Prof. Dr. Miguel Marques

Artificial Intelligence for Integrated Material Science
Ruhr-Universität Bochum

Contact
Author IDs

Hub
  • Machine learning guided high-throughput search of non-oxide garnets
    Schmidt, Jonathan and Wang, Hai-Chen and Schmidt, Georg and Marques, Miguel A. L.
    npj Computational Materials 9 (2023)
    Garnets have found important applications in modern technologies including magnetorestriction, spintronics, lithium batteries, etc. The overwhelming majority of experimentally known garnets are oxides, while explorations (experimental or theoretical) for the rest of the chemical space have been limited in scope. A key issue is that the garnet structure has a large primitive unit cell, requiring a substantial amount of computational resources. To perform a comprehensive search of the complete chemical space for new garnets, we combine recent progress in graph neural networks with high-throughput calculations. We apply the machine learning model to identify the potentially (meta-)stable garnet systems before performing systematic density-functional calculations to validate the predictions. We discover more than 600 ternary garnets with distances to the convex hull below 100 meV ⋅ atom−1. This includes sulfide, nitride, and halide garnets. We analyze their electronic structure and discuss the connection between the value of the electronic band gap and charge balance. © 2023, The Author(s).
    view abstract10.1038/s41524-023-01009-4
  • Machine-Learning-Assisted Determination of the Global Zero-Temperature Phase Diagram of Materials
    Schmidt, Jonathan and Hoffmann, Noah and Wang, Hai-Chen and Borlido, Pedro and Carriço, Pedro J. M. A. and Cerqueira, Tiago F. T. and Botti, Silvana and Marques, Miguel A. L.
    Advanced Materials 35 (2023)
    Crystal-graph attention neural networks have emerged recently as remarkable tools for the prediction of thermodynamic stability. The efficacy of their learning capabilities and their reliability is however subject to the quantity and quality of the data they are fed. Previous networks exhibit strong biases due to the inhomogeneity of the training data. Here a high-quality dataset is engineered to provide a better balance across chemical and crystal-symmetry space. Crystal-graph neural networks trained with this dataset show unprecedented generalization accuracy. Such networks are applied to perform machine-learning-assisted high-throughput searches of stable materials, spanning 1 billion candidates. In this way, the number of vertices of the global T = 0 K phase diagram is increased by 30% and find more than ≈150 000 compounds with a distance to the convex hull of stability of less than 50 meV atom−1. The discovered materials are then accessed for applications, identifying compounds with extreme values of a few properties, such as superconductivity, superhardness, and giant gap-deformation potentials. © 2023 The Authors. Advanced Materials published by Wiley-VCH GmbH.
    view abstract10.1002/adma.202210788
  • Symmetry-based computational search for novel binary and ternary 2D materials
    Wang, Hai-Chen and Schmidt, Jonathan and Marques, Miguel A L and Wirtz, Ludger and Romero, Aldo H
    2D Materials 10 (2023)
    We present a symmetry-based systematic approach to explore the structural and compositional richness of two-dimensional materials. We use a 'combinatorial engine' that constructs candidate compounds by occupying all possible Wyckoff positions for a certain space group with combinations of chemical elements. These combinations are restricted by imposing charge neutrality and the Pauling test for electronegativities. The structures are then pre-optimized with a specially crafted universal neural-network force-field, before a final step of geometry optimization using density-functional theory is performed. In this way we unveil an unprecedented variety of two-dimensional materials, covering the whole periodic table in more than 30 different stoichiometries of form A n B m or A n B m C k . Among the discovered structures, we find examples that can be built by decorating nearly all Platonic and Archimedean tessellations as well as their dual Laves or Catalan tilings. We also obtain a rich, and unexpected, polymorphism for some specific compounds. We further accelerate the exploration of the chemical space of two-dimensional materials by employing machine-learning-accelerated prototype search, based on the structural types discovered in the systematic search. In total, we obtain around 6500 compounds, not present in previous available databases of 2D materials, with a distance to the convex hull of thermodynamic stability smaller than 250 meV/atom. © 2023 The Author(s). Published by IOP Publishing Ltd.
    view abstract10.1088/2053-1583/accc43
  • Roadmap on Machine learning in electronic structure
    Kulik, H.J. and Hammerschmidt, T. and Schmidt, J. and Botti, S. and Marques, M.A.L. and Boley, M. and Scheffler, M. and Todorović, M. and Rinke, P. and Oses, C. and Smolyanyuk, A. and Curtarolo, S. and Tkatchenko, A. and Bartók, A.P. and Manzhos, S. and Ihara, M. and Carrington, T. and Behler, J. and Isayev, O. and Veit, M. and Grisafi, A. and Nigam, J. and Ceriotti, M. and Schütt, K.T. and Westermayr, J. and Gastegger, M. and Maurer, R.J. and Kalita, B. and Burke, K. and Nagai, R. and Akashi, R. and Sugino, O. and Hermann, J. and Noé, F. and Pilati, S. and Draxl, C. and Kuban, M. and Rigamonti, S. and Scheidgen, M. and Esters, M. and Hicks, D. and Toher, C. and Balachandran, P.V. and Tamblyn, I. and Whitelam, S. and Bellinger, C. and Ghiringhelli, L.M.
    Electronic Structure 4 (2022)
    view abstract10.1088/2516-1075/ac572f
  • Detection of Cu2Zn5SnSe8 and Cu2Zn6SnSe9 phases in co-evaporated Cu2ZnSnSe4 thin-films
    Schwarz, T. and Marques, M.A.L. and Botti, S. and Mousel, M. and Redinger, A. and Siebentritt, S. and Cojocaru-Mirédin, O. and Raabe, D. and Choi, P.-P.
    Applied Physics Letters 107 (2015)
    Cu2ZnSnSe4 thin-films for photovoltaic applications are investigated using combined atom probe tomography and ab initio density functional theory. The atom probe studies reveal nano-sized grains of Cu2Zn5SnSe8 and Cu2Zn6SnSe9 composition, which cannot be assigned to any known phase reported in the literature. Both phases are considered to be metastable, as density functional theory calculations yield positive energy differences with respect to the decomposition into Cu2ZnSnSe4 and ZnSe. Among the conceivable crystal structures for both phases, a distorted zinc-blende structure shows the lowest energy, which is a few tens of meV below the energy of a wurtzite structure. A band gap of 1.1 eV is calculated for both the Cu2Zn5SnSe8 and Cu2Zn6SnSe9 phases. Possible effects of these phases on solar cell performance are discussed. © 2015 AIP Publishing LLC.
    view abstract10.1063/1.4934847

« back