Here - PMOD/WRC

ight half is uncoated.
Measurements
So far, only a limited number of measurements have
been made. We will present the preliminary results and
some preliminary conclusions. More extensive
measurements are needed to make definite suggestions and
conclusions.
A measurement sequence determining responsivitiy change
with applied oxide bias at selected wavelengths is
presented in fig. 2. The responsivity change saturates with
oxide bias approaching 30 V. The saturation level is a
measure of the IQD caused by recombination losses
between the oxide and the pn junction of the diode.
to the reflectance from the surface of the detector caused
most likely by differences in the oxide thickness. In the
next generation of these detectors a different process for
the growth of the oxide layer is needed.
The detectors and the thin layer of gold have a relatively
large surface. A criterion to succeed with the proposed
method is that the gold bias saturates the IQE uniformly.
We measured the difference in uniformity with 30 V gold
bias and without bias. The results are presented in fig. 4
and shows that the change is relatively uniform taken into
account that a semiconductor laser was used in the
characterization.
Change in responsivity [%]
5
4
3
2
1
400
420
440
460
0
0 10 20 30 40
Applied bias of gold [V]
Figure 2. The responsivity change with applied oxide bias
at selected wavelengths between 400 and 460 nm is
presented.
The uniformity of the detector responsivity is critical in
order to achieve the highest possible accuracy. The
uniformity of the detector measured with a semiconductor
laser at 405 nm. Simultaneously, the spatial reflectance
was measured allowing decomposition of the loss
mechanisms and determination of the source of the
nonuniformity. In fig. 3 the measured uniformity (a) and
spatial reflectance (b) is presented.
Figure 4. The uniformity change with an applied oxide
bias of 30 V.
Preliminary conclusions
So far, promising results of the principles have been
obtained. The responsivity changes with oxide bias have
been as predicted. The uniformity of the detectors was
poorer than expected, making the ones tested so far
unsuited as single detector standards. However, since the
reason for the nonuniformity is variations in the
reflectance, mounting two detectors in a wedge trap
configuration is likely to make the detectors suitable for
calibration purposes as well. It is highly likely that the
uniformity can be improved in the next generation of
detectors. More extensive measurements and comparison
to the CR will be presented at the conference. In addition,
necessary corrections and caution with the method will be
presented.
Acknowledgments This project is partly financed by
the Nordic Innovations Center (NICe).
(a)
Figure 3. The detector uniformity (a) and spatial
reflectance (b) is presented.
We note that the nonuniformity as shown in fig. 3 (a) is in
the order of 4 %. This high nonuniformity does not make
the detector suited as a standard. The uniformity
measurement shows that in areas where the responsivity is
high we have a dip in the reflectance and vice versa.
Therefore, the reason for the nonuniformity is mainly due
(b)
References
Zalewski, E. F., Geist, J., Silicon photodiode absolute spectral
response self-calibration, Appl. Optics, 19, 1214 – 1216, 1980.
Verdebout, J., Booker R. L., Degradation of native oxide
passivated silicon photodiodes by repeated oxide bias, J. Appl.
Phys., 55, 406-412, 1984.
Gentile T. R., Houston J. M., Cromer C. L., Realization of a scale
of absolute spectral response using the National Institute of
Standards and Technology high-accuracy cryogenic
radiometer, Appl. Optics, 35, 4392-4403, 1996.
Haapalinna A., Kärhä P., Ikonen E., Spectral reflectance of
silicon photodiodes, Appl. Optics, 37, 729-732, 1998.
64