Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

MLIP-MC: A Framework for Adsorption Simulations using Machine-Learned Interatomic Potentials

Connor W. Edwards, Fengxu Yang, Konstantin Stracke, Jack D. Evans

TL;DR

This work introduces MLIP-MC, an open-source Python framework to conduct GCMC simulations with MLIPs, and uses this framework to benchmark a series of universal models, including MACE-MP-0, ORB-v3, and fairchem ODAC, for CO2 adsorption on ZIF-8, ZIF-4, and Mg-MOF-74.

Abstract

Grand canonical Monte Carlo (GCMC) simulations are essential for screening metal-organic frameworks (MOFs) for gas adsorption, yet their accuracy is limited by underlying interatomic potentials. Universal machine-learned interatomic potentials (MLIPs), trained on diverse chemical datasets, promise zero-shot prediction without system-specific training. We introduce MLIP-MC, an open-source Python framework to conduct GCMC simulations with MLIPs, and use this framework to benchmark a series of universal models, including MACE-MP-0, ORB-v3, and fairchem ODAC, for CO2 adsorption on ZIF-8, ZIF-4, and Mg-MOF-74. All universal models exhibit systematic biases, consistently over- or underestimating adsorption energetics. Crucially, accuracy depends on training data composition: only models trained on MOF-adsorbate interactions achieve reasonable agreement with a density functional theory derived reference. Errors grow linearly with CO2 uptake, reflecting compounding inaccuracies in adsorbate-adsorbate interactions. Our results demonstrate that current universal MLIPs require finetuning for quantitative adsorption predictions and demonstrate the power of MLIP-MC to rapidly test models.

MLIP-MC: A Framework for Adsorption Simulations using Machine-Learned Interatomic Potentials

TL;DR

This work introduces MLIP-MC, an open-source Python framework to conduct GCMC simulations with MLIPs, and uses this framework to benchmark a series of universal models, including MACE-MP-0, ORB-v3, and fairchem ODAC, for CO2 adsorption on ZIF-8, ZIF-4, and Mg-MOF-74.

Abstract

Grand canonical Monte Carlo (GCMC) simulations are essential for screening metal-organic frameworks (MOFs) for gas adsorption, yet their accuracy is limited by underlying interatomic potentials. Universal machine-learned interatomic potentials (MLIPs), trained on diverse chemical datasets, promise zero-shot prediction without system-specific training. We introduce MLIP-MC, an open-source Python framework to conduct GCMC simulations with MLIPs, and use this framework to benchmark a series of universal models, including MACE-MP-0, ORB-v3, and fairchem ODAC, for CO2 adsorption on ZIF-8, ZIF-4, and Mg-MOF-74. All universal models exhibit systematic biases, consistently over- or underestimating adsorption energetics. Crucially, accuracy depends on training data composition: only models trained on MOF-adsorbate interactions achieve reasonable agreement with a density functional theory derived reference. Errors grow linearly with CO2 uptake, reflecting compounding inaccuracies in adsorbate-adsorbate interactions. Our results demonstrate that current universal MLIPs require finetuning for quantitative adsorption predictions and demonstrate the power of MLIP-MC to rapidly test models.
Paper Structure (9 sections, 2 equations, 4 figures, 1 table)

This paper contains 9 sections, 2 equations, 4 figures, 1 table.

Table of Contents

  1. Introduction
  2. Computational methods
  3. MLIP-MC framework
  4. Simulation details
  5. Results and discussion
  6. Isosteric heat of adsorption at infinite dilution
  7. Adsorption isotherms
  8. Shortcomings of universal models
  9. Summary and Outlook

Figures (4)

  • Figure 1: Isosteric heat of adsorption of CO2 at infinite dilution ($Q_{st, dilution}$) on a) ZIF-8, b) ZIF-4 and c) Mg-MOF-74 calculated by Widom insertion for a series of universal models. Reference data from the finetuned model of Goeminne et al.10.1021/acs.jctc.3c00495
  • Figure 2: Isosteric heat of adsorption ($Q_{st}$) for CO2 on a) ZIF-8, b) ZIF-4 and c) Mg-MOF-74 computed by GCMC simulations for a series of universal models. Reference data from a finetuned model of Goeminne et al.10.1021/acs.jctc.3c00495
  • Figure 3: CO2 adsorption isotherms for a) ZIF-8, b) ZIF-4 and c) Mg-MOF-74 calculated using GCMC simulations for a series of universal models. Reference data from the finetuned model of Goeminne et al.10.1021/acs.jctc.3c00495
  • Figure 4: Error in interaction energy, model predicted relative to the reference DFT reported by Goeminne et al.10.1021/acs.jctc.3c00495, with increasing CO2 molecules for the adsorption on ZIF-8.