The Electronic Structure, Metastability and Transport Properties of Optimized Amorphous Silicon-Germanium Alloys
- PDF / 503,203 Bytes
- 6 Pages / 414.72 x 648 pts Page_size
- 35 Downloads / 212 Views
achieve uniform carrier generation rates. Light exposure for the Uni-Solar samples alloys was at 6 W/cm 2 for 70 hours, and for the Harvard's samples 1.7 W/cm 2 for 264 hours. Four methods were utilized to 103 O characterize the defect structure: DriveLevel Capacitance Profiling (DLCP) [9], ""TPC Transient Photocapacitance and Transient 2 Photocurrent spectroscopy (TPC & TPI) o PI ., 10 [101 and Modulated Photocurrent (MPC) _ [11]. The ambipolar diffusion lengths were z x=0.68 • obtained using the Steady-State 0• 10oPhotoncarrier Grating method (SSPG) [12].
6
The SSPG measurements for the Harvard samples were carried out at Harvard University, while those for the Uni-Solar
10
samples were conducted
at The Hebrew
University in Israel.
z
3. RESULTS AND DISCUSSION 3.1 Defect Studies
Z1i 1
x=0.35
•-•
The combination of TPC and TPI spectra are able to distinguish between the OPTICAL ENERGY (eV) optically generated electron current from the hole current [10,13]. By comparing the Fig. 1: The TPC and TPI spectra of one Harvard TPC and TPI spectrum of intrinsic samples sample with 68at.% Ge and one Uni-Solar with the doped samples, we were previously sample with 35at.% Ge. The model calculations able to identify significant positively charged are represented by solid thin lines. One filled defects in Uni-Solar's intrinsic a-Si_,xGex:H subband is used to fit the spectra of Harvard alloys [13,14]. Further more, we found that samples while, for Uni-Solar samples, one filled this D+ defect density matched the defect subband and one unoccupied subband must be density obtained from our DLCP used in the model calculations. measurements for all the Uni-Solar intrinsic samples in both the annealed and lightsoaking states. Thus, we have inferred that the defect sub-band from the DLCP measurements most likely corresponds to D- sub-band density [13]. The defect structure of the Harvard's samples, on the other hand, appears to be different. Figure 1 displays the TPC and TPI spectrum of one Harvard and one Uni-Solar intrinsic sample. It appears that there is only one dominant optical defect transition (D--4D++e-, as we argue below) for the Harvard samples, resulting in a perfect overlap of the TPC and TPI sectra outside the band tail region. That is, if there were significant concentrations of D and D defects, this would result in a significant optical defect transition of the type D++e---->D0 . However, because the hole current contributes with an opposite sign to the TPC and TPI spectrum, the two spectra would then deviate at moderate optical energies in a manner similar to the Uni-Solar sample spectra. This is not observed for any of the Harvard samples. Thus, in contrast to the Uni-Solar samples, this suggests that the Do charge state dominates in the Harvard samples. This hypothesis is supported by a second experimental observation. The defect sub-band observed in MPC measurement had been identified as arising from the Do defect sub-band in a-Si:H [15]. For Uni-Solar's samples, this MPC defect sub-band exhibits a very sma
Data Loading...
