Improving GaP Solar Cell Performance by Passivating the Surface Using AlxGa1-xP Epi-Layer

Abstract

A good candidate for the top junction cell in a multijunction solar cell system is the GaP solar cell because of its proper wide band gap. Here, for the first time, we passivate the front surface of these GaP solar cells with an AlGaP layer. To study the passivation effect of this layer, we design a novel growth procedure via liquid phase epitaxy. X-Ray diffraction results show that the resulting passivation epitaxial layer is of good quality. Integrated quantum efficiency measurements show an 18% increase in current due to the AlGaP. The current-voltage measurements indicate that with this AlGaP surface passivation layer, the GaP solar cell's efficiency is 2.90%. This is an improvement over previously reported results for GaP solar cells.

Effects of growth temperature and device structure on GaP solar cells grown by molecular beam epitaxy

“…14 Current collection in GaP solar cells has proven even more problematic, considering that only 27% of the semi-empirical limit of J SC has been achieved and that internal quantum efficiencies (IQE) of prior cells have not exceeded $67% at any single wavelength. 15 While GaP is an indirect semiconductor, the X valley minima at 2.26 eV (k ¼ 549 nm) are relatively close in energy to the strongly absorbing, direct C valley at 2.78 eV (k ¼ 446 nm). The indirect band structure leads to poor absorption for k ¼ 446-549 nm, necessitating thicker bases than the 1-4 lm typically used in III-V cells.…”

“…16 Given that high-energy photons should be predominantly absorbed near the surface of the material, the low IQE previously observed at short wavelengths likely results from a combination of low minority carrier diffusion length and high surface recombination velocity. Most previous GaP solar cells were not able to capitalize on carriers generated by absorption at the C valley due to the use of 0.5-1.0 lm n þ emitters, 13,15,[17][18][19] which are much thicker than the 0.1-0.2 lm emitters typical of other III-V devices. [20][21][22][23] The hole diffusion length (L p ) in n þ -GaP has been modeled to be in the range of 110-200 nm, 13,15,17,18 and it can therefore be expected that a thinner emitter would lead to enhanced collection of carriers generated by short-wavelength photons.…”

“…Most previous GaP solar cells were not able to capitalize on carriers generated by absorption at the C valley due to the use of 0.5-1.0 lm n þ emitters, 13,15,[17][18][19] which are much thicker than the 0.1-0.2 lm emitters typical of other III-V devices. [20][21][22][23] The hole diffusion length (L p ) in n þ -GaP has been modeled to be in the range of 110-200 nm, 13,15,17,18 and it can therefore be expected that a thinner emitter would lead to enhanced collection of carriers generated by short-wavelength photons. The considerable challenges with both V OC and J SC motivate further study of both growth conditions and device designs for GaP solar cells.…”

“…The considerable challenges with both V OC and J SC motivate further study of both growth conditions and device designs for GaP solar cells. While GaP cells have been created by a variety of techniques, including diffusion, 11,12 liquid phase epitaxy, 13,15,[17][18][19] organometallic vapor phase epitaxy, 10 and molecular beam epitaxy (MBE), 24,25 no studies on the effect of growth temperature or emitter thickness on cell performance have been reported.…”

“…A high-bandgap window layer is typical in solar cells made from other III-V materials, including InGaP, 6 GaAs, 30,31 and InGaAs 32 to reduce surface recombination and increase minority carrier diffusion length in the emitter. Unlike the few previous reports that incorporated Al-rich window layers into GaP solar cell design, 10,15 we employed a thin 20 nm window with a low 25% Al content in our solar cells to mitigate potential issues with oxidation.…”