This thesis analyses the world of thin-film solar cells focusing on two kind of technologies,\nthe first based on copper-indium-gallium diselenide, also called CIGS, the\nsecond one on cadmium telluride.\nThe aim of this work in particular is giving suggestions on how to increase the\nconversion efficiency of these devices by improving knowledge and understanding of\ntheir physical behavior.\nThin-film solar cells based on CIGS or CdTe absorbers have been studied in research\nlaboratories for at least 20 years, but only recently entered the phase of industrial\nproduction. While still lagging behind their Si-based competitors in terms\nof efficiency and dollars/Watt, these technologies are poised to conquer increasing\nmarket shares, and perhaps even market dominance. Thin-film solar cells based on\nCu(In,Ga)Se2 (CIGS) show record efficiencies among thin-film technologies, with\nmanufacturers introducing mass production processes yielding cells with efficiencies\nin the 13-15% range. Lab specimens can provide power conversion efficiency as\nhigh as 20%, despite the poly-crystalline structure of the semiconductor thin film.\nSeveral papers were published on material growth, processing and characterization\non one side, and on the performance of finished cells and modules on the other, but\nthere is still a gap to fill in between: some of the specific features of CIGS cells,\nspecifically those pertaining to the behavior of grain boundaries and hetero-junctions,\nare still under debate, and a complete understanding of the relationship between material\ncharacteristics and cell behavior is not available yet.