# Hydrogen liquid-liquid transition from first principles and machine learning

https://mdr.nims.go.jp/datasets/b3cac5fb-4d6a-47f2-8be4-9df0f77bf815

## File

- [2502.02447v2.pdf](https://mdr.nims.go.jp/filesets/71a89a09-19a7-4d3b-b2b4-d9961e5e69db/download) ([Detail](https://mdr.nims.go.jp/filesets/71a89a09-19a7-4d3b-b2b4-d9961e5e69db.md))

## Id

b3cac5fb-4d6a-47f2-8be4-9df0f77bf815

## Local identifier



## Visibility

open_to_public

## State

published

## Created at

2025-12-01T07:33:47.265910Z

## Updated at

2025-12-01T23:30:10.386167Z

## Published at

2025-12-01T23:23:29.805276Z

## Doi



## First published url

https://doi.org/10.1103/pbrk-3zgd

## Date published

2025-09-24

## Recorded date published

2025-9

## Resource type

journal_article

## Manuscript type

accepted_manuscript

## Collection



## Title

- title: Hydrogen liquid-liquid transition from first principles and machine learning
  title_type: original
  lang: en

## Description

- description: "The molecular-to-atomic liquid-liquid transition (LLT) in high-pressure
    hydrogen is a fundamental\r\ntopic touching domains from planetary science to
    materials modeling. Yet, the nature of the\r\nLLT is still under debate. To resolve
    it, numerical simulations must cover length and time scales\r\nspanning several
    orders of magnitude. We overcome these size and time limitations by constructing
    a fast and accurate machine-learning interatomic potential (MLIP) built on the
    MACE neural network architecture. The MLIP is trained on Perdew-Burke-Ernzerhof
    (PBE) density functional calculations and uses a modified loss function correcting
    for an energy bias in the molecular phase. Classical and path-integral molecular
    dynamics driven by this MLIP show that the LLT is always supercritical above the
    melting temperature. The position of the corresponding Widom line agrees with
    previous ab initio PBE calculations, which in contrast predicted a first-order
    LLT. According to our calculations, the crossover line becomes a first-order transition
    only inside the molecular crystal region. These results call for a reconsideration
    of the LLT picture previously drawn."
  description_type: abstract
  lang: und

## Creator

- name: Giacomo Tenti
  role: author
- name: Bastian Jäckl
  role: author
- name: Kousuke Nakano
  role: author
  orcid: https://orcid.org/0000-0001-7756-4355
  organization: National Institute for Materials Science
- name: Matthias Rupp
  role: author
- name: Michele Casula
  role: author

## Contact agent



## Publisher

organization: American Physical Society (APS)

## Managing organization



## Keyword

- subject: Machine learning potential
  schema: not_defined
- subject: Hydrogen liquid-liquid transition
  schema: not_defined

## Rights

- identifier: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

## Other identifier(s)



## Data origin

- data_origin_type: other

## Embargo



## Journal

- title: Physical Review B
  issn: 1550235X
  volume: '112'
  issue: '10'
  article_number: '104208'

## Conference



## Related item



## Funding

- identifier: JPMXS0320220025
  funder_name: Ministry of Education, Culture, Sports, Science and Technology
- identifier: JPMJBY24F3
  funder_name: Japan Science and Technology Corporation
- funder_name: European High Performance Computing Joint Undertaking
- identifier: HORIZON-EUROHPC-JU-2022-INCO-04
  funder_name: Horizon Foundation for New Jersey
- identifier: '952165'
  funder_name: Horizon Foundation for New Jersey

## Instrument



## Instrument operator



## Instrument managing organization



## Measurement method



## Specimen



## Chemical composition



## Structure for specimen



## Structural feature for specimen



## Specific property for specimen



## Process for specimen treatment



## Computational method



## Energy level/transition state



## Software



## Custom property



## Fileset

- id: 71a89a09-19a7-4d3b-b2b4-d9961e5e69db
  filename: 2502.02447v2.pdf
  content_type: application/pdf
  size: 2091142
  md5: 5b9d32d49c3054ca850c336dab42ad6b

## Thumbnail

fileset_id: 71a89a09-19a7-4d3b-b2b4-d9961e5e69db
filename: 2502.02447v2.pdf