Modeling materials and processes in hybrid/organic photovoltaics: from dye-sensitized to perovskite solar cells.

Abstract

CONSPECTUS: Over the last 2 decades, researchers have invested enormous research effort into hybrid/organic photovoltaics, leading to the recent launch of the first commercial products that use this technology. Dye-sensitized solar cells (DSCs) have shown clear advantages over competing technologies. The top certified efficiency of DSCs exceeds 11%, and the laboratory-cell efficiency is greater than 13%. In 2012, the first reports of high efficiency solid-state DSCs based on organohalide lead perovskites completely revolutionized the field. These materials are used as light absorbers in DSCs and as light-harvesting materials and electron conductors in meso-superstructured and flat heterojunction solar cells and show certified efficiencies that exceed 17%. To effectively compete with conventional photovoltaics, emerging technologies such as DSCs need to achieve higher efficiency and stability, while maintaining low production costs. Many of the advances in the DSC field have relied on the computational design and screening of new materials, with researchers examining material characteristics that can improve device performance or stability. Suitable modeling strategies allow researchers to observe the otherwise inaccessible but crucial heterointerfaces that control the operation of DSCs, offering the opportunity to develop new and more efficient materials and optimize processes. In this Account, we present a unified view of recent computational modeling research examining DSCs, illustrating how the principles and simulation tools used for these systems can also be adapted to study the emerging field of perovskite solar cells. Researchers have widely applied first-principles modeling to the DSC field and, more recently, to perovskite-based solar cells. DFT/TDDFT methods provide the basic framework to describe most of the desired materials and interfacial properties, and Car-Parrinello molecular dynamics allow researchers the further ability to sample local minima and dynamical fluctuations at finite temperatures. However, conventional DFT/TDDFT has some limitations, which can be overcome in part by tailored solutions or using many body perturbation theory within the GW approach, which is however more computationally intensive. Relativistic effects, such as spin-orbit coupling, are also included in simulations since they are fundamental for addressing systems that contain heavy atoms. We illustrate the performance of the proposed simulation toolbox along with the fundamental modeling strategies using selected examples of relevant isolated device constituents, including dye and perovskite absorbers, metal-oxide surfaces and nanoparticles, and hole transporters. We critically assess the accuracy of various computational approaches against the related experimental data. We analyze the representative interfaces that control the operational mechanism of the devices, including dye-sensitized TiO2/hole transporter and organohalide lead perovskite/TiO2, and the results reveal fundamental aspects of the device's operational mechanism. Although the modeling of DSCs is relatively mature, the recent "perovskite storm" has presented new problems and new modeling challenges, such as understanding exciton formation and dissociation at interfaces and carrier recombination in these materials.