The Role of Deep Defect Relaxation Dynamics in Optical Processes in Hydrogenated Amorphous Silicon
- PDF / 464,821 Bytes
- 6 Pages / 414.72 x 648 pts Page_size
- 93 Downloads / 183 Views
FIG. 1. Arrhenius plot of the
conductance component of the ac admittance at four measurement frequencies. The Schottky diode a-Si:H sample was measured under a dc bias of -1.5V. At
sufficiently high frequency or
low temperature all curvesZ become identical and indicate
that the bulk film has a
C-)
102 Ea =0.725eV W C 00 10' EOOOOOO nkBE [ IqGAIgeF 1,,((o)l .Isin
' with EO = kBT log(o/v) (-o}
(1)
where on is the electron capture cross-section, is the electron thermal velocity, q is the electronic charge, G 1 is the sinusoidal component of the generation rate at angular frequency co, Pe is the free electron mobility, and F is the electric field. It is this quantity we have plotted as a function of the light modulation frequency in Fig. 2(a) for five measurement temperatures. In Fig. 2(b) the same quantity is plotted as a function of ECo using a value for the thermal emission prefactor, v, that is identical to that determined from early capacitance transient measurements for n-type a-Si:H [10]. The excellent agreement to thermally activated behavior with a peak energy near 0.63 eV is clearly evident. Because the value of v from the capacitance studies is known unambiguously to correspond to DIDo transitions [11], we believe that the MPC spectrum must be identified with the same transition. Thus, this spectrum indicates a band of DO states lying roughly 0.09eV above EF which occasionally capture (and then re-emit) the excess photoelectrons provided by the modulated light. The transient effects associated with turning on and off a 1.9eV source of bias light (with a photon flux of 1013 crn 2 s"1) on the ac admittance and on the MPC signal is shown in Figs. 3 and 4, respectively. At sufficiently high frequencies the admittance measurements, according to our previous discussion, simply reflect the changes in the conductivity in the undepleted portion of the film. Indeed, for frequencies larger than about 50Hz, all the curves are essentially identical to within a constant factor. The transient photoconductivity indicated is, indeed, quite consistent with previous measurements on coplanarsamples [3,4,5]. In particular, our curves clearly display the long time photoconductivity decay (LTPD) after the bias illumination has been switched off. 222
2 Ci,
(a)
101
-
380K
---
360K
-0-
320K 300K
S340K (b)
2 _j
100
Cf) 0~
1 0-1
0I-I
2
0 101
104 103 102 FREQUENCY (Hz)
105 0.4 0.5
0.6
0.7 0.8
0.9
ENERGY BELOW Ec (eV)
FIG. 2. (a). Modulated photocurrent spectra vs. frequency for five measurement temperatures. The signal is obtained form the real and imaginary components of the MPC response as given by Eq. (1). (b) The same data plotted against E,,, where the exponential prefactor, v, has the value 2 x 108 T 2 s0K 2 . Good activated behavior of the MPC spectra is indicated, with a 0.63eV characteristic energy at the peak. 10-7
1.2
300K~
a) E
-•.46 3
02
C 0
zz C 0
Data Loading...
