skip to main content
DOE PAGES title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: Predicting the dissolution kinetics of silicate glasses by topology-informed machine learning

Abstract

Machine learning (ML) regression methods are promising tools to develop models predicting the properties of materials by learning from existing databases. However, although ML models are usually good at interpolating data, they often do not offer reliable extrapolations and can violate the laws of physics. Here, to address the limitations of traditional ML, we introduce a “topology-informed ML” paradigm—wherein some features of the network topology (rather than traditional descriptors) are used as fingerprint for ML models—and apply this method to predict the forward (stage I) dissolution rate of a series of silicate glasses. We demonstrate that relying on a topological description of the atomic network (i) increases the accuracy of the predictions, (ii) enhances the simplicity and interpretability of the predictive models, (iii) reduces the need for large training sets, and (iv) improves the ability of the models to extrapolate predictions far from their training sets. As such, topology-informed ML can overcome the limitations facing traditional ML (e.g., accuracy vs. simplicity tradeoff) and offers a promising route to predict the properties of materials in a robust fashion.

Authors:
 [1];  [1]; ORCiD logo [2]; ORCiD logo [3];  [4]; ORCiD logo [5];  [1]
  1. Univ. of California, Los Angeles, CA (United States)
  2. Univ. of California, Los Angeles, CA (United States); India Inst. of Technology Delhi (India)
  3. Aalborg Univ. (Denmark)
  4. Pacific Northwest National Lab. (PNNL), Richland, WA (United States)
  5. Alternative Energies and Atomic Energy Commission (CEA), Bagnols-sur-Ceze (France)
Publication Date:
Research Org.:
Alternative Energies and Atomic Energy Commission (CEA), Cadarache (France)
Sponsoring Org.:
USDOE Office of Science (SC), Basic Energy Sciences (BES); National Science Foundation (NSF)
OSTI Identifier:
1667366
Grant/Contract Number:  
SC0016584; 1562066
Resource Type:
Accepted Manuscript
Journal Name:
npj Materials Degradation
Additional Journal Information:
Journal Volume: 3; Journal Issue: 1; Journal ID: ISSN 2397-2106
Publisher:
Springer
Country of Publication:
United States
Language:
English
Subject:
36 MATERIALS SCIENCE; condensed-matter physics; glasses

Citation Formats

Liu, Han, Zhang, Tony, Anoop Krishnan, N. M., Smedskjaer, Morten M., Ryan, Joseph V., Gin, Stéṕhane, and Bauchy, Mathieu. Predicting the dissolution kinetics of silicate glasses by topology-informed machine learning. United States: N. p., 2019. Web. doi:10.1038/s41529-019-0094-1.
Liu, Han, Zhang, Tony, Anoop Krishnan, N. M., Smedskjaer, Morten M., Ryan, Joseph V., Gin, Stéṕhane, & Bauchy, Mathieu. Predicting the dissolution kinetics of silicate glasses by topology-informed machine learning. United States. doi:10.1038/s41529-019-0094-1.
Liu, Han, Zhang, Tony, Anoop Krishnan, N. M., Smedskjaer, Morten M., Ryan, Joseph V., Gin, Stéṕhane, and Bauchy, Mathieu. Thu . "Predicting the dissolution kinetics of silicate glasses by topology-informed machine learning". United States. doi:10.1038/s41529-019-0094-1. https://www.osti.gov/servlets/purl/1667366.
@article{osti_1667366,
title = {Predicting the dissolution kinetics of silicate glasses by topology-informed machine learning},
author = {Liu, Han and Zhang, Tony and Anoop Krishnan, N. M. and Smedskjaer, Morten M. and Ryan, Joseph V. and Gin, Stéṕhane and Bauchy, Mathieu},
abstractNote = {Machine learning (ML) regression methods are promising tools to develop models predicting the properties of materials by learning from existing databases. However, although ML models are usually good at interpolating data, they often do not offer reliable extrapolations and can violate the laws of physics. Here, to address the limitations of traditional ML, we introduce a “topology-informed ML” paradigm—wherein some features of the network topology (rather than traditional descriptors) are used as fingerprint for ML models—and apply this method to predict the forward (stage I) dissolution rate of a series of silicate glasses. We demonstrate that relying on a topological description of the atomic network (i) increases the accuracy of the predictions, (ii) enhances the simplicity and interpretability of the predictive models, (iii) reduces the need for large training sets, and (iv) improves the ability of the models to extrapolate predictions far from their training sets. As such, topology-informed ML can overcome the limitations facing traditional ML (e.g., accuracy vs. simplicity tradeoff) and offers a promising route to predict the properties of materials in a robust fashion.},
doi = {10.1038/s41529-019-0094-1},
journal = {npj Materials Degradation},
number = 1,
volume = 3,
place = {United States},
year = {2019},
month = {8}
}

Journal Article:
Free Publicly Available Full Text
Publisher's Version of Record

Save / Share:

Works referenced in this record:

Accelerating the Design of Functional Glasses through Modeling
journal, June 2016


Solubility of glasses in the system P2O5–CaO–MgO–Na2O–TiO2: Experimental and modeling using artificial neural networks
journal, March 2007


Predicting glass transition temperatures using neural networks
journal, October 2018


Predicting the dissolution kinetics of silicate glasses using machine learning
journal, May 2018


Structure-property relationships from universal signatures of plasticity in disordered solids
journal, November 2017


Machine learning in materials informatics: recent applications and prospects
journal, December 2017

  • Ramprasad, Rampi; Batra, Rohit; Pilania, Ghanshyam
  • npj Computational Materials, Vol. 3, Issue 1
  • DOI: 10.1038/s41524-017-0056-5

Current Understanding and Remaining Challenges in Modeling Long-Term Degradation of Borosilicate Nuclear Waste Glasses
journal, November 2013

  • Vienna, John D.; Ryan, Joseph V.; Gin, Stéphane
  • International Journal of Applied Glass Science, Vol. 4, Issue 4
  • DOI: 10.1111/ijag.12050

Nuclear Waste Glasses - How Durable?
journal, December 2006


Durable Glass for Thousands of Years
journal, March 2010


Structure of International Simple Glass and properties of passivating layer formed in circumneutral pH conditions
journal, February 2018


Thermodynamic and kinetic constraints on reaction rates among minerals and aqueous solutions. II. Rate constants, effective surface area, and the hydrolysis of feldspar
journal, December 1984

  • Helgeson, Harold C.; Murphy, William M.; Aagaard, Per
  • Geochimica et Cosmochimica Acta, Vol. 48, Issue 12
  • DOI: 10.1016/0016-7037(84)90294-1

Diffusion-controlled reaction of water with glass
journal, April 1983


Investigating structural features which control the dissolution of bioactive phosphate glasses: Beyond the network connectivity
journal, January 2016


Decoding the glass genome
journal, April 2018


Effects of glass structure on the corrosion behavior of sodium-aluminosilicate glasses
journal, December 1997


Dissolution of albite glass and crystal
journal, August 2000

  • Hamilton, James P.; Pantano, Carlo G.; Brantley, Susan L.
  • Geochimica et Cosmochimica Acta, Vol. 64, Issue 15
  • DOI: 10.1016/S0016-7037(00)00388-4

Dissolution of nepheline, jadeite and albite glasses: toward better models for aluminosilicate dissolution
journal, November 2001

  • Hamilton, James P.; Brantley, Susan L.; Pantano, Carlo G.
  • Geochimica et Cosmochimica Acta, Vol. 65, Issue 21
  • DOI: 10.1016/S0016-7037(01)00724-4

General models for estimating daily global solar radiation for different solar radiation zones in mainland China
journal, June 2013


Impacts of glass composition, pH, and temperature on glass forward dissolution rate
journal, August 2018

  • Vienna, John D.; Neeway, James J.; Ryan, Joseph V.
  • npj Materials Degradation, Vol. 2, Issue 1
  • DOI: 10.1038/s41529-018-0042-5

Topological Control on Silicates’ Dissolution Kinetics
journal, April 2016


An experimental study of the dissolution rates of simulated aluminoborosilicate waste glasses as a function of pH and temperature under dilute conditions
journal, September 2008


Dissolution Kinetics of Hot Compressed Oxide Glasses
journal, September 2017

  • Mascaraque, Nerea; Bauchy, Mathieu; Fierro, José Luis G.
  • The Journal of Physical Chemistry B, Vol. 121, Issue 38
  • DOI: 10.1021/acs.jpcb.7b04535

Thermodynamic and kinetic constraints on reaction rates among minerals and aqueous solutions; I, Theoretical considerations
journal, March 1982


Topology of covalent non-crystalline solids I: Short-range order in chalcogenide alloys
journal, October 1979


Topology of covalent non-crystalline solids II: Medium-range order in chalcogenide alloys and ASi(Ge)
journal, January 1981


Prediction of Glass Hardness Using Temperature-Dependent Constraint Theory
journal, September 2010


Fracture toughness anomalies: Viewpoint of topological constraint theory
journal, December 2016


Topological Control on the Structural Relaxation of Atomic Networks under Stress
journal, July 2017


Composition dependence of glass transition temperature and fragility. I. A topological model incorporating temperature-dependent constraints
journal, March 2009

  • Gupta, Prabhat K.; Mauro, John C.
  • The Journal of Chemical Physics, Vol. 130, Issue 9
  • DOI: 10.1063/1.3077168

Composition dependence of glass transition temperature and fragility. II. A topological model of alkali borate liquids
journal, June 2009

  • Mauro, John C.; Gupta, Prabhat K.; Loucks, Roger J.
  • The Journal of Chemical Physics, Vol. 130, Issue 23
  • DOI: 10.1063/1.3152432

Prediction of the Young's modulus of silicate glasses by topological constraint theory
journal, June 2019


Direct Experimental Evidence for Differing Reactivity Alterations of Minerals following Irradiation: The Case of Calcite and Quartz
journal, January 2016

  • Pignatelli, Isabella; Kumar, Aditya; Field, Kevin G.
  • Scientific Reports, Vol. 6, Issue 1
  • DOI: 10.1038/srep20155

Topological controls on the dissolution kinetics of glassy aluminosilicates
journal, August 2017

  • Oey, Tandré; Kumar, Aditya; Pignatelli, Isabella
  • Journal of the American Ceramic Society, Vol. 100, Issue 12
  • DOI: 10.1111/jace.15122

Rate controls on silicate dissolution in cementitious environments
journal, September 2017


The role of the network-modifier's field-strength in the chemical durability of aluminoborate glasses
journal, February 2019


Effects of Irradiation on Albite’s Chemical Durability
journal, September 2017

  • Hsiao, Yi-Hsuan; La Plante, Erika Callagon; Krishnan, N. M. Anoop
  • The Journal of Physical Chemistry A, Vol. 121, Issue 41
  • DOI: 10.1021/acs.jpca.7b05098

Correlating the Network Topology of Oxide Glasses with their Chemical Durability
journal, January 2017

  • Mascaraque, Nerea; Bauchy, Mathieu; Smedskjaer, Morten M.
  • The Journal of Physical Chemistry B, Vol. 121, Issue 5
  • DOI: 10.1021/acs.jpcb.6b11371

Role of Electrochemical Surface Potential and Irradiation on Garnet-Type Almandine’s Dissolution Kinetics
journal, March 2018

  • Hsiao, Yi-Hsuan; La Plante, Erika Callagon; Krishnan, N. M. Anoop
  • The Journal of Physical Chemistry C, Vol. 122, Issue 30
  • DOI: 10.1021/acs.jpcc.8b04459

Accuracy vs. Simplicity: A Complex Trade-Off
journal, January 2002

  • Aragones, Enriqueta; Gilboa, Itzhak; Postlewaite, Andrew
  • SSRN Electronic Journal
  • DOI: 10.2139/ssrn.332382

Artificial neural networks (the multilayer perceptron)—a review of applications in the atmospheric sciences
journal, August 1998


Structure of boroaluminosilicate glasses: Impact of [Al 2 O 3 ]/[SiO 2 ] ratio on the structural role of sodium
journal, August 2012