Low-cost self-driving labs (SDLs) offer faster prototyping, low-risk hands-on experience, and a test bed for sophisticated experimental planning software which helps us develop state-of-the-art SDLs.

Diffusion Models outperform Generative Adversarial Networks (GANs) and Wasserstein GANs in material discovery.

Over the past 150 years, our ability to produce and transform engineered materials has been responsible for our current high standards of living, especially in developed economies. However, we must carefully think of the effects our addiction to creating and using materials at this fast rate will have on the future generations. The way we currently make and use materials detrimentally affects the planet Earth, creating many severe environmental problems. It affects the next generations by putting in danger the future of the economy, energy, and climate. We are at the point where something must drastically change, and it must change now. We must create more sustainable materials alternatives using natural raw materials and inspiration from nature while making sure not to deplete important resources, i.e. in competition with the food chain supply. We must use less materials, eliminate the use of toxic materials and create a circular materials economy where reuse and recycle are priorities. We must develop sustainable methods for materials recycling and encourage design for disassembly. We must look across the whole materials life cycle from raw resources till end of life and apply thorough life cycle assessments (LCAs) based on reliable and relevant data to quantify sustainability. We need to seriously start thinking of where our future materials will come from and how could we track them, given that we are confronted with resource scarcity and geographical constrains. This is particularly important for the development of new and sustainable energy technologies, key to our transition to net zero. Currently ‘critical materials’ are central components of sustainable energy systems because they are the best performing. A few examples include the permanent magnets based on rare earth metals (Dy, Nd, Pr) used in wind turbines, Li and Co in Li-ion batteries, Pt and Ir in fuel cells and electrolysers, Si in solar cells just to mention a few. These materials are classified as ‘critical’ by the European Union and Department of Energy. Except in sustainable energy, materials are also key components in packaging, construction, and textile industry along with many other industrial sectors. This roadmap authored by prominent researchers working across disciplines in the very important field of sustainable materials is intended to highlight the outstanding issues that must be addressed and provide an insight into the pathways towards solving them adopted by the sustainable materials community. In compiling this roadmap, we hope to aid the development of the wider sustainable materials research community, providing a guide for academia, industry, government, and funding agencies in this critically important and rapidly developing research space which is key to future sustainability.

We present Descending from Stochastic Clustering Variance Regression (DiSCoVeR) (https://www.github.com/sparks-baird/mat_discover), a Python tool for identifying and assessing high-performing, chemically unique compositions relative to existing compounds using a combination of a chemical distance metric, density-aware dimensionality reduction, clustering, and a regression model. In this work, we create pairwise distance matrices between compounds via Element Mover's Distance (ElMD) and use these to create 2D density-aware embeddings for chemical compositions via Density-preserving Uniform Manifold Approximation and Projection (DensMAP). Because ElMD assigns distances between compounds that are more chemically intuitive than Euclidean-based distances, the compounds can then be clustered into chemically homogeneous clusters via Hierarchical Density-based Spatial Clustering of Applications with Noise (HDBSCAN*). In combination with performance predictions via Compositionally-Restricted Attention-Based Network (CrabNet), we introduce several new metrics for materials discovery and validate DiSCoVeR on Materials Project bulk moduli using compound-wise and cluster-wise validation methods. We visualize these via multi-objective Pareto front plots and assign a weighted score to each composition that encompasses the trade-off between performance and density-based chemical uniqueness. In addition to density-based metrics, we explore an additional uniqueness proxy related to property gradients in DensMAP space. As a validation study, we use DiSCoVeR to screen materials for both performance and uniqueness to extrapolate to new chemical spaces. Top-10 rankings are provided for the compound-wise density and property gradient uniqueness proxies. Top-ranked compounds can be further curated via literature searches, physics-based simulations, and/or experimental synthesis. Finally, we compare DiSCoVeR against the naive baseline of random search for several parameter combinations in an adaptive design scheme. To our knowledge, this is the first time automated screening has been performed with explicit emphasis on discovering high-performing, novel materials.