Facilitating the application of machine learning (ML) to materials science problems requires enhancing the data ecosystem to enable discovery and collection of data from many sources, automated dissemination of new data across the ecosystem, and the connecting of data with materials-specific ML models. Here, we present two projects, the Materials Data Facility (MDF) and the Data and Learning Hub for Science (DLHub), that address these needs. We use examples to show how MDF and DLHub capabilities can be leveraged to link data with ML models and how users can access those capabilities through web and programmatic interfaces.

Synthetic biology combines engineering and biology to produce artificial systems with programmable features. Specifically, engineered microenvironments have advanced immensely over the past few decades, owing in part to the merging of materials with biologic mimetic structures. In this review, the authors adapt a traditional definition of community ecology to describe “cellular ecology,” or the study of the distribution of cell populations and interactions within their microenvironment. The authors discuss two exemplar hydrogel platforms: (1) self-assembling peptide hydrogels and (2) poly(ethylene) glycol hydrogels and describe future opportunities for merging smart material design and synthetic biology within the scope of multicellular platforms.

The growing application of data-driven analytics in materials science has led to the rise of materials informatics. Within the arena of data analytics, deep learning has emerged as a game-changing technique in the last few years, enabling numerous real-world applications, such as self-driving cars. In this paper, the authors present an overview of deep learning, its advantages, challenges, and recent applications on different types of materials data. The increasingly availability of materials databases and big data in general, along with groundbreaking advances in deep learning offers a lot of promise to accelerate the discovery, design, and deployment of next-generation materials.

With rapid growth of human population and decreasing labefaction of our environment, the usable fresh water is facing severe pollution and global shortage. Bio-inspired engineering and biotemplate-directed engineering thus offer great promise in clean water generation, including desalination, decontamination, and disinfection. This perspective begins with an introduction of solar energy-based interfacial evaporation system inspired by the natural systems of organisms, and then provides a review of the development and recent progress of the interfacial evaporation system for clean water generation. The long-term outlook in this field of clean water generation using bio-inspired interfacial systems is also discussed.

Immune cells can be genetically engineered with a synthetic chimeric antigen receptor (CAR) to eliminate cancer cells, but clinical efficacy in solid tumors has been disappointing due in part to the immunosuppressive tumor microenvironment (TME). Additionally, the cost and logistical issues of personalized medicine necessitate the creation of an off-the-shelf CAR therapy. Synthetic biology tools were implemented in addressing these problems: an anti-mesothelin CAR, membrane-bound IL-15/IL-15Rα complex, and inducible caspase 9 “kill switch” were expressed in natural killer cells for tumor-targeting capabilities, immunostimulatory effects, and safety in treating a preclinical model of ovarian cancer with a renewable, allogenic cell therapy.

The authors showcase the potential of symbolic regression as an analytic method for use in materials research. First, the authors briefly describe the current state-of-the-art method, genetic programming-based symbolic regression (GPSR), and recent advances in symbolic regression techniques. Next, the authors discuss industrial applications of symbolic regression and its potential applications in materials science. The authors then present two GPSR use-cases: formulating a transformation kinetics law and showing the learning scheme discovers the well-known Johnson–Mehl–Avrami–Kolmogorov form, and learning the Landau free energy functional form for the displacive tilt transition in perovskite LaNiO3. Finally, the authors propose that symbolic regression techniques should be considered by materials scientists as an alternative to other machine learning-based regression models for learning from data.

This paper demonstrates the application of Natural Language Processing (NLP) tools to explore large libraries of documents and to correlate heuristic associations between text descriptions in figure captions with interpretations of images and figures. The use of visualization tools based on NLP methods permits one to quickly assess the extent of the research described in the literature related to a specific topic. The authors demonstrate how the use of NLP methods on only the figure captions without having to navigate the entire text of a document can provide an accelerated assessment of the literature in a given domain.

The recent observations of ferromagnetic order in several two-dimensional (2D) materials have generated an enormous interest in the physical mechanisms underlying 2D magnetism. In the present Prospective Article, we show that Density Functional Theory combined with either classical Monte Carlo simulations or renormalized spin-wave theory can predict Curie temperatures for ferromagnetic insulators that are in quantitative agreement with experiments. The case of materials with in-plane anisotropy is then discussed, and it is argued that finite size effects may lead to observable magnetic order in macroscopic samples even if long range magnetic order is forbidden by the Mermin–Wagner theorem.

Protein-based materials are a powerful instrument for a new generation of biological materials, with many chemical and mechanical capabilities. Through the manipulation of DNA, researchers can design proteins at the molecular level, engineering a vast array of structural building blocks. However, our capability to rationally design and predict the properties of such materials is limited by the vastness of possible sequence space. Directed evolution has emerged as a powerful tool to improve biological systems through mutation and selection, presenting another avenue to produce novel protein materials. In this prospective review, we discuss the application of directed evolution for protein materials, reviewing current examples and developments that could facilitate the evolution of protein for material applications.

Machine learning (ML) has revolutionized disciplines within materials science that have been able to generate sufficiently large datasets to utilize algorithms based on statistical inference, but for many important classes of materials the datasets remain small. However, a rapidly growing number of approaches to embedding domain knowledge of materials systems are reducing data requirements and allowing broader applications of ML. Furthermore, these hybrid approaches improve the interpretability of the predictions, allowing for greater physical insights into the factors that determine material properties. This review introduces a number of these strategies, providing examples of how they were implemented in ML algorithms and discussing the materials systems to which they were applied.

Nanoscale magnetization modulation by electric field enables the construction of low-power spintronic devices for information storage applications and, etc. Electric field-induced ion migration can introduce desired changes in the material's stoichiometry, defect profile, and lattice structure, which in turn provides a versatile and convenient means to modify the materials’ chemical-physical properties at the nanoscale and in situ. In this review, we provide a brief overview on the recent study on nanoscale magnetization modulation driven by electric field-induced migration of ionic species either within the switching material or from external sources. The formation of magnetic conductive filaments that exhibit magnetoresistance behaviors in resistive switching memory via foreign metal ion migration and redox activities is also discussed. Combining the magnetoresistance and quantized conductance switching of the magnetic nanopoint contact structure may provide a future high-performance device for non-von Neumann computing architectures.

Nanohybrids containing graphene and bismuth ferrite have been actively employed as efficient photo-catalysts these days owing to the low rate of charge carrier's (e−–h+) recombination, moderate surface area with a suitable range of band-gaps. We have synthesized nanohybrids of graphene oxide (GO) and doped BiFeO3 using a co-precipitation method and the doping elements were lanthanum and manganese, hence called BLFMO/GO nanohybrids. The surface area of BLFMO [La = 15% increased from 6.8 m2/g (for pure) to 62.68 m2/g (in nanohybrid)]. Also, the bandgap of the BLFMO/GO nanohybrid reduced significantly up to 1.75 eV. The resulting BLFMO/GO nanohybrid represents significantly higher catalytic activity (96% in 30 min) than the pure BiFeO3 (30% in 30 min).

Polymer redox-active materials (redoxmers) have numerous applications in the emerging electrochemical energy storage systems due to their structural versatility, fast-cycling ability, high theoretical capacity as electrode materials, sustainability, and recyclability. This review examines recent developments in improving the cycling performance of such materials and provides a vista on the future research directions.

The use of statistical/machine learning (ML) approaches to materials science is experiencing explosive growth. Here, we review recent work focusing on the generation and application of libraries from both experiment and theoretical tools. The library data enables classical correlative ML and also opens the pathway for exploration of underlying causative physical behaviors. We highlight key advances facilitated by this approach and illustrate how modeling, macroscopic experiments, and imaging can be combined to accelerate the understanding and development of new materials systems. These developments point toward a data-driven future wherein knowledge can be aggregated and synthesized, accelerating the advancement of materials science.

Here, we report that a marine sandworm Nereis virens jaw protein, Nvjp1, nucleates hemozoin with similar activity as the native parasite hemozoin protein, HisRPII. X-ray diffraction and scanning electron microscopy confirm the identity of the hemozoin produced from Nvjp1-containing reactions. Finally, we observed that nAl assembled with hemozoin from Nvjp1 reactions has a substantially higher energetic output when compared to analogous thermite from the synthetic standard or HisRPII-nucleated hemozoin. Our results demonstrate that a marine sandworm protein can nucleate malaria pigment and set the stage for engineering recombinant hemozoin production for nanoenergetic applications.

Computational acceleration of performance metric-based materials discovery via high-throughput screening and machine learning methods is becoming widespread. Nevertheless, development and optimization of the opto-electronic properties that depend on dilute concentrations of point defects in new materials have not significantly benefited from these advances. Here, the authors present an informatics and simulation suite to computationally accelerate these processes. This will enable faster and more fundamental materials research, and reduce the cost and time associated with the materials development cycle. Analogous to the new avenues enabled by current first-principles-based property databases, this type of framework will open entire new research frontiers as it proliferates.

Polarized resonant soft x-ray scattering (RSoXS) can leverage contrast between domains in mass, chemical bonds, and chain orientation to characterize polymer thin films. Although mass and chemical contrast can be predicted from the absorption spectra, the origins of orientational contrast are not fully understood. For example, in the case where one domain is crystalline, it is not obvious how the total crystallinity will affect the anisotropy. Using a simple calculation, we predict polarized RSoXS of model structures and compare with experiment. Through experiments and simulations, we demonstrate that scattering anisotropy is not monotonically related to local order but instead peaks when 50% of domains have aligned chains, i.e., when the sample is 50% crystalline.

The complex interaction between hematopoietic stem cells (HSCs) and their microenvironment in the human bone marrow ensures a life-long blood production by balancing stem cell maintenance and differentiation. This so-called HSC niche can be disturbed by malignant diseases. Investigating their consequences on hematopoiesis requires a deep understanding of how the niches function in health and disease. To facilitate this, biomimetic models of the bone marrow are needed to analyze HSC maintenance and hematopoiesis under steady state and diseased conditions. Here, 3D bone marrow models, their fabrication methods (including 3D bioprinting), and implementations recapturing bone marrow functions in health and diseases are presented.

Porphyrins absorb light to initiate photocatalytic activity. The complex, asymmetric structures of natural porphyrins such as heme, chlorophyll, and their derivatives hold unique interest. A platform for biosynthesis of porphyrins in Escherichia coli is developed with the aim of producing a variety of porphyrins for examining their photocatalytic properties within a porous material. Bioderived protoporphyrin IX is tethered inside the highly porous metal-organic framework (MOF) NU-1000 via solvent-assisted ligand incorporation. This MOF catalyzes the photocatalytic oxidation of 2-chloroethyl ethyl sulfide with improved performance over an expanded range of the visible spectrum when compared to unmodified NU-1000.