Allen M. Hermann - University of Arkansas Physics Department

-PHYSICA 8IAnisotropic Seebeck coefficient in Tl2Ba2CuO6-, single crystalsW. Kiehl &*, H.M. Duan b, A.M. HermannUniuersity of Arkansas, Fayetteuille, AR 72701. USAColorado School of Mines. Gola'en. CO 80401. USAUniversity of Colorado. Boulder, CO 80309, USAReceived 16 May 1995Anisotropic Seebeck coefficients have been carefully measured in TI,Ba,CuO, single crystals with differing oxygendo~lng levels. In-plane Seebeck coefficients are small in magnitude (- 5 fl/K), negative at room temperature and have adccrc.u~ng magnitude as the temperature decreases. C-axis Seebeck coefficients are larger in magnitude (- 20 O/K) andPsltlve. The absolute magnitude of the Seebeck coefficients (dl directions) increases as the oxygen (hole) doping levels'"reae In these overdoped materials. In samples with high transition temperatures, a crossover from negative to positive"Pars - 15 K above the superanducting transition temperature. The good ~0rfelation between bulk compound data andI1 P'opnarcly weighted ,"-plane and c-aris crystal data is shown The data are discussed in tenns of single-band andI;l~"ln.l models but none are identified as appropriate for these complex materials.1. IntroductionSince its discovery [I], TI,Ba,CuO, (TI-2201)ha been one of the most interesting high-tempemsuperconductorsystems available for study. It isow of the simplest HTSC crystal structures contain-'% Only or:e CU-0 plane per formula unit [21. The'\':V-'rc~*ductin~ transition temperature, T,, of thiscan be tuned over a large temperature range(' [o 92 K) by simple oxygen or inert gas (N,, He*&) alnealing [3] (oxygen acts as a flopant). In thispaper we present and discuss measurements of thean60tmpic Seebeck effect of TI-2201 single crystalsxs a function of T, (hole doping).fie first Seebeck measurements on m-2201 were1 Rponed by Dyadakin et a]. [4]. They show forICaponding author.one bulk ceramic sample with T, < 40 K. Weeks eta]. [5] reported measurements on bulk samples withT,'s of 80 and 60 K. These were supplemented bymeasurements on Ce doped TI-2201 compounds [6].Obertelli et al. measured Seebeck data on bulk samplesof TI-2201 with T,'s ranging fmm 0 to - 80 K[7]. Zhang et al. reported anisotropic data on onecrystal with T, - 80 K as grown in our lab [8]. Wenow report anisotropic measurements on crystals ofvarying T, from the same growth batch as thatreported in Ref. [8].2 ExperimentalThe crystals were grown using a self-flux method[9] and were selected for lack of growth steps on thesurface, lack of intergrowth lines, thinness, and sharpedges. Dimensions were approximately 1 mm X 0.5

PHYSICAL REVIEW B VOLUME 45, NUMBER 10 1 MARCH 1992-11Flux-creep activation energies in T12B~CaCu20, single crystalsM. NikoloPhysics Department, University of Colorado, Boulder, Colorado 80309W. KiehlPhysics Deparrment, University of Arkansas, Fayerteville, Arkansas 72701and Physics Department, University of Colorado, Boulder, Colorado 80309H. M. Duan and A. M. HermannPhysics Department, University of Colorado, Boulder, Colorado 80309(Received 2 May 1991; revised manuscript received 15 October 1991)The measured frequency shin in temperature of the ac susceptibility TlzBa,CaCu20, single crystals ismodeled by an Arrhenius-like expression to give fluxcreep activation energies at various applied ac magneticfields ranging from 15 to 200 Oe. The activation energies range from 288*14 meV in an applied acmagnetic 6eld of 15 Oe to 25f 3 meV in 200 Oe. Considerable flux creep is observed. It is estimated thatIfcI at a temperature of about 100 K is approximately 10 Oe. The results are interpreted in terms of Andersonflux-creep model and the activation-energy values repofled here are similar to those deducedfrom magnetic-fielddependent resistivity measurements on the TI-based single crystals, textured films,and bulk samples. The field dependence of the activation energies is observed to vary as 1 /a. Thislow-field dependence is explained by a scaling argument where the vortex volume is approximated by$ao, with 5 the coherence length in the ab plane and a. the Abrikosov vortex-lattice spacing along the caxis of the crystal lattice.EXPERIMENTAL DETAILSFlux pinning plays a key role in realizing high criticalcurrent densities in high temperature superconductorsand much attention has been focused on this subject.Palstra et al.' measured the resistive transition in a singlecrystal of B~,S~,C~CU~O~-~ in large dc magnetic fieldsand showed that flux creep was thermally activated, asdefined by the Arrhenius resistivity expressionp(T,H )=poexp[- U(H )/kT], with magnetic fielddependentpinning energies ranging from 30 to 300 meVnear T,. Using the same technique, Sun et al., showedthat the activation energies in ErBazCu307-6 were muchgreater when the applied magnetic field was perpendicularto the c axis than when it was parallel.Yeshurun and ~alozemoff~ used the method of magneticrelaxation, introduced by Kim et ~ 1 and . ~ Anderson,'in single YBa2Cu307-, crystals to derive the pinningpotentials. Ferrari et ~ 1 derived . ~ a distributionfunction of the pinning energies for a YBa2Cu307-, thinfilm and for a B~,S~,C~CU,O,+~ crystal by measuring theflux noise using a low-noise 4-K SQUID. Nikolo andGoldfarb7 and Miiller, Nikolo, and Driversp9 measuredthe intergranular activation energies in Y-Ba-Cu-0 and(Bi,Pb)-Sr-Ca-Cu-0 bulk samples by the method of a frequencyshift in ac susceptibility. Here we use this lattermethod and Anderson's theory of flux creep to find theflux-creep activation energies in Tl2Ba2CaCuzO, (2:2: 1 :2phase) single crystals in low ac magnetic fields.Volume susceptibility was measured with an ac susceptometerspecifically designed to pick up a small signalfrom a single crystal. Two sensing coils, connected in oppositionto pick up the crystal's signal induced by the primarycoil, were used. The sample remained stationaryinside one of the sensing coils with the signal from thesample superimposed on the background. Only the real(inductive) part, whose signal is considerably larger thanthat of the imaginary part, was measured. Severalbackground-only runs were made to confirm the transition.Only samples where the transition was quite distinctwere included in the analysis.To maximize the sensitivity for a single small crystal,the smallest possible dimensions of the sample holder andthe sensing coils had to be selected. The smallest temperaturediode available is the Lake Shore DT-450 and thesample holder was designed around it. Figure 1 showsthe dimensions of the sample holder configuration. Thepick-up coils had a 5-mm inner diameter and were woundwith a 420-gauge (0.003-in.-diam) wire. A StanfordResearch Systems lock-in amplifier, model SR-530 wasused to detect the signal from the pick-up coils. The temperaturewas maintained using an Omega model CYC91controller with a Nichrome heater wire. The heater coilwas oppositely wound to eliminate its magnetic effects.The primary coil was driven with a California Instruments3213-K high-power constant rms current source.5641 @ 1992 The American Physical Society

PHYSICAL REVIEW B VOLUME 43, NUMBER 16 1 JUNE 1991Anisotropic resistivity and paraconductivity of T12Ba2CaCu208 single crystalsH. M. DuanPhysics Department, Uniuersity of Colorado, Boulder, Colorado 80309W. KiehlPhysics Department, University of Arkansas, Fayetteville, Arkansas 72701C. DongPhysics Department, University of Colorado, Boulder, Colorado 80309A. W. CordesChemistry Department, University of Arkansas, Fayetteville, Arkansas 72701M. J. Saeed and D. L. ViarPhysics Department, University of Arkansas, Fayetteville, Arkansas 72701A. M. HermannPhysics Department, Uniuersity of Colorado, Boulder, Colorado 80309(Received 21 June 1990; revised manuscript received 13 February 1991)The anisotropic resistivity and paraconductivity of T12Ba2CaCu208 single crystals were measured.The in-plane and c-axis resistivities are, respectively, -7X and lo-' cm at room temperature.The c-axis resistivity decreases with decreasing temperature in the temperature range fromroom temperature down to the superconducting transition temperature, and the second derivativeof p, is negative, which is not characteristic of YBa2Cu,0, and Bi2Sr2,2Cao,8Cu208 single crystals, inwhich the second derivative is either positive or zero. The in-plane paraconductivity is obtainedfrom the in-plane resistivity, and a method without the critical temperature as an adjustable parameteris employed to obtain the dimensionality. The data show that the superconductivity is two dimensional,but a crossover from two to three dimensions has also been observed near the transitiontemperature in some samples.INTRODUCTIONSince the discovery of high-temperature superconductivityin the doped copper-oxide systems, there has beenmuch interest in the usual properties of these systems intheir normal state. The highly anisotropic electronicstructure of the carrier planes and the interaction betweenthe planes are very interesting aspects of the natureof these systems. In the normal state, the resistivity ratiopc./pp (where p, and pp are the resistivities along the caxls and in the a-b planes, respectively) was reported tobe as high as lo2- lo5, and thermal excitation transport inthe normal state was observed in the c axis of YBa2Cu307and Bi2Sr2,,Ca0, ,Cu20, crystals. '-' The dimensionalityof the superconductivity has been discussed in relation toparaconductivity measurements on bulk, thin-film, andsingle-crystal samples of YBa2Cu307 (Refs. 6-8) and~i~~r~,~~a~,,Cu~0,.~In the T1-Ba-Ca-Cu-0 system, measurementswere carried out on polycrystalline thin-filmand bulk samples. 9*'0In this paper, we report the results of anisotropic resistivityand paraconductivity measurements on single crystalsof Tl2Ba2CaCu2O8 (2:2:1:2). The pp versus T behaviorshows an increase with increasing temperature and isapproximately linear above 220 K. Below 220 K, thecurve gradually deviates from the linear behavior. In ourexperiments, p, decreases as the temperature decreases,and the p, versus T curve is not linear, the second derivativeis negative. The pc-T curve saturates as temperatureincreases, this may be explained by a resistivity model ofa metallic system where the mean free path is close to theMott limit. We believe the electronic band structure of2:2:1:2 corresponds to that of a three-dimensional metalwith a large anisotropic effective mass. In some crystals,paraconductivity may be fit by a two-dimensional theoryin the temperature range from 150 to - 2 K above transi-tion temperature (-102 K), but in other samples, acrossover from two- or three-dimensional behavioroccurs at -5.5 K above the transition temperature.EXPERIMENTTwo different processes were used to grow the singlecrystals. In the first process ( A ), we used a Ba2Ca2Cu,0,precursor. Appropriate amounts of BaO, CaO, and CuO(each 99.99% pure) were mixed and ground to form amixture with a nominal stoichiometry of Ba2Ca2Cu30,.The mixture was heated in air for 24 h at 920°C, with anintermediate grinding carried out in order to facilitate acomplete reaction. The reacted mixture was then ground12 925 @ 1991 The American Physical Society

Modem Physics Lcttcrs B, Vol. 4. No. 20 (1990) 1281-128763 World Scientific Publishing Company100 K SUPERCONDUCTING PHASE IN THETI-Pb-Sr-R-Cu-0 SYSTEM WITH R = RARE EARTHS2. Z. SHENG. Y. XIN. and J. M. MEASONDt7partmc.rrt of Phy.vic.c, L'nivenirjb qf .4rknnnrns, Fuyertmilk .4R 72 701. t< S.4Received 13 August 1990Superconductivity around 100 K was observed in the TI-Pb-Sr-R-Cu-0 system(R- rare earths) by resistiincc and ac susceptibility measurements. Powder X-ray diffractionanalyses showed that the phase responsible for the suprconductivity is of a tetragonalstructure with a general formula (TI, ..sPb,,NSr3-,,R~v)Cu0 (132 phase).Discoveries of the 90 K TI-Ba-Cu-0 system'.2 and 120 K TI-Ba-Ca-Cu-0has lead to a number of new Tl-based superconducting system~.~-"nour own work, we found that the T1-Sr-Ca-R-Cu-0 system (R rare earths) issuperconducting around 90 K.7 Later, we found that the Ca-free TI-Sr-R-Cu-0system was also superconducting at temperatures up to about 90~~-'"Furtherwork showed that when T1 was partially substituted by Pb, the TI-Pb-Sr--R-Cu-0 system exhibited superconductivity at even higher temperaturesaround 100 K. In this letter, we report the observation of 100 K superconductivityin the TI-Pb-Sr-R-Cu-0 samples and the phase identification based onpowder X-ray diffraction data. In the present experiments Pr and Tb were usedas the representatives of the light rare earths and the heavy rare earths,respectively. Other rare earths, including Y, also work, and the results will bepublished in a separate paper.TI-Pb-Sr-R-Cu-0 samples were prepared using high-purity T1203, Pb02,SrO (SrCO,, or ST(N03)2), R2O3 (except Ce02, Pr60,,, and Tb4O7), and CuO. Ina typical procedure, component compounds with a certain metallic atom ratiowere completely mixed and ground. The powder was pressed into a pellet witha diameter of either 7 mm or t 5 mm and a thickness of 1-2 mm. The pellet wasplaced in an alumina boat, and was heated in a preheated tube furnace at1000-1050°C with flowing oxygen for 2-30 minutes. The samples were theneither quenched or furnace cooled. The samples weR examined by resistance andac susceptibility measurements and by powder X-ray diffraction analyses usingthe apparatuses and methods we described previou~ly.~~'~Figure 1 shows resistance-temperature curves for nominal samplesT11.2(Sr2.6Pro.4)Cu,0,0 (sample A) and (T~-,P~,~SKS~~.~P~O.~)~ZO~(sample B).These two samples were prepared in the same batch. Although they exhibit&PACS Nos.: 74.10 + v, 74.70.Ya.

Solid State Communications, Vol. 73, No. 1, pp. 53-57, 1990.Printed in Great Britain.0038-1098/90 $3.00 + .00Pergamon Press plcTHERMOELECTRIC POWER OF TI-Ca-Ba-Cu-O AND Pr-TI-Sr-Ca-Cu-0HIGH Tc SUPERCONDUflORSA.K. Bhatnagar*', R. Pan', D.G. Naugle', P.J. SquattritoZ and A. Cleariield2'Physics Department; Chemistry Department, Texas A & M University, College Station, TX 77843 USAandZ.Z. Sheng, Q.A. Shams and A.M. HermannDepartment of Physics, University of Arkansas, Fayetteville, AR 72701 USA(Received 28 July 1989 by M. Cardona)The resistance and the thermoelectric power of thallium based high T,superconductors Tl-Ca-Ba-Cu-O and Pr-11-Sr-Ca-Cu-O have beenmeasured in the temperature interval 77-300 K. The resistive behaviourof Tl-Ca-Ba-Cu-0 is found to be metallic while that of Pr-TI-Sr-Ca-Cu-0 shows a mixed metallic and semiconductive behaviour. Thethermoelectric power of both superconductors is positive throughoutthe temperature interval. The thermoelectric power of the TI-Ca-Ba-Cu-0 superconductor decreases linearly with increasing temperatureabove the superconducting transition temperature, T,, below which itis zero. The thermoelectric power of the Pr-11-Sr-Ca-Cu-O sampleshows a maximum near 180K. The temperature dependence of thethermoelectric power of these samples, above T,, is quite different thanthat observed for RBa,Cu307-, superconductors.TNTRODUrnONAFTER THE discovery [l] of RBa2Cu30,-, high T,.;!ipcrconductors with T, = 90K where R stands for ar;!re earth clement, a higher Tc of about 100-125K*;is obtained in TI-Ca-Ba-Cu-0 superconductors(21. In particular, 11,Ca,Ba,Cu,O,, was found to havea T, of 120-125K. Recently, a new family of thalliumoxide bascd superconductors, R-TI-Sr-Ca-Cu-0,h:lvc bcen synthesized with T, between 80 and 90 K [3].In this scrics the Pr-TI-Sr-Ca-Cu-0 system is thefirst oxide supcrconducting system containing Pr with7; above 77 K.The structure of thallium based oxide supercontluctorsis tctragonal and consists of double, for*l12Ca, Ua,Cu,O,, or triple, for Tl,Ca,Ba,Cu3010,CuO, sheets, with Cu in square planar coordination,stacking alternately with sheets of Ba ions and doublesheets of Tl-0 atoms, and with Ca between Cu-Osheets [4J. Neutron diffraction results [5j indicate apossible substitution of Ca on the TI sites and viceversa in TIFazBa,Cu308 (referred to hereafter as the2 122 phase), and only Ca subsitution onthe TI sites* Present address School of Physics, University ofHyderabad, Hyderabad 500 134, Andhra Pradesh,India.in TI,CaZBa,Cu3O,,, (referred to hereafter as the 2 22 3phase). Appreciable disorder of the oxygen atoms inthe Tl-0 sheets was observed for both compounds,and evidence of a small oxygen deficiency in the 22 23TI compound was found (x = 9.86 instead of 10,where x is the number of oxygen atoms). UnlikeRBa,Cu307-, superconductors, there are no Cu-0chains in Tl-Ca-Ba-Cu-0 superconductors. Thus,Cu-0 planes seems to be the primary requirement forsuperconductivity in the cuprate superconductors.Further, unlike RBa,Cu307-, superconductors, theoxygen content in TI-Ca-Ba-Cu-0 superconductorsremains very close to the stoichiometric composition.This difference in structure as well as absence ofoxygen deficiency is expected to result in differentbehaviour of some important physical properties ofTl-Ca-Ba-Cu-O superconductors in comparisonwith those of RBazCu307-, superconductors.Electronic transport properties of RBa2Cu307-,,in particular those of YBa,Cu,O,-,, superconductorshave been investigated extensively to understand thenormal state behaviour and to obtain some clue to thepossible interaction responsible for superconductivityat such high temperatures. There have been veryfew reports on electronic transport properties ofTI-Ca-Ba-Cu-0 superconductors. Besides the usualresistance measurements to determine the resistive

lEEE TRANSACTIONS ON MAGNETICS, VOL. 25. NO. 2, MARCH 1989JOSEPHSON JUNCTIONS WITH BULK YBa2Cu30;r, Bi2SrzCaCu2%, AND T12Ca2Ba~Cu3010+xW. Eidelloth, F. S. Barnes, S. Geller, K.-Y. WuDepartment of Electrical and Computer EngineeringUniversity of Colorado, Boulder, CO 80309-0425andZ. Z. Sheng, A. M. HermannDepartment of PhysilsUniversity of Arkansas, Fayetteville, AR 72701AbstractThe characteristics of point contacts withcombinations of Pb, YBazCu307.,: (YBCO), Bi2Sr2CaCu2OX(BSCCO), and Tl2Ca2Ba2Cu3OlO+,: (TCBCO) have beenmeasured in a shielded liquid helium dewar. By formingseveral junctions of a kind we have extracted typicalfeatures of the current-voltage characteristics for eachcombination. Zero-voltage currents were observed in allcases. Diffraction patterns in the dependence of thecritical currents on the applied magnetic field were foundin some junctions of each kind. Therefore electron-pairsrather than single electrons are responsible for thecurrent transport in all three oxide superconductorsmentioned above. For point contacts using the high-Tcmaterials only, both homo- and hetero-junctions wereinvestigated. We observed Josephson effects above liquidnitrogen temperatures. Whereas YBCO and TCBCO werefound to perform very similarly, junctions with BSCCOyielded considerably different results. All junctionsbehaved like weak link bridge structures.IntroductionSuperconducting point contacts are realized bypressing a sharp piece of superconducting materialagainst a flat one. The resulting structure dependsstrongly on the mechanical properties of the material aswell as the mechanical pressure between them; thereforethe underlying physics is, in general, not well defined.The two superconductors are usually linked by severalpaths, either metallic or through oxide barriers. 1 In spiteof the uncertain weak coupling mechanism, pointcontact structures have been used extensively in the pastand play an important role in applications. Pointcontacts can be readily formed with various materials.We studied point contacts with electrodes made ofPb, YBCO, BSCCO, or TCBCO in various combinations.By forming several point contacts of a kind we couldobserve features which were common to all junctions ofa particular combination of electrodes. In this way wewere able to compare junctions formed with the abovementioned oxide superconductors qualitatively and tosome extent also quantitatively.The su~erconductine com~oundsTo form the counter electrodes, we used half-diskshapedpellets with radii of about 9 mm and thicknessesof about 2 mm. The flat part of the half disk was polishedwith sandpaper. The surface of the specimen was cleanedwith alcohol before each experiment.Point contacts with Pb and high-Tc materialA needle-shaped piece of Pb was pressed againstthe flat and polished surface of a piece of bulk YBCO,BSCCO, or TCBCO. The mechanical pressure, as well asthe location of the Pb-needle on the flat counter piece ofhigh-Tc material, could be adjusted externally. Allelectrical measurements were done by a four-point probetechnique. Two leads were soldered to the Pb; two leadswere attached to the respective oxide superconductor bysilver paint. A current source was connected tdtwo leads(one to the Pb, one m the counter electrode) as'in Fig. 1,and produced a low frequency (-1 Hz) triangle waveform.The voltage was measured across the other two leads.Thus I-V curves could be directly observed on anoscilloscope. A small coil was mounted on top of thepoint contact to expose the junctions to a dc magneticfield up to 2-10-5 T.holders'=*co ntereyectrode v 1A -Ax-in utosciloscopey-inputoscilloscopeFig.1: Arrangement to form point contacts and measureBoth the YBCO and the BSCCO samples were their 1-v characteristics.made at the Universitv of Colorado. Thev showedtransition temperatures (R=O) of 94 K ind 74 K This arrangement was mounted on one end of aThe TCBCO were made at the one meter long stainless steel tube, which could beUniversity of Arkansas and exhibited a Tc of 125 K. inserted into a magnetically shielded liquid heliumdewar. On the tube's other end a filter box allowedexternal electrical connections. Two rods were accessibleManuscript received August 22, 1988.through quick couplings and allowed external control of0018-9464/89/03W39501.0001989 IEEEAulhorized licensed use limited to: University of Mansas. Downloaded on June 3,2009 at 1556 from IEEE Xplore. Resbidions apply.

Preparation of high 7, TI-Ba=Ca=Cu-0 thin films by pulsed laser evaporationand TI, 0, vapor processing8. Johs, D. Thompson, N. J. lanno, and John A. WoollamDepartment of Electrical Engineering and Center for Microelectronic and Optical Muterials Research,University of Nebmska-Lincoln, Lincoln, Nebraska 68588-0511S. H. LiouBehhn Labomtory of Physics, and Center for Microelectronic and Optical Materials Research, Universityof Nebraska-Lincoln, Lincoln, Nebraska 68588-0111A. M. Hermann, Z. Z. Sheng, W. Kiehl, Q. Shams, X. Fei, L. Sheng, and Y. H. LiuDepclrtment of Physics, University of Arkansas, Fuyetteuille, Arkansas 72701(Received 13 December 1988; accepted for publication 7 March 1989)Ti-Ba-Ca-Cu-0 superconducting thin films with zero-resistance temperatures up to 115 Khave beer. prepared using a T12 0, vapor process on Ba-Ca-Cu-0 precursor thin films. TheBa-Ca-Cu-0 thin films were made by laser deposition on Y-stabilized ZrO, substrates. Thistechnique minimizes problems caused by the toxicity of T120,, and its subsequentdecomposition to the volatile and toxic T1,O upon heating. Therefore, it may have practicalapplication in the fabrication of high T, TI-Ba-Ca-Cu-0 superconducting thin-film devices.With the advantages of ewe of formation, the highesttemperature ( 120 K) T1-Ba-Ca-Cu-0 superconductors maybe the first superconductors for practical commercial applicationsat liquid-nitrogen temperatures. However, the hightoxicity of the thallium compounds commonly used in superconductorpreparation has introduced safety concerns inthe fabrication of T1-based superconductors. In order to alleviatethis problem, we have developed a Ti, 0, vapor process'which consists of two steps: (a) preparation of Ba-Ca-Cu-0 precursors by pulsed laser e~aporation,~ and (b)Tl, 0, vapor processing of the precursors. Vapor processinghas produced high quality TI-Ba-Ca-Cu-0 superconductorsin bulk, thick wire, and thick-film form.' In this letter wereport that we have successfully used these techniques forfabricating TI-Ba-Ca-Cu-0 superconducting thin films withzero-resistance temperatures up to 115 K, which is muchhigher than that previously reported by Qiu and Shih'usinga similar processing method.A Ba-Ca-Cu oxide target with a given nominal compositionwas prepared using the method described in Ref. 1. Theoxide film was then deposited by laser deposition onto an Y-stabilized ZrO, substrate at 20 'C using a frequency-doubledNd-YAG laser operating at 523 nm and 10 Hz,' forming aBa-Ca-Cu-0 thin film with a thickness of 2-3 pm. This precursorthin film was subjected to aT120, vapcr process similarto that described in Kef. i. Briefly, a platinum boat containinga small amount of T1,0, (typically 0.1-0.2 g) wasplaced in a quartz boat. The Ra-Ca-Cu-0 precursor thin filmwas then put above the platinum boat. The qqcartz boat withthe contents was put into a tube furnace, which had beenheated to 900 "C and was heated for about 3 min in flowingj5oxygen followcd by furnace cooling.The as-deposited film was a dull black color and an x-a z -ray diffraction scan of the film did not exhibit any peaksBa,Ca,Cu,O, target will be designated by a TI-233 prefix.Figure 1 shows the temperature dependencies of resistancefor a TI, 0, vapor-processed T1-223 film and a T1,0, vaporprocessedT1-233 film. Their zero-resistance temperaturesare 115 and 104 K, respectively.The superconducting behavior of vapor-processed thinfilms depends strongly oc processing conditions. The resistanceof film T1-223B as shown in Fig. 2 reached zero only at80 K since it was heated at too high a temperature (about920 "C) for 3 min. The film Tl-233B did not show a transitionabove liquid-nitrogen temperalure as seen in Fig. 2. Thisfilm was heated at too low a temperature (850 'C) and fortoo short a time (2 min). After reheating at 9C0 "C for 3 min,this film reached zero resistance at 104 K, as also seen in Fig.2. This zero-resistance temperature is identical to film T1-233A, which underwent a.single thallination step.An x-ray diffraction scan of sampie TI-223A is shown inFig. 3. The scan shows that the film is multiphase with ac-axis orientation. The dominant phase is theTi, Ba,Ca, Cu, 0, with the indexed (001) peaks correspondingto a lattice parameter of c = 29.2 b;.' TheT1, Ba,Ca, Cu,O, and TI, Ba2 Ca, Cu,O, phases are alsolo y r.f-- 7--0 50 100 I50 200 2Xl 300other than that of the substrate, indicating that the precursorfilm was amorphous or extremely fine grained. After TI, 0,TEMPERATURE lK1vapor processing, a superconducting thin was formed' FIG, 1, 'kmperdure dcpmdences of redstane for a TI,C), vapor pro-Films from a Baz Ca2 Cu3 O7 target be cessed Bn,Ca, Cu,O, thin film (TI-223A) and u TI, 0, vapor-prmssednated by a T!-223 prefix while films deposited from aB,,c~,cu;o, thillfilm (TI-233~).1810 Appl. Phys. Lett. 54 (18), 1 May 1989 0003-6951 /88/18: 81 0-02$01 .OO @I 1989 American Institute of Physics 1810

-Holland, AmsterdamERCONDUCTIVITY ABOUT 120 K IN THE 71-BiSr-Cn-Cu-0SYSTEMNG, 2.2. SHENG, X. FEI, L. SHENG, J.H. WANG and A.M. HERMANNiversity ofArkansas, Fayelforille, AR 72701, USAmanuscript received 24 September 1989i;A number of the samples in the TI-Bi-Sr-Ca-Cu-0 system were prepared by a solid state reaction method and were characterizedby X-ray powder diffraction, resistivity and AC magnetic susceptibility measurements. A multi-phased sample with nominalcomposition nBiSrlCalCu@Io is superconducting with a T, (onset) = 119 K and T. (R=O) = 106 K. X-ray diffraction analysisof the.samples shows that the crystal structure of the 120 K superconducting phase is different from that of known n-based andBi-based superconducting compounds. Compared with the leaddoped Bi-Sr-Ca-Cu-O compounds, T1-BiSr-Ca-Cu-O compoundshave an advantage of easier forination.1. IntroductionElemental composition of the high-Tc materials,from La-Sr-Cu-0, Y-Ba-Cu-0 to Bi-Sr-Ca-Cu-0 and T1-Ba-Ca-Cu-0, can be described by a generalformula:where M represents RE, T1, Pb, Bi or their combinations;A represents the alkaline earth elements Ca,SF, Ba or their combinations.We believe that new high-Tc superconductors canbe found in the systems described by the formula ( 1 )with new combinations of the elements, Therefore,a complex research project,was planned to explorenew high-Tc materials within the framework of thisformula. Hundreds of samples have been preparedqeording to the general farmula and resistance andAC susceptibility measurzments were made to chara~terizetheir electronic properties. Initially, we improvedthe superconducting properties of some pre-viously studied Sr-Ca-Cu-0 based materials. Anumber of interesting results were obtained, for example,achievement of zero resistance temperaturesat 58 K and 80 K were obtained in the Tl-Sr-Ca-Cu-0 and La-Tl-Sr-Ca-Cu-0 materials respectively.The crystal structure of the superconducting phase inthe RE-T1-Sr-Ca-Cu-0 system was determined tobe of the so-called 1212-type [ 11. Superconductivityabove 100 K in the Tl-Pb-Sr-Ca-Cu-O and Tl-Pb-RE-Sr-Ca-Cu-0 systems were confirmed; superconductivitywith Tc= 82 K in the Tb-Pb-Sr-Ca-Cu-0 system was observed [2]. We subsequentlyfound a new series of high-Tc ( Tc from 40 K to 70 K)compounds in the TI-RE-Sr-Cu-0 system(RE = La, Pr, Ce) without calcium [3]. In this communication,we report the preparation, X-ray characterizationand superconducting properties of T1-Bi-Sr-Ca-Cu-0 compounds with Tc near 120 K. Itshould be pointed out that a few groups previouslystudied the superconductors containing both Tl andBi. Haldar et al. [4] have obtained a T1-Bi-Sr-Ca-Cu-O compound with a Tc(R=O) at 75 K [4].Recently, Li and Greenblatt preparedTb.,Bb.,Sr,CaCu20, using Sr4T120, as a precursor.This compound, with a 1212-type crystal structure,shows a higher Tc (95 K) than that of other knownisostructural compounds [5]. The authors speculatedthat the superconducting temperature is notonly the function of the number of Cu-0 layers perunit cell, but is also a function of the hole providingcapacity of the Tl(Bi, Pb)-0 coupling sheets in thestructure. We tried to prepare the 1223 or 2223 phasein the T1-Bi-Sr-Ca-Cu-0 systems so as to obtainhigher Tc compounds in this system.092.1-4534/89/$03.50 O Elsevier Science Publishers B.V.( North-Holland Physics Publishing Division )

Physica C 160 ( 1989) 489-496North-Holland, AmsterdamDOUBLESTEP BEHAVIOR OF CRITICAL CURRENT VS. MAGNETIC FIELD IN Y-, Bi- ANDTI-BASED BULK HIGH-T, SUPERCONDUCKIRSJ. W. EKIN and T.M. LARSONElectromagnetic Technology Division, National Institute of Stundardr and Technology, Boulder, CO 80303. US4A.M. HERMANN and Z.Z. SHENGDepartment of Physics, University ofArkamas, Foy~rteville. AK 72701, USAK. TOGANO and H. KUMAKURANational Research Imtitute for hfetaL~, 1-2-1, Sengen. Tsukuba-Ci& Ibaraki 305, lawnReceived 31 May 1989Revised manuscript received 19 July 1989A double step characteristic is observed at 76 K io the transport critical m o t as a function of magnetic field ( lo-' T to 10 T)in bulk sintered Y-, Bi- and Tl-based high-T, superconducting mat&&. The low-fidd, aeplikedrop in thecritical cumnt wilyJ, commences at magoetic fieldr Bbctween about 0.3 and 2 mT. This is followed by a plateau eonof ~hdvely coastant crilidcurrent extending from about 30 to 300 mT, and then a second drop at fidds beween about 0.3 and 10 f. Thev fcatum occurfor all three superconductor materia and arc interpaed respectively as a self-field/wcak-link regime, a rumant percolationpath regime and a flux-flow/upper~ritiQ1-field M e . The sharpness of the transition of the v~!t~gC-amnt ( v-1) characteristic,represented by the transition parameter n (i.e., Vql"), has a similar double-sup shape % a fundion of mgnetic fielddirectly corresponding to the features of the JJB) characteristic.1. IntroductionWith the recent discovery of TI- [ 1 ] and Bi-based[2] high-T, superconductors, the practical questionarises as to the character of the transport critical currentin bulk samples of these materials. In particular,is the weak link limitation on the transport criticalcurrent of the granular YIBa2Cu307-d system [3-51at very low magnetic fields ( c about I0 mT) as muchof a limit in these more recently discovered higher-T, materials? Also, what is the character of the higher-T, materials at high fields?In this article [6], the transport critical current inTI-, Bi- and Y-based superconducton has been measuredover five orders of magnitude of magnetic fieldfrom T to 10 T in a single field sweep. The resultsshow that the transport critical current density,J,, at liquid-nitrogen temperature in all three materialsystems shows a similar w a g e drop-off withmagnetic field, B(B=p&); that is, the J,(B) de-pendence has a double-step character. The first dropin J, amounts to between about one and two ordersof magnitude and occurs at magnetic fields between0.3 and 30 mT. This common characteristic is interpretedhere as evidence for a weak-link limitationon the transport critical current of all three systems.This is followed by a plateau regime at magnetic fieldsbetween about 30 mT and 0.3 T where J, is relativelyindependent of magnetic field The value ofJ,in the plateau regime is strongly dependent on thequality of each sample. The second drop in J, occursbetween about 0.3 and I0 T. We interpret this as afl ux-flow/upper-critical-field regime. For the TI andBi systems, rhe drop is much more precipitous thanfor the Y,Ba2Cu307-a (YBCO) system. Finally weshow that, for polycrystalline materials, the drop a ppears to be dominated by thevery low values ( 1-7 I)of the anisotropic upper critical field B& (field pnpendicularto the Cu-O plans). The implication ofthese results is that high-T+ materials may have to09214534/89/$03.500 W e r Science Publishers B.V.( North-Holland Physics Publishing Division )&-?.

PhysicaC 160 (1989) 170-176North-Holland. AmsterdamPRESSURE STUDY OF T, AND JOSEPHSON ABSORPTION IN YBazC~307-d ANDTl-Ca-Ba-Cu-O SUPERCONDUCTORSJ. STANKOWSKI ', B. CZYZAK a, M. KRUPSKI a, J. BASZYNSKI ", T. DATTACarmen ALMASAN b, 2.2. SHENG and A.M. HERMANNImtifute of i\loleculnr Physics. Polish Academy of Sciences. Poznan. Smoluchowskiego 17. PolandUni~ersily ofSouth Carolina. Columbia. SC 29208. USAUniversity ofArkamas, Fayeffwille, AR 72701, USAReceived 22 March 1989Revised manuscript received 22 June 1989The effm of pmsure on high-T. supcrconducton was studied by Josephson absorption as a new and very accurate method fordetermination of the critical temperature. The uncertainty in Tiwas less than 0.06 KWe propose our method as a new stmdardol T, measurements. The pmsure derivative dT,/dpquals 1.06 K GPa-' for YBazCu30,-d and 2.20 K GPa-' for TI-Ca-Ba-Cu-O ceramics. Absence of thermal hysteresis was observed. The change of microwave absorption under hydrostatic preuurewas also investigated.I. IntroductionShortly after the discovery of high-temperature superconductivity,the effect of hydrostatic pressure onthe critical temperature Tc was investigated. ForLaBaCuO Chu et al. obtained a value of 6.3 K GPa-'for the dTc/dp slope [ 11. Later results for YBaCuOgave about 2-1 5 K pa-' (2-61, moreover, someof rhern were of opposite sign [7]; for low pressuredT,/dp was negative, and for pressures ranging from200 hlPa to 400 MPa, dT,/dp was positive. The valueof the slope dTc/dp depends upon the relative valueof resistivity at which the experiment is performed[8]. For the onset temperature T, and for the midpointtemperature T,,,,, the derivatives are dT,/dp= 1 .O and TC.,,,/dp=0.5 KIGPa- I, respectively.Recently published &ta [9] showed, that the valueof dTc/dp for YBa2Cu3O7-s is equal to 0.805 KGPa- ' for pressures lower than 14.9 GPa. Such largediscrepancies between the published values dTc/dphave stimulated us to develop a more accuraternerhod for determination of the phase transitiontemperature, proposed in our previous study of theinternal Josephson absorption [10,1 I].0921-4534189lS03.50 0 Elsevier Science Publishers B.V.( N~rth-Hobd Physics Publishing Division )2. Experimental details and resultsThe YBa2Cu30,dd sample was prepared by thepowder method using Y203, BaCO, and CuO powdersas starting materials. The powders were wellmixed and pulverized in an agate mortar prior to firingat 920°C for 4 h in oxygen. The firing and pulverizationwere repeated twice to improve the homogeneity.The powders were cold pressed intopellets at about 500 MPa, and sintered at 970°C for4 h in oxygen then cooled in a furnace. The powderX-ray diffraction peaks show characteristics of a wellcrystallizedmaterial, with no impurity phases prrs/ent. The structure was orthorhombic with lattice parametersa=381.7pm, b=388.7 pm c= 1167 pmand .k0.15. This orthorhombicity and absence of EPRICU+~ signal leads to the conclusion that single-phasesamples were studied. The multi-phase l7-Ca-Ba-Cu-O sample was prepared from T1203, Ba(NO,),iCaO and CuO using the standard method desmiin the literature 122-241. Measurements of microiwave Josephson absorption were canied out bymeans of a new version [ 12,131 of a high pmsureEPR-probe (fig. 1) attached to the EPR RADI-OPAN SEIX spccuometer. The pressure chamberi'1

130-K T1-Ca-Ba-Cu-0 SUPERCONDUCTORSA.M.HERMANN AND Z.Z.SHENGDepartment of Phyaica, University of Arkansas,Fayetteville. Arkanaaa 72701, USABbstract We have diacovered reproducible and stablebulk superconductivity in the T1-Ca-Ba-Cu-0 eyatemabove 130 K with zero reaiatance above 120 I. Magneticand electronic tranaport propertiee includingthermoelectric power of the new auperconductors arepreaented. The T1-Ca-Ba-Cu-0 ayatem contain6 125-K2223 (T1:Ca:Ba:Cu) phaae and 105-K 2122 phaae, whilethe Ca-free T1-Ba-Cu-0 ayatem conaiata of 80-K 2021phaae, auggeating that the addition of each Cu and Calayer increaaea tranaition temperature by about 20 I.We also preaent recent Josephaon junction studies onthe 2223 phase ahowing the existence of electron pairaupercurrenta and demonatrating weak link behavior at77K.INTRODUCTIONDiacoveriea of 30-K La-Ba-Cu-0 auperconductor 111 and 90-KY-Be-Cu-0 superconductor 121 have atimulated a worldwiderace for higher temperature auperconductors. Breakthroughawere recently made by the diacoveriea of the 90-I T1-Ba-Cu-0ayatem 13.41, 110-K Bi-Ca-Sr-Cu-0 system 15.63 and 120-K T1-Ca-Be-Cu-0 ayatem [7-91. The T1-Ba-Cu-0 syatem is the firstrare earth-free ayatem which reachea eero reelstance aboveliquid nitrogen boiling point [lo], while the T1-Ca-Be-CU-0ayatem la the firat ayatem which reachea eero realstanceabove 100 K and ha8 the highest zero-reaiatance temperature(125 K). In thia paper, we present the preparationprocedurea and electronic propertlea of the 120-I T1-Ca-Ba-Cu-0 superconductors. We also diacuaa Josephson junctionatudiea on T1-Ca-Ba-Cu-0 which ahow the existence ofelectron pair aupercurrenta and weak link behavior at 77K.T1-Ca-Ba-Cu-0 superconductive compounds are eaay toayntheaize; there are many ways to make good-qualityauperconducting aamplea. One of typical procedures inpreparing the T1-Ca-Ba-Cu-0 samples which we use is thefollowing. Appropriate amounta of TlaOs, CaO and Ba-Cuoxide (depending on the deaired atoichiometry) werecompletely mixed and ground, and preaaed into a pellet witha diameter of 7 rnm and a thickness of 1-2 mm. The pelletwaa then put into a tube furnace which had been heated to880-910 oC, and waa heated for 2-5 minutes in flowingoxygen, followed by furnace cooling to below 200 oC.Quenching in air from 900 oC to room temperature depreaaesTc only slightly. A number of aamplea with differentatoichiometry, including a aeries of aamplea with nominalcompoaitiona of TlzCarBaCusO'l+r*X with y = 1, 1.5. 2. 3 andMRS Int'l. Mtg. on Adv. Mats. Vol. 6 01989 Matdals Research Sodety.

I58~ 120 160 200TEMPERATURE (K)PHYSICAL REVIEW B VOLUME 40, NUMBER 10 1 OCTOBER 1989Ultrasonic attenuation measurements on tballim-based high-temperature superconductorsD. 0. Pederson. A. El Ali, Z. Z. Sheng, and A. M. HermannPhysics Department, University of Arkansas, Fayerteville, Arkansas 72701(Received 24 April 1989; revised manuscript received 12 June 1989)We report ultrasonic attenuation measurements on thallium-based high-temperature superconductorswith a transition temperature near 110 K. Measurements were camed out between 300K and liquid-nitrogen temperature. Several peaks were observed with one peak appearing nearT,. At the critical temperature, no sharp drop was noted within our experimental resolution. Theattenuation increases with decreasing temperature below T,. Similar behavior of attenuationmeasured in Y-based compounds was also observed.Since the discovery of high-temperature superconductivity,' several groups reported measurements of ultrasonicattenuation in both La compounds2-' and Y-based superconductors.3-'6 Many theories have becn suggested tounderstand the behavior of both systems. I7-l9 None ofthe reported experimental results show a sharp drop of attenuationnear T, as would be expected in a normalphonon-mediated BCS-type superconductor. After thediscovery of the thallium-based superconductors by Shengand ~ermann,~~.~' extensive studies were carried out tocharacterize these new compounds. No measurements ofultrasonic attenuation on thallium-based superconductorshave been previously published although preliminary resultshave been p~esented.~~.~~ The pnsent work reportsmeasurements of ultrasonic attenuation on T1-based superconductorsbetween 300 K and liquid-nitrogen temperature.The previously reported ultrasonic attenuation measurementsdo not exhibit a sharp drop near the transitiontemperature as predicted by the BCS theory. Most previouswork on Y-based superconductors finds that the ultrasonicattenuation generally decreases as the temperatureis decreased below 150 ~ . ~ - ' 9 ~ . ~ . ' ~ 9 ' ~ - ' ~ ~ ' Near ~ T,,peaks are often ~bserved~.~-~.'%'' while in only two cases aclear minimum is ob~erved~,~ (Ref. 7 has a peak at 10MHz and a minimum at 30 MHz).The samples studied were prepared by mixing appropriateamounts of CaO, BaCO3, and CuO. These powderswere mixed and heat treated for 24 h to form the prtcursorCa2Ba2Cu30x. CazBazCu,Ox was ground and mixedwith Tlz03. The mixed powder was pressed into pellets of1.27 cm diam and 0.7- 1.0 cm thickness. The pellets wereput in a preheated furnace at 890-910°C, heated for 3-5min in flowing Oz and furnace cooled to room temperature.Preliminary analysis showed that the resistance of thesesamples is zero at 110 K. A standard four-probe rcsistancemeasurement on a sample similarly prepared to thatsample used in our attenuation measurements is shown inFig. 1. X-ray analysis of similarly prepared samplesshowed that they were multiphase with a dominant Tlz-Gal BazCu20, phase.Samples were polished to 1-2-pm finish, with planeparallel faccs. A 36O Y-cut, lithium-niobate transducerwas bonded to one face with Nonaq Stopcock Grease. AMatec MBS-8000 system which utilizes a phase-detectiontechnique, was used with a single-longitudinal-transducersetup. No thermal-expansion corrections were made.Samples were cooled from room temperature at 80 K andTEMPERATURE (K)FIG. 1. dc resistance-temperature measurements on a similarlyprepared sample to that used for attenuation measurements.FIG. 2. Temperature dependence of attenuation in a TIbadsuperconductor. Shown is the cooling down from roomtemperature to 80 K (open circles) and the warming up from 80to 220 K (filled circles).73 13 01989 The American Physical Society

aC 158 (1989) 507-510Hohd, Amsterdam90 K BULK SUPERCONDUCIlVITY IN THE TI-Ba-Ce-Cu-0 SYSTEMJ.H. WANG ', Z.Z. SHENG, C. DONG, X. FEI, L. SHENG and A.M. HERMANNDepartment of Physics, University ofArkansas. F~yetteville, Arkansas 72701, USAEZ.X. ZHAOInstitute of Physics, Chinese Academy of Sciences, Beoing, P.R. China, Received 21 March 1989Revised manusaipt meived 14 April 1989Bulk superconductivity at 90 K has been observed resistively and by AC susceptibility in the TI-Ba-Ce-Cu-0 system. Theresults am easily reproduced, and the samples are stable. Powder X-ray diffraction data are presented and discussed,Discoveries of the TI-Ba-Cu-0 and Tl-Ba-Cau-0superconducting systems [ 1-5 ] have led toentification of a number of Tl-based supercongcompounds, including the series Tl,-n-lCunOl.Sm+2n+l with m=2 16-91 and with[lO,ll, and n=1-5 [6-111. TheBa2Ca2Cu3010 phase has the highest reproducibleo resistance temperature of 125 K to date,5@,9]. In order to clarify the high T, oxide surconductingmechanism and search for new anden higher temperature superconducting systems,we have carried out extensive elemental substitunsfor the Tl-based superconducting systems. In aof experiments, we found that when Ca2+ in theBa-Ca-Cu-0 system is totally replaced by Y3+,e new Tl-Ba-Y-Cu-O system forms a semiconctorTlBa2YCu206., ( 1212 phase). Since Tl-2CaCu206.s and 1212 phases in some other sysemsare superconducting, we have replaced Y3+ ine TI-Ba-Y-Cu-O system-with Ce4+, to study thesubstitution effects of a 4+ valence rare earth onelectronic properties and structure of the system. Asa result, we found that the Tl-Ba-Ce-Cu-0 systemis superconducting at 90 K. The results are easily reproduced,and the samples are stable in ambient atmosphere.In this paper, we report preparation pro-* Also Institute. ~hihese Academy of Sciences, Beijing, People'sRepublic of China.cedure, resistance and AC susceptibilitymeasurements, and powder X-ray diffraction resultsfor the new superconducting system.The Tl-Ba-Ce-Cu-0 samples were prepared usinghigh-purity T1203, Ba02, Ce02, and CuO. In atypical procedure, appropriate amounts of n2O3,BaO,, Ce02, and CuO were completely mixed,ground, and pressed into a pellet with a diameter of7 mm and a thickness of 1-2 mm. The pellet was putin an alumina crucible. The crucible and its contentswere put in a tube furnace, which had been heatedto 900-925 "C, and heated in flowing O2 for about3-5 minutes, followed by furnace cooling. NominalTl,2Ba2CeCu30,.3 samples prepared using the aboveprocedure are superconducting at 90 K. The sampleshave a dark grey color, and are dense. Superconductingbehavior of the samples is sensitive to bothheating temperature and heating duration. However,the above procedure does change significantly TIcontent, for example, Tl in a nominalT12.2Ba2CeCu309.3 sample, after processing, decreasesfrom 2.2 to about 1.75, assuming that Ba, Ce,and Cu remain unchanged. Quenching depresses T,only slightly. Samples prepared in Ar atmosphere arenot superconducting.Resistance (AC, 27 Hz) was measured by thestandard four-probe technique with silver paste contacts.AC (27 Hz) susceptibility was measured usingthe technique similar to that by Norton [ 121. Both0921-4534/89/$03.50 O Elsevier Science Publishers B.V.( North-Holland Physics Publishing Division)

- Modem Physics Letters B Vol. 3, No. 3 (1989) 249-2550 World Scientific Publishing Company. -4I*PREPARATION AND CHARACTERIZATION OF UNUSUALLEVITATION T1-Ba-Ca-Cu-0 SAMPLESZ. Z. SHENG, Y. H. LIU, X. FEI, L. SHENG, C. DONG, W. G. HARTER, Iand A. M. HERMANNDepartment of Physics, Universily of Arkansas, Fayetleville, Arkansas 72701, USA1D. C. VIER and S. SCHULTZDepartment of Physics, University of Calfirrtia, San Diego, ta Jolla, Cali/ornia 92093. USAandS. B. OSEROFFDepartment of Physics, San Diego State University, San Diego, Calijornia 92182, USAReceived 10 November 1988A preparation procedure for the TI-Ba-Ca-Cu-0 superconducting samples which can belevitated above or beneath a magnet is described. X-ray powder diffraction data for thesesamples are presented and discussed. Their electronic and magnetic properties arepresented. The explanation for the unusual levitation is briefly discussed.I1IDiscoveries of the 90 K TI-Ba-Cu-0 system'12 and the 120 K TI-Ba-Ca-Cu-0 superconducting system3 have led to the identification of varioussuperconducting compoundsc8 and the observation of superconductivity in newTI-based superconducting systems: Tl-Sr-Ca-C~-0,~-'l TI-Pb-Sr-Ca-Cu-012and TI-R-Sr-Ca-Cu-0 with R -- rare-earths.I3 In addition to this progress,which has contributed to the understanding of the mechanism of high temperatureoxide superconductivity and to the search for higher temperature superconductors,an interesting levitation phenomenon - a superconductor may belevitated above, to the side of, or beneath a magnet - has been observed for someTl-Ba-Ca-Cu-0 samples.14 This kind of levitation is denoted as unusual inorder to distinguish it from the normal levitation (only above a magnet). Previousexamples of the unusual levitation were achieved for some Ago-dopedYBa2Cu,0, samples." While the rare-earth based-superconductors which displaythis unusual levitation must be doped with silver oxide, the TI-based superconductorsshowing this behavior need to be specially prepared. In this letter, wedescribe the preparation procedure of the unusual TI-based samples. Accordingto our knowledge, this is the first description of a preparation procedure for highqualityunusual TI-based samples which can be fabricated with great consistency249

High Temperature Superconductivity. The First Two Years. Proceedings of a conference, 1989, p 175-80120-K TL-CA-BA-CU-0 BULK SUPERCONDUCTORSA-M-HERMANN AND Z.Z.SHENGDepartment of Physics, University of Arkansas,Fayetteville, Arkansas 72701, USAAbstract We have discovered reproduci ble and stablebulksuperconductivi ty in the TI-Ca-Ba-Cu-0 systemabove 130 K, with zero resistance above 120 K. Magneticand electronic transport propert'i es including thermoelectricpower of the new superconductors are presented. TheTI-Ca-Ba-Cu-0 system contains the 125-K 2223 (TI: Ca:Ba:Gi)phase and the 105-K 2122 phase, while the Ca-freeTI-Ba-Cu-0 system consists of the 80-K 2021 phase, suggestingthat the addition of each Cu and Ca layerincreases the transition temperature by about 20 K.INTRODUCTIONDiscoveries of the 30-K La-Ba-Cu-0 superconductor [ll and the 50-KY -Ba-Cu-0 superconductor [2] have st imul ated a worl dwi derace for higher temperature superconductors. Breakthroughswere recently made b.y the discoveries of the 90-K TI-Ba-Cu-0system PA] the 11&K Bi -Ca-Sr-Cu-0 system L5,6] and the 12&K TI -Ca-Ba-Cu-0 system [7-91. The TI-Ba-Cu-0 system is the firstrare earth-free system which reaches zero resistance aboveliquid nitrogen boiling point, while the TI-Ca-Ba-Cu-0system is the first system which reaches zero resistanceabove 100 K and has the highest zero-resistance temperature(125 K). In this paper, we present the preparationprocedures and some properties of the 120-K TI-Ca-Ba-Cu-0superconductors.PREPARATIONTI -Ca-Ba-Cu-0 superconductive compounds form easi 1 y; thereare many ways to make good-qua1 i ty superconducti ng sampl es.One of the typical procedures in preparing the TI-Ca-Ba-Cu-0

PHYSICAL REVIEW B VOLUME 39. NUMBER 4 1 FEBRUARY 1989Superconductivity above 77 K in the R-TI&-Ca-Cu-0 system (R represents rare earths)Z. Z. Sheng, L. Sheng, X. Fei, and A. M. HermannDepartment of Physics. University of Arkansas, Fayettevllle. Arkansas 72701(Received 19 September 1988; revised manuscript received 24 October 1988)Superconductivity at 80-90 K has been resistively and magnetically observed in the R-TI-Sr-Ca-Cu-0 systems (R represents the rare earths including Y, La, Pr, Nd, Sm, Eu, Gd. Tb, Dy,Ho, Er. Tm, Yb, and Lu and excepting Ce) in which five metal elements all are required for thesuperconductivity. In particular, the ~r-TI-Sr-~a-CU-0 system and Tb-T1-Sr-Ca-Cu-0 systemare the first superconducting systems above 77 K in which Pr or Tb are required for the superconductivity.We believe that even higher critical temperatures may be achieved by optimizing thestarting composition and preparation conditions, and/or by further elemental substitutions.Discoveries of the Tl-based superconducting system~'-~have provided increased T,, with zero-resistancetemperatures as high as 125 K .~-~ A number of TI-basedsuperconducting compounds have been identified, includingthe series TImBa2Can-~C~nOl.5m+2n+l with m-2(Refs. 6- 11) and 1 ,I2*l3 and n - 1-4. '-I3 The T, of thesecompounds increases with n as well as m. These trendshave stimulated a strong search for higher T, TI-based superconductorsby increasing n and/or m. On the otherhand, search for new and even higher temperature T1-based superconducting systems has been also widely carriedout. In extensive elemental substitutions, we.have observedsuperconductivity at 20 and 70 K in the Tl-Sr-Ca-Cu-0 system.14 Although a sharp drop in resistance at,about 100 K was observed in our early experiments and asimilar superconductivity was reported by otherthese results are difficult to reproduce, suggestingthat the formation of this high-temperature superconductingphase is very sensitive to preparation procedure.Recently, Subramanian et a1. l7 found that additionof Pb into TI-Sr-Ca-Cu-0 systern results in formationof high temperature T1-Pb-Sr-Ca-Cu-0 superconductingphases. Two phases, with T, 80-90 K for(Tlo.sPbo.s)Sr2CaCu207 and 120 K for (Tlo.5Pb0.5)-Sr2CazCu309, have been identified. We have obtainedsimilar results. '' In this paper, we report superconductivityat 80-90 K in the R-T1-Sr-Ca-Cu-oxide systems with Rrepresenting the rare earths including Y, La, Pr, Nd, Sm,Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb, and Lu and exceptingCe. In these systems, five metal elements all are requiredfor the 80-90 K superconductivity. In particular, the Pr-TI-Sr-Ca-Cu-0 system and Tb-T1-Sr-Ca-Cu-0 systernare the first superconducting systems above 77 K in whichPr or Tb are required for the superconductivity. We believethat higher Tc may be achieved by optimizing startingcomposition and preparation conditions. We also believethat new, even higher temperature superconductingsystems may be discovered by further elemental substitutionsfor these systems.The R-T1-Sr-Ca-Cu-oxide samples were prepared usinghigh-purity T1203, SrC03, CaC03, CuO, and rare-earthoxides (R203 with exceptions of Ce02 and Tb4O7). First,appropriate amounts of SrCO3, CaC03, and CuO weremixed and ground. The mixture was heated at 950°C inair for at least 24 h with several intermediate grindings.The resulting Sr-Ca-Cu-0 powder served as master material.In a typical procedure, appropriate amounts of Sr-Ca-Cu-0 powder, T1203, and a rareearth oxide weremixed, ground, and pressed into a pellet with a diameterof 7 mm and a thickness of 1-2 mm. The pellet was putinto a furnace, which had been heated to 900°C, and washeated in flowing oxygen for 5 min, followed by furnacecooling. Both low-frequency ac resistance and dc resistancewere measured by the standard four-probe techniquewith silver-paste contact. Low-frequency (5-kHz)ac susce tibility was measured using the technique ofNorton. I!Figure 1 shows ac resistance versus temperature curvesdown to 77 K for samples whose starting compositionswere given by RT12Sr2Ca2Cu3011.5 (hereafter, R1:2:2:2:3), where R represents the rare earths Y, La, Ce,Pr, Nd, Sm. Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb, and Lu.All samples except R -Ce showed a sharp drop in resistancein 'the range of 80-90 K. Resistance temperaturevariations in the normal state were linear for most sam-TEMPERATURE (K)FIG. 1. Temperature dependences of the ac resistance for R1 :2:2:2:3 samples prepared using the typical procedure.1989 The American Physical Society

International Symposium on New Developments in Applied Superconductivity, 1989, P 5-10HIGHEST TEMPERATURE (120 K) TL-BA-CA-CU-0 SUPERCOFIDUCTING SS{STEM5A.M.Hermannand Z.Z.ShengDepartment of Ph sics, University of ArkansasFayettevil y e, Arkansas 72701, USA--Abstract The highest temperature (120 K) Tl-Ba-Ca-Cu-0 superconductingsystem comprises a number of superconducting compounds. The TI-Ba-Ca-Cu-0superconductors are easily made. The structure, electronic and magneticroperties are presented. An unusual levitation phenomenon, in whlch the T1-[a-~a-CU-0 superconductor can be suspended above, below, or to the side of amagnet, is discussed. A new TI 0 -vapor process for fabricating the TI-Ba-Ca-Cu-0 superconductors is descgized.IntroductionDiscoveries of 30-K La-Ba-Cu-0 superconductor [I] and 90-K Y-Ba-Cu-0superconductor [2] have stimulated a worldwide race for new and even highertemperature superconductors. Breakthrou hs were made by the discoveries ofthe 90-K TI-Ba-Cu-0 system [3,4], 110-K I i-Sr-Ca-Cu-0 system [5,6] and 120-KTI-Ba-Ca-Cu-0 system 7-91, Recently, high temperature superconductivity wasalso observed in the f 1-Sr-Ca-Cu-0 s stem [lo-121, and in the M-TI-Sr-Ca-Cu-0with M = Pb [13,14] and rare earths f15]. In thi ., paper, we presentoreparation procedures, structure, and some propeeties of the 120-K T1 -Ba-Ca-:u-0 superconductors. We discuss an unusual 1evi:ation phenomenon of the T1-Ba-Ca-Cu-0 superconductor due to flux pinning [le]. Finally, we present anew T1203-vapor-process [17] which allows the hishest temperature TI-Ba-Ca-Cu-0 superconductors to be easily made in the f0.m of complex bulkcom onents, wires and fibers, and thick and thin films, and minimizespro g lems caused by toxicity and volatility of Tl starting compounds.PreparationTI-Ba-Ca-Cu-0 superconductive corr~poundr form easil ; there are many waysto make good-quality superconductin Sam lej. One of t e typical proceduresin preparing the TI-Ba-Ca-Cu-0 samp 9 es w f: ich we use is the following. Ba-Ca-Cu-oxides are first prepared using the method similar to that we previouslydeveloped for preparation of Ba-Cu-oxides [18,19]. Ap ropriate amounts ofBaCO CaO or CaC03), and CuO are mixed, ground, and eated at 925-950 OC inair f~r ag24- 8 hour with several intermediate grindings. The resultinguniform black material is ground and served as master material. Ap ropriateamounts of T1203 and Ba-Ca-Cu oxide (depending on the desired stoic ! iometry)are completely mixed and ground, and pressed into a pellet with a diameter of7 mm and a thickness of 1-2 mrn. The pellet is then put into a tube furnacewhich had been heated to 880-910 OC, and is heated for 2-5 minutes in flowingoxygen, foll owed by furnace cool ing to be1 ow 200 OC.StructureThe TI-Ba-Ca-Cu-0 system can form a number OF superconducting phases..wo phases, T1 Ba Ca Cu 0 O,X (2223) and Tl2Ba2CalCu2Osfx (2212), were firstidentified [201. '~h; 232j superconductor has a 3.85 x .85 x 36.25 Atetragonal unit cell. The 2212 superconductor has a 3.85 x 3.85 x 29.55 Atetra onal unit cell [20,21]. The 2223 phase is related to 2212 by additionof ex ? ra calcium and copper layers. In addition, the su erconducting phasein the Ca-free TI-Ba-Cu-0 system is T12Ba2Cu06+x (2201) !20,22]. Fig. 1X

APPLIED PHYSICS COMMUNICATIONS, 9(4), 2Lj-255 (1989-90)Effects of Ag Addition to the TI - Ba - Ca - Cu - 0Superconducting SystemD. ~arsht, F. ~rarnrnashf, J. ~ennettt, R. ~ eck~, A.R. El lit,D.E. weekst f , and A.M. ~ennanntfiThe electronic and structural properties of TI - based supcrconducton with Ago ;iddedir: 3e range from 2% to 20% by weight are studied. Ln all cases Ago addition decreases the2:-.:unt of superconducting phase produced. Small amounts of Ago addition appear to slightlyir:.:rease zero resistance temperatures and in the case of the 2% sample, dramatically increase thecr.;icd current. In samples with larger amounts of Ago addition unreacted Ca2Ba~Cuj07 is.jbservcd as well as silver med. The bulk nsistiviry of all the samples is greater than an undopcdcontrol sample with the exception of the 208 sarnple with a resistivity of l/7 of the control smplc.In all cases silver oxide addition did not destroy superconductivity.I) Introduction tThe development of technological applications that incorporate the recently discoveredYBa2CugOy (YBCO). Bi-PbCa-Cu-0 (BPCCO). and TI-Ba-Ca-Cu-0 WCCO) superconductingPhysics Department. University of Arkansas, Fayetteville Ar,72701tf Hebrew University, Jerusalem 91Y04. IsraelPhysics Department. University of Colorado. Boulder, CO 80303COpflghl O 1990 by Marcel Dckkcr. Inc.

C161 (1989)257-261Holland, Amsterdam.*'gsSUPERCONDUCIIVITY ABOUT 120 K IN THE TI-BiSr-Ca-Cu-0 SYSTEMC. DONG, Z.Z. SHENG, X. FEI, L. SHENG, J.H. WANG and A.M. HERMANNDepartment ofPhysics, University ofArkanss. Fayetteville, AR 72701, USAReceived 10 July 1989Revised manuscript received 24 September 1989A number of tbe samples in the TI-Bi-Sr-Ca-Cu-O system wen prepared by g solid state reaction metbod and werc charsmrizedby X-ray powder ditfraction, resistivity and AC magnetic susceptibility muuurements. A multi-phaud samph witb n~rPinalcomposition TlBiSr2CapCu,0,0 is superconducting with a T, (onset) = 119 K and T. (R= 0) = 106 K. X-ray diffraction Bnalysisof the.samples shows that the crystal structure of the 120 K superconducting phase is different fmrn that of Wwq TI-W andBi-based superconducting compounds. Compared with the leaddoped Sir-Ca-Cu-O compounds, TI-Bi4r-Q-Cu-O campoundshave an advantage of easier formation.1. Introduction, Elemental composition of the high-Tc materials,ffom La-Sr-Cu-0, Y-Ba-Cu-0 to Bi-Sr-Ca-Cu-0 and TI-Ba-Ca-Cu-0, can be described by a generalformula:M-A~u-0 (1)where M represents RE, TI, Pb, Bi or their combinations;A represents the alkaline earth elements Ca,Sr, Ba or their combinations.We believe that new high-T, superconductors canbe found in the systems described by the formula ( 1 )with new combinations of the elements' Therefore,a complex research project, was planned to explorenew high-Tc materials within the framework of thisformula. Hundreds of samples have been preparedaccording to the general f~nnuland resistance andAC susceptibility measurgments were made to characterizetheir electronic properties. Initially, we improvedthe superconducting properties of some pre-viously studied Sr-Ca-Cu-0 based materials. Anumber of interesting results were obtained, for example,achievement of zero resistance temperaturesat 58 K and 80 K were obtained in the TI-Sr-Ca-Cu-0 and La-T1-Sr-Ca-Cu-O materials respectively.The crystal structure of the superconducting phase inthe RE-T1-Sr-Ca-Cu-O system was determined tobe of the so-called 1212-type [ 1 1. Superconductivityabove 100 Kin the TI-Pb-Sr-Ca-Cu-O and TI-Pb-.RE-Sr-Ca-Cu-O systems were confirmad; super-opnductivity with Tc= 82 K in the ~b-~&s-~a~Cp-0 system was observed [q]. Wa oubaequpntlyfound a new series of high-T, ( T, from 40 IS t~ 79 K)compounds in the lJ-RS4r-Cu-Os system(RE=La, Pr, Ce) without calcium [31. In thia aommuoication,we report the p&paration, X-ray qhvacterizationand superconducting propcrliet $F T\-Bi-Sr-Ca-Cu-0 compounds with Tc neq 13QK I1should be painted out that a few groupr'gtqyi~~dystudied the supercoaductors containing bd$ TI s ~ dBi. Haldar et al. [4] have obt&cd a Tl-Bi-Specti-Cu-O compound with a T,(.&=$),&,Recently, mql? #@WlaTfo.,Bi,,.,Sr2CaCu20y using Sr4T1207 as aThis compound, with a 1 2 12-type crystal structure,shows a higher Tc (95 K) than tbat of other knowisostructural compounds [5). The authors $muliltedthat the superconducting temperature is nstonly the function of the number d Cu-O layers perunit cell, but is also a function of the hole providingcapacity of the Tl(Bi, Pb)-0 coupling sheets in thestructure. We tried to prepare the 1223 or 2223 phasein the Tl-Bi-Sr-Ca-Cu-0 systems so as to obtainhigher T, oompqpqds in this system.092.1-4534/89/$03.50 O Elsevier Science Publishers Y.V.( North-Holland Physics Publishing Division )

APPLIED PHYSICS COHHUNICATIONS, 9 ( 1~2) , 27-33 (1989)SUPERCONDUCTIVITY AT 82 K IN THE TB-PB-SR-CA-CU-0 SYSTEMZ.Z.Sheng, C.Dong, X.Fei, Y.H.Liu, L.Sheng, and A.M.HennannDepartment of PhysicsUniversity of ArkansasFayetteville, AR 72701Superconductivity at 82 K has been observed by lowfrequency ac susceptibility and resistance measurements inthe Tb-Pb-Sr-Ca-Cu-0 system. The powder x-ray diffractionpattern, compared with related published data, can be indexedfrom an orthorhombic lattice with a=5.38 A, b=5.42 A, and-15.71 A,

APPLIED PHYSICS COHHUNICATIONS, 9(3), 129-146 (1989)Formation of Ti-Ca-Ba-Cu-0 Superconducting Thick Filmsby Vapor Evaporation of 1'1203 into Rolled Ca2Ba2Cu30~J.D.Smith, Q.A.Shams, M.J.Saeed, D.Marsh, F.Arammash,J. Bennett, Z.Z.Sheng, and A.M.HerrnannDepartment of PhysicsUniversity of ArkansasFayetteville, AR 72701M.J.Dreiling and M.D.PhilllpsPhillips Petroleum CompanyBartlesville, OK 74004Abstract:Using a roller technique, Ba2Ca2Cu30x thick filmprecursors (about 100 microns in thickness)were prepared. These films were then treated bya vapor process using TI203 to formsuperconducting TI-Ba-Ca-Cu-0 films. Thesuperconducting onset temperature of thesefilm is about 120K with a zero-resistancetemperature ranging from 91K to 111K. X-raySEM, and electron mlcroprobe anaylses of thesefilms are presented, and, on the basis of theseanalyses, crystallite growth will be discussed.

TI-Based Cu-0 High Temperature SuperconductorsA.M. HERMANN and Z.Z. SHENGDepartment of Physics, University of Arkansas, Fayetteville, AR 72701, USA~dvances in Superconductivity. Proc. of the 1" International Symposium on Superconductivity, 1989, p 763-71Abstract We have discovered reproducible and stable bulk superconductivity in the T1-Cah-Cu-06~6t5m with zero resistance above 120 K. Msgnetic and electronic tran6px-tproperties including thermoelectric power and critical current of the new superconductorsare presented. The T1-Ca-Ba-Cu-0 system can form a number of superconducting cow&,including TlmCan-iBazCunOi.~mtzn+l with m = 1 and 2, and n = 1, 2, 3, and 4. The Tcincreases with m and n. The 2223 (T1:Ca:Ba:Cu) phase has the highest Tc (125 K) to date.We present Josephson junction studies on the 2223 phase showing the existence of electronwir supercurrents and demonstrating weak link behavior at 77 K. We discuss an unusuallevitation phenomenon in which the 2223 phase can be suspended, above, below, or to theside of a ring shaped magnet. We also discuss a new T1-based superconducting system, inwhich Sr replaces the Ba, the T1-Ca-Sr-Cu-0 system.Discoveries of 30-K La-Ba-Cu-0 superconductor I:. and 90-K Y-Ba-Cu-0 superconductor [Z]have stimulated a worldwide race for higher tempera' :re superconductors. Breakthroughs wererecently made by the discoveries of the 90-K T1-Ba-'G-0 system [3,4], 110-K Bi-Ca-Sr-Cu-0 system[5,61 and 120-K TI-Ca-Ba-Cu-0 system [7-91. The T1 3-Cu-0 system is the first rare earth-freecystem which reaches zero resistance above liquid r.?rogen boiling point, while the T1-Ca-Ba-Cu-0 system is the first system which reaches zero reL jtanCe above 100 K and has the highest zero-"esistance temperature (125 K). In this paper, we -resent the preparation procedures and 60~~3properties of the 120-K T1-Ca-Ba-Cu-0 superconduct -3, we discuss unusual levitation phenomenadue to flux pinning, and we present Josephson jun-- -on studies on T1-Ca-Ba-Cu-0 which show theexistence of electron pair supercurrents and weak mk behavior at 77 K. We also presentpreliminary data on a new T1-based superconducting system, T1-Ca-Sr-Cu-0.PREPARATIONT1-Ca-J3a-CII-0 superconductive compounds i nn easily; there are many ways to make gccdqualitysuperconducting samples. One of typizal procedures in preparing the TI-Ca-Ba-Cu-0cm~ples which we use is the following. Appr.:priate amounts of T1203, CaO and Ba-Cu oxide(depending on the desired stoichiometry) wer2 completely mixed and ground, and pressed into awllet with a diameter of 7 rnm and a thickness of 1-2 m. The pellet was then mt into a tubefurnace which had been heated to 8811-910 or, and was heated for 2-5 minutes in flowing oxygen,followed by furnace cooling to below 200 oC. Quenching in air from 900 oC to room temperaturedepresses Tc only slightly. A number of samples with different stoichiometry, including a~eries of samples with nominal compositions of TlzCayBaCu307+y+x with y = 1, 1.5, 2, 3 and 4,were prepared. Samples with nominal com~ositions of Tli.eaCaBaCu307.~tx, TlzCaBazCuzOstx andTlzCazBazCusOl otx were also synthesized. All samples readily levitate over a magnetic field .The strong Meissner effect observed is due to both a large volume fraction of superconductivephase and a higher transition temperature. Recently, most samples were prepared using precursorBa-Ca-Cu-oxides.The TI-Ca-Ba-Cu-0 system can form a number of superconducting phases. Two phases,T1zCazBazCu~C)iotx (2223) and TlzCaiBazCuzOetx (2122), were first identified [lo]. The 2223cupx-conductor hasp 3.85 x 3.85 x 36.25 A tetragonal unit cell. The 2122 superconductor, whichappears to be structurally related to BizCaiSrzCuzOstx, has a 3.85 x 3.85 x 29.55 A tetragonalunit cell [10,111. The 2223 phase is related to 2122 by addition of extra calcium and copper'syers. In addition, the superconducting phase in the Ca-free T1-Ba-Cu-0 system is TlzBazCuOstx.2021) [10,121. Fig. 1 shows schematically the arrangements of metallic atoms in these threeT1-based superconducting phases. The 2021 phase has a zero-resistance temperature of abut 80K, whereas the 2122 and 2223 phases have zero-resistance temperatures 108 K and 125 K,reepectively [lo-151. It aPFears that the addition of a Ca and Cu layer increases the

Solid State Communicatiane, Vol. 71, No. 9, pp. 739-741. 1989.Printed in Crest Rritain.0038-1098/89$3-OO+.OOPergawn Preee plcSUPERCONDUCTIVITY IN THE R-TI-Sr-CU-0 SYSTEH WITH R = RARE EARTHSZ.Z.Sheng, C.Dong, X.Fel, Y.H.Liu, L.Sheng,J.H.Wang and A.H.HermannDepartment of Physics, University of Arkansas, Fayetteville, AR 72701, USA(Received January 3. 1989, in revised 'form on June 19. IS89 by R. H. Silsbee)1 Superconductivity above 40 K has been observed by low frequencyac susceptlbil lty and resistance measurements in the R-TI -Sr-CU-0system wlth R = La, Pr, Nd, and Y. Powder x-ray diffractionanalyses for a nominal La2Tl Sr Cu30 0 sample and a nominalPr TI lr CU 0 sample show4 tiat thy m nearly single phasean$ cfn 2e fnigxed to a tetragonal unit cell with a-3.775kO.005 Aand c+35.35+0.05 A, and with a=3.762+0.005 h and c-35.34i0.05 A,respect1 vel y .;6hDiscoveries of the 90 K TI-Ba-Cu-0 systemResistance (ac, 27 Hz) was measured by thej [1,2] and 120 K TI-Ba-Ca-Cu-0 system 13-51 have standard four-probe technique with silver paste"cad not only to the identification of a number contacts. AC (500 Hz) susceptibility wasf of TI-containing superconducting compounds.measured using the technique similar to that by5 including T1,Ba Ca - Cu 01 with -1-2, Norton [17]. Powder X-ray diffraction wasand n-1-5 [6-llf, 6uI aPs0'~~f6:~discoveries of carried out with CU-K* radiation with use of a( a series of Sr-Ca-CU-0-based superconducting DIANO DTM 1057 diffractometer.?L systems: the 70 K TI-Sr-Ca-Cu-0 system 1121, the Properties of the R-TI-Sr-Cu-0 samples% 115 K Pb-TI-Sr-Ca-Cu-0 system [13], the 90 K depended strongly on both the rare earths and theR-TI -Sr-Ca-Cu-0 system wlth R = rare earths [14], preparation procedures. Superconducting samplesk; and the 70 K Pb-R-Sr-Ca-Cu-0 system [15,16]. with R = La and Pr were prepared using both3 Further experiments reported in this paper on the procedures, and superconducting samples with R =90 K TI-R-Sr-Ca-Cu-0 system have showed that the Nd and Y were made only using the firstR-T1 -Sr-CU-D system (without Ca) is also procedure. Neither procedure has unambiguouslysuperconducting at lower temperatures (about 40 produced superconducting samples with rare earthsK). i@like the R-TI-Sr-Ca-Cu-0 system in which other than La, Pr, Nd, and Y. Fig. 1 showsthe rare earths do not influence significantly resi stance-temperature dependences for norni nalthe superconducting behavior, the rare earths In La2TlpSr Cu 0 1, nominal Pr2T12Sr2Cu301 , andthe R-TI-Sr-Cu-0 system change greatly thenominal b~j~b C U ~ O 5 ~ sampler. ~esisf ancesof the system. So far only samples tenperature varfat~oni of these samples areLa, Pr, Nd, and Y have been observed to strictly linear in the nomal state up to roombe superconducting unambiguously above IS K. In temperature. The superconducti ng transitions ofthis paper we present preparation procedures, ac the La sample and the Pr sample are fairly sharp.suscepti bi 1 i ty and resistance measurements, andpowder x-ray diffraction data.The R-TI -Sr-Cu-oxide samples ware prepared 700using high-purity T1203, SrO ,(or SrC03). CuO and600rare earth oxides (R20 with exceptions of CeO2-and TblO7). Two proce~urerrere used. In onec E 500procedure, component oxides with a certain molarratio *ere completely mixed, ground, and pressed :, 400into a pellet with a diameter of 7 mm and athickness of 1-2 m. The ellet was heated in a 3 300gtube furnace at about 930 C for 3-5 minutes inflowing oxygen, foll owed by furnace-cool ing or g 200uenching. In a second procedure, a Sr-Cu-0 wrecursor was first prepared by mixing and 0: 100.rinding appropriate arnounts of SrCO qnd CuO,nd heating at 950 OC in air for at feast 24 0 .ours with several intermediate grfndings. The 0 50 100 150esulting Sr-Cu-0 powder served as masteraterial. Appropriate amounts of Sr-Cu-0 powder,TEMPERATURE (K)1 03, and a rare earth oxide were mixed, groundn$ pressed into a pellet. The pellet was heated Figure 1 Resi stance-temperaturen a tube furnace at 900 OC for 5 minutes in dependences for a nominal La T12Sr2Cug0 samplelowing oxygen, followed by furnace-cooling or (La), a nominal Pr TI ~ r ~ sample ~ u fbr), ~ 6 and ~i nominal ~ d ~ 1 ~ ~ sample r ~ ~ (~d). u ~ 6 ~ , ~7 39

Science and Technology of Thin Film Superconductors, Proceedings of the Conference, 1989, p53 1-9THE HIGH TC TL-BA-CA-CU-0 SUPERCONDUCTING SYSTEMA.M.Hermannand Z.Z.ShengDepartment of PhysicsUniversity of ArkansasFayettevil le, AR 72701INTRODUCTIONDiscoveries of 30-K La-Ba-Cu-0 superconductor [l] and 90-K Y-Ba-Cu-0superconductor [2] have stimulated a worldwide race for new and evenhigher temperature superconductors. Breakthroughs were made by thediscoveries of the 90-K TI -Ba-Cu-0 system [3,4], 110-K Bi-Sr-Ca-Cu-0system [5,6] and 120-K TI -Ba-Ca-Cu-0 system [7-91. Recently, hightemperature superconductivity was also observed in the TI-Sr-Ca-Cu-0system 110-121, and in the M-TI-Sr-Ca-Cu-0 with M = Pb 113,141 and rareearths [IS]. In this paper, we present preparation procedures, structure,and some properties of the 120-K TI -Ba-Ca-Cu-0 superconductors. Wediscuss an unusual levitation phenomenon of the TI-Ba-Ca-Cu-0superconductor due to flux pinning [16]. Finally, we present a new T1203-vapor-process [17] which allows the highest temperature TI -Ba-Ca-Cu-0superconductors to be easily made in the forms of complex bulk components,wires and fibers, and thick and thin films, and minimizes problems causedby toxicity and volatility of TI starting compounds. Recent results onTI203 vapoer-processi ng of thin film Ba-Ca-Cu-0 precursors are included.PREPARATIONl'l -Ba-Ca-Cu-0 superconductive compounds form easily; there are manyways to make good-qua1 i ty superconducting samples. One of the typicalprocedures in preparing the TI-Ba-Ca-Cu-0 samples which we use is thefollowing. Ba-Ca-Cu-oxides are first prepared using the method similar tothat we previously developed for preparation of Ba-Cu-oxides [18,19].Appropriate amounts of BaC03, CaO (or CaCO ), and CuO are mixed, ground,and heated at 925-950 OC in air for 24-48 iour with several intermediategrindings. The resulting uniform black material is ground and served asmaster material. Appropriate amounts of T1203 and Ba-Ca-Cu oxide(depending on the desired stoichlometry) are completely mixed and ground,and pressed into a pellet with a diameter of 7 mm and a thickness of 1-2mm. The pellet is then put into a tube furnace which had been heated to880-910 OC, and is heated for 2-5 minutes in flowing oxygen, followed byfurnace cooling to below 200 OC.STRUCTUREThe T1 -Ba-Ca-Cu-0 system can form a number of superconducting phases.

APRIL 1988 -LEllERS TO NATUREh i O Q O O e O O Q O aReceived 26 February; accepted 18 March 19881. Dorrian, J. F., Newnham, R. E., Smith, D. K & Kay, M. 1. Ferroelectrics 3, 17-27 (1971).2. Michel, C. el of. Phys. 868, 421-423 (1987).3. Maeda, H., Tanaka, Y., Fukutomi, M. & Asano, T. lap. J. appl. Phys. Lell. 27 (in lhe press).4. Tarascon. J. M. er 01. preprint.5. Zandbergen H. W., Groen W. A,, van Tendeloo G., van Landuyt J. & Amelinckx S. SolidSt. Commun. 66, 397-403 (1988).6. Huang, Y. K, er 01. I. Crysl. Growl11 (submitted).7. Kadowaki, K., Huang Y. K., van Sprang M. & Menovsky A. A. Physica 1458. 1-4 (1987).8. Teske C. L. & Mueller-Buschbaurn H. Z. Anorg. Algem. Chem. 379,234-241 (1910).9. Zandbergen. H. W.. Gronsky R., Wang K & Thomas G. Norure 331, 596-598 0988).Bulk superconductivity inT12CaBa2Cu,0,+, up to 120 KL. Gao*, Z. J. Huang*, R. L. Meng*, P. H. Hor*,J. Bechtold*, Y. Y. Sun*, C. W. Chu*, Z. Z. Shengt& A. M. Hermannt*Department of Physics, Texas Center for Superconductivity andSpace Vacuum Epitaxy Center, University of Houston, Houston,Texas 77004, USA+Department of Physics, University of Arkansas, Fayetteville,Arkansas 72701, USA.- -a:ions of the structures of a,a,Cu,O,,+,; and c, Bi. ,Sr,Caternobserved in Fig. 4. T he unitcells are indicated by a full drawn line.bin problem with the 110-K superconductor seems to$culty of synthesis. This is in accordance with the longructures of the two structure types, reported here. Inng periodic structures are more difficult to synthesize.the main problem in the synthesis is believed to bethe temperature region in which these structures areId the synthesis route used so far. Starting with Bi203$has to be reacted at -800 "C to prevent Bi203 fromhis temperature Bi2Sr2CaCu208+, is readily formed.s that, once the material is heated at 88O0C,!Os+, has to react with the remaining Ca compounds~pounds. Such a process will take a very long timekr of days) because of the large particle sizes andstable compoI00 K in the system TI-Ca-Ba-Cu-0 (TCBCO).Two superconducting phases were immediately identified2 asT12CaBa2Cu208+6 (2122) and T~,C~,B~,CU~O,~+~ (2223). Herewe confirm that zero resistance is achievable above 100 K in2122-TCBCO synthesized by afast solid-state reaction, and reporta small positive pressure dependence of the transition temperature,as in ~ i~CaSr~Cu~0~+~~.Samples were prepared at Houston by reacting appropriateamounts of T120,, CaO, BaCO, and CuO in two steps. Extremecare was taken, because of the toxicity of T1 and T130,. ABa2Cu305 powder was made by heating a mixture of BaC0,and CuO powders in air at 920 "C for 2 h. The resulting powderwas mixed with an appropriate amount of T1203 and pressedinto pellets of diameter 5 mm and thickness 1.5 mm. The pelletswere put into a furnace which had been heated to 880-950°C,and were heated for 2-5min in flowing oxygen. Finally, thepellets were quenched in air to room temperature by takingthem from the furnace or cooled to room temperature in thefurnace by turning off the power with the furnace door slightlyopen or closed. X-ray powder diffraction was done with a RigakuDIMAX IIIB automated diffractometer with a Cu source. Barshapedsamples measuring 4 x 1.5 x 1.5 mm were cut from thepellets for electrical and magnetic measurements.Seven TCBCO samples with nominal compositionTl2Ca2Ba2Cu3Ol0+, were synthesized under different conditionsFig. 1 X-ray powder diffraction pattern of TCBCO sample 3. Thepeaks arising from the 2122 phase are indexed in a subcell of5.459~ 5.459 x 29.53 ti.

PHYSICAL REVIEW B VOLUME 38, NUMBER 10 1 OCTOBER 1988Thermoelectric power of the TI-Ca-Ba-Cu-0 superconductorN. Mitra and J. TrefnyDepartment of Physics, Colorado School of Mines. Golden, Colorado 80401B. Yarar and G. PineDepartment of Materials Science and Metallurgical Engineering.Colorado School of Mines. Golden, Colorado 80401Z. Z. Sheng and A. M. HermannDepartment of Physics, University of Arkansas. Fayetteville. Arkansas 72701(Received 11 March 1988)We report measurements of the thermoelectric power of the high-temperature TI-Ca-Ba-Cu-0superconductor. The data indicate a transition which is centered at 118 k 3 K, in agreement withmagnetic measurements and published resistivity results. The thermoelectric power is positiveand, aside from the high transition temperature, is remarkably similar in magnitude and temperaturedependence to that of the rare-earth-based oxide superconductors which have been reportedpreviously.Since the discovery of high-temperature oxide superconductorsin 1986 by Bednorz and Miiller,' at least threefamilies of such materials have been identified. The firstgroup, containing La-Ba-Cu-0 and its derivatives, exhibitssuperconductivity in the range of 30-40 K. The secondgroup, discovered by Chu and co-workers,' is described byR-Ba-Cu-0, where R represents one of the rare-earth elementsor yttrium. These materials exhibit transitions inthe 90-K range. Recently, a new family has beendiscovered with even higher transition temperatures in therange of 115- 120 K. Remarkably, the new materials donot contain yttrium or any of the rare-earth elements.One type, which has been studied by several workers,j-'contains the element bismuth. A second type, first an-nounced by Sheng and ~ermann,~ and later confirmed byother^,^ is distinguished by the presence of thallium. Thepurpose of this note is to describe recent measurements ofthe thermoelectric vower of this new material.The samples were prepared at the University of Arkansasfrom BazCu305 and oxides of thallium and calcium.The powders were mixed, ground, pressed into pellets, andthen heated to about 900°C in flowing oxygen for 3 min.The sample was furnace cooled by turning off the furnacefor 1 h and then removed. These procedures have beendescribed in more detail in previous p~blications.~'~The apparatus and procedures for the thermoelectricpower measurements have also been described in an earlierpaper.9 A differential method was employed in whicha voltage difference is measured across the sample inresponse to an imposed temperature difference. Two setsof reference leads (Constantan and Chromel) were usedand the voltages, after correcting for the thermoelectricpower of the leads, were averaged to obtain the final results.The measurements were made in vacuum as thesample slowly warmed from liquid-nitrogen temperatureto ambient.Figure l(a) shows the thermoelectric power data for asample with nominal composition T12,2Ca2BazCu3010.3+..Energy-dispersive x-ray measurements on this sample indicateda cornposition of T~Z.ZOC~~.~~B~~.I~CU~,~ZFigurel(b) shows thermoelectric power data for a Y-Ba-Cu-0 sample for comparison. In either case the normalstatethermoelectric power is positive, indicating dominanthole conduction, and drops rapidly to zero at the superconductivetransition. The apparent transition for bothsamples is somewhat broadened by the temperaturedifference of several degrees which was imposed duringthe measurements. The midpoint of the transition wasdetermined to lie at 118 f 3 K. This is in good agreementwith data taken at I MHz by a magnetic induction technique''which showed an onset of screening at about 115K. The midpoint of the thermoelectric transition for theY-Ba-Cu-0 sample is 93 f 2 K.Resistance was measured by the standard four-probetechnique using silver-paste contacts. Figure 2 shows thetemperature dependence of the resistance for a piece ofthe sample of Fig. 1 (a). From this curve it is seen that theonset of the resistive transition is about 120 K, the midpointis 110 K, and the sample reaches zero resistance(less than n cm) at 104 K.At least three separate ranges of temperaturedependentbehavior are apparent in the thermoelectricpower data of Fig. I. Below the transition, this quantity isrequired to vanish on thermodynamic grounds. From thetransition temperature to about 175 K in both samples,the thermoelectric power is an increasing function of temperature.Finally, from 175 K to room temperature, weobserve a linear decrease with increasing temperature.The pronounced peak near 175 K in the data for bothsamples is believed to be intrinsic. Careful calibration ofthe thermoelectric power of the reference leads, as well asmeasurements on nonsuperwnductive materials in thistemperature range, have revealed no other explanation forthis feature.Several reports on the thermoelectric powers of oxidesuperconductors have appeared in the literat~re.''-'~ In- 38 7064@ 1988 The American Physical Society

Levitation effects involving high Tc thallium based superconductorsWiiliam G. Harter, A. M. Herrnann, and Z. 2. ShengDepurtment of Physics, Uniurrsity of Arkansas, Fayettcui!lr, Arkunsus 72701(Received 3 1 May 1988; accepted for pubiication 29 July 1988)The thallium based superconductor TI,Ca,Ba,Cu,O,, , has been shown to exhibit very stableand unusual levitation equilibria in various arrangements involving this material andpc-manent magnets. Attractive and repulsive forces are evident in experiments in whichsamples are levitated above and below magnets. Photographs of these experiments andapproximate quantitative discussions of the results are given.The recent discovery of thallium based superconductors'" with critical temperatures greater than 100 K hasstimulated renewed interest in applications of materials withunilsual magnetic properties. Of particular interest is theability of these materials to be suspended or levitated in thepresence of permanent magnets or to suspend magnetsthemselves. Levitatiorl of rectangular magnets over rareearthbased superconductors has been discussed by Hellmanef ol.%nd by Williams and Matey." Hellman et ol.' analyzedand measured the point of equilibrium in the vertica! directionfor a magnet over a flat rare-earth based superconductor,and they gave a qualitative discussion of the curioushorizontal stability. Williams and Matcyf'also noted the curioushorizontal stability and measured the frequency of torsionaland translational motions of a magnet around its horizontalequilibrium above a flat superconductor.For the thallium based superconductors, even morestriking examples of stable levitation occur when either themagnet or the superconductor are suspended. There appearslo he a conticuous range of stable levitation positions for amagnet above a thallium based superconducting disk, andthese disks may themselves be levitated above, to the side of,or heiow a ring magnet as shown in photographs given in thefollowing. Previous examples of inverted levitation wereachieved by Peters el al.' using a rare-earth based material,but in that case it was necessary to dope the material withsilver oxide.A spinning cylindrical Co-Sm permanent magnet isbe tilted to any angle within about 40" of the horizontal ar.d itwill usually stay there.The stabi!ity of each position and orientation can beroughly tested by simply picking up the sample while it levitatesthe magnet and tilting or shaking the sample. Surprisingly,for some positions, the magnet remains near its equi-!ibrium point even when the sample is tilted by more than45". The critical angle of repose depends on the orientationand position of the magnet, and provides an easy way to findthe horizontal components of force needed to change theequilibrium position given the mass (m) or weight (mg) ofthe magnet.- . (1)fcrlricvl - ,g sin @2"'i""lThe critical angle and fcrce ore greatest when the magnet iscentered parallel to the sample end is tilted dong its axis. It isalso greatest when the point of equilibrium is as close aspossible to the surface of the magnet.The levitation of two disks of the thallium based superconductorby a ring magnet is shown in Figs. 2 and 3. Theorientation in each figure can be verified by the falling of thecooled vapor which is visible in the lower part of the photographs.In each case the magnet is gripped in a pair of handheldinsulating pliers and is viewed edge-on and the samplesappear to be suspended in magnetic packets. The Co-Sm ringmagnet has an outer diameter of 1.9 cm, an inner diameter cf1.0 cm, and it is 0.6 cm thick. The suspended thallium based0.2 cm above the sample, exterlds up to about 0.4 cm, andinclude ali points withill about 0. IS cm of the edge of thesample. In other words, the magnet can be placed so that itrotates stably with one end very near the edge of the sampleor in betwen. In 2*ldition, the orientation of $he 110.1. Rotaliq nyn* suspend& above a thaI1i1~m based smagnet may beset somewhat arbitrarily. At each point it can behaves like a frictionless braring.~ p ~ r c ~ ~ ~ ~11 19 Appl. Phys. Lett. 53 (12). 19 September 1988 0003-6951/88/381119-03$01.00 (A 1988 American Institute of Physics 11 19

Structural and elemental analysis of melt-processible high-temperaturesuperconductors by surface science and x-ray dlff raction measurementsA. M. Hermann, Z. Z. Sheng, A. El Ali, and G. D. MooneyDepartment of Physics J. William Fulbright College of Arts and Sciences, University of Arkansas,FayetteuiIle, Arkansas 72701A. J. Nelson, J. Goral, and L. L. KazmerskiSolar Energy Researrh Institute, &/den, Colorado 80401(Received 3 hne 1988; accepted for publication 24 July 1988)We report the results of scanning electron microscopy (SEM), microprobe analysis(wavelength dispersive and energy dispersive spectsoscopies), secondary-ion massspectroscopy (SIMS), and x-ray diffraction measurements on a series of melt-processed highcritical-temperature superconductors. The superconductors were synthesized through a meltreaction between TbBa&u,O, -. and nonstoichiometric rare-earth Ba-Cu-0 oxides. The meltprocessedsamples are dense, largely void-free superconductors with critical temperatures inthe 90-95 K range. Microprobe and SIMS data show the absence of Tb in surfaces of theinteriors of samples which had been fractured to expose the interior regions. Comparativemicroprobe and SEM data are included for conventionally prepared sintered samples.Comparative x-ray diffraction studies are presented which show that the melt-processedsamples studied have diffraction patterns nearly identical to those of sintered (rareearth)B+Cu,O, -. samples. In two of the three Y-based melt-processed superconductorsstudied, the compound stoichiometry from rnicroprobe analysis is Y ,Ba,Cu,O, .-. Theevidence presented indicates that the TbBa,Cu,O, -,decomposes during the melt reaction.I. INTRODUCTIONThe recent discovery of superconductors whose criticaltemperature exceeds 90 K1 has caused a scientific rush tofind processible materials capable of carrying large currentdensities and of being manufactured into useful superconductingcomponents for magnets, power transmission cables,levitated vehicles, and a plethora of other applicationsonly now being conceived.We have discovered'.' a melt process that producesdense rare-earth-based high-temperature superconductorsthat are vold free. This paper reports the characterization ofthese superconductors by scanning electron microscopy(SEM), microprobe analysis, secondary-ion mass spectroscopy(SIMS), and x-ray diffraction (XRD). Resistivitytemperaturevariations showing the superconducting criticaltemperatures are also included.H. EXPERIMENTThe buik superconductors were formed in a melt processin which molten TbBa,Cil,O,- , was reacted at 950 'Cwith sintered (rare earth) ,,,Ba,,, CuO, _, where the rarezarths studied to date include Y, Ho, Nd, Sm, Gd, Er, Tm,Yb, and Lu. Since Y is typical of the rare earths studied, weemphasize here the measurements on Y-based compoundsas sintered starting materials. To form the TbBa,Cu,O,- .,mixtures of Tb407, BaCO,, and CuO powders were groundin an agate mortar and heated in air at 890-910 "C for 6-8 hin a tube furnace to prepare a powder of nominal compositionTbBa2Cu30,.,. To form the Y-based starting compounds,mixtures of Y203, BaCO, and CuO were treatedsimilarly to form a powder of nomina: compositionY,,2B~,8C~03.6. The heated mixtures were then regroundand, in certain cases, pressed into pellets of diameter 4 or 4 in.In one set of experiments, the powdered Tb and Y compoundwere pressed separately into pellets and heated (Tb pellet ontop of Y pellet) to 920-960 "C in flowing O2 for 12-24 h in atube furnace. The sample was then cooled in the furnace to200 "C in 1-2 h before being brought to room temperature.The Y pellet is initially green, the Tb pellet yellowbrown.After the melt reaction and cooling, the interfacialmelt-recrystallized layer (several millimeters thick) isblack. According to a weight balance experiment, if the Tbcompound had melted uniformly, the melt-recrystallized su-perconductor would have a composition ofTh.,, Y, Ba,,, Cu30, (we discuss decomposition of theTb compound in a later section). We have found that elementalCu can a h replace the CuO in the starting compound,and that a powder of the Y (or another rare earth)compound can be substituted for a pellet in the process. Theresultant melt-resolidified portions were cut away from anyunmelted portions of the Y pellet, as necessary.The SEM measurements were made on a PHI model600 scanning auger microprobe having a base pressure ofI X 10-lo Torr. SIMS measurements were made with a Perkin-Elmer/PhysicalElectronics SIMS I1 spectrometer. Anelectrostatic extraction lens was used at the front end of thequadrupole to increase sensitivity.Microprobe analysis by wavelength dispersive spectroscopywas performed using a Cameca MBX spectrometer atan incident energy of either 10 or I5 keV. (The latter issufficient to stimulate Tb emission, but too energetic forhighly accurate oxygen analysis. )The analysis point was disk shaped with adiameter of 20pm. Five or six randomly chosen (unless specified otherwise)points were sampled on each surface studied. Full5056 J. Appl. Phys. 61 (10). 15 November 1988 0021-6979/88/225056-05$0e.40 @ 1968 Americen Institute of Physics 5056DOWnlaaded 21 Mar 2008.10 .. 1.3D~~XW~aa66~~6Red~i~t~ill:Uti~~aa~~bJ~~tt..t9.A~P...Ii~(?n.s.e.~.r:,.~~~~y~Ight;see http:iljap.aip.orgljapl opyright jsp...5. ..

with the data; its intensity becomes too large. This curvean predicted by the Solomon and Garcia model, and af 6 less than the model results reported. Curve A, whichlargest mixing ratios of the three above 25 km, is lesse event of 23 and 24 September, and find that theshape is consistent with N20 emission from a thinh a peak mixing ratio of -40 p.p.b.v. centred at anse seen in the other observations in this series. Ourobservations" during this same event showus increase in ozone took place in the 30-40 kmin the latitudinal gradient in N20, which would produce anincrease in the Antarctic N20 concentration.In our papers discussing our measurements of C10 in~ntarctica'~.'~, we present data showing chlorine to be stronglyinvolved in the ozone depletion mechanism at low altitudes.The very low mixing ratios of N20 that we report here furtherdemonstrate the unusual and unexpected nature of the Antarcticspringtime stratosphere.We thank the NSF's Division of Polar Programs for its supportthrough all phases of our Antarctic work, and the manyindividuals in Antarctic Services, Inc., the US Navy SupportForce and the NSF whose interest and enthusiasm smoothedour way. We thank Dr R. L. Jones for supplying us with informationon the trajectory of air over McMurdo during 23-24 September1986, and NASA's Upper Atmospheric Research Programand the Chemical Manufacturer's Association for the supportwhich made it possible to carry out this research.Received 19 May; accepted 17 December 19871. World Meteorological OrganizationJNASA, 7'he Stratosphere, 1981; 7'heoly and Measurement(Report No. 16 of the WMO Global Ozone Research and Monitoring Project,NASA Headquarters, Washington, 1986).2. Farman, J. C., Gardiner, B. G. & Shanklin, J. D. Nature 315, 207-210 (1985).3. Goldan, P. D., Kuster, W. C.. Albritton, D. L., & Schmeltekopf, A. L. I. geophys. Res. 85,413-423 (1980) 86, 5385-5386 (correction) (1981).4. Vedder, J. F., Inn, E. C. Y., Tyson, B. J., Boitnott, C. A. & O'Hara, D. J. geophys. Res. 86,7363-7368 (1981).5. Jones, R. L. & F'yle, J. A. I. geophys. Res. 89, 5263-5279 (1984).6. Mahlman, J. D., Levy, H. I1 & Moxim, W. J. J. geophys. Res. 91, 2681-2686 (1986).7. KO, M. K. W., Tung, K. K., We~nstein, D. K. & Sze, N. D. J. geophys. Res. 90, 2313-2329(1985).8. Solomon, S. & Garcia, R. R. I. geophys. Res. 89, 11633-11644 (1984).9. Schmeltekopf, A. L. el of. I. Armos. Sci. 34, 729-736 (1977).10. Goldman, A., Fernald, F. G., Murcray, F J., Murcray, F. H. & Murcray, D. G. I. quanr.Specl. Rod Trans. 29, 189-204 (1983).11. Parrish, A., deZafra, R. L., Solomon, P. M. & Barrett, J. W. Radio SCI. (in the press).12. Hofmann, D. J., Harder, J. W., Rolf, S. R. & Rosen, J. M. Nature 326, 59 (1987)13. deZafra, R. L. er of. Norure 328.408-411 (1987).14. Solomon, P. M. er of. Norure 328, 411-414 (1987).15. Connor, B. J. er of. Geophys Res Lerr. 14, 1254 (1987).16. Tung, K. K., KO, M. K. W., Rodriguez, J. M. & Sze, N. D. Nature 322, 811 (1986)17. Tung, K. K. Geophys. Res. Lerr. 13, 1308-I311 (1986).18. Connor, B. el of. J. geophys. Res. 92, 13221-13230 (1987).19. Mount, G., Sanders, R., Schmeltekopf, A. & Solomon, S. I. geophys. Res. 92, 8320-8327(1987).20. Kmeger. A. J.. Schoeberl, M. R. & Stolarski. R. S. Geophys. Res. Lerr. 14, 527-530 (1987).21 Mahlman, J. D. & Fels. S. B. Geophys. Rcs. Lerr. 13, 1316-1319 (1986).Dobson units (DU) or less; thus thethe N20 depletion region appears to belarger than, the ozone hole.either induced by transrcticwinter. Indeed, astolysis and by reactionnd upper stratosphere, one wouldring the winter months of darkness.aspects of the Antarctic vortex aredynamics peculiar toSuperconductivity in therare-earth-free T1-Ba-Cu-0 systemabove liquid-nitrogen temperatureZ. Z. Sheng & A. M. HermannDepartment of Physics, University of Arkansas, Fayetteville,Arkansas 72701, USAThe initial discovery by Bednorz and ~uller' of 35-K superconductivityin the La-Ba-Cu-0 system has stimulated worldwide activityin searching for higher-temperature superconductors. Elementalsubstitution has proved to be most effective in raising transitiontemperature. Substitution of Sr for Ba has produced 40-K superconductivity2-5,and substitution of Y for La has produced a newhigh-temperature superconductor with transition temperatureabove liquid-nitrogen temperature6. A class of superconductingcompounds of the form RBa,Cu30,-, has been explored by furthersubstitutions of other rare earths (Y is considered in the rare-earth[R] category here) for y7-13. TO date, a rare earth, an alkalineearth, copper and oxygen have been required for all hightemperatures~~erconductors'~~'~. (Zhang et al.I4 reported 90-Ksuperconductivity in the Th-Ba-Pb(Zr)-Cu-0 system. Pan et al."reported 50-K superconductivity in the Y-Ba-Ag-0 system. AsTh is a member of the actinide series which belongs to the sameGroup 3B in the periodic table as the lanthanide series and Ag

Bulk superconductivity at 120 K in theT1-CaIBa-Cu-0 systemZ. Z. Sheng & A. M. HermannDepartment of Physics, University of Arkansas, Fayetteville,Arkansas 72701, USAThe discovery of 30-K superconductivity in the La-Ba-Cu-0svsteml and 90-K su~erconductivitv in the Y-Ba-Cu-0 svstemZsiimulated a worldwide search for even higher-temperatureHuperconductors.Unfortunately, most of the higher-temperature transitionsreported in the past year have proved 6 be unstable,irreproducible, or not due to bulk ~u~erconductivit~~~. Recently,we and co-~orkers'.~ reported superconductivity above 90 K in anew TI-Ba-Cu-0 system, and pointed out that elemental snbstitutionsin this system may lead to even higher-temperature superconductivity.Here we report stable and reproducible bulk superconductivitywith an onset at 120 K and zero resistance above 100 Kin the TI-Ca/Ba-Cu-0 system. This transition temperature ismuch higher than those observed for typical rare-earth-containingsuperconductors, and the onset temperatures are comparable tothat in the Bi-CaISr-Cu-0 system, as reported in refs 10 and 11(received after submission of this paper).A typical procedure for preparing I-Ca/Ba-Cu-0 samplesis as follows. Appropriate amounts of m2o3, CaO and BaCu,04for a certain nominal composition (for example,Tl,CaBaCu308+,), were completely mixed, ground and pressedinto a pellet with a diameter of 7 mm and a thickness of 2-3 mm.The pellet was then put into a tube furnace, which had beenheated to 880-910 "C, and was heated for 3-5 min in flowingoxygen. The sample was then taken from the furnace andquenched to room temperature in air. Some heated sampleswere furnace-cooled to room temperature, and some quenchedsamples were subsequently annealed at 450 "C in flowing oxygenfor several hours. We use BaCu304 as a starting material becauseit has the lowest melting point among the Ba-Cu oxides, andmolten BaCu304 is extremely fluid1'; we believe that melt-solidreactions take place more completely than do typical solid-statereactions. The BaCu,04 used in this experiment was preparedaccording to the procedure described in refs 8, 9 and 11.Resistance was measured by the standard four-probe technique,with silver paste contacts. Rectangular bars of typicalFig. 2A TI-Ca/Ba-Cu-0 sample levitating over a magimmersion in liquid nitrogen.size 7x3 x 1 mm were cut from the pellets formeasurements. The resistance of the Tl-Ca/Ba-Cuusuallystarts to drop sharply at -120 K, aagreement with the onset as determined by the intersthe slopes in the transition region and in the normal reabove the transition. Figure 1 shows resistance plottetemperature for a furnace-cooled f12Cal,5BaCu308.5(open circles) and a Tl, 8,CaBaCu307,8+, sample an450°C (solid line). For comparison, Fig. 1 also sbehaviour of a EuBa2Cu,07-, sample (dashed line) wcarefully prepared using a typical solid-state sinter tobserved for high-quality rare-earth-containingfor example, refs 13 and 14), and the zero-resistancis higher than those attained so far in the Bisystem'0"'.The superconducting transition of the Tl-Cu-0 samples is rather broad (usually >20 K), suggestizero resistance could be reached at higher temperatimprovement of the preparation procedure.Qualitative magnetic examinations of T1-Ca/Basamples showed strong diamagnetic repulsion. Figure 2a sample with a nominal composition of T12Ca2BaClevitating over a magnetic field of -5,000 G after immersliquid nitrogen. Figure 3 shows the temperature depende0 . . . . . . . . . . . . . . . . . . . .4 &O n- -75 0 100 150 200 250 300Temperature (K)Fig. I Krblbtancr verbus temperature for T1ZCa,,5BaC~308.5+x(circles) and TI,,,,CaBaCu,0,,8+, (solid line), compared withEuBa,Cu,O,-, (dashed line).- -1.: -2t. m-3m -.2 -45- 6Fig. 3 Temperature dependence of the a.c. susceptibility ofT12Ca, B~CU~OR~,, sample.T( K

T12Q3 vapor process of making TI-Ba-Ca-Cu-Q superconductorsZ. Z. Sheng, L. Sheng, H. M. Su, and A. M. HermannDcpnrlrnenr cf Physics, Uniucnify of Arkonsos. Fiyerreuille, Arkansas 72701(Received 30 August 1988; accepted for publication 2 Ncvember 1988)A new process for making TI-Ra-Ca-Cu-0 superconductors which is based on reactionsbetwcen solid precursor Ba-Ca-Cu-oxides and vapor T1,0, has becn developed. The T1,03vapor process allows TI-Bz-Ca-Cu-0 superconductors to be easily made in the forms ofcomplex bulk components, wires and fibcrs, and thick and thin films, and minimizes problemscaused by toxicity and volatility of TI starting conlpounds.Discovery of thc TI-Ba-Ca-Cu-0 superconducting system'has provided increased i", with zero-resistance temperatureup to 125 K."' Structural and compositional analyseshave shown that the T1-Ra-Ca-Cu-0 system can form anumber of superconducting compounds, includingTi,,,BatCa,, . , Cu,,O1 . , ., , , , with m = 2 ( R C 4-9) ~ and1 (Refs. 10and 11) and n = 14 (Refs. 4-11). The 1; ofthese compounds increases with n as wcll as m. Thesc trendshave stimulated a strong search for higher T, TI-based superconductorsby increasing n and/or m. For many applicntions,the T, of a superconductor usually must be 1/2-1/4higher than operating tempcrature. ' I With theadvantages ofhigh T,. and easy formation, the TI-Ba-Ca-Cu-0 superconductorsmay be the first liquid-nitrogen temperature superconductorsfor practical commerciai applications. However,the high toxicity and volatility of the TI starting compoundshave introduced safcty concerns in thc fabrication of TIbasedsuperconductors. In extensive preparation experimentson TI-based superconductors, we have found thatTlzO, evaporates above its 717 'C melting point, and the vaporrcacts with solid Ba-Ca-Cu-oxides, brming high qualityT1-Ba-Ca-Cu-0 superconductors. This vapor-solid reactionhas simplified the making of TI-Ba-Ca-Cu-0 supcrconductorsto the making of Ba-Ca-Cu-oxides, and this minimizesthe toxicity problem of TI starting u)mpounds. In particular,this Tilo, vapor proces allows the TI-Ba-Ca-Cu-0 superconductorsto be easily made in the forms of complex bulkcomponents, wires arid fibers, and thick and thin films byfabrication of the precursor Ba-Ca-Cu-oxides and subsequentintroduction of TI. This technique may also be of importancein understanding of the formation of TI-Ba-Ca-Cu-0 superconductors, and the search for new, even highertemperature, T1-based superconductors.Ba-Ca-Cu-oxides were first prepared using the methodsimilar to that we previously devcloped for preparation ofBa-Cu-oxides. ".I4 Appropriate amounts of BaCO,, CaO (orCaCO,), and CuO wcre mixed, ground, and hcated el 925-950 "C in air for 24-48 h with sevcral intermcdiatc grindings.The resulting uniform black material was ground and scrvedas master material. An appropriate amount of the Ra-Ca-Cu-oxide powder wa pressed into a pellet, and thc pellet washeated at 925-950 "C in flowing oxygen for 5-10 min andthen was air coolcd. A small platinum boat was put in aquartz boat, and a small amount of TI, 0, (typically 0.1-0.2g) was put into the platinum boat. The heated Ba-Ca-Cu-0pellet was put above the platinum boat. The quartz boat withthe contenls was put into a tflbc furnace, which had beenhcated to 9W925 "C, and was heated for about 3 min inflowing oxygen followed by furnace cooling.After the above heating treatment, the T1,0, in theplatinum boat had completely evaporated, and the T1,0,vapor processed Ba-Ca-Cu-0 pellet fo-med a layer of TI-Ba-Ca-Cu-0 superconducting cornpound(s) on its bottom sJrface.Figure 1 shows temperature dependences of resistancefor some TI,O, vapor processed Ba-Ca-Cu-0 precursors:(a) BaCa,Cu,O,, (b) Ba2CaCu20,, and (c) Ba2Ca2Cu307.Thcir zero-resistance temperatures are 110, 96, and 105 K,respectively, which are comparable to those of correspondingsintered samples.In principle, this technique has simplified the preparationof TI-Ba-Ca-Cu-0 to thc preparation of Ba-Ca-Cu-0.Therefore, this technique allows TI-Ba-Ca-Cu-0 superconductingcomponents, such as bulk components, wires andfibers, thick and thin films, to be easily made. In particular,T1,0, vapor processcd molten Ba-Ca-Cu-oxides are also superconducting,and thus this technique al!ows superconductingcomponents to bc easily made in arbitrary shape.Figure 2 shows resistance-temperature dependences for (a)a TI10, vapor processed Ba2Ca,Cu,07 thick wire and for(5) a TlzO, vapor processed Ba2Ca,Cu,0, thick film. Thethick film was prepared by first melting Ba,Ca,Cu,07 on apiatinum substrate (heating at 980 "C in flowing oxygen for5 min) and then subjecting to T&O, vapor processing. Itreached zero resistance et 11 I K. This 2-mm-wide film wasTEMPERATURE (K)FIG. 1. Tcmpemture dcpcndenccs of rcsistancc for T1,0, vapor prwcssalD~-a-CU-o precursors: (a) B:rCa,Cu,O.,, (I,) Ba:C*Cu,O,, and :c)E~~,cI~~cu,o,.2686 Appl. Phys. Lett. 53 (26), 26 December 1988 0003-6951 /88/522686-03501.00 (9 1988 American lnstittite of Physics 2686

PHYSICAL REVIEW B VOLUME 38, NUMBER 10 1 OCTOBER 1988Superconductivity in the 1lSr-Ca-Cu-0 systemZ. Z. Sheng and A. M. HermannDeportment of Physics, University of Arkonsos, Foyetteville, Arkonsm 72701D. C. Vier and S. SchultzDeportment of Physics. University of Coli/ornio. Son Diego, La Jollo, Colifornio 92093S. B. OseroffDeportment of Physics, Son Diego Stote University, Son Diego. Colifornio 92182D. J. George and R. M. HazenGeophysical Loborotory, Corncgie Institution of Woshington. 2801 Upton Street, N. W., Woshington. D.C. 200011-2898(Received 9 June 1988)Superconductivity at 20 K has been unambiguously observed in the TI-Sr-Ca-Cu-0 system.Superconductivity at 70 K was also observed in the same system. These observations may providea new insight into the mechanism of oxide superconductivity.Discoveries of TI-based superconducting systems'-5and Bi-based superconducting systems6-' have not onlyset new Tc rccords with zero resistance up to 125 K, butalso have provided a new insight into the mechanism ofhigh-Tc oxide superconductivity. Compositional analysesand structure determinations of the new generation ofrarecarth-free supercond~ctors~-'~ have showed that (a)there are no Cu-0 chains, but Cu-0 sheets are apparentlyresponsible for superconductivity in these new systems,and thus Cu-0 chains which were thought important inrarecarth-based 1 :2:3 compounds are not required for cupratesuperconductivity; (b) T, increases with the numberof Cu-0 layers per unit cell, the T12Ba2Ca2Cu30~o+x havingthe highest Tc (zero resistance at 125 K) has threeCu-0 layers; and (c) buckling of Cu-0 sheets of superconductingcompounds may lead to decrease of T, (Refs.12 and 14).The existing experimental data also indicate thatgroup-IIA elements are required for oxide superconductivity,playing perhaps a subtle role. Until now, Ba hasbeen required for the TI-based cuprate superconductors,whereas Sr is required for the Bi-based cuprate superconductors.At present, theorists are not able to predictwhether superconductive compounds can be formed withsubstitution of Sr for Ba in the TI-based cuprate superconductingsystems, or substitution of Ba for Sr in the Bibasedcuprate superconducting systems. In extensive elementalsubstitutions for the TI-based superconducting systems,we have observed 20-K superconductivity and possible70-K superconductivity in the TI-Sr-Ca-Cu-0 system.I6The samples are stable and the results are reproducible.The discovery of the TI-Sr-Ca-Cu-0 superconductingsystem bridges the TI-Ba-Ca-Cu-0 and Bi-Sr-Ca-Cu-0 superconductors, and should provide furtherclues to an understanding of the mechanism of high T, oxidesuperconductivity.A typical procedure of preparing TI-Sr-Ca-Cu-0 superconductingsamples is the following. Appropriateamounts of TI203 (99.999% pure), SrO (99.5% pure),CaO (99.95% pure), and CuO (certificate reagent) weremixed and ground completely to obtain a powder with certainnominal compositions (for example, TIzSrzCaz-CU~O~~+~). The powder was pressed into a pellet with adiameter of 7 mm and a thickness of 1-2 mm. The pelletwas then put into a tube furnace which had been heated to900-950°C, and was heated for 3-5 min in flowing oxygen,followed by furnace cooling or quenching. Since asmall amount ( < 1%) of barium entering into the TI-Ca-Cu-0 system could lead to observation of superconductivityabove liquid-nitrogen temperature," great care waspaid in avoiding contamination with barium. Qualitativechemical analyses were performed for samples 1 and 2(both with a nominal composition ofTlzSr2Ca2Cu30~o+,)using a JSM model 35 scanning electron microscope.operated at 20 kV and 0.0 1 PA beam current. The majorelements in both samplea were T1. Sr, Ca, and Cu. SampleI, in addition, contains a trace ( < 2%) of Ag, perhapsfrom silver paste or from capsule contamination. Neithersample contains detectable barium (

APPLIED PHYSICS COHNUNICATIONS, 8(2&3), 127-138 (1988)RESISTANCE FLUCTUATIONS IN THE Y -8A-CU-0-FSYSTEMZ.Z.Sheng, W.Kieh1, D.Marsh, D.Mooney,F.Aramnash, A.E1 All, and A.M.HermsnnDepartment of Physics,University of Arkansas, Fayetteville, AR 72701, USAM.W.Austin and H-LedbetterNatIonal Bureau of Standards, Boulder, CO 80303, USA-AbstractTWO batches of samples with nominal composition of YBa2C~j06.5,~F2~(~4-2) were prepared. All fluorine-containing samples were mu1 tiphase.Resistance-temperature relationships of these samples depended strongly ontheir fluorine content and their preparation conditions.Reslstance-temperature anomalies were ob~erved for sane samples with large values of x.Combining transport measurements with x-ray powder dlffractlon data, vieattribute these anomalies either to existence of htgh temperaturesuperconductive phasets) In the Y-Ba-Cu-0-F systm, or to resistancefluctuations, resul tlng from the presence of multiple phases.After the discovery of 90 K superconduc-tors In Y-Ba-Cu-0system (I),superconductors with even hlgher transition temperatures have been actlvelysought.Trace superconductivity at 240 K was reported In Y-Ba-Cu-0 systemCopyrbht b 1988 by Mamsl Ddrksr, lnc.

VOLUME 60, NUMBER 10 PHYSICAL REVIEW LETTERS 7 MARCH 1988Superconductivity at 90 K in the TI-Ba-Cu-0 System2.2. Sheng, A. M. Hermann, and A. El AliDepartment of Physics, Uniuersily of Arkansas. Fayetteville. Arkansas 72701C. Almasan, J. Estrada, and T. DattaDepartment of Physics. University of South Carolina, Columbia, South Carolina 29208andR. J. MatsonSolar Energy Research Institute, Golden, Colorado 80481(Received 22 January 1988)Stable and reproducible superconductivity has been unambiguously observed in a new Tl-Ba-Cu-0system containing no group-IIIB elements. Resistance in these rare-earth-free oxides starts to dropsharply above 90 K and reaches zero at 81 K. Meissner flux expulsion and diamagnetic shielding showbulk superconductivity. Further substitutions of elements for this system may lead to even highertemperaturesuperconductors.PACS numbers 74.10-t~. 74.70.YaThe high-temperature superconducting La-Ba-Cu-0system discovered by Bednorz and Miiller ' has a specialfeature; that is, the elements in this system can be replacedpartially or completely by other elements, producingeven higher-temperature superconductors. Substitutionof Sr for Ba has produced 40-K superconductivity.'v3Substitution of Y for La has produced differentcrystal structure with an increased transition temperatureof 90 K .~ The compound responsible for the hightemperaturesuperconductivity in the latter case was rapidlyidentified as YBa2Cu307-, with perovskite structure.A large class of 90-K superconducting compoundsof the form RBazCu3O7-, with R-La, Nd, Sm, Eu, Gd,Dy, Ho, Er. Tm Yb, and Lu were prepared by furthersubstitutions of other rare earths (Y is considered in therare-earth category in this Letter) for Y in YBa2Cu3-07 -,. 6*7 Later, 90-K superconductivity in the radioactiveTh-Ba-Pb(Zr)-Cu-0 system was reported.' Recently,an unstable 50-K superconductor, YBa2Ag307-,,was also rep~rted.~ To date, three of the elements belongingrespectively to groups IIIB, IIA, and IB are requiredfor all high-temperature oxide superconductors(the only exception is 40-K superconductivity ofLa2Cu04), 'O." and only partial substitutions of otherelements have led to superconductivity, but with nosignificant increase in transition temperature.In this Letter, we report a new Tl-Ba-Cu-0 superconductingsystem which contains no rare-earth elements.Resistance measurements show that this superconductingsystem has an onset temperature above 90 K with a zeroresistance (< n cm in resistivity) at 81 K. Magneticmeasurements confirm the superconducting transition.The samples are stable in air at ambient temperaturefor at least two months (according to resistancetemperatureremeasurements) and their preparation iseasily reproduced. We believe that the discovery of thisnew superconductive system will provide deeper insightinto mechanism of high-temperature superconductivityand may lead to even higher-temperature superconductorsby variations of composition and preparation procedures,and/or by further elemental substitutions.Samples were prepared from 99.999%-pure T1203,99.999%-pure BaC03, and certificate reagent CuO. Atypical procedure for the preparation of T1-Ba-Cu-0samples, in which some principles of the melt processwhich we developed for preparing high-quality meltprocessibleRBazCu307-, superconductors'2~'3 are used,is as follows: Appropriate amounts of BaC03 and CuOwere mixed and ground with an agate mortar and pestle,and heated in air at 925°C for more than 24 h withseveral intermediate grindings to obtain a uniformblack Ba-Cu-oxide powder (for example, BaCu304 orBazCu3Os). Appropriate amounts of T1203 and Ba-Cuoxidepowder with a certain nominal composition (forexample, T~B~CU~O~.~+, and T11.SBaZC~307.3+x, hereafterdenoted by 1 - 13 and 1-5-23. respectively), werecompletely mixed and ground, and pressed into a pelletwith a diameter of 7 mm and a thickness of 1-2 mm. Aquartz boat containing the pellet was then put into atube furnace. which had been heated to 880-9 10 "C. andwas heated for 2-5 min in flowing oxygen. As soon as ithad slightly melted, the sample was taken out from thefurnace and quenched in air to room temperature. Thesamples are black, porous. and multiphase. It must beemphasized that slight melting of the samples is crucialto a sharp superconducting transition above liquidnitrogentemperature.Electron-probe microanalysis showed considerablecomposition disparity between points in the samplesstudied, which can be a result of either actual differences8 1988 The American Physical Society

- 138LETERST0Bulk superconductivity at 120 K in theTI-Ca/Ba-Cu-0 system2. Z. Sheng & A. M. HermannDepartment of Physics, University of Arkansas, Fayetteville,Arkansas 72701, USAThe discovery of 30-K superconductivity in the La-Ba-Cu-0system1 and 90-K superconductivity in the Y-Ba-Cu-0 system2stimulated a worldwide search for even higher-temperature superconductors.Unfortunately, most of the higher-temperature transitionsreported in the past year have proved to be unstable,irreproducible, or not due to bulk supercond~ctivit~~~. Recently,we and co-worker~~.~ reported superconductivity above 90 K in anew T1-Ba-Cu-0 system, and pointed out that elemental substitutionsin this system may lead to even higher-temperature super- Fig. 2 A TI-Ca/Ba-Cu-0 sample levitating over a magconductivity. Here we report stable and reproducible bulk super-immersion in liquid nitrogen.conductivity with an onset at 120 K and zero resistance above 100 Kin the TI-Ca/Ba-Cu-0 system. This transition temperature ismuch higher than those observed for typical rare-earth-containingsuperconductors, and the onset temperatures are comparable tothat in the Bi-Ca/Sr-Cu-0 system, as reported in refs 10 and 11(received after submission of this paper).- 110 K. We estimate the onseA typical procedure for preparing TI-Ca/Ba-Cu-0 samples at which the curvature in the ris as follows. Appropriate amounts of TI#,, CaO and BaCu30,for a certain nominal composition (for example,T1,CaBaCu308+,), were completely mixed, ground and pressed the slopes in the transition region and in the normal reinto a pellet with a diameter of 7 mm and a thickness of 2-3 mm. above the transition. Figure 1 shows resistance plotteThe pellet was then put into a tube furnace, which had been temperature for a furnace-cooled Tl2Ca,.,BaCu3Oheated to 880-910 "C, and was heated for 3-5 min in flowing (open circles) and a Tl,,,,CaBaCu,O,,,+, sampleoxygen. The sample was then taken from the furnace and 450°C (solid line). For comparison, Fig. 1 alsoquenched to room temperature in air. Some heated samples behaviour of a EuBa,Cu,O,_, sample (dashed line)were furnace-cooled to room temperature, and some quenched carefully prepared using a typical solid-state sintersamples were subsequently annealed at 450 "C in flowing oxygen technique. The transition midpoint and zero-resistanfor several hours. We use BaCu,O, as a starting material becauseit has the lowest melting point among the Ba-Cu oxides, andmolten BaCu304 is extremely fluid12; we believe that melt-solidreactions take place more completely than do typical solid-state observed for high-quality rarereactions.The BaCu304 used in this experiment was prepared for example, refs 13 and 14), andaccording to the procedure described in refs 8, 9 and 11.Resistance was measured by the standard four-probe technique,with silver paste contacts. Rectangular bars of typical8 - 1 1 1 1 1 1 1 1r Izero resistance could be reached at higher temperatimprovement of the preparation procedure.Qualitative magnetic examinations of T7 ---C: 5-E0-4 -0)06 -0+.- "3 2:U)c?2 -1 -00o00-0 0o liquid nitrogen. Figure 3 shows the temperature depenthe a.c. susceptibility of a sample of nominal comTI,Ca, ,B~CU,O,~+,, as determined by the techniq0 o 00 -0 ort ton". The onset temperature as determined by the dev00-0.. .......... . . . . . . . .-0-750 100 150 200 250 300Temperature (K)Fig. 1 Ke>~btance versus temperature for T1,Ca,.,BaCu3O8.,+,(circles) and Tl,,,,CaBaCu,O,,,+, (solid line), compared withEuBa,Cu,O,-, (dashed line).

IVorld Congress on Superconductivity, 1988, p 365-7690 K TL-BA-CU-0 AND 120 K TL-CA-BA-CU-0 8ULK SUPERCONDUCTORSZ. Z. Sheng and A. M. HerrnannDepartment of Physi csUniversity of ArkansasFayetteville, Arkansas 72701, USA---------------------------------------------------------------We have discovered a reproducible and stable bulk superconductorabove 90 K with zero resistance at 81 K in the 1-1-Ba-Cu-Csystem. !3y appropriate elemental substitutions in th's 'system,we have observed reproducible and stable bulk superconductivityin the TI-Ca-Ba-CA-0 system above 120 K, which is much higherthan that observed for any existing rare earth-contatningsuperconductor. The T7-Ca-Ea-Cu-0 superconductor samples areeasily made and readily levitate over a magnetic field. Webe1 ieve that further e! emental substitutions in the thalliumcontainingsystems may produce even higher temperaturesuperconductors.---------------------------------------------------------------1 IntroductionDiscoveries of 30 K sucarconductivity in the La-Ba-Cu-0 system[l] and 90 K superconductiv- c,y in the Y-Ba-Cu-0 system [2] havestimulated a worldwide race for higher temperature superconductors.In spite of many efforts ;r the past year, stable and reproduciblesuperconductivity has rema 7ed at the level of 90 K in the RE(rareearth)-Ba-Cu-0 system. E*- .:kthroughs were recently made by thediscoveries of the 90 K Ti-3a-Cu-0 system [3,4], 110 K Bi-Ca-Sr-Cu-0system [5,6] and 120 K T7 - .a-Ba-Cu-C system [7,8]. The T1 -Ba-Cu-0system is the first non-rz-.-earth system which reaches zeroresistance above liquid n':rogen temperature, whereas the TI-Ca-Ba-Cu-0 system is the first systtm which reaches zero resistance above 100K, and has the highest zerl-resistance temperature (125 K). In thispaper, we present the pi-c 3ration procedures and properties of the newTI-based superconductc- and briefly discuss further exploration foreven higher temperatur. superconductors.2. The 90 K TI-Ba-Cu- SuperconductorSamples were pr~xred using 99.999 % pure T1203, 99.999 % pureBaCO?, and certifica-.e reagent CuO. Since T1203 has a relatively lowme1 ting point of 71- OC and starts to decompose at a temperature aslow as 100 OC, typi :a1 procedures used for preparing REBa2Cu307-Xsuperconductors do not apply to the TI-Ba-Cu-0 system. In ourpreparation of the TI-Ba-Cu-0 superconductor, we used short high' 3erature heating and quenching. High temperature is required for

PHYSICAL REVIEW B VOLUME 37, NUMBER 16 1 JUNE 1988Magnetization of the 120-K TI-Ca-Ba-Cu-0 superconductorA. M. Hermann and 2. 2. ShengDepartment of Physics, University of Arkansas, Fayetteville. Arkansas 72701D. C. Vier and S. SchultzDepartment of Physics, University of CaIi/ornia, Son Diego. La Jolla, Cali/ornia 92093S. B. OseroffDepartment of Physics, Sun Diego State University, Sun Diego, Calqornia 92182(Received 29 February 1988; revised manuscript received 26 April 1988)dc magnetization measurements on the new TI-Ca-Ba-Cu-0 superconductor show superconductivitywith an onset temperature as high as 118 K, 24 K higher than that of a high-quality Eu-BazCu~07-, sample. The resistive onsets of these new superconductors are near 140 K. Themagnetic data show complete diamagnetic flux exclusion and 10-15% Meissner expulsion at 10K.Discoveries of 30-K superconductivity in the La-Ba-Cu-0 system1 and 90-K superconductivity in the Y-Ba-Cu-0 system2 have stimulated a worldwide race forhigher-temperature superconductors. In spite of manyefforts in the past year, stable and reproducible superconductivityhas remained at the level of 90 K in the R-Ba-Cu-0 system (R equals rare earth). Breakthroughs wererecently made in rare-earth-free superconductors by thediscoveries of the 90-K T1-Ba-Cu-0 system,h4 110-K Bi-Ca-Sr-Cu-0 system, 'v6 and 120-K TI-Ca-Ba-Cu-0 system.7-9The TI-Ca-Ba-Cu-0 superconducting sampleshave onset and zero-resistance temperatures much higherthan those of the Bi-Ca-Sr-Cu-0 ~ystem.~-~ Two superconductingphases have been identifi~i.~ One has a compositionof T12Ca2Ba2Cu3010+, (2:2:2:3 phase), andanother of TlzCaBa2Cu~0~+, (2:1:2:2 phase, which issimilar to Bi2CaSr2Cu208+,). In this Brief Report we reportmagnetization and resistance data on the TI-Ca-Ba-Cu-0 samples. The onset temperature of magnetization is117-118 K, 23 K higher than that of a highqualityE~Ba~Cu30.r-, sample.Two samples with nominal compositions T12Ca2Ba-Cu3Og+, and T12Ca4BaCu301 I+, used in the present experimentswere pre ared using TI203, CaO, and BaCu304in the same batch. BP Both were heated at 9W°C for 3 minand furnace cooled. The resistance-temperature dependencesof these two samples are shown in Fig. 1. Theresistance was measured by the standard four-probe techniquewith silver-paste contacts. These two samples bothhave onset temperatures near 140 K (defined by thesmallest curvature of the resistance-temperature curve lo),midpoint of 127 K, and zero-resistance temperature at121 K. Zero-resistance data correspond to resistivitiesless than 10 -6 n cm. The difference of calcium contentsin these two samples does not seem to have significantlychanged their superconducting behavior, which dependsstrongly on preparation conditions of the samples. Notethat the Tl2CazBaCu~Og+, sample consists of approximately80% of the new 2:2:2:3 superconducting phases9Magnetization measurements were performed utilizinga superconducting quantum interference device (SQUID)magnetometer manufactured by BTI Corp., San Diego,CA. Figures 2 and 3 show dc magnetization (field cooledand zero field cooled) as a function of temperature for anapplied field of I mT for samples of Tl2Ca2BaCu309+,and T12Ca4BaCu3Ol I +, respectively, whose resistive behavioris shown in Fig. 1. The Meissner flux expulsion is10- 15 % of the diamagnetic flux exclusion; the magnitudeof the diamagnetic shielding is large and consistent withcomplete flux exclusion from the samples at 10 K. Thisbehavior is similar to that observed in Tl-~a-CU-o,~ La-B~(s~)-cu-o," and Y-Ba-Cu-0 samples.12 Explanationssuggested to date include existence of a superconductingglass state," anisotropy effects,I3 and superconductivityconfined to thick shells around normal grains. l2At this stage, we hesitate to speculate on differentiationbetween these causes, and only note that our samples wereTENPERANRE (K)FIG. 1. Resistance-temperature dependences of samplesTlzCazBaCu~Op+, (triangles) and TlzCarBaCu~011+, (circles).- 37 9742@ 1988 The American Physical Society

DownSosded 29 Mar 2008 to 130.984,237.6. Redistribution subject to AIP license or copyright; see http:l1apl.aip.org/apUcopyr1 ht.jspBNew 120 K TI-Ca-Ba-Cu-0 superconductor2. Z. Sheng, W. Kiehl, J. Bennett, A. El Ali, D. Marsh, G. D. Mooney, F. Arammash,J. Smith, D. Viar, and A. M. HermannDrparrmenr of Phy.sics, Universi?y ufArkansas, Fayetteoille, Arkansas 72701(Received 26 February 1988; accepted for publication 30 March 1988)Electronic and magnetic properties of the new 120 K TI-Ca-Ba-Cu-0 superco~~ductor arepresented. Kesistance-temperature variations, ac susceptibility, and the (low) magnctic fielddependence of the critical current density are reported. The new superconductor can beprcpared in a molten state, which should allow processing leading to high critical currentdensity. Preliminary x-ray diffraction data are also presented.Discoveries of 30 K superconductivity in the La-Ba-Cu-0 system' and 90 K superconductivity in the Y-Ba-CJu-0system2 have stimulated a worldwide race for higher tcmperaturesupcrconductors. In spite of much effort in the pastyear, stable and reproducible supercorlductivity has remainedat the level of 90 K in the RE (rare earth)-Ba-Cu-Osystem. Breakthroughs were recently made by the discovcricsof thc 90 K TI-Ra-Cu-0 system,'-' 1 I0 K Bi-Ca-Sr-Cu-0system,'*" and 120 K TI-Ca-Ra-Cu-0 system.' In this letter,we report electronic and magnetic data on the r~cw superconductingT1-Ca-Ba-Cu-0 system, and presect preliminarycritical current density data and x-ray diffraction patterns ofthe new gelieration of high-temperature superconductors incompariscn with those of the conventi:)nal RE-Ba-Cu-0systems.T1-Ca-Ba-Cu-0 samples used in the present experimcntswere prcpared using the procedlire similar to that wepreviously described.' Briefly, appropriate amounts ofT1,0,, CaO, and BaCu,O, (cr Ba,Cu,O,) were complctelymixed and ground, and pressed into a pellet. The pellet wasthen put into a tube furnace, which had been heated to 880-920 'C, and was heated for 3-5 min in Rowing oxygen. Thesample was then hrnace cooled to room temperature inabout I h. The BaCu,O, and Ra,Cu,O, were prepeied accordingto the procedure we previously developed." Ascries of samples with nomind compositions ofT1,CayBaCu,O, ., + , with y = 1, 1.5. 2, 3, and 4 was prepared.Samples with nominal composition ofTl2Ca,Ba,Cu,O,,, + , were also synthesized. All samplesreadily levitate over a magnetic field. Because our samplesare multiphase, the strong Mcissner effect observed is due toboth a lnrge volume fraction of superconductive phase and ton higher transition temperature. Furnace cooling slightlyimproves the su~erconducting behavior of the samples. Thisindicates that the TI-Ca-Ba-Cu-0 superconducting compoundis stable over a relatively large range of temperatures,and thus the T1-Ca-Ba-Cu-0 superconductor may be of importancein many applications. We believe that annealing inoxygen is not required for these samples and that thc improvementof the superconducting behavior is mainly due tothe decrease in damage to the crystalline structure by slowcooling.Resistance was measured by the standard four-probetechnique with silver paste contacts. Figure 1 shows resistance-temperaturedependence for a sample with y = 4. Thezero-resistancc temperature of this sample is 106 K, threedegrees higher than that which we reported previously for asample with y = 1.5 (see Ref. 7). Usually, the onset temperatureof the sharp drop in resistance, defined by smallestcurvature of resistance-temperature curve,9 is above 120 K,and we beiieve that a higher temperature for zero resistancecould be achieved by improvement of preparation prwedure.''Figure 2 shows 5 kHz ac susceptibility for a sample withy = 4 as measured by the techniqge of Norton." The onsettemperature of susceptibility drop of this sample is about 117K (as seer, in the high gain inset), and the lower temperatureleveling occurs at abo~t 85 I

APPLIED PHYSICS COMHUNICATIONS, 7(4), 237-253 (1987)ICRONIG-PENNEY BAND STRUClXlU3 FOR ONE DIMENSIONALSUPERLArnCEsWilliam G. Hartex, G. David Mooney, and Allen M HcrmannDepamnent of PhysicsJ. William Fdbright College of Arts and SciencesUniversity of ArhcqFayettcville, AR 72701~etailtd calculations of high order band-gap boundaries rn given as fnnctions ofpotential barrier heights and widths for a range of superlattice structures.Prominent features of the structures include a multitude of band-gaps above thebeer as well as many band crossings An int-tation of the crossings pmvidesan appro xi ma^^ fomiuIa for estimating th

h;. -APPLIED PHYSICS CONMUNICATIONS, 7 (4), 275-283 (1987)A. H. Hermann, 2. Sheq, W. Kiehl, D. wrsh, P. Ar-ah, A. El Ali,D. Wney, and L. ShengDeparbwnt of physicsuniversity of ArkansasFayetteville, Arkansas 72701J. A. mollam and A.Depar mnt of Electrical mgineer inguniversity of NebraskaLincoln, Nebraska 68588-0511iTe rrepot the preparation and chsrecterizatlon of a mlt-processiblehigh-tm&erature aupereonductor bad on a melt reaction batween 'IbBa2Cu307-xand YlS2 Ba0.8Cu34y at 950°c. he samples have sharp superconductingtransitions above 90K as determined by resistivity and AC susceptibilitymeasurerrats. x-ray diffraction shows crystal strwtures nearly identicalto thorn of Y Ba2Cu307, sintered superconductore. Ttb? dense void-freemelt-processible superconductore have potential for large single crystalgrowth and for high critical currents with low20at processing.

Melt-processible rare earth-Ba-Cu-0 superconductors based on moltenBa-Cu oxidesA. M. Hermann and 2.2. ShengDepartment of Physics. Uniuenity of Arkansas, Fayettevilk Arkansav 72701(Received 16 September 1987; accepted for publication 12 October 1987)Ba-Cu oxides with certain compositions were found to be meltable or partly meltable at arelatively low temperature of 950 "C. High-quality, dense, and void-free high-temperaturesuperconductors based on rare earth(=)-Ba-Cu oxides of the form REBa,Cu,O,, ,,havebeen made by reactions between molten Ba-Cu oxiderr and solid RE-containing materials at thesaine temperature. This has allowed superconductive components to be made in a wide verietyof shapes. This melt process will significantly alter oxide supercor,ductor preparationprocedures, and will allow growth of large sing!e crystals aiding in the understanding of themechanism of superconductivity and in the search for superconductivity at even highertransition temperatures.The discovery of 35 K superconductivity in theLa-Ba-Cu-0 system by Bednorz and hhUerl was soon followedby the finding of 90 K superconductivity in the Y-Ba-Cu-0 sy~tem.~ The scperconduct.ive compoundYBa2Cu30,, .,, was rapidly identif~ed,~ and a series ofREBa,Cu, O,, + , superconductors, with RE = La, Nd,Sm, Eu, Gd, Dy, Ha, Er, Tm, Yb, and Lu, was prepared witha similar transition temperat~re.~ This breakthrough ofabove-liquid-nitrogen temperature superconductivity hassimulated strong activity in theoretical and experimental researchaimed at finding superconductivity at even highertransitioa temperatures, up to or even above room temperature.Superconductors of the type REBa, Cu, O,, + , are conventionallyprepared as ceramics by solid-state reactions,and useful forms are fabricated by compression. The ceramicmaterials are porous, and the preparation process is relativelyintractable for manufacturabiljty. In a series of experimentson substitution of rare earths for Y inYBa, Cu3 O,., +. , we found that TbBa, Cu, O,, , melted(pdy) at 950 "G--a typical temperature at which otherrare earth-bariumcopper oxides form sintered superconductorsby solid-state reaction, and that the moltenTbBa, Cu, O,, + , reacts with Y ,,, Bq,, CuO,, + . to form ahigh-quality superconduct~r.~ In this letter we report a newfinding, that Ba-Cu oxides with certain compositions aremeltable or partly nieltable at 950 "C, and that the moltenBa-Cu oxides react with RE-containing materials, such asRE oxides, RE-Ba- and Mi-Cu-binary oxides, or RE-Ba-Cu-ternary oxides, to form melt-processible high-quality,dense, and void-free supermnductors of theREBa2Cu,0,., +, class. This finding has ellowed high-qualitysuperconductor components of arbitrary shape to be fabricatedat a relatively low temperature. We believe the meltprocess will significantly alter the superconductor samplepreparation procedures used by researchers in this field, Adis ultimately expected to be of importance in the understandingof the mechanism of high-temperature oxide superconductivity,and in the search for new superconductors witheven higher transition temperatures.A typicd procedrm for preparing this kind d meltprocessibiesuperconductor is as follows: appropriateamounts of BaCO, and CuO were mixed and ground in anagate mortar, and heated in air at 900 'C for 12 h to obtain amaterial with nominal composition of Ba, Cu, 0,. The heatedmixt- re was reground and pressed into a pellet. Similarly,Y,0, , BaCO, , and CuO were mixed, ground, and heated inair at 950°C for 12 h to obtain the material with nominalcomposition of Y,,, Bbs CuO,, ,. The heated mixture waspowdered. The Ba,Cu,O, peilet was put on theY,, Ba,,s CUO,~ + powder, heated in flowing 4 at 950 "Cfor 24 h in a tube furnace, and then at 650 "C for 4 h.Finally, the materials were furnace cooled in flowing 0,to less than 200 "C in 1-2 h before they were removed fromthe furnace. After the heating at 950°C. the gray-blackB+Cu,Os pellet has partly melted. A black melt-recrystal-M-like superconductive chunk (usually a half-spherewith a diameter, for example, about 1 cm) was formed,which was embedded in excess dark-greenY,.2B%.8Cu03.6.+ x powder. The Y,.ZB%.8Cu0,.6 .+ x pow-der was then removed. Rectangular paralleiepiped sampleswere cut with a saw to typical dimensions of 1 x 2 x 7 mrn.Several samples were prepared frcm interior cuts of theoriginal half-sphere.The x-ray powder diffraction pattern of this mdt-processible material was the same as that ofYB~,CU,O,,~ ,. with two additional lines !at d = 2.52 A(71 t ) and 2.33 A ( t 11 ) 1 characteristic of a small amount ofCuO present. This melt-processible material has a density ofabout 6.4 g/cm3, close to the value of 6.3 g/cm3 forYBa,Cu,O,., +, predicted from theory.' This materialcould be easily levitated in a magnetic field, and showed astrong diamagnetic flux exclusion above liquid-nitrogentemperature, comparable to that of any sample we have carefullyprepared using the conventional sinter method.' Thissample had a midpoint transition temperature of 93 K and atransition widt.5 of about 2 K, as measured by a standardfour-probe method.5 Figure 1 shows a typical variation ofresistance as a function of temperature dqwn to liquid-nitrogentemperature.According to Cava et the Y,, Ba,,,,CuO,, ,. consistsof 1/3 ofblack superconductive phase YBa, Cu, 06.,and 2/3 of nonsupercorrductive green phase Y,BaCu05, Inthe above example, formation of the high-quality black+ ,1854 Appl. Phys. Lett. 51 (22),30 November 1087 0003-6951 /87/481854-03501.00 O 1887 Arnerlcan Institute of Physics 1854

APPLIED PHYSICS COUUUNICATIONS, 1(4), 285-290 (1987)Microstructural Analyis of Y-Ba-0.1-0 mylBrconductDes by Light MicrOWpyH. Sreshta, Department of Mechanical EngineeringandA. M. wrmann aM Z. Z. shwrg, Department of Physic6university of ArkansasFayetteville, AR 72701The roan temperature microstcucture of einterd Y-Ba-'3-0superconductors as determined by light rnicro~~~py is reported in thie paper.The aintered superconductors contain poroai ty in a platelet nultrix Ofsupeeconluctor: phase. qhe platelet growth habit suggeats liquid-phaseSinter ing is a daninant mechanibn for growth.4 Microatructural analysis using light microscogy has been a pwerW -18for quantifying features such as grain size, grain shape, phase size, andphase shape in metallic system. In ceramic ba& systems, the urn of lightmicroscopy has ~een limited to materiale that are capable of reflectinglight. Struere Inc. (1) have ahown that light microscopy can be used toinvestigate the configuration of silicon wafers, determine the integrity ofitb interface bond wtueen alunina and a metallic coating, and esthte the