Surface Photo-Voltage Spectroscopy: A Versatile Technique …
629
consequence of the signal from SCR1 becoming dominant [108]. Note that SPS
signal plotted in Fig. 9 shows the absolute value of the SPV. In actual measurements,
the non-LT sample exhibits the change in sign of the signal near 1.51 eV. This is in
agreement with the earlier results obtained on CdTe device structure by the inclusion
of a surface space charge region [109] and confirmed by its elimination due to various
surface treatments [108, 110]. The elimination of this feature in the LT samples is an
evidence that the surface states have been passivated by the laser treatment and the
Fermi level pinning at the surface is now significantly reduced.
The hypothesis that the laser treatment passivate the surface is also supported
by the SPS phase spectra and the chopping frequency dependent measurements as
described earlier. It has also been confirmed by the Photoluminescence and TimeResolved Photoluminescence measurements (TRPL) on the samples before and after
the laser treatment and the results are shown to be in agreement with the above
conclusion.
Thus, using surface photovoltage spectroscopy, we could successfully demonstrate that a modest laser treatment of 250 W/cm
2 with a dwell time of 20 s is
sufficient to reduce the effects of Fermi level pinning at the surface due to surface
defects.
5.4 Minority Carrier Diffusion Length and Carrier Life
Times
Surface photovoltage (SPV) measurements have been used extensively to measure
the minority carrier diffusion length, L, particularly for solar cell applications [111–
115]. It must be noted here that the measurement of diffusion length requires the
L to be much larger than the thickness of the semiconductor materials involved.
Further, the L should also be much larger than the depletion width in order to reliably
measure it. Although these requirements may not be an issue in wafer level, they
certainly are matter of concern when it comes to the thin film devices. Toušek,
et al. made an interesting attempt to generalize the analysis for silicon samples
[109] and subsequently for thin film CdTe [108]. They used the SPS measurement
in combination with the constant photocurrent method to determine the diffusion
length in thin film CdS/CdTe solar cells structures. They presented a model with one
SCR at the CdTe/CdS junction and theoretically calculated the SPV signal without
constraints of absorption coefficients for the incident radiation and thickness of the
sample. In order to exclude the effects of surface recombination velocity on their
measurements, they measured the SPV spectra from the back surface as well. In
case of SPS measured on back surface under illumination without the ohmic contact
there facilitates measurements which are predominantly governed by the diffusion
length. On the other hand, the standard measurement using light penetrating from
the CdS side strongly depends on the thickness of the SCR. They could successfully
explain their measurements and determine the minority carriers diffusion length
629
consequence of the signal from SCR1 becoming dominant [108]. Note that SPS
signal plotted in Fig. 9 shows the absolute value of the SPV. In actual measurements,
the non-LT sample exhibits the change in sign of the signal near 1.51 eV. This is in
agreement with the earlier results obtained on CdTe device structure by the inclusion
of a surface space charge region [109] and confirmed by its elimination due to various
surface treatments [108, 110]. The elimination of this feature in the LT samples is an
evidence that the surface states have been passivated by the laser treatment and the
Fermi level pinning at the surface is now significantly reduced.
The hypothesis that the laser treatment passivate the surface is also supported
by the SPS phase spectra and the chopping frequency dependent measurements as
described earlier. It has also been confirmed by the Photoluminescence and TimeResolved Photoluminescence measurements (TRPL) on the samples before and after
the laser treatment and the results are shown to be in agreement with the above
conclusion.
Thus, using surface photovoltage spectroscopy, we could successfully demonstrate that a modest laser treatment of 250 W/cm
2 with a dwell time of 20 s is
sufficient to reduce the effects of Fermi level pinning at the surface due to surface
defects.
5.4 Minority Carrier Diffusion Length and Carrier Life
Times
Surface photovoltage (SPV) measurements have been used extensively to measure
the minority carrier diffusion length, L, particularly for solar cell applications [111–
115]. It must be noted here that the measurement of diffusion length requires the
L to be much larger than the thickness of the semiconductor materials involved.
Further, the L should also be much larger than the depletion width in order to reliably
measure it. Although these requirements may not be an issue in wafer level, they
certainly are matter of concern when it comes to the thin film devices. Toušek,
et al. made an interesting attempt to generalize the analysis for silicon samples
[109] and subsequently for thin film CdTe [108]. They used the SPS measurement
in combination with the constant photocurrent method to determine the diffusion
length in thin film CdS/CdTe solar cells structures. They presented a model with one
SCR at the CdTe/CdS junction and theoretically calculated the SPV signal without
constraints of absorption coefficients for the incident radiation and thickness of the
sample. In order to exclude the effects of surface recombination velocity on their
measurements, they measured the SPV spectra from the back surface as well. In
case of SPS measured on back surface under illumination without the ohmic contact
there facilitates measurements which are predominantly governed by the diffusion
length. On the other hand, the standard measurement using light penetrating from
the CdS side strongly depends on the thickness of the SCR. They could successfully
explain their measurements and determine the minority carriers diffusion length