Highlights

  • Using machine learning to study the microscopic behavior of a solid-state electrolyte

    Scientists in Michele Ceriotti's lab at EPFL have used machine learning to paint a more precise picture of how charge transport happens in lithium thiophosphate,  a promising material for solid-state batteries. The group is now perfecting the study of this material by adding an analysis of thermal transport, which will be the topic of a new publication.

  • GPT-3 transforms chemical research

    A study by MARVEL researchers in Berend Smit's laboratory at EPFL shows that large language models such as GPT-3 can be used to simplify the application of artificial intelligence to chemical analysis, improving accuracy while drastically reducing the amount of data needed for training.  The model, trained with relatively few Q&As, correctly answered over 95% of very diverse chemical problems. The method is as easy as conducting a literature search, and is applicable to various chemical problems.

  • In search of muons: why they switch sites in antiferromagnetic oxides

    A study involving MARVEL scientists and just published in Physical Review Letters has found that in manganese oxide, a textbook antiferromagnetic material,  the site of an implanted spin-polarized muon is not well identified, but can change due to a previously neglected effect: magnetostriction.

  • An in-depth look at a high-temperature superconducting nickelate

    A new study by Philipp Werner's group at the University of Fribourg makes advancements in the theoretical description of the correlated electron state of La3Ni2O7,  which has recently been described as a superconductor with a critical temperature around 80 K. The use of  GW+EDMFT, a method developed within MARVEL over the last ten years, was crucial for the study.

  • A “gold standard” for computational materials science codes

    Scientists from NCCR MARVEL led the most comprehensive verification effort so far on computer codes for materials simulations, providing their colleagues with a reference dataset and a set of guidelines for assessing and improving existing and future codes.

  • Automated, bespoke Wannier functions for all materials

    Two newly-released articles by MARVEL members Junfeng Qiao, Giovanni Pizzi and Nicola Marzari provide scientists with very robust and reliable algorithms that standardize and automatize the process of obtaining  Wannier functions for a given material, a much used tool for computational condensed matter physics and materials science.  To validate their algorithms, the scientists first chose four or five typical materials to explain how the methods work and reproduce what chemists would already guess about the materials. Then they stress-tested these algorithms on a larger set of materials to collect statistics and compare to previous approaches.

  • The semi-metal that wasn’t there

    Scientists have been looking for real-world examples of Weyl semi-metals, that are topological materials with unique transport, optical and thermoelectric behavior. Many computational and experimental papers had described a compound of europium, cadmium and arsenic, EuCd2As2, as a Weyl semi-metal. But a new study just published by an international research team led by MARVEL’s Ana Akrap has found that it is instead a magnetic semiconductor.

  • Tackling excited states: Koopmans functionals now available as an open-source software package

    A 10-year effort has led to a theoretical framework and a software package, called koopmans, that allows to obtain reliable spectral properties of molecules and materials with density functional theory. The framework is described in a new article by MARVEL researchers at EPFL, and last month a week-long school co-organised by MARVEL in Pavia saw students learning about Koopmans functionals and trying out the code.

  • How to make better electrical contacts with graphene nanoribbons

    A research group including MARVEL scientists has developed, studied computationally and tested a transistor based on strips of graphene with widths of only a few nanometers, thanks to a new fabrication technique that overcomes the current challenges in making electrical contacts between graphene nanoribbons and electrodes.  

  • Computational Model Paves the Way for More Efficient Energy Systems

    EPFL researchers make theoretical breakthrough in thermoelectric material to better harness waste heat for sustainable energy.

  • Over 3,000 bidimensional materials are now in the Materials Cloud database

    The collection of 2D materials, first initiated  in 2018, has been expanded with 1,252 new monolayers that could be exfoliated from existing tridimensional structures.

  • A compass to explore covalent organic frameworks in search of good photocatalysts.

    A new study by Berend Smit’s group at EPFL introduces a new computational framework that allows to screen large numbers of Covalent Organic Frameworks (COFs) in a fast and efficient way, to pre-select the best candidates for specific photocatalytic applications, such as water splitting. Starting from a set of 419 COFs for which there are reported experiments, the workflow allowed the selection of 13 candidate materials for water splitting.