Atomic Layer Deposition of Thin Film Indium Oxysulfide - A Non-Toxic Electron Transport Layer for Chalcogenide Solar Cells

Abstract

CZT(S,Se) (Cu2ZnSn(SxSe1-x)4) has emerged as an earth-abundant, non-toxic alternative to thin-film photovoltaic technologies based on CIG(S,Se) (Cu(In,Ga)(S,Se)2) and CdTe. Devices employing CZT(S,Se), however, suffer from poor voltage extraction, reaching only 60 % of the Shockley-Queisser limit. The low photovoltage results primarily from interfacial recombination at the absorber-electron conductor junction, owing to mismatch in conduction band energies, lack of conformality, unpassivated defects, and elemental interdiffusion. The best existing heterojunction devices employ CdS as the electron transport layer. We have explored In2(O,S)3 films prepared by atomic layer deposition (ALD) as an electron transport layer for CZT(S,Se). The motivation for this thesis was to tune the bands of In2(O,S)3 by controlling the S:O ratio. This would result in higher collection of photo-generated electrons and higher open circuit voltage on coupling In2(O,S)3 with CZT(S,Se). We initially selected In2S3 on grounds that In3+ ion possibly has lower diffusivity than Cd2+ which should limit elemental interdiffusion. It has been previously established in literature by solution deposition that incorporation of oxygen into In2S3 allows us to obtain band positions at the p-n junction that should enable good minority carrier extraction. Oxygen incorporation also increases the band gap and therefore the optical transparence. Unfortunately, solution-based methods leave a large number of uncontrollable hydroxyl groups in the film, which affords poor control over electrical properties of the film. We now report that atomic layer deposition (ALD) using alternate cycles of tris(N,N'-diisopropylformamidinato) indium(III) (indium formamidinate), water, and hydrogen sulfide at 200 oC results in growth of a pure In2(O,S)3 film (free of halides, carbon) with close control in-situ over sulfur to oxygen ratio. We arrived at conditions for growth of In2(O,S)3 by studying ALD of binaries In2S3 and In2O3 independently. The choice of precursor for growing In2(O,S)3 was made on the basis of kinetics of the growth process of the aforementioned binaries. As we use water as the oxygen source, there is incorporation of adventitious hydroxyl groups. We are able to control the content of hydroxyl groups in the In2(O,S)3 film though, by varying number of water containing sub-cycles. In2(O,S)3 films exhibited an indirect band gap higher than In2S3 which minimizes absorption losses in the electron transport layer. A reasonably high mobility was obtained with wide and tunable range of carrier concentrations over 5 orders of magnitude. To limit recombination, fewer charge carriers are targeted in In2(O,S)3 close to its junction with absorber while more carriers are targeted adjacent the transparent conducting oxide in the solar cell. Band offset measurements of In2(O,S)3 with reference to CZT(S,Se) by x-ray photoelectron spectroscopy indicate that oxygen incorporation (and resultant decrease in S:O ratio) increases conduction band offset relative to the pure sulfide. We identify In2(O,S)3 with oxygen contents between 4-33 at. % as the optimal composition in combination with CZT(S,Se), to potentially boost the open circuit voltage of the solar cells.