- Details
- Last Updated: Saturday, 24 December 2022 19:13
Characterization and Modeling of Defects in Polycrystalline Silicon
Goal: Characterization and modeling of Polycrystalline Silicon Defects using a Modified Light Beam Induced Current (LBIC) to help optimizing Solar Cell Screen Printing fabrication process.
Solar grade cast polycrystalline silicon wafers and multi-crystalline silicon layers have high density of extended defects that are notorious for a strong reduction of solar cell efficiency. Recombination centers and traps (defects decorated with impurities, dangling bonds,...) sink the photogenerated current and shunt the cell junction. Laser/Electron Beam Induced Current (LBIC / EBIC) are excellent techniques for detecting these centers and fully characterizing their recombination activity. They were originally designed to analyze grain boundaries (GBs), precipitates, and dislocations.
To understand the carrier dynamics effected by extended defects (planar, line, large precipitates, gigantic voids,...), I developed an original physics model for minority carrier recombination, based on distributions of carrier trapping and recombination centers within extended defects. This allowed detailed characterization of carrier recombination by GBs, solar cell surface, and layer interfaces. The model was based on experimental observations; it considers active extended defects as arrays of zero dimensional recombination centers. Continuity equations were written and special boundary conditions were developed. Green’s Function formalism was employed to solve them. Crystal orientations of adjacent grains and coincidence lattice allowed to address those boundaries that can be decorated with impurities and expected to be electrically active. Since the system combines various symmetries and a large number of parameters, the problem was solved numerically. Excellent matching with experimental data was obtained. This work clarified the distribution of recombination centers within the GBs, which largely varies depending on the nature of the GB, the thermal processes it has undergone, and the contamination/doping. Similar work was done for dislocations vertical (threading) and parallel to the junction (misfit dislocation).
To enable characterization of polycrystalline silicon solar cells during processing, the LBIC technique was augmented by simultaneously mapping the local reflectance and the induced current in the test device. “As-processed cells” have roughness and are intentionally textured. I referred to the new approach as “Reflectance Coupled to LBIC (RC‑LBIC)”. The technique removes complication of the LBIC signal due to variations of photon absorption. Furthermore, study of reflected beam provides valuable information on recombination centers at the surface. Using RC-LBIC, textured polycrystalline silicon cells were successfully characterized in a non-invasive fashion, whereas traditionally, solar cells should be polished to obtain meaningful LBIC signal. RC-LBIC was used to calculate accurately the recombination velocity at extended defects.
RC-LBIC was also utilized to characterize electrically active defects in polycrystalline silicon solar cells that were passivated by either molecular hydrogen or H-plasma. The passivation performance was measured by the reduction of GB and dislocation recombination velocities. Effect of passivation on traps related to dangling bonds and attached impurities and clusters was studied through the photocurrent and fill factor of I-V characteristics of passivated solar cells.
RC-LBIC was also instrumental for characterizing bevel-polished samples from cells treated with “aluminum back side gettering”, where we evaluated the depth and uniformity of the back surface field (BSF) in screen-printed silicon cells. The data was used to optimize the solar rear electric contact as well. RC-LBIC was also instrumental for analyzing the effects of wet chemical etching on the removal of surface states in polycristalline silicon cells, and the resulting change in carrier recombination at the solar cell surface.
The RC-LBIC based methodology has benefited the analysis of polycrystalline solar cell fabrication processes. This was crucial for the development of screen-printing technology, which has become the most used technology by the entire PV industry.