The Competition Between Cation‐Anion and Cation‐Triglyme Interaction in Solvate Ionic Liquids Probed by Far Infrared Spectroscopy and Molecular Dynamics Simulations

Abstract

Glyme‐based electrolyte solutions provide new concepts for developing suitable lithium‐ion batteries. The so‐called solvate ionic liquids (SILs) are promising electrolytes. They are most efficient in equimolar mixtures of lithium bis(trifluoromethanesulfonyl)imide ([Li][NTf₂]) and glyme, wherein the [Li]⁺ cation is supposedly fully solvated by glyme molecules. Here, we performed far (FIR) and mid (MIR) infrared spectroscopy for probing the solvation and local structures around the [Li]⁺ ions. In particular, we studied the competition between the triglyme molecule (G3) and the salt anions for the coordination to the lithium cations with increasing [Li][NTf₂] concentration. The formation of nano structures in the [Li][NTf₂]:G3 mixtures is discussed in terms of contact (CIP) and solvent‐separated (SIP) ion pairs in solution. At low salt concentrations, the [Li]⁺ cations are solvated by two triglyme molecules resulting in SIPs only. With increasing salt concentration, [Li]⁺ is predominantly solvated by one triglyme molecule as [Li(triglyme)₁]⁺ but still remains in contact to one of the four oxygen atoms of the [NTf₂]⁻ anion. Molecular dynamics (MD) simulations provide a molecular picture of the [Li][NTf₂]:G3 mixtures that supports the conclusions drawn from the experimental findings.

Far infrared (FIR) spectroscopy is a suitable method for probing the local structures in solvate ionic liquids. For mixtures of lithium bis(trifluoromethanesulfonyl)imide ([Li][NTf₂]) and triglyme (G3), the FIR spectra provide detailed information about the contacts between the [Li]⁺ cations and either the oxygen atoms of the G3 molecules or those of the [NTf₂]⁻ anions. Molecular dynamics (MD) simulations deliver a molecular picture of the [Li][NTf₂]:G3 complexes.

Introduction

Ionic liquids (ILs) are suitable electrolytes for fuel cells, solar cells, super capacitors and batteries. The low vapour pressure, high thermal and electrochemical stability as well as the low inflammability of ILs are tunable properties for developing further improved electrolyte systems.[ ¹ , ² , ³ , ⁴ , ⁵ , ⁶ , ⁷ ] The aim is to obtain new electrolytes that allow using cathodes with higher energy, lower cost or being more environmentally friendly. The widespread use of lithium‐ion (Li‐ion) batteries has been driving a notable technological change and is currently enabling a transition to more environmentally sustainable automotive mobility. However, the possible combinations of ions have advantages and disadvantages with regard to the development of applicable electrolyte systems. The presence of lithium ions in the electrolyte is particularly important in the development of safe electrolytes. It is a well‐known fact that when the IL is doped with a Li salt, the viscosity of the electrolyte increases dramatically due to the aggregation of the ions, resulting in an electrolyte with reduced ionic conductivity. In addition, the strong interaction between the ions leads to a low Li transference number and a strong polarization during charge and discharge.[ ⁸ , ⁹ ]

[Li][NTf₂] dissolved in small oligoethers such as triglyme C₈H₁₈O₄ (G3) are regarded as a new class of ionic liquids, the so‐called solvate ionic liquids (SILs).[ ¹⁰ , ¹¹ , ¹² , ¹³ , ¹⁴ , ¹⁵ , ¹⁶ , ¹⁷ , ¹⁸ , ¹⁹ , ²⁰ ] It has been shown that in the 1 : 1 mixtures, a single triglyme molecule wraps around a lithium ion forming crown‐ether like [Li(triglyme)₁]⁺ complex cations, finally leading to the SIL [Li(triglyme)₁][NTf₂].[ ²¹ , ²² , ²³ , ²⁴ , ²⁵ , ²⁶ ] However, it is not quite clear whether the 1 : 1 complex represents a solvent‐separated ion pair (SIP) or a contact ion pair (CIP), wherein the [Li]⁺ ion is not fully solvated by the triglyme molecule. Moreover, there is no systematic study showing the solvation of the [Li]⁺ ions with step by step adding the salt [Li][NTf₂] to the pure triglyme system. For that purpose, we first studied eight mixtures ranging from 1 : 19 up to 1 : 1 mole fractions of [Li][NTf₂] in triglyme by far infrared spectroscopy (FIR). Based on earlier studies on pure ILs, we hoped to obtain important structural information about how the individual components (cation, anion and solvent molecule) in the SILs vibrate against each other (see Scheme 1). In general, FIR spectra show very broad and overlapping vibrational bands that are highly unspecific and hardly allow for an assignment to a cation‐anion, cation‐molecule and molecule‐molecule stretching vibration. This is different for Li‐based electrolytes such as [Li][NTf₂]:G3 mixtures. The light Li cation oscillates against the comparatively heavy anion or triglyme molecule so that, according to the equation for the harmonic oscillator, the vibrational modes are determined by the force constants as well as by the reduced masses via $\tilde{\nu }=\frac{1}{2\pi c}\sqrt{k/\mu }$ , where c is the speed of light, k the force constant and μ the reduced mass.[ ²⁷ , ²⁸ , ²⁹ ] Due to the large mass of both the [NTf₂]⁻ anion and the triglyme molecule, the mass of the [Li]⁺ cation serves as good approximation for the reduced mass. The measured frequency in the FIR is therefore a direct measure of the force constant, and thus, the strength of the interaction between the [Li]⁺ cations and its oscillation partners. Consequently, we measured far infrared (FIR) spectra accompanied by mid infrared (MIR) spectra for observing the direct and indirect vibrational signatures of the constituents present in the [Li][NTf₂]:G3 mixtures. The spectral assignment is supported by quantum chemical calculations for a variety of [Li][NTf₂]:G3 complexes representing all binding possibilities within the system. The cluster distribution depending on the G3 concentration obtained from deconvoluting the FIR and MIR spectra nicely agree with [Li]⁺‐O(G3) and [Li]⁺‐O([NTf₂]⁻) radial distribution functions and calculated coordination numbers defining SIP and CIP complexes.

Scheme 1.

Potential complexes in the SIL [Li][NTf₂] : G3 mixtures with varying salt concentration [Li][NTf₂]. a) at low salt concentration: the [Li]⁺ cation is solvated by two G3 molecules (SIP), b) at medium salt concentration: the [Li]⁺ cation is fully solvated by a single G3 molecule (SIP), c) at medium salt concentration: the [Li]⁺ cation is mainly solvated by a single G3 molecule but additionally interacts with the [NTf₂]⁻ counter ion (CIP). d) at high salt concentration: the [Li]⁺ cation is fully solvated by [NTf₂]⁻ anions solely (CIP).

Experimental Methods

Far and Mid Infrared Spectroscopy

The FIR measurements were performed with a Bruker VERTEX 70 FTIR spectrometer. The instrument is equipped with an extension for measurements in the far infrared region. This equipment consists of a multilayer mylar beam splitter, a room temperature DLATGS detector with preamplifier and polyethylene (PE) windows for the internal optical path. The accessible spectral region for this configuration lies between 30 cm⁻¹ and 680 cm⁻¹. Further improvement could be achieved by using a high pressure mercury lamp and a silica beam splitter. The reason why mercury lamps have proved to be so successful for FIR is that emission from the plasma reinforces the emission from the hot quartz envelope of the lamp. This configuration allowed measurements down to 10 cm⁻¹ and significantly better signal‐to‐noise ratios compared to the above given configuration. Background spectra have been recorded for the triglyme solvent.

In all cases, the sample temperature was maintained by an external Haake DC 30/ K 20 bath chiller and recorded with a NiCrNi thermocouple attached directly to the cell.

Mid infrared measurements were performed with a Bruker Vector 22 FTIR spectrometer. An L.O.T.‐Oriel variable‐temperature cell equipped with CaF₂ windows having a path length of 0.05 mm was used for the variable‐temperature experiments. Temperatures were maintained with an external Haake C25P cryostat and were monitored with a thermocouple attached directly to the cell. For each spectrum 128 scans were recorded at a spectral resolution of 1 cm⁻¹. Solvent subtraction was carried out by using reference spectra obtained at exactly the same temperatures as the sample spectra. The sample chamber was purged with dry air during data collection.

Theoretical Methods

DFT Calculations

We calculated several [Li][NTf₂]:G3 complexes at the B3LYP−D3/6‐31+G* level considering Grimme's D3 dispersion correction with the Gaussian 09 program.[ ³⁰ , ³¹ , ³² , ³³ ] All configurations were fully optimized followed by frequency calculations. The obtained vibrational frequencies were all positive, showing that we calculated at least local minimum structures. The geometry‐optimized CIP and SIP configurations which were the lowest in energy are given in the Supporting Information.

MD Simulations

We have performed MD simulations of mixtures of [Li][NTf₂] and triglyme at 300 K. For the description of the lithium ions, we used the forcefield for monovalent alkali cations by Joung and Cheatham. ^[34] Besides, the NGOLP forcefield was applied for the [NTf₂]⁻ anions.[ ³⁵ , ³⁶ ] As a force field for triglyme, we have relied on the modified TraPPE‐UA force field of Fischer et al.[ ³⁷ , ³⁸ ] which has recently been compared favorably to various other forcefields. ^[39] The simulations were carried out under NpT conditions at a pressure of 1 bar and using mixing ratios of x([Li][NTf₂]) =0.037–0.5 (see Table 1). Hereby, the sum of all [Li][NTf₂] ion pairs (IPs) and triglyme molecules was kept constant at 1080 particles. The systems were equilibrated for 20 ns each. Then, MD simulations of 100 ns length each were performed using GROMACS 2019.6.[ ⁴⁰ , ⁴¹ , ⁴² , ⁴³ , ⁴⁴ , ⁴⁵ , ⁴⁶ , ⁴⁷ ] The integration time step for all simulations was 2 fs. The temperature of the simulated systems was controlled by employing the Nosé–Hoover thermostat[ ⁴⁸ , ⁴⁹ ] with a coupling time τ _T=1.0 ps. Additionally, the Parrinello‐Rahman barostat[ ⁵⁰ , ⁵¹ ] was applied with a coupling time τ_p =2.0 ps. The electrostatic interactions were treated by the smooth particle mesh Ewald summation. ^[52] The Ewald convergence parameter was set to a relative accuracy of the Ewald sum of 10⁻⁵. For the van der Waals interactions, a Lennard‐Jones cut‐off scheme with a cut‐off radius of 0.9 nm was used. All bond lengths were kept fixed during the simulation run, and the distance constraints were solved by means of the LINCS procedure. ^[53] The computations of the properties from MD simulations were done using our home‐built software package MDorado based on the MDAnalysis, ^[54] NumPy, ^[55] and Scipy ^[56] frameworks. MDorado is available via GitHub.

Table 1.

Simulated mixtures of [Li][NTf₂] and triglyme (G3).

x([Li][NTf2])	MD	IR	ratio [Li][NTf2]:G3	N MD([Li][NTf2])	N MD(G3)
0.037	x		1 : 26	40	1040
0.050		x	1 : 19	‐	‐
0.066	x		1 : 14	72	1008
0.100		x	1 : 9	‐	‐
0.111	x	x	1 : 8	120	960
0.167		x	1 : 5	‐	‐
0.200	x	x	1 : 4	216	864
0.250		x	1 : 3	‐	‐
0.286		x	1 : 2.5	‐	‐
0.333	x	x	1 : 2	360	720
0.400	x	x	1 : 1.5	432	648
0.444		x	1 : 1.25	‐	‐
0.450	x		1 : 1.22	486	594
0.500	x	x	1 : 1	540	540

Results and Discussion

We measured the far‐infrared (FIR) spectra of the [Li][NTf₂]:G3 mixtures with ratios 1 : 19, 1 : 9, 1 : 5, 1 : 4, 1 : 3, 1 : 1 : 2.5, 1 : 2 and 1 : 1 related to salt mole fractions of x=0.05, 0.11, 0.20, 0.25, 0.33, 0.40, 0.44 and 0.5, respectively. All spectra were recorded at 303 K. For clarity, we give all the ratios and concentrations of the mixtures [Li][NTf₂]:G3 we used in the IR experiments and/or in the MD simulations in Table 1. The pure G3 sample was used for background measurements and has been subtracted from the sample spectra in order to remove the pure G3 spectral contributions. Thus, if there are any pure G3 contributions in the spectral range of interest, they occur as increasing negative features with decreasing G3 concentration in the mixtures. The measured low frequency spectra in the spectral range between 150 cm⁻¹ and 480 cm⁻¹ are shown in Figure 1. The vibrational band slightly above 200 cm⁻¹ belongs to the wagging modes of the O=S=O groups in the [NTf₂]⁻ anion and thus, continuously increases with increasing salt concentration.[ ⁵⁷ , ⁵⁸ ] Supported by DFT calculations (see SI), we assign the other low frequency bands between 250 cm⁻¹ and 480 cm⁻¹ to vibrational motions between lithium and the G3 oxygens ([Li]⁺‐O(G3)) or between lithium and the [NTf₂]⁻ oxygen atoms ([Li]⁺‐O([NTf₂]⁻)). These vibrational modes show varying concentration dependencies. The [Li]⁺‐O(G3) contributions indicate the existence of SIPs, whereas the [Li]⁺‐O([NTf₂]⁻) show the presence of CIPs with increasing salt concentration. ^[59]

Figure 1.

Far‐infrared (FIR) spectra of the [Li][NTf₂]/triglyme mixtures recorded at 303 K. The pure triglyme sample was used for background measurements. The low frequency spectra are measured as a function of salt concentration and cover the spectral range between 150 cm⁻¹ and 480 cm⁻¹, respectively. The vibrational band slightly above 200 cm⁻¹ belongs to the wagging modes of the O=S=O groups in the [NTf₂]⁻ anion and thus, continuously rises with increasing salt concentration, whereas the other vibrational bands are related to [Li]⁺‐O(G3) and [Li]⁺‐O([NTf₂]⁻) vibrational modes show varying concentration dependencies.

The Far‐infrared (FIR) spectra of the equimolar [Li][NTf₂]/triglyme mixture (black curve) in Figure 2 cover the spectral range between 220 cm⁻¹ and 480 cm⁻¹, only showing bands related to [Li]⁺‐O(G3) and [Li]⁺‐O([NTf₂]⁻) interactions. The spectra of the 1 : 19 and 1 : 4 mixtures are multiplied by 19 and 4, respectively, for enhancing the vibrational bands (see Figure 2a). By doing so, we would like to demonstrate that three of the frequency bands were already present at low salt concentrations, and that one contribution (indicated by the turquoise color bar) clearly decreases with increasing salt concentration. The multiplied 1 : 19 und 1 : 4 spectra were then subtracted from the [Li][NTf₂]:G3 1 : 1 spectrum. As a result, we observe in Figure 2b that with increasing salt concentration up to the 1 : 1 mixtures, the vibrational band slightly above 400 cm⁻¹ (turquoise line) decreases, whereas all the other vibrational bands (indicated by the blue, green and orange lines) increase (see Figure 2c). Using the derived frequencies of the main four vibrational bands from this approach, we deconvoluted the far‐infrared (FIR) spectrum of the [Li][NTf₂]:G3 1 : 1 mixture at 303 K: As shown in Figure 3, the three vibrational bands (indicated by the blue, green and orange areas) rise with increasing salt concentration, whereas one (turquoise area) is clearly decreasing, indicating differently‐solvated [Li]⁺ cations for low and high salt concentrations. These three vibrational features are illustrated in Figure 4 and assigned as follows: The vibrational band at around 328 cm⁻¹ indicates the motion of the [Li]⁺ cation towards one of the oxygen atoms within a onefold G3‐solvated [Li]⁺ environment (see Figure 4a). The vibrational band at 376 cm⁻¹ is assigned to the (G3)O⋅⋅⋅[Li]⁺ ⋅⋅⋅O(G3) interaction of the [Li]⁺ cation between two oxygen atoms O(G3) on opposite sites within the solvate molecule as shown in Figure 4b. Finally, the vibrational band around 437 cm⁻¹ is related to the motion of a [Li]⁺ cation which is on one hand solvated by a G3 molecule, but on the other hand also interacting with one of the four oxygen atoms of the [NTf₂]⁻ anion ([Li]⁺ ⋅⋅⋅O([NTf₂]⁻)) (see Figure 4c). It forms a solvated contact ion pair (CIP), whereas in all the other complexes the [Li]⁺ cations and the [NTf₂]⁻ anions are separated by the G3 molecules resulting in solvent separated ion pairs (SIP). The band with the negative intensity describes the [Li]⁺ ⋅⋅⋅O(G3) interaction in the G3 rich region, wherein a [Li]⁺ cation is solvated by two G3 molecules (see Scheme 1a). The number of these 1 : 2 complexes should automatically decrease with decreasing G3 concentration by adding salt. This assignment is confirmed by frequency calculations at the density functional theory (DFT) level on assorted complexes (see SI).

Figure 2.

a) Far‐infrared (FIR) spectrum of the [Li][NTf₂]/triglyme 1 : 1 mixture (black curve) covering the spectral range between 220 cm⁻¹ and 480 cm⁻¹, only showing bands related to [Li]⁺‐O(G3) and [Li]⁺‐O([NTf₂]⁻) interaction. The spectrum of the 1 : 19 and 1 : 4 mixtures are multiplied by 19 and 4, respectively, to demonstrate that all vibrational features are present already at low salt concentrations, and that one contribution (indicated by the turquoise vertical line) obviously decreases with higher salt concentration. b) The spectra of the 1 : 4 mixture (×4, orange line) and the 1 : 19 mixture (×19, red line) are subtracted from the [Li][NTf₂]/triglyme 1 : 1 spectrum. As a result, we clearly observe that with increasing salt concentration up to the 1 : 1 mixtures, the vibrational band slightly above 400 cm⁻¹ decreases (turquoise line). c) In contrast to the turqoise‐marked vibrational band, all the other vibrational bands (indicated by the blue, green and orange lines) increase.

Figure 3.

Deconvoluted far‐infrared (FIR) spectrum of the [Li][NTf₂]/triglyme 1 : 1 mixture at 303 K: Three vibrational bands (indicated by the blue, green and orange lines) increase with increasing salt concentration whereas one (turquoise line) is clearly decreasing, indicating differently‐solvated [Li]⁺ cations at low than at high salt concentrations.

Figure 4.

FIR symmetric and asymmetric stretching modes (ν_σ) for the [Li]⁺‐G3 and [Li]⁺‐[NTf₂]⁻ interaction in the [Li(G3)₁][NTf₂] complexes.

We then plotted the IR intensities of all observed vibrational bands against the [Li][NTf₂] salt concentration as depicted in Figure 5 (see SI for plotted data values). Up to 33 mol % salt corresponding to a 1 : 2 ratio of [Li][NTf₂]:G3, we observe a moderate increase for the vibrational bands at 328 cm⁻¹, 376 cm⁻¹ and 437 cm⁻¹ with similar slope and a related decrease of the intensity of the band at 396 cm⁻¹, respectively. Beyond the 1 : 2 ratio for [Li][NTf₂]:G3, we observe a drastic change of the complex distributions. Obviously, the 1 : 2 complexes, wherein one [Li]⁺ cation is solvated by two G3 molecules, are continuously replaced by the other complexes, wherein the [Li]⁺ cation is interacting with only one G3 molecule and/or the [NTf₂]⁻ anion. Fewer G3 molecules are available and the formation of 1 : 2 complexes is suppressed. Consequently, the intensity of this band strongly decreases with increasing salt concentration for the benefit of all the other bands, showing similar increasing IR intensities for increasing [Li]⁺‐O(G3) and [Li]⁺‐O([NTf₂]⁻) interactions.

Figure 5.

Intensities of the vibrational bands deconvoluted from the FIR spectra measured as a function of salt concentration: Three vibrational bands increase with increasing salt concentration whereas one is clearly decreasing, indicating differently‐solvated [Li]⁺ cations at low than at high salt concentrations. Obviously, at low salt concentrations, the intensities change moderately whereas above the 1 : 2 salt/triglyme ratio, strongly enhanced vibrational bands suggest the occurrence of increasing amounts of new [Li]⁺‐O(G3) and [Li]⁺‐O([NTf₂]⁻⁾ complexes.

We observe a similar trend in the mid IR spectra between 1000 cm⁻¹ and 1400 cm⁻¹ wavenumbers as shown in Figure 6. The pure G3 spectra are used as background, resulting in a negative contribution slightly above 1100 cm⁻¹. The other vibrational bands increase with higher salt concentration indicating the formation of new complexes in the [Li][NTf₂]:G3 mixtures. Whereas the vibrational bands below 1100 cm⁻¹ and above 1300 cm⁻¹ show a continuous increase with higher salt concentration, an additional vibrational band appears at 1209 cm⁻¹ characterized by a strong increase in intensity. Our DFT calculations suggest that this band is clearly assigned to the S=O stretch frequencies of the [NTf₂]⁻ anion with strongly increased intensity due to interaction with the [Li]⁺ cation und thus, the formation of CIPs. This finding for the intramolecular S=O stretching frequencies is related to our earlier observation of increasing [Li]⁺ ⋅⋅⋅O([NTf₂]⁻) intermolecular interaction in the low frequency range. Using the same procedure for obtaining difference spectra at low and high salt concentrations, we multiplied the IR spectra for the 1 : 19 complexes by a factor of 19 and subtracted the resulting spectrum from that of the 1 : 1 complex. The difference spectra in Figure 7 show a significantly enhanced band at 1209 cm⁻¹ which we refer to the stretching frequencies of the S=O bonds interacting with the [Li]⁺ cations. The intensity of this band increases with increasing salt concentration in the same way as shown before for the intermolecular vibrational modes. The particularly strong increase of the intramolecular ([NTf₂]⁻)S=O⋅⋅⋅[Li]⁺ band beyond the 1 : 3 ratios in the [Li][NTf₂]:G3 mixtures results from decreasing availability of G3 molecules to form 1 : 2 complexes.

Figure 6.

Mid‐infrared (MIR) spectra of the [Li][NTf₂]/triglyme mixtures recorded at 303 K. The pure triglyme sample was used for background measurements indicated by the band with negative intensity around 1110 cm⁻¹. The IR spectra are measured as a function of salt concentration and cover the spectral range between 1000 cm⁻¹ and 1400 cm⁻¹.

Figure 7.

a) Mid‐infrared (FIR) spectrum of the [Li][NTf₂]:G3 1 : 1 mixture (black line) covering the spectral range between 1150 cm⁻¹ and 1380 cm⁻¹, showing bands related to [Li]⁺‐O(G3) and [Li]⁺‐O([NTf₂]⁻) interactions. The spectrum of the 1 : 19 mixture (red line) is multiplied by 19 to show the enhanced vibrational features (dashed red line), which are already present at low salt concentrations. b) This spectrum (×19, dashed red line) is subtracted from the 1 : 1 [Li][NTf₂]:G3 spectrum, resulting in a difference spectrum given by the violet dashed line. c) The difference spectrum shows that the vibrational band at 1209 cm⁻¹ representing the S=O stretch frequencies of the [NTf₂]⁻ anion is strongly enhanced due to interaction with the [Li]⁺ cation, indicating the formation of CIPs in accord with the observation of increasing [Li]⁺⋅⋅⋅O([NTf₂]⁻) intermolecular interaction observed in the low frequency range.

We performed molecular dynamics (MD) simulations for validating the interpretation of the far and mid IR spectra of the [Li][NTf₂]:G3 mixtures. We show the pair correlation function [Li]⁺ ⋅⋅⋅O(G3) and [Li]⁺ ⋅⋅⋅O([NTf₂]⁻) in Figures 8 and 9, respectively, along with the logarithmic scale for pronouncing effect. The logarithmic pair correlation function for [Li]⁺ ⋅⋅⋅O(G3) in Figure 8b highlights two significant peaks for the lithium‐oxygen distances in the lithium‐triglyme complexes. At a mixing ratio of 1 : 8 of [Li][NTf₂] and G3, the mixture demonstrates two strong peaks at 2.0 Å and 3.0 Å, whereby the first peak can be assigned to G3 oxygen atoms coordinating more strongly and the latter peak origins from weakly coordinating oxygen atoms of the complex‐forming triglyme molecules. The latter case can be mainly observed for the 2 : 1 complex cations of triglyme and lithium at high triglyme concentrations. Hence, by increasing the lithium salt concentration, the peak at 3.0 Å diminishes and from the 1.5 : 1 mixture onwards, the 2 : 1 complexes are mainly replaced by 1 : 1 complexes which are characterised by strongly coordinating oxygen atoms solely with Li−O distances of 2.0 Å. Finally, for the 1 : 1 mixture, the vast majority of Li⁺ cations are onefold‐triglyme coordinated and thus, the peak at 3.0 Å fully disappears.

Figure 8.

a) Radial‐distribution function, g(r), plotted versus r([Li]⁺‐O[G3]) indicating the distances between the [Li]⁺ cation and oxygen atoms of the G3 molecules. In the 8 : 1 mixture, most of the [Li]⁺ cations are solvated by two G3 molecules resulting in a large peak at 2.0 Å for the stronger‐solvated oxygen atoms and a smaller peak at 3.0 Å for weaker‐solvated oxygen atoms of the coordinating G3 molecules. b) The same radial‐distribution function, g(r), given at logarithmic scale for better analysis of the second peak at 3.0 Å. This peak disappears with increasing salt concentration. Finally, in the 1 : 1 mixture the Li+ cations are mainly solvated by single G3 molecules resulting in only one ([Li]⁺‐O[G3]) distance at 2.0 Å. Thus, with increasing salt concentration, the 2 : 1 complexes are replaced by 1 : 1 complexes, with a significant change from the 1.5 : 1 mixture onwards.

Figure 9.

a) Radial‐distribution function, g(r), plotted versus r([Li]⁺‐O[NTf₂]⁻) indicating the distances between the Li⁺ cation and oxygen atoms of the [NTf₂] ⁻ anions. In the 8 : 1 and 4 : 1 mixtures, the [Li]⁺ cations are fully solvated by G3 molecules and only a few cations are interacting with oxygen atoms of the [NTf₂]⁻ anions. b) The same radial‐distribution function, g(r), given at logarithmic scale for better analysis of the second peak at 4.2 Å. This peak increases significantly with increasing salt concentration from 2 : 1 mixtures onwards, similar to the first peak. Thus, beyond the 2 : 1 mixtures, the [Li]⁺ cations are solvated by one G3 molecule, and additionally, interact with the [NTf₂]⁻ anions via one strongly‐bound oxygen giving the first peak at 2.1 Å. The neighboring oxygen atom of the sulfonyl group SO₂ of the anion shows up at 4.2 Å.

From the logarithmic pair correlation functions of [Li]⁺ ⋅⋅⋅O([NTf₂]⁻) in Figure 9b can be seen that the 8 : 1 and 2 : 1 mixtures do not display strong peaks at small distances between lithium and the oxygen atoms of the [NTf₂]⁻ anion. In these mixtures, the majority of the cations are fully coordinated by triglyme and the anions are located further away from the cations leading to SIPs. Here, only very few cations interact directly with the anions. From a triglyme‐to‐salt ratio of 2 : 1 onwards, two characteristic peaks at 2.1 Å and 4.2 Å increase significantly. For higher salt concentrations, the Li⁺ cations are solvated by one G3 molecule, and additionally, interact with the [NTf₂]⁻ anions. Therefore, the first peak can be assigned to the coordinating oxygen atom of an anion while the latter distance belongs to the neighboring oxygen atom of the sulfonyl group SO₂ of the anion.

The structural findings from the pair correlation functions are summarised in Figure 10 in which the percentage of one‐fold and two‐fold triglyme coordinated complex cations as well as cations without triglyme interaction are illustrated dependent on the mixing ratio of triglyme and lithium salt (for plotted data values see SI). At low salt concentrations with a mole fraction of the salt of less than 0.3, the lithium cation is solvated by two G3 molecules as shown in Scheme 1a. If more salt is added to the mixtures, the 2 : 1 complexes are replaced by 1 : 1 complexes. In the latter complexes, [Li]⁺ cation is fully solvated by a single G3 molecule (see Scheme 1b,c). Even for the equimolar mixture of triglyme and lithium salt, only a few cations are surrounded solely by anions. If the amount of the lithium salt in the mixture is much higher than the amount of triglyme molecules, triglyme cannot compete with the anions anymore and we mostly find CIPs as illustrated in Scheme 1d.

Figure 10.

Quotient of [Li(G3) _n ]⁺ complexes with one, two or more coordinated G3 molecules, N(n), and the total number of coordinated G3 molecules, N(tot), plotted against the salt concentration. In the mixtures up to x(IP)=0.3, where IP denotes the [Li][NTf₂] ion pairs, the [Li]⁺ cation is solvated by two G3 molecules. With decreasing G3 concentration, the 2 : 1 complexes are replaced by 1 : 1 complexes, wherein the [Li]⁺ cation is fully solvated by a single G3 molecule.

Conclusions

We conclude that far infrared (FIR) spectroscopy is a suitable method for probing the solvation and local structures in so‐called solvate ionic liquids. For mixtures of lithium bis(trifluoromethanesulfonyl)imide ([Li][NTf₂]) and triglyme (G3), the low frequency vibrational modes provide detailed information about the local contacts between the [Li]⁺ cations and either the oxygen atoms of the G3 molecules or those of the [NTf₂]⁻ anions. Due to the large mass of the [NTf₂]⁻ anion and the G3 molecule, the mass of the [Li]⁺ cation is a good approximation for the reduced mass. Thus, the measured frequency provides direct information about the force constant und a good estimate for the interaction strength between [Li]⁺ and its oscillation partners. That allowed us to discuss the specific complexes present in the [Li][NTf₂]:G3 mixtures in terms of contact ion pairs (CIP) or solvent ‐separated ion pairs (SIP) in solution. At low salt concentrations, the [Li]⁺ cations are solvated by two triglyme molecules resulting in SIPs only. With increasing salt concentration, the [Li]⁺ cation is solvated by only one triglyme molecule as [Li(triglyme)₁]⁺, but shows additional contact to one of the four oxygen atoms of the [NTf₂]⁻ anion, resulting in the formation of CIPs. Molecular dynamics (MD) simulations provide a molecular picture of the [Li][NTf₂]:G3 mixtures that fully supports the conclusions drawn from the experimental findings. Above [Li]⁺ concentrations in the 1 : 3 [Li][NTf₂]:G3 mixtures, the CIP concentration strongly increases on the debit of SIP species.

Conflict of Interests

The authors declare no conflict of interest.

Supporting information

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer reviewed and may be re‐organized for online delivery, but are not copy‐edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

Supporting Information

Kristin Philipp J., Fumino K., Appelhagen A., Paschek D., Ludwig R., ChemPhysChem 2025, 26, e202400991. 10.1002/cphc.202400991

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.