Effect of Relative Mass on Ion Velocity Cross-Correlations in Ionic Liquids and Molten Salts: Different Perspectives in Different Reference Frames.

In electrolytes, the self- and interdiffusion coefficients, transport numbers, and electrical conductivity can be used to determine velocity cross-correlation coefficients (VCC) that are also accessible through molecular dynamics simulations. In an ionic liquid or molten salt, there are only three, corresponding to correlations between the velocities of distinct ion pairs (cation-anion, cation-cation, and anion-anion) averaged over both the ensemble and time, calculable from experimental ion self-diffusion coefficients and the electrolyte conductivity. Most usually, the mass-fixed frame of reference (with velocities relative to that of the center of mass of the system) is used to discuss the VCC and the distinct diffusion coefficients (DDC) derived from them. Recent work in the literature has suggested a dependence of the DDC on the ratio of the anion to cation mass. Here, we demonstrate, using our own and literature transport property data for a large number of ionic liquids and molten salts, that the trends observed depend on the particular choice of velocity reference frame, mass-, number-, or volume-fixed. The perception of ion-ion interactions may be distorted in the mass- and volume fixed frames when the co-ions have very different masses or volumes, particularly for systems containing light lithium ions.

A volumetric and intra-diffusion study of solutions of AlCl3 in two ionic liquids - [C2TMEDA][Tf2N] and [C4mpyr][Tf2N].

Intra-diffusion coefficients (DSi) have been measured for the ionic liquid constituent ions and aluminium-containing species in aluminium chloride (AlCl3) solutions in the ionic liquids 1-(2-dimethyl-aminoethyl)-dimethylethylammonium bis(trifluoromethylsulfonyl)amide ([C2TMEDA][Tf2N]) and N-butyl-N-methylpyrrolidinium bis(trifluoromethylsulfonyl)amide ([C4mpyr][Tf2N]), to investigate whether spectroscopically detected interactions between the ions and AlCl3 affect these properties. Such electrolyte solutions are of interest for the electrowinning of aluminium. The temperature, composition and molar volume dependences are investigated. Apparent (Vϕ,1) and partial molar (V1) volumes for AlCl3 have been calculated from solution densities. For [C2TMEDA][Tf2N] solutions, Vϕ,1 increases with increasing solute concentration; for [C4mpyr][Tf2N] solutions, it decreases. In pure [C2TMEDA][Tf2N], the cation diffuses more quickly than the anion, but this changes as the AlCl3 concentration increases. In the [C4mpyr][Tf2N] solutions, the intra-diffusion coefficient ratio remains equal to that for the pure ionic liquid and the aluminium species diffuses at approximately the same rate as the anion at each composition. The intra-diffusion coefficients can be fitted to the Ertl-Dullien free volume power law by superposing the iso-concentration curves with concentration dependent, but temperature independent, molar volume offsets. This suggests that they are primarily dependent on the molar volume and secondarily on a colligative thermodynamic factor due to dilution by AlCl3. AlCl3 complexation by [Tf2N]- and [C2TMEDA]+, confirmed by 27Al, 15N and 19F NMR spectroscopy, seems to play a minor role. Our results indicate that the application of free volume theories might be fruitful in the study of the transport properties of ionic liquid solutions and mixtures.

Does [Tf2N]- slither? Equivalence of cation and anion self-diffusion activation volumes in 1-ethyl-3-methylimidazolium bis(trifluoromethylsulfonyl)amide.

New high-pressure self-diffusion data are reported for the ionic liquid 1-ethyl-3-methylimidazolium bis(trifluoromethylsulfonyl)amide ([EMIM][Tf2N]) at pressures up to 363 MPa in the temperature range 288-348 K. The cation and anion activation volumes derived from these are found to be equal at a fixed temperature, within experimental error, in contradiction to a report in the literature that they differ significantly. Self-diffusion activation volumes derived from our earlier high-pressure diffusion studies also show equality for the respective cations and anions of bis(trifluoromethylsulfonyl)amide, tetrafluoroborate and hexafluorophosphate salts with various cations. Stokes-Einstein-Sutherland analysis and density scaling are applied to the [EMIM][Tf2N] self-diffusion measurements and support the conclusion that pressure effects both cation and anion mass (and hence charge) transport in the same way. The density scaling parameters are consistent with the theoretical predictions of Knudsen et al. and agree with that for the viscosity, as for other ionic liquids.

Thermodynamic or density scaling of the electrical conductivity of molten salts.

Thermodynamic or density scaling of high-pressure conductivities and molar conductivities of the high-temperature molten salts NaOH, and the alkali chlorides, bromides, and nitrates, from Na to Cs, taken from the literature, is found to be consistent with the simulations of Knudsen, Niss, and Bailey (KNB). They used a simple model fluid of point particles interacting through an interionic potential with a repulsive inverse power law part varying as r-9 and an attractive Coulombic part. This yields values between the limits 0.33-3 for the scaling parameter, γ. The Coulombic potential reduces the scaling parameter to values much lower than are normally found for molecular liquids, and KNB used this to explain the low values typically found for ionic liquids. Here, it is shown that the high-temperature molten salts examined behave similarly.

Correction: Self-diffusion, velocity cross-correlation, distinct diffusion and resistance coefficients of the ionic liquid [BMIM][Tf2N] at high pressure.

Correction for 'Self-diffusion, velocity cross-correlation, distinct diffusion and resistance coefficients of the ionic liquid [BMIM][Tf2N] at high pressure' by Kenneth R. Harris et al., Phys. Chem. Chem. Phys., 2015, 17, 23977-23993, DOI: 10.1039/C5CP04277A.

Thermodynamic or density scaling of the thermal conductivity of liquids.

Thermodynamic or density scaling is applied to thermal conductivity (λ) data from the literature for the model Lennard-Jones (12-6) fluid; the noble gases neon to xenon; nitrogen, ethene, and carbon dioxide as examples of linear molecules; the quasi-spherical molecules methane and carbon tetrachloride; the flexible chain molecules n-hexane and n-octane; the planar toluene and m-xylene; the cyclic methylcyclohexane; the polar R132a and chlorobenzene; and ammonia and methanol as H-bonded fluids. Only data expressed as Rosenfeld reduced properties could be scaled successfully. Two different methods were used to obtain the scaling parameter γ, one based on polynomial fits to the group (TVγ) and the other based on the Avramov equation. The two methods agree well, except for λ of CCl4. γ for the thermal conductivity is similar to those for the viscosity and self-diffusion coefficient for the smaller molecules. It is significantly larger for the Lennard-Jones fluid, possibly due to a different dependence on packing fraction, and much larger for polyatomic molecules where heat transfer through internal modes may have an additional effect. Methanol and ammonia, where energy can be transmitted through intermolecular hydrogen bonding, could not be scaled. This work is intended as a practical attempt to examine thermodynamic scaling of the thermal conductivity of real fluids. The divergence of the scaling parameters for different properties is unexpected, suggesting that refinement of theory is required to rationalize this result. For the Lennard-Jones fluid, the Ohtori-Iishi version of the Stokes-Einstein-Sutherland relation applies at high densities in the liquid and supercritical region.

On the Use of the Angell-Walden Equation To Determine the "Ionicity" of Molten Salts and Ionic Liquids.

In this work, the Angell analysis of Walden plots of the conductivity of ionic liquids and other electrolytes against viscosity is used to examine simple molten salts at high temperatures, a test that does not appear to have been made previously. It is found that many simple salts such as alkali metal fluorides and chlorides are predicted to be "superionic" as their Walden plots fall above the arbitrary reference line introduced by Angell, which passes through the datum point for 1 M aqueous KCl at 25 °C. This contradicts long-standing molecular dynamics evidence in the literature showing that these salts conduct simply by ion migration in an electric field. Zinc chloride is also predicted to be "ideal", whereas one would expect it to be "subionic" in Angell's terminology given that it is an associated salt. Results for certain protic ionic liquids are also contradictory. Therefore, Angell-Walden analyses of this type do not convey any useful information other than a qualitative ranking of the conductivity of similar ionic liquids at a given viscosity and their use for estimating "ionicity" is best discontinued. It cannot and should not be used for classifying the interactions in ionic liquids. Instead, it is argued that an examination of Laity resistance coefficients is more useful in any discussion of true association in molten salts and ionic liquids where known examples show negative like-ion resistance coefficients with NE deviation parameters close to unity. Such an approach could be more fruitful in understanding the transport properties of molten salts and ionic liquids rather than simple comparisons of viscosity and conductivity.

Possible Proton Conduction Mechanism in Pseudo-Protic Ionic Liquids: A Concept of Specific Proton Conduction.

In a previous work, we have found that the pseudo-protic ionic liquid N-methylimidazolium acetate, [C1HIm][OAc] or [Hmim][OAc], mainly consists of the electrically neutral molecular species N-methylimidazole, C1Im, and acetic acid, AcOH, even though the mixture has significant ionic conductivity. This system was revisited by employing isotopic substitution Raman spectroscopy (ISRS) and pulsed field gradient (PFG) NMR self-diffusion measurements. The ISRS and PFG-NMR results obtained fully confirm our earlier findings. In particular, the self-diffusion coefficient of the hydroxyl hydrogen atom in AcOH is identical to that of the methyl hydrogen atoms within the experimental uncertainty, consistent with very little ionization. Therefore, a proton conduction mechanism similar to the Grotthuss mechanism for aqueous acid solutions is postulated to be responsible for the observed electrical conductivity. Laity resistance coefficients (rij) are calculated from the transport properties, and the negative values obtained for the like-ion interactions are consistent with the pseudo-ionic liquid description, that is, the mixture is indeed a very weak electrolyte. The structure and rotational dynamics of the mixture were also investigated using high-energy X-ray total scattering experiments, molecular dynamics simulations, and dielectric relaxation spectroscopy. Based on a comparison of activation energies and the well-known linear free energy relationship between the kinetics and thermodynamics of autoprotolysis, we propose for [C1HIm][OAc] a Grotthus-type proton conduction mechanism involving fast AcOH/AcO- rotation as a decisive step.

Correction: Comment on "Negative effective Li transference numbers in Li salt/ionic liquid mixtures: does Li drift in the "Wrong" direction?" by M. Gouverneur, F. Schmidt and M. Schönhoff, Phys. Chem. Chem. Phys., 2018, 20, 7470.

Correction for 'Comment on "Negative effective Li transference numbers in Li salt/ionic liquid mixtures: does Li drift in the "Wrong" direction?" by M. Gouverneur, F. Schmidt and M. Schönhoff, Phys. Chem. Chem. Phys., 2018, 20, 7470' by Kenneth R. Harris, Phys. Chem. Chem. Phys., 2018, 20, 30041-30045.

Comment on "Negative effective Li transference numbers in Li salt/ionic liquid mixtures: does Li drift in the "Wrong" direction?" by M. Gouverneur, F. Schmidt and M. Schönhoff, Phys. Chem. Chem. Phys., 2018, 20, 7470.

Gouverneur et al. have recently reported "effective transport numbers" for mixtures of lithium and 1-ethyl-3-methylimidazolium salts with common (fluorinated) anions using 7Li, 1H and 19F and electrophoretic NMR to determine the electrophoretic mobilities of all three ionic species. The "effective transport number" for lithium is small, but negative. From this they deduce that the Li+ ions are each associated with two or more anions to form negatively charged complexes. However this interpretation may be incorrect: only a single independent transport number can be measured in such a system as the three ion fluxes are not independent. One ion flow must define the reference frame and then the transport numbers for the other two ions must sum to unity. Electrophoretic NMR appears to produce what are called "external" ion mobilities and transport numbers in the notation used by Klemm and Haase for molten salts. These are defined in the laboratory frame of reference and can depend on the boundary conditions of the experiment. Simple relations exist for their conversion to "internal" transport numbers where ion mobilities for two ions are given relative to that of the third, analogous to the more familiar Hittorf transport numbers of ions in electrolyte solutions which are given in the "solvent-fixed" frame of reference, i.e. relative to the flow of solvent. It is not unusual for a cation external transport number to be negative in molten salt mixtures, e.g. (LiNO3 + AgNO3) in a Hittorf experiment employing nitrate electrodes whereas true ion association would produce negative internal transport numbers. In the examples studied by Gouverneur et al. the cation internal transport numbers are both positive. Those for Li+ are also very small, and close to zero within experimental error. This may simply reflect that the mixtures employed are dilute in lithium ions.

The importance of transport property studies for battery electrolytes: revisiting the transport properties of lithium-N-methyl-N-propylpyrrolidinium bis(fluorosulfonyl)imide mixtures.

Transport properties are examined in some detail for samples of the low temperature molten salt N-propyl-N-methyl pyrrolidinium bis(fluorosulfonyl)imide [Pyr13][FSI] from two different commercial suppliers. A similar set of data is presented for two different concentrations of binary lithium-[Pyr13][FSI] salt mixtures from one supplier. A new and significantly different production process is used for the synthesis of Li[FSI] as well as the [Pyr13]+ salt used in the mixtures. Results for the viscosity, conductivity, and self-diffusion coefficients, together with the density and expansivity and apparent molar volume, are reported over the temperature range of (0 to 80) °C. The data for neat [Pyr13][FSI] are discussed in the context of velocity cross correlation (VCC or fij) and Laity resistance (rij) coefficients. Unusually, f+- ∼ f++ < f--. The three resistance coefficients are of similar magnitude indicating all three ion-ion interactions contribute to the transport properties, not just the cation-anion interaction. The composition dependence of the transport properties is compared to previously reported data for the same and related compounds: in contrast to high-temperature molten salt mixtures, this is an exponential dependence. The Nernst-Einstein parameter Δ, which contains information on the correlations of the ionic velocities and is determined by differences in the VCC for the various ion-ion combinations, was calculated for both the neat ionic liquid and its binary mixture. It increases with increasing lithium concentration. The new data set also allows some conclusions with regards to the lithium-[FSI]- coordination environment.

Revised and Extended Values for Self-Diffusion Coefficients of 1-Alkyl-3-methylimidazolium Tetrafluoroborates and Hexafluorophosphates: Relations between the Transport Properties.

Earlier measurements of the self-diffusion coefficients of 1-alkyl-3-methylimidazolium (or [RMIM], R = alkyl) tetrafluoroborates and hexafluorophosphates have been revised and extended to 90 °C. The main changes are to DS+ and DS- for [HMIM][PF6] ([C6C1Im][PF6]) and DS- for [OMIM][BF4] ([C8C1Im][BF4]). New atmospheric pressure self-diffusion, density, and conductivity data are provided for [HMIM][BF4] ([C6C1Im][BF4]). Velocity cross-correlation, distinct diffusion, and Laity resistance coefficients have been calculated. There is no evidence for ion association. A new relation between the Nernst-Einstein deviation parameter (Δ) and resistance coefficients (rij) is derived; Δ tends toward 0.5 when the like-ion rii are much smaller than the unlike-ion rij, i.e., when the counterion interactions dominate. [OMIM]+ ion salts approach this limit. Stokes-Einstein-Sutherland and Walden plots overlap almost quantitatively for [BF4]-, [PF6]-, and Cl- salts with a common [RMIM]+ cation. That is, in thermodynamic states that have the same viscosity, the salt molar conductivities and hence ionic electrical mobilities of, say, [BMIM][BF4] and [BMIM][PF6] are almost equal, as are the corresponding Brownian or diffusive mobilities, (DSi/RT), for the cation, and also for these three small anions.

Can the Transport Properties of Molten Salts and Ionic Liquids Be Used To Determine Ion Association?

There have long been arguments supporting the concept of ion association in molten salts and ionic liquids, largely based on differences between the conductivity and that predicted from self-diffusion coefficients by the Nernst-Einstein equation for noninteracting ions. It is known from molecular dynamics simulations that even simple models based on charged hard spheres show such a difference due to the (anti)-correlation of ion motions. Formally this is expressed as a difference between the velocity cross-correlation coefficient of the oppositely charged ions and the mean of those for the two like-charged ions. This article examines molten salt and ionic liquid transport property data, comparing simple and model associated salts (ZnCl2, PbCl2, and TlCl) including weakly dissociated molecular liquids (H2O, HCOOH, H2SO4). Analysis employing Laity resistance coefficients (rij) shows that the common ion-association rationalization is flawed, consistent with recent direct measurements of the degree of ionicity in ionic liquid chlorides and with theoretical studies. However, the protic ionic liquids [PyrOMe][BF4] and [DBUH][CH3SO3] have larger than usual NE deviation parameters (>0.5), and large negative like-ion rii, analogous to those of ZnCl2. Structural, spectroscopic, and theoretical studies are suggested to determine whether these are indeed genuine examples of association.

Self-diffusion, velocity cross-correlation, distinct diffusion and resistance coefficients of the ionic liquid [BMIM][Tf2N] at high pressure.

Ion self-diffusion coefficients (DSi) have been measured for the ionic liquid 1-butyl-3-methylimidazolium bis(trifluoromethanesulfonyl)amide [BMIM][Tf2N] at pressures to 200 MPa between 25 and 75 °C and at 0.1 MPa between 10 and 90 °C. Self-diffusion coefficients are reported for 1-ethyl-, 1-hexyl- and 1-octyl-3-methylimidazolium bis(trifluoromethanesulfonyl)amide salts at 0.1 MPa, supplemented by viscosity, electrical conductivity and density measurements. Velocity cross-correlation (VCC, fij) and distinct diffusion coefficients (D) are calculated from the data. Both DSi and D are analysed in terms of (fractional) Stokes-Einstein-Sutherland (SES) equations. SES and Walden plots show almost identical slopes, with high-pressure isotherms and the atmospheric pressure isobar falling on common, single lines for each property for [BMIM][Tf2N]. SES plots for the anion self-diffusion coefficients for the [RMIM][Tf2N] (R = alkyl) series are coincident, whereas those for the cations depend on their alkyl substitution, as do the Walden plots. In common with other [Tf2N](-) salts, the VCC follow the order f-- < f++ < f+-. The Nernst-Einstein deviation parameter Δ for [BMIM][Tf2N] is independent of temperature and pressure. Those for the other [Tf2N](-) salts are independent of temperature. Δ increases in magnitude with increasing alkyl chain length on the cation. The transport properties of [BMIM][Tf2N] are re-examined in terms of density scaling using reduced conductivities and reduced molar conductivities for the first time. Identical scaling parameters (γ) are obtained for the several reduced transport properties. This result is supported by data for other ionic liquids. It is suggested that the γ for ionic liquids may depend on packing fraction.

Viscosity scaling of the self-diffusion and velocity cross-correlation coefficients of two functionalised ionic liquids and of their non-functionalized analogues.

Ion self-diffusion coefficients have been measured for ionic liquids based on the cations N-acetoxyethyl-N,N-dimethyl-N-ethylammonium ([N(112,2OCO1)](+)) and its non-functionalised analogue, N,N-dimethyl-N-ethyl-N-pentylammonium ([N1125](+)), and N,N-dimethyl-N-ethyl-N-methoxyethoxyethylammonium ([N(112,2O2O1)](+)), and its analogue, N,N-dimethyl-N-ethyl-N-heptylammonium ([N1127](+)) and the bis(trifluoromethanesulfonyl)amide anion. The functionalised chain on an ammonium cation has the same length, in terms of the number of atoms, as the non-functionalised chain of the corresponding analogue. For [N(112,2OCO1)][Tf2N] and [N1127][Tf2N], the cation and anion self-diffusion coefficients are equal, within experimental error, whereas for [N1125][Tf2N], the cation diffuses more quickly, and for [N(112,2O2O1)][Tf2N], it is the anion that diffuses more quickly than the ether-functionalised cation. But these differences are relatively small, just beyond experimental error. The data are used to calculate velocity cross-correlation coefficients (VCC or f(ij)) and distinct diffusion coefficients (D(ij)(d)). Both the self-diffusion and distinct diffusion coefficients are analysed in terms of (fractional) Stokes-Einstein-Sutherland equations. Though the self-diffusion coefficients, as with the conductivity and viscosity, show marked differences in absolute terms between the functionalised and non-functionalised forms, being higher for the ethoxy-substituted IL and lower for the acetoxy-substituted IL, these are largely removed by scaling with the viscosity. Thus the transport properties are better understood as functions of the viscosity rather than the temperature and density, per se. The presence of the alkoxy-substituted side chains is known to change the local mesoscopic liquid structure, but it appears once this is done, the transport properties scale correspondingly. In the case of the acetoxy-substituted IL, this is also largely the case, but the Nernst-Einstein deviation parameter, Δ, which depends on the difference between the anion-cation VCC and the mean of the cation-cation and anion-anion VCCs, is smaller than that of its analogue salt, and also temperature dependent.

Transport, electrochemical and thermophysical properties of two N-donor-functionalised ionic liquids.

Two N-donor-functionalised ionic liquids (ILs), 1-ethyl-1,4-dimethylpiperazinium bis(trifluoromethylsulfonyl)amide (1) and 1-(2-dimethylaminoethyl)-dimethylethylammonium bis(trifluoromethylsulfonyl)amide (2), were synthesised and their electrochemical and transport properties measured. The data were compared with the benchmark system, N-butyl-N-methylpyrrolidinium bis(trifluoromethylsulfonyl)amide (3). Marked differences in thermal and electrochemical stability were observed between the two tertiary-amine-functionalised salts and the non-functionalised benchmark. The former are up to 170 K and 2 V less stable than the structural counterpart lacking a tertiary amine function. The ion self-diffusion coefficients (Di ) and molar conductivities (Λ) are higher for the IL with an open-chain cation (2) than that with a cyclic cation (1), but less than that with a non-functionalised, heterocyclic cation (3). The viscosities (η) show the opposite behaviour. The Walden [Λ[proportionality](1/η)(t) ] and Stokes-Einstein [Di /T)[proportionality](1/η)(t) ] exponents, t, are very similar for the three salts, 0.93-0.98 (±0.05); that is, the self-diffusion coefficients and conductivity are set by η. The Di for 1 and 2 are the same, within experimental error, at the same viscosity, whereas Λ for 1 is approximately 13% higher than that of 2. The diffusion and molar conductivity data are consistent, with a slope of 0.98±0.05 for a plot of ln(ΛT) against ln(D+ +D- ). The Nernst-Einstein deviation parameters (Δ) are such that the mean of the two like-ion VCCs is greater than that of the unlike ions. The values of Δ are 0.31, 0.36 and 0.42 for 3, 1 and 2, respectively, as is typical for ILs, but there is some subtlety in the ion interactions given 2 has the largest value. The distinct diffusion coefficients (DDC) follow the order D(d)__ < D(d)++ < D(d)+_, as is common for [Tf2N](-) salts. The ion motions are not correlated as in an electrolyte solution: instead, there is greater anti-correlation between the velocities of a given anion and the overall ensemble of anions in comparison to those for the cationic analogue, the anti-correlation for the velocities of which is in turn greater than that for a given ion and the ensemble of oppositely charged ions, an observation that is due to the requirement for the conservation of momentum in the system. The DDC also show fractional SE behaviour with t~0.95.

On the density scaling of pVT data and transport properties for molecular and ionic liquids.

In this work, a general equation of state (EOS) recently derived by Grzybowski et al. [Phys. Rev. E 83, 041505 (2011)] is applied to 51 molecular and ionic liquids in order to perform density scaling of pVT data employing the scaling exponent γ(EOS). It is found that the scaling is excellent in most cases examined. γ(EOS) values range from 6.1 for ammonia to 13.3 for the ionic liquid [C(4)C(1)im][BF(4)]. These γ(EOS) values are compared with results recently reported by us [E. R. López, A. S. Pensado, M. J. P. Comuñas, A. A. H. Pádua, J. Fernández, and K. R. Harris, J. Chem. Phys. 134, 144507 (2011)] for the scaling exponent γ obtained for several different transport properties, namely, the viscosity, self-diffusion coefficient, and electrical conductivity. For the majority of the compounds examined, γ(EOS) > γ, but for hexane, heptane, octane, cyclopentane, cyclohexane, CCl(4), dimethyl carbonate, m-xylene, and decalin, γ(EOS) < γ. In addition, we find that the γ(EOS) values are very much higher than those of γ for alcohols, pentaerythritol esters, and ionic liquids. For viscosities and the self-diffusion coefficient-temperature ratio, we have tested the relation linking EOS and dynamic scaling parameters, proposed by Paluch et al. [J. Phys. Chem. Lett. 1, 987-992 (2010)] and Grzybowski et al. [J. Chem. Phys. 133, 161101 (2010); Phys. Rev. E 82, 013501 (2010)], that is, γ = (γ(EOS)/φ) + γ(G), where φ is the stretching parameter of the modified Avramov relation for the density scaling of a transport property, and γ(G) is the Grüneisen constant. This relationship is based on data for structural relaxation times near the glass transition temperature for seven molecular liquids, including glass formers, and a single ionic liquid. For all the compounds examined in our much larger database the ratio (γ(EOS)/φ) is actually higher than γ, with the only exceptions of propylene carbonate and 1-methylnaphthalene. Therefore, it seems the relation proposed by Paluch et al. applies only in certain cases, and is really not generally applicable to liquid transport properties such as viscosities, self-diffusion coefficients or electrical conductivities when examined over broad ranges of temperature and pressure.

High pressure studies of the transport properties of ionic liquids.

High pressure measurements have been made of viscosities, ion self-diffusion coefficients and electrical conductivities of ionic liquids, mainly of imidazolium salts. We review how these properties have been analysed in terms of the empirical Stokes-Einstein, Walden and Nernst-Einstein equations, and examine trends revealed by the phenomenological approach of velocity correlation coefficients and the more general theory of density scaling. Finally we examine the possibility of dynamic crossover in the transport properties of ionic liquids.