Specialized computational chemistry packages have permanently reshaped the landscape of chemical and materials science by providing tools to support and guide experimental efforts and for the prediction of atomistic and electronic properties. In this regard, electronic structure packages have played a special role by using first-principle-driven methodologies to model complex chemical and materials processes. Over the past few decades, the rapid development of computing technologies and the tremendous increase in computational power have offered a unique chance to study complex transformations using sophisticated and predictive many-body techniques that describe correlated behavior of electrons in molecular and condensed phase systems at different levels of theory. In enabling these simulations, novel parallel algorithms have been able to take advantage of computational resources to address the polynomial scaling of electronic structure methods. In this paper, we briefly review the NWChem computational chemistry suite, including its history, design principles, parallel tools, current capabilities, outreach, and outlook.

Optimizing proton conduction in solids remains the most promising solution for achieving intermediate temperature (∼750−1000 K) solid oxide fuel cell devices, and enabling selective membranes for H 2 separation. Proton conduction, a thermally activated process, exhibits its highest rates in yttrium (Y) acceptor doped BaZrO 3 at an optimal doping level of 20% Y. The presence of extended defects such as grain boundaries has typically generated a wide variability in reported conductivity values. This has hindered a fundamental mechanistic understanding of how (acceptor) doping levels correlate with the activation energy of protons to produce an optimal doping level for fast proton transport. While isolated dopants have been suggested as the primary source of proton trapping, our results indicate that it is the local dopantdensity that matters. Here, we show that increasing the local dopant density promotes localized lattice distortions in the presence of point defects such as oxygen-vacancies or proton interstitials. An increasing distortion amplitude traps the point defects more strongly in the form of polarons, forming defect-clusters at higher concentrations. This leads to a monotonic increase in the activation energy (and hence a decrease in proton mobility) as observed in our measurements. The optimum doping level can now be explained as a competition between increasing proton concentration with doping levels and increasing activation energy due to defect-clusters formed by defect-polarons. Based on our findings, we demonstrate how to improve proton conductivity in doped BaZrO 3 , by inhibiting this dopant-lattice polaronic interaction. This approach should be generally applicable for ionic conduction in perovskite oxides such as oxygen-ion conduction in solid-oxide fuel cells and alkali-ion conduction in solid-state batteries where carriers might get trapped as defect-polarons.

Heterogeneities such as point defects, inherent to material systems, can profoundly influence material functionalities critical for numerous energy applications. This influence in principle can be identified and quantified through development of large defect data sets which we call the defect genome, employing high-throughput ab initio calculations. However, high-throughput screening of material models with point defects dramatically increases the computational complexity and chemical search space, creating major impediments toward developing a defect genome. In this work, we overcome these impediments by employing computationally tractable ab initio models driven by highly scalable workflows, to study formation and interaction of various point defects (e.g., O vacancies, H interstitials, and Y substitutional dopant), in over 80 cubic perovskites, for potential proton-conducting ceramic fuel cell (PCFC) applications. The resulting defect data sets identify several promising perovskite compounds that can exhibit high proton conductivity. Furthermore, the data sets also enable us to identify and explain, insightful and novel correlations among defect energies, material identities, and defect-induced local structural distortions. Such defect data sets and resultant correlations are necessary to build statistical machine learning models, which are required to accelerate discovery of new materials.