1c:I[419806,["/_next/static/chunks/bb923814e1f4d2e9.js","/_next/static/chunks/dd3ee9f5b7716ca6.js","/_next/static/chunks/4204a0ae82542ca8.js","/_next/static/chunks/1270dead247da766.js","/_next/static/chunks/6e1aa89a642f40be.js","/_next/static/chunks/2d7aa753e81324a8.js","/_next/static/chunks/8ffd0b71567920d8.js","/_next/static/chunks/67ea6dfd21f16c5f.js","/_next/static/chunks/24eb4379f68013fd.js","/_next/static/chunks/b8cd775ffdf935bd.js","/_next/static/chunks/983103cd9eb7b428.js","/_next/static/chunks/263f2cdcb9aa2341.js","/_next/static/chunks/294b9525248c8e5f.js"],"default"] 1d:T42b,The use of machine learning methods for accelerating the design of crystalline materials usually requires manually constructed feature vectors or complex transformation of atom coordinates to input the crystal structure, which either constrains the model to certain crystal types or makes it difficult to provide chemical insights. Here, we develop a crystal graph convolutional neural networks framework to directly learn material properties from the connection of atoms in the crystal, providing a universal and interpretable representation of crystalline materials. Our method provides a highly accurate prediction of density functional theory calculated properties for eight different properties of crystals with various structure types and compositions after being trained with 10^{4} data points. Further, our framework is interpretable because one can extract the contributions from local chemical environments to global properties. Using an example of perovskites, we show how this information can be utilized to discover empirical rules for materials design.1e:T4da,Theoretical studies of the possibility of forming ammonia electrochemically at ambient temperature and pressure are presented. Density functional theory calculations were used in combination with the computational standard hydrogen electrode to calculate the free energy profile for the reduction of N(2) admolecules and N adatoms on several close-packed and stepped transition metal surfaces in contact with an acidic electrolyte. Trends in the catalytic activity were calculated for a range of transition metal surfaces and applied potentials under the assumption that the activation energy barrier scales with the free energy difference in each elementary step. The most active surfaces, on top of the volcano diagrams, are Mo, Fe, Rh, and Ru, but hydrogen gas formation will be a competing reaction reducing the faradaic efficiency for ammonia production. Since the early transition metal surfaces such as Sc, Y, Ti, and Zr bind N-adatoms more strongly than H-adatoms, a significant production of ammonia compared with hydrogen gas can be expected on those metal electrodes when a bias of -1 V to -1.5 V vs. SHE is applied. Defect-free surfaces of the early transition metals are catalytically more active than their stepped counterparts.1f:T447,Currently, the only compounds known to support fixation and functionalization of dinitrogen (N₂) under nonmatrix conditions are based on metals. Here we present the observation of N₂ binding and reduction by a nonmetal, specifically a dicoordinate borylene. Depending on the reaction conditions under which potassium graphite is introduced as a reductant, N₂ binding to two borylene units results in either neutral (B₂N₂) or dianionic ([B₂N₂]^2-) products that can be interconverted by respective exposure to further reductant or to air. The ¹⁵N isotopologues of the neutral and dianionic molecules were prepared with ¹⁵N-labeled dinitrogen, allowing observation of the nitrogen nuclei by ¹⁵N nuclear magnetic resonance spectroscopy. Protonation of the dianionic compound with distilled water furnishes a diradical product with a central hydrazido B₂N₂H₂ unit. All three products were characterized spectroscopically and crystallographically.

	Year	Citations

References