Skip to main content
NCBI home page
ACS AuthorChoice logoLink to ACS AuthorChoice
. 2024 Jul 5;128(28):6876–6884. doi: 10.1021/acs.jpcb.4c02497

Evidence for NMR Relaxation Enhancement in a Protic Ionic Liquid by the Movement of Protons Independent of the Translational Diffusion of Cations

Magdalena Knapkiewicz 1, Iga Jankowska 1, Jolanta Swiergiel 1, Jadwiga Tritt-Goc 1,*
PMCID: PMC11264263  PMID: 38968545

Abstract

graphic file with name jp4c02497_0008.jpg

The molecular dynamics, thermal stability, and ionic conductivity were studied in the protic ionic liquid 1-methylimidazolium bis(trifluoromethylsulfonyl)imide ([MIm][TFSI]). The relaxation of the 1H spin–lattice of cations in the measured frequency range (10 kHz to 20 MHz) and temperature (298 to 343 K) is sensitive mainly to slow processes occurring in the molecular dynamics of protic ionic liquid and dominated by the contribution of intermolecular translational diffusion. Molecular rotations give only a constant contribution and become significant in the higher frequency range. An interesting feature is the observed enhancement of the 1H spin–lattice relaxation below 0.03 MHz attributed to the exchange of protons (order of 10–5 s) between imidazolium cations. The measurements of the self-diffusion coefficient of hydrogen atoms of cation from 298 to 343 K additionally confirm the observed phenomenon. The coefficient for exchangeable protons −NH is higher than for the cation. The nuclear magnetic resonance (NMR) experiments provide unambiguous evidence for proton transport decoupled from molecular diffusion of ions and support the conclusion that the charge transport mechanism in the studied PIL includes contributions from both the vehicular and Grotthus mechanisms. The protic ionic liquid is thermally stable to about 573 K as shown by thermogravimetric analysis and its electrical conductivity is 5 × 10–2 S/cm at 423 K.

1. Introduction

The research conducted in our laboratory is aimed at designing and obtaining new ionically conductive materials for potential applications in ecological energy sources in the intermediate temperature range of 373 to 473 K. We paid great attention to obtaining composites of cellulose and nanocellulose with heterocyclic molecules containing nitrogen. The best electrical conductivity result of about 10–1 S/m at a temperature of 413 K under anhydrous conditions, was obtained for nanocrystalline cellulose composite with imidazole.1 Unfortunately, the thermal stability of this material is not sufficient for use, e.g., as a membrane in fuel cells. For this reason, we decided to replace imidazole acting as a charge carrier in cellulose composites with another carrier, i.e., a protic ionic liquid (PIL).2,3 PILs are a subclass of ionic liquids characterized by low vapor pressure, high thermal stability, and high ionic conductivity, and additionally have an exchangeable proton, usually supported on a cation.2−4 The reports of biopolymers with PILs are unique but very promising and were first presented by Danyliv in 20215 and recently by us.6 We decided to use protic ionic liquid to overcome the problem of the high viscosity of most “conventional” ionic liquids which makes charge transport usually too slow for electrochemical applications. In PILs, charge transport comes not only from the movement of large ions (translational diffusion of ions) but also from the transport of small and light protons (proton hopping, independent of ionic species’ motions).7 In this way, charge transport can be separated from mass transport, unlike in an aprotic ionic liquid, and the conductivity can be increased. We used imidazole-based protic ionic liquids to prepare cellulose/PIL composites because of their higher thermal stability compared to ammonia-based ionic liquids8 and immiscibility with water.9 The last property is important when we think about the use of ionic liquids in membranes because it prevents the ionic liquid from being washed out of the membrane by water, which is a product of the oxygen reduction reaction at the cathode in the fuel cell, even in the case of an electrolyte that is conductive under anhydrous conditions.

The idea of obtaining a conductive composite with a protic ionic liquid was the inspiration to research the selected ionic liquid 1-methylimidazolium bis(trifluoromethylsulfonyl) imide ([MIm][TFSI]) itself. The cation of this ionic liquid consists of a 5-membrane ring with two nitrogen atoms and three carbon atoms, i.e., a derivative of imidazole, with methyl group substituted on one nitrogen atom and, unlike other members of this type of cations, only one hydrogen on the other nitrogen. Thanks to this active (exchangeable) proton the [MIm][TFSI] is the protic ionic liquid. The exchangeable proton on the cation can form hydrogen bonds to one electronegative atom of the anion (e.g., N, O, or F) or cation. The formation of such a network of hydrogen bonds enables the proton hopping and their transport for a long-range distance unrelated to ion diffusion, i.e., the Grotthuss mechanism.7 In general, there are two possible contributions to charge transport in PILs: the movement of ions (anion and cation) and the transportation of small and light protons.7 However, whether protons can move faster and independently of cations is a matter of current debate. It is a challenge to provide unambiguous experimental evidence for this process.

The studies of a wide series of 1-alkylimidazolium bis(trifluoromethylsulfonyl)imide ionic liquids with the different alkyl chain lengths (from n = 2 to 12) on the imidazolium cation have shown that −NH proton on the cation form stronger H-bonds with the anion for longer alkyl chain and that the ionic conductivity also depends on the alkyl chain length.10 The cation-independent transport of protons in PILs has so far been studied using Pulse Field Gradient NMR (PFG NMR) methods,11 which allow for the determination of the self-diffusion coefficients of protons of the alkyl chain, imidazolium ring and NH group of PIL.10,12,13 The values were very similar for these three types of hydrogen atoms and based on this similarity, a separate movement of the exchangeable protons in the −NH position from the diffusion of the parent cations was excluded in pure PILs. On the other hand, evidence of a proton motion decoupled from molecular diffusion of ions was reported by Martinelli et al.14 in the mixture of protic ionic liquid and imidazole but in the low concentration range of imidazole.

Here, we present an investigation of molecular dynamics in 1-methylimidazolium bis(trifluoromethylsulfonyl) imide ([MIm][TFSI]) at the relevant time- and length scales. That was achieved by measuring the diffusion of cations with PFG NMR methods and the proton spin–lattice relaxation time as a function of the magnetic field by the Fast-Field Cycling NMR method (FFC NMR).11,15,16 These nuclear magnetic resonance (NMR) methods are the most useful tools in the study of ion dynamics because they are nondestructive, nucleus specific, and allow access to the motions in a broad frequency range through the determination of self-diffusion coefficient, D, and spin–lattice relaxation times, T1. The self-diffusion coefficient provides access to motions in the time scale 1–100 ms and μm length range thus probing the long-range dynamics. The T1 gives information about molecular motions near the inverse of the resonant frequency. With commercially available spectrometers (e.g., Spin Master, Stelar Mede, Italy) dynamics of liquids can be studied in the broad frequency range between 1 kHz and 40 MHz, referring to the 1H resonance frequency but it is possible to go down to 3 Hz through FFC NMR as shown by Fujara et al.16 Thus, T1 relaxation measurement for given temperatures allows probing the short-range dynamics in one single experiment. The T1 relaxation is measured as a function of the magnetic field but is usually presented as a spin–lattice relaxation rate R1 (R1 = 1/T1) vs Larmor frequency. The plot is named the relaxation dispersion profile or relaxation profile. The obtained relaxation profiles, assuming appropriate theoretical models, will allow determining the correlation times of all types of motions affecting the nuclear relaxation in the PILs. For [MIm][TFSI] we determined parameters characterizing the translational and rotational movements of the cation, but more importantly, we also documented the occurrence of proton motion. The dispersion observed in the relaxation profiles below 0.03 MHz at each temperature studied is due to the slow proton exchange (hopping) between the imidazolium cations. To the best of our knowledge, this is the first evidence of a proton motion decoupled from molecular diffusion of cation in bulk protic ionic liquid.

The article also pays attention to the characterization of the thermal properties and electrical conductivity of 1-methylimidazolium bis(trifluoromethylsulfonyl)imide.

2. Materials and Methods

2.1. Materials

The studied ionic liquid 1-methylimidazolium bis(trifluoromethylsulfonyl) imide ([MIm][TFSI]) of a purity grade of 98% was purchased from IoLiTec (Ionic Liquids Technologies GmbH). At room temperature, the chemical compound was in the form of a white powder.

2.2. Thermogravimetric Analysis (TGA/DTG)

Thermogravimetric analysis and its derivative (TGA/DTG) measurements were investigated using a PerkinElmer TGA8000 apparatus. The measurements were carried out under an N2 atmosphere in the temperature range of 303–973 at 10 K/min. The sample of PIL was placed in a platinum pan. Based on the results, were determined: the onset temperature (Tonset) of the decomposition process (5% mass loss) and the maximum decomposition peak (Tmax).

2.3. Conductivity Measurements

Measurements of the complex dielectric permittivity were performed in the frequency region of 100 Hz–5 MHz by using an HP 4194A impedance/gain analyzer. A homemade measuring capacitor consisted of three plane electrodes (the surface of about 1.2 cm2): one central and two grounded on each side, with a distance between them of about 1 mm. The shape of the capacitor electrodes is rectangular and they are made with a gold-plated copper. The probing electric field intensity, E, was equal to about 1 V/mm. The measurements were performed in the temperature range from 298 to 423 K. At first, the samples were cooled down from 343 to 295 K and next the measurements were performed for increasing temperature. The temperature of the measuring cell was controlled with a “Scientific Instruments” device, model 9700, within ±2 × 10–3 °C.

2.4. NMR Experiments

Proton NMR spin–lattice relaxation measurements were performed on a SpinMaster NMR relaxometer from Stelar Company (Mede, Italy) with a B0 magnetic field of 0.5 T. The system enables measurements of relaxation profiles of samples from 10 kHz to the maximum operating magnetic field of 20 MHz (1H Larmor frequency). Below 10 MHz, the prepolarization sequence with the polarizing magnetic field corresponding to 20 MHz was applied for a time of 5T1. Above 10 MHz, the nonpolarized sequence was used. For each resonance frequency, 16 magnetization values versus time in a logarithmic time scale have been recorded. The switching time of the magnet was set to 3 ms. More details on the FFC NMR technique can be found elsewhere.11,15,16 The measured NMR signal comes only from the [MIm]+ cation because the anion of the tested liquid does not contain hydrogen atoms and turned out to be single-exponential for all measured temperatures. Thus, the single T1 relaxation times were calculated with an error of about 5%. 1H-T1 relaxation times of [MIm][TFSI] were studied in the temperature range from 298 up to 343 K (Tm = 328 K). The temperature of the samples was controlled to an accuracy better than 0.5 K. The samples without drying were used for NMR measurements. The powder was transferred directly from the manufacturer’s vial to the NMR tube and sealed immediately. We assume that the water content, if any, was very low because the sample’s high-resolution 1H NMR spectra (Figure 7) did not show any detectable peaks other than the main peaks assigned to the PIL.

Figure 7.

Figure 7

Open in a new tab

Self-diffusion coefficients for [MIm]+ cation (Dcat) and the exchangeable NH proton (DNH) were measured as a function of temperature by PFG NMR. For comparison, the Dcat values calculated from NMR relaxometry were added at temperatures of 298, 308, 333, and 343 K.

The additional T1 measurements at a proton resonance frequency of 500 MHz were performed with a Bruker Avance III HD spectrometer for all studied temperatures. This spectrometer was also used to carry on the pulse-field gradient-type diffusion experiments. The self-diffusion coefficient for the cation of [MIm][TFSI] was obtained with a stimulated echo pulse sequence and calculated by fitting the Stejskal-Tanner equation17 to the signal’s intensity attenuation data according to

2.4. 1

where S and S0 are echo signal intensities, with and without magnetic field gradient pulse applied, respectively, γ is the gyromagnetic ratio, δ is the gradient pulse duration, and D is the self-diffusion coefficient. The measurements were carried out as a function of the magnetic field gradient strength, g, which changed in 32 steps, and the parameters δ and Δ were equal to 1 and 20 ms, respectively, and remained unchanged during the measurements. The measurements were performed for the same temperatures as the T1 measurements.

3. Theory of Spin-Lattice Relaxation

Oscillating electromagnetic fields near nuclear spins are crucial for NMR transitions between spin energy levels. They arise from interactions between spins and the random motions of molecules, which causes the interactions to fluctuate in time. From the theory of nuclear magnetic relaxation in ionic liquids, the most important interactions responsible for the relaxation processes are direct dipole–dipole interactions between nearby nuclear spins11,18−21 although Coulombic interactions, along with van der Waals, hydrogen bonds, and π–π interactions dominate between IL ions.

According to spin relaxation theory, relaxation rates are linear combinations of spectral density functions which are defined as the Fourier transform of the corresponding correlation functions that characterize the dynamic processes leading to stochastic fluctuations of dipole–dipole magnetic interactions and thus causing relaxation processes.11,18−21 The mathematical form of the correlation function and therefore of the spectral density depends on the mechanism of the motion. Therefore, relaxation dispersion profiles, i.e., the relaxation rate of the spin–lattice as a function of the resonance frequency, are a direct fingerprint of this mechanism.

The relaxation dispersion in bulk ionic liquids can be fully described by models taking rotational and translational motions into account. One advantage of NMR relaxometry data analysis is its possibility to separate these contributions, and the total relaxation rate can be written as a sum of these processes

3. 2

where ω0 = 2πν0 = γB0 denotes the proton Larmor frequency in the magnetic field B0 and γ is the gyromagnetic factor. The T1intra expresses the intramolecular magnetic dipole interactions among protons within the same molecules (ions) modulated by local molecular reorientations. The contribution of T1inter relaxation comes from intermolecular magnetic dipolar proton interactions (cation–cation and/or cation–anion), which fluctuate in time as a result of the relative translational self-diffusion of molecules.

The relaxation rate of the homonuclear spin–lattice for intramolecular dipolar interactions, 1/T1intra, is given by the Bloembergen, Purcell, and Pound equation,18 which takes the following form, assuming the Lorentz form of the spectral density

3. 3

where τrot denotes a single rotational correlation time characterizing the fluctuation of the dipole–dipole interactions, μ0 is the magnetic susceptibility of vacuum, ℏ = h/2π where h is the Planck constant, and rHH is the intramolecular distance. The parameters before the bracket define the dipole–dipole interaction constant, marked with the symbol Dintra. The homonuclear intermolecular contribution to the spin–lattice relaxation, 1/T1inter, is given by the following equation18

3. 4

where Jinter(ω) is the translational spectral density function, Inline graphic where nH is the number of nuclei per unit volume, NA is the Avogadro number, ρ is the density of the ionic liquid, and M is the molecular mass. The most commonly used model to describe the effect of translational diffusion on the spin–lattice relaxation time is the force-free hard-sphere model20,21 with Jinter(ω) given as

3. 5

where u is an integration variable, dcc is the distance of the closest approach of interacting species, and Inline graphic is the translational correlation time. The parameter D12 is a relative translational diffusion coefficient expressed as the sum of the self-diffusion coefficient of the interacting ions D12 = D1 + D2. In the present case, identical ions (cations) are interacting and thus the final relation is D12 = 2D. Equation 4, after inserting eq 5, can be rewritten as

3. 6

In the protic ionic liquid, [MIm][TFSI], only the MIm+ cation contains 1H nuclei, and only TFSI– anion contains 19F nuclei. Therefore, we can independently obtain information about the dynamics of ions. However, only the cation dynamics are the subject of our research. The investigation on imidazolium-based PILs with anion containing 19F showed that in the case of proton spin–lattice relaxation, the heteronuclear translational contribution of fluorine anions can be neglected22−24 and the 1H relaxation dispersion can be described taking into account only the intra- and intermolecular interactions of protons.

The contribution, 1/T1ex, to the total relaxation rate due to proton exchange16,25 gives a Lorentz-shaped term

3. 7

where the τex is the proton exchange time and δ2 is the mean quadratic interaction strength between the 1H and interacting spins.

The total proton spin–lattice relaxation rate 1/T1total, of studied PIL consists of a relaxation rate, of 1/T1intra, due to the intramolecular reorientations of cations, 1/T1inter due to the translational diffusion of cations, and 1/T1ex due to the proton exchange

3. 8

After inserting the appropriate relaxation contributions given by eqs 3, 6, and 7 into eq 8, we obtain the final form of the equation that was used to analyze the experimental relaxation dispersion profiles.

4. Results and Discussion

4.1. Thermal Properties

The thermal stability of the [MIm][TFSI] was investigated by thermogravimetric analysis, which provides information on the mass loss of materials with increasing temperature. Figure 1 depicts the TGA thermogram and its corresponding DTG curve for [MIm][TFSI]. PIL is thermally stable to about 573 K. The process of decomposition of [MIm][TFSI] displays a mass loss step above 573 K, with a maximum decomposition temperature of 698 K. The onset decomposition temperature is 583 K. The temperature maximum for [MIm][TFSI] is similar to other alkyl-imidazolium salts.26

Figure 1.

Figure 1

Open in a new tab

TGA thermogram with the corresponding derivative DTG for pure [MIm][TFSI] measured at a 10 K/min heating rate.

4.2. Ionic Conductivity

Figure 2a presents the imaginary part of the dielectric spectra, ε″, of [MIm][TFSI] recorded in the frequency range from 100 Hz to 5 MHz, at different temperatures, and in solid and liquid phases. In the temperature range studied, the PIL exhibits the solid-to-liquid phase transition. The melting process in the PIL manifests itself as a significant change in the character of the permittivity temperature dependence.

Figure 2.

Figure 2

Open in a new tab

Imaginary part of the dielectric spectra [MIm][TFSI] as a function of frequency (f) for different temperatures (a) and the corresponding real part of conductivity spectra (b).

The imaginary part of the dielectric spectrum, for low enough frequencies in comparison to those where the maximum of the dielectric loss of the studied liquid occurs, can be expressed by reduced Debye equation: ε″ = σDC/ωε0 where σDC is called a direct current ionic conductivity. According to this relation, ε″(see Figure 2a) should present straight lines of slope −1 (on a log–log scale). The data ε″ presented in Figure 2a can be transformed into the real part of conductivity spectra, σ′, according to the relation where ε0 = 8.85 pF/m is the permittivity of free space, see Figure 2b. The value of σDC was calculated from the linear dependence of log ε″ vs log f which occurs in the high-frequency range (from about 25 kHz to 5 MHz) above 323 K and in the low-frequency range (from about 100 Hz to 15 kHz) in the temperature range of 298–318 K. For real samples, some deviation from Ohm law is observed as seen in Figure 2 at low frequencies for high temperatures and at high frequencies for low temperatures. The decrease in the slope of the logε″ vs. logf dependence observed in the low frequencies (for high temperatures) manifests itself as a decrease in the electric conductivity of [MIm][TFSI]. The effect is a consequence of the double-layer formation near the blocking electrodes of the measuring capacitor.27 The increase of σ′(f) observed at higher frequencies (for low temperatures) is usually associated with relaxation processes. In the case of ionic liquids, these may be reorientation movements of ions.28

The physical quantity σDC obtained in our experiment for solid and liquid phases in [MIm][TFSI] is presented in Figure 3 as a function of temperature. A clear solid–liquid transition is visible at 328 K. This result is in line with previously reported by Abdurrokhman10 that only the PILs with very short or very long alkyl chains have clear solid–liquid transitions. The studied PIL shows the highest conductivity among this type of PILs and confirms the previously demonstrated relationship that the conductivity decreases with the increase of the alkyl chain.10

Figure 3.

Figure 3

Open in a new tab

Temperature dependence of the dc-conductivity for the protic ionic liquid [MIm][TFSI]. The figure also shows the temperatures for which a full analysis of the dispersion profiles of NMR relaxation times was performed.

4.3. Cation Dynamics

A series of 1H NMR spin–lattice relaxation times were measured for the cation of the protic ionic liquid [MIm][TFSI] in the temperature range from 298 to 343 K.

The temperature range was selected to cover the solid and liquid phases of the PIL and the transition region. The relaxation dispersion profiles are shown in Figure 4 as a function of the external magnetic field expressed in the Larmor frequency units.

Figure 4.

Figure 4

Open in a new tab

1H relaxation dispersion profiles for [MIm]+ cation in the temperature range from 298 to 343 K.

The relaxation rate, R1, strongly depends on the temperature and becomes shorter when the temperature increases. A significant increase in R1 is observed in all dispersion profiles below 0.03 MHz. Above this value, R1 gradually decreases as the Larmor frequency increases. No indication of sudden changes in the relaxation value was observed, which would be a signature of the phase transition process from the solid to the liquid state (328 K). The same applies to the shape of the relaxation dispersion profile, which remains similar in both phases. Figure 5a,b presents a sample fitting to the experimental data collected at selected temperatures, showing the separate contributions to the 1H spin–lattice relaxation. The experimental data also contained the R1 values measured at 500 MHz. These data were of great importance to the analysis of the results because they allowed for a more faithful reproduction of the shape of the relaxation profile. The spin–lattice relaxation in the entire range of studied frequencies was analyzed based on the model given by eq 8. The solid lines in Figure 5 present the best fit of the model to the experimental data, while the dashed lines reflect the contribution of individual relaxation mechanisms. We used a web-based solution for professional model fitting29 created and described by Sebastiao30 to analyze the data.

Figure 5.

Figure 5

Open in a new tab

1H relaxation dispersion profiles for [MIm]+ cation in a solid (a, c) and liquid phase (b, d).

The solid lines represent the best fit of eq 8 to the experimental points with the fitting parameters given in Table 1. The dashed lines represent contributions due to translational diffusion, rotational dynamics, and proton exchange.

Table 1. Parameters characterizing, the rotational, τrot, and translational dynamics, τtrans, and proton exchange, τex, between cations [MIm]+ along with the distance of the closest cation approach, dcc.a.

T (K) τrot (s) dcc (m) τtrans (s) τex (s)
298 1.06 × 10–10 (±0.13 × 10–11) 4.95 × 10–10 (±0.10 × 10–10) 5.94 × 10–9 (±0.01 × 10–10) 5.18 × 10–5 (±0.37 × 10–5)
308 6.21 × 10–11 (±0.47 × 10–11) 4.95 × 10–10 (fixed) 4.01 × 10–9 (±0.02 × 10–10) 5.11 × 10–5 (±0.14 × 10–5)
333 4.10 × 10–11 (±0.13 × 10–11) 0.95 × 10–10 (fixed) 2.01 × 10–9 (±0.01 × 10–10) 6.41 × 10–5 (±0.15 × 10–5)
343 3.69 × 10–11 (±0.04 × 10–11) 4.95 × 10–10 (fixed) 1.56 × 10–9 (±0.55 × 10–10) 3.32 × 10–5 (±0.09 × 10–5)
Open in a new tab
a

The 298–308 K temperature range covers the PIL solid phase, while the 333–343 K range covers the liquid phase.

In the studied frequency range, the relaxation is dominated by homonuclear intermolecular relaxation due to translational diffusion (H–H). In agreement with recent investigations on imidazolium-based ILs with anions containing 19F, the contribution due to heteronuclear relaxation (H–F) for 1H can be ignored.22−24 Generally, 1H relaxation rates also include contributions due to 1H–19F dipole–dipole interactions modulated in time by the relative translational diffusion of ions. We are aware that not considering this contribution in the analysis of dispersion curves is a simplification, but taking them into account means the need to determine three additional parameters in the fitting procedure. The experimental data were very well reproduced without the contribution of 1H–19F relaxation, as seen in Figure 5, which justifies our simplification. The contribution of molecular rotations which modulate the intramolecular interactions gives only a constant contribution and becomes significant in the higher frequency range. At the lowest tested frequencies, below 0.03 MHz, the influence of an additional relaxation mechanism becomes visible. Unfortunately, we could only study this process in the narrow frequency range due to hardware limitations. We postulate that the origin of the relaxation enhancement observed at lower frequencies is due to the magnetic interaction of 1H–14N modulated by the slow process of proton exchange (proton transfer) between imidazolium cations: a proton from one cation (−NH proton) “visits” 14N on the side of another cation and leads to the local fluctuation of the field, which gives a contribution to the spectral density at low frequencies. The concept of pairing ions with opposite charges is well-accepted in ionic liquids. The association of like-charged ions seems unlikely. However, there is already experimental evidence supporting such possibilities. Direct spectroscopic evidence for H-bonded cation–cation clusters similar to those known for water and alcohols was reported for the ionic liquid, 1-(2-hydroxyethyl)-3-methylimidazolium tetrafluoroborate.31 Mele et al. reported NOE contacts between protons of imidazolium cations.32 A gradual self-association of the cations was also observed while studying the structure and properties of 4-oxopiperidinium salts [OC5 H8 NH2]X for a series of anions X(−) of decreasing basicity.33

To eliminate the number of fitting parameters in eq 8, we calculated the value of the NH parameter, which is 8.07 × 1028 m–3. Based on the equilibrium geometry of a single imidazolium cation and using the relation Inline graphic were rIS the distance between all protons within the cation and N(N – 1)/2 is the number of IS pairs within the cation, the intramolecular distance rHH of 2.14 × 10–10m was calculated, which gives the Dintra constant in eq 3 of 1.34 × 109 Hz2. The best-fitting parameters of eq 8 to the experimental data from Figure 5 are collected in Table 1.

The proposed NMR relaxation model fits well with the experimental relaxation profiles and the values of parameters τrot,τtrans, and dcc are consistent with those obtained for similar cations.10,12,23,24 The parameter τex, can be compared with previously published data for proton exchange in different materials. To our knowledge, proton exchange by the FFC NMR method has only been evidenced in water as dispersion of the proton spin–lattice relaxation time observed below 104 Hz.16,25 It was assumed that the origin of this dispersion is due to the slow exchange of protons between different oxygen environments, which modulates the magnetic 1H–17O interaction. The fitted parameter τex was equal to 3.4 × 10–4 s at 296 K and 1.4 × 10–4 s at 353 K.16,25 In our case τex is the order of 10–5 s. This means that the proton exchange process in the tested PIL is faster than in water and thus contributes to relaxation at higher frequencies than in water. The proton exchange time in studied PIL correlates well with the value estimated for proton exchange between imidazole in imidazole-doped cellulose, obtained from the analysis of the line shape of solid-state NMR spectra.34

So far, experimental evidence for cation-independent proton transport in PILs comes mainly from measurements of the self-diffusion coefficient of cation hydrogen atoms by PFG NMR.12,13 Such evidence is considered to be a higher DNH value measured for the exchangeable proton than for other protons or the cation as a whole.

The 1D proton NMR spectrum of [MIm][TFSI] at 343 K (liquid phase) along with the chemical structure and peak assignment is shown in Figure 6.

Figure 6.

Figure 6

Open in a new tab

Liquid state 1H NMR spectrum of [MIm]+ cation in the [MIm][TFSI] protic ionic liquid. The NMR peak assignment is consistent with the labeling in the PIL molecular structure in the figure.

It is easy to see that the signal of the exchangeable proton −NH is broader than the signal of the nonexchangeable protons from C2H, C4H, and C5H. Separate 1H NMR resonances from Figure 6 were used to calculate the self-diffusion coefficient of the cation as whole molecules, Dcat, which is the average of the values obtained from the ring protons. The DNH is calculated from the resonance assigned to the NH groups. The values of Dcat and DNH show a similar temperature dependence, with DNH values becoming increasingly higher than Dcat with increasing temperature (see Figure 7). The larger DNH than Dcat can be explained by assuming that in [MIm][TFSI] protons diffuse faster than the [MIm] cation. This means that DNH contains a contribution from Dcat and DH, where DH is the proton self-diffusion coefficient. Therefore, the charge transport mechanism in the studied PIL includes contributions from both the vehicular and Grotthus mechanisms. Our results are contrary to those obtained for other protic ionic liquids based on imidazolium and triazolium13 where the reported values of the self-diffusion coefficients were very similar for all hydrogen atoms, which indicated the diffusion of an exchangeable proton together with the cation.

It is worth comparing the translational self-diffusion coefficient of the cation obtained directly from PFG NMR with those obtained indirectly from the relaxation data analysis. We calculated the relative diffusion coefficient of [MIm] cation in ([MIm][TFSI] taking into account parameters characterizing the translational dynamics of the cation (τtrans and dcc from Table 1) and the relationship Inline graphic. In the case of uncorrelated motion, the relative diffusion coefficient is given as a sum of the self-diffusion coefficient (measured by NMR gradient methods) of interacting ions D12 = D1 + D2) in our case cations of studied PIL. Thus, for identical cations [MIm] the relative diffusion coefficient is twice as large as the self-diffusion coefficient (D12 = 2Dcat). The obtained values are denoted as Dcat from relaxometry (being, in fact, equal to half of the relative translational diffusion coefficient) and plotted in Figure 7.

For a given temperature, the Dcat values from relaxometry should be equal to the Dcat values from PFG NMR, assuming no correlation in cation dynamics. In Figure 7, the values agree well at lower temperatures (298 and 333 K) but differ at high temperatures (333 and 343 K)—the values from the diffusometry measurements are larger. This discrepancy may suggest that the relative cation–cation translational motion becomes more correlated with increasing temperature.35

5. Conclusions

The studied PIL is thermally stable up to approximately 573 K and its electrical conductivity is of the order of 5 × 10–2 S/cm at 423 K. Such parameters are satisfactory if we take into account the possibility of using this ionic liquid to obtain membranes, among others, for fuel cells operating in the intermediate temperature range of 373 to 473 K. For this temperature range, there are still no efficient membrane materials and in this respect, PIL has the potential to fill this gap.

We believe that the most interesting result of the presented research concerns the dynamics of the ionic liquid. In the range of magnetic field strengths, where the presented measurements were made using the FFC NMR method, it was shown that NMR relaxation is sensitive mainly to slow processes occurring in the molecular dynamics of protic ionic liquid [MIm][TFSI]. The relaxation is dominated by the contribution of intermolecular translational diffusion, enhanced at the lowest magnetic field by an additional mechanism, i.e., proton exchange between slowly diffusing cations. To date, most of the reported experimental evidence for cation-independent proton transport comes from PFG NMR measurements of the self-diffusion coefficient of hydrogen atoms. Based on this method, no separate movement of exchangeable protons in the −NH position from the diffusion of parent cations was detected for any of the protic ionic liquids studied so far.10,12,13 Therefore, we believe that the observed low-frequency dispersion of the spin–lattice relaxation times of cationic protons is the first experimental evidence of this phenomenon in PILs. The observation of independent proton transport in PIL was possible due to the sensitivity of the FFC NMR method. The observed phenomenon was additionally supported by measurements of the self-diffusion coefficient using the PFG NMR method. The measured DNH for the exchangeable protons was higher than for the cation as a whole. Finally, we can conclude that NH protons can diffuse via the Grotthuss mechanism and therefore faster than the cationic molecule. Thus, the charge transport mechanism in the studied PIL includes contributions from both the vehicular and Grotthus mechanisms.

Author Contributions

The manuscript was written through the contributions of all authors. All authors have approved the final version of the manuscript.

The authors declare no competing financial interest.

References

  1. Tritt-Goc J.; Lindner Ł.; Bielejewski M.; Markiewicz E.; Pankiewicz R. Proton conductivity and proton dynamics in nanocrystalline cellulose functionalized with imidazole. Carbohydr. Polym. 2019, 225, 115196 10.1016/j.carbpol.2019.115196. [DOI] [PubMed] [Google Scholar]
  2. Greaves T. L.; Drummond C. J. Protic Ionic Liquids: Properties and Applications. Chem. Rev. 2008, 108, 206–237. 10.1021/cr068040u. [DOI] [PubMed] [Google Scholar]
  3. Greaves T. L.; Drummond C. J. Protic Ionic Liquids: Evolving Structure–Property Relationships and Expanding Applications. Chem. Rev. 2015, 115, 11379–11448. 10.1021/acs.chemrev.5b00158. [DOI] [PubMed] [Google Scholar]
  4. Armand M.; Endres F.; Macfarlane D. R.; Ohno H.; Scrosat B. Ionic-liquid materials for the electrochemical challenges of the future. Nat. Mater. 2009, 8, 621–629. 10.1038/nmat2448. [DOI] [PubMed] [Google Scholar]
  5. Danyliv O.; Strach M.; Nechyporchuk O.; Nypelö T.; Martinelli A. Self-Standing, Robust Membranes Made of Cellulose Nanocrystals (CNCs) and a Protic Ionic Liquid: Toward Sustainable Electrolytes for Fuel Cells. ACS Appl. Energy Mater. 2021, 4, 6474–6485. 10.1021/acsaem.1c00452. [DOI] [Google Scholar]
  6. Jankowska I.; Bielejewski M.; Ławniczak P.; Pankiewicz R.; Tritt-Goc J. Solid electrolytes based on cellulose nanocrystals with protic ionic liquid for next-generation fuel cells. Cellulose 2024, 31, 427–444. 10.1007/s10570-023-05640-8. [DOI] [Google Scholar]
  7. Ingenmey J.; Gehrke S.; Kirchner B. How to Harvest Grotthuss Diffusion in Protic Ionic Liquid Electrolyte Systems. ChemSusChem 2018, 11, 1900–1910. 10.1002/cssc.201800436. [DOI] [PubMed] [Google Scholar]
  8. MacFarlane D. R.; Forsyth S. A.; Golding J.; Deacon G. B. Ionic liquids based on imidazolium, ammonium and pyrrolidinium salts of the dicyanamide anion. Green Chem. 2002, 4, 444–448. 10.1039/b205641k. [DOI] [Google Scholar]
  9. Danyliv O.; Martinelli A. Nafion/Protic Ionic Liquid Blends: Nanoscale Organization and Transport Properties. J. Phys. Chem. C 2019, 123 (23), 14813–14824. 10.1021/acs.jpcc.9b02874. [DOI] [Google Scholar]
  10. Abdurrokhman I.; Elamin K.; Danyliv O.; Hasani M.; Swenson J.; Martinelli A. Protic Ionic Liquids Based on the Alkyl-Imidazolium Cation: Effect of the Alkyl Chain Length on Structure and Dynamics. J. Phys. Chem. B 2019, 123, 4044–4054. 10.1021/acs.jpcb.9b01274. [DOI] [PubMed] [Google Scholar]
  11. Kimmich R.NMR—Tomography, Diffusometry, Relaxometry; Springer: Heidelberg, 1997. [Google Scholar]
  12. Hasani M.; Evans L.; Martinelli A. Molecular dynamics involving proton exchange of a protic ionic liquid-water mixture studied by NMR spectroscopy. Phys. Chem. Chem. Phys. 2019, 21, 22014–22021. 10.1039/C9CP03563J. [DOI] [PubMed] [Google Scholar]
  13. Morais E. M.; Idstrom A.; Evans L.; Martinelli A. Transport properties of protic ionic liquids based on triazolium and imidazolium: development of an air-free conductivity setup. Molecules 2023, 28, 5147 10.3390/molecules28135147. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Yaghini N.; Gomez-Gonzalez V.; Varela M.; Martinelli A. Structural origin of proton mobility in a protic ionic liquid/imidazole mixture: insight from computational and experimental results. Phys. Chem. Chem. Phys. 2016, 18, 23195–23206. 10.1039/C6CP03058K. [DOI] [PubMed] [Google Scholar]
  15. Kimmich R.; Anoardo E. Field-cycling NMR relaxometry. Prog. Nucl. Magn. Reson. Spectrosc. 2004, 44, 257–320. 10.1016/j.pnmrs.2004.03.002. [DOI] [PubMed] [Google Scholar]
  16. Kresse B.; Becher M.; Privalov A. F.; Hofman M.; Rossler E. A.; Vogel M.; Fujara F. 1H NMR at Larmor frequency down to 3 kHz by means of Field-Cycling. J. Magn. Reson. 2017, 227, 79–85. 10.1016/j.jmr.2017.02.002. [DOI] [PubMed] [Google Scholar]
  17. Stejskal E. O.; Tanner J. E. Spin diffusion measurements: spin echoes in the presence of a time-dependent field gradient. J. Chem. Phys. 1965, 42, 288–292. 10.1063/1.1695690. [DOI] [Google Scholar]
  18. Abragam A.The Principles of Nuclear Magnetism; Clarendon Press, 1961. [Google Scholar]
  19. Kowalewski J.; Maler M.. Nuclear Spin Relaxation in Liquids: Theory, Experiments, and Applications, 2nd ed.; CRC Press: Boca Raton, FL, 2017. [Google Scholar]
  20. Ayant Y.; Belorizky E.; Alizon J.; Gallice J. Calculation of spectral density resulting from random translational movement with relaxation by magnetic dipolar interaction in liquids. J. Phys. I 1975, 36, 991–1004. [Google Scholar]
  21. Hwang L. P.; Freed J. H. Dynamic effects of pair correlation functions on spin relaxation by translational diffusion in liquids. J. Chem. Phys. 1975, 63, 4017–4025. 10.1063/1.431841. [DOI] [Google Scholar]
  22. Kruk D.; Meier R.; Rachocki A.; Korpała A.; Singh R. K.; Rossler E. A. Determining diffusion coefficients of ionic liquids by means of field cycling nuclear magnetic resonance relaxometry. J. Chem. Phys. 2014, 140, 244509 10.1063/1.4882064. [DOI] [PubMed] [Google Scholar]
  23. Ordikhani-Seyedler A.; Stapf S.; Mattea C. Nuclear magnetic relaxation and diffusion study of the ionic liquids 1-ethyl-and 1-butyl-3-methylimidazolium bis(trifluoromethylsulfonyl)imide confined in porous glass. Magn. Reson. Chem. 2019, 59, 818–829. [DOI] [PubMed] [Google Scholar]
  24. Seyedlar A. O.; Stapf S.; Mattea C. Dynamics of the ionic liquid 1-butyl-3-methyl-imidazolium bis(trifluoromethylsulphonyl)imide studied by nuclear magnetic resonance dispersion and diffusion. Phys. Chem. Chem. Phys. 2015, 17, 1653–1659. 10.1039/C4CP04178J. [DOI] [PubMed] [Google Scholar]
  25. Graf V.; Noack F.; Bene G. J. Proton spin T1 relaxation dispersion in liquid H2O by slow proton exchange. J. Chem. Phys. 1980, 72, 861–863. 10.1063/1.439240. [DOI] [Google Scholar]
  26. Awad W. H.; Gilman J. W.; Nydena M.; Harris R. H.; Sutto T. E.; Callahan J.; Trulove P. C.; DeLong H. C.; Fox D. M. Thermal degradation studies of alkyl-imidazolium salts and their application in nanocomposites. Thermochim. Acta 2004, 409, 3–11. 10.1016/S0040-6031(03)00334-4. [DOI] [Google Scholar]
  27. Świergiel J.; Płowaś I.; Jadżyn J. Temperature and Concentration Dependences of the Electric Conductivity of Dimethyl Sulfoxide + Ammonium Nitrate Electrolytes. J. Chem. Eng. Data 2013, 58, 2302–2306. 10.1021/je400402n. [DOI] [Google Scholar]
  28. Vavra S.; Ahlberg E.; Martinelli A. Charge transfer and electrical double layer of an amphiphilic protic ionic liquid in bulk and when confined in nanochannels. Phys. Chem. Chem. Phys. 2022, 24, 24469–24479. 10.1039/D2CP01634F. [DOI] [PubMed] [Google Scholar]
  29. Sebastiao P. J.; Beira M.; Kumar A.; Cordeiro R.. fitteia (TM), version 56, 2009–2024, http://fitteia.org.
  30. Sebastião P. J. The art of model fitting to experimental results. Eur. J. Phys. 2014, 35, 015017 10.1088/0143-0807/35/1/015017. [DOI] [Google Scholar]
  31. Knorr A.; Ludwig R. Cation-cation clusters in ionic liquids: Cooperative hydrogen bonding overcomes like-charge repulsion. Sci. Rep. 2015, 5, 17505 10.1038/srep17505. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Mele A.; Romano G.; Giannone M.; Ragg E.; Fromza G.; Raos G.; Marcon V. The Local Structure of Ionic Liquids: Cation-Cation NOE Interactions and Internuclear Distances in Neat [BMIM] [BF4] and [BDMIM][BF4]. Angew. Chem., Int. Ed. 2006, 45, 1123–1126. 10.1002/anie.200503745. [DOI] [PubMed] [Google Scholar]
  33. Gamrad W.; Dreier A.; Goddard R.; Pörschke K.-R. Cation-Cation Pairing by N-C-H···O Hydrogen Bonds. Angew. Chem., Int. Ed. 2015, 54, 4482–4487. 10.1002/anie.201408278. [DOI] [PubMed] [Google Scholar]
  34. Zhao L.; Smolarkiewicz I.; Limbach H. H.; Breitzke H.; Pogorzelec-Glaser K.; Pankiewicz R.; et al. Imidazole-doped cellulose as membrane for fuel cells: Structural and dynamic insights from solid-state NMR. J. Phys. Chem. C 2016, 120, 19574–19585. 10.1021/acs.jpcc.6b07049. [DOI] [Google Scholar]
  35. Kruk D.; Masiewicz E.; Lotarska S.; Markiewicz R.; Jurga S. Correlated dynamics in ionic liquids by means of NMR relaxometry: Butyltriethylammonium bis(trifluoromethanesulfonyl)imide as an example. Int. J. Mol. Sci. 2021, 22, 9117 10.3390/ijms22179117. [DOI] [PMC free article] [PubMed] [Google Scholar]

Articles from The Journal of Physical Chemistry. B are provided here courtesy of American Chemical Society

RESOURCES