Benchmarking machine learning methods for modeling physical properties of ionic liquids

Prediction and Interpretability of Melting Points of Ionic Liquids Using Graph Neural Networks

Ionic liquids (ILs) have wide and promising applications in fields such as chemical engineering, energy, and the environment. However, the melting points (MPs) of ILs are one of the most crucial properties affecting their applications. The MPs of ILs are affected by various factors, and tuning these in a laboratory is time-consuming and costly. Therefore, an accurate and efficient method is required to predict the desired MPs in the design of novel targeted ILs. In this study, three descriptor-based machine learning (DBML) models and eight graph neural network (GNN) models were proposed to predict the MPs of ILs. Fingerprints and molecular graphs were used to represent molecules for the DBML and GNNs, respectively. The GNN models demonstrated performance superior to that of the DBML models. Among all of the examined models, the graph convolutional model exhibited the best performance with high accuracy (root-mean-squared error = 37.06, mean absolute error = 28.79, and correlation coefficient = 0.76). Benefiting from molecular graph representation, we built a GNN-based interpretable model to reveal the atomistic contribution to the MPs of ILs using a data-driven procedure. According to our interpretable model, amino groups, S+, N+, and P+ would increase the MPs of ILs, while the negatively charged halogen atoms, S-, and N- would decrease the MPs of ILs. The results of this study provide new insight into the rapid screening and synthesis of targeted ILs with appropriate MPs.

Computer-Aided Molecular Design of Ionic Liquids as Advanced Process Media: A Review from Fundamentals to Applications

The unique physicochemical properties, flexible structural tunability, and giant chemical space of ionic liquids (ILs) provide them a great opportunity to match different target properties to work as advanced process media. The crux of the matter is how to efficiently and reliably tailor suitable ILs toward a specific application. In this regard, the computer-aided molecular design (CAMD) approach has been widely adapted to cover this family of high-profile chemicals, that is, to perform computer-aided IL design (CAILD). This review discusses the past developments that have contributed to the state-of-the-art of CAILD and provides a perspective about how future works could pursue the acceleration of the practical application of ILs. In a broad context of CAILD, key aspects related to the forward structure-property modeling and reverse molecular design of ILs are overviewed. For the former forward task, diverse IL molecular representations, modeling algorithms, as well as representative models on physical properties, thermodynamic properties, among others of ILs are introduced. For the latter reverse task, representative works formulating different molecular design scenarios are summarized. Beyond the substantial progress made, some future perspectives to move CAILD a step forward are finally provided.

Application of machine learning techniques to the modeling of solubility of sugar alcohols in ionic liquids

The current trend of chemical industries demands green processing, in particular with employing natural substances such as sugar-derived compounds. This matter has encouraged academic and industrial sections to seek new alternatives for extracting these materials. Ionic liquids (ILs) are currently paving the way for efficient extraction processes. To this end, accurate estimation of solubility data is of great importance. This study relies on machine learning methods for modeling the solubility data of sugar alcohols (SAs) in ILs. An initial relevancy analysis approved that the SA-IL equilibrium governs by the temperature, density and molecular weight of ILs, as well as the molecular weight, fusion temperature, and fusion enthalpy of SAs. Also, temperature and fusion temperature have the strongest influence on the SAs solubility in ILs. The performance of artificial neural networks (ANNs), least-squares support vector regression (LSSVR), and adaptive neuro-fuzzy inference systems (ANFIS) to predict SA solubility in ILs were compared utilizing a large databank (647 data points of 19 SAs and 21 ILs). Among the investigated models, ANFIS offered the best accuracy with an average absolute relative deviation (AARD%) of 7.43% and a coefficient of determination (R²) of 0.98359. The best performance of the ANFIS model was obtained with a cluster center radius of 0.435 when trained with 85% of the databank. Further analyses of the ANFIS model based on the leverage method revealed that this model is reliable enough due to its high level of coverage and wide range of applicability. Accordingly, this model can be effectively utilized in modeling the solubilities of SAs in ILs.

Insights into modeling refractive index of ionic liquids using chemical structure-based machine learning methods

Ionic liquids (ILs) have drawn much attention due to their extensive applications and environment-friendly nature. Refractive index prediction is valuable for ILs quality control and property characterization. This paper aims to predict refractive indices of pure ILs and identify factors influencing refractive index changes. Six chemical structure-based machine learning models called eXtreme Gradient Boosting (XGBoost), Light Gradient Boosting Machine (LightGBM), Categorical Boosting (CatBoost), Convolutional Neural Network (CNN), Adaptive Boosting-Decision Tree (Ada-DT), and Adaptive Boosting-Support Vector Machine (Ada-SVM) were developed to achieve this goal. An enormous dataset containing 6098 data points of 483 different ILs was exploited to train the machine learning models. Each data point's chemical substructures, temperature, and wavelength were considered for the models' inputs. Including wavelength as input is unprecedented among predictions done by machine learning methods. The results show that the best model was CatBoost, followed by XGBoost, LightGBM, Ada-DT, CNN, and Ada-SVM. The R² and average absolute percent relative error (AAPRE) of the best model were 0.9973 and 0.0545, respectively. Comparing this study's models with the literature shows two advantages regarding the dataset's abundance and prediction accuracy. This study also reveals that the presence of the -F substructure in an ionic liquid has the most influence on its refractive index among all inputs. It was also found that the refractive index of imidazolium-based ILs increases with increasing alkyl chain length. In conclusion, chemical structure-based machine learning methods provide promising insights into predicting the refractive index of ILs in terms of accuracy and comprehensiveness.

Conductivity prediction model for ionic liquids using machine learning

Ionic liquids (ILs) are salts, composed of asymmetric cations and anions, typically existing as liquids at ambient temperatures. They have found widespread applications in energy storage devices, dye-sensitized solar cells, and sensors because of their high ionic conductivity and inherent thermal stability. However, measuring the conductivity of ILs by physical methods is time-consuming and expensive, whereas the use of computational screening and testing methods can be rapid and effective. In this study, we used experimentally measured and published data to construct a deep neural network capable of making rapid and accurate predictions of the conductivity of ILs. The neural network is trained on 406 unique and chemically diverse ILs. This model is one of the most chemically diverse conductivity prediction models to date and improves on previous studies that are constrained by the availability of data, the environmental conditions, or the IL base. Feature engineering techniques were employed to identify key chemo-structural characteristics that correlate positively or negatively with the ionic conductivity. These features are capable of being used as guidelines to design and synthesize new highly conductive ILs. This work shows the potential for machine-learning models to accelerate the rate of identification and testing of tailored, high-conductivity ILs.

Thermoresponsive Ionic Liquid/Water Mixtures: From Nanostructuring to Phase Separation

The thermodynamics, structures, and applications of thermoresponsive systems, consisting primarily of water solutions of organic salts, are reviewed. The focus is on organic salts of low melting temperatures, belonging to the ionic liquid (IL) family. The thermo-responsiveness is represented by a temperature driven transition between a homogeneous liquid state and a biphasic state, comprising an IL-rich phase and a solvent-rich phase, divided by a relatively sharp interface. Demixing occurs either with decreasing temperatures, developing from an upper critical solution temperature (UCST), or, less often, with increasing temperatures, arising from a lower critical solution temperature (LCST). In the former case, the enthalpy and entropy of mixing are both positive, and enthalpy prevails at low T. In the latter case, the enthalpy and entropy of mixing are both negative, and entropy drives the demixing with increasing T. Experiments and computer simulations highlight the contiguity of these phase separations with the nanoscale inhomogeneity (nanostructuring), displayed by several ILs and IL solutions. Current applications in extraction, separation, and catalysis are briefly reviewed. Moreover, future applications in forward osmosis desalination, low-enthalpy thermal storage, and water harvesting from the atmosphere are discussed in more detail.