The next generation solar cell materials have attracted tremendous research to improve their performance. In these materials, chalcogenides materials, inorganic perovskite and newly developed organometal halide perovskite have demonstrated their potential usage as solar cells owing to their exceptional properties to absorb the light and transform the light energy to current. Hence, understanding and improving these properties can promote further material design strategies for higher performance but lower the cost. Density functional theory is a widely used accurate calculation method to compute various physical properties of a material in an efficient way. In this thesis, we mainly use the density functional theory method to explore the light-matter interaction and its effect to the material's application as a solar cell. Alkali-metal chalcogenides have been found to exhibit appropriate band gaps for solar cells. We find that the volume compression can substantially enhance the optical dielectric function and the absorption coefficient intrinsically. The density function calculation and the tight-binding model show that this structure-property relation is mainly owing to the wavefunction phase change by compression, where the one-dimensional atomic chains play a significant role to relate the optical absorption and the structural change. But the high absorption does not guarantee high power conversion efficiency. This is because the excited carrier need to diffuse to the electrodes before they recombine. Organometal halide perovskites are found to have very large diffusion length and the long carrier lifetime. But the mechanism for such phenomena is still unknown. Here, by studying the structural change to the band structure and spin using CH3NH3PbI3 as an example, we find that the strong Rashba effect contributes to the long carrier lifetime by creating spin-forbidden electronic transitions, which slows down the radiative recombination and enhance the carrier lifetime. Furthermore, to study the spatial disorder effect to the electronic structure, we develop a large-scale tight-binding model which can highlight the structural disorder but still compute the band structure efficiency for very large systems. We find that the spatial disorder can create localized changes. These charge localization are spatially separated for valence band minimum and conduction band maximum. Therefore, their recombination will be further slowed down due to such spatial separation. In addition to these solar cell mechanism, we also studied the non-linear optical effect (bulk photovoltaic effect) in inorganic semiconductors. In this thesis, I use the example of CH3NH3PbI3 to illustrate its bulk photovoltaic effect responses. It is found that this material can generate more than three times large photo-current than the prototypical material BiFeO3, although its polarization is only less than one tenth of BiFeO3. We think this is due to its delocalized electronic structure of the band edges. The effect of Cl to the bulk photovoltaic response is also studied, we find that the apical substitution of I to Cl can enhance the response owing to the larger polarization. The bulk photovoltaic response of other materials such as LiAsSe2, BiFeO3 are compared, and we generalize the strategies to design new materials with better performance.