Learning a reactive potential for silica-water through uncertainty attribution


JSON Export

{
  "id": "1816", 
  "updated": "2023-07-06T10:21:59.088110+00:00", 
  "metadata": {
    "version": 1, 
    "contributors": [
      {
        "givennames": "Swagata", 
        "affiliations": [
          "Department of Materials Science and Engineering, Massachusetts Institute of Technology, Cambridge, MA, USA"
        ], 
        "familyname": "Roy"
      }, 
      {
        "givennames": "Johannes", 
        "affiliations": [
          "Evonik Operations GmbH, Marl, Germany"
        ], 
        "familyname": "D\u00fcrholt"
      }, 
      {
        "givennames": "Thomas", 
        "affiliations": [
          "Evonik Operations GmbH, Marl, Germany"
        ], 
        "familyname": "Asche"
      }, 
      {
        "givennames": "Federico", 
        "affiliations": [
          "IBM Research, Zurich, Switzerland"
        ], 
        "familyname": "Zipoli"
      }, 
      {
        "givennames": "Rafael", 
        "affiliations": [
          "Department of Materials Science and Engineering, Massachusetts Institute of Technology, Cambridge, MA, USA"
        ], 
        "email": "rafagb@mit.edu", 
        "familyname": "G\u00f3mez-Bombarelli"
      }
    ], 
    "title": "Learning a reactive potential for silica-water through uncertainty attribution", 
    "_oai": {
      "id": "oai:materialscloud.org:1816"
    }, 
    "keywords": [
      "free energy", 
      "neural network potential", 
      "chemical reactions", 
      "silicates", 
      "Active learning"
    ], 
    "publication_date": "Jul 06, 2023, 12:21:59", 
    "_files": [
      {
        "key": "silica_Painn_model.zip", 
        "description": "Neural network interatomic potential (PaiNN \tarXiv:2102.03150 ) for silica-water. Predicts energy and forces with units kcal/mol and Kcal/mol-\u00c5 respectively. Code for training and running MD simulations are found at https://github.com/learningmatter-mit/NeuralForceField", 
        "checksum": "md5:54896b829ec630425f13e43dfa57049a", 
        "size": 473848819
      }, 
      {
        "key": "README.md", 
        "description": "README", 
        "checksum": "md5:0170f0f718b1cb66f52236a07bb44991", 
        "size": 1682
      }
    ], 
    "references": [
      {
        "citation": "S.Roy, J.D\u00fcrholt, T.Asche, F.Zipoli, R.G\u00f3mez-Bombarelli, arXiv:2307.01705", 
        "url": "http://arxiv.org/abs/2307.01705", 
        "type": "Preprint"
      }
    ], 
    "description": "The reactivity of silicates in aqueous solution is relevant to various chemistries ranging from silicate minerals in geology, to the C-S-H phase in cement, nanoporous zeolite catalysts, or highly porous precipitated silica. While simulations of chemical reactions can provide insight at the molecular level, balancing accuracy and scale in reactive simulations in the condensed phase is a challenge. Here, we demonstrate how a machine-learning reactive interatomic potential can accurately capture silicate-water reactivity. The model was trained on a new dataset comprising 400,000 energies and forces of molecular clusters at the \ud835\udf14-B97XD\\def2-TVZP level. To ensure the robustness of the model, we introduce a new and general active learning strategy based on the attribution of the model uncertainty, that automatically isolates uncertain regions of bulk simulations to be calculated as small-sized clusters. Our trained potential is found to reproduce static and dynamic properties of liquid water and solid crystalline silicates, despite having been trained exclusively on cluster data. Furthermore, we utilize enhanced sampling simulations to recover the self-ionization reactivity of water accurately, and the acidity of silicate oligomers, and lastly study the silicate dimerization reaction in a water solution at neutral conditions and find that the reaction occurs through a flanking mechanism.", 
    "status": "published", 
    "license": "Creative Commons Attribution 4.0 International", 
    "conceptrecid": "1815", 
    "is_last": true, 
    "mcid": "2023.106", 
    "edited_by": 576, 
    "id": "1816", 
    "owner": 1074, 
    "license_addendum": null, 
    "doi": "10.24435/materialscloud:61-x8"
  }, 
  "revision": 9, 
  "created": "2023-07-02T16:42:54.007480+00:00"
}