Accelerating the progress toward realizing metal halide perovskite solar cells with improved efficiency, stability and reliability requires a deeper understanding of the thin film formation processes. This paper investigates the impact of rationally selected chemical additives in precursor solutions on the nucleation and growth of metal halide perovskite thin films. Computational screening was performed to guide the selection of tetrahydrothiophene oxide as an additive with stronger solvation efficacy than all other commonly used solvents. In situ grazing incidence X-ray diffraction measurements show that the additives suppress the formation of homogeneous nuclei as well as crystalline intermediate structures. Instead, heterogeneous nucleation on the substrate surface and growth of a thin film with a strongly preferential crystallographic orientation occur directly from the precursor solution. Density functional theory calculations show that the crystallographic orientation of the thin films can be tuned by altering the surface energies with the chemical additives. The crystallographic orientation of the thin films is found to have a significant impact on the open circuit voltage of solar cell devices, highlighting the importance of controlling the metal halide perovskite thin film orientation for improved solar cell efficiency.
Accelerated searches, made possible by machine learning techniques, are of growing interest in materials discovery. A suitable case involves the solution processing of components that ultimately form thin films of solar cell materials known as hybrid organic–inorganic perovskites (HOIPs). The number of molecular species that combine in solution to form these films constitutes an overwhelmingly large “compositional” space (at times, exceeding 500,000 possible combinations). Selecting a HOIP with desirable characteristics involves choosing different cations, halides, and solvent blends from a diverse palette of options. An unguided search by experimental investigations or molecular simulations is prohibitively expensive. In this work, we propose a Bayesian optimization method that uses an application-specific kernel to overcome challenges where data is scarce, and in which the search space is given by binary variables indicating whether a constituent is present or not. We demonstrate that the proposed approach identifies HOIPs with the targeted maximum intermolecular binding energy between HOIP salt and solvent at considerably lower cost than previous state-of-the-art Bayesian optimization methodology and at a fraction of the time (less than 10%) needed to complete an exhaustive search. We find an optimal composition within 15 ± 10 iterations in a HOIP compositional space containing 72 combinations, and within 31 ± 9 iterations when considering mixed halides (240 combinations). Exhaustive quantum mechanical simulations of all possible combinations were used to validate the optimal prediction from a Bayesian optimization approach. This paper demonstrates the potential of the Bayesian optimization methodology reported here for new materials discovery.
The interlayer stacking of thirty-three two-dimensional covalent organic frameworks (2D COFs) that derive from five boron-containing, “connectors” and seven hydrocarbon “linkers” was evaluated with a Molecular Mechanics (MM) approach, employing the MM3 potential of Allinger et al. Several of these COFs have been synthesized previously, whereas others are proposed structures. This comprehensive study allowed us to classify the nature of COF stacking in terms of potential energy surfaces (PES) and, further, to determine the origin of these surfaces in terms of combinations of repulsive electrostatic and attractive dispersion forces. None of these 2D COFs stack in the exact eclipsed configuration traditionally assigned to these materials; instead, adjacent layers adopt surprisingly similar small offsets (1.5–2.8 Å). For verification, PES of selected, important interlayer offsets were obtained using more accurate Density Functional Theory (DFT) calculations and shown to produce similar offsets to those derived from MM. The symmetry of the 2D COF networks allows the offsets to be accommodated in several degenerate directions, precluding the formation of interlayer stacking patterns with pervasive long-range order, such as staircase, helical, or zigzag arrangements. Although these offsets are often small relative to the size of the COF unit cell, they are sufficient to impede the filling of COF pores with large “guests”, such as fullerenes. Finally, we use a combination of topological, compositional, and interaction parameters to provide a simple correlation that allows researchers to predict inter-layer offsets and interaction energies without performing any molecular simulations. Uncovering such design principles should pave the way to target the creation of COFs with prescribed structures and charge transport properties.