Marco Klähn

What determines the miscibility of ionic liquids with water? Identification of the underlying factors to enable a straightforward prediction.

Whether an ionic liquid (IL) is water-miscible or immiscible depends on the particular ions that constitute it. We propose an explanation, based on molecular simulations, how ions determine the miscibility of ILs and suggest a straightforward and computationally inexpensive method to predict the miscibility of arbitrary new ILs. The influence of ions on the solvation of water is analyzed by comparing molecular dynamics simulations of water in 9 different ILs with varying cation and anion constituents. The solvation of water in ILs is found to depend primarily on the electrostatic water-ion interaction strength, which, in turn, is determined mainly by two factors: primarily, by the size of the ions and secondarily by the amount of charge on the ion surface that is coordinated with water. It is demonstrated that large ions lead to weaker interactions with water, due to the involved delocalization of the ion charge. A large charge on the ion surface, which is determined by the chemical structure of the ion, strengthens water-ion interactions. We observe that whenever the interaction strength of water with ions exceeds a certain threshold, an IL becomes water-miscible. On the basis of these findings, a simple equation is derived that estimates the water-ion interaction strength. With this equation it is possible to predict most of the observed water-miscibilities of a sample of 83 ILs correctly. A linear increase of the water saturation concentration with the estimated water-ion interaction strength is observed in water-immiscible ILs, which can be utilized to predict the water concentration in new ILs.

A Force Field for Guanidinium-Based Ionic Liquids That Utilizes the Electron Charge Distribution of the Actual Liquid: A Molecular Simulation Study

We propose a new all-atom force field for guanidinium-based ionic liquids (GILs) which is based on the charge distribution in the actual liquid. It comprises all cations that can be built by attaching alkyl chains of variable length to an acyclic or cyclic guanidinium compound and that are paired with nitrate or perchlorate anions. We based the parametrization of the force field on liquid-phase charge distributions to improve the prediction of energetic and dynamic properties of GILs. The impact of electron charge transfer and polarization on various properties of GILs is systematically assessed. A significant average electron charge transfer between -0.12 and -0.06 e from anions to the central guanidinium group of the cations and a strong polarization of acyclic cations are observed by applying a combined quantum mechanical/molecular mechanical (QM/MM) approach. Molecular dynamics simulations of GILs are performed, utilizing the proposed force field. Derived structures approach the accuracy of QM/MM structures, and a previously reported crystal structure remains stable throughout the simulations. Mass densities are reproduced with a deviation of only 1.4% from experimental data. The calculated melting point of a GIL crystal deviates only 8% from the measured value. Self-diffusion coefficients of various GILs are reported, and a comparison with a diffusion coefficient derived from experimental data indicates that the values are within a reasonable range. We observe that the melting point of a GIL crystal was lowered up to 60 K and that diffusion coefficients are substantially increased by a factor of up to 3.5 upon consideration of charge transfer and polarization. The results demonstrate that liquid-phase partial charges are capable of improving the quality of ionic liquid force fields substantially and that their utilization led to a model that can be applied to predict structural, energetic, and dynamic properties of GILs.

What Determines CO2 Solubility in Ionic Liquids? A Molecular Simulation Study

Molecular dynamics (MD) simulations of 10 different pure and CO2-saturated ionic liquids are performed to identify the factors that determine CO2 solubility. Imidazolium-based cations with varying alkyl chain length and functionalization are paired with anions of different hydrophobicity and size. Simulations are carried out with an empirical force field based on liquid-phase charges. The partial molar volume of CO2 in ionic liquids (ILs) varies from 30 to 40 cm(3)/mol. This indicates that slight ion displacements are necessary to enable CO2 insertions. However, the absorption of CO2 does not affect the overall organization of ions in the ILs as demonstrated by almost equal cation-anion radial distribution functions of pure ILs and ILs saturated with CO2. The solubility of CO2 in ILs is not influenced by direct CO2-ion interactions. Instead, a strong correlation between the ratio of unoccupied space in pure ILs and their ability to absorb CO2 is found. This preformed unoccupied space is regularly dispersed throughout the ILs and needs to be expanded by slight ion displacements to accommodate CO2. The amount of preformed unoccupied space is a good indicator for ion cohesion in ILs. Weak electrostatic cation-anion interaction densities in ILs, i.e., weak ion cohesion, leads to larger average distances between ions and hence to more unoccupied space. Weak ion cohesion facilitates ion displacement to enable an expansion of empty space to accommodate CO2. Moreover, it is demonstrated that the strength of ion cohesion is primarily determined by the ion density, which in turn is given by the ion sizes. Ion cohesion is influenced additionally to a smaller extent by local electrostatic interactions among ion moieties between which CO2 is inserted and which do not depend on the ion density. Overall, the factors that determine the solubility of CO2 in ILs are identified consistently across a large variety of constituting ions through MD simulations.

Impact of ionic liquids in aqueous solution on bacterial plasma membranes studied with molecular dynamics simulations.

The impact of five different imidazolium-based ionic liquids (ILs) diluted in water on the properties of a bacterial plasma membrane is investigated using molecular dynamics (MD) simulations. Cations considered are 1-octyl-3-methylimidazolium (OMIM), 1-octyloxymethyl-3-methylimidazolium (OXMIM), and 1-tetradecyl-3-methylimidazolium (TDMIM), as well as the anions chloride and lactate. The atomistic model of the membrane bilayer is designed to reproduce the lipid composition of the plasma membrane of Gram-negative Escherichia coli. Spontaneous insertion of cations into the membrane is observed in all ILs. Substantially more insertions of OMIM than of OXMIM occur and the presence of chloride reduces cation insertions compared to lactate. In contrast, anions do not adsorb onto the membrane surface nor diffuse into the bilayer. Once inserted, cations are oriented in parallel to membrane lipids with cation alkyl tails embedded into the hydrophobic membrane core, while the imidazolium-ring remains mostly exposed to the solvent. Such inserted cations are strongly associated with one to two phospholipids in the membrane. The overall order of lipids decreased after OMIM and OXMIM insertions, while on the contrary the order of lipids in the vicinity of TDMIM increased. The short alkyl tails of OMIM and OXMIM generate voids in the bilayer that are filled by curling lipids. This cation induced lipid disorder also reduces the average membrane thickness. This effect is not observed after TDMIM insertions due to the similar length of cation alkyl chain and the fatty acids of the lipids. This lipid-mimicking behavior of inserted TDMIM indicates a high membrane affinity of this cation that could lead to an enhanced accumulation of cations in the membrane over time. Overall, the simulations reveal how cations are inserted into the bacterial membrane and how such insertions change its properties. Moreover, the different roles of cations and anions are highlighted and the fundamental importance of cation alkyl chain length and its functionalization is demonstrated.

Extraction of tryptophan with ionic liquids studied with molecular dynamics simulations.

Extraction of amino acids from aqueous solutions with ionic liquids (ILs) in biphasic systems is analyzed with molecular dynamics (MD) simulations. Extraction of tryptophan (TRP) with the imidazolium-based ILs [C(4)mim][PF(6)], [C(8)mim][PF(6)], and [C(8)mim][BF(4)] are considered as model cases. Solvation free energies of TRP are calculated with MD simulations and thermodynamic integration in combination with an empirical force field, whose parametrization is based on the liquid-phase charge distribution of the ILs. Calculated solvation free energies reproduce successfully all observed experimental trends according to the previously reported partition of TRP between water and IL phases. Water is present in ILs as a cosolvent, due to direct contact with the aqueous phase during extraction, and is found to play a major role in the extraction of TRP. Water improves solvation of cationic TRP by 7.8 and 5.1 kcal/mol in [C(4)mim][PF(6)] and [C(8)mim][PF(6)], respectively, which is in the case of [C(4)mim][PF(6)] sufficient to extract TRP. Extraction in [C(8)mim][PF(6)] is not feasible, since the hydrophobic octyl groups of the cations limit the water concentration in the IL. The solvation of cationic TRP is 2.4 kcal/mol less favorable in [C(8)mim][PF(6)] than in [C(4)mim][PF(6)]. Water improves the solvation of TRP in ILs mostly through dipole-dipole interactions with the polar backbone of TRP. Extraction is most efficient with [C(8)mim][BF(4)], where hydrophilic BF(4)(-) anions substantially increase the water concentration in the IL. Additionally, stronger direct electrostatic interactions of TRP with BF(4)(-) anions improve its solvation in the IL further. The solvation of cationic TRP in [C(8)mim][BF(4)] is 3.4 kcal/mol more favorable than in [C(8)mim][PF(6)]. Overall, the extractive power of the ILs correlates with the water saturation concentration of the IL phase, which in turn is determined by the hydrophilicity of the constituting ions. The results of this work identify relations between the extraction performance of ILs and the basic chemical properties of the ions, which provide guidelines that could contribute to the design of improved novel ILs for amino acid extraction.

A model for self-diffusion of guanidinium-based ionic liquids: a molecular simulation study.

We propose a novel self-diffusion model for ionic liquids on an atomic level of detail. The model is derived from molecular dynamics simulations of guanidinium-based ionic liquids (GILs) as a model case. The simulations are based on an empirical molecular mechanical force field, which has been developed in our preceding work, and it relies on the charge distribution in the actual liquid. The simulated GILs consist of acyclic and cyclic cations that were paired with nitrate and perchlorate anions. Self-diffusion coefficients are calculated at different temperatures from which diffusive activation energies between 32-40 kJ/mol are derived. Vaporization enthalpies between 174-212 kJ/mol are calculated, and their strong connection with diffusive activation energies is demonstrated. An observed formation of cavities in GILs of up to 6.5% of the total volume does not facilitate self-diffusion. Instead, the diffusion of ions is found to be determined primarily by interactions with their immediate environment via electrostatic attraction between cation hydrogen and anion oxygen atoms. The calculated average time between single diffusive transitions varies between 58-107 ps and determines the speed of diffusion, in contrast to diffusive displacement distances, which were found to be similar in all simulated GILs. All simulations indicate that ions diffuse by using a brachiation type of movement: a diffusive transition is initiated by cleaving close contacts to a coordinated counterion, after which the ion diffuses only about 2 A until new close contacts are formed with another counterion in its vicinity. The proposed diffusion model links all calculated energetic and dynamic properties of GILs consistently and explains their molecular origin. The validity of the model is confirmed by providing an explanation for the variation of measured ratios of self-diffusion coefficients of cations and paired anions over a wide range of values, encompassing various ionic liquid classes as well as the simulated GILs. The proposed diffusion model facilitates the qualitative a priori prediction of the impact of ion modifications on the diffusive characteristics of new ionic liquids.

Proton transfer between tryptophan and ionic liquid solvents studied with molecular dynamics simulations.

The reaction free energies and associated pK(a) values for proton transfer from positively charged tryptophan (HTrp(+)) to the two pure ionic liquids (ILs) BMIM-PF6 and BMIM-BF4 are derived from molecular simulations. IL solvation effects are examined with molecular dynamics simulations together with an empirical force field in which the average charge distribution in the actual IL is taken into account. A combination of molecular mechanical and quantum mechanical description (QM/MM) is used to examine the protonation of the anion constituents of the ILs. A dissociation of the protonated anions is observed into hydrogen fluoride and BF3 or PF5. Finally, pK(a) values of 16.5 and 21.5 in BMIM-BF4 and BMIM-PF6, respectively, are found for proton transfer from HTrp(+) to PF6(-) and BF4(-) anions, which indicates that a deprotonation of HTrp(+) is highly unfavorable compared to aqueous solutions. An examination of the contributions to the reaction free energies demonstrates that a deprotonation of tryptophan is impeded because two ions need to be annihilated for the reaction to occur: HTrp(+) and an anion. While the solvation effects induced by the two ILs are similar, the low proton acceptance of PF6(-) anions leads to the larger pK(a) value in BMIM-PF6. Also, estimates suggest that IL-induced pK(a) shifts are comparably small in proton transfer reactions where the total number of ions remains unchanged. For the first time, pK(a) values of acids were determined computationally in ILs. The obtained results elucidate the role of solvation effects on proton transfer between amino acids and ILs and improve our understanding of the observed pH memory of proteins that are solvated in ILs.

Protein-like dynamics of polycarbonate polymers in water.

The dynamics of amphiphilic peptide-mimicking polycarbonate polymers are investigated, considering variations in polymer length, monomer sequence, and monomer modification. The polymers are simulated in aqueous solution with atomistic molecular dynamics simulations and an empirical force field. Various structural polymer properties, interaction strengths, and solvation free energies are derived. It is found that water is a less favorable solvent for these polymers than for peptides. Moreover, polymers readily adopt irreversibly a compact state that consists of a variety of distinct compact conformations that are adopted through frequent transitions. Furthermore, the polymers exhibit a strong propensity to form large aggregates. The driving forces for these processes appear to be a hydrophobic effect and more favorable polymer-solvent interactions of aggregates that overcome the otherwise strong mutual repulsion between the positively charged polymers. Replacing hydrophobic residues with polar side chains destabilizes the compact conformations of the polymers. Our results also indicate that the monomer sequence has little effect on the overall solvation properties of the polymer molecule. However, the sequence influences flexibility and compactness of the monomer in solution. Overall, the results of this work confirm the protein-like characteristics of these polymers and elucidate the role of single residues in influencing the structure and aggregation in aqueous solution.