Publications

Abstract Heat capacity, a crucial physical property for chemical processes, is often understudied in Deep Eutectic Solvents (DESs), which in turn are promising green alternatives to environmentally hazardous conventional solvents. This work addresses this gap by developing a machine learning model to predict DES heat capacity and identify key structural features influencing it. We employed a dataset of 530 DESs with corresponding experimental heat capacity values. Quantum-chemical COSMO-RS-based descriptors, capturing detailed information about DES structures, were calculated for each data point. Various machine learning algorithms, namely k-Nearest Neighbours (kNN), Random Forests (RF), Neural Network Multilayer Perceptron (MLP), and Support Vector Machines (SVM) were explored alongside a linear model (Multiple Linear Regression, MLR). Hyperparameter optimisation ensured all models were fine-tuned for optimal performance. The most successful model, based on the MLP technique, achieved remarkably low Average Absolute Relative Deviation (AARD) values of 0.500 % and 3.999 % for the training and test sets, respectively. This signifies a significant improvement in prediction accuracy compared to traditional methods. Furthermore, by applying a SHapley Additive exPlanations (SHAP) analysis, we identified the most crucial structural factors within DES components that govern their heat capacity. This comprehensive investigation offers valuable insights that can pave the way for an efficient design of novel DESs in the future. © 2024 The Author(s)

Abstract Magnetic ionic liquids (MILs) stand out as a remarkable subclass of ionic liquids (ILs), combining the desirable features of traditional ILs with the unique ability to respond to external magnetic fields. The incorporation of paramagnetic species into their structures endows them with additional attractive features, including thermochromic behavior and luminescence. These exceptional properties position MILs as highly promising materials for diverse applications, such as gas capture, DNA extractions, and sensing technologies. The present Review synthesizes key experimental findings, offering insights into the structural, thermal, magnetic, and optical properties across various MIL families. Special emphasis is placed on unraveling the influence of different paramagnetic species on MILs' behavior and functionality. Additionally, the Review highlights recent advancements in computational approaches applied to MIL research. By leveraging molecular dynamics (MD) simulations and density functional theory (DFT) calculations, these computational techniques have provided invaluable insights into the underlying mechanisms governing MILs' behavior, facilitating accurate property predictions. In conclusion, this Review provides a comprehensive overview of the current state of research on MILs, showcasing their special properties and potential applications while highlighting the indispensable role of computational methods in unraveling the complexities of these intriguing materials. The Review concludes with a forward-looking perspective on the future directions of research in the field of magnetic ionic liquids.

Abstract Electrolytes play a crucial role in enhancing the performance of energy storage devices, including batteries and supercapacitors. However, traditional electrolytes, such as aqueous solutions, organic solvents, and ionic liquids, exhibit inherent limitations and challenges. Deep eutectic solvents have recently emerged as promising alternatives due to their environmentally friendly nature and favorable properties. Despite their widespread applications in various domains, their potential as electrolytes remains relatively underexplored. This study investigates three distinct types of deep eutectic solvents derived from different isomers of butanediols combined with choline chloride. Ab initio molecular dynamics simulations are employed to analyze the microstructure of these deep eutectic solvents, focusing on non-covalent electrostatic interactions, hydrogen bonding patterns, and vibrational spectra. The results reveal significant differences in the structural configuration of hydrogen bond acceptors and hydrogen bond donors and their interactions within the deep eutectic solvents. Specifically, the positioning of functional groups in hydrogen bond donors significantly impacts the hydrogen bonding network and the interaction with monoatomic ions. Moreover, the vibrational spectra analysis highlights the existence of hydrogen bonds involving stretching modes of the OH group, as evidenced by redshift deviations. Overall, this study provides valuable insights into the unique features of deep eutectic solvents as potential electrolytes for energy storage applications. The comprehensive analysis of their microstructure and vibrational properties enhances our understanding of deep eutectic solvent utilization and opens avenues for further research in sustainable energy storage.

Abstract Technologies involving a solvent|surface interface, such as nanotechnology, electrochemistry, and energy storage applications, are actively pursuing ecologically responsible and sustainable development practices. In response to this pressing need, deep eutectic solvents have emerged as a promising solution to bridge the gap between technological requirements and environmental concerns. In this work, we present the results of a molecular dynamics simulation study of the interface between a monocrystalline gold surface and the deep eutectic solvent ethaline, where a molar ratio of 1:2 choline chloride:ethylene glycol was used for ethaline. The simulations covered a range of temperatures from 313 K to 343 K and applied charge values ranging from 0 to +/- 24 mu C cm-2. Several key interfacial properties were thoroughly analyzed, including among others, charge density profiles, radial distribution functions, hydrogen bond close contacts, and molecular orientation. Additionally, we examined how the differential capacitance varied upon the applied potential. Our findings reveal that, at neutral surfaces, all components of the solvent are present in the innermost layer, with ethylene glycol molecules being the most prevalent, followed by choline cations and a residual amount of chloride anions. For lower applied charges, this mixed composition at the boundary layer persists, despite the growing accumulation of ionic species with charges opposite to that of the electrode. As surface polarization increases, unique innermost boundary layers composed exclusively of one of the ionic species and the hydrogen bond donor molecules are observed, forming a multilayer structure, with subsequent layers enriched of paired counterions. Interestingly, even at higher applied charges, choline cations and ethylene glycol molecules tended to orient themselves in a parallel fashion toward the electrodes. Differential capacitance curves exhibited a camel-shaped behavior, suggesting a complex interplay of electrochemical processes at the DES|Au(100) interface. In summary, our study provides valuable insights into the interfacial properties of deep eutectic solvents on gold surfaces and their response to changes in temperature and potential, which are crucial for understanding and optimizing deep eutectic solventbased electrochemical systems.

Abstract New molecular dynamics (MD) simulations and experimental data on a deep eutectic solvent, propeline, composed by choline chloride, ChCl, and propylene glycol, PG, in a molar ratio of 1:2 are reported in this work. The experimental physicochemical properties (density, viscosity and self-diffusion coefficients) were used as support in the development of a new OPLS based force field model (FFM) for propeline. Validation of the new force field was established both through measuring physicochemical properties over a range of temperatures (298.15-373.15 K) and by comparison with experimental and simulated data of ethaline (ChCl:ethylene glycol, at a molar ration of 1:2). Classical MD simulations using the new FFM led to good agreement between experimental and simulated data. Structural properties, namely radial and spatial distribution functions, coordination numbers, and hydrogen bonding were analyzed. Moreover, it was found that the interactions between the anion, Cl-, and the hydrogen bond donor (HBD) form a network that is immutable with increasing temperature. The higher prevalence of anion-HBD hydrogen bonds is likely the major reason for the relatively high viscosity of propeline.

Abstract Despite a growing number of research reports on neat room temperature ionic liquids (RTILs) and their mixtures with molecular solvents in recent years, understanding and rationalising of such systems is still a challenge. In this work, we performed a classical molecular dynamics simulation study of the pure components - 1-ethyl-3-methylimidazolium thiocyanate (C2C1 imSCN), methanol, and ethanol - and their binary mixtures at room temperature. Thermodynamic (density and heats of vaporization), transport (viscosity and self-diffusion coefficients) and structural (in terms of radial, angular and spatial distributions) properties were analysed. It was found, that with the decrease of RTIL content, the ions self-diffusion coefficients notably increase, reaching higher values in the C2C1 imSCN-MeOH system. Density and viscosity follow the opposite trend, reaching their minimum at lower RTIL mole fraction. Negative deviations of excess molar volume from ideality in the studied mixtures with minima at similar to 0.2-03 mole fraction of RTIL suggest the strongest ion-molecular interactions at this mixture composition. A careful analysis at the molecular level revealed that introducing of alcohols to both systems weakens the inter-ionic H-bonding network, particularly, at low RTIL content. The cation-cation arrangement was found to lose its characteristic above/below orientation in neat RTIL and become disordered at low RTIL content. As to the tail length of the selected alcohols, this was found to have an insignificant effect on the structural properties of the addressed systems.

Abstract Mixing of ionic liquids provides new opportunities for their tuning, enabling the applications of ionic liquid mixtures to expand. At the same time, the genesis of the fundamental properties of ionic liquid mixtures is still poorly understood. In this study we carried out a molecular dynamics simulation of binary mixtures of 1-buthyl-3-methylimidazolium hexafluorophosphate, 1-buthyl-3-methylimidazolium bis(trifluoromethylsulfonyl)imide, and 1-buthyl-3-methylimidazolium tris(perfluoroethyl)trifluorophosphate ([C(4)mim][PF6] + [C(4)mim][NTf2], [C(4)mim][PF6] + [C(4)mim][FAP], [C(4)mim][FAP] + [C(4)mim][NTf2]) in a wide concentration range at 303.15 K and complemented it with quantum mechanical calculations. Three pure ionic liquids underwent the same kind of analysis for comparison purposes. We found that the addition of the [FAP](-)-anion to a mixture enhances the segregation of non-polar domains and weakens the hydrogen-bond network. The H-bonds in the studied mixtures are rather weak, as follows from QTAIM analysis, with the rarest occurrence for the [FAP](-)-anion. The competition of two anions in the mixtures for the most acidic hydrogen of the 1-butyl-3-methylimidazolium cation is reported. In most of the cases, the smaller anion ([PF6](-) or [NTf2](-)) with stronger charge concentration displaces the bigger one ([NTf2](-) or [FAP](-)) from the preferred coordination site. The existing nano-segregation in some mixtures notably slows down ion diffusion. Our results show that the differences in anion size, shape and nature are the main reasons for nano-segregation and the non-ideal behavior of ionic liquid mixtures.

Abstract In this work we developed a new force field model (FFM) for propylene glycol (PG) based on the OPLS all-atom potential. The OPLS potential was refined using quantum chemical calculations, taking into account the densities and self-diffusion coefficients. The validation of this new FFM was carried out based on a wide range of physicochemical properties, such as density, enthalpy of vaporization, self-diffusion coefficients, isothermal compressibility, surface tension, and shear viscosity. The molecular dynamics (MD) simulations were performed over a large range of temperatures (293.15-373.15 K). The comparison with other force field models, such as OPLS, CHARMM27, and GAFF, revealed a large improvement of the results, allowing a better agreement with experimental data. Specific structural properties (radial distribution functions, hydrogen bonding and spatial distribution functions) were then analyzed in order to support the adequacy of the proposed FFM. Pure propylene glycol forms a continuous phase, displaying no microstructures. It is shown that the developed FFM gives rise to suitable results not only for pure propylene glycol but also for mixtures by testing its behavior for a 50 mol % aqueous propylene glycol solution. Furthermore, it is demonstrated that the addition of water to the PG phase produces a homogeneous solution and that the hydration interactions prevail over the propylene glycol self-association interactions.

Abstract In this work, we combined various parameters found in the literature for the choline cation, chloride anion, and ethylene glycol to set up force field models (FFMs) for a eutectic mixture, namely, ethaline (1:2 choline chloride/ethylene glycol (ChCl:2EG)). The validation of these models was carried out on the basis of physical and chemical properties, such as the density, expansion coefficient, enthalpy of vaporization, self-diffusion coefficients, isothermal compressibility, surface tension, and shear viscosity. After the initial evaluation of the FFMs, a refinement was found necessary and accomplished by taking into account polarization effects in a mean-field manner. This was achieved by rescaling the electrostatic charges of the ions based on partial charges derived from ab initio molecular dynamics (MD) simulations of the bulk system. Classical all-atom MD simulations performed over a large range of temperatures (298.15-373.15 K) using the refined FFMs clearly showed improved results, allowing a better prediction of experimental properties. Specific structural properties (radial distribution functions and hydrogen bonding) were then analyzed in order to support the adequacy of the proposed refinement. The final selected FFM leads to excellent agreement between simulated and experimental data on dynamic and structural properties. Moreover, compared to the previously reported force field model (Perkins, S. L.; Painter, P.; Colina, C. M. Experimental and Computational Studies of Choline Chloride-Based Deep Eutectic Solvents. J. Chem. Eng. Data 2014, 59, 3652-3662), a 10% improvement in simulated transport properties, i.e., self-diffusion coefficients, was achieved. The isothermal compressibility, surface tension, and shear viscosity for ethaline are accessed in MD simulations for the first time.