Skip to content
DraconiticResearch
Search Sources API
190,335 studies31 sourcesUpdated 19 min ago
Peer-reviewedCC BY-NC-NDNature communications

Predicting phase transitions across temperature, pressure, and chemical potential using exponentially tilted thermodynamic maps.

  • Suemin Lee1,2
  • Ruiyu Wang3
  • Lukas Herron1,2
  • Pratyush Tiwary4,5,6
6 affiliations
  1. 1Biophysics Program, Institute for Physical Science and Technology, University of Maryland, College Park, MD, USA.
  2. 2University of Maryland Institute for Health Computing, Bethesda, MD, USA.
  3. 3Institute for Physical Science and Technology, University of Maryland, College Park, MD, USA.
  4. 4Biophysics Program, Institute for Physical Science and Technology, University of Maryland, College Park, MD, USA. [email protected].
  5. 5University of Maryland Institute for Health Computing, Bethesda, MD, USA. [email protected].
  6. 6Department of Chemistry and Biochemistry, University of Maryland, College Park, MD, USA. [email protected].
Read full textPDFDOICite

Abstract

Predicting and characterizing phase transitions is crucial for understanding generic physical phenomena such as crystallization, protein folding, and others. However, directly observing phase transitions is not always easy, and we often have limited observations far from the phase boundary, measured under specific thermodynamic conditions. In this study, we propose a statistical-physics- and generative-AI-driven framework that uses such limited information to generate samples of different phases under arbitrary thermodynamic conditions, which we call Exponentially Tilted Thermodynamic Maps (expTM). The central idea is to map collected data into a simple, tractable prior expressed as an exponentially tilted Gaussian. We demonstrate how the variance and mean of the prior can be correlated with pairs of thermodynamic control variables, including temperature, pressure, and chemical potential. This gives us the ability to generate thermodynamically correct samples under any values of the control variables. To demonstrate the practical applicability of this approach, we use expTM to sample a lattice gas model with the grand canonical ensemble, capturing phase transitions under varying chemical potentials and temperatures. We further demonstrate how expTM can model the isothermal-isobaric ensemble, with which we predict different phases of CO₂ under varying pressure conditions. Both examples are trained on very limited data far from the phase boundary. These results establish expTM as a robust tool for understanding phase transitions across diverse thermodynamic conditions requiring only a small number of observations.

Related

  1. Excursion-Resolved Thermodynamic Inference without Observing Reverse Transitions
    Jie Gu
    arXivPreprintCC BY 4.0Full textStatistical & nonlinear physicsarXiv
  2. Structural, electronic, spectroscopic, and thermodynamic properties of gentisic acid: solvent and temperature effects using DFT and TD-DFT approaches
    Eshetu Diriba Kena, Alemu Kebede Hordofa, Dereje Gelanu Dadi +2
    Scientific ReportsPeer-reviewedCC BY-NC-ND 4.0Full textAtomic, molecular & optical physicsCrossref
  3. Thermodynamic geometry as the missing link: toward a unified framework for black hole first-order phase transitions
    Shi-Hao Zhang, Jing-Fei Zhang, Xin Zhang
    arXivPreprintCC BY 4.0Full textGravitation & general relativityHigh-energy & particle physicsarXiv
  4. A thermodynamic reference for field-induced phenomena in α-RuCl3: two coexisting polymorphs account for multiple magnetic phase transitions
    Hamza Nasir, Daniel Balazs, Muhammad Nauman +5
    arXivPreprintCC BY 4.0Full textCondensed matterarXiv
  5. A multiphysics-coupled DeepONet framework for predicting sodium dendrite growth in wide-temperature sodium-ion capacitors
    Yifan Hao, Jing Yang
    Electrochemistry CommunicationsPeer-reviewedCC BY 4.0Full textStatistical & nonlinear physicsCrossref

Cite

BibTeXRIS
BibTeX
@article{lee2026predicting,
  title = {{Predicting phase transitions across temperature, pressure, and chemical potential using exponentially tilted thermodynamic maps.}},
  author = {Suemin Lee and Ruiyu Wang and Lukas Herron and Pratyush Tiwary},
  journal = {Nature communications},
  year = {2026},
  month = sep,
  doi = {10.1038/s41467-026-77612-y},
  url = {https://europepmc.org/article/MED/42875434},
  note = {CC BY-NC-ND},
}
RIS
TY  - JOUR
TI  - Predicting phase transitions across temperature, pressure, and chemical potential using exponentially tilted thermodynamic maps.
AU  - Suemin Lee
AU  - Ruiyu Wang
AU  - Lukas Herron
AU  - Pratyush Tiwary
PY  - 2026
DA  - 2026/09/14
JO  - Nature communications
DO  - 10.1038/s41467-026-77612-y
UR  - https://europepmc.org/article/MED/42875434
AB  - Predicting and characterizing phase transitions is crucial for understanding generic physical phenomena such as crystallization, protein folding, and others. However, directly observing phase transitions is not always easy, and we often have limited observations far from the phase boundary, measured under specific thermodynamic conditions. In this study, we propose a statistical-physics- and generative-AI-driven framework that uses such limited information to generate samples of different phases under arbitrary thermodynamic conditions, which we call Exponentially Tilted Thermodynamic Maps (expTM). The central idea is to map collected data into a simple, tractable prior expressed as an exponentially tilted Gaussian. We demonstrate how the variance and mean of the prior can be correlated with pairs of thermodynamic control variables, including temperature, pressure, and chemical potential. This gives us the ability to generate thermodynamically correct samples under any values of the control variables. To demonstrate the practical applicability of this approach, we use expTM to sample a lattice gas model with the grand canonical ensemble, capturing phase transitions under varying chemical potentials and temperatures. We further demonstrate how expTM can model the isothermal-isobaric ensemble, with which we predict different phases of CO₂ under varying pressure conditions. Both examples are trained on very limited data far from the phase boundary. These results establish expTM as a robust tool for understanding phase transitions across diverse thermodynamic conditions requiring only a small number of observations.
SN  - 2041-1723
ER  -