A fast, robust pipeline for strain mapping of crystalline materials is important for many technological applications. Scanning electron nanodiffraction allows us to calculate strain maps with high accuracy and spatial resolutions, but this technique is limited when the electron beam undergoes multiple scattering. Deep-learning methods have the potential to invert these complex signals, but require a large number of training examples. We implement a Fourier space, complex-valued deep-neural network, FCU-Net, to invert highly nonlinear electron diffraction patterns into the corresponding quantitative structure factor images. FCU-Net was trained using over 200,000 unique simulated dynamical diffraction patterns from different combinations of crystal structures, orientations, thicknesses, and microscope parameters, which are augmented with experimental artifacts. We evaluated FCU-Net against simulated and experimental datasets, where it substantially outperforms conventional analysis methods. Our code, models, and training library are open-source and may be adapted to different diffraction measurement problems.

Hybrid organic-inorganic perovskites are emerging semiconductors for cheap and efficient photovoltaics and light-emitting devices. Different from conventional inorganic semiconductors, hybrid perovskites consist of coexisting organic and inorganic sub-lattices, which present disparate atomic masses and bond strengths. The nanoscopic interpenetration of these disparate components, which lack strong electronic and vibrational coupling, presents fundamental challenges to the understanding of charge and heat dissipation. Here we study phonon population and equilibration processes in methylammonium lead iodide (MAPbI₃) by transiently probing the vibrational modes of the organic sub-lattice following above-bandgap optical excitation. We observe inter-sub-lattice thermal equilibration on timescales ranging from hundreds of picoseconds to a couple of nanoseconds. As supported by a two-temperature model based on first-principles calculations, the slow thermal equilibration is attributable to the sequential phonon populations of the inorganic and organic sub-lattices, respectively. The observed long-lasting thermal non-equilibrium offers insights into thermal transport and heat management of the emergent hybrid material class.

Oxygen redox at high voltage has emerged as a transformative paradigm for high-energy battery cathodes such as layered transition-metal oxides by offering extra capacity beyond conventional transition-metal redox. However, these cathodes suffer from voltage hysteresis, voltage fade and capacity drop upon cycling. Single-crystalline cathodes have recently shown some improvements, but these challenges remain. Here we reveal the fundamental origin of oxygen redox instability to be from the domain boundaries that are present in single-crystalline cathode particles. By investigating single-crystalline cathodes with different domain boundaries structures, we show that the elimination of domain boundaries enhances the reversible lattice oxygen redox while inhibiting the irreversible oxygen release. This leads to significantly suppressed structural degradation and improved mechanical integrity during battery cycling and abuse heating. The robust oxygen redox enabled through domain boundary control provides practical opportunities towards high-energy, long-cycling, safe batteries.