This is an exciting time for the development of solar cell materials. Quantum dot solar cells made from lead chalcogenides (PbS/Se) continue to achieve new levels of efficiency each year, reaching 9% in May 2014. An even newer class of materials, lead halide (PbI/Cl) perovskites, has meanwhile emerged and overtaken PbS quantum dots to become the fastest-advancing type of solar cell material ever. PbI perovskite efficiencies have reached 20%, and startups have arisen proposing that commercialization will occur within only three years.
One reason for this enthusiasm is that the hybrid organic-inorganic nature of these materials, PbS/Se quantum dots and PbI/Cl perovskites, allows them to be solution-processed at low temperature. This potentially makes their energy and capital requirements much lower than those of purely inorganic materials such as silicon. But the hybrid organic-inorganic character of these materials also hinders the research which is necessary before these materials fulfill their promise. The most well-tried models for molecular interactions are either organic (OPLS) or inorganic (Stillinger-Weber), but not both. The models that can handle both types, such as REAX, are extremely complex as a result, and difficult to adapt to a many-element system such as PbS/Se nanocrystals or PbI/Cl perovskites. This undermines our ability to study the early stages of nucleation of these materials from the solution, which is critical to successful solution processing.
We have created an all-atom reactive molecular model for lead/organic hybrid systems. Our model uses a combination of Morse potentials and Coulombic charges to make efficient, easily parameterized representation of the complex inorganic interactions. This form merges with other popular force fields for organic molecules, allowing the hybrid system to include any of the vast array of available organic ligands. Using such a force field for the quantum dot system, we show how the quantum dot synthesis proceeds in a manner similar to polymerization, and why this process tends to produce dots of uniform size. For the colloidal perovskite system, we study how the colloidal particles join at the surface to form a complete layer. We also model two surfaces being pushed together and laminating, a process potentially important in perovskite processing. Finally, we show atomic interdiffusion at the heterojunction and ion diffusion across the interface, determining the free energy barriers to transport, which have never before been accessible to simulation or experiment. By providing insight into the earliest phases of the nucleation and growth of quantum dots and perovskites, this research will allow experimenters to select optimal conditions and increase the robustness and scale of their techniques. This enhanced understanding of the atomic-level structure may greatly assist the development of these exciting materials.