Search

Machine learning (ML) is gaining popularity as a tool for materials scientists to accelerate computation, automate data analysis, and predict materials properties. The representation of input material features is critical to the accuracy, interpretability, and generalizability of data-driven models for scientific research. In this Perspective, we discuss a few central challenges faced by ML practitioners in developing meaningful representations, including handling the complexity of real-world industry-relevant materials, combining theory and experimental data sources, and describing scientific phenomena across timescales and length scales. We present several promising directions for future research: devising representations of varied experimental conditions and observations, the need to find ways to integrate machine learning into laboratory practices, and making multi-scale informatics toolkits to bridge the gaps between atoms, materials, and devices.

We describe a single crystal X-ray diffraction study and computational analysis of three guanidinium (Gua) based low-dimensional iodoplumbates with one edge-sharing and two corner-sharing octahedral connectivities, respectively. (Gua)3PbI5, which is reported for the first time, has a 1D corner-sharing octahedral chain structure. GuaPbI3 adopts a 1D edge-sharing octahedral chain structure in preference to structures that are either 3D and corner-sharing (i.e., perovskite) or 1D and face-sharing. (Gua)2PbI4 exhibits 2D corner-sharing octahedral connectivity in agreement with previous work. Density functional theory calculations are used to gain insight into the relative stabilities of the three polymorphs of GuaPbI3 and to assess how the connectivity and dimensionality of the octahedral framework influence the electronic structure of each of the hybrid perovskites studied.

Organic-inorganic halide perovskites, especially methylammonium lead halide, have recently led to remarkable advances in photovoltaic devices. However, due to environmental and stability concerns around the use of lead, research into lead-free perovskite structures has been attracting increasing attention. In this study, a layered perovskite-like architecture, (NH4)3Bi2I9, is prepared from solution and the structure solved by single crystal X-ray diffraction. The band gap, which is estimated to be 2.04 eV using UV-visible spectroscopy, is lower than that of CH3NH3PbBr3. The energy-minimized structure obtained from first principles calculations is in excellent agreement with the X-ray results and establishes the locations of the hydrogen atoms. The calculations also point to a significant lone pair effect on the bismuth ion. Single crystal and powder conductivity measurements are performed to examine the potential application of (NH4)3Bi2I9 as an alternative to the lead containing perovskites.