîëþì³íåñöåíö³¿ (ÔË), ùî âèêëèêàíà íàÿâí³ñòþ<br />
àíñàìáëþ Si-êëàñòåð³â. Ðîçïîä³ë ðîçì³ð³â Si-êëàñòåð³â â àíñàìáë³ àïðîêñèìîâàíèé ó ðàìêàõ ðîçïîä³ëó Ãàóññà. Ó ðîáîò³<br />
ïîêàçàíî, ùî ñïåêòðè ÔË ìàþòü àñèìåòð³þ ï³êà â îáëàñòü ÷åðâîíîãî çñóâó, ùî óçãîäæóºòüñÿ ç åêñïåðèìåíòîì; ÷åðâîíèé çñóâ<br />
ï³êà ñïåêòðà ÔË âèçíà÷àºòüñÿ ðîçïîä³ëîì ïî ðîçì³ðàõ êëàñòåð³â, ùî äîçâîëÿº îäåðæàòè ô³çè÷íî ðîçóìíó ³íôîðìàö³þ ïðî<br />
åíåðã³þ çâ’ÿçêó.<br />
Êëþ÷îâ³ ñëîâà: êëàñòåðè, ôîòîëþì³íåñöåíö³ÿ, ðîçïîä³ë.<br />
135
136<br />
UDC 621.382<br />
I. M. VIKULIN, 1 SH. D. KURMASHEV, P. YU. MARKOLENKO, 1 P. P. GECHEV<br />
Odessa A.S.Popov National Academy of Communications<br />
1 Odessa I.I.Mechnicov National University<br />
Odessa, 65026, Ukraine, e-mail: kurm@mail.css.od.ua. Tel. (0482) — 746-66-58.<br />
RADIATION IMMUNITY OF THE PLANAR n-p-n-TRANSISTORS<br />
Influence of streams of electrons, neutrons and γ–quantum is investigational on the amplification<br />
factor of bipolar planar-epitaxial transistors. It is shown that a preliminary thermal-electric-train<br />
allows to increase of radiation immunity in 2–3 times.<br />
1. INTRODUCTION<br />
The action of radiation around to solids brings to<br />
the origin of a number of effects: excitation of atoms<br />
and their ionization, nuclear transmutation birth of<br />
pair, is an electron-positron, displacement of atoms<br />
from the knots of crystalline grate in interstice space<br />
and other The change of electrophysics properties of<br />
silicon under the action of radiation (this, as a rule,<br />
compensation of conductivity [1]) associates both<br />
with the process of origin of initially-stable defects<br />
(divacancies, tetravacancies and other is similar defects)<br />
and second — as a result of quasi-chemical reactions<br />
(complexes of vacancies with admixtures by<br />
the alloying and concomitant admixtures of oxygen,<br />
carbon and other[2]). In particular, in n-Si, alloyed<br />
phosphorus, non-equilibrium vacancies, entering<br />
into the quasi-chemical reaction with alloying (phosphorus)<br />
or with base-line (oxygen, carbon) admixtures,<br />
and also between itself, form the second radiation<br />
defects[3].<br />
Most effectively the reactions take place in the<br />
conditions of irradiation of γ–quantum Ñî 60 , by electrons<br />
with energy 1 MeV, low-energy protons and<br />
other by particles energy of which in order of size is<br />
comparable with energy of origin of defects (by different<br />
estimations she makes for silicon 145-250 keV).<br />
In scientific literature in detail all sides of work of<br />
bipolar transistors are lighted up at the irradiation. It is<br />
shown that none of parameters characteristic for them<br />
remains here unchanging. Nevertheless, distinctions<br />
appear substantial in the change of properties at the<br />
irradiation of alloyed transistors and structures got a<br />
diffusive method, in particular, planar-epitaxial transistors.<br />
In the real work the comparative change of amplifying<br />
properties of the planar transistors of the same<br />
type is considered at influence of radiation by fast<br />
electrons, γ–quantum, fast neutrons.<br />
2. METHOD OF EXPERIMENT<br />
The experimental samples silicon n-p-n-transistors<br />
were made on planar-epitaxial technology. Specific<br />
resistance of initial epitaxial film ρ= 0.5 Ω∙cm.<br />
Before the leadthrough of thermal operations silicon<br />
epitaxial structures 15KEP was processed by boiling<br />
5 min. in solution of Í Î + Í ÎÍ + Í Î (4:1:1).<br />
2 2 2 2<br />
Then conducted oxidization of standards in a stove at<br />
Ò=(1200±3) 0Ñ. . Thickness of the growth layer SiO2 it was 0.6 μ. This layer served as a mask at diffusion<br />
of the boron. After creation of oxyde masking film<br />
photolithography was conducted for diffusion of the<br />
coniferous boron (base). Boron spin-on was carried<br />
out at Ò=9400Ñ in the atmosphere of argon for the receipt<br />
of sheet resistivity R =60 Ω/�. For forming of<br />
s<br />
base layer with the necessary distributing of concentration<br />
of dopant the drivi-in of dopant of the boron<br />
was conducted at Ò=11500Ñ. Concentration of alloy-<br />
ing dopant in a base ~1.5∙10 18 cm 2<br />
− , depth of diffusive<br />
transition was ~3.5 μ. Photolithography was further<br />
conducted under the emitter of transistor. The region<br />
of emitter was created by diffusion of phosphorus from<br />
PCl 3 at Ò=1050 0 Ñ in the atmosphere of oxygen. Depth<br />
of emitter diffusion was 2-3 μ. The thickness of base<br />
made ~1.5 μ. For forming of contact in area of base<br />
and emitter photolithography was conducted under<br />
the contacts. After it on the plate of silicon by cathode-ray<br />
evaporation coating the layer of aluminium<br />
in thick 1.5 μ. After lithography on metallization aluminium<br />
was firing at Ò=550 0 Ñ in the atmosphere of<br />
argon during 15 min. The got structures were pressurized<br />
in a glass-to-metal corps.<br />
The irradiation of structures was made by the<br />
stream of fast electrons Fe on the linear accelerating<br />
”Electronics” as ELU-4. Energy of electrons 5.0<br />
MeV, current of irradiation 0.1-1.0 mA. The irradiation<br />
of samples was made at Ò=60 0 Ñ.<br />
The irradiation by the streams of neutrons F n with<br />
Å=1.5 MeV was made in the horizontal channel of reactor<br />
of ÂÂÐ-M. The thermal-neutron were chopped<br />
off by a cadmium filter. The fluence was 10 11 -10 15<br />
n∙cm -2 .<br />
The irradiation of γ–quantum was made from the<br />
source of Ñî 60 . Intensity of γ–quantum (dose of D γ )<br />
was 3500 R/s, middle energy of quantum — 1.25 MeV.<br />
The temperature of samples in the process of irradiation<br />
changed in an interval (293-330) 0 C.<br />
Alloying of single-crystals planar-epitaxial of silicon<br />
by the ions of the boron with energy 30-100 keV<br />
to the doses 10 14 -10 18 i∙cm -2 it was conducted on the<br />
accelerating setting “Vesuvius-1À” with the division<br />
on the masses.<br />
© I. M. Vikulin, Sh. D. Kurmashev, P. Yu. Markolenko, P. P. Gechev, 2009
3. RESULTS AND DISCUSSION<br />
On descriptions of semiconductor elements determination<br />
of maximum streams is the practical purpose<br />
of any research of influencing of radiations at which<br />
an element falls out. There is a research task also development<br />
of technological methods of increase of<br />
these maximum values.<br />
Amplifying properties of transistor are characterized<br />
by h 21e , the transfer coefficient of current h 21e<br />
equal to attitude of change of output current toward<br />
the change of entrance in a scheme common-emitter<br />
( h 21e = Ic I b ). In this case we examine a transfer coefficient<br />
exactly h 21e , as it answers most amplification<br />
factor of power of transistor. As is generally known,<br />
it concernes by efficiency of emitter, coefficient of<br />
transfer and efficiency of collector.<br />
In a scheme common-emitter h 21e measured a<br />
transfer (strengthening) current h21e at tension on a<br />
collector U с = 3V and current of collector I c =3 A.<br />
Temperature of measurings T = 293K . Measurings<br />
were conducted for three parties of transistors (for<br />
three types of transfer got out with the identical value<br />
h 21e , each of which was exposed to the rays by a certain<br />
stream in a range 10 12 - 10 15 cm 2 − , whereupon<br />
extracted a transistor from the chamber of irradiation<br />
and measured the value h 21e .<br />
It was studied to dependence of mobility (fig.1),<br />
time of life of charge curriers (fig.2), and also specific<br />
resistance of collector, base and emitter (fig.3) from a<br />
radiation stream.<br />
Fig. 1. Dependence of mobility of electrons in epitaxial layer<br />
of silicon with ρ= 0.5 Ω∙cm from the stream of electrons (1), fast<br />
neutrons (2) and γ–quantum (3)<br />
The slump of mobility and of life time of charge<br />
curriers under the action of radiation decreases length<br />
of diffusion L , and accordingly the transfer coeffi-<br />
n,p<br />
cient of current of transistor, as 21e<br />
h ~ L n,p /w 2 . At the<br />
same time the thickness of base w changes at the irradiation.<br />
The growth of specific resistance of collector,<br />
base and emitter (fig. 3) corresponds to decreases of<br />
concentration of charge curriers in a semiconductor<br />
at the irradiation [1]. It results in displacement of collector<br />
deep into semiconductor, i.e. increase w. It is<br />
special shows up in structures with a thin base, as in<br />
our case (w≈1.5 μ).<br />
Fig. 2. Change of life time of electrons in the p-base (1) and<br />
holes in the n-collector (2)<br />
Fig. 3. Change of specific resistance collector, base and emitter<br />
at the irradiation by neutrons: 1– n-collector ( ρ= 10 Ω⋅cm),<br />
2 — p-base ( ρ=10 –2 Ω⋅cm), 3 — n-emitter ( ρ=10 –3 Ω⋅cm)<br />
On a fig. 4 dependence of the transfer coefficient<br />
of current h 21e of planar transistors is shown on the<br />
stream of electrons, neutrons and γ–quantum. It ensues<br />
from the graphs, that thresholds values of radiation<br />
streams which the sharp decreasing of strengthening<br />
of transistor is after, make accordingly: for the<br />
14<br />
2<br />
stream of electrons 5 ∙10<br />
e ∙ cm − , for the neutrons of<br />
10 13 2<br />
n cm − , for γ–quantum 10 6 R . Physical reason<br />
of decreasing h 21e under the action of radiation is the<br />
origin of defects in the crystalline structure of semiconductor,<br />
which shunt emitter p-n-junctions and reduce<br />
its coefficient of injection, and also decreasing of<br />
life time of charge curriers in the base of transistor.<br />
The graphs of fig. 4 correlates with the dependencis<br />
of pictures 1, 2 and 3. It ensues from these dataes, that<br />
with the increase of maximum frequency of transistors<br />
their radiation immunity is increased. We will remind,<br />
that maximum decreased is frequency on which the<br />
transfer coefficient of current h 21e diminishes in 2<br />
one times. In the scheme of including common-emitter<br />
maximum frequency concernes by effective time of<br />
life of transmitters of charge (inversely proportional).<br />
At planar, i.e. diffusive transistor of n-p-n-type maximum<br />
frequency higher, than at driveable, therefore<br />
and radiation immunity higher.<br />
137
Fig. 4. Influencing of streams of electrons, neutrons and<br />
γ–quantum on the transfer coefficient of current h 21e<br />
For the increase of radiation immunity of transistors<br />
we used the method of thermal-electric-train<br />
(ÒEÒ). Co-operation of radiation streams with a<br />
semiconductor it is possible to examine by consisting<br />
of two stages. On the first stage (at small energies of<br />
stream) a radiation operates on technological defects<br />
appearing in material at creation of transistor, and on<br />
the second (at large energies) the own defects of crystalline<br />
structure are created. The creature of method<br />
consists of that at ÒÝÒ technological defects and parameters<br />
of transistor collapse at the small streams of<br />
radiation does not change.<br />
138<br />
UDC 621.382<br />
I. M. Vikulin, Sh. D. Kurmashev, P. Yu. Mfrkolenko, P. P. Gechev<br />
RADIATION IMMUNITY OF THE PLANAR n-p-n-TRANSISTORS<br />
Experimental verification of method of ÒEÒ was<br />
carried out as follows. Transistors were heated to<br />
0<br />
125 C and was simultaneously exposed to periodic<br />
electric influence with power in 2.5 time greater, than<br />
passport maximal power. Self-control was conducted<br />
at maximal influence about 5 min. with a period 20<br />
min. Research of influencing of radiation streams on<br />
the parameters of transistors, that after ÒEÒ during<br />
40–60 hours thresholds values of radiation streams resulting<br />
in the sharp decreased h 21e , is increased more<br />
than in 2 times. The got results talk about the change<br />
of terms of second radiation residual damage as a result<br />
preliminary thermal and ÒEÒ treatments of transistors<br />
structures.<br />
As the structure of investigational transistors does<br />
not differ from transistors in the integrated circuits,<br />
phototransistors and etc, the considered method of<br />
increase of radiation immunity is applicable to these<br />
semiconductor devices.<br />
4. CONCLUSIONS<br />
The thresholds streams of electrons, neutrons, γ–<br />
14<br />
quantum for a planar n-p-n-transistor make 5 ∙10<br />
e<br />
2<br />
cm − , 10 13 2<br />
n cm − and 10 6 R , accordingly. For the increase<br />
of radiation immunity of transistors the method<br />
of thermal-electric-train can be used.<br />
References<br />
1. Áðóäíûé Â.Í. Ìîäåëü ñàìîêîìïåíñàöèè è ñòàáèëèçàöèÿ<br />
óðîâíÿ Ôåðìè â îáëó÷åííûõ ïîëóïðîâîäíèêàõ/<br />
Â.Í.Áðóäíûé, Í.Ã.Êîëèí, Ë.Ñ.Ñìèðíîâ// Ôèçèêà è òåõíèêà<br />
ïîëóïðîâîäíèêîâ. — 2007. — Ò.41, ¹9. — Ñ. 1031-<br />
1040.<br />
2. Ìóðèí Ë.È. Áèñòàáèëüíîñòü è ýëåêòðè÷åñêàÿ àêòèâíîñòü<br />
êîìïëåêñà âàêàíñèÿ-äâà àòîìà êèñëîðîäà â<br />
êðåìíèè/Ë.È.Ìóðèí, Â.Ï.Ìàðêåâè÷, È.Ô.Ìåäâåäåâà,<br />
Ë.Äîáî÷åâñêèé // Ôèçèêà è òåõíèêà ïîëóïðîâîäíèêîâ.<br />
— 2006. — Ò.40, ¹11. — Ñ. 1316-1320.<br />
3. Ïàãàâà Ò.À. Âëèÿíèå òåìïåðàòóðû îáëó÷åíèÿ íà ýôôåêòèâíîñòü<br />
ââåäåíèÿ ìóëüòèâàêàíñèîííûõ äåôåêòîâ â<br />
êðèñòàëëàõ n-Si/ Ò.À.Ïàãàâà // Ôèçèêà è òåõíèêà ïîëóïðîâîäíèêîâ.<br />
— 2006. — Ò.40, ¹8. — Ñ. 919-921.<br />
Abstract<br />
Influence of streams of electrons, neutrons and γ–quantum is investigational on the amplification factor of bipolar planar-epitaxial<br />
transistors. It is shown that a preliminary thermal-electric-train allows to increase of radiation immunity in 2–3 times.<br />
Key words: neytrons, radiation immuniti, transistors.<br />
ÓÄÊ 621.382<br />
È. Ì. Âèêóëèí, Ø. Ä. Êóðìàøåâ, Ï. Þ. Ìàðêîëåíêî, Ï. Ï. Ãå÷åâ<br />
ÐÀÄÈÀÖÈÎÍÍÀß ÑÒÎÉÊÎÑÒÜ ÏËÀÍÀÐÍÛÕ n-p-n-ÒÐÀÍÇÈÑÒÎÐÎÂ<br />
Ðåçþìå<br />
Èññëåäîâàëîñü âëèÿíèå ïîòîêîâ ýëåêòðîíîâ, íåéòðîíîâ è γ-êâàíòîâ íà êîýôôèöèåíò óñèëåíèÿ áèïîëÿðíûõ ïëàíàðíîýïèòàêñèàëüíûõ<br />
òðàíçèñòîðîâ. Ïîêàçàíî, ÷òî ïðåäâàðèòåëüíàÿ ýëåêòðîòåðìîòðåíèðîâêà ïîçâîëÿåò óâåëè÷èòü ðàäèàöèîííóþ<br />
óñòîé÷èâîñòü ïðèáîðîâ â 2 ðàçà.<br />
Êëþ÷åâûå ñëîâà: íåéòðîíû, ðàäèàöèîííàÿ ñòîéêîñòü, òðàíçèñòîð.
ÓÄÊ 621.382<br />
². Ì. ³êóë³í, Ø. Ä. Êóðìàøåâ, Ï. Þ. Ìàðêîëåíêî, Ï. Ï. Ãå÷åâ<br />
ÐÀIJÀÖ²ÉÍÀ ÑÒ²ÉʲÑÒÜ ÏËÀÍÀÐÍÈÕ n-p-n -ÒÐÀÍÇÈÑÒÎвÂ<br />
Ðåçþìå<br />
Äîñë³äæóâàâñÿ âïëèâ ïîòîê³â åëåêòðîí³â, íåéòðîí³â ³ γ –êâàíò³â íà êîåô³ö³ºíò ï³äñèëåííÿ á³ïîëÿðíèõ ïëàíàðíî-åï³òàêñ³éíèõ<br />
òðàíçèñòîð³â. Ïîêàçàíî, ùî ïîïåðåäíº åëåêòðîòåðìîòðåíóâàííÿ äîçâîëÿº ï³äâèùèòè ðàä³àö³éíó ñò³éê³ñòü ïðèëàä³â â<br />
2 ðàçè.<br />
Êëþ÷îâ³ ñëîâà: íåéòðîíè, ðàä³àö³éíà ñò³éê³ñòü, òðàíçèñòîðè.<br />
139
140<br />
UDC.539.125.5<br />
K. V. AVDONIN<br />
The Kiev national university of technologies and design,<br />
01011, Ukraine, Kiev, Nemirovich’s street, 2,<br />
Ph. 256-29-27, 419-32-59<br />
E-mail: info@izdo-knutd.com.ua<br />
BUILD-UP OF WAVE FUNCTIONS OF THE PARTICLE<br />
IN THE MODELLING PERIODIC FIELD<br />
A movement pattern of charged free particles inside of a crystal is proposed in this work. Effective<br />
potential field of interaction between particles and crystal lattice point is considered in two presentations:<br />
as three-dimensional Fourier series and as the presentation of effective field, in which its periodicity<br />
is made by Heviside’s function. Wave functions are found by composition of individual solutions<br />
of Schrödinger equation in the form of functional series. A possibility of existence of energy levels of<br />
foreign particle’s condition in crystal appropriate to them is researched.<br />
1. INTRODUCTION<br />
The Overall objective of paper was examination of<br />
an opportunity to find particular solutions of a steadystate<br />
Schrodinger equation: ϕ (r)<br />
2<br />
⎛ ⎞<br />
h<br />
⎜− Δ+ W(r) ⎟ψ<br />
(r) = Eϕ(r).<br />
(1)<br />
⎝ 2μ<br />
⎠<br />
In the form of functional numbers by means of a<br />
method created in paper [1], for a derived periodic field<br />
(2) and a modelling field (38), research of existence of<br />
corresponding power levels, and also opportunities of<br />
reception of distribution of probability for directions<br />
of movement of a particle. One of last papers devoted<br />
to the general features of a power spectrum of a particle<br />
in a periodic field, is paper [2] which result confirms<br />
the character of the power spectrum received in given<br />
paper. Advantage of an offered method of search is its<br />
relative simplicity and besides it gives new opportunities<br />
for a computer simulation of processes occuring<br />
in real crystals.<br />
2. SEARCH OF WAVE FUNCTIONS<br />
OF A PARTICLE IN THE PERIODIC FIELD<br />
WHICH HAS BEEN RESOLVED<br />
IN A FOURIER SERIES<br />
The most often the periodic field is represented in<br />
the form of a Fourier series:<br />
W(r) = a +∑ a exp( ikr).<br />
(2)<br />
0 k<br />
k≠0 After substitution of potential energy in a Schrodinger<br />
equation (1) in the form of a Fourier series (2)<br />
it would be:<br />
{ } 2<br />
m<br />
Δ+<br />
Or in the other form:<br />
ϕ (r),<br />
(3)<br />
Δϕ (r) = f (r) ϕ (r),<br />
Where the labels are used as follows:<br />
(4)<br />
f(r) = A +<br />
2 2μ<br />
m = 2 ( E−a0) ,<br />
h<br />
2μ<br />
Ak = a 2 k,<br />
h<br />
A exp( ikr) = A exp( ikr),<br />
0 k k<br />
k≠0 k<br />
(5)<br />
∑ ∑ (6)<br />
Let’s choose function, too periodic, with a Fourier<br />
analysis: h (r)<br />
h(r) = ∑ Hнexp( ivr),<br />
н≠0 Satisfying to a requirement:<br />
Δ h(r) = A0h(r), Then the particular solution of the equation (6)<br />
can be constructed in such form:<br />
Where:<br />
1<br />
n<br />
∑ ( ) n (7)<br />
ϕ (r) = h(r) −D (r, ν ) + −1 D (r, ν),<br />
D<br />
∑∑<br />
∞<br />
n=<br />
2<br />
н k00 1<br />
н k0≠0 2<br />
k0<br />
+ν<br />
{ ( +ν)<br />
}<br />
H A exp i k r<br />
;<br />
sα = k0 + k 1+<br />
... + k α.<br />
In fact, we shall work functional Laplassa on function<br />
(7):<br />
Δϕ (r) =Δh−Δ D1+ΔD2 −Δ D3 +ΔD4 − ... (8)<br />
As:<br />
∑<br />
н≠0 { } ∑<br />
k0≠0 0 { }<br />
h(r) ∑ Ak exp{ ik}<br />
0<br />
0r<br />
;<br />
ΔD (r, ν ) = − H exp iνrA exp ikr−<br />
1н k 0<br />
−<br />
k0≠0 ⎧⎪ ⎫⎪<br />
ΔDn(r, ν ) = − ⎨∑Ak exp{ ik}<br />
1<br />
1r<br />
n−<br />
n−⎬×<br />
⎪⎩kn−1⎪⎭ × ∑∑∑<br />
н k0≠0 k1 Hн∏ n−1exp{ i(<br />
sn−2 +ν)<br />
r}<br />
... ∑<br />
=<br />
2 2 2<br />
kn−2<br />
s0 +ν s 1+ν ⋅... ⋅ sn−<br />
2 +ν<br />
=−f(r) D (r, ν).<br />
n−1<br />
© K. V. Avdonin, 2009
{ i 0( − 0)<br />
+ i ( − ) }<br />
That of expression (8) follows, that:<br />
Δϕ (r) = h(r) A0 + h(r) ∑ Ak exp{ ik}<br />
0<br />
0r−<br />
k≠0 − f(r) D1(r) + f(r) D2(r) − f(r) D3(r)<br />
+ ... =<br />
= f(r) ( h− D1+ D2 − D3 + ... ) = f(r)<br />
ϕ(r).<br />
For an important special case:<br />
h(r) = exp( imr<br />
) .<br />
The solution (7) will look like:<br />
Ak exp{ i(<br />
k ) }<br />
0<br />
0 + m r<br />
ϕ (r) = exp{ imr}<br />
− ∑<br />
+<br />
2<br />
k≠0<br />
k0+ m<br />
(9)<br />
n<br />
∞<br />
( −1) ∏ n−1exp{ i(<br />
sn−1+ m) r}<br />
+ ∑∑∑ ... ∑<br />
.<br />
2 2 2<br />
n=<br />
2 k0≠0 k1 kn−1<br />
s0 + m s1+ m ⋅... ⋅ sn−1+ m<br />
Grouping addends on degrees of coefficient (see<br />
appendix A), we have: 0 A<br />
⎧ ⎪ Ak exp{ i(<br />
k ) }<br />
0<br />
0 + m r<br />
ϕ= exp{ imr}<br />
−⎨1− ∑<br />
+<br />
2<br />
⎪⎩ k≠0<br />
k0 + m + A0<br />
∏ n−1exp{ i(<br />
sn + m) r<br />
⎫ } ⎪<br />
+ ∑ ∑ −... .<br />
2 2 ⎬ (10)<br />
k0≠0 k1≠0( s0 + m + A0)( s1+ m + A0)<br />
⎪<br />
⎭<br />
A series (10) has a singularity at, that impedes<br />
search of wave functions generally. s j = 0 Let Fourier<br />
analysis of a wave function looks like:<br />
ϕ (r) =∑ Bkexp( ikr).<br />
k<br />
Then the functional series (10) will be iteration of<br />
a relation:<br />
Ak B { ( ) }<br />
0 k exp i k0 + k r<br />
ϕ= exp( imr<br />
) −∑<br />
. 2<br />
k 0 ,k k0+ k −m<br />
(11)<br />
By a direct substitution it is possible to prove (see<br />
addition), that the relation (11) satisfies the equation<br />
(5). Substituting in expression of a Fourier analysis<br />
(13) coefficients in the form of:<br />
λ0<br />
1<br />
Ak = u( с ) { }<br />
0<br />
0 exp ik0с0 dс<br />
0.<br />
V ∫<br />
0 0<br />
(12)<br />
Where: 0 0 0 0 ;<br />
1 exp с k r с<br />
Ω= ∑<br />
=<br />
2 2<br />
VV 0 kk 0 k-k0<br />
− m<br />
⎛ 1 ⎞ 1 exp с { ik0(<br />
r − 0)<br />
}<br />
= ⎜ ∑exp с { ik0(<br />
r − 0) }<br />
2 2<br />
V<br />
⎟ ∑<br />
=<br />
⎝ k ⎠V0k<br />
k 0 0 − m<br />
1 exp с { ik0(<br />
r − 0)<br />
}<br />
=δ(<br />
с - с 0 ) ∑ . (14)<br />
2 2<br />
V0 k k 0 0 − m<br />
Substituting the integrand total in the form of (14)<br />
in the equation (13) we have:<br />
λ<br />
imr<br />
ϕ (r) = e + ∫ G0( r, ρ) u( ρ) ϕ( ρ) dρ.<br />
0<br />
V =λ xλyλ z V =λxλyλ z and — a vector<br />
of translation for functions that accordingly, we shall<br />
gain:<br />
λ0 λ<br />
ϕ (r) = exp{ imr } −∫∫ Ωu( ρ0) ϕ( ρ) dρ0dρ. (13)<br />
0 0<br />
Where:<br />
1 exp{ ik0( r −ρ 0)<br />
+ ik(<br />
r −ρ)<br />
}<br />
Ω= ∑<br />
.<br />
2 2<br />
VV 0 kk 0 k-k0<br />
− m<br />
It is obvious, that translations vectors and should<br />
multiple. λλ 0 Here two cases are possible: 1);<br />
λ 0 = nλ<br />
2), where — natural number n. In the first<br />
case for the integrand total of expression (13) there is<br />
an opportunity to shift the beginning of summation<br />
on a vector to a vector then the integrand total would<br />
be equal:<br />
% (15)<br />
Where:<br />
1 exp{ ik<br />
0 (r }<br />
G%<br />
−ρ<br />
0 (r, ρ ) = − ∑<br />
. (16)<br />
2 2<br />
V0 k k 0 0 − m<br />
In the second case for the integrand total of expression<br />
(13) there is an opportunity to shift the beginning<br />
of summation on a vector to a vector then the<br />
integrand total is as follows: Ω kk0<br />
1 exp с { i k k0( r − с 0)<br />
+ i ( − ) }<br />
Ω= ∑<br />
=<br />
2 2<br />
VV 0 kk 0 k-k0<br />
− m<br />
⎛ 1 ⎞ 1 exp с { ik(<br />
r − ) }<br />
= ⎜ ∑exp с с{<br />
ik0(<br />
− 0) } ⎟<br />
2 2<br />
V ⎟ ∑<br />
=<br />
⎝ 0 k V 0<br />
⎠ k k − m<br />
1 exp с { ik(<br />
r − 0 ) }<br />
=δ( с−с 0 ) ⋅ ∑ . (17)<br />
2 2<br />
V k k − m<br />
Substituting the integrand total in the form of (17)<br />
in the equation (13) we have:<br />
λ0<br />
imr<br />
ϕ (r) = e + ∫ G(r, ρ) u() ρ ϕ( ρ) dρ.<br />
0<br />
% Where:<br />
(18)<br />
1 exp{ ik(r<br />
}<br />
G%<br />
−ρ<br />
0 (r, ρ ) = − ∑<br />
. (19)<br />
2<br />
V k k0−m By a direct substitution we can convince, that functional<br />
action on functions (16) and (19) gives a Dirac<br />
delta function. ( ) 2<br />
Δ+ m Thus, functions also play a<br />
role of source functions for the equation (3). Spectral<br />
decompositions of a source function (16), (19) are already<br />
known (for example, from [4]). A vector and a<br />
wave function as arise from the equations (15); (18)<br />
have an identical translation vector, that is:<br />
⎛2π 2π 2π<br />
⎞<br />
m = ⎜ lx; ly; lz<br />
.<br />
⎜<br />
⎟<br />
⎝λx λy λz<br />
⎠<br />
(20)<br />
Where: lx, ly, l z — integers.<br />
Let’s view a boundary case, in other words, we<br />
shall change Fourier series by a Fourier integral in expressions<br />
(16), (19). After Integrating, by means of the<br />
calculus of residues, we gain:<br />
1 exp { ik(r<br />
−ρ)<br />
}<br />
G(r,<br />
ρ ) = − ∑ . (21)<br />
2 2<br />
V k k − m<br />
Expression (21) coincides with a source function<br />
for the acyclic field, found in paper [5]. Source func-<br />
141
tions in the equations (15), (18) have a singularity at,<br />
accordingly k0= m k = m . Let — the complete set<br />
of the vectors equal to a vector modulo. We can put out<br />
singularity for source functions in expressions (16),<br />
(19) by transducing source functions as follows:<br />
N 1 imr%<br />
j<br />
G(r, ρ ) = − ∑ exp 2 { imj(r<br />
−ρ) } −<br />
V j=<br />
1 2m<br />
1 exp{ ik(r<br />
−ρ)<br />
}<br />
− ∑<br />
.<br />
(22)<br />
2 2<br />
V k k − m<br />
k ≠ m<br />
N 1 imr%<br />
j<br />
G0(r, ρ ) = − ∑ exp 2 { imj(r<br />
−ρ) } −<br />
V0 j=<br />
1 2m<br />
1 exp{ ik<br />
0 (r −ρ)<br />
}<br />
− ∑<br />
. (23)<br />
2 2<br />
V0 k k0−m k ≠ m<br />
Where:<br />
r% = r −λn;<br />
л n = { nzλx; nyλy; nzλ<br />
z}<br />
— the peak translation<br />
vector, contained in radius-vector.<br />
As is shown in appendix D, the functions, definiendums<br />
by (22), (23) satisfied relations:<br />
142<br />
2 ( m ) G(<br />
)<br />
Δ+ r, ρ =δ(r −ρ ). (24)<br />
Thus, the transduced source functions (22), (23)<br />
can be used in integral representations of a Schrodinger<br />
equation (15); (18), that is:<br />
λ<br />
ϕ (r) = exp imr + G (r, ρ) u ρ ϕ( ρ) dρ.<br />
{ } ( )<br />
∫<br />
0<br />
0<br />
(25)<br />
λ0<br />
{ } ( )<br />
ϕ (r) = exp imr + ∫ G(r, ρ) u<br />
0<br />
ρ ϕ( ρ) dρ.<br />
(26)<br />
The homogeneous equations corresponding the<br />
equations (25); (26), can be presented in the form<br />
of:<br />
N imr%<br />
j<br />
ϕ (r) = −∑ exp 2 { imjr} P(<br />
m j)<br />
−<br />
j=<br />
1 2m<br />
exp{ ikr}<br />
− ∑ P(k).<br />
2 2<br />
k − m<br />
(27)<br />
k<br />
k ≠ m<br />
N imr%<br />
j<br />
ϕ (r) = −∑ exp 2 { imjr} P0(<br />
m j)<br />
−<br />
j=<br />
1 2m<br />
exp{ ik0r}<br />
− ∑ P(k).<br />
2 2<br />
k k − m<br />
k ≠ m<br />
Where:<br />
λ0<br />
0<br />
(28)<br />
1<br />
P(k) = u() ρ ϕ( ρ)exp( −ik ρ) dρ.<br />
V ∫ (29)<br />
λ0<br />
1<br />
P (k) = ∫ u() ρ ϕ( ρ)exp( −ik ρ) dρ.<br />
(30)<br />
0 0<br />
V0<br />
0<br />
After both parts of the equations (27), (28) were<br />
multiplied on functions:<br />
1<br />
u(r)exp{ − ik′<br />
r } .<br />
V<br />
Accordingly, and having integrated, we shall gain<br />
systems of the homogeneous linear equations:<br />
Where:<br />
( ( ′ ) +δk,k′<br />
)<br />
k<br />
( ( ′ ) +δ ′ )<br />
∑ Q k,k P(k).<br />
(31)<br />
∑ Q0 k,k 0 0 k 0,k P<br />
0 0(k). 0 (32)<br />
k0<br />
N<br />
N<br />
⎛ Ck,k ′ ⎞ A%<br />
k,k ′<br />
Q = ∑⎜δ k,m 2 2 2 { 1 k,m } .<br />
j ⎟+<br />
∏ −δ (33)<br />
j<br />
j= 1 ⎝ 2m<br />
⎠ k − m j=<br />
1<br />
N<br />
N<br />
⎛ Ck 0′ ,k⎞ A%<br />
0 k 0′ ,k0<br />
Q0<br />
= ∑⎜δ k { }<br />
0 ,m 1 2 ⎟+<br />
2 2<br />
k,m .<br />
j ∏ −δ (34)<br />
j<br />
j= 1 ⎝ 2m<br />
⎠ k0−m j=<br />
1<br />
λ0<br />
1<br />
A% k,k′<br />
= u() r exp{ −i( k −k′<br />
) r} dr.<br />
V ∫<br />
0<br />
λ0<br />
i<br />
Ck,k′ = kru() r exp{ i( k k′ ) r} dr.<br />
V ∫ − −<br />
0<br />
Then, according to the general theory of integral<br />
equations, the discrete energy distribution can be found,<br />
putting to sero systems determinants (31), (32):<br />
( ) k,k<br />
k,k′ +δ = 0.<br />
(35)<br />
Q ′<br />
Q ( 0 0) k 0,k0′ k,k′ +δ = 0.<br />
(36)<br />
Except for relations (35), (36) for finding of a discrete<br />
spectrum we can use a requirement of equality to<br />
zero of a determinant of Fredholm corresponding the<br />
equations (25), (26). If a determinant of systems (31),<br />
(32) or a determinant of Fredholm of the equations<br />
(25), (26) are not equal to zero, that, by consecutive<br />
iterations, from integral equations (25), (26) it is possible<br />
to find the wave functions corresponding an ionization<br />
continuum.<br />
3. WAVE FUNCTIONS OF A PARTICLE<br />
IN A MODELLING PERIODIC FIELD<br />
The modelling periodic field can be constructed<br />
as follows: all space should be broken into unit cells<br />
⎛a b c⎞<br />
with the size ⎜ ; ; ⎟.<br />
To put to each cell two vec-<br />
⎝2 2 2⎠<br />
tors in conformity: n = { nx; ny; nz}<br />
— an integer vector<br />
which defines a position of a cell in space and a vector<br />
which defines a potential energy of a particle in the<br />
given cell:<br />
{ α }<br />
б б<br />
wn(r) = h0 +θn(r) ⋅hexp β<br />
б<br />
Rn<br />
. (37)<br />
Where: h 0 And — the real constants;<br />
α1 α1<br />
R = x− T R = y− T .<br />
α1 α1<br />
n ;<br />
x nx<br />
ny ny<br />
The set of values of a vector is defined by equalities:<br />
б<br />
⎧+<br />
1<br />
α=⎨ .<br />
⎩−1<br />
Components of a vector for all cells are equal on<br />
absolute size, but different on a sign, that is: б в<br />
в б = { α1βx, α2βy, α3β z}<br />
, äå βx, βy, β z — the real<br />
stationary values.
The translation vector provides a value set identical<br />
to all cells for exponential curve argument in expression<br />
(37), and Heaviside vector function chooses<br />
for each unit cell an exponential curve with a corre-<br />
б б<br />
sponding vector T nбθn(r) β . The obvious view of vectors<br />
is given in addition C. Particle potential energy<br />
can be written down in the form of:<br />
∞<br />
W(r) = ∑∑ w (r).<br />
(38)<br />
n=<br />
0б<br />
The characteristic feature of the created model of<br />
a field is that the field in each cell consists only of one<br />
exponential curve and a constant component h 0 . At<br />
transition to the next cell the sign of one of components<br />
of vectors changes on opposite, and the module<br />
of component does not change в б . The exponent of<br />
exponential curve in expression (37) is a real number.<br />
However if to resolve function (38) in a Fourier series<br />
we shall gain the real translation vector for a reciprocal<br />
lattice that will be coordinated with the standard<br />
theory. Thus, there is no necessity to write down a<br />
Schrodinger equation for all space, it is enough to<br />
write it down for one cell and to impose corresponding<br />
boundary condition on its solutions. We shall write<br />
down a Schrodinger equation for one cell:<br />
б<br />
n<br />
2<br />
⎛ h<br />
б ⎞ б б<br />
− Δ+ wn ϕ n = Eϕn<br />
⎜ (r) ⎟ (r) (r). (39)<br />
⎝ 2μ<br />
⎠<br />
Let’s substitute potential energy (38) in the equation<br />
(39) and write down it in the form of:<br />
б б<br />
Δϕ n(r) = f (r) ϕ n(r).<br />
(40)<br />
Where such table of symbols are used:<br />
2 2μ<br />
2μ<br />
m = 2 ( E−h0) , A0 =− 2 ( E−h0) ,<br />
h<br />
h<br />
2μ<br />
α<br />
A= hf 2 () r = A0 + Aexp(<br />
βαRn)<br />
(41)<br />
h<br />
Particular solution of the equation (40) can<br />
be found using the solution (10) if to figure:<br />
k0 = k1 = k 2 = ... = −iβ α .<br />
It will be such:<br />
∞<br />
αα , ′ imбr<br />
ϕ n = e + ∑<br />
k = 1<br />
α { ( α′ α n ) }<br />
k k<br />
( −1) A exp i m r −ikβ R<br />
k<br />
∏(<br />
mα′<br />
−ilβα) l = 1<br />
Where: m б′ ( ′ 1m , ′ x 2my, ′ 3mz)<br />
.<br />
2<br />
.<br />
(42)<br />
= α α α With the view of<br />
satisfying the wave function found from the equation<br />
(40), to boundary conditions it is necessary to choose<br />
it in the form of such linear combination of particular<br />
solutions (42) as follows:<br />
б б,б′<br />
ϕ n = ∑ ϕn<br />
.<br />
б′<br />
Boundary conditions look like:<br />
(43)<br />
( na yz) ( na yz)<br />
ϕ ± , , =ϕ ± , , ;<br />
α1, α2, α3 −α1, α2, α3<br />
n x n<br />
x<br />
α1, α2, α⎛ na 3 x ⎞ −α1, α2, α⎛<br />
na<br />
3 x ⎞<br />
ϕ n ⎜± yz , =ϕ n ± yz , ;<br />
2<br />
⎟ ⎜<br />
2<br />
⎟<br />
⎝ ⎠ ⎝ ⎠<br />
( x, nyb, z) ( x, nyb, z)<br />
;<br />
ϕ ± =ϕ ±<br />
α1, α2, α3 α1, −α2, α3<br />
n n<br />
nb<br />
1, 2, y nb<br />
α α α⎛ ⎞ 3 α1, −α2, α⎛<br />
y ⎞<br />
3<br />
ϕ n ⎜x, ± , z⎟=ϕ n ⎜x, ± , z⎟;<br />
⎝ 2 ⎠ ⎝ 2 ⎠<br />
α1, α2, α3 ϕ xy , , ± nc<br />
α1, α2, −α3<br />
=ϕ xy , , ± nc;<br />
( ) ( )<br />
n z n<br />
z<br />
α1, α2, α⎛ nc 3 z ⎞ α1, α2, −α⎛<br />
nc<br />
3<br />
z ⎞<br />
ϕ n ⎜xy , , ± =ϕ n xy , , ± ;<br />
2<br />
⎟ ⎜<br />
2<br />
⎟<br />
⎝ ⎠ ⎝ ⎠<br />
We gain the next expressions from them:<br />
⎧1⎫ sin ⎨ ( α 1ma x +α 2mb<br />
y ) ⎬=<br />
0;<br />
⎩2⎭ ⎧1⎫ sin ⎨ ( α 2mb y +α 3mc<br />
z ) ⎬=<br />
0;<br />
⎩2⎭ ⎧1⎫ sin ⎨ ( α 3mc z +α 1ma<br />
z ) ⎬=<br />
0; (44)<br />
⎩2⎭ From which arises, that builders of a wave vector<br />
are real numbers, which quantize as follows:<br />
2πN<br />
2πN<br />
x<br />
y 2πN<br />
z<br />
mx<br />
= ; my<br />
= ; mz<br />
= . (45)<br />
a b c<br />
Where: N x — the arbitrary integers.<br />
Then the energy distribution corresponding the<br />
constructed wave function will be discrete:<br />
2 2 2 2 2<br />
2π<br />
h ⎛ N N x y N ⎞<br />
z<br />
EN= h0+<br />
⎜ + + .<br />
2 2 2 ⎟ (46)<br />
μ ⎜ a b c ⎟<br />
⎝ ⎠<br />
There is shown in appendix A, that the functional<br />
series of type (10) converge under condition of:<br />
A < m −iβ .<br />
(47)<br />
0<br />
2<br />
α′ α<br />
The inequality (47) can be copied in the form of:<br />
2 2<br />
h β<br />
h0 < E < h0<br />
+ . (48)<br />
2<br />
4μ( 1−2cos θ)<br />
Where: θ — It is a corner between vectors that m б .<br />
From a relation (48) arises existence of the forbidden<br />
and allowed directions of a particle wave vector<br />
so as the corner cannot accept value for which<br />
1<br />
θ cos θ ≥ . Thus, the particle can transfer in one<br />
2<br />
of the next cells with different probability. We can see<br />
from figure 1, that if the particle has energy in an interval:<br />
2 2<br />
h β<br />
h0 < E < .<br />
(49)<br />
4μ<br />
That the greatest quantity of the allowed moving<br />
directions exists for it . If value of energy will exceed<br />
critical value, that, with magnification of energy, the<br />
quantity of possible trajectories of a motion will sharply<br />
decrease and will aspire to zero for the energy considerably<br />
exceeding critical value.<br />
2 2<br />
h β<br />
E = .<br />
4μ<br />
Thus, in a modelling field, the particle actually<br />
has a narrow band , with degree of, a band of the allowed<br />
values of energy, definiendums (46) and (49)<br />
−1 −1<br />
~10 − 10 eB . In paper [6] on the basis of the ob-<br />
143
servational effects the possible model of energy states (46). It is shown, that a discrete series of energy levels<br />
of an amorphous cadmium diphosphide has been de- has a upper bound which depends on a direction of a<br />
veloped which will be coordinated with effects of ex- wave vector. In the created modelling potential field<br />
periment. The approach was used at build-up of this there is an opportunity of search of probability distri-<br />
model according to which the amorphous substance is bution for possible trajectories of a particle motion.<br />
considered as a crystal periodic field of which gets new The constructed modelling field is useful for exami-<br />
properties, owing to rather major concentration of nation of features of electromigration mechanisms in<br />
flaws. Infringement of stehiometry in (surplus) leads<br />
to formation of donor type centre in which the electron<br />
simultaneously is both in a field of flaws, and in a<br />
amorphous semiconductors.<br />
periodic field of a crystal CdP2Cd . Volumetric periodicity<br />
of an arrangement of such centres as experiment<br />
5. ADDITIONS<br />
has proved, leads to occurrence of the new periodic<br />
field, one of main features is existence of periodically<br />
located planes of reflecting symmetry that is characteristic<br />
for the modelling potential created in given<br />
paper. It enables to explain, leaning on an energy distribution<br />
of a particle in a modelling field, occurrence<br />
of the non-local energy levels in a forbidden region<br />
A. Expression transformation (9)<br />
Let’s group summands of series (9) on degrees of<br />
coefficient: 0<br />
amorphous CdP 2 . The flaws related to surplus, are the<br />
double donors, capable to donate one or two electrons<br />
Cd. They should create very narrow bands that will be<br />
coordinated with the effect gained from created model.<br />
Besides prompt diminution of quantity of the allowed<br />
particle mechanical trajectory, with increasing of its<br />
energy, can explain occurrence of conductance channels<br />
of in amorphous CdP 2 . The motion of particles<br />
with high energy on rather small fields of substance<br />
volume raises their temperature and as consequence,<br />
can cause local changes of structure amorphous CdP 2 .<br />
That is, the created modelling field gives one of possible<br />
explanations to a switching effect in amorphous<br />
semiconductors, supplementing the principal causes<br />
of its occurrence viewed in paper [7].<br />
E<br />
A<br />
⎧⎪ Ak exp{ ik}<br />
0<br />
0r<br />
imr ϕ= e ⎨1−<br />
∑<br />
+ 2<br />
⎪⎩ k0≠0 k+m 0<br />
Ak exp{ ik} { }<br />
0 0l Ak A<br />
0 k1exp is1r<br />
+ A0<br />
∑ + 2 ∑ ∑<br />
−<br />
2 2<br />
k0≠0 k+m 0<br />
k0≠0 k1≠0 k+m 0 s+m 1<br />
Ak exp{ ik} { }<br />
0 0r Ak A<br />
0 k1exp is<br />
2<br />
1r<br />
−A0∑ − A<br />
6 0∑<br />
∑ r 2 r − 4<br />
k0≠0 k+m 0<br />
k0≠0 k1≠0 k+m 0 s+m 1<br />
Ak A { }<br />
0 k1exp is1r<br />
−A0∑ ∑ r 4 r − 2<br />
k0≠0 k1≠0 k+m 0 s+m 1<br />
∏ 2exp{ is2r}<br />
⎫⎪<br />
− ∑ ∑ ∑<br />
+ ...<br />
2 2 2 ⎬ =<br />
k0≠0 k1≠0 k2≠0 k+m 0 s+m 1 s+m 2 ⎪⎭<br />
Ak exp{ ik} ( )<br />
0<br />
0rS1 k0<br />
= exp{ imr}{<br />
1−<br />
∑<br />
+<br />
2<br />
k ≠0<br />
k+m<br />
−π<br />
144<br />
3π<br />
−<br />
4<br />
π<br />
−<br />
2<br />
π<br />
−<br />
4<br />
2 2<br />
β<br />
h0<br />
+<br />
4μ<br />
�<br />
h0<br />
0<br />
π<br />
4<br />
π 3π<br />
π<br />
2 4<br />
Fig. 1. Dependence of the greatest possible energy of a particle<br />
on a corner, a definiendum (40). θ<br />
4. EFFECTS<br />
For search of the particle wave functions having<br />
an ionization continuum in a modelling periodic field,<br />
from integral equations of Fredholm type (25) and<br />
(26), source functions (22), (23) are offered, which allow<br />
to carry out search of wave functions with a continuance<br />
not only equal but also multiple to a continuance<br />
of a potential field. The possible discrete energy<br />
distribution can be defined from relations (35), (36).<br />
The modelling periodic field for which wave functions<br />
of a particle in an analytical view are gained is<br />
constructed. The energy distribution of a particle, in<br />
a modelling potential field, is defined by expression<br />
θ<br />
∑ ∑<br />
k0≠0 k1≠0 0<br />
∏<br />
0<br />
( )<br />
isr 1<br />
1e S2<br />
k 0,k1 2 2<br />
+ −<br />
k +m k +k +m<br />
0 0 1<br />
∏ 2exp{ is3r} S3(<br />
k 0,k 1,k2) ⎫⎪<br />
∑ ∑ ∑ ... (50)<br />
2 2 2<br />
− + ⎬<br />
k0≠0 k1≠0 k2≠0 k+m 0 s+m 1 s+m 2 ⎪⎭<br />
Let’s find quantities: S j( k 0,k 1,..,k j−<br />
1)<br />
S<br />
A A<br />
= − + −<br />
1( k0) 1<br />
0<br />
2<br />
k0 + m<br />
2<br />
0<br />
4<br />
k0 + m<br />
3 4<br />
A0 A0<br />
− + −...<br />
6 8<br />
k0 + m k0 + m<br />
(51)<br />
A series (51) is a geometrical progression which<br />
will converge under condition of:<br />
A 0 < k0 + m .<br />
(52)<br />
If the requirement (52) is carried out, quantity will<br />
k S<br />
be equal: ( )<br />
1 0<br />
S<br />
2<br />
k0+ m<br />
1( 0) =<br />
2<br />
k0 + m + A0<br />
k .<br />
2<br />
(53)<br />
Assumed, that the condition (52) is satisfied, we<br />
S<br />
k,k<br />
shall evaluate a value: ( )<br />
2 0 1
S<br />
A A A<br />
= − − + +<br />
1( k,k 0 1) 1<br />
0<br />
2<br />
k0 + m<br />
0<br />
2<br />
s1+ m<br />
2<br />
0<br />
4<br />
k0 + m<br />
2 3 3<br />
A0 A0 A0<br />
+ − − + ... +<br />
4 6 6<br />
s1+ m s1+ m k0 + m<br />
A ⎡ A A ⎤<br />
+ × ⎢1 − − + ... ⎥ =<br />
k m s m ⎢⎣ k m s m ⎥⎦<br />
0 +<br />
2<br />
0<br />
2<br />
1+ 2<br />
0<br />
0 +<br />
2<br />
0<br />
r 2<br />
1+<br />
2<br />
s1+ m A0<br />
= − +<br />
2 2<br />
s1+ m + A0 k0 + m + A0<br />
2<br />
A0<br />
+<br />
S<br />
2 2 2( k,k 0 1)<br />
.<br />
k0 + m s1+ m<br />
There is the linear algebraic equation for finding<br />
S k,k<br />
of it: ( )<br />
2 0 1<br />
S2(<br />
k,k 0 1) 2<br />
s1+ m<br />
2<br />
s1+ m + A0<br />
A0 2<br />
A0<br />
0 +<br />
2<br />
+ A0<br />
0 +<br />
2<br />
1+<br />
2<br />
= −<br />
( )<br />
−<br />
k m<br />
+<br />
k m s m<br />
S2<br />
k,k 0 1 . (54)<br />
We have from the equation (54) as follows:<br />
II<br />
k,k .<br />
0 1<br />
S2(<br />
0 1)<br />
=<br />
FF 0 1<br />
2<br />
Ij = s j + m<br />
Where:<br />
Similarly we find other values of sequence:<br />
j−1<br />
Il<br />
S j( k 0,k 1,...,k j−1)<br />
= ∏ . (55)<br />
l = 0 Fl<br />
Substituting expression (55) in (50) we gain:<br />
{ }<br />
⎛ Ak exp ik<br />
0<br />
0r<br />
ϕ= exp{ imr}<br />
⎜1−<br />
+<br />
⎜ ∑<br />
⎝ k0<br />
≠0<br />
F0<br />
∏ 1exp{ is1r}<br />
⎞<br />
∑ ∑ ... . (56)<br />
+ − ⎟<br />
k0≠0 k1≠0<br />
FF ⎟⎠<br />
0 1<br />
B. Relation check (13)<br />
Acting on both parts of expression (13) by func-<br />
2<br />
Δ+ m<br />
tional we have: ( )<br />
( )<br />
k 0 ,k<br />
k 0 ,k<br />
∑<br />
{ ( + ) }<br />
2 2 k0k 0<br />
Δ+ m ϕ=−m −<br />
2 2<br />
k 0 ,k k0+ k −m<br />
( Δ exp{ ( k + k)<br />
} )<br />
k0k 0<br />
2 2<br />
k0+ k −m<br />
{ ( + ) }<br />
2 k0k 0<br />
k 0 ,k<br />
2 2<br />
k0+ k −m<br />
k0k 0 +<br />
2<br />
0 +<br />
2 2<br />
−m<br />
{ ( ) }<br />
k0k 0<br />
A B exp i k k r<br />
A B i<br />
− =<br />
∑<br />
A B exp i k k r<br />
=− m<br />
+<br />
∑<br />
A B k<br />
+ ∑<br />
k 0 ,kk<br />
k<br />
k<br />
exp{ i(<br />
k0 + k) r}<br />
=<br />
= A B exp i k + k r = u(r)<br />
ϕ(r).<br />
∑<br />
C. The functions used in model of a periodic field<br />
The label θξ () — corresponds to function of<br />
Heaviside. The table of symbols — corresponds to<br />
Heaviside function.<br />
⎧+<br />
1<br />
б = ( α1; α2; α 3)<br />
; α i = ⎨ ;<br />
⎩−1<br />
б б б<br />
w (r) =θ (r) Aexp β R (r) ;<br />
б<br />
n<br />
( α )<br />
n n n<br />
α1 α2<br />
α3<br />
( Rn x R )<br />
x n y R<br />
y n z<br />
z<br />
R (r) = ( ); ( ); ( ) ;<br />
( x y z)<br />
б α1 α2<br />
α3<br />
в = α β , α β , α β ; θ (r) =θ ( x) θ ( y) θ ( z);<br />
б 1 2 3<br />
n<br />
( 1+α<br />
)<br />
nx ny nz<br />
α ⎛<br />
1<br />
1 ⎞<br />
Tn =γ 1 ( ) ;<br />
x xa⎜nx + −αθx<br />
⎟<br />
⎝ 2 ⎠<br />
α ⎛ ( 1+α<br />
)<br />
2<br />
2 ⎞<br />
Tn =γ 2 ( ) ;<br />
y yb⎜ny + −αθ y ⎟<br />
⎝ 2<br />
⎠<br />
α ⎛ ( 1+α<br />
)<br />
3<br />
3 ⎞<br />
Tn =γ 3 ( ) ;<br />
z zc⎜nz + −αθ z ⎟<br />
⎝ 2<br />
⎠<br />
γ x , are functions of a coordinates sign, accordingly;<br />
( 1−α<br />
)<br />
α ⎧⎪ 1<br />
1 ⎫⎪<br />
θ n ( x) = ⎨ +α1θ( x) ⎬θ(<br />
x − an )<br />
x<br />
x ×<br />
⎪⎩ 2 ⎪⎭<br />
( 1+α<br />
)<br />
⎛ ⎛ 1 ⎞ ⎞ ⎧⎪ 1 ⎫⎪<br />
×θ ⎜a⎜nx+ ⎟−<br />
x ⎟+<br />
⎨ − α1θ ( x)<br />
⎬×<br />
⎝ ⎝ 2⎠ ⎠ ⎪⎩ 2 ⎪⎭<br />
⎛ ⎛ 1 ⎞⎞<br />
×θ⎜ x − a⎜nx + ⎟⎟θ<br />
( a( nx + 1 ) − x)<br />
;<br />
⎝ ⎝ 2 ⎠⎠<br />
α ⎧⎪( 1−α<br />
)<br />
2<br />
2 ⎫⎪<br />
θ n ( y) = ⎨ +α2θ( y) ⎬θ(<br />
y − bn )<br />
y<br />
y ×<br />
⎪⎩ 2<br />
⎪⎭<br />
⎛ ⎛ 1 ⎞ ⎞ ⎧⎪( 1+α2)<br />
⎫⎪<br />
×θ ⎜b⎜ny+ ⎟−<br />
y ⎟+<br />
⎨ − α2θ ( y)<br />
⎬×<br />
⎝ ⎝ 2⎠ ⎠ ⎪⎩ 2<br />
⎪⎭<br />
⎛ ⎛ 1 ⎞⎞<br />
×θ⎜ y − b⎜ny + ⎟⎟θ<br />
( b( ny + 1 ) − y)<br />
;<br />
⎝ ⎝ 2 ⎠⎠<br />
( 1−α<br />
)<br />
α ⎧⎪ 3<br />
3 ⎫⎪<br />
θ n ( z) = ⎨ +α3θ( z) ⎬θ(<br />
z − cn )<br />
z<br />
z ×<br />
⎪⎩ 2 ⎪⎭<br />
( 1+α<br />
)<br />
⎛ ⎛ 1 ⎞ ⎞ ⎧⎪ ×θ ⎜c⎜nz+ ⎟−<br />
z ⎟+<br />
⎨<br />
⎝ ⎝ 2⎠ ⎠ ⎪⎩ 2<br />
3 ⎫⎪<br />
− α3θ ( z)<br />
⎬×<br />
⎪⎭<br />
⎛ ⎛ 1 ⎞⎞<br />
×θ⎜ z − c⎜nz + ⎟⎟θ<br />
( c( nz + 1 ) − z)<br />
.<br />
⎝ ⎝ 2 ⎠⎠<br />
D. The proof of relations (24)<br />
(57)<br />
Operating with the functional on function (22) we<br />
2<br />
Δ+ m<br />
gained: ( )<br />
G<br />
( r, )<br />
( mrexp % j { im<br />
j(<br />
r ) } )<br />
N Δ −ρ<br />
ρ = − −<br />
∑<br />
0<br />
j=<br />
1<br />
2<br />
2Vm<br />
N<br />
2<br />
( mrexp %<br />
j { m j(<br />
r−ρ)<br />
} )<br />
m<br />
−∑ j=<br />
1<br />
i<br />
2<br />
2Vm<br />
−<br />
⎛ 2<br />
⎞<br />
1 ⎜ Δexp{ ik( r −ρ) } m exp{ ik(<br />
r −ρ)<br />
} ⎟<br />
− 2 2 2 2<br />
V ⎜ ∑ + ∑<br />
k k m k k m ⎟<br />
=<br />
⎜ − − ⎟<br />
⎝ k ≠ m k ≠ m<br />
⎠<br />
145
146<br />
N 1 1<br />
= ∑exp{ im j { r −ρ } + ∑ exp{ ik(<br />
r −ρ ) } =<br />
V V<br />
j=<br />
1 k<br />
k ≠ m<br />
1<br />
= ∑exp{<br />
ik(<br />
r −ρ ) } =δ( r −ρ)<br />
.<br />
V k<br />
Similarly we can gain a relation (24) for function<br />
G ρ .<br />
( )<br />
0 r,<br />
References<br />
1. Àâäîí³í Ê. Â. Çíàõîäæåííÿ ðîçâ’ÿçê³â ³íòåãðàëüíèõ òà<br />
äèôåðåíö³àëüíèõ ð³âíÿíü çà äîïîìîãîþ îïåðàòîð³â<br />
çì³ùåííÿ // Ìàòåð³àëè ì³æíàðîäíî¿ íàóêîâî-òåõí³÷íî¿<br />
êîíôåðåíö³¿, ïðèñâÿ÷åí³é 90 — ð³÷÷þ ç äíÿ íàðîäæåííÿ<br />
Ã. À. ϳñêîðñüêîãî, 2003., Ñ. 139 — 142.<br />
2. Áîðèñîâ Ä. È., Ãàäûëüøèí Ð. Ð. Î ñïåêòðå îïåðàòîðà<br />
Øðåäèíãåðà ñ áûñòðî îñöèëëèðóþùèì ïîòåíöèàëîì //<br />
ÒÌÔ, ò.147, ¹1, 2006.<br />
3. Äæîíñ Ã. Òåîðèÿ çîí Áðèëëþýíà è ýëåêòðîííûå<br />
ñîñòîÿíèÿ â êðèñòàëëàõ. — Ì., Ìèð, 2000. — 564 ñ.<br />
4. Ìåíòêîâñêèé Þ. Ë. ×àñòèöà â ÿäåðíî-êóëîíîâñêîì<br />
ïîëå. — Ì., Íàóêà, 2002. — 216 ñ.<br />
5. Áîãîëþáîâ Í. Í., Øèðêîâ Ä. Â. Ââåäåíèå â òåîðèþ<br />
êâàíòîâàííûõ ïîëåé. — Ì., 2001. — 442 ñ.<br />
6. Àâäîí³í Ê. Â., Êëèìåíêî À. Ï., Ëàïøèí Â. Ô., Â. Ê.<br />
Ìàêñèìîâ Â. Ê. Åëåêòðîíí³ ÿâèùà ïåðåíîñó â øàð³<br />
óëüòðàäèñïåðñíîãî äèôîñô³äó êàäì³ÿ // Òåçè äîïîâ³äåé<br />
XXII íàóêîâî¿ êîíôåðåíö³¿ êðà¿í ÑÍÄ “Äèñïåðñí³<br />
ñèñòåìè”, Îäåñà, 2006., Ñ.25 — 26.<br />
7. Êîñòûë¸â Ñ. À., Øêóò Â. À. Ýëåêòðîííîå ïåðåêëþ÷åíèå â àìîðôíûõ ïîëóïðîâîäíèêàõ. — Ê., Íàóêîâà äóìêà, 1999. — 204<br />
ñ.<br />
UDC.539.125.5<br />
K. V. Avdonin<br />
BUILD-UP OF WAVE FUNCTIONS OF THE PARTICLE IN THE MODELLING PERIODIC FIELD<br />
Abstract<br />
A movement pattern of charged free particles inside of a crystal is proposed in this work. Effective potential field of interaction<br />
between particles and crystal lattice point is considered in two presentations: as three-dimensional Fourier series and as the presentation<br />
of effective field, in which its periodicity is made by Heviside’s function. Wave functions are found by composition of individual solutions<br />
of Schrödinger equation in the form of functional series. A possibility of existence of energy levels of foreign particle’s condition in<br />
crystal appropriate to them is researched.<br />
Key words: individual solutions, periodic field.<br />
ÓÄÊ.539.125.5<br />
Ê. Â. Àâäîíèí<br />
ÏÎÑÒÐÎÅÍÈÅ ÂÎËÍÎÂÛÕ ÔÓÍÊÖÈÉ ×ÀÑÒÈÖÛ Â ÌÎÄÅËÜÍÎÌ ÏÅÐÈÎÄÈ×ÅÑÊÎÌ ÏÎËÅ<br />
Ðåçþìå<br />
 äàííîé ðàáîòå ðàññìàòðèâàåòñÿ äâèæåíèå ÷àñòèöû â ìîäåëüíîì ïåðèîäè÷åñêîì ïîëå, êîòîðîå ðàññìàòðèâàåòñÿ â òðàäèöèîííîì<br />
âèäå, òî åñòü â ïðåäñòàâëåíèè ðÿäà Ôóðüå, è â òàêîì ïðåäñòàâëåíèè, â êîòîðîì åãî ïåðèîäè÷íîñòü ñîçäà¸òñÿ ôóíêöèÿìè<br />
Õåâèñàéäà. Ðàññìîòðåíà âîçìîæíîñòü íàõîæäåíèÿ âîëíîâûõ ôóíêöèé ÷àñòèöû ïóò¸ì ïîñòðîåíèÿ ÷àñòíûõ ðåøåíèé<br />
êëàññè÷åñêîãî âîëíîâîãî óðàâíåíèÿ â âèäå ôóíêöèîíàëüíûõ ðÿäîâ.<br />
Êëþ÷åâûå ñëîâà: âîëíîâûå ôóíêöèè, ìîäåëüíîå ïåðèîäè÷åñêîå ïîëå.<br />
ÓÄÊ.539.125.5<br />
Ê. Â. Àâäîí³í<br />
ÏÎÁÓÄÎÂÀ ÕÂÈËÜÎÂÈÕ ÔÓÍÊÖ²É ÄËß ×ÀÑÒÈÍÊÈ Ó ÌÎÄÅËÜÍÎÌÓ ÏÅвÎÄÈ×ÍÎÌÓ ÏÎ˲<br />
Ðåçþìå<br />
 äàí³é ðîáîò³ ðîçãëÿäàºòüñÿ ðóõ çàðÿäæåíî¿ ÷àñòèíêè ó ìîäåëüíîìó ïåð³îäè÷íîìó ïîë³, ÿêå ïðåäñòàâëåíå ó òðàäèö³éíîìó<br />
âèãëÿä³, òîáòî ó âèãëÿä³ ðÿäó Ôóð’º, òà ó òàêîìó ïðåäñòàâëåíí³, â ÿêîìó éîãî ïåð³îäè÷í³ñòü ñòâîðþºòüñÿ ôóíêö³ÿìè Õåâ³ñàéäà.<br />
Ðîçãëÿíóòà ìîæëèâ³ñòü çíàõîäæåííÿ õâèëüîâèõ ôóíêö³é ÷àñòèíêè ó ìîäåëüíîìó ïîë³ øëÿõîì ïîáóäîâè ÷àñòèííèõ<br />
ðîçâ’ÿçê³â êëàñè÷íîãî õâèëüîâîãî ð³âíÿííÿ ó âèãëÿä³ ôóíêö³îíàëüíèõ ðÿä³â.<br />
Êëþ÷îâ³ ñëîâà: õâèëüîâ³ ôóíêö³¿, ìîäåëüíå ïåð³îäè÷íå ïîëå.
UDC 621.315.59<br />
V. A. ZAVADSKY, G. S. POPIK<br />
Odessa National Maritime Academy, 8, Didrikhsona, Odessa, Ukraine.<br />
e-mail: vaaz@ukr.net<br />
MODIFICATION OF PARAMETERS IR-FOTODETECTORS BY HIGH-<br />
ENERGY PARTICLES<br />
Irradiation influence by high-energy particles, by γ-quantums and fast electrons on parameters<br />
of IR-photodetectors on the basis of triple semi-conductor compound cadmium-mercury-tellurium<br />
(CMT) is investigated.<br />
Radiating firmness and modification possibility electro- and photo-electric parameters are investigated:<br />
the Hall constant, photosensitivity, dark resistance and current of photoconductive cell by<br />
CMT.<br />
INTRODUCTION<br />
The development and prospective trends of solid<br />
state electronics are concerned with mulicomponent<br />
single and polycrystalline materials. Mulicomponent<br />
materials can be prepared as blocks and films<br />
with different thickness. CdHgTe is one of above<br />
mentioned materials. The most useful application of<br />
CdHgTe is fabrication of IR- detectors, what forms<br />
the base of optoelectronics. This material is also<br />
prospective compound for fabrication of semiconductor<br />
image signal devices and charge connection<br />
devices [1].<br />
Electrical properties of CdHgTe are affected by<br />
biographic point defects (vacancies, complexes, and<br />
interstate atoms), longing (dislocations, grain boundaries,<br />
metal inclusions) and non desirable admixtures.<br />
Because of strong interactions and non ideal control<br />
all defect parameters CdHgTe has great deviation of<br />
properties. At the same time, the devices, based on<br />
CdHgTe, have non stable parameters what decreases<br />
their reliability and working time.<br />
Mulicomponent materials can be prepared as<br />
blocks and films with different thickness. The fabrication<br />
methods, their electrical and structural properties<br />
have been deeply studied. The applications of<br />
mulicomponent materials in quantum and nano electronics,<br />
optoelectronics and electro technique have<br />
also been worked out. However, radiation physics and<br />
technology of mulicomponent materials are under the<br />
level of their practical applications.<br />
EXPERIMENTAL<br />
Investigation of n-type Cd x Hg 1-x Te solid state solutions<br />
treated by electron and γ-radiation showed<br />
that Hall’s mobility and resistivity temperature dependences<br />
didn’t change at 77-300 Ê after irradiation with<br />
electrons, having enegry 4,5 ÌeV on linear accelerator<br />
with intensity 10 12 cm -2 with streams 10 15 — 5∙10 16<br />
cm -2 . Treatment with γ-radiation from Co 60 source<br />
with doze of 10 6 — 10 7 rentgen led to binding of the<br />
Hall’s mobility curve at 100-110 Ê. Additional irra-<br />
© V. A. Zavadsky, G. S. Popik, 2009<br />
diation shifted the binding to the higher temperature<br />
range.<br />
The objects of research were IR-photresistors<br />
based on single crystalline CdHgTe. Taking in account<br />
the quality of bulk single crystals on structural parameters<br />
(dislocation density, Hg vacancy concentration)<br />
and conductivity (charge carriers concentration, their<br />
mobility) is better than the quality of epitaxial layers,<br />
bulk n-type single crystals with x=0,2 have been chosen<br />
for photo resistor fabrication [2].<br />
The source of high energy electrons (3,5 MeV)<br />
was linear accelerator Electronica ELA-4. Integral<br />
electrons stream was calculated according to [2]:<br />
Ô = 6,2∙1012I t ý<br />
where I — current of high energy electrons stream,<br />
ý<br />
μÀ, t — exposition time, sec.<br />
The treatment profile has been chosen with restriction<br />
of samples heating up to 600Ñ. The investigation of γ-radiation influence to Cd-<br />
HgTe lattice constant resulted to conclusion that the<br />
most radiative stable compound is the one with the<br />
lowest Hg vacancy concentration, i.e. having Hg limit<br />
of homogeniousity [3].<br />
Investigation of natural and radiative ageing of<br />
CdHgTe epitaxial layers (by means of Hall’s mobility<br />
temperature dependence measurements) showed<br />
that the maximum of this dependence enhanced and<br />
shifted to higher temperatures for samples, naturally<br />
aged for 1,5 years (fig.1) [4].<br />
Treatment with 3,6 . 108 rentgen γ-radiation<br />
changed the material properties in a such way that<br />
in 7 hour of ageing after irradiation the maximum of<br />
R (1/T) shifted to lower temperatures and in 30 hours<br />
x<br />
it position was almost constant (fig.2). Ageing of the<br />
irradiated samples during 1-7 days entailed with decreasing<br />
resistivity and deterioration of photo electrical<br />
parameters (fig.3) [4].<br />
Under investigation of non stability and hysteretic<br />
phenomena in MOS structures based on CdHgTe<br />
it was found that the change of their characteristics<br />
is tailored by the oxide-semiconductor interface and<br />
neighbor regions where surface states and traps in<br />
semiconductor band gap are located [5].<br />
147
148<br />
Rx, cm 3 , C -1<br />
Fig. 1. Temperature dependence of R x epitaxial layers CdHgTe<br />
[4]: 1 — before ageing; 2 — after ageing during 1,5 years<br />
Rx, cm 3 , C -1<br />
Fig. 2. Temperature dependence of R x for epitaxial layers after<br />
3,6 . 10 8 rentgen irradiation with γ-radiation [4]: 1- before irradiation;<br />
2 — in 7 hours after iradiation; 3 — in 31 hours after iradiation;<br />
4 — in 55 hours after iradiation ÷åðåç<br />
Fig. 3. Phot response of the samples befor and after 3,6 . 10 8<br />
rentgen irradiation with γ-radiation: 1 — before irradiation; 2 — in<br />
1 day; 3 — in 2 days; 4 — in 7 days<br />
RESULTS AND DISCUSSION<br />
The influence of high energy electrons (25 ÌeV)<br />
on electro physical properties of n and p-type Cd Hg x 1-<br />
Te (x = 0,2) has been investigated. Irradiation of pxtype<br />
samples with 1019 m-2 stream resulted in inversion<br />
point shift to lower temperatures approximately to<br />
10 0C what points to the increase of donor concentration.<br />
The n-type samples with initial concentration<br />
of charge carriers 2∙1021 m-3 have been exposed<br />
consequently to 1019 è 1020 m-2 at 77 K. Annealing<br />
was observed at room temperature. It was found that<br />
with increase of irradiation doze electron concentration<br />
increased and resistivity decreased. The mobility<br />
decreased at the initial point but it increased starting<br />
with stream 1019 m-2 . The charge carrier concentration<br />
increased proportionally to square root from electron<br />
stream. The annealing at room temperature pointed to<br />
the fact that charge carrier concentration changed exponentially<br />
with time, i.e. according to the first order<br />
reaction [8, 9].<br />
The irradiation of the same samples with stream<br />
3∙1019 m-2 showed that electro physical parameters of<br />
the samples have not been changed up to stream 1019 m-2 as in the first case, but then the charge carrier<br />
concentration sharply increased. It says about thermal<br />
stability of some defects, formed under irradiation.<br />
The Hall’s constant increased after 15 hours annealing.<br />
Its temperature dependence pointed to the<br />
presence of admixture conductance band. Temperature<br />
dependence of charge carrier mobility μ(T) was<br />
also changed, and its decline in low temperature region<br />
pointed to stronger scattering than as for point<br />
defects. The dependence R(T) has not been changed<br />
in 10 days after irradiation whereas admixture conductance<br />
band disappeared. Next changes of parameters<br />
at room temperature have not been found, so additional<br />
20 minutes annealing in vacuum at 100 and<br />
1500Ñ has been performed. After that charge carrier<br />
concentration was restored and the mobility restored<br />
partially [8].<br />
Under irradiation of Cd Hg Te with low en-<br />
0,2 0,8<br />
ergy radiation Frenkel’s pair formation is the post<br />
possible process. The initial defect structure restored<br />
immediately because of Frenkel’s pairs after<br />
the irradiation off. Their concentration is compared<br />
with interstate Hg atoms concentration, what can<br />
be recognized as increase of electron concentration.<br />
The structure restore rate is defined by diffusion<br />
coefficient of the most mobile defect (Hg ), which<br />
i<br />
depends on temperature and material structure perfection<br />
[6].<br />
The increase of irradiation stream and temperature<br />
can bring to Frenkel’s pairs concentration growth<br />
and point defect complexes formation. The recombination<br />
of the latest needs higher temperatures.<br />
Investigation of γ-radiation influence on lattice<br />
constant of powder materials (CdHg)Te showed that<br />
the most radioactive stable material is the solution of<br />
those components on Hg — the boundary of homogeneity<br />
region, i.e. with minimal concentration of<br />
Hg vacancies. The change of solid state properties<br />
can be concerned with radioactive defect formation,<br />
similar to high temperature defects. They represent
difficult Te x complexes with concentration, proportional<br />
Hg vacancy concentration. The separation Te x<br />
phase takes place with the increase of Hg vacancy<br />
concentration [7].<br />
It has been studied for photo resistors the modification<br />
of two main parameters dark current I T and<br />
dark resistivity with fast electrons irradiation.<br />
Measurement errors of the above mentioned parameters<br />
of photo resistors were up to ± 5%, with<br />
probability P=0,95. The results of measurements and<br />
calculations are presented in fig.4. Those results show<br />
following statements:<br />
– fast electron irradiation with 10 13 -10 15 cm -2 doze<br />
led to decrease of dark resistivity and increase of dark<br />
current ;<br />
– relative changes of dark resistivity and dark current<br />
bigger at 80 K;<br />
– For doze higher 10 15 cm -2 relative changes of<br />
photo resistors parameters differ little.<br />
Fig. 4. Doze dependence of photo resistor’s dark current (�)<br />
and dark resistivity (�) at 120 K (---) and at 80K (—)<br />
UDC 621.315.59<br />
V. A. Zavadsky, G. S. Popik<br />
CONCLUSION<br />
Irradiation of photo detectors based on CdHgTe<br />
changes their parameters: sensitivity limit (for its enhance<br />
it is necessary to decrease charge carrier concentration<br />
and increase effective quantum efficiency),<br />
recovery time (for its decrease it is worth to decline<br />
charge carrier life time ) and photo resistor spectral<br />
sensitivity range.<br />
References<br />
1. Ëåíêîâ Ñ. Â., Ìîêðèöêèé Â. À., Ãàðêàâåíêî À. Ñ., Çóáàðåâ<br />
Â. Â., Çàâàäñêèé Â. À. Ðàäèàöèîííîå óïðàâëåíèå<br />
ñâîéñòâàìè ìàòåðèàëîâ è èçäåëèé îïòî- è ìèêðîýëåêòðîíèêè.<br />
Ìîíîãðàôèÿ. — Îäåññà: 2003. — 345ñ.<br />
2. Çàâàäñêèé Â. À., Ìàñåíêî Á. Ï. Âëèÿíèå îáëó÷åíèÿ íà<br />
ïàðàìåòðû êðåìíèåâûõ ýëåìåíòîâ/ Â ñá.: Ìîëîäåæü<br />
òðåòüåãî òûñÿ÷åëåòèÿ: ãóìàíèòàðíûå ïðîáëåìû è ïóòè<br />
èõ ðåøåíèÿ. Ñåð. Ýêîíîìèêà, ìîäåëèðîâàíèå òåõíè-<br />
÷åñêèõ è îáùåñòâåííûõ ïðîöåññîâ, èíôîðìàöèîëîãèÿ,<br />
ýêîëîãèÿ. — Îäåññà: 2003. — Ò. Ç. — Ñ. 236-241.<br />
3. Èíäåíáàóì Ã.Â. è äð. Âëèÿíèå îáëó÷åíèÿ ãàììà-êâàíòàìè<br />
íà ïåðèîä ðåøåòêè òâåðäûõ ðàñòâîðîâ Cd Hg Te // Òåç.<br />
x 1-x<br />
äîêë. I Âñåñîþçí. íàó÷í.-òåõí. êîíô. “Ïîëó÷åíèå è ñâîéñòâà<br />
ïîëóïðîâîäíèêîâûõ ñîåäèíåíèé òèïà ÀÏÂYI è AIYBYI è<br />
òâåðäûõ ðàñòâîðîâ íà èõ îñíîâå. — Ì., 1997. — Ñ. 115.<br />
4. Îòðàáîòêà íàó÷íûõ îñíîâ òåõíîëîãèè è ïàðàìåòðîâ<br />
ïðîöåññà ãëóáîêîé î÷èñòêè èñõîäíûõ âåùåñòâ, ñèíòåçà<br />
ñîåäèíåíèé è âûðàùèâàíèÿ ñîâåðøåííûõ ìîíîêðèñòàëëîâ<br />
è ïëåíîê õàëüêîãåíèäîâ è îêèñëîâ òÿæåëûõ<br />
öâåòíûõ ìåòàëëîâ: Îò÷åò î ÍÈÐ (/ Ìîñêîâñêèé èí-ò<br />
ñòàëè è ñïëàâîâ Ðóê. Âàíþêîâ À.Â. — ¹ ÃÐ 77067934;<br />
Èíâ. ¹ Â787041. — Ì., 1999.<br />
5. Ëèòîâ÷åíêî Â.Ã. Òðåõñëîéíàÿ ìîäåëü ñòðóêòóð ïîëóïðîâîäíèê-äèýëåêòðèê<br />
// Ïîëóïðîâîäíèêîâàÿ òåõíèêà è<br />
ìèêðîýëåêòðîíèêà. — 1998. — Âûï. 12. — Ñ. 3-15.<br />
6. Êðåãåð Ô. Õèìèÿ íåñîâåðøåííûõ êðèñòàëëîâ. — Ì.:<br />
Ìèð, 1999. — 218 ñ.<br />
7. Çàèòîâ Ô.À. è äð. Ïîëóïðîâîäíèêè ñ óçêîé çàïðåùåííîé<br />
çîíîé è ïîëóìåòàëëû // Ìàòåðèàëû IY Âñåñîþçíîãî ñèìïîçèóìà.<br />
— Ëüâîâ: Âèùà øêîëà, 1995. — 4.5. — Ñ. 14-17.<br />
8. Çàèòîâ Ô.À., Ìóõèíà Î.Â., Ïîëÿêîâ À.ß. Îáëó÷åíèå<br />
òâåðäîãî ðàñòâîðà CdTe — HgTe ýëåêòðîíàìè ñ ýíåðãèåé<br />
25 ÌýÂ.: Ñá. “Òåõíèêà ðàäèàöèîííîãî ýêñïåðèìåíòà”.<br />
— Ì.: Àòîìèçäàò, 1997. — Âûï. 5. — Ñ. 34.<br />
9. Çàèòîâ Ô.À. è äð. Äåéñòâèå ïðîíèêàþùåé ðàäèàöèè íà<br />
ýëåêòðîôèçè÷åñêèå ïàðàìåòðû ïîëóïðîâîäíèêîâîãî<br />
ñïëàâà Cd Hg Te– Ê.: 1996. — 56 ñ. (Ïðåïð. Ðàäèàöè-<br />
x 1-x<br />
îííûå ýôôåêòû â ïîëóïðîâîäíèêîâûõ ñîåäèíåíèÿõ<br />
ÊÈßÈ ÀÍ Óêðàèíû, 76-22).<br />
MODIFICATION OF PARAMETERS IR-FOTODETECTORS BY HIGH-ENERGY PARTICLES<br />
Abstract<br />
Irradiation influence by high-energy particles, by γ-quantums and fast electrons on parameters of IR-photodetectors on the basis<br />
of triple semi-conductor compound cadmium-mercury-tellurium (CMT) is investigated.<br />
Radiating firmness and modification possibility electro- and photo-electric parameters are investigated: the Hall constant, photosensitivity,<br />
dark resistance and current of photoconductive cell by CMT.<br />
Key words: high energy particles, three elements compound, photo resistor, modification of parameters, irradiation influence.<br />
149
150<br />
ÓÄÊ 621.315.59<br />
Â. À. Çàâàäñêèé, Ã. Ñ. Ïîïèê<br />
ÌÎÄÈÔÈÊÀÖÈß ÏÀÐÀÌÅÒÐΠÈÊ-ÔÎÒÎÏÐÈÅÌÍÈÊΠÂÛÑÎÊÎÝÍÅÐÃÅÒÈ×ÅÑÊÈÌÈ ×ÀÑÒÈÖÀÌÈ<br />
Ðåçþìå<br />
Èññëåäîâàíî âëèÿíèå îáëó÷åíèÿ âûñîêîýíåðãåòè÷åñêèìè ÷àñòèöàìè, γ-êâàíòàìè è áûñòðûìè ýëåêòðîíàìè íà ïàðàìåòðû<br />
ÈÊ-ôîòîïðèåìíèêîâ íà îñíîâå òðîéíîãî ïîëóïðîâîäíèêîâîãî ñîåäèíåíèÿ êàäìèé-ðòóòü-òåëëóð (ÊÐÒ).<br />
Èññëåäîâàíû ðàäèàöèîííàÿ ñòîéêîñòü è âîçìîæíîñòü ìîäèôèêàöèè ýëåêòðî- è ôîòîýëåêòðè÷åñêèõ ïàðàìåòðîâ: ïîñòîÿííîé<br />
Õîëëà, ôîòî÷óâñòâèòåëüíîñòè, òåìíîâûõ ñîïðîòèâëåíèé è òîêà ôîòîðåçèñòîðîâ íà ÊÐÒ.<br />
Êëþ÷åâûå ñëîâà: âûñîêîýíåðãåòè÷åñêèå ÷àñòèöû, òðîéíîå ñîåäèíåíèå, ôîòîðåçèñòîð, ìîäèôèêàöèÿ ïàðàìåòðîâ, âëèÿíèå<br />
îáëó÷åíèÿ.<br />
ÓÄÊ 621.315.59<br />
Â. À. Çàâàäñüêèé, Ã. Ñ. Ïîï³ê<br />
ÌÎÄÈÔ²ÊÀÖ²ß ÏÀÐÀÌÅÒв ²×-ÔÎÒÎÏÐÈÉÌÀײ ÂÈÑÎÊÎÅÍÅÐÃÅÒÈ×ÍÈÌÈ ×ÀÑÒÊÀÌÈ<br />
Ðåçþìå<br />
Äîñë³äæåíî âïëèâ îïðîì³íåííÿ âèñîêîåíåðãåòè÷íèìè ÷àñòêàìè, γ-êâàíòàìè òà øâèäêèìè åëåêòðîíàìè íà ïàðàìåòðè<br />
²×-ôîòîïðèéìà÷³â íà îñíîâ³ ïîòð³éíîãî íàï³âïðîâ³äíèêîâîãî ç’ºäíàííÿ êàäì³é-ðòóòü-òåëóð (ÊÐÒ).<br />
Äîñë³äæåíà ðàä³àö³éíà ñò³éê³ñòü ³ ìîæëèâ³ñòü ìîäèô³êàö³¿ åëåêòðî- òà ôîòîåëåêòðè÷íèõ ïàðàìåòð³â: ïîñò³éíî¿ Õîëëà,<br />
ôîòî÷óòëèâîñò³, òåìíîâèõ îïîð³â ³ ñòðóìó ôîòîðåçèñòîð³â íà ÊÐÒ.<br />
Êëþ÷îâ³ ñëîâà: âèñîêîåíåðãåòè÷í³ ÷àñòèíêè, ïîòð³éíå ç’ºäíàííÿ, ôîòîðåçèñòîð, ìîäèô³êàö³ÿ ïàðàìåòð³â, âïëèâ îïðîì³íåííÿ.
“Photoelectronics” Articles publishes the papers<br />
which contain information about scientific research<br />
and technical designs in the following areas:<br />
• Physics of semiconductors;<br />
• Physics of microelectronic devices;<br />
• Linear and non-linear optics of solids;<br />
• Optoelectronics and optoelectronic devices;<br />
• Quantum electronics;<br />
• Sensorics.<br />
“Photoelectronics” Articles is defined by the decision<br />
of the Highest Certifying Commission as the<br />
specialized edition for physical-mathematical and<br />
technical sciences and published and printed at the<br />
expense of budget items of Odessa I.I. Mechnikov National<br />
University.<br />
“Photoelectronics” Articles is published in English.<br />
Authors send two copies of papers in English. The<br />
texts are accompanied by 3.5” diskette with text file,<br />
tables and figures. Electronic copy of a material can be<br />
sent by e-mail to the Editorial Board and should meet<br />
the following requirements:<br />
1.The following formats are applicable for texts —<br />
MS Word (rtf, doc).<br />
2. Figures should be made in formats — EPS,<br />
TIFF, BMP, PCX, JPG, GIF, WMF, MS Word I MS<br />
Giaf, Micro Calc Origin (opj). Figures made by packets<br />
of mathematical and statistic processing should be<br />
converted into the foregoing graphic formats.<br />
 ñáîðíèêå “Ôîòîýëåêòðîíèêà” ïå÷àòàþòñÿ<br />
ñòàòüè, êîòîðûå ñîäåðæàò ñâåäåíèÿ î íàó÷íûõ èññëåäîâàíèÿõ<br />
è òåõíè÷åñêèõ ðàçðàáîòêàõ â íàïðàâëåíèÿõ:<br />
* ôèçèêà ïîëóïðîâîäíèêîâ;<br />
* ãåòåðî- è íèçêîðàçìåðíûå ñòðóêòóðû;<br />
* ôèçèêà ìèêðîýëåêòðîííûõ ïðèáîðîâ;<br />
* ëèíåéíàÿ è íåëèíåéíàÿ îïòèêà òâåðäîãî òåëà;<br />
* îïòîýëåêòðîíèêà è îïòîýëåêòðîííûå ïðèáîðû;<br />
* êâàíòîâàÿ ýëåêòðîíèêà;<br />
* ñåíñîðèêà.<br />
Ñáîðíèê “Ôîòîýëåêòðîíèêà” èçäà¸òñÿ íà àíãëèéñêîì<br />
ÿçûêå. Ðóêîïèñü ïîäàåòñÿ àâòîðîì â<br />
äâóõ ýêçåìïëÿðàõ íà àíãëèéñêîì è ðóññêîì ÿçûêàõ.<br />
Ê ðóêîïèñè ïðèëàãàþòñÿ äèñêåòà ñ òåêñòîâûì<br />
ôàéëîì è ðèñóíêàìè. Ýëåêòðîííàÿ êîïèÿ<br />
ìàòåðèàëà ìîæåò áûòü ïðèñëàíà â ðåäàêöèþ ïî<br />
ýëåêòðîííîé ïî÷òå.<br />
Ýëåêòðîííàÿ êîïèÿ ñòàòüè äîëæíà îòâå÷àòü<br />
ñëåäóþùèì òðåáîâàíèÿì:<br />
1. Ýëåêòðîííàÿ êîïèÿ (èëè äèñêåòà) ìàòåðèàëà<br />
ïðèñûëàåòñÿ îäíîâðåìåííî ñ òâåðäîé êîïèåé<br />
òåêñòà è ðèñóíêîâ.<br />
2. Äëÿ òåêñòà äîïóñòèìû ñëåäóþùèå ôîðìàòû<br />
— MS Word (rtf, doc).<br />
INFORMATION FOR CONTRIBUTORS<br />
OF “<strong>PHOTOELECTRONICS</strong>” ARTICLES<br />
ÈÍÔÎÐÌÀÖÈß ÄËß ÀÂÒÎÐÎÂ<br />
ÑÁÎÐÍÈÊÀ “ÔÎÒÎÝËÅÊÒÐÎÍÈÊÀ”<br />
The papers should be sent to the address:<br />
Kutalova M. I., Physical Faculty of Odessa<br />
I. I. Mechnikov National University, 42 Pastera str,<br />
65026 Odessa, Ukraine, e-mail: wadz@mail.ru, tel.<br />
+38-0482-7266356. Information is on the site:<br />
http://www.photoelectronics.onu.edu.ua<br />
The title page should contain:<br />
1. Codes of PACS<br />
2. Surnames and initials of authors<br />
3. TITLE OF PAPER<br />
4. Name of institution, full postal address, number<br />
of telephone and fax, electronic address<br />
An abstract of paper should be not more than 200<br />
words. Before a text of summary a title of paper, surnames<br />
and initials of authors should be placed.<br />
Equations are printed in MS Equation Editor.<br />
References should be printed in double space and<br />
should be numbered in square brackets consecutively<br />
throughout the text. References for literature published<br />
in 1999-2009 years are preferential.<br />
Illustrations will be scanned for digital reproduction.<br />
Only high-quality illustrations will be taken for<br />
publication. Legends and symbols should be printed<br />
inside. Neither negatives, nor slides will be taken for<br />
publication. All figures (illustrations) should be numbered<br />
in the sequence of their record in text.<br />
For additional information please contact with the<br />
Editorial Board.<br />
3. Ðèñóíêè ïðèíèìàþòñÿ â ôîðìàòàõ — EPS,<br />
TIFF, BMP, PCX, JPG, GIF, CDR, WMF, MS Word<br />
È MS Giaf, Micro Calc Origin (opj). Ðèñóíêè, âûïîëíåííûå<br />
ïàêåòàìè ìàòåìàòè÷åñêîé è ñòàòèñòè-<br />
÷åñêîé îáðàáîòêè, äîëæíû áûòü êîíâåðòèðóåìûå<br />
â âûøåóêàçàííûå ãðàôè÷åñêèå ôîðìàòû.<br />
Ðóêîïèñè ïðèñûëàþòñÿ ïî àäðåñó:<br />
Îòâ. ñåêð. Êóòàëîâîé Ì. È., óë. Ïàñòåðà, 42,<br />
ôèç. ôàê. ÎÍÓ, ã. Îäåññà, 65026 Å-mail: wadz@<br />
mail.ru, òåë. 0482 — 726 6356 .<br />
Àííîòàöèè ñòàòåé ñá.”Ôîòîýëåêòðîíèêà” íàõîäÿòñÿ<br />
íà ñàéòå: http://photoelectronics.onu.edu.ua<br />
Ê ðóêîïèñè ïðèëàãàåòñÿ:<br />
Òèòóëüíàÿ ñòðàíèöà:<br />
1. Êîäû ÐÀÑ è ÓÄÊ. Äîïóñêàåòñÿ èñïîëüçîâàíèå<br />
íåñêîëüêèõ øèôðîâ, êîòîðûå ðàçäåëÿþòñÿ<br />
çàïÿòîé.  ñëó÷àå, êîãäà àâòîðîì (àâòîðàìè)<br />
íå áóäåò óêàçàí íè îäèí øèôð, ðåäàêöèÿ æóðíàëà<br />
óñòàíàâëèâàåò øèôð ñòàòüè ïî ñâîåìó âûáîðó.<br />
2. Ôàìèëèÿ (à) è èíèöèàëû àâòîðà (åë).<br />
3. Ó÷ðåæäåíèå, ïîëíûé ïî÷òîâûé àäðåñ, íîìåð<br />
òåëåôîíà, íîìåð ôàêñà, àäðåñà ýëåêòðîííîé<br />
ïî÷òû äëÿ êàæäîãî èç àâòîðîâ.<br />
4. Íàçâàíèå ñòàòüè.<br />
151
5. Ðåçþìå îáúåìîì äî 200 ñëîâ ïèøåòñÿ íà àíãëèéñêîì,<br />
ðóññêîì ÿçûêàõ è (äëÿ àâòîðîâ èç Óêðàèíû)<br />
— íà óêðàèíñêîì.<br />
Òåêñò äîëæåí ïå÷àòàòüñÿ øðèôòîì 14 ïóíêòîâ<br />
÷åðåç äâà èíòåðâàëà íà áåëîé áóìàãå ôîðìàòà À4.<br />
Íàçâàíèå ñòàòüè, à òàêæå çàãîëîâêè ïîäðàçäåëîâ<br />
ïå÷àòàþòñÿ ïðîïèñíûìè áóêâàìè è îòìå÷àþòñÿ<br />
ïîëóæèðíûì øðèôòîì.<br />
Óðàâíåíèÿ íåîáõîäèìî ïå÷àòàòü â ðåäàêòîðå<br />
ôîðìóë MS Equation Editor. Íåîáõîäèìî äàâàòü<br />
îïðåäåëåíèå âåëè÷èí, êîòîðûå ïîÿâëÿþòñÿ â òåêñòå<br />
âïåðâûå.<br />
Òàáëèöû ïîäàþòñÿ íà îòäåëüíûõ ñòðàíèöàõ.<br />
Äîëæíû áûòü âûïîëíåíû â ñîîòâåòñòâóþùèõ<br />
òàáëè÷íûõ ðåäàêòîðàõ èëè ïðåäñòàâëåíû â òåêñòîâîì<br />
âèäå ñ èñïîëüçîâàíèåì ðàçäåëèòåëåé (òî÷êà,<br />
çàïÿòàÿ, çàïÿòàÿ ñ òî÷êîé, çíàê òàáóëÿöèè).<br />
Ññûëêè íà ëèòåðàòóðó äîëæíû ïå÷àòàòüñÿ<br />
÷åðåç äâà èíòåðâàëà, íóìåðîâàòüñÿ â êâàäðàòíûõ<br />
ñêîáêàõ (â íîðìàëüíîì ïîëîæåíèè) ïîñëåäîâàòåëüíî,<br />
â ïîðÿäêå èõ ïîÿâëåíèÿ â òåêñòå ñòàòüè.<br />
Ññûëàòüñÿ íåîáõîäèìî íà ëèòåðàòóðó, êîòîðàÿ èçäàíà<br />
ïîçäíåå 2000 ãîäà. Äëÿ ññûëîê èñïîëüçóþòñÿ<br />
ñëåäóþùèå ôîðìàòû:<br />
Êíèãè. Àâòîð(û) (èíèöèàëû, ïîòîì ôàìèëèè),<br />
íàçâàíèå êíèãè êóðñèâîì, èçäàòåëüñòâî, ãîðîä<br />
è ãîä èçäàíèÿ. (Ïðè ññûëêå íà ãëàâó êíèãè, óêàçûâàåòñÿ<br />
íàçâàíèå ãëàâû, íàçâàíèå êíèãè êóðñèâîì,<br />
íîìåðà ñòðàíèö). Ïðèìåð, J. A. Hall, Imaging<br />
tubes, Chap 14 ø The Infrared Handbook, Eds W. W.<br />
 æóðíàë³ “Ôîòîåëåêòðîí³êà” äðóêóþòüñÿ<br />
ñòàòò³ òà êîðîòê³, ÿê³ ì³ñòÿòü â³äîìîñò³ ïðî íàóêîâ³<br />
äîñë³äæåííÿ òà òåõí³÷í³ ðîçðîáêè ó íàïðÿìêàõ:<br />
* ô³çèêà íàï³âïðîâ³äíèê³â;<br />
* ãåòåðî- òà íèçüêîâèì³ðí³ ñòðóêòóðè;<br />
* ô³çèêà ì³êðîåëåêòðîííèõ ïðèëàä³â;<br />
* ë³í³éíà òà íåë³í³éíà îïòèêà òâåðäîãî ò³ëà;<br />
* îïòîåëåêòðîí³êà òà îïòîåëåêòðîíí³ ïðèëàäè;<br />
* êâàíòîâà åëåêòðîí³êà;<br />
* ñåíñîðèêà.<br />
Æóðíàë “Ôîòîåëåêòðîí³êà” âèäàºòüñÿ àíãë³éñüêîþ<br />
ìîâîþ. Ðóêîïèñ ïîäàºòüñÿ àâòîðîì ó<br />
äâîõ ïðèì³ðíèêàõ àíãë³éñüêîþ ìîâîþ. Äî ðóêîïèñó<br />
äîäàºòüñÿ äèñêåòà ç òåêñòîâèì ôàéëîì ³ ìàëþíêàìè<br />
Åëåêòðîííà êîï³ÿ ìàòåð³àëó ìîæå áóòè<br />
íàä³ñëàíà äî ðåäàêö³¿ åëåêòðîííîþ ïîøòîþ.<br />
Åëåêòðîííà êîï³ÿ ñòàòò³ ïîâèííà â³äïîâ³äàòè<br />
íàñòóïíèì âèìîãàì:<br />
1. Åëåêòðîííà êîï³ÿ (àáî äèñêåòà) ìàòåð³àëó<br />
íàäñèëàºòüñÿ îäíî÷àñíî ç òâåðäîþ êîﳺþ òåêñòó<br />
òà ìàëþíê³â.<br />
2. Äëÿ òåêñòó ïðèïóñòèì³ íàñòóïí³ ôîðìàòè —<br />
MS Word (rtf, doc).<br />
3. Ìàëþíêè ïðèéìàþòüñÿ ó ôîðìàòàõ — EPS,<br />
TIFF, BMP, PCX, JPG, GIF, CDR, WMF, MS Word<br />
² MS Giaf, Micro Calc Origin (opj). Ìàëþíêè, âèêîíàí³<br />
ïàêåòàìè ìàòåìàòè÷íî¿ òà ñòàòèñòè÷íî¿<br />
152<br />
²ÍÔÎÐÌÀÖ²ß ÄËß ÀÂÒÎвÂ<br />
ÇÁ²ÐÍÈÊÀ “ÔÎÒÎÅËÅÊÒÐÎͲÊÀ”<br />
Wolfe, Î J’ Zissis. pp. 132-176, ERIM, Arm Arbor, MI<br />
(1978).<br />
Æóðíàëû. Àâòîð(û) (èíèöèàëû, ïîòîì ôàìèëèè),<br />
íàçâàíèå ñòàòüè, íàçâàíèå æóðíàëà êóðñèâîì<br />
(èñïîëüçóþòñÿ àááðåâèàòóðû òîëüêî äëÿ èçâåñòíûõ<br />
æóðíàëîâ), íîìåð òîìà è âûïóñêà, íîìåð<br />
ñòðàíèö è ãîä èçäàíèÿ. Ïðèìåð, N. Blutzer and<br />
A. S. Jensen, Current readout of infrared detectors //<br />
Opt Eng 26(3), pp. 241-248.<br />
Ïîäïèñè ê ðèñóíêàì è òàáëèöàì ïå÷àòàþòñÿ â<br />
ðóêîïèñè ïîñëå ëèòåðàòóðíûõ ññûëîê ÷åðåç äâà<br />
èíòåðâàëà.<br />
Èëëþñòðàöèè áóäóò ñêàíèðîâàòüñÿ öèôðîâûì<br />
ñêàíåðîì. Ïðèíèìàþòñÿ â ïå÷àòü òîëüêî âûñîêîêà÷åñòâåííûå<br />
èëëþñòðàöèè. Ïîäïèñè è ñèìâîëû<br />
äîëæíû áûòü âïå÷àòàíû. Íå ïðèíèìàþòñÿ â ïå-<br />
÷àòü íåãàòèâû, ñëàéäû, òðàíñïîðàíòû.<br />
Ðèñóíêè äîëæíû èìåòü ñîîòâåòñòâóþùèé ê<br />
ôîðìàòó æóðíàëà ðàçìåð — íå áîëüøå 160x200 ìì.<br />
Òåêñò íà ðèñóíêàõ äîëæåí âûïîëíÿòüñÿ øðèôòîì<br />
12 ïóíêòîâ. Íà ãðàôèêàõ åäèíèöû èçìåðåíèÿ<br />
óêàçûâàþòñÿ ÷åðåç çàïÿòóþ (à íå â ñêîáêàõ). Âñå<br />
ðèñóíêè (èëëþñòðàöèè) íóìåðóþòñÿ â ïîðÿäêå èõ<br />
ðàçìåùåíèÿ â òåêñòå. Íå äîïóñêàåòñÿ âíîñèòü íîìåð<br />
è ïîäïèñü íåïîñðåäñòâåííî íà ðèñóíêàõ.<br />
Ðåçþìå îáúåìîì äî 200 ñëîâ ïèøåòñÿ íà àíãëèéñêîì,<br />
ðóññêîì ÿçûêàõ è íà óêðàèíñêîì (äëÿ<br />
àâòîðîâ èç Óêðàèíû). Ïåðåä òåêñòîì ðåçþìå ñîîòâåòñòâóþùèì<br />
ÿçûêîì óêàçûâàþòñÿ ÓÄÊ, ôàìèëèè<br />
è èíèöèàëû âñåõ àâòîðîâ, íàçâàíèå ñòàòüè.<br />
îáðîáêè ïîâèíí³ áóòè êîíâåðòîâàí³ ó âèùåâêàçàí³<br />
ãðàô³÷í³ ôîðìàòè ³ ðîçòàøîâàí³ ó òåêñò³ ñòàòò³,<br />
çã³äíî çì³ñòó.<br />
Ðóêîïèñè íàäñèëàþòüñÿ íà àäðåñó:<br />
³äï. ñåêð. Êóòàëîâ³é Ì. ²., âóë. Ïàñòåðà, 42,<br />
ô³ç. ôàê. ÎÍÓ, ì.Îäåñà, 65026 Å-mail: wadz@mail.<br />
ru, òåë. 0482-266356.<br />
Äî ðóêîïèñó äîäàºòüñÿ:<br />
1. Êîäè ÐÀÑ òà ÓÄÊ. Äîïóñêàºòüñÿ âèêîðèñòàííÿ<br />
äåê³ëüêîõ øèôð³â, ùî ðîçä³ëÿþòüñÿ êîìîþ.<br />
Ó âèïàäêó, êîëè àâòîðîì (àâòîðàìè) íå áóäå<br />
âêàçàíî æîäåí øèôð, ðåäàêö³ÿ æóðíàëó âñòàíîâëþº<br />
øèôð ñòàòò³ çà ñâî¿ì âèáîðîì.<br />
2. Ïð³çâèùå òà ³í³ö³àëè àâòîðà.<br />
3 Óñòàíîâà, ïîâíà ïîøòîâà àäðåñà, íîìåð òåëåôîíó,<br />
íîìåð ôàêñó, àäðåñà åëåêòðîííî¿ ïîøòè<br />
äëÿ êîæíîãî ç àâòîð³â.<br />
4. Íàçâà ñòàòò³.<br />
Ðåçþìå îá’ºìîì äî 200 ñë³â ïèøåòüñÿ àíãë³éñüêîþ,<br />
óêðà¿íñüêîþ òà ðîñ³éñüêîþ ìîâàìè. Ïåðåä<br />
òåêñòîì ðåçþìå â³äïîâ³äíîþ ìîâîþ âêàçóþòüñÿ<br />
êîä, íàçâà ñòàòò³, ïð³çâèùà òà ³í³ö³àëè âñ³õ àâòîð³â.<br />
Òåêñò ïîâèíåí äðóêóâàòèñÿ øðèôòîì 12 ïóíêò³â<br />
÷åðåç äâà ³íòåðâàëè íà á³ëîìó ïàïåð³ ôîðìàòó<br />
À4. Íàçâà ñòàòò³, à òàêîæ çàãîëîâêè ï³äðîçä³ë³â<br />
äðóêóþòüñÿ ïðîïèñíèìè ë³òåðàìè ³ â³äçíà÷àþòüñÿ<br />
íàï³âæèðíèì øðèôòîì.
гâíÿííÿ íåîáõ³äíî äðóêóâàòè ó ðåäàêòîð³<br />
ôîðìóë MS Equation Editor. Íåîáõ³äíî äàâàòè<br />
âèçíà÷åííÿ âåëè÷èí, ùî ç’ÿâëÿþòüñÿ â òåêñò³<br />
âïåðøå.<br />
Òàáëèö³ ïîäàþòüñÿ íà îêðåìèõ ñòîð³íêàõ. Ïîâèíí³<br />
áóòè âèêîíàí³ ó â³äïîâ³äíèõ òàáëè÷íèõ ðåäàêòîðàõ<br />
àáî ïðåäñòàâëåí³ ó òåêñòîâîìó âèãëÿä³ ç<br />
âèêîðèñòàííÿì ðîçä³ëüíèê³â (êðàïêà, êîìà, êîìà<br />
ç êðàïêîþ, çíàê òàáóëÿö³¿).<br />
Ïîñèëàííÿ íà ë³òåðàòóðó ïîâèíí³ äðóêóâàòèñÿ<br />
÷åðåç äâà ³íòåðâàëè, íóìåðóâàòèñü â êâàäðàòíèõ<br />
äóæêàõ (ó íîðìàëüíîìó ïîëîæåíí³) ïîñë³äîâíî ó<br />
ïîðÿäêó ¿õ ïîÿâè â òåêñò³ ñòàòò³.<br />
Êíèãè. Àâòîð(è) (³í³ö³àëè, ïîò³ì ïð³çâèùà),<br />
íàçâà êíèãè êóðñèâîì, âèäàâíèöòâî, ì³ñòî ³ ð³ê âèäàííÿ.<br />
(Ïðè ïîñèëàíí³ íà ãëàâó êíèãè, âêàçóºòüñÿ<br />
íàçâà ãëàâè, íàçâà êíèãè êóðñèâîì, íîìåðè ñòîð³íîê).<br />
Ïðèêëàä: J. A. Hall, Imaging tubes, Chap 14 ø<br />
The Infrared Handbook, Eds W. W. Wolfe, Î J’ Zissis.<br />
pp. 132-176, ERIM, Arm Arbor, MI (1978).<br />
Æóðíàëè ( ×àñîïèñè). Àâòîð(è) (³í³ö³àëè, ïîò³ì<br />
ïð³çâèùà), íàçâà ñòàòò³, íàçâà æóðíàëó êóðñèâîì<br />
(âèêîðèñòîâóþòüñÿ àáðåâ³àòóðè ò³ëüêè äëÿ<br />
â³äîìèõ æóðíàë³â), íîìåð òîìó ³ âèïóñêó, íîìåð<br />
ñòîð³íîê ³ ð³ê âèäàííÿ. Ïðèêëàä: N. Blutzer and<br />
A. S. Jensen, Current readout of infrared detectors //<br />
Opt Eng 26(3), pp. 241-248 (1987).<br />
²ëþñòðàö³¿ áóäóòü ñêàíóâàòèñÿ öèôðîâèì ñêàíåðîì.<br />
Ïðèéìàþòüñÿ äî äðóêó ò³ëüêè âèñîêîÿê³ñí³<br />
³ëþñòðàö³¿. ϳäïèñè ³ ñèìâîëè ïîâèíí³ áóòè<br />
âäðóêîâàí³. Íå ïðèéìàþòüñÿ äî äðóêó íåãàòèâè,<br />
ñëàéäè, òðàíñïàðàíòè.<br />
Ðèñóíêè ïîâèíí³ ìàòè â³äïîâ³äíèé äî ôîðìàòó<br />
æóðíàëó ðîçì³ð íå á³ëüøå 160x200 ìì. Òåêñò<br />
íà ðèñóíêàõ ïîâèíåí âèêîíóâàòèñü øðèôòîì<br />
10 ïóíêò³â. Íà ãðàô³êàõ îäèíèö³ âèì³ðó âêàçóþòüñÿ<br />
÷åðåç êîìó (à íå â äóæêàõ). Óñ³ ðèñóíêè (³ëþñòðàö³¿)<br />
íóìåðóþòüñÿ â ïîðÿäêó ¿õ ðîçì³ùåííÿ â<br />
òåêñò³. Ðåêîìåíäîâàíî ïîñèëàòèñü íà ë³òåðàòóðó,<br />
ÿêà íàäðóêîâàíà ç 2000 ïî 2009 ð³ê.<br />
153
Íàóêîâå âèäàííÿ<br />
�<br />
ÔÎÒÎÅËÅÊÒÐÎͲÊÀ<br />
�<br />
̳æâóç³âñüêèé íàóêîâèé çá³ðíèê<br />
�<br />
¹ 18’2009<br />
Àíãë³éñüêîþ ìîâîþ<br />
Ãîëîâíèé ðåäàêòîð Â. À. Ñìèíòèíà<br />
³äïîâ³äàëüíèé ñåêðåòàð Ì. ². Êóòàëîâà<br />
Çäàíî ó âèðîáíèöòâî 26.06.2009. ϳäïèñàíî äî äðóêó 17.08.2009.<br />
Ôîðìàò 60õ84/8. Ïàï³ð îôñåòíèé. Ãàðí³òóðà «Newton». Äðóê îôñåòíèé.<br />
Óì. äðóê. àðê. 17,90. Òèðàæ 100 ïðèì. Çàì. ¹ 324.<br />
Âèäàâíèöòâî ³ äðóêàðíÿ «Àñòðîïðèíò»<br />
65091, ì. Îäåñà, âóë. Ðàçóìîâñüêà, 21<br />
Òåë.: (0482) 37-07-95, 37-24-26, 33-07-17, 37-14-25<br />
www.astroprint.odessa.ua; www.fotoalbom-odessa.com<br />
Ñâ³äîöòâî ñóá’ºêòà âèäàâíè÷î¿ ñïðàâè ÄÊ ¹ 1373 â³ä 28.05.2003 ð.