The development of new materials typically requires an iterative sequence of synthesis and characterization, but high-performance computing (HPC) adds another dimension to the process: materials can be synthesized and/or characterized virtually as well, and it is often an overlapping quilt of data from these four aspects of design that is used to develop a new material. This is made possible, in large measure, by the algorithms and hardware collectively referred to as HPC. Prominent within this developing approach to materials design is the increasingly important role that quantum mechanical analysis techniques have come to play. These techniques are reviewed with an emphasis on their application to materials design. This issue of MRS Bulletin highlights specific examples of how such HPC tools are used to advance energy science research in the areas of nuclear fission, electrochemical batteries, photovoltaic energy conversion, hydrocarbon catalysis, hydrogen storage, clathrate hydrates, and nuclear fusion.

The behavior of nuclear fuel in a reactor is a complex phenomenon that is influenced by a large number of materials properties, which include thermomechanical strength, chemical stability, microstructure, and defects. As a consequence, a comprehensive understanding of the fuel material behavior presents a significant modeling challenge, which must be mastered to improve the efficiency and reliability of current nuclear reactors. It is also essential to the development of advanced fuel materials for next-generation reactors. Over the last two decades, the use of density functional theory (DFT) has greatly contributed to our understanding by providing profound information on nuclear fuel materials, ranging from fundamental properties of f-electron systems to thermomechanical materials properties. This article briefly summarizes the main achievements of this first-principles computational methodology as it applies to nuclear fuel materials. Also, the current status of first-principles modeling is discussed, considering existing limitations and drawbacks such as size limitation and the added complexity associated with high temperature analysis. Finally, the future role of DFT modeling in the nuclear fuels industry is put into perspective.

Energy storage is a critical hurdle to the success of many clean energy technologies. Batteries with high energy density, good safety, and low cost can enable more efficient vehicles with electrified drive trains, such as hybrid electric vehicles, electric vehicles, and plug-in hybrid electric vehicles. They can also provide energy storage for intermittent energy sources, such as wind and solar. Today, and for the foreseeable future, rechargeable lithium batteries deliver the highest energy per unit weight or volume at reasonable cost. Many of the important properties of battery materials can be calculated with first-principles methods, making lithium batteries fertile ground for computational materials design. In this article, we review the successes and opportunities in using first-principles computations in the battery field. We also highlight some technical challenges facing the accurate modeling of battery materials.

The drive to make solar energy competitive with conventional energy sources has prompted the investigation of new photoconversion technologies, often referred to as third-generation photovoltaics, which have both lower cost and improved efficiency compared to existing technologies. In that framework, nanostructured materials, such as nanocrystals, nanowires, and nanotubes, occupy a prominent place because of their potential advantages over crystalline or thin-film photovoltaics technologies—high tunability of the bandgap via size control, strong band-edge absorption coefficient, efficient multiple-exciton generation by a single photon, and possibly high up-conversion efficiency. The ability to control the size, shape, composition, and surface termination of nanostructures provides new degrees of freedom that are inaccessible in conventional solar cell architectures. At the same time, the ability to explore this vast configuration space by synthesis and characterization alone is limited, which makes computational interrogation of the electronic and optical properties of nanostructures particularly valuable. In recent years, the convergence of new algorithms and new computational capabilities has made it possible for the first time to perform accurate electronic-structure calculations for large nanostructures. This article reviews recent developments in both semi-empirical and first-principles atomistic electronic structure methods that have led to accurate predictions and to a better understanding of carrier generation, relaxation, and recombination processes in nanostructured materials.

Hydrogen is considered by some to be a promising non-CO2-emitting energy carrier for the future. However, to realize a hydrogen economy, there are several technological barriers to overcome. Currently, safe and efficient storage of hydrogen is a bottleneck in the practical usage of hydrogen for fuels. In this article, we present a review on the first-principles computational approach in designing hydrogen storage materials with an emphasis on molecular hydrogen storage in nanostructured materials. Given the limitation of pristine nanostructures for room-temperature hydrogen storage, the strategy of decorating the backbone structure of the nanostructure with transition metal atoms in order to enhance the hydrogen adsorption energy is addressed, and the interplay between the Coulomb interactions and the so-called Kubas interaction (nondissociative weak chemisorption via electron donation and back-donation channels) has been studied. The influence of electron spin on the hydrogen binding energy, problems of metal clustering and oxidation, and the structural instability that may arise during hydrogen sorption are also discussed. We address the limitations and challenges in the development of high-capacity hydrogen storage materials and provide perspectives for how computational materials design can help cope with those problems.

The energy science of clathrate hydrates is a rapidly expanding field, with high-performance computing (HPC) playing an ever-growing role to help understand the molecular processes and properties that drive clathrate hydrates to nucleate and grow into crystalline, amorphous, or mixed structures, their non-stoichiometric nature upon formation, the formation mechanism from homogeneous and heterogeneous nucleation, and their stability and limits of metastability. Many of the questions that HPC can help to answer about hydrates are intractable experimentally because of the difficulty of measurements at the length (nanometers) and time (nanoseconds) scales imposed by the fundamental phenomena at the molecular level. At the same time, the length and time scales that are accessible by simulations pose limitations on what can be studied (e.g., phase equilibria and metastability, nucleation mechanisms, non-stoichiometry) and how it can be studied (e.g., Monte Carlo, molecular dynamics, metadynamics, transition path sampling, thermodynamic integration). Ultimately, the energy science of clathrate hydrates will benefit from HPC by gaining insight into the detailed mechanism for formation, dissociation, and stability.

Future energy production and storage in the chemical and refinery industries, stationary power generation, and transportation sectors will employ a diverse suite of technologies, including renewables, such as biomass, untapped energy resources, and processes with improved energy efficiency. Heterogeneous nanocatalysts will play an ever-increasing role in these technologies. Increased precision in molecular architecture over multiple length scales and/or tailored multi-functionality will often be needed in these materials. Advances in computational-based discovery of such nanomaterials are described through examples that predict the molecular architecture of emergent catalytic materials and reveal mechanisms of colloidal metal nanoparticle growth.

The plasma facing components, first wall, and blanket systems of future tokamak-based fusion power plants arguably represent the single greatest materials engineering challenge of all time. Indeed, the United States National Academy of Engineering has recently ranked the quest for fusion as one of the top grand challenges for engineering in the 21st century. These challenges are even more pronounced by the lack of experimental testing facilities that replicate the extreme operating environment involving simultaneous high heat and particle fluxes, large time-varying stresses, corrosive chemical environments, and large fluxes of 14-MeV peaked fusion neutrons. Fortunately, recent innovations in computational modeling techniques, increasingly powerful high-performance and massively parallel computing platforms, and improved analytical experimental characterization tools provide the means to develop self-consistent, experimentally validated models of materials performance and degradation in the fusion energy environment. This article will describe the challenges associated with modeling the performance of plasma facing component and structural materials in a fusion materials environment, the opportunities to utilize high-performance computing, and two examples of recent progress.

Energy storage can help integrate wind and solar power into the electric grid.