Development of High Band Gap Absorber and Buffer Materials for Thin Film Solar Cell Applications

CuInGaSe2 (CIGS) device efficiencies are the highest of the thin film absorber materials (vs. CdTe, alpha-Si, CuInSe2). However, the band gap of the highest efficiency CIGS cells deviates from the expected ideal value predicted by models [1]. Widening the band gap to the theoretically ideal value is one way to increase cell efficiencies. Widening the band gap can be accomplished in two ways; by finding a solution to the Ga-related defects which limit the open circuit voltage at high Ga ratios, or by utilizing different elemental combinations to form an alternative high band gap photoactive Cu-chalcopyrite (which includes any combination of the cations Cu, Al, Ga, and In along with the anions S, Se, and Te). This thesis focuses on the second option, substituting aluminum for gallium in the chalcopyrite lattice to form a CuInAlSe2 (CIAS) film using a sputtering and selenization approach.;Both sequential and co-sputtering of metal precursors is performed. Indium was found to be very mobile during both sputtering processes, with a tendency to diffuse to the film surface even when deposited as the base layer in a sequential sputtering process. Elemental diffusion was controlled to a degree using thicker Cu top layer in co-sputtering. The greater thermal conductivity of stainless steel foil (16 W/mK) vs. glass (0.9-1.3 W/mK) can also be used to limit indium diffusion, by keeping the substrate cooler during sputtering. In both sputtering methods aluminum is deposited oxygen-free by capping the film with a Cu capping layer in combination with controlling the indium diffusion.;Selenization of metal precursor films is completed using two different techniques. The first is a thermal evaporation approach from a heated box source (method 1- reactive thermal evaporation (RTE-Se)). The second is batch selenization using a heated tube furnace (method 2 . batch selenization). Some batch selenized precursors were capped with ∼ 1mum of selenium. In both selenization methods elemental selenium is used at the selenium source.;In method 1 films selenized above 500°C showed low Al incorporation and phase separation. Films selenized with a Se depositional rate of 12 A/s showed poor adhesion compared to samples selenized at 4 A/s. Segregation of aluminum towards the back contact as well as oxygen incorporation appears to cause adhesion loss in extreme cases, and poor interface electrical characteristics in others. The maximum device efficiency measured for method 1 was 5.2% under AM1.5 for a device with ∼ 2 at. % aluminum.;For method 2, samples deposited on glass demonstrated poor adhesion and similar attributes to the RTE-Se samples. No improvements were seen with the additional Se capping layer on the film. Metal foil samples show improved adhesion vs. glass samples deposited under the same conditions. Samples still showed oxidation of aluminum at the Mo interface. Increasing the temperature to 550°C resulted in the loss of Mo adhesion due to excessive MoSe 2 formation. Samples selenized for 90 minutes at 520°C showed decreased adhesion compared to those selenized for 40 minutes. Again excess MoSe 2 growth was seen, though not to the extent of the samples selenized at 550°C.;The effective heat of formation model suggests the low Al incorporation found in all films is due the favorable formation of InSe vs. In2Se3. It predicts InSe will form first at the growing film interface, and this phase is then consumed in the formation of CuInSe2 at temperatures above 220°C. This results in the consumption of In before the (Al,In)2Se 3 phase can form, and therefore minimal formation of the CuInAlSe 2 chalcopyrite phase. Due to minimal (Al,In)2Se3 formation, free Al is able to react with H2O or form Al2Se 3 which is also able to react with H2O. Both reactions result in the formation of Al2O3 and H2Se. This Al2O3 forms a resistive barrier at the back contact, which results in loss of adhesion in high Al films, and an I-V "roll-over" effect in low Al devices.;In addition to CIAS studies, a study of alternative buffer layers to chemical bath deposited cadmium sulfide was completed. Best known processes were used to deposit cadmium sulfide, zinc sulfide, and indium sulfide by chemical bath deposition to examine their basic materials characteristics important to thin film device efficiency. In terms of surface coverage, light transmission, and ease of deposition cadmium sulfide was found be the superior film compared to the alternative sulfides. Future plans to deposit films on CIGS devices should shed more light on the efficiency vs. cost of these alternative buffer layers.