Performance of annealed hybrid silicon heterojunctions: A numerical computer study

Abstract

The performance of the standard hydrogenated amorphous silicon carbon–crystalline silicon solar cell is extensively compared with the performance of a hybrid structure subjected to a high-temperature annealing processing. Our analysis indicates that high-temperature-annealed heterojunctions show more robustness in the presence of energy offsets and defective amorphous-crystalline interfaces. Annealed hybrid cells are also less vulnerable to the negative impact of amorphous silicon carbon doped layers with poor electrical properties. Furthermore, annealed structures have the potential to generate higher efficiencies than conventional heterojunctions regardless of the wafer quality. The presence of boron at the amorphous-crystalline interface and in the wafer front region plays an important role in annealed hybrid structures that are made with low-quality wafers or where there is a highly defective amorphous-crystalline interface. In this scenario, a linear boron profile in the wafer front region is more appropriate, for which there is an optimum thickness. For low defect amorphous-crystalline interfaces and high-quality wafers, a boron exponential profile is more appropriate when boron creates additional defects in the front region of the wafer. The shape of the boron profile becomes less relevant when the boron does not add additional defects to the front region of high-quality wafers and when the amorphous-crystalline interface is low defect or defect-free.

Further insight on recombination losses in the intrinsic layer of a-Si:H solar cells using computer modeling tools

“…6,7 Our software D-AMPS-1D (Refs. [8][9][10][11][12] is based on the AMPS-1D source code, (b) some of their electrical parameters are unusual, (c) fittings of experimental J-V curves under illuminated conditions are not fully shown: only V OC and the ratio between the maximum power density P MAX and the photocurrent J P measured at a reverse voltage of À2 V were really matched, (d) no fittings of dark J-V curves were included, (e) red laser instead of the AM1.5 light source was used, and (f) the maximum power density of the as-deposited a-Si:H cells is believed to have achieved the hole-mobility limit established by valence band-tail trapping; i.e., defect states did not limit their efficiency.…”

“…D-AMPS is an updated version of the well-known software AMPS that includes extra features like amphoteric states, Defect Pool-model (DPM), and Pool-Frenkel effect, a simplified treatment of light scattering, among others. [8][9][10][11][12] A. Solar cell devices: Experimental and simulated performances…”

“…Undiluted Naturally, electrical parameters used by different groups working on a-Si:H solar cell modeling show some spread. [11][12][13][14][16][17][18][19][20][21][22] However, the electrical parameters of Liang and Schiff 4,15,23 have two unusual features: a considerably higher density of states at mobility edges (G AO ¼ 1.6 Â 10 22 cm À3 and G DO ¼ 6 Â 10 21 cm À3 ) with narrower tails (E A ¼ 20 meV and E D ¼ 40 meV) and poor free carrier mobilities (l N ¼ 2 cm 2 V À1 s À1 and l P ¼ 0.3 cm 2 V À1 s À1 ). These unusual parameters have a significant impact on the predicted J-V characteristics.…”