In this work we target on accurately predicting energy conversion efficiency of dye-sensitized solar cells (DSC) using parameter-free first principles simulations. We present a set of algorithms, mostly based on solo first principles calculations within the framework of density functional theory, to accurately calculate key properties in energy conversion including sunlight absorption, electron injection, electron–hole recombination, open circuit voltages, and so on. We choose two series of donor-π-acceptor dyes with detailed experimental photovoltaic data as prototype examples to show how these algorithms work. Key parameters experimentally measured for DSC devices can be nicely reproduced by first-principles with as less empirical inputs as possible. For instance, short circuit current of model dyes can be well reproduced by precisely calculating their absorption spectra and charge separation/recombination rates. Open circuit voltages are evaluated through interface band offsets, namely, the difference between the Fermi level of electrons in TiO2 and the redox potential of the electrolyte, after modification with empirical formulas. In these procedures the critical photoelectron injection and recombination dynamics are calculated by real-time excited state electronic dynamics simulations. Estimated solar cell efficiency reproduces corresponding experimental values, with errors usually below 1–2%. Device characteristics such as light harvesting efficiency, incident photon-to-electron conversion efficiency, and the current–voltage characteristics can also be well reproduced and compared with experiment. Thus, we develop a systematic ab initio approach to predict solar cell efficiency and photovoltaic performance of DSC, which enables large-scale efficient dye screening and optimization through high-throughput first principles calculations with only a few parameters taken from experimental settings for electrode and electrolyte toward a renewable energy based society.
Read full abstract