622
A. Jariwala et al.
sample. In case of super bandgap regime at 1.7 eV, the penetration depth of incident
light is hardly about 100 nm owing to very high absorption coefficient of CdTe
(10
5 cm
−1 ) in this regime. Thus, the fast component must originate from SCR1, and
hence reasonably be assigned to surface recombinations. In case of the LT sample,
the signal from SCR1 is not appreciable anymore. Hence, the slower values with τ
of 1.6 and 3.1 ms for both sub- and super-bandgap regimes, respectively, must both
be corresponding to SCR2. Interestingly, in the non-LT sample under sub-bandgap
illumination, τ is very similar at 2.0 ms, suggesting that the SCR2 dominates the SPS
dynamics in the sub-bandgap regime. This is realistic because the electron-hole pairs
are generated quite consistently in the sub-bandgap regime, and also the depletion
region at SCR2 is broader than that of SCR1. Using these data and discussion, it has
been successfully corroborated that the laser annealing treatment drastically reduces
the Fermi level pinning at the surface.
Thus, it is shown that the chopping frequency is a handy tool to understand the
charge-carrier generation and separation processes in semiconductor surfaces as well
as bulk.
4.3 Illumination Intensity Dependence of SPS
One more variants of the SPS measurements is the illumination intensity. The SPS
measurements can be performed by varying the illumination intensity in order to
distinguish various SPV mechanisms that respond differently to the change in illumination. Jing Liqiang et al. have demonstrated the use of illumination intensity
dependent SPS measurements to identify bulk states and the surface states [78]
Their analysis reveals that the charge dynamics at the surface states and that in the
bulk states behave drastically differently in the intensity dependent SPS. Hence, with
appropriate design of experiment, it is feasible to exploit this difference to distinguish
between the bulk states and the surface states. They have supported their conclusions
by experimental results on two samples of InP: one with a high density of bulk states
whereas the other one with a high surface states density. They have shown that a
soft saturation regime is characteristic of bulk states, whereas a superlinear SPV as
a function of the illumination intensity is characteristic of surface states. Thus, the
intensity-resolved SPS could be a handy tool to distinguish between surface and
bulk states. Sharma et al. [79] have also used illumination intensity dependent SPS
measurements on SI-GaAs samples in both, super-bandgap as well as sub-bandgap
regimes. They reported unusual variation of SPV signal as a function of illumination
intensity in the sub-bandgap region at 887 nm wavelength. They show that the SPV
signal start decreasing with the increasing illumination intensity beyond a certain
value of intensity in the SI-GaAs samples measured at 28 and 67 Hz chopping
frequencies. They successfully explained these variations by using three different
SPV mechanisms in the sample, out of which two have strong intensity dependence
and opposite phases in their contribution to the measured SPV signals. They could
A. Jariwala et al.
sample. In case of super bandgap regime at 1.7 eV, the penetration depth of incident
light is hardly about 100 nm owing to very high absorption coefficient of CdTe
(10
5 cm
−1 ) in this regime. Thus, the fast component must originate from SCR1, and
hence reasonably be assigned to surface recombinations. In case of the LT sample,
the signal from SCR1 is not appreciable anymore. Hence, the slower values with τ
of 1.6 and 3.1 ms for both sub- and super-bandgap regimes, respectively, must both
be corresponding to SCR2. Interestingly, in the non-LT sample under sub-bandgap
illumination, τ is very similar at 2.0 ms, suggesting that the SCR2 dominates the SPS
dynamics in the sub-bandgap regime. This is realistic because the electron-hole pairs
are generated quite consistently in the sub-bandgap regime, and also the depletion
region at SCR2 is broader than that of SCR1. Using these data and discussion, it has
been successfully corroborated that the laser annealing treatment drastically reduces
the Fermi level pinning at the surface.
Thus, it is shown that the chopping frequency is a handy tool to understand the
charge-carrier generation and separation processes in semiconductor surfaces as well
as bulk.
4.3 Illumination Intensity Dependence of SPS
One more variants of the SPS measurements is the illumination intensity. The SPS
measurements can be performed by varying the illumination intensity in order to
distinguish various SPV mechanisms that respond differently to the change in illumination. Jing Liqiang et al. have demonstrated the use of illumination intensity
dependent SPS measurements to identify bulk states and the surface states [78]
Their analysis reveals that the charge dynamics at the surface states and that in the
bulk states behave drastically differently in the intensity dependent SPS. Hence, with
appropriate design of experiment, it is feasible to exploit this difference to distinguish
between the bulk states and the surface states. They have supported their conclusions
by experimental results on two samples of InP: one with a high density of bulk states
whereas the other one with a high surface states density. They have shown that a
soft saturation regime is characteristic of bulk states, whereas a superlinear SPV as
a function of the illumination intensity is characteristic of surface states. Thus, the
intensity-resolved SPS could be a handy tool to distinguish between surface and
bulk states. Sharma et al. [79] have also used illumination intensity dependent SPS
measurements on SI-GaAs samples in both, super-bandgap as well as sub-bandgap
regimes. They reported unusual variation of SPV signal as a function of illumination
intensity in the sub-bandgap region at 887 nm wavelength. They show that the SPV
signal start decreasing with the increasing illumination intensity beyond a certain
value of intensity in the SI-GaAs samples measured at 28 and 67 Hz chopping
frequencies. They successfully explained these variations by using three different
SPV mechanisms in the sample, out of which two have strong intensity dependence
and opposite phases in their contribution to the measured SPV signals. They could