Highlights

More than 30 research institutions including NCCR MARVEL have come together to form the Open Databases Integration for Materials Design consortium and develop an API specification enabling seamless access and interoperability among materials databases. The paper “OPTIMADE, an API for exchanging materials data,” published today in Nature Research’s journal Scientific Data, presents the OPTIMADE API specification, illustrates its use and discusses future prospects and ongoing development. 

Chemical engineers at EPFL have developed a machine-learning model that can predict a compound’s oxidation state, a property that is so essential that many chemists argue it must be included in the periodic table.

All so-far experimentally determined Weyl semimetals (WSMs) have featured Weyl points—crossings of linearly dispersing energy bands of three-dimensional crystals—that appear in pairs in the momentum space. Now, NCCR MARVEL researchers have combined experimental and theoretical methods to show that the WSM platinum gallium PtGa features a singular, unpaired Weyl point surrounded by closed Weyl nodal walls (WNW). This is the first time that an unpaired electronic WP has been observed experimentally and the first time that electronic topological WNWs/surfaces have been experimentally reported in crystalline solids. The findings of teams led by Prof. Oleg Yazyev, Chair of Computational Condensed Matter Physics at EPFL and Prof. Ming Shi of the Spectroscopy of Novel Materials group at the Paul Scherrer Institute extend our understanding of basic topological physics and the application of Weyl semimetals into spintronics.

The modelling of some molecular systems requires an explicit description of changes in the electronic structure. While quantum mechanics/molecular mechanics (QM/MM) molecular dynamics (MD) simulations have been developed for such cases, they are computationally expensive compared to fully classical simulations. In a paper recently published in the Journal of Chemical Theory and Computation, ETH Prof. Sereina Riniker, group leader in NCCR MARVEL Design and Discovery Project 1, and colleagues Lennard Böselt and Moritz Thürlemann show how using a machine learning (ML) approach that combines high-dimensional neural network potentials (HDNNPs) and Δ-learning reduces the cost of modelling while retaining accuracy in (QM)ML/MM MD simulations of condensed-phase systems.

Two recently published papers from a collaboration between two NCCR MARVEL labs have identified a new type of defect as the most common source of disorder in on-surface synthesized graphene nanoribbons (GNRs), a novel class of carbon-based materials that may prove extremely useful in next-generation electronic devices. Combining scanning probe microscopy with first-principles calculations allowed the researchers to identify the atomic structure of these so-called "bite" defects and to investigate their effect on quantum electronic transport in two different types of graphene nanoribbon. They also established guidelines for minimizing the detrimental impact of these defects on electronic transport and proposed defective zigzag-edged nanoribbons as suitable platforms for certain applications in spintronics.

Perovskite solar cells (PSCs) are among the most promising and cheapest photovoltaic technologies now available, but widespread application has been hampered by issues linked to long-term stability and processability. In the paper A combined molecular dynamics and experimental study of two-step process enabling low-temperature formation of phase-pure α-FAPbI3, recently published in Science Advances, researchers including Prof. Michele Parrinello, professor of computational Sciences at the Università della Svizzera italiana and ETHZ as well as Group Leader in NCCR MARVEL's Design & Discovery Project 1, and Paramvir Ahlawat,  a PhD student in the EPFL Lab of Prof. Ursula Roethlisberger, address this problem with a combined experimental and simulation study that could improve the design of industrial-scale processing techniques for MAPbI3 and FAPbI3, two lead perovskites.

Two-dimensional superconductors have drawn considerable attention both for the fundamental physics they display as well as for potential applications in fields such as quantum computing. Although considerable efforts have been made to identify them, materials with high transition temperatures have been hard to find. Materials that feature both superconductivity and non-trivial band topology, a combination that could potentially give rise to exotic states of matter, have proven even more elusive. In the paper Prediction of phonon-mediated superconductivity with high critical temperature in the two-dimensional topological semimetal W2N3 , recently published in Nano Letters, researchers predict just such a material in the easily exfoliable, topologically non-trivial 2D semimetal W2N3.

Combing through the “vast virtual set of all conceivable stable combinations of elements and structural configurations which form matter” is a daunting task in materials design. In the paper Simplifying inverse material design problems for fixed lattices with alchemical chirality, Anatole von Lilienfeld, professor at the Institute of Physical Chemistry at the University of Basel/University of Vienna and project leader of NCCR MARVEL Incubator Project 2 and postdoc Guido Falk von Rudorff, show how the concept of 4-dimensional chirality resulting from an alchemical reflection plane in the nuclear charge space allows researchers to break down combinatorial scaling. This alchemical chirality and the simplifications that it allows deepen our understanding of chemical compound space and enable researchers to establish new trends ‘on-the-fly,’ without resorting to empirical observation. The paper is forthcoming in Science Advances. 

Researchers led by Oleg Yazyev, head of the Chair of Computational Condensed Matter Physics at EPFL, have determined a straightforward method for probing the topological character of electronic bands in two-dimensional Moiré superlattices using Landau level sequences. The results can be easily extended to other twisted graphene multilayers and h-BN/graphene heterostructures, making the approach a powerful tool for detecting non-trivial valley band topology. The paper Landau Levels as a Probe for Band Topology in Graphene Moiré Superlattices was recently published in Physical Review Letters.

Pinning down the nature of bulk hydrated electrons—extra electrons solvated in liquid water—has proven difficult experimentally because of their short lifetime and high reactivity. Theoretical exploration has been limited by the high level of electronic structure theory needed to achieve predictive accuracy. Now, joint work from teams at the University of Zurich and EPFL and colleagues has resulted in a highly accurate machine-learning (ML) model that is inexpensive enough to allow for a full quantum statistical and dynamical description, giving an accurate determination of the structure, diffusion mechanisms, and vibrational spectroscopy of the solvated electron. This new approach, outlined in the Nature Communications paper Simulating the Ghost: Quantum Dynamics of the Solvated Electron, could also be applied to excited states and quasiparticles such as polarons and would allow for high-accuracy simulations at a moderate price.

The combination of electronic structure calculations and machine learning (ML) techniques has become commonplace in atomistic modelling—ML interatomic potentials, for example, can now describe the potential energy surface of a material across many phases, including a wide range of defects. Looking ahead, however, it is the calculation of ML models that can predict properties beyond the interactions between atoms that might eventually allow integrated machine learning models to replace costly electronic structure calculations entirely. In the paper Learning the electronic density of states in condensed matter, recently published in Physical Review B, researchers led by Michele Ceriotti, head of EPFL’s Laboratory of Computational Science and Modelling, have taken a step in that direction with a new ML framework for predicting the electronic density of states (DOS). The technique has already been applied to understand electronic transitions in dense amorphous silicon, in a paper that came out in Nature today. 

The NCCR MARVEL has ‘Materials’ revolution’ in its name for a reason: it seeks to radically transform and accelerate the process of materials discovery and design. It seeks to do this by taking a computational approach, built on a platform of database-driven, high-throughput quantum simulations: if we could compute the properties of every material using electronic structure calculations, and if we could use that information to screen big databases for materials with just the right characteristics for a given purpose — then that would put us at an enormous advantage in the search for new and better materials, avoiding much of the trial and error that comes with experiments in the lab. Such a project requires robust computer resources. For one, the sort of calculations that MARVEL research relies on take a lot of computing time and data storage; as those calculations get ever more sophisticated, demands on computing power will only grow. For another, to turn the MARVEL project into a true revolution, computer resources must be organised in such a way that they can serve as a platform for the broader materials science community.