Physics-guided cascade prediction of CO₂ solubility in ionic liquids for extrapolation to unseen ionic structures

BAO Yingzheng; ZHOU Huairong; YANG Yong; WANG Dongliang; YANG Qingchun

doi:10.11949/0438-1157.20260686

您当前的位置：

首页 >

文章列表页 >

Physics-guided cascade prediction of CO₂ solubility in ionic liquids for extrapolation to unseen ionic structures

更新时间：2026-07-17

- Physics-guided cascade prediction of CO₂ solubility in ionic liquids for extrapolation to unseen ionic structures
- CIESC Journal (2026)
- 作者机构：
  
  1.兰州理工大学石油化工学院，甘肃兰州 730050
  2.合肥工业大学化学与化工学院，安徽合肥 230009
- 作者简介：
- 基金信息：
- DOI：10.11949/0438-1157.20260686
  CLC： O645.12;TQ028.2;TP181
- Received：18 May 2026，
  
  Revised：2026-07-17，
  
  Accepted：17 July 2026，
- 稿件说明：
移动端阅览
BAO Yingzheng, ZHOU Huairong, YANG Yong, et al. Physics-guided cascade prediction of CO₂ solubility in ionic liquids for extrapolation to unseen ionic structures[J/OL]. CIESC Journal, 2026.
DOI：

BAO Yingzheng, ZHOU Huairong, YANG Yong, et al. Physics-guided cascade prediction of CO₂ solubility in ionic liquids for extrapolation to unseen ionic structures[J/OL]. CIESC Journal, 2026. DOI： 10.11949/0438-1157.20260686.

摘要

针对统一回归模型难以同时表征低溶解度尾部、低压近Henry区和高压非理想区的问题，构建融合COSMO-RS Henry先验、Apparent Henry校准、分类门控和压力路由的物理引导级联模型。基于932124个COSMO-RS Henry数据和297种离子液体的22294个实验数据进行建模。分类门控后，低压和高压分支测试集决定系数（coefficient of determination，R²）分别为0.905和0.865；完整模型在离子液体配对（ionic-liquid pair，IL-pair）、Cation和Anion划分下的R²分别为0.823、0.591和0.780，较最佳基线提高0.189、0.048和0.473。该模型可用于高CO₂溶解度离子液体的计算初筛，但仍需独立实验及多物性评价。

Abstract

A physics-guided cascade model integrating a COSMO-RS Henry prior

apparent-Henry calibration

classifiCation gating

and pressure-based routing was developed to address the difficulty of simultaneously describing the low-solubility tail

the low-pressure near-Henry region

and the high-pressure non-ideal region using a unified regression model. The framework was constructed using 932

124 COSMO-RS Henry-constant data points and 22

294 experimental CO₂-solubility data points for 297 ionic liquids. After classifiCation gating

the low- and high-pressure branches achieved test R² values of 0.905 and 0.865

respectively. The complete cascade model achieved test R² values of 0.823

0.591

and 0.780 under the IL-pair

Cation

and Anion splits

representing improvements

关键词

Keywords

references

崔国凯 , 吕书贞 , 王键吉 . 功能化离子液体在二氧化碳吸收分离中的应用 [J ] . 化工学报 , 2020 , 71 ( 1 ): 16 - 25 .

Cui G K , Lyu S Z , Wang J J . Functional ionic liquids for carbon dioxide capture and separation [J ] . CIESC Journal , 2020 , 71 ( 1 ): 16 - 25 .

Kuroki N , Suzuki Y , Kodama D , et al . Machine learning-boosted design of ionic liquids for CO 2 absorption and experimental verifiCation [J ] . The Journal of Physical Chemistry B , 2023 , 127 ( 9 ): 2022 - 2027 .

Ali M , Sarwar T , Mubarak N M , et al . Prediction of CO 2 solubility in Ionic liquids for CO 2 capture using deep learning models [J ] . Scientific Reports , 2024 , 14 : 14730 .

Laakso J P , Gorji A E , Uusi-Kyyny P , et al . Machine learning model ing of the CO 2 solubility in ionic liquids by using σ-profile descriptors [J ] . Chemical Engineering Science , 2025 , 307 : 121226 .

Kong X Y , Qiu Y X , Wang K , et al . Deep learning driven ionic liquid screening for toluene and styrene capture [J ] . Chemical Engineering Science , 2026 , 320 : 122724 .

张瑞娜 , 田媛 , 葛春亮 , 等 . 人工智能机器学习助力离子液体CO 2 捕集性能分析 [J ] . 科学通报 , 2025 , 70 ( 26 ): 4462 - 4472 .

Zhang R N , Tian Y , Ge C L , et al . Artificial intelligence and machine learning for analysis of CO 2 capture by ionic liquids [J ] . Chinese Science Bulletin , 2025 , 70 ( 26 ): 4462 - 4472 .

张威 , 张瑞娜 , 葛春亮 , 等 . 功能化离子液体活性位点强化CO 2 捕集 [J ] . 低碳化学与化工 , 2024 , 49 ( 7 ): 23 - 33, 41 .

Zhang W , Zhang R N , Ge C L , et al . IntensifiCation of CO 2 capture by active sites in functional ionic liquids [J ] . Low-Carbon Chemistry and Chemical Engineering , 2024 , 49 ( 7 ): 23 - 33, 41 .

Lei Z G , Dai C N , Chen B H . Gas solubility in ionic liquids [J ] . Chemical Reviews , 2014 , 114 ( 2 ): 1289 - 1326 .

Eckert F , Klamt A . Fast solvent screening via quantum chemistry: COSMO-RS approach [J ] . AIChE Journal , 2002 , 48 ( 2 ): 369 - 385 .

Zhang X C , Liu Z P , Wang W C . Screening of ionic liquids to capture CO 2 by COSMO-RS and experiments [J ] . AIChE Journal , 2008 , 54 ( 10 ): 2717 - 2728 .

Sistla Y S , Khanna A . Validation and prediction of the temperature-dependent henry’s constant for CO 2 –ionic liquid systems using the conductor-like screening model for realistic solvation (COSMO-RS) [J ] . Journal of Chemical & Engineering Data , 2011 , 56 ( 11 ): 4045 - 4060 .

Moya C , Gonzalez-Miquel M , Rodriguez F , et al . Non-ideal behavior of ionic liquid mixtures to enhance CO 2 capture [J ] . Fluid Phase Equilibria , 2017 , 450 : 175 - 183 .

Kapoor U , Shah J K . Molecular origins of the apparent ideal CO 2 solubilities in binary ionic liquid mixtures [J ] . The Journal of Physical Chemistry B , 2018 , 122 ( 42 ): 9763 - 9774 .

Karmakar A , Mukundan R . Modeling solubility of CO 2 gas in room temperature ionic liquids using the COSMOSAC-LANL model: a first principles study [J ] . Physical Chemistry Chemical Physics , 2019 , 21 ( 35 ): 19667 - 19685 .

Liu X Y , O’Harra K E , Bara J E , et al . Solubility behavior of CO 2 in ionic liquids based on ionic polarity index analyses [J ] . The Journal of Physical Chemistry B , 2021 , 125 ( 14 ): 3665 - 3676 .

王伟彬 , 银建中 , 孙丽华 , 等 . CO 2 /离子液体体系热力学性质的分子动力学模拟 [J ] . 物理化学学报 , 2009 , 25 ( 11 ): 2291 - 2295 .

Wang W B , Yin J Z , Sun L H , et al . Molecular dynamics simulation of thermodynamic properties for CO 2 /ionic liquid systems [J ] . Acta Physico-Chimica Sinica , 2009 , 25 ( 11 ): 2291 - 2295 .

Rogers D , Hahn M . Extended-connectivity fingerprints [J ] . Journal of Chemical Information and Modeling , 2010 , 50 ( 5 ): 742 - 754 .

Dong Q , Muzny C D , Kazakov A , et al . ILThermo: a free-access web database for thermodynamic properties of ionic liquids [J ] . Journal of Chemical & Engineering Data , 2007 , 52 ( 4 ): 1151 - 1159 .

Chen X X , Chen G Z , Xie K C , et al . Exploring the chemical space of ionic liquids for CO 2 dissolution through generative machine learning models [J ] . Green Chemical Engineering , 2025 , 6 ( 3 ): 335 - 343 .

Klamt A , Eckert F . COSMO-RS: a novel and efficien t method for the a priori prediction of thermophysical data of liquids [J ] . Fluid Phase Equilibria , 2000 , 172 ( 1 ): 43 - 72 .

Fan J , Dai Z X , Cao J , et al . Hybrid data-driven and physics-based modeling for viscosity prediction of ionic liquids [J ] . Green Energy & Environment , 2024 , 9 ( 12 ): 1878 - 1890 .

Blanchard L A , Gu Z Y , Brennecke J F . High-pressure phase behavior of ionic liquid/CO 2 systems [J ] . The Journal of Physical Chemistry B , 2001 , 105 ( 12 ): 2437 - 2444 .

Baltus R E , Culbertson B H , Dai S , et al . Low-pressure solubility of carbon dioxide in room-temperature ionic liquids measured with a quartz crystal microbalance [J ] . The Journal of Physical Chemistry B , 2004 , 108 ( 2 ): 721 - 727 .

Chen T Q , Guestrin C . XGBoost: a scalable tree boosting system [C ] // Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining . San Francisco California USA . ACM , 2016 : 785 - 794 .

Akiba T , Sano S , Yanase T , et al . Optuna: a next-generation hyperparameter optimization framework [C ] // Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining . Anchorage AK USA . ACM , 2019 : 2623 - 2631 .

Specht T , Nagda M , Fellenz S , et al . HANNA: hard-constraint neural network for consistent activity coefficient prediction [J ] . Chemical Science , 2024 , 15 ( 47 ): 19777 - 19786 .

Qin H L , Wang K , Ma X F , et al . Predicting the solubility of CO 2 and N 2 in ionic liquids based on COSMO-RS and machine learning [J ] . Frontiers in Chemistry , 2024 , 12 : 1480468 .

Shiflett M B , Yokozeki A . Solubilities and diffusivities of carbon dioxide in ionic liquids: [bmim ] [PF 6 ] and [bmim ] [BF4 ] [J ] . Industrial & Engineering Chemistry Research, 2005, 44 ( 12 ): 4453 - 4464 .

Breiman L . Random forests [J ] . Machine Learning , 2001 , 45 ( 1 ): 5 - 32 .

Cadena C , Anthony J L , Shah J K , et al . Why is CO 2 so soluble in imidazolium-based ionic liquids [J ] . Journal of the American Chemical Society , 2004 , 126 ( 16 ): 5300 - 5308 .

Yang K , Swanson K , Jin W G , et al . Analyzing learned molecular representations for property prediction [J ] . Journal of Chemical Information and Modeling , 2019 , 59 ( 8 ): 3370 - 3388 .

吕超 , 张习文 , 金理健 , 等 . 新型两相吸收剂-离子液体系统高效捕获CO 2 [J ] . 化工进展 , 2023 , 42 ( 6 ): 3226 - 3232 .

Lü C , Zhang X W , Jin L J , et al . Efficient capture of CO 2 by a new biphasic solvent-ionic liquid system [J ] . Chemical Industry and Engineering Progress , 2023 , 42 ( 6 ): 3226 - 3232 .

Views

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

Functionalized ionic liquids based on [EMIM][OAC] for vapor-liquid phase equilibrium of acetonitrile and water by COSMO-RS method

One-step extraction and separation of high purity chitin based on choline ionic liquid

Extraction of bicyclic S/N-compounds from FCC diesel with ionic liquid: experimental and molecular insight

Hypergolic ignition characteristics and propulsion performance of imidazolium dicyanamide ionic liquids blended with furfuryl alcohol

Recent progress on application of COSMO-RS model in screening of ionic liquids/deep eutectic solvents

Related Author

LI Tingting

YANG Qing

PENG Changjun

LIU Honglai

FENG Mi

ZHANG Jie

LYU Xingmei

JIANG Siqi

Related Institution

State Key Laboratory of Chemical Engineering, Department of Chemistry, East China University of Science and Technology

Institute of Process Engineering, Chinese Academy of Sciences

Longzihu New Energy Laboratory, Zhengzhou Institute of Emerging Industrial Technology

Beijing BW Techtextile Company Limited by Shares

Hubei Key Laboratory of Biomass Fibers and Eco-dyeing & Finishing, College of Chemistry and Chemical Engineering, Wuhan Textile University

⁰