A MICROSTRUCTURAL COMPARISON OF Cu(In,Ga)Se2 THIN FILMS GROWN FROM CuxSe AND (In,Ga) 2 Se3 PRECURSORS ANDREW M. GABOR*, J. R. TUTTLE*, D. S. ALBIN*, R. MATSON*, A. FRANZ*, D. W. NILES*, M. A. CONTRERAS*, A. M. HERMANN**, R. NOUFI* *National Renewable Energy Laboratory, 1617 Cole Blvd., Golden, CO 80301 "**University of Colorado, Boulder, CO 80309-0390 ABSTRACT We fabricated CulnSe2 and Cu(In,Ga)Se2 thin films by two different pathways using physical vapor deposition. In the first we formed a Cu-Se precursor and then reacted it with a flux of (In,Ga) + Se. These films had large grains but were too rough for optimal device performance. In the other pathway, we first formed a smooth precursor of (In,Ga) 2 Se3 and then exposed it to a flux of Cu+Se. We overshot the optimal film composition to allow recrystallization of the film by a secondary CuxSe phase. We then consumed the excess CuxSe in a third stage deposition of (In,Ga) + Se. The recrystallization step increased the grain sizes, and the resulting films remained smooth. Photovoltaic solar cells made from these films have produced the highest total-area efficiencies of any non-single-crystal, thin-film solar cell. INTRODUCTION Thin films of CulnSe2 (CIS) and related semiconductor alloys are among the most promising materials for use as light absorbing layers in thin-film photovoltaic (PV) solar cells. Significant advances have been made in cell efficiency recently with the alloying of CIS with the higher band- gap CuGaSe2 (CGS) to form Cu(In,Ga)Se2 (CIGS) [1,2]. This alloying raises the band gap of the absorber to a better match with the solar spectrum, and by varying the Ga content as a function of film depth, one can engineer graded-band-gap structures to aid cell performance. A common method of deposition has been coevaporation from the elemental sources. This has been successful, but a simpler approach where fewer elements are deposited at a time would be attractive from a manufacturing perspective. Selenization of the metallic precursors by H 2 Se gas or Se vapor is potentially scalable, but problems associated with film adhesion and phase separation of Cu-In-Ga intermetallic compounds complicate this chemical pathway. Partly motivated by the difficulties of selenization, we have chosen a different simplifying approach where we prevent Cu-(In,Ga) compound formation by separating the deposition into stages where we coevaporate Cu+Se or In+Ga+Se. In doing this, we can benefit from the properties of the precursor and intermediate or secondary phases. Although we use physical vapor deposition (PVD) as flexible tool to explore different pathways, the results should apply to various manufacturing scenarios. In an open system with sufficient Se activity, compositions of Cu, In, and Se tend to fall on the Cu2Se - In2Se3 pseudobinary tie line. The Cu 2 Se - In2Se3 pseudobinary phase diagram (see Fig. 1) shows that for Cu-rich compositions ([Cu] > [In]), the excess Cu will exist in the Cu 2 Se phase. However, with high enough temperatures and Se activity, this Cu2Se could be converted to a liquid CuxSe phase (x<2). Different groups have taken advantage of the liquid-phase- assisted growth potential in this two-phase region by initiating the deposition with a Cu-rich flux of Cu+In+Ga+Se to form a large-grain precursor [2,4,5]. These groups then introduce a Cu- poor flux to consume the excess CuxSe and grow additional CIGS quasi-epitaxially onto the precursor. The final composition for a good device must be slightly Cu-poor, because any remaining CuxSe will short the device. Well developed models for such growth commencing from the Cu-rich regime have already been presented [5,6]. For this work, we pursued two main paths toward the formation of CIGS. For pathway #1 we took the idea behind the Cu-rich precursor methods to its logical extreme and started with a precursor of only Cu+Se. We then reacted it with a flux of In+Ga+Se to form CIGS. For pathway #2 we started with an (In,Ga)2Se 3 precursor and exposed it to a flux of Cu+Se. In a 143 Mat. Res. Soc. Symp. Proc. Vol. 343. 01994 Materials Research Society