Interface Effects on Photocarrier Transport in a-Si:H
- PDF / 272,785 Bytes
- 5 Pages / 420.48 x 639 pts Page_size
- 16 Downloads / 238 Views
INTERFACE EFFECTS ON PHOTOCARRIER TRANSPORT IN a-Si:H P. D. Persans', D. Arnzen', G. Possin'*, L. D'Anna*, X. S. Zhao* and K. Breton' * Physics Department and Center for Integrated Electronics, Rensselaer Polytechnic Institute, Troy, NY 12180 ** General Electric Corporate Research and Development Laboratories, Schenectady, NY ABSTRACT We discuss the absorption-length and film thickness dependence of photoconductivity in amorphous silicon using measurements of the ambipolar diffusion length and the Debye screening length. INTRODUCTION The surface and interface properties of a-Si:H thin film devices can control the performance of such devices. It is clear that space-charge layers control dark conductivity and can affect photoconductivity in coplanar structures [1]. For example, a decrease in photoconductivity is usually observed for photocarriers generated near an interface. It has been proposed that this decrease is either due to surface recombination [2,3] or to band bending due to pinning of the surface by defects [4]. In this paper we address the question of the mechanisms for surface effects on photoconductivity by comparing ambipolar diffusion length, Debye screening length, interface defect densities, the chemical potential dependence of photoresponse, the thickness dependence of photoresponse and the energy dependence of photoconductivity in the high photon energy regime. EXPERIMENTAL DETAILS Intrinsic a-Si:H samples between 0.1 and 1 pm in thickness were prepared by rf plasma-assisted deposition from pure silane onto glass substrates heated to 3000C. Further information on the preparation and properties of similar samples can be found elsewhere [5]. Contacts for coplanar photoconductivity were painted with carbon paint with approximately 2 mm separation. Photoconductivity was measured using monochromatic 2 light chopped at ;u10 Hz. A weak (Frzl0'3 photons/s/cm ) dc bias light was incident on the sample in order to linearize the ac response and to keep the response time constant as a function of incident ac flux. The magnitude of the photocurrent response was also kept approximately constant by inserting neutral density filters in the excitation beam and varying the slit width. The deflection medium for PDS was CC14. All spectra were normalized in real time by splitting off a portion of the incident monochromatic flux to a flat-response power detector. The ambipolar diffusion length was measured using the steady-state photocarrier grating technique [6,7]. The output from a vertically- polarized 5 mW HeNe laser was spread to about 5 mm diameter and split into a cw "strong" beam ( 20 mW/cm2) 2 and a chopped "weak" beam (13 Hz, 0.1 mW/cm ). The two beams were recombined
Mat. Res. Soc. Symp. Proc. Vol. 219. @1991 Materials Research Society
838
on the sample to create an interference grating. The polarization of the strong beam could be rotated using a 1/2 wave plate so that the two beams could be polarized parallel or perpendicular to one another, giving either interference or no interference. The modulation
Data Loading...
