Discovery of Sustainable Refrigerants through Physics-Informed RL Fine-Tuning of Sequence Models

• URL: http://arxiv.org/abs/2509.19588v1
• Date: Tue, 23 Sep 2025 21:24:35 GMT
• Title: Discovery of Sustainable Refrigerants through Physics-Informed RL Fine-Tuning of Sequence Models
• Authors: Adrien Goldszal, Diego Calanzone, Vincent Taboga, Pierre-Luc Bacon,
• Abstract summary: Refgen is a generative pipeline that integrates machine learning with physics-grounded inductive biases.<n>We present Refgen, a generative pipeline that integrates machine learning with physics-grounded inductive biases.
• Score: 12.04169043797071
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Most refrigerants currently used in air-conditioning systems, such as hydrofluorocarbons, are potent greenhouse gases and are being phased down. Large-scale molecular screening has been applied to the search for alternatives, but in practice only about 300 refrigerants are known, and only a few additional candidates have been suggested without experimental validation. This scarcity of reliable data limits the effectiveness of purely data-driven methods. We present Refgen, a generative pipeline that integrates machine learning with physics-grounded inductive biases. Alongside fine-tuning for valid molecular generation, Refgen incorporates predictive models for critical properties, equations of state, thermochemical polynomials, and full vapor compression cycle simulations. These models enable reinforcement learning fine-tuning under thermodynamic constraints, enforcing consistency and guiding discovery toward molecules that balance efficiency, safety, and environmental impact. By embedding physics into the learning process, Refgen leverages scarce data effectively and enables de novo refrigerant discovery beyond the known set of compounds.

Related papers

• Refining Machine Learning Potentials through Thermodynamic Theory of Phase Transitions [0.0]
  This work proposes a fine-tuning strategy via top-down learning to correct wrongly predicted transition temperatures.<n>We demonstrate that our approach can accurately correct the phase diagram of pure Titanium in a pressure range of up to 5 GPa.<n>Our approach is model-agnostic, applicable to multi-component systems with solid-solid and solid-liquid transitions.
  arXiv  Detail & Related papers  (2025-12-03T17:06:26Z)
• Foundation Models for Discovery and Exploration in Chemical Space [57.97784111110166]
  MIST is a family of molecular foundation models trained on large unlabeled datasets.<n>We demonstrate the ability of these models to solve real-world problems across chemical space.
  arXiv  Detail & Related papers  (2025-10-20T17:56:01Z)
• Iterative Distillation for Reward-Guided Fine-Tuning of Diffusion Models in Biomolecular Design [58.8094854658848]
  We address the problem of fine-tuning diffusion models for reward-guided generation in biomolecular design.<n>We propose an iterative distillation-based fine-tuning framework that enables diffusion models to optimize for arbitrary reward functions.<n>Our off-policy formulation, combined with KL divergence minimization, enhances training stability and sample efficiency compared to existing RL-based methods.
  arXiv  Detail & Related papers  (2025-07-01T05:55:28Z)
• ChemKANs for Combustion Chemistry Modeling and Acceleration [0.0]
  ChemKAN is a novel neural network framework for model inference and simulation acceleration for combustion chemistry.<n>ChemKAN's structure augments the generic Kolmogorov Arnold Network Ordinary Differential Equations (KAN-ODEs) with knowledge of the information flow through the relevant kinetic and thermodynamic laws.<n>We benchmark the robustness of ChemKANs to sparse data containing up to 15% added noise, and superfluously large network parameterizations.
  arXiv  Detail & Related papers  (2025-04-17T01:53:28Z)
• A physics-based data-driven model for CO$_2$ gas diffusion electrodes to drive automated laboratories [34.82692226532414]
  electrochemical reduction of atmospheric CO$$ into high-energy molecules with renewable energy is a promising avenue for energy storage.<n>We present a modeling framework to efficiently explore the high-dimensional parameter space of GDE designs in an active learning context.<n>We deploy the model in a simulated active learning setup with real electrochemical data gathered by the AdaCarbon automated laboratory.
  arXiv  Detail & Related papers  (2025-02-10T10:20:10Z)
• Conditional Synthesis of 3D Molecules with Time Correction Sampler [58.0834973489875]
  Time-Aware Conditional Synthesis (TACS) is a novel approach to conditional generation on diffusion models. It integrates adaptively controlled plug-and-play "online" guidance into a diffusion model, driving samples toward the desired properties.
  arXiv  Detail & Related papers  (2024-11-01T12:59:25Z)
• Physics-guided weak-form discovery of reduced-order models for trapped ultracold hydrodynamics [0.0]
  We study the relaxation of a highly collisional, ultracold but nondegenerate gas of polar molecules. The gas is subject to fluid-gas coupled dynamics that lead to a breakdown of first-order hydrodynamics. We present substantially improved reduced-order models for these same observables.
  arXiv  Detail & Related papers  (2024-06-11T17:50:04Z)
• GFlowNets for AI-Driven Scientific Discovery [74.27219800878304]
  We present a new probabilistic machine learning framework called GFlowNets. GFlowNets can be applied in the modeling, hypotheses generation and experimental design stages of the experimental science loop. We argue that GFlowNets can become a valuable tool for AI-driven scientific discovery.
  arXiv  Detail & Related papers  (2023-02-01T17:29:43Z)
• Prediction of liquid fuel properties using machine learning models with Gaussian processes and probabilistic conditional generative learning [56.67751936864119]
  The present work aims to construct cheap-to-compute machine learning (ML) models to act as closure equations for predicting the physical properties of alternative fuels. Those models can be trained using the database from MD simulations and/or experimental measurements in a data-fusion-fidelity approach. The results show that ML models can predict accurately the fuel properties of a wide range of pressure and temperature conditions.
  arXiv  Detail & Related papers  (2021-10-18T14:43:50Z)
• Physics-informed CoKriging model of a redox flow battery [68.8204255655161]
  Redox flow batteries (RFBs) offer the capability to store large amounts of energy cheaply and efficiently. There is a need for fast and accurate models of the charge-discharge curve of a RFB to potentially improve the battery capacity and performance. We develop a multifidelity model for predicting the charge-discharge curve of a RFB.
  arXiv  Detail & Related papers  (2021-06-17T00:49:55Z)
• Interpretable Data-driven Methods for Subgrid-scale Closure in LES for Transcritical LOX/GCH4 Combustion [0.0]
  The objective of this study is to assess stress models from conventional physics-driven approaches and an interpretable machine learning algorithm. The accuracy of the random-forest regressor decreased when physics-based constraints are applied to the feature set.
  arXiv  Detail & Related papers  (2021-03-11T00:54:50Z)
• Modeling nanoconfinement effects using active learning [1.58501955991843]
  Predicting the spatial configuration of gas molecules in nanopores of shale formations is crucial for fluid flow forecasting and hydrocarbon reserves estimation. We present a method for building and training physics-based deep learning surrogate models to predict molecular configurations of gas inside nanopores.
  arXiv  Detail & Related papers  (2020-05-06T04:01:53Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.