Machine learning approaches to understand and predict rate constants for organic processes in mixtures containing ionic liquids.

The ability to tailor the constituent ions in ionic liquids (ILs) is highly advantageous as it provides access to solvents with a range of physicochemical properties. However, this benefit also leads to large compositional spaces that need to be explored to optimise systems, often involving time consuming experimental work. The use of machine learning methods is an effective way to gain insight based on existing data, to develop structure-property relationships and to allow the prediction of ionic liquid properties. Here we have applied machine learning models to experimentally determined rate constants of a representative organic process (the reaction of pyridine with benzyl bromide) in IL-acetonitrile mixtures. Multiple linear regression (MLREM) and artificial neural networks (BRANNLP) were both able to model the data well. The MLREM model was able to identify the structural features on the cations and anions that had the greatest effect on the rate constant. Secondly, predictive MLREM and BRANNLP models were developed from the full initial set of rate constant data. From these models, a large number of predictions (>9000) of rate constant were made for mixtures of different ionic liquids, at different proportions of ionic liquid and molecular solvent, at different temperatures. A selection of these predictions were tested experimentally, including through the preparation of novel ionic liquids, with overall good agreement between the predicted and experimental data. This study highlights the benefits of using machine learning methods on kinetic data in ionic liquid mixtures to enable the development of rigorous structure-property relationships across multiple variables simultaneously, and to predict properties of new ILs and experimental conditions.

Controlling the outcome of SN2 reactions in ionic liquids: from rational data set design to predictive linear regression models.

Rate constants for a bimolecular nucleophilic substitution (SN2) process in a range of ionic liquids are correlated with calculated parameters associated with the charge localisation on the cation of the ionic liquid (including the molecular electrostatic potential). Simple linear regression models proved effective, though the interdependency of the descriptors needs to be taken into account when considering generality. A series of ionic liquids were then prepared and evaluated as solvents for the same process; this data set was rationally chosen to incorporate homologous series (to evaluate systematic variation) and functionalities not available in the original data set. These new data were used to evaluate and refine the original models, which were expanded to include simple artificial neural networks. Along with showing the importance of an appropriate data set and the perils of overfitting, the work demonstrates that such models can be used to reliably predict ionic liquid solvent effects on an organic process, within the limits of the data set.

Understanding the effects of solvate ionic liquids as solvents on substitution processes.

The effects of solvate ionic liquids as solvents have been considered for two substitution processes where the solvent effects of typical ionic liquids have been extensively investigated previously; the bimolecular nucleophilic substitution (SN2) reaction between pyridine and benzyl bromide and the nucleophilic aromatic substitution (SNAr) reaction between ethanol and 1-fluoro-2,4-dinitrobenzene. It was found that use of solvate ionic liquids gave rise to similar trends in the activation parameters for both substitution processes as typical ionic liquids, implying the microscopic interactions responsible for the effects were the same. However, different effects on the rate constants compared to typical ionic liquids were observed due to the changes in the balance of enthalpic and entropic contributions to the observed rate constants. From these data it is clear that the reaction outcome for both of these substitution reactions fall within the 'predictive framework' established in previous studies with a cautionary tale or two of their own to add to the general knowledge of ionic liquid solvent effects for these processes, particularly with respect to potential reactivity of the solvate ionic liquids themselves.

NMR Diffusion Measurements as a Simple Method to Examine Solvent-Solvent and Solvent-Solute Interactions in Mixtures of the Ionic Liquid [Bmim][N(SO2 CF3 )2 ] and Acetonitrile.

The self-diffusion coefficients of each component in mixtures of 1-butyl-3-methylimidazolium bis(trifluoromethanesulfonyl)imide ([Bmim][N(SO2 CF3 )2 ]) and acetonitrile were determined. The results suggest that the hydrodynamic boundary conditions change from "stick" to "slip" as the solvent composition transitions from "ionic liquid dissolved in acetonitrile" (χIL <0.4) to "acetonitrile dissolved in ionic liquid" (χIL >0.4). At higher χIL , the acetonitrile species are affected by "cage" and "jump" events, as the acetonitrile molecules reside nearer to the charged centre on the ions than in the "non-polar" regions. The self-diffusion coefficients of hexan-1-amine, dipropylamine, 1-hexanol and dipropylether in mixtures of [Bmim][N(SO2 CF3 )2 ] and acetonitrile were determined. In general, the nitrogen-containing solutes were found to diffuse slower than the oxygen-containing solutes; this indicates that there are greater ionic liquid-N interactions than ionic liquid-O interactions. This work demonstrates that the self-diffusion coefficients of species can provide valuable information about solvent-solvent and solvent-solute interactions in mixtures containing an ionic liquid.

Developing principles for predicting ionic liquid effects on reaction outcome. A demonstration using a simple condensation reaction.

The effects of a series of ionic liquids, with systematic variations in the cation, on the condensation of an alkyl amine with an aromatic aldehyde were investigated, and the outcomes compared with those predicted based on related reactions. The addition of ionic liquids increased the observed rate constant; the mole fraction dependence of this increase was qualitatively consistent with predictions. Temperature-dependent kinetic analyses were used to demonstrate that the microscopic origins of the effects were as forecast, though the relative weighting of enthalpic and entropic contributions was dependent on the salt used.

The Dependence of Ionic Liquid Solvent Effects on the Nucleophilic Heteroatom in SN Ar Reactions. Highlighting the Potential for Control of Selectivity.

Nucleophilic aromatic substitution (SN Ar) reactions of 1-fluoro-4-nitrobenzene using similar nitrogen and sulfur nucleophiles were studied through extensive kinetic analysis in mixtures containing ionic liquids. The interactions of the ionic liquid components with the starting materials and transition state for each process were investigated in an attempt to construct a broad predictive framework for how ionic liquids affect reaction outcome. It was found that, based on the activation parameters, the microscopic interactions and thus the ionic liquid solvent effect were different for each of the nucleophiles considered. The results from this study suggest that it may be possible to rationally select a given ionic liquid mixture to selectively control reaction outcome of an SN Ar reaction where multiple nucleophiles are present.

Investigating Variation of the Pnicogen Nucleophilic Heteroatom on Ionic Liquid Solvent Effects in Bimolecular Nucleophilic Substitution Processes.

A series of nucleophiles containing Group 15 nucleophilic heteroatoms has been used to expand and develop the current understanding of ionic liquid solvent effects on bimolecular nucleophilic substitution processes. It was found that when using arsenic-, antimony- and bismuth-based nucleophiles, rate constant enhancement was observed for all solvent compositions containing ionic liquids. This rate constant enhancement was driven by ionic liquid/transition state interactions, which contrasts with previous studies on earlier Group 15 nucleophiles. This study provides a holistic understanding and augments the predictive framework for the effects of ionic liquids on bimolecular nucleophilic substitution processes, with the potential for these periodic trends to be broadly applied.

Ionic Liquids as Solvents for SN 2 Processes. Demonstration of the Complex Interplay of Interactions Resulting in the Observed Solvent Effects.

Bimolecular nucleophilic substitution reactions between triphenylphosphine and benzylic electrophiles have been examined in an ionic liquid to probe interactions with species along the reaction coordinate. Trends in the rate constant were found on both varying the leaving group and the electronic nature of the aromatic ring. In all the cases considered, interactions between the components of the ionic liquid and the transition state were shown to be more significant in determining reaction outcome than previously observed for this class of reaction. This demonstrates the importance of considering interactions of the ionic liquid components with all species along the reaction coordinate when investigating the origin of ionic liquid solvent effects, along with how such effects might be exploited.