Research data infrastructure for high-throughput experimental materials science

A method to computationally screen for tunable properties of crystalline alloys

Conventionally, high-throughput computational materials searches start from an input set of bulk compounds extracted from material databases, but, in contrast, many real functional materials are heavily engineered mixtures of compounds rather than single bulk compounds. We present a framework and open-source code to automatically construct and analyze possible alloys and solid solutions from a set of existing experimental or calculated ordered compounds, without requiring additional metadata except crystal structure. As a demonstration, we apply this framework to all compounds in the Materials Project to create a new, publicly available database of > 600,000 unique "alloy pair" entries that can be used to search for materials with tunable properties. We exemplify this approach by searching for transparent conductors and reveal candidates that might have been excluded in a traditional screening. This work lays a foundation from which materials databases can go beyond stoichiometric compounds and approach a more realistic description of compositionally tunable materials.

The Materials Provenance Store

We present a database resulting from high throughput experimentation, primarily on metal oxide solid state materials. The central relational database, the Materials Provenance Store (MPS), manages the metadata and experimental provenance from acquisition of raw materials, through synthesis, to a broad range of materials characterization techniques. Given the primary research goal of materials discovery of solar fuels materials, many of the characterization experiments involve electrochemistry, along with optical, structural, and compositional characterizations. The MPS is populated with all information required for executing common data queries, which typically do not involve direct query of raw data. The result is a database file that can be distributed to users so that they can independently execute queries and subsequently download the data of interest. We propose this strategy as an approach to manage the highly heterogeneous and distributed data that arises from materials science experiments, as demonstrated by the management of over 30 million experiments run on over 12 million samples in the present MPS release.

Is Cu3-x P a Semiconductor, a Metal, or a Semimetal?

Despite the recent surge in interest in Cu_3-x P for catalysis, batteries, and plasmonics, the electronic nature of Cu_3-x P remains unclear. Some studies have shown evidence of semiconducting behavior, whereas others have argued that Cu_3-x P is a metallic compound. Here, we attempt to resolve this dilemma on the basis of combinatorial thin-film experiments, electronic structure calculations, and semiclassical Boltzmann transport theory. We find strong evidence that stoichiometric, defect-free Cu₃P is an intrinsic semimetal, i.e., a material with a small overlap between the valence and the conduction band. On the other hand, experimentally realizable Cu_3-x P films are always p-type semimetals natively doped by copper vacancies regardless of x. It is not implausible that Cu_3-x P samples with very small characteristic sizes (such as small nanoparticles) are semiconductors due to quantum confinement effects that result in the opening of a band gap. We observe high hole mobilities (276 cm²/(V s)) in Cu_3-x P films at low temperatures, pointing to low ionized impurity scattering rates in spite of a high doping density. We report an optical effect equivalent to the Burstein-Moss shift, and we assign an infrared absorption peak to bulk interband transitions rather than to a surface plasmon resonance. From a materials processing perspective, this study demonstrates the suitability of reactive sputter deposition for detailed high-throughput studies of emerging metal phosphides.

High-Throughput Selection and Experimental Realization of Two New Ce-Based Nitride Perovskites: CeMoN3 and CeWN3

Nitride perovskites have only been experimentally realized in very few cases despite the widespread existence and commercial importance of perovskite materials. From oxide perovskites used in ultrasonics to halide perovskites that have revolutionized the photovoltaics industry, the discovery of new perovskite materials has historically impacted a wide number of fields. Here, we add two new perovskites, CeWN₃ and CeMoN₃, to the list of experimentally realized perovskite nitrides using high-throughput computational screening and subsequent high-throughput thin film growth techniques. Candidate compositions are first down-selected using a tolerance factor and then thermochemical stability. A novel competing fluorite-family phase is identified for both material systems, which we hypothesize is a transient intermediate phase that crystallizes during the evolution from an amorphous material to a stable perovskite. Different processing routes to overcome the competing fluorite phase and obtain phase-pure nitride perovskites are demonstrated for the CeMoN_3-x and CeWN_3-x material systems, which provide a starting point for the development of future nitride perovskites. Additionally, we find that these new perovskite phases have interesting low-temperature magnetic behavior: CeMoN_3-x orders antiferromagnetically below T _N ≈ 8 K with indications of strong magnetic frustration, while CeWN_3-x exhibits no long-range order down to T = 2 K but has strong antiferromagnetic correlations. This work demonstrates the importance and effectiveness of using high-throughput techniques, both computational and experimental: they are integral to optimize the process of realizing two entirely novel nitride perovskites.

Zinc Titanium Nitride Semiconductor toward Durable Photoelectrochemical Applications

Photoelectrochemical fuel generation is a promising route to sustainable liquid fuels produced from water and captured carbon dioxide with sunlight as the energy input. Development of these technologies requires photoelectrode materials that are both photocatalytically active and operationally stable in harsh oxidative and/or reductive electrochemical environments. Such photocatalysts can be discovered based on co-design principles, wherein design for stability is based on the propensity for the photocatalyst to self-passivate under operating conditions and design for photoactivity is based on the ability to integrate the photocatalyst with established semiconductor substrates. Here, we report on the synthesis and characterization of zinc titanium nitride (ZnTiN₂) that follows these design rules by having a wurtzite-derived crystal structure and showing self-passivating surface oxides created by electrochemical polarization. The sputtered ZnTiN₂ thin films have optical absorption onsets below 2 eV and n-type electrical conduction of 3 S/cm. The band gap of this material is reduced from the 3.36 eV theoretical value by cation-site disorder, and the impact of cation antisites on the band structure of ZnTiN₂ is explored using density functional theory. Under electrochemical polarization, the ZnTiN₂ surfaces have TiO₂- or ZnO-like character, consistent with Materials Project Pourbaix calculations predicting the formation of stable solid phases under near-neutral pH. These results show that ZnTiN₂ is a promising candidate for photoelectrochemical liquid fuel generation and demonstrate a new materials design approach to other photoelectrodes with self-passivating native operational surface chemistry.

Crystallize It before It Diffuses: Kinetic Stabilization of Thin-Film Phosphorus-Rich Semiconductor CuP2

Numerous phosphorus-rich metal phosphides containing both P–P bonds and metal–P bonds are known from the solid-state chemistry literature. A method to grow these materials in thin-film form would be desirable, as thin films are required in many applications and they are an ideal platform for high-throughput studies. In addition, the high density and smooth surfaces achievable in thin films are a significant advantage for characterization of transport and optical properties. Despite these benefits, there is hardly any published work on even the simplest binary phosphorus-rich phosphide films. Here, we demonstrate growth of single-phase CuP₂ films by a two-step process involving reactive sputtering of amorphous CuP_2+x and rapid annealing in an inert atmosphere. At the crystallization temperature, CuP₂ is thermodynamically unstable with respect to Cu₃P and P₄. However, CuP₂ can be stabilized if the amorphous precursors are mixed on the atomic scale and are sufficiently close to the desired composition (neither too P poor nor too P rich). Fast formation of polycrystalline CuP₂, combined with a short annealing time, makes it possible to bypass the diffusion processes responsible for decomposition. We find that thin-film CuP₂ is a 1.5 eV band gap semiconductor with interesting properties, such as a high optical absorption coefficient (above 10⁵ cm^–1), low thermal conductivity (1.1 W/(K m)), and composition-insensitive electrical conductivity (around 1 S/cm). We anticipate that our processing route can be extended to other phosphorus-rich phosphides that are still awaiting thin-film synthesis and will lead to a more complete understanding of these materials and of their potential applications.

Materials Data toward Machine Learning: Advances and Challenges

Machine learning (ML) is believed to have enabled a paradigm shift in materials research, and in practice, ML has demonstrated its power in speeding up the cost-efficient discovery of new materials and autonomizing materials laboratories. In this Perspective, current research progress in materials data which are the backbones of ML are reviewed, focusing on high-throughput data generation, standardized data storage, and data representation. More importantly, the challenging issues in materials data that should be overcome to unlock the full potential of ML in materials research and development, including classic 5V (volume, velocity, variety, veracity, and value) issues, 3M (multicomponent, multiscale, and multistage) challenges, co-mining of experimental and computational data, and materials data toward transferable/explainable ML or causal ML, are discussed.

Addressing the critical need for open experimental databases in materials science

With the HTEM, an open online database containing experimental synthesis and characterization data of thin film inorganic materials, Talley et al. (2021) lay a foundation for a new era of high-throughput materials design.