The design of new materials for more efficient production and use of sustainable and clean energy is of utmost importance for the standard of living all over the World the coming years. Efficient utilization of solar energy can take many forms including transforming the energy of the light into heat, electrical energy, or fuels.

In the talk I shall describe computational efforts to idenfity or design new materials for efficient light absorption with particular focus on light-induced splitting of water but also with an eye towards other light absorbing devices like PV cells. The materials have to obey a number of criteria in order to work for water splitting depending on the particular design of the device. We consider in particular stability, including corrosion resistance in water, appropriate bandgap and bandstructure for visible light absorption, |and an adequate line-up of band edges to the water redox potential.

We have considered several classes of materials with most emphasis on the cubic perovskite structure and derivatives like double perovskites and layered perovskites (Ruddlesden-Popper and Dion-Jacobson phases) with 52 different metallic elements and different anions (O, N, S, F). Also a range of Sn and Pb based organic and inorganic perovskites have been considered with different combinations of the anions I, Br, and Cl.

Furthermore the screening has been expanded to 2400 materials form the ICSD database (http://icsd.fiz-karlsruhe.de) included in the Materials Project database (https://materialsproject.org/). These materials are thus known to exist in Nature but many of their properties like their bandgaps are not known from experiment. The systems are used to evaluate the possibilities for finding materials for different tandem solar cell designs with or without water protection layers.

The possibility of tuning band gaps by atomically well-defined stacking of perovskite layers is illustrated by results for compounds composed of BaSnO3 and BaTaO2N. It is shown that the band gap can vary by about 1 eV and that the gap formation can be understood based on electronic confinement and tunneling. As expected the light absorption efficiency depends strongly on the character of the band-edge states.

Finally, we consider band gap tuning by applying strain to different Sn-containing semiconductors. For some systems the gap decreases dramatically with tensile strain while for others the behavior is more modest and can have either sign. The behavior can be understood from role played in the electronic structure by the Sn s-states.