Translational diffusion of components in concentrated LiTFSI solutions in acetonitrile according to multinuclear pulse NMR data

doi:10.21203/rs.3.rs-6959449/v1

Download PDF

Research Article

Translational diffusion of components in concentrated LiTFSI solutions in acetonitrile according to multinuclear pulse NMR data

https://doi.org/10.21203/rs.3.rs-6959449/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 21 Jul, 2025

Read the published version in Applied Magnetic Resonance →

You are reading this latest preprint version

The diffusion coefficients of solvent molecules, cations and anions in concentrated “LiTFSI–acetonitrile” solutions were measured in the temperature range from − 15 to + 75^oC. The activation energies (E_a) for diffusion motion of each species were calculated. The features of translational motion in this system are discussed. The data indicate the existence of some predominant structural composition with the molar ratio “salt/ACN”≈1/4.5. The acetonitrile molecules move more freely in the structural "mosaic" than solvated ions. A new approach to describing the concentration dependencies of diffusion coefficients in the case of fast both translational motion and exchange between substructures is suggested.

The search for new, more efficient and/or object-oriented liquid electrolytes for lithium batteries (LIBs) is one of the urgent tasks of modern electrochemical and, generally, physico-chemical research. Thus, it was found that with certain mutual combinations of salts and solvent it is possible to increase the content of ions to such an extent that their amount is comparable to the number of solvent molecules, but these objects remain in the liquid state at room temperature. It is currently believed that due to incomplete solvation of cations/anions caused by insufficient amount of solvent molecules, these highly concentrated electrolytes (HCE), unlike related dilute solutions, exhibit beneficial properties, such as wider electrochemical stability windows [1–4], high values of Li⁺ cation transport numbers [5, 6], etc.

The study of highly concentrated electrolytes of interest to electrochemistry began in the 1960s. At that time, the electrolytes studied were based on water as a solvent and the salts were “common” ones: NaCl, LiCl, CaBr₂, etc. The salts used in those early studies allowed reaching molar “salt/water” ratios not more than 1:10 (close to saturation).

Interest in HCE has recently increased due to the use of highly soluble lithium salts with TFSI- and FSI- anions in electrochemical energy storage systems in general and, in particular, in lithium-ion batteries. The commonly used solvents in HCEs are acetonitrile, carbonates (ethylene carbonate, dimethyl carbonate, propylene carbonate), and water. A number of studies have been conducted in this direction over the past decade, see, for example, [7–10] and references therein. In these works, the authors investigated structure and dynamics of highly concentrated electrolytes using different research methods (SAXS, QENS, MD and other techniques).

Analyzing the published results, it can be concluded that there are still quite wide areas for new research in this area, the results of which, combined with the available data, will lead to a better understanding of the structure and dynamics of solutions relevant for electrochemistry. We chose LiTFSI solutions in acetonitrile (ACN) for the study, and the NMR-diffusometry method for characterizing the behavior of components in selected objects. The NMR allows one to measure diffusion coefficients (D) for each component in a solution and, thus, to reach more detailed information on the interrelation of transport properties of ions and solvent molecules. The NMR-diffusomertry was successfully applied to a lot of liquid systems (including ionic liquids and HCE/WIS/SIS systems [11‑17]) but as far as we know no studies were undertaken for “LiTFSI-acetonitrile” solutions. In some aspects, the work is a continuation and development of the research presented in Ref. [16, 17].

2.1 Sample preparation

Dry acetonitrile ACN (Sigma-Aldrich, ≤ 10 ppm water content) and lithium bis(trifluoromethanesulfonyl)imide LiTFSI (Acros Organic, 99%) were used for sample preparation. All samples were prepared using the mass method, where an accurately weighed amount of LiTFSI was dissolved in ACN at 25°C. The characteristics of the five prepared samples are presented in Table 1.

Table 1

The characteristics of the samples investigated (m is the molality, i. e. the number of salt moles per 1kg of a solvent).
Sample designation	m (LiTFSI)	$\:{m}_{ACN}/\:{m}_{LiTFSI}$
1m LiTFSI	1.0	24.4
5m LiTFSI	5.0	4.9
6m LiTFSI	6.0	4.1
8m LiTFSI	8.3	2.9
10m LiTFSI	10.0	2.4

Before the diffusion coefficient measurements, ¹H NMR spectra were recorded (500 MHz) for the prepared samples. The spectra exhibit a dominant signal from the CH₃-group of acetonitrile and a very weak signal from the water admixture. The integral ratio I(H₂O)/I(CH₃) is below 0.003, which indicates that water is present in the systems in insignificant quantities.

2.2 NMR measurements

Among all experimental methods for studying diffusion processes, the NMR one occupies a prominent place, since this method is non-contact and non-destructive, does not require the insertion of any material "tag", allows selective measurements of diffusion coefficients of components in a wide range of values [18–20]. There exist a number of NMR methods for the determination of diffusion coefficients, the Pulsed Field Gradient Stimulated Echo method is used in this work: dstebpgp3s — Double Stimulated Echo with bipolar gradients and 3 spoil-pulses (the magnitude of magnetic gradients is 0 ÷ 50 Gauss/cm, diffusion time is 50 ms). The error in determining the diffusion coefficients was no more than 3%. The temperature range used was from − 15 to + 75^oC. A control experiment with samples of different lengths (standard filling and 1 cm high) showed that convection had insignificant effect on the results of determining diffusion coefficients (see Fig. S2).

All NMR experiments were conducted on the basis of Center for Magnetic Resonance in Saint-Petersburg State University using Bruker Avance III 500 MHz.

Fig. 1-3 show the typical dependences of the diffusion coefficients of the ¹H, ⁷Li and ¹⁹F nuclei on temperature in the “LiTFSI – ACN” solutions (1, 5, and 10 mol/kg ACN). The full set of data is presented in Supplementary Information (SI).

As it is obvious from the figures, the diffusion coefficients (CDs) of cations and anions are close but not identical for the minimum salt concentration studied (1m). This is indirectly consistent with the findings of the work Ref. [9] that at low salt concentrations (i.e., with a sufficient number of acetonitrile molecules, see Table 1), the TFSI¯ anion prefers to form weak associates with acetonitrile rather than ion pairs with the cation. However, for high concentrations (5 ÷ 10 m) these DCs practically coincide, which means a strong correlation of the translational motion of ions in the solutions studied. The coincidence of the CDs of the cation and the anion indicates, apparently, the joint diffusion of the counterions in a cluster or domain (like an ionic mosaic with solvent molecules included in it). As it was previously noted in our work [17], we emphasize that it cannot be a simple contact ion pair, since in this case the solution would have low electrical conductivity (which is not confirmed in experiments).

The CDs of acetonitrile significantly exceed the CDs of ions at all concentrations. This means that acetonitrile molecules move more freely in the "mosaic" than solvated ions, that is, the manifestation of the individual behavior of namely acetonitrile molecules differs from the collective one.

Since the graphs in Fig. 1–3 have a rectilinear appearance, one can assume that the activation model is applicable to the objects under study and the activation energy (E_a) for diffusion motion can be calculated. Since during diffusion over the experiment (50 ms), the components of the solution undergo multiple transitions between different substructures (fast exchange), the E_a parameter makes sense of an average parameter that allows us to characterize the average energy barriers for the movement of components in the solution. The calculated E_a values are shown in Table 2 and graphically in Fig. 4.

Table 2

The values of activation energies (kJ/mol) for LiTFSI solutions in acetonitrile.
	1m	5m	6m	8.3m	10m
ACN	10.1 ± 0.2	19.1 ± 0.2	21.5 ± 0.3	27.3 ± 0.4	31.8 ± 0.7
⁷Li	11.3 ± 0.1	20.4 ± 0.2	22.6 ± 0.2	28.7 ± 0.5	32.6 ± 0.5
¹⁹F	11.6 ± 0.3	20.9 ± 0.7	22.9 ± 0.8	29.6 ± 1.7	32.7 ± 0.9

Before discussing the concentration dependences of the diffusion coefficients, it is advisable to make a few remarks. If there are several states of molecules and/or ions in a system, and a fast (on the time scale of the experiment) molecular exchange is realized between these states, then the diffusion coefficient $\:{D}_{exp}$ measured in the experiment is usually expressed by a formula, which in the case of two (for simplicity) states has the form (see, for example, [21]):

$$\:{D}_{exp}={p}_{1}{D}_{1}+{p}_{2}{D}_{2}\:,\:\text{w}\text{i}\text{t}\text{h}\:{p}_{1}+{p}_{2}=1,\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:$$

where $\:{D}_{\text{1,2}\:\:}$ are the diffusion coefficients in states 1 and 2; $\:{p}_{\text{1,2}\:\:}$ is the relative content of diffusing species in states 1 and 2. In a more convenient form, expression (1) can be rewritten as:

$$\:{\:\:\:\:\:\:\:\:\:\:\:\:\:\:D}_{exp}={D}_{1}+{p}_{2}\left({D}_{2}-{D}_{1}\right).$$

It can be seen from formula (2) that if, for example, $\:{D}_{1}\gg\:\:$ $\:{D}_{2},$ then the rapidity of the averaged diffusion process is determined to a large extent by the value of $\:{D}_{1}$ ("faster" structure), which is surprising (both from the point of view of logic and experience).

On the other hand, the nuclear magnetic relaxation times (T₁ and T₂) are an important source of information about molecular motion. It is proved [18–20], that in the case of rapid exchange, the NMR relaxation rates (1/T_1,2) are averaged according to a formula similar to (1):

(1/T_1,2)_exp = p₁ (1/T_1,2)₁ + p₂ (1/T_1,2)₂. (3)

In case of fast movement 1/T_1,2 = const/D [18–20], and we come to a contradiction between expressions (1, 2) and (3). Since expression (3) leads to a more natural deduction that the average motion under the conditions discussed reflects a greater influence of the slow process, one can conclude: to discuss the effect of increasing the salt content (in our example, p₂), it is advisable to use the dependence of 1/$\:{D}_{exp}$ on the salt concentration expressed in relative units (or in units of molality). It should be noted that earlier one of the authors of this work developed and successfully applied the approach for studying the microstructure of electrolyte solutions based on the investigation of concentration and temperature dependences of the rates of magnetic relaxation of solvent (and solute) nuclei. A formula of type (3) (for an arbitrary number of states) was used to interpret the experimental data [22]. The same approach was used in the study of “hydration shells" of CH₂ groups of ω-amino acids in aqueous solutions [23].

Figures 5 and 6 present the dependences of 1/$\:{D}_{exp}\:$on the concentration of LiTFSI for the nuclei of solvent molecules ¹H (ACN) and ⁷Li, ¹⁹F of salt ions. The full set of data is presented in SI. All the graphs show some patterns that are particularly pronounced at low temperatures:

the diffusion coefficients of all components decrease by almost 2 orders of magnitude with an increase in the concentration of LiTFSI to 10 mol/kg ACN, and the diffusion coefficients of solvent molecules are approximately in 2–3 times greater than those for ions (in the entire studied concentration range);

the graphs show a bend in the region of 5.5 mol/kg of ACN, that is, at the molar ratio “salt/ACN”≈1/4.5, that indicates the existence of some predominant structural composition in the system.

These results confirm the conclusions noted during the discussion of temperature dependences.

The multinuclear NMR-diffusomertry (¹H nuclei for solvent molecules, ⁷Li and ¹⁹F for salt ions) was applied to the investigation of molecular mobility of all components in concentrated “LiTFSI–acetonitrile” solutions. The evolution of diffusion coefficients in the temperature range from − 15 to + 75^oC is traced. The results obtained indicate that the increase in salt concentration leads to the mobile ionic network ("mosaic") that causes a drastic dynamical slowdown. This model is confirmed by the fact that the activation energies (E_a) for diffusion motion of acetonitrile molecules, cations and anions are identical. Besides, the diffusion coefficients of cations and anions are very close over the entire concentration range. The data indicate the existence of some predominant structural composition with the molar ratio “salt/ACN” ≈ 1/4.5. Nevertheless, the diffusion coefficients of acetonitrile significantly exceed ones of ions at all concentrations. This means that acetonitrile molecules move more freely in the structural "mosaic" than solvated ions, i.e., the manifestation of the individual behavior of namely acetonitrile molecules differs from the collective one. To specify the structural models, it is planned to conduct computer modeling of microstructure (quantum chemistry and modeling of molecular dynamics).

We hope that our experimental data and model discussions will help in the development of safe and promising electrolytes with better electrochemical characteristics.

Acknowledgements

NMR measurements were carrying out in the Center for Magnetic Resonance of Research Park of St. Petersburg State University

Contributions

Kirill A. Mukhin: experiment, data processing, design of the manuscript;

Vladimir V. Matveev: idea of experiment, discussion of results, writing;

Vladimir I. Chizhik: supervision, idea of interpreting concentration dependencies, writing and editing.

Ethical Approval

Not applicable to this study.

Funding

This work was supported by the Russian Science Foundation (project № 23-23-00049).

Availability of data and materials

Not applicable to this study.

Alvarado J., Schroeder M. A., Zhang M. et al. (2018). A carbonate-free, sulfone-based electrolyte for high-voltage Li-ion batteries. Materials Today, 21(4), 341–353. http://dx.doi.org/10.1016/j.mattod.2018.02.005
Nilsson V., Younesi R., Brandell D., Edström K. and Johansson P. (2018). Critical evaluation of the stability of highly concentrated LiTFSI - Acetonitrile electrolytes vs. graphite, lithium metal and LiFePO 4 electrodes. Journal of Power Sources, 384, 334–341. https://doi.org/10.1016/j.jpowsour.2018.03.019
Dudley J. T., Wilkinson D. P., Thomas G., et al. (1991). Conductivity of electrolytes for rechargeable lithium batteries. Journal of Power Sources, 35(1), 59–82. https://doi.org/10.1016/0378-7753(91)80004-H
Petibon R., Aiken C. P., Ma L., Xiong D. and Dahn J. R. (2015). The use of ethyl acetate as a sole solvent in highly concentrated electrolyte for Li-ion batteries. Electrochimica Acta, 154, 287–293. https://doi.org/10.1016/j.electacta.2014.12.093
Iamprasertkun P., Ejigu A. and Dryfe R. (2020). Understanding the Electrochemistry of “Water-in-salt” Electrolytes: Basal Plane Highly Ordered Pyrolytic Graphite as a Model System. Chemical Science. https://doi.org/10.1039/D0SC01754J
Montenegro M. I. and Pletcher D. (1986). The determination of the kinetics of electron transfer using fast sweep cyclic voltammetry at microdisc electrodes. Journal of Electroanalytical Chemistry and Interfacial Electrochemistry, 200(1-2), 371–374. https://doi.org/10.1016/0022-0728(86)90069-0
Lundin F., Aguilera L., Hansen H. W., et al. (2021). Structure and dynamics of highly concentrated LiTFSI/acetonitrile electrolytes. Physical Chemistry Chemical Physics, 23(25), 13819–13826. https://doi.org/10.1039/D1CP02006D
Nilsson V., Kotronia A., Lacey M. J., Edstrom, K. and Johansson P. (2020). Highly concentrated LiTFSI – EC electrolytes for Li metal batteries. ACS Applied Energy Materials, 3, 200-207. https://doi.org/10.1021/acsaem.9b01203
Xinyi L., Shao-Chun L., Soenke S., Randall E. W., Y Z and Tao L. (2023). Relationship of the Molecular Structure and Transport Properties of Imide-Based Lithium Salts of “Acetonitrile/Water-in-Salt” Electrolytes. Chemistry of Materials, 35(16), 6415–6422. https://doi.org/10.1021/acs.chemmater.3c01148
Ma L. and Jiang J. (2024). Vehicular Motions Dominate the Ion Transport in Concentrated LiTFSI Aqueous Solutions? J Phys Chem Lett., 15(17), 4531-4537. https://doi.org/10.1021/acs.jpclett.4c00791
Damodaran K. (2022). Recent advances in NMR spectroscopy of ionic liquids. Progress in Nuclear Magnetic Resonance Spectroscopy, 129, 1–27. https://doi.org/10.1016/j.pnmrs.2021.12.001
Kumar V., Reddy R. R., Kumar B. V. N. P. et al. (2019). Lithium speciation in the LiPF₆/PC electrolyte studied by two-dimensional heteronuclear overhauser enhancement and pulse-field gradient diffusometry NMR. J. Phys. Chem. C, 123, 9661–9672. https://doi.org/10.1021/acs.jpcc.8b11599
Baranauskaite V., Pestova O., Vovk M., Matveev V. and Lähderanta E. (2019). Local dynamics in LiCl–CsCl–D₂O water-in-salt solutions according to NMR relaxation. Physical Chemistry Chemical Physics. https://doi.org/10.1039/C9CP03878G
Ishikawa T., Sudoh T., Shigenobu K. et al. (2025). Weakly coordinating monoether-based concentrated electrolytes: effects of frustrated Li ion coordination on ion transport and Li metal battery performance. Electrochimica Acta, 527, 146234. https://doi.org/10.1016/j.electacta.2025.146234
Tyutyukin K.V., Ievlev A.V., Matveev V.V. et al. (2024). Molecular Mobility Study of 1-Butyl-1-Methylpyrrolidinium Bis(trifluoromethylsulfonyl)imide Ionic Liquid by NMR Diffusometry. Appl Magn Reson., 55, 785–794. https://doi.org/10.1007/s00723-024-01678-4
Matveev V.V., Pestova O.N., Tyutyukin K.V. and Chizhik V.I. (2023). Molecular Mobility in Mixed “Water-in-Salt” Solutions of LiOAc and KOAc According to NMR Data. Applied Magnetic Resonance, 54(10), 971–978. https://doi.org/10.1007/s00723-023-01558-3
Mukhin K.A., Pestova O.N., Matveev V.V. and Chizhik V.I. (2024). Translational Mobility in Ternary Systems “Lithium Acetate–Cesium Acetate–Water” According to PFG NMR Data. Applied Magnetic Resonance, 55(8), 775–783. https://doi.org/10.1007/s00723-024-01670-y
Abragam A. (1961). The Principles of Nuclear Magnetism. Oxford University Press, Amen House, Londond.
Slichter C. P. (1980). Principles of Magnetic Resonance. Springer-Verlag Berlin Heidelberg New York, 655 pp.
Chizhik V.I., Chernyshev Y. S., Donets A. V. et al. (2014). Magnetic Resonance and Its Applications. Springer-Verlag. 782 pp. https://doi.org/10.1007/978-3-319-05299-1
Wijesekera D., Stait‐Gardner T., Gupta A. et al. (2018). A complete derivation of the Kärger equations for analyzing NMR diffusion measurements of exchanging systems. Concepts in Magnetic Resonance Part A, 47A(2), e21468. https://doi.org/10.1002/cmr.a.21468
Chizhik V. I. (1997). NMR relaxation and microstructure of aqueous electrolytesolutions. Molecular Physics, 90:4, 653-660. Doi: 10.1080/002689797172363
Rabdano S. O., Donets A. V., Vovk M. A. et al. (2015). “Hydration Shells” of CH₂Groups of ω-Amino Acids as Studied by Deuteron NMR Relaxation. Journal of Physical Chemistry B, 119, 13358–13366. https://doi.org/10.1021/acs.jpcb.5b06584

No competing interests reported.

SI2025.docx

Download PDF

Journal Publication

published 21 Jul, 2025

Read the published version in Applied Magnetic Resonance →

Editorial decision: Accepted
09 Jul, 2025
Reviews received at journal
06 Jul, 2025
Reviews received at journal
04 Jul, 2025
Reviewers agreed at journal
30 Jun, 2025
Reviewers agreed at journal
27 Jun, 2025
Reviewers invited by journal
27 Jun, 2025
Editor assigned by journal
25 Jun, 2025
Submission checks completed at journal
24 Jun, 2025
First submitted to journal
23 Jun, 2025

You are reading this latest preprint version

Translational diffusion of components in concentrated LiTFSI solutions in acetonitrile according to multinuclear pulse NMR data

Status:

Journal Publication

Version 1

Abstract

Figures

1 Introduction

2 Experimental

2.1 Sample preparation

2.2 NMR measurements

3 Results and discussion

4 Conclusion

Declarations

References

Additional Declarations

Supplementary Files

Status:

Journal Publication

Version 1