Solar Cells using Congruent Thermal Evaporation

Although highly valuable as an R&D tool, ALD is a rather slow and thus expensive fabrication method for growing fi lms on the order of 1 μm thick, posing a challenge to industrial scaleup. Hence, there is a need to transfer learnings from ALD to higher-throughput manufacturing techniques without compromising quality. We fi rst demonstrate the ease of phase purifi cation of commercial SnS powder by comparing the X-ray diffraction (XRD) patterns of as-purchased powder, post-annealed powder, and an evaporated thin fi lm ( Figure 1 a). The purchased SnS powder (Alfa Aesar, batch # Lot K17U030) is nominally 99.5% pure Sn and S, i.e., not distinguishing between SnS and the phase impurities Sn 2 S 3 and SnS 2 . In the XRD pattern of the purchased powder we detect signifi cant quantities of the sulfur-rich phases Sn 2 S 3 (55% by weight) and SnS 2 (2%) in addition to the desired SnS phase (43%) (Figure 1 a, black line). We anneal this precursor powder in a dedicated tube furnace under vacuum (approximately 15 mTorr) at 500 °C for 60 minutes to eliminate high vapour pressure contaminants and to achieve phase purity (Figure 1 a, red line). The annealed SnS powder is then transferred into the deposition chamber for SnS fi lm growth by congruent evaporation (Figure 1 a, dark yellow line). Our XRD data indicate a substantial improvement in phase purity upon annealing the as-purchased powder. We see no peaks from the Sn 2 S 3 or SnS 2 phases for the annealed powder, and the sensitivity of our measurement puts an upper limit of 0.2% on Sn 2 S 3 phase fraction. The schematics on the right (Figure 1 b through 1 d) illustrate the different SnS feedstock conditions in line with the XRD spectra. Without treatment of the as-purchased powder, thermal evaporation of powder with sulfur-rich phase impurities would also yield a sulfur-rich SnS fi lm. We demonstrate the congruent evaporation process by comparing our deposition rates as a function of source temperature ( T ) to the published equilibrium pressure of SnS(g) over SnS(s). We calculate an equivalent SnS(g) vapour pressure from measured deposition rates using the Langmuir equation and assuming that the coeffi cient of evaporation α e = 1. [ 15 ] The agreement between our data and published equilibrium pressures of SnS(g) over SnS(s) shown in Figure 2 demonstrates deposition by congruent evaporation. The systematic difference between our data and the equilibrium SnS(g) vapour pressure The demand for low-cost and scalable renewable energy continues to spur research in thin-fi lm photovoltaics (PV), in particular those with chalcogenide semiconductor absorbers, which can combine high power conversion effi ciencies (PCE) with effi cient materials utilization. Solar cells based on cadmium telluride (CdTe) have achieved PCEs up to 20.4% and are produced at gigawatt-per-year (GW/yr) levels, [ 1 ] while PV cells based on copper (indium, gallium) (diselenide, disulfi de) (CIGS) reach up to 20.9% PCE and are expected to enter GW/ yr-level production in the near future. [ 2 ]