What problem does this paper attempt to address?

The core problem that this paper attempts to solve is: **How to develop a reliable and efficient computational model for predicting the solubility of carbon dioxide (CO₂) in reactive deep eutectic solvents (DESs)**. The solution to this problem is of great significance for accelerating the design and application of reactive deep eutectic solvents to achieve efficient carbon capture. ### Background and Problem Description: 1. **Environmental Challenges**: CO₂ is a harmful greenhouse gas and a major contributor to global warming. To address this environmental challenge, using deep eutectic solvents (DESs) as an environmentally friendly and sustainable medium for effective CO₂ capture has become a promising method. 2. **Limitations of Existing Technologies**: - Physical DESs (which capture CO₂ through physical adsorption) have a relatively low CO₂ absorption capacity. - Reactive DESs (which capture CO₂ by forming chemical bonds) perform better, but currently there is a lack of accurate computational models to predict the solubility of CO₂ in these solvents. 3. **Research Objectives**: Develop a machine - learning (ML) - based model for accurately predicting the solubility of CO₂ in reactive DESs, thereby accelerating the design and application of these solvents. ### Specific Problem Decomposition: 1. **Insufficient Data and Experimental Costs**: Due to the extremely high structural diversity of reactive DESs, relying solely on experimental methods to screen a large number of potential candidate solvents is both time - consuming and expensive. 2. **Deficiencies of Existing Models**: Although the traditional COSMO - RS model can calculate the σ - profile characteristics of the molecular surface, it has a large deviation (up to 195%) when predicting the CO₂ solubility in reactive DESs. 3. **Model Requirements**: A new type of model that combines physically - driven input features (such as COSMO - RS - derived σ - profile features) and machine - learning algorithms is required to improve the prediction accuracy. ### Solution Overview: The paper solves the above problems through the following steps: 1. **Data Collection**: 214 data points were collected from published studies, covering the CO₂ solubility of 149 different reactive DESs at different temperatures, pressures, and molar ratios. 2. **Input Feature Selection**: - Use the σ - profile descriptors calculated by COSMO - RS to quantify the relative probability of molecular surface segments having a specific shielding charge density. - Introduce other physicochemical properties (such as pKa, viscosity, temperature, and pressure) as input features. 3. **Model Development**: Multiple machine - learning algorithms (including artificial neural networks (ANN), support vector regression (SVR), random forest (RF), and gradient - boosted trees (GBT)) were compared, and it was finally found that the ANN model provided the most accurate prediction results (with an average absolute relative deviation of 2.94%). 4. **Model Validation**: The performance of the model was evaluated through statistical indicators (such as the coefficient of determination R², the average absolute relative deviation AARD, the root - mean - square error RMSE, etc.), and it was confirmed that the ANN model performed best on the test set. ### Summary: This paper aims to solve the current difficult problem of predicting the solubility of CO₂ in reactive deep eutectic solvents by developing a hybrid model based on physically - driven features and machine - learning, thereby providing a more efficient and economical solution for carbon - capture technologies.

Physics-Based Machine Learning Models Predict Carbon Dioxide Solubility in Chemically Reactive Deep Eutectic Solvents

Accurate Prediction of Carbon Dioxide Capture by Deep Eutectic Solvents using Quantum Chemistry and a Neural Network

Mathematical Modeling of the Solubility of Carbon Dioxide in Deep Eutectic Solvents

Application of machine learning methods for estimating and comparing the sulfur dioxide absorption capacity of a variety of deep eutectic solvents

Accurate Machine Learning for Predicting the Viscosities of Deep Eutectic Solvents

Computational simulation using machine learning models in prediction of CO2 absorption in environmental applications

Machine learning boosted eutectic solvent design for CO2 capture with experimental validation

Extensive Data Analysis and Modelling of Carbon Dioxide Solubility in Ionic Liquids Using Chemical Structure-Based Ensemble Learning Approaches

Machine learning approach for mapping the heat capacity of deep eutectic solvents for sustainable energy applications

High-Pressure Carbon Dioxide Solubility in a Deep Eutectic Solvent (Choline Chloride/MDEA) + Sulfolane - Experimental Study and Thermodynamic Modeling Using PC-SAFT Equation of State

Comprehensive Evaluation of a Deep Eutectic Solvent Based CO 2 Capture Process through Experiment and Simulation

Rigorous Connectionist Models to Predict Carbon Dioxide Solubility in Various Ionic Liquids

Assessing the Feasibility of Deep Eutectic Solvents For CO2 Capture From Molecular And Process Modeling

Predicting the Surface Tension of Deep Eutectic Solvents Using Artificial Neural Networks

Fibromatosis of the male breast: A case report with immunohistochemistry study and review of the literature.

DESignSolvents: An Open Platform for Search and Prediction of the Physicochemical Properties of Deep Eutectic Solvents

Novel hybrid QSPR-GPR approach for modeling of carbon dioxide capture using deep eutectic solvents

Exploring the performance of machine learning models to predict carbon monoxide solubility in underground pure/saline water

Prediction carbon dioxide solubility in ionic liquids based on deep learning

Estimating the density of deep eutectic solvents applying supervised machine learning techniques

New molecular structure based models for estimation of the CO2 solubility in different choline chloride-based deep eutectic solvents (DESs)