The use of layered perovskites is an important strategy to improve the stability of hybrid perovskite materials and their optoelectronic devices. However, tailoring their properties requires accurate structure determination at the atomic scale, which is a challenge for conventional diffraction-based techniques. We demonstrate the use of nuclear magnetic resonance (NMR) crystallography in determining the structure of layered hybrid perovskites for a mixed-spacer model composed of 2-phenylethylammonium (PEA + ) and 2-(perfluorophenyl)ethylammonium (FEA + ) moieties, revealing nanoscale phase segregation. Moreover, we illustrate the application of this structure in perovskite solar cells with power conversion efficiencies that exceed 21%, accompanied by enhanced operational stability.

Knowledge of the structure of amorphous solids can direct, for example, the optimization of pharmaceutical formulations, but atomic-level structure determination in amorphous molecular solids has so far not been possible. Solid-state nuclear magnetic resonance (NMR) is among the most popular methods to characterize amorphous materials, and molecular dynamics (MD) simulations can help describe the structure of disordered materials. However, directly relating MD to NMR experiments in molecular solids has been out of reach until now because of the large size of these simulations. Here, using a machine learning model of chemical shifts, we determine the atomic-level structure of the hydrated amorphous drug AZD5718 by combining dynamic nuclear polarization-enhanced solid-state NMR experiments with predicted chemical shifts for MD simulations of large systems. From these amorphous structures we then identify H-bonding motifs and relate them to local intermolecular complex formation energies.

The development of magic-angle spinning dynamic nuclear polarization (MAS DNP) has allowed atomic-level characterization of materials for which conventional solid-state NMR is impractical due to the lack of sensitivity. The rapid progress of MAS DNP has been largely enabled through the understanding of rational design concepts for more efficient polarizing agents (PAs). Here, we identify a new design principle which has so far been overlooked. We find that the local geometry around the unpaired electron can change the DNP enhancement by an order of magnitude for two otherwise identical conformers. We present a set of 13 new stable mono-and dinitroxide PAs for MAS DNP NMR where this principle is demonstrated. The radicals are divided into two groups of isomers, named open (O-) and closed (C-), based on the ring conformations in the vicinity of the N−O bond. In all cases, the open conformers exhibit dramatically improved DNP performance as compared to the closed counterparts. In particular, a new urea-based biradical named HydrOPol and a mononitroxide O-MbPyTol yield enhancements of 330 ± 60 and 119 ± 25, respectively, at 9.4 T and 100 K, which are the highest enhancements reported so far in the aqueous solvents used here. We find that while the conformational changes do not significantly affect electron spin−spin distances, they do affect the distribution of the exchange couplings in these biradicals. Electron spin echo envelope modulation (ESEEM) experiments suggest that the improved performance of the open conformers is correlated with higher solvent accessibility.

The increasing urgency to make chemical processes more environmentally friendly while continuing to derive the chemicals required for modern society from renewable resources requires the development of a forthcoming generation of synthetic processes and the catalysts needed to facilitate these reactions. Recently, applications of machine-learning (ML) algorithms involving catalysis have begun to appear with increasing frequency, as they constitute an attractive pathway both for discovering prospective species and identifying trends surrounding catalytic behavior, principally because the number of potential catalysts that can be examined greatly exceeds those found in more traditional experimental or theoretical approaches. Here, we harness a data-driven approach powered by ML in tandem with molecular volcano plots to estimate the activity of over 143,000 homogeneous nickel catalysts bearing phosphine and N-heterocyclic carbene ligands for the reductive C(sp 2 )−O cleavage reaction in aryl ether compounds, an important step in the degradation of biomass (lignin) into industrially useful feedstock chemicals. Our computational workflow reveals that a vast majority of Ni-phosphine and Ni-carbene catalysts are not ideally tuned to facilitate this reaction. An analysis of those species identified as being the most promising uncovers a clear catalytic design strategy that can be exploited in an experimental setting to enhance the rate of reductive C(sp 2 )−O cleavage of aryl ether compounds.

To alleviate the computational cost associated with on-the-fly ab initio semiclassical calculations of molecular spectra, we propose the single-Hessian thawed Gaussian approximation, in which the Hessian of the potential energy at all points along an anharmonic classical trajectory is approximated by a constant matrix. The spectra obtained with this approximation are compared with the exact quantum spectra of a one-dimensional Morse potential and with the experimental spectra of ammonia and quinquethiophene. In all cases, the single-Hessian version performs almost as well as the much more expensive on-the-fly ab initio thawed Gaussian approximation and significantly better than the global harmonic schemes. Remarkably, unlike the thawed Gaussian approximation, the proposed method conserves energy exactly, despite the time dependence of the corresponding effective Hamiltonian, and, in addition, can be mapped to a higher-dimensional time-independent classical Hamiltonian system.We also provide a detailed comparison with several related approximations used for accelerating prefactor calculations in semiclassical simulations.

Determination of the three-dimensional atomic-level structure of powdered solids is one of the key goals in current chemistry. Solid-state NMR chemical shifts can be used to solve this problem, but they are limited by the high computational cost associated with crystal structure prediction methods and density functional theory chemical shift calculations. Here, we successfully determine the crystal structures of ampicillin, piroxicam, cocaine, and two polymorphs of the drug molecule AZD8329 using on-the-fly generated machine-learned isotropic chemical shifts to directly guide a Monte Carlo-based structure determination process starting from a random gas-phase conformation.

Photoinduced halide segregation hinders widespread application of three-dimensional (3D) mixed-halide perovskites. Much less is known about this phenomenon in lower-dimensional systems.Here, we study photoinduced halide segregation in lower-dimensional mixed iodide-bromide perovskites (PEA 2 MA n−1 Pb n (Br x I 1−x ) 3n+1 , with PEA + : phenethylammonium and MA + : methylammonium) through time-dependent photoluminescence (PL) spectroscopy. We show that layered two-dimensional (2D) structures render additional stability against the demixing of halide phases under illumination. We ascribe this behavior to reduced halide mobility due to the intrinsic heterogeneity of 2D mixed-halide perovskites, which we demonstrate via 207 Pb solid-state NMR. However, the dimensionality of the 2D phase is critical in regulating photostability. By tracking the PL of multidimensional perovskite films under illumination, we find that while halide segregation is largely inhibited in 2D perovskites (n = 1), it is not suppressed in quasi-2D phases (n = 2), which display a behavior intermediate between 2D and 3D and a peculiar absence of halide redistribution in the dark that is only induced at higher temperature for the quasi-2D phase.

Machine-learned chemical shifts enable probabilistic assignment in organic solids without knowledge of the structure.