New Network Polymer Electrolytes Based on Ionic Liquid and SiO2 Nanoparticles for Energy Storage Systems

Elementary processes of electro mass transfer in the nanocomposite polymer electrolyte system by pulse field gradient, spin echo NMR spectroscopy and the high-resolution NMR method together with electrochemical impedance spectroscopy are examined. The new nanocomposite polymer gel electrolytes consisted of polyethylene glycol diacrylate (PEGDA), salt LiBF4 and 1—ethyl—3—methylimidazolium tetrafluoroborate (EMIBF4) and SiO2 nanoparticles. Kinetics of the PEGDA matrix formation was studied by isothermal calorimetry. The flexible polymer–ionic liquid films were studied by IRFT spectroscopy, differential scanning calorimetry and temperature gravimetric analysis. The total conductivity in these systems was about 10−4 S cm−1 (−40 °C), 10−3 S cm−1 (25 °C) and 10−2 S cm−1 (100 °C). The method of quantum-chemical modeling of the interaction of SiO2 nanoparticles with ions showed the advantage of the mixed adsorption process, in which a negatively charged surface layer is formed from Li+ BF4— ions on silicon dioxide particles and then from ions of the ionic liquid EMI+ BF4−. These electrolytes are promising for use both in lithium power sources and in supercapacitors. The paper shows preliminary tests of a lithium cell with an organic electrode based on a pentaazapentacene derivative for 110 charge–discharge cycles.


Introduction
In recent years, ionic liquids (ILs) have been increasingly used as components of polymer electrolytes for energy storage systems [1,2]. ILs have a number of advantages, such as low flammability, low vapor pressure and a wide thermal, chemical and electrochemical stability window [3].
He et al. [14] found that ILs are utilized as synthesis and dispersion media for nanoparticles as well as for surface functionalization. Ionic liquid and nanoparticle hybrid systems are governed by a combined effect of several intermolecular interactions between their constituents, for each interaction, including van der Waals, electrostatic, steric and hydrogen bonding. Various self-organized structures based on nanoparticles in ionic liquids are generated as a result of a balance of these intermolecular interactions. These structures, including nanoparticle-stabilized ionic-liquid-containing emulsions, possess properties of both ionic liquids and nanoparticles, which render them useful as novel materials, especially in electrochemical applications.
Unlike traditional salts, ionic liquids usually consist of large asymmetric polyatomic ions with an ionic radius above about 5 to 10 times that of monatomic ions such as Li + or Na + . The large ionic size of the ion liquids increases the average distance between the charge cation and anion centers, reducing the electrostatic interaction strength. The ions or ion clusters are attracted to the nanoparticle surface by electrostatic forces [14]. The particles of metals have great attraction. The ionic liquid cations are attracted to the surface of a negatively charged nanoparticle to form a positive ion layer, and then counter ions form a second layer on the nanoparticle surface by electrostatic attraction.
In addition, there are nanoparticles that interact less with IL ions. One such example is colloidal fumed silica SiO 2 . Lithium salt additives are used to stabilize them, which is a requirement for the use of such systems in lithium-ion batteries. Thus, Nordström et al. [15] investigated the stability and interactions in dispersions of colloidal fumed silica SiO 2 (Aerosil 200, Evonik Resource Efficiency GmbH, Antwerp, Belgium) and the ionic liquid 1-butyl-3-methylimidazolium tetraflouroborate (BMIBF 4 ) as a function of the Li salt concentration (LiBF 4 ). The increased stability with the addition of Li salt was found by Raman spectroscopy which is explained by the formation of a more stable solvation layer, where Li ions accumulate on the surface.
Electrolytes based on ionic liquids are mainly used in supercapacitors [16][17][18], where ionic liquid ions are charge carriers. There are also polymer electrolytes based on ILs for lithium current sources. Here, the competitive ion transport of lithium cations and IL cations has strong influences on the electrochemical processes [19][20][21][22][23].
The addition of different ionic liquids has various effects on the properties of Li + ion conductive polymer electrolytes. In addition, the physical properties of ILs (in particular, viscosity and dielectric constant) have an important role in the structure design and conducting properties of polymer electrolytes. Low viscosity leads to an increase in the segmental mobility of the polymer chains. On the other hand, a high dielectric constant of the ionic liquid increases ion pair dissociation and, therefore, is the cause of the charge carrier concentration. Both of the above aspects contribute to the increase in polymer gel electrolyte ionic conductivity.
This study deals with the synthesis and study of an electro mass transfer of new network polymer electrolytes based on an EMIBF 4 ionic liquid with the introduction of various amounts of fumed silica SiO 2 (Aerosil 380) nanoparticles (0, 2, 4, 6 wt.%). The SiO 2 nanoparticles have a very small particle size of 7 nm with a narrow distribution and a very developed surface of 380 g/m 2 due to porosity. On such a surface, lithium salt molecules can be adsorbed and dissociated into free ions [26][27][28][29].
We have previously studied polymer gel electrolytes based on polyethylene glycol diacrylate (PEGDA) as a polymer network, which showed good properties as a threedimensional network with ethylene oxide units, which retain well a large amount of a liquid phase, consisting of both traditional aprotic solvents and ionic liquids such as EMIBF 4 Membranes 2023, 13, 548 3 of 21 and 1-butyl-3-methylimidazolium tetrafluoroborate BMIBF 4 [4,30]. The mechanism of ionic and molecular transport in new nanocomposite systems based on fumed SiO 2 was investigated by the NMR method, electrochemical impedance spectroscopy and quantumchemical modeling, which are the most informative methods for such complex systems.
The structures of PEGDA and EMIBF 4 are shown in Figure 1.
phase, consisting of both traditional aprotic solvents and ionic liquids such as EMIBF4 1-butyl-3-methylimidazolium tetrafluoroborate BMIBF4 [4,30]. The mechanism ionic and molecular transport in new nanocomposite systems based on fumed SiO2 investigated by the NMR method, electrochemical impedance spectroscopy and qu tum-chemical modeling, which are the most informative methods for such complex tems.

PEGDA Kinetics of Radical Polymerization
The kinetics of radical polymerization of PEGDA in the presence of the ionic liq salt LiBF4, EC, SiO2 nanopowder and benzoyl peroxide as an initiator was studied by thermal calorimetry on a DAK-1-1 differential automatic calorimeter (EZAN, C nogolovka, Russia) at 60 °C. The reaction mixture was placed into glass ampoules for orimetric measurements and sealed.

Synthesis of Nanocomposite Polymer Electrolytes
Nanocomposite polymer electrolytes (NPEs) were synthesized by the rad polymerization of PEGDA in the presence of the radical initiator PB without an inert vent.
The composition of the polymerizable mixture was as follows: PEGDA, LiBF4, EMIBF4, SiO2 and 1 wt.% PB for the whole sample. The compositions of NPEs are liste Table 1 in molar proportions of the components and in Table S1, ESI in mass perc where the synthesis procedure in detail (ESI) is given.  (1)(2)(3)(4)(5) indicate the sites of the 1 H and 13 C atoms (for the description of the NMR spectra).

PEGDA Kinetics of Radical Polymerization
The kinetics of radical polymerization of PEGDA in the presence of the ionic liquid, salt LiBF 4 , EC, SiO 2 nanopowder and benzoyl peroxide as an initiator was studied by isothermal calorimetry on a DAK-1-1 differential automatic calorimeter (EZAN, Chernogolovka, Russia) at 60 • C. The reaction mixture was placed into glass ampoules for calorimetric measurements and sealed.

Synthesis of Nanocomposite Polymer Electrolytes
Nanocomposite polymer electrolytes (NPEs) were synthesized by the radical polymerization of PEGDA in the presence of the radical initiator PB without an inert solvent.
The SEM image of the initial SiO 2 powder is shown in Figure 2a.  To study samples by NMR method, the NPEs were synthesized in closed glass capillaries with a diameter of d = 4 mm, l = 50 mm. The capsules with NPEs were closed and placed in standard 5 mm ampoules for NMR examination.
The SEM image of the initial SiO2 powder is shown in Figure 2a. An optical photo of the final nanocomposite electrolyte film is shown in Figure 2b. It can be seen from Figure 2b that the film is transparent.

Differential Scanning Calorimetry (DSC) Method
The glassy transition temperature in the temperature range from −150 to 50 °С and the homophase nature of the NPEs were determined from the differential scanning calorimetry (DSC) data obtained on a DSC 822e Mettler-Toledo instrument (Kutznacht an der Zürichsee, Switzerland) with the Star software at a scanning rate of 5 deg min −1 .

Fourier Transform Infrared Spectroscopy (FTIR)
The FTIR spectra of thin-film electrolyte samples and initial components (PEGDA, IL and EC) were recorded on an IRTracer-100 FTIR spectrometer (Shimadzu, Germany) at room temperature in a wave number range of 400-4000 cm −1 with a spectral resolution of 2-4 cm −1 .

Thermogravimetric Analysis (TGA) Method
The TGA data for the samples were obtained on a TGA/SDTA851 Mettler-Toledo instrument (China) in the temperature range from 30 to 150 °С at a heating rate of 5 deg min −1 . An optical photo of the final nanocomposite electrolyte film is shown in Figure 2b. It can be seen from Figure 2b that the film is transparent.

Differential Scanning Calorimetry (DSC) Method
The glassy transition temperature in the temperature range from −150 to 50 • C and the homophase nature of the NPEs were determined from the differential scanning calorimetry (DSC) data obtained on a DSC 822e Mettler-Toledo instrument (Kutznacht an der Zürichsee, Switzerland) with the Star software at a scanning rate of 5 deg min −1 .

Fourier Transform Infrared Spectroscopy (FTIR)
The FTIR spectra of thin-film electrolyte samples and initial components (PEGDA, IL and EC) were recorded on an IRTracer-100 FTIR spectrometer (Shimadzu, Germany) at room temperature in a wave number range of 400-4000 cm −1 with a spectral resolution of 2-4 cm −1 .

Thermogravimetric Analysis (TGA) Method
The TGA data for the samples were obtained on a TGA/SDTA851 Mettler-Toledo instrument (China) in the temperature range from 30 to 150 • C at a heating rate of 5 deg min −1 .

High-Resolution NMR
High-resolution spectra for 1 H, 7 Li, 11 B, 13 C and 19 F were recorded on a Bruker Avance III 500 MHz NMR spectrometer. The measurements at frequencies of 500, 194, 160, 126 and 471 MHz for 1 H, 7 Li, 11 B, 13 C and 19 F, respectively, were carried out at room temperature (22 ± 1 • C). The chemical shift scale was calibrated with the DMSO-d6 signal in the capillary as an external standard (2.50 ppm for 1 H). The 1 H, 7 Li and 19 F NMR spectra were obtained using the standard sequence π/2 pulses, FID. No signal accumulation was applied. To obtain the 13 C NMR spectra, a standard sequence from the TopSpin (Bruker, Billerica, MA, USA) zgpg30 library was used. The sequence is an accumulation of signals from 30 • pulses with the suppression of the 1 H spin-spin interaction for the duration of all of the experimental times. The number of repetitions was ns = 512, and the delay between the repetition sequence was d1 = 1.0 s. For interpretation of 1 H and 13 C, twodimensional 13 C-1 H HSQC correlation spectra were recorded (standard pulse sequence from the TopSpin library (Bruker)).

NMR with Pulsed Field Gradient
The NMR measurements on a Bruker Avance-III 400 MHz NMR spectrometer equipped with the diff60 gradient unit (the maximum field gradient amplitude was 30 T m −1 ) were carried out at the temperature 22 ± 1 • C. The NMR measurements of 1 H (diffusion of solvent molecules EC and ionic liquid IL, EMI − ), 7 Li (diffusion of lithium cations) and 19 F (diffusion of anions BF 4 − ) were carried out with operating frequencies of 400, 155.5 and 376.5 MHz, respectively. The stimulated spin echo sequence was applied. The details of self-diffusion coefficient measurements are given in [31,32]. The experimental NMR parameters of pulse sequences were the following: π/2 pulse was 9 µs ( 1 H), 9 µs ( 7 Li) and 10 µs ( 19 F); gradient pulse duration time δ was 1.0 ( 1 H), 1.0 ( 7 Li) and 3.0 ( 19 F) ms; diffusion time was 19.7 ( 1 H), 19.7 ( 7 Li) and 49.0 ( 19 F) ms; repetition time 3 s; and the diffusion 32 steps with maximum field gradient amplitude g were 3.5 ( 1 H), 11.5 ( 7 Li) and 4.0 (19F) T m −1 . The strength of the gradient changed linearly. The measurement error of the self-diffusion coefficients was 5%. The temperature dependences of the diffusion coefficients were measured at the temperature range from 22 to 60 • C.

Electrochemical Methods
To measure the conductivity of NPE film samples by the electrochemical impedance method in symmetrical stainless steel cells (SS)//SS with an area equal to 0.2 cm 2 , a Z-2000 impedance meter (Elins, Chernogolovka, Russia) was used in the frequency range from 1 Hz to 2 MHz with a signal amplitude of 10 mV. The cell impedance was detected in the temperature range from −40 to 100 • C. Four measurements were carried out for each sample. The measurement error was not higher than 2%.
Symmetrical cells with Li metal and LiOTAP organic cathodes were assembled in coin-type cells CR2032. To measure the resistance of the boundary of NPE/electrode by the electrochemical impedance method in symmetrical cells Li/Li and LiOTAP/LiOTAP, a Z-2000 impedance meter was used analogically.
The electrochemical performance of the Li//LiOTAP batteries was evaluated using a BTS-5 V 10 mA battery analyzer (Neware Technology Ltd., Shenzhen, China) by performing charge/discharge cycling at current rates C/2 in a range of 0.7-3.5 V. LiOTAP was synthesized and characterized in [33]. The electrochemical performance of LiOTAP was evaluated in coin-type lithium batteries. The cathode composition comprised 45 wt.% of LiOTAP, 50 wt.% of conductive carbon black (Timical Super C65) and 5 wt.% of PVDF polymer binder (Kynar Flex HSV 900, Arkema, Colombes, France). The procedure for assembling cells with a polymer electrolyte differed from that mentioned in [33] in that an NPE film was placed instead of a separator with a liquid electrolyte.

Quantum-Chemical Modeling
The structure of complexes of different ions with solvent molecules and SiO 2 was studied using the nonempirical Perdew-Burke-Erzernhof (PBE) exchange-correlation functional [34] [35]. The geometry of larger systems containing a counterion and additional solvent molecules was optimized using the effective Hamiltonian method [36] taking into account the van der Waals interaction. The Priroda package [37] was used for all the calculations carried out at the Joint Supercomputer Center of the Russian Academy of Sciences.

Quantum-Chemical Modeling
The structure of complexes of different ions with solvent molecules and SiO2 was studied using the nonempirical Perdew-Burke-Erzernhof (PBE) exchange-correlation functional [34] [35]. The geometry of larger systems containing a counterion and additional solvent molecules was optimized using the effective Hamiltonian method [36] taking into account the van der Waals interaction. The Priroda package [37] was used for all the calculations carried out at the Joint Supercomputer Center of the Russian Academy of Sciences. The kinetic curves ( Figure 3a) followed that the main part of the polymerization of the studied compositions takes about 3 h. To accelerate the synthesis time of finished electrolyte films, it is necessary to carry out polymerization in a stepwise mode (60, 70 and 80 °C). This is because an increase in temperature by 10° leads to an increase in the polymerization rate by several times. Therefore, the post-polymerization time reduced, and the polymer electrolytes with a maximum conversion were obtained.

DSC of NPEs
All compositions of NPEs and initial ionic liquid were studied by DSC. The ionic liquid had only one phase transition; the glass transition temperature Tg = −103 °C. All NPE samples had two Tg values that characterized EMIBF4 and PEGDA. The results are presented in Table 2. Figure 4 shows the DSC diagrams of EMIBF4, NPE1, NPE4 and NPE5 as an example.  The kinetic curves ( Figure 3a) followed that the main part of the polymerization of the studied compositions takes about 3 h. To accelerate the synthesis time of finished electrolyte films, it is necessary to carry out polymerization in a stepwise mode (60, 70 and 80 • C). This is because an increase in temperature by 10 • leads to an increase in the polymerization rate by several times. Therefore, the post-polymerization time reduced, and the polymer electrolytes with a maximum conversion were obtained.

DSC of NPEs
All compositions of NPEs and initial ionic liquid were studied by DSC. The ionic liquid had only one phase transition; the glass transition temperature T g = −103 • C. All NPE samples had two T g values that characterized EMIBF 4 and PEGDA. The results are presented in Table 2. Figure 4 shows the DSC diagrams of EMIBF 4 , NPE1, NPE4 and NPE5 as an example.
Note. T 01 is the onset of the relaxation transition; T g1 is the temperature of the first relaxation transition (relaxation transition of the crosslinked polymer matrix); ∆T 1 is the range of the first relaxation transition; T g2 is the glass transition temperature of the second relaxation transition; ∆T 2 is the range of the second relaxation transition. Note. T01 is the onset of the relaxation transition; Tg1 is the temperature of the first relaxation trans tion (relaxation transition of the crosslinked polymer matrix); ΔT1 is the range of the first relaxatio transition; Tg2 is the glass transition temperature of the second relaxation transition; ΔT2 is the rang of the second relaxation transition.

FTIR Аnalysis of NPEs
The peak of the carbonyl group of PEGDA at 1721 cm −1 shifts to a range of 1733 cm − which is caused by the three-dimensional crosslinking of diacrylate in a medium of a larg amount of the liquid phase as shown by the quantum-chemical modeling of PEGDA cross linking ( Figure S1, ESI) and confirmed experimentally in [38].
The peak of the carbonyl group of ethylene carbonate in the composition o PEGDA-LiBF4-3EC undergoes a strong shift. This is due to the formation of a solvat shell of the lithium cation by EC molecules. The theoretical IR spectra of the LiBF4-3EC solvate showed this effect ( Figure S2, ESI).
The peaks of the carbonyl group of EC upon the addition of SiO2 nanoparticles retur to their original position ( Figure 5). Most likely, they came out of the EC solvate environ ment. As can be seen in Figure S3, ESI showed an enlarged spectrum of the carbony group.

FTIR Analysis of NPEs
The peak of the carbonyl group of PEGDA at 1721 cm −1 shifts to a range of 1733 cm −1 , which is caused by the three-dimensional crosslinking of diacrylate in a medium of a large amount of the liquid phase as shown by the quantum-chemical modeling of PEGDA crosslinking ( Figure S1, ESI) and confirmed experimentally in [38].
The peak of the carbonyl group of ethylene carbonate in the composition of PEGDA-LiBF 4 -3EC undergoes a strong shift. This is due to the formation of a solvate shell of the lithium cation by EC molecules. The theoretical IR spectra of the LiBF 4 -3EC solvate showed this effect ( Figure S2, ESI).
The peaks of the carbonyl group of EC upon the addition of SiO 2 nanoparticles return to their original position ( Figure 5). Most likely, they came out of the EC solvate environment. As can be seen in Figure S3, ESI showed an enlarged spectrum of the carbonyl group.

TGA of NPEs
The TGA dependences of all NPE compositions are shown in Figure 6. It is seen from Figure 6a that the polymer electrolytes are stable up to 100 °С. Instrumental inaccuracy is the mass loss of up to 1%. A slight weight loss may indicate a loss of moisture. Moisture during preparation for the study could get into the sample. The end

TGA of NPEs
The TGA dependences of all NPE compositions are shown in Figure 6.

TGA of NPEs
The TGA dependences of all NPE compositions are shown in Figure 6. It is seen from Figure 6a that the polymer electrolytes are stable up to 100 °С. Instrumental inaccuracy is the mass loss of up to 1%. A slight weight loss may indicate a loss of moisture. Moisture during preparation for the study could get into the sample. The end It is seen from Figure 6a that the polymer electrolytes are stable up to 100 • C. Instrumental inaccuracy is the mass loss of up to 1%. A slight weight loss may indicate a loss of moisture. Moisture during preparation for the study could get into the sample. The end of the first stage of the TGA diagram at 100 • C confirmed this. In addition, when examining samples by IR spectroscopy before and after heating up to 150 • C, it was shown that the main peaks of all components did not change ( Figure S4, ESI). Figure 6b shows the TGA diagrams up to 600 • C of the NPE4 and NPE5 compositions with maximum conductivity. Figure 6b shows the multistage character of sample weight loss. This explains the gradual loss of each component. The loss of ethylene carbonate (bp = 248 • C) occurs first. The ionic liquid, apparently, decomposes together with the polymer matrix at 390 • C. Silicon dioxide remains in the residue. Under extreme conditions, it is possible that SiO 2 will insulate between the electrodes (if they are stable up to these temperatures).

High-Resolution NMR
The 1 H and 13 C NMR spectra to check the purity and confirm the NPE compositions were recorded. The 1 H and 13 C NMR spectra for all NPEs compared to the ionic liquid are shown in Figures 7 and 8, respectively. The 1 H and 13 C spectra differ in integral signal intensities due to different molar ratios of EMIBF 4 to solvent. The 7 Li, 19 F and 11 B NMR spectra were also obtained ( Figures S5-S7, ESI). of the first stage of the TGA diagram at 100 °C confirmed this. In addition, when examining samples by IR spectroscopy before and after heating up to 150 °C, it was shown that the main peaks of all components did not change ( Figure S4, ESI). Figure 6b shows the TGA diagrams up to 600 °C of the NPE4 and NPE5 compositions with maximum conductivity. Figure 6b shows the multistage character of sample weight loss. This explains the gradual loss of each component. The loss of ethylene carbonate (bp = 248 °C) occurs first. The ionic liquid, apparently, decomposes together with the polymer matrix at 390 °С. Silicon dioxide remains in the residue. Under extreme conditions, it is possible that SiO2 will insulate between the electrodes (if they are stable up to these temperatures).

High-Resolution NMR
The 1 H and 13 C NMR spectra to check the purity and confirm the NPE compositions were recorded. The 1 H and 13 C NMR spectra for all NPEs compared to the ionic liquid are shown in Figures 7 and 8, respectively. The 1 H and 13 C spectra differ in integral signal intensities due to different molar ratios of EMIBF4 to solvent. The 7 Li, 19 F and 11 B NMR spectra were also obtained ( Figures S5-S7, ESI).    Figure 7 shows that the signals in the 1 H NMR spectra of the polymer electrolytes are significantly broader than those in the pure EMIBF4. The signal of ethylene carbonate is also broadened (~4 ppm). The signal broadening is caused by the formation of a branched network polymer structure formed by PEGDA, which considerably impedes the chaotic motion of EMIBF4 and ЕС [39]. The 1 H NMR spectrum of the electrolyte exhibits a very broad signal from -O-CH2-CH2-O-of the polymer matrix with a maximum at ~3 ppm. This signal in the 13 C-1 H HSQC spectrum correlates with the 13 С signal at 69.2 ppm in the 13 C NMR spectrum ( Figure S8, ESI).
A 2D DOSY spectrum to confirm signal assignment in the proton spectrum was also recorded (Figure 9).  Figure 7 shows that the signals in the 1 H NMR spectra of the polymer electrolytes are significantly broader than those in the pure EMIBF 4 . The signal of ethylene carbonate is also broadened (~4 ppm). The signal broadening is caused by the formation of a branched network polymer structure formed by PEGDA, which considerably impedes the chaotic motion of EMIBF 4 and EC [39]. The 1 H NMR spectrum of the electrolyte exhibits a very broad signal from -O-CH 2 -CH 2 -O-of the polymer matrix with a maximum at 3 ppm. This signal in the 13 C-1 H HSQC spectrum correlates with the 13 C signal at 69.2 ppm in the 13 C NMR spectrum ( Figure S8, ESI).
A 2D DOSY spectrum to confirm signal assignment in the proton spectrum was also recorded (Figure 9).
A spectrum DOSY signal in the spectrum of a mixture of molecules depending on the diffusion coefficients (Y-axis, Ds) allows them to separate. The EC solvent molecule is smaller than the IL and therefore more mobile. Thus, the signals from the less mobile IL (Ds(IL)) and the signal from the EC (Ds(EC)) from Figure 9 are easy to separate.

Self-Diffusion Coefficients (SDCs) According to the PFG NMR Data
The SDCs on 1 H, 7 Li and 19 F for all the NPE compositions were measured by NMR with PGF. The diffusion decays on all nuclei of all compositions were exponential ( Figure S9, ESI). The measurements of the self-diffusion coefficients D s on 1 H make it possible to determine the mobility of EMIBF 4 and EC (the analysis of diffusion decays of the signals from the ionic liquid or ethylene carbonate solvent allows one to estimate their mobilities separately). The D s on 7 Li corresponds to the mobility of lithium cations, and that on 19 F corresponds to the mobility of the BF 4 − anion. A spectrum DOSY signal in the spectrum of a mixture of molecules depending on the diffusion coefficients (Y-axis, Ds) allows them to separate. The EC solvent molecule is smaller than the IL and therefore more mobile. Thus, the signals from the less mobile IL (Ds(IL)) and the signal from the EC (Ds(EC)) from Figure 9 are easy to separate.

Self-Diffusion Coefficients (SDCs) According to the PFG NMR Data
The SDCs on 1 H, 7 Li and 19 F for all the NPE compositions were measured by NMR with PGF. The diffusion decays on all nuclei of all compositions were exponential ( Figure  S9, ESI). The measurements of the self-diffusion coefficients Ds on 1 H make it possible to determine the mobility of EMIBF4 and EC (the analysis of diffusion decays of the signals from the ionic liquid or ethylene carbonate solvent allows one to estimate their mobilities separately). The Ds on 7 Li corresponds to the mobility of lithium cations, and that on 19 F corresponds to the mobility of the BF4 − anion.

Self-Diffusion Coefficients on 1 H Nucleus
The results of measuring Ds for the NPE1-5 compositions are given in Table 3. The Ds values for pure ionic liquid EMIBF4 are presented for comparison. Table 3 shows that the mobility of EC and IL molecules increases with an increase in the fraction of IL in the polymer electrolyte.

Self-Diffusion Coefficients on 1 H Nucleus
The results of measuring D s for the NPE1-5 compositions are given in Table 3. The D s values for pure ionic liquid EMIBF 4 are presented for comparison.  Table 3 shows that the mobility of EC and IL molecules increases with an increase in the fraction of IL in the polymer electrolyte.
The D s of EC increases by more than six times with an increase in the mass content of IL from 0 to 6% (compositions NPE1-4). In this case, the D s of IL increases by four times. This is probably related to the "loosening" of the electrolyte polymer network upon the introduction of IL molecules. In this case, the D s of pure IL is more than an order of magnitude higher than the D s of IL in the polymer electrolyte (NPE 2). An increase in the content of the SiO 2 additive at the same mass content of IL (NPE4 and NPE5) does not lead to a significant change in the mobility of the NPE components. The temperature dependences of the self-diffusion coefficients D s on 1 H for EMIBF 4 and EC molecules were measured in the range from 22 to 60 • C ( Figure 10). Self-diffusion coefficient temperature D s (T) dependences are approximated by the Arrhenius equitation: where D 0 is the temperature independent value, R is a gas constant, and T is the absolute temperature. E a is self-diffusion activation energy. The temperature dependences of the self-diffusion coefficients Ds on 1 H for EMIBF4 and EC molecules were measured in the range from 22 to 60 °С ( Figure 10). Self-diffusion coefficient temperature Ds(T) dependences are approximated by the Arrhenius equitation: where D0 is the temperature independent value, R is a gas constant, and T is the absolute temperature. Ea is self-diffusion activation energy.
(a) (b) The dependences are Arrhenius. The activation energies of diffusion were calculated (Table 3). It is shown that an increase in the IL content in the composition of the polymer electrolyte leads to a decrease in the activation energy of the diffusion of EC and IL molecules up to ~40 to ~30 kJ mol −1 . The activation energy of the diffusion of the pure ionic liquid was 21 kJ mol −1 .

Self-Diffusion Coefficients on 7 Li Nucleus
The results of measuring Ds ( 7 Li) for the NPE1-5 compositions are given in Table 4. An increase in the SDC for Li + cations with an increase in the IL content in the polymer electrolyte, as in the case of the molecular mobility of the IL and EC components, was observed. The Ds of lithium cations increases by more than an order of magnitude (with an increase in the mass content of IL from 0 to 6 wt.%).
An increase in the addition of SiO2 from 2 to 6 wt.% (transition from composition 4 to composition 5) leads to a slight increase in the Ds of lithium cations. Note that the SDC (Li + ) is ten times lower than the SDC of EMIBF4 despite the small size of the lithium cation compared to the IL molecule. The temperature dependences of the self-diffusion coefficients Ds on 7 Li were measured in the range from 22 to 60 °С (Figure 11a). The dependences are Arrhenius. The activation energies of diffusion were calculated (Table 3). It is shown that an increase in the IL content in the composition of the polymer electrolyte leads to a decrease in the activation energy of the diffusion of EC and IL molecules up to~40 to~30 kJ mol −1 . The activation energy of the diffusion of the pure ionic liquid was 21 kJ mol −1 .

Self-Diffusion Coefficients on 7 Li Nucleus
The results of measuring D s ( 7 Li) for the NPE1-5 compositions are given in Table 4. An increase in the SDC for Li + cations with an increase in the IL content in the polymer electrolyte, as in the case of the molecular mobility of the IL and EC components, was observed. The D s of lithium cations increases by more than an order of magnitude (with an increase in the mass content of IL from 0 to 6 wt.%).
An increase in the addition of SiO 2 from 2 to 6 wt.% (transition from composition 4 to composition 5) leads to a slight increase in the D s of lithium cations. Note that the SDC (Li + ) is ten times lower than the SDC of EMIBF 4 despite the small size of the lithium cation compared to the IL molecule. The temperature dependences of the self-diffusion coefficients D s on 7 Li were measured in the range from 22 to 60 • C (Figure 11a). Membranes 2023, 13, x FOR PEER REVIEW 13 of 21 (a) (b) Figure 11. Temperature dependences of the diffusion coefficients on (a) 7 Li and (b) 19 F nuclei.
The dependences are Arrhenius. The activation energies of diffusion were calculated (Table 4). Both Ea values for 7 Li and 19 F nuclei increase upon passing from the NPE5 to NPE1 composition (with a decrease in the IL content in the composite). It is shown that an increase in the IL content in the composition of the polymer electrolyte leads to a decrease in the activation energy of the diffusion of Li + from ~50 to ~35 kJ mol −1 .  Figure S7, ESI) shows one singlet, which is a superposition of the signals from the lithium salt anion BF4 − and the second BF4 − in the EMIBF4 ionic liquid. This was caused by the rapid chemical exchange of BF4 − . Thus, the SDCs measured on the 19 F nuclei are the weighted average of the mobility of the BF4 − and BF4 − anions in EMIBF4. Table 4 shows that the mobility of BF4 − increases by an order of magnitude with an increase in the mass content of IL from 0 to 6% (compositions NPE1-4). An increase in the content of the SiO2 additive at the same mass content of IL (compositions 4 and 5) does not lead to a significant change in the mobility of BF4 − cations in the NPEs. The temperature dependences of the self-diffusion coefficients Ds on 19 F were measured in the range from 22 to 60 °С (Figure 11b). The dependences are Arrhenius. The diffusion activation energies calculated by Formula (1) are presented in Table 4. It is shown that an increase in the IL content in the composition of the polymer electrolyte leads to a decrease in the activation energy of the diffusion of BF4 − from ~37 to ~27 kJ mol −1 .
Thus, IL molecules contribute to an increase in the mobility of all components in the polymer matrix. The SDCs of Li + and BF4 − increase by an order of magnitude with an increase in the mass content of IL from 0 to 6% (compositions NPE1-4). The dependences are Arrhenius. The activation energies of diffusion were calculated (Table 4). Both E a values for 7 Li and 19 F nuclei increase upon passing from the NPE5 to NPE1 composition (with a decrease in the IL content in the composite). It is shown that an increase in the IL content in the composition of the polymer electrolyte leads to a decrease in the activation energy of the diffusion of Li + from~50 to~35 kJ mol −1 .

Self-Diffusion Coefficients on 19 F Nucleus
The 19 F NMR spectrum ( Figure S7, ESI) shows one singlet, which is a superposition of the signals from the lithium salt anion BF 4 − and the second BF 4 − in the EMIBF 4 ionic liquid. This was caused by the rapid chemical exchange of BF 4 − . Thus, the SDCs measured on the 19 F nuclei are the weighted average of the mobility of the BF 4 − and BF 4 − anions in EMIBF 4 . Table 4 shows that the mobility of BF 4 − increases by an order of magnitude with an increase in the mass content of IL from 0 to 6% (compositions NPE1-4). An increase in the content of the SiO 2 additive at the same mass content of IL (compositions 4 and 5) does not lead to a significant change in the mobility of BF 4 − cations in the NPEs. The temperature dependences of the self-diffusion coefficients D s on 19 F were measured in the range from 22 to 60 • C (Figure 11b). The dependences are Arrhenius. The diffusion activation energies calculated by Formula (1) are presented in Table 4. It is shown that an increase in the IL content in the composition of the polymer electrolyte leads to a decrease in the activation energy of the diffusion of BF 4 − from~37 to~27 kJ mol −1 . Thus, IL molecules contribute to an increase in the mobility of all components in the polymer matrix. The SDCs of Li + and BF 4 − increase by an order of magnitude with an increase in the mass content of IL from 0 to 6% (compositions NPE1-4).
The Li + cation has the lowest diffusion coefficient D s . According to the obtained data, D s (EMI + ) > Ds (BF 4 − ) >> Ds(Li + ). Increasing the addition of SiO 2 from 2 to 6 wt.% (transition from composition 4 to composition 5) leads to a slight increase in the D s of lithium cations.

NPE Conductivity
The conductivity of the obtained NPE samples was measured by the electrochemical impedance method in symmetrical SS//SS cells in the temperature range from −40 • C to 100 • C. The typical Nyquist plots of the cells are shown in Figure S10, ESI. The measurement results are given in Table S2, ESI and in Figure 12

NPE Conductivity
The conductivity of the obtained NPE samples was measured by the electrochemical impedance method in symmetrical SS//SS cells in the temperature range from −40 °С to 100 °С. The typical Nyquist plots of the cells are shown in Figure S10, ESI. The measurement results are given in Table S2, ESI and in Figure 12. The Arrhenius temperature dependence of the conductivity for all compositions (Figure 12) had a break in the temperature range from 15 to 25 °С, and, hence, the effective activation energy of conductivity was calculated in two ranges ( Table 5). The activation energies of conductivity and diffusion are compared. For the NPE1 composition, the activation energy of conductivity is noticeably lower than the activation energy of 7 Li diffusion and comparable with the activation energies of 19 F diffusion. With an increase in ionic liquid content, there is the same tendency of activation energies, but Ea values for conductivity decrease most strongly, 2.5 times. It is seen from Figure 12 and Table 5 that the composition with 6 mol IL (NPE4 and NPE5) has the highest conductivity and the most effective activation energy among all thin-film electrolytes. However, these values are higher for the NPE5 composition (with 6 wt.% SiO2). This indicates the positive contribution of SiO2 nanoparticles.

Electrochemical Study of NPEs in Li//LiOTAP Cells
In this work, battery prototypes with a cathode based on the lithium salt of the tetraazapentacene derivative LiOTAP were assembled. The electrochemical reduction and oxidation of LiOTAP are shown in Scheme 1. Each molecule of LiOTAP can undergo eight- The Arrhenius temperature dependence of the conductivity for all compositions ( Figure 12) had a break in the temperature range from 15 to 25 • C, and, hence, the effective activation energy of conductivity was calculated in two ranges ( Table 5). The activation energies of conductivity and diffusion are compared. For the NPE1 composition, the activation energy of conductivity is noticeably lower than the activation energy of 7 Li diffusion and comparable with the activation energies of 19 F diffusion. With an increase in ionic liquid content, there is the same tendency of activation energies, but E a values for conductivity decrease most strongly, 2.5 times. It is seen from Figure 12 and Table 5 that the composition with 6 mol IL (NPE4 and NPE5) has the highest conductivity and the most effective activation energy among all thin-film electrolytes. However, these values are higher for the NPE5 composition (with 6 wt.% SiO 2 ). This indicates the positive contribution of SiO 2 nanoparticles.

Electrochemical Study of NPEs in Li//LiOTAP Cells
In this work, battery prototypes with a cathode based on the lithium salt of the tetraazapentacene derivative LiOTAP were assembled. The electrochemical reduction and oxidation of LiOTAP are shown in Scheme 1. Each molecule of LiOTAP can undergo eightelectron oxidation also releasing eight Li + cations, which corresponds to the theoretical specific capacity of 468 mA h g −1 .
FOR PEER REVIEW 15 of 21 electron oxidation also releasing eight Li + cations, which corresponds to the theoretical specific capacity of 468 mA h g −1 .
Scheme 1. Electrochemical reduction and oxidation of LiOTAP.
First, the compatibility of NPE3-5 with Li-anode and LiOTAP-cathode materials was investigated by the electrochemical impedance method. The method of "liquid-phase therapy" for LiOTAP//LiOTAP cells was applied. The liquid electrolyte 1M LiTFSI in Scheme 1. Electrochemical reduction and oxidation of LiOTAP.
First, the compatibility of NPE3-5 with Li-anode and LiOTAP-cathode materials was investigated by the electrochemical impedance method. The method of "liquid-phase therapy" for LiOTAP//LiOTAP cells was applied. The liquid electrolyte 1M LiTFSI in DOL/DME (1:1 vol.) was used similar to [39]. The results are shown in Figure 13.

Scheme 1. Electrochemical reduction and oxidation of LiOTAP.
First, the compatibility of NPE3-5 with Li-anode and LiOTAP-cathode materials was investigated by the electrochemical impedance method. The method of "liquid-phase therapy" for LiOTAP//LiOTAP cells was applied. The liquid electrolyte 1M LiTFSI in DOL/DME (1:1 vol.) was used similar to [39]. The results are shown in Figure 13.     Figure 14a. Figure 14b shows the charge-discharge profiles of Li/NPE3/LiOTAP cells at C/2 current for the 1-20, 100 cycle numbers. Figure S12 (ESI) shows the Coulomb efficiency (CE) during the cycling of this cell. Figure 14 shows that the LiOTAP organic cathode material loses its capacity in the first cycles but then stabilizes at 150 mAh g −1 . This substance belongs to the class of "small molecules" and is able to dissolve in the process of charge-discharge. The effect of LiOTAP dissolution in a liquid electrolyte was observed in [33]. However, with the use of a polymer electrolyte, no LiOTAP dissolution was observed in this work. This was confirmed by IR spectroscopy in the study of NPE4 samples before and after charge-discharges cycling. The opening of the cell after cycling was made in an Ar glove box. The lithium anode remained a shiny metal. No characteristic peaks of the IR spectra of the LiOTAP material in the film composition were observed ( Figure S13, ESI).
A significant excess of the Coulomb efficiency above 100% can be seen in Figure S12, ESI. Obviously, this effect is due to SiO 2 nanoparticles, since in their absence, the reversible operation of the Li/EMIBF 4 /LiOTAP cell was fixed, although short-lived ( Figure S14, ESI). We attribute this effect to the reduction reaction of the EMI + cation on the surface of the SiO 2 nanoparticle. Quantum-chemical calculations show a significant binding energy (49.4 kcal mol −1 or 2.14 eV) of two radicals formed during the reduction of EMI + ( Figure  S15a, ESI). An excess amount of discharge capacity over charging capacity (Figure 14b) occurs at low potentials of 0.7 V. Apparently, here, the activity of an ionic liquid with a carbon material (50 wt.% of cathode) similar to supercapacitors or a dual graphite battery manifests itself [40][41][42][43][44]. The Coulomb efficiency would be 100% if EMIBF 4 Figure 14 shows that the LiOTAP organic cathode material loses its capacity in the first cycles but then stabilizes at 150 mAh g −1 . This substance belongs to the class of "small molecules" and is able to dissolve in the process of charge-discharge. The effect of LiOTAP dissolution in a liquid electrolyte was observed in [33]. However, with the use of a polymer electrolyte, no LiOTAP dissolution was observed in this work. This was confirmed by IR spectroscopy in the study of NPE4 samples before and after charge-discharges cycling. The opening of the cell after cycling was made in an Ar glove box. The lithium anode remained a shiny metal. No characteristic peaks of the IR spectra of the LiOTAP material in the film composition were observed ( Figure S13, ESI).
A significant excess of the Coulomb efficiency above 100% can be seen in Figure S12, ESI. Obviously, this effect is due to SiO2 nanoparticles, since in their absence, the reversible operation of the Li/EMIBF4/LiOTAP cell was fixed, although short-lived ( Figure S14, ESI). We attribute this effect to the reduction reaction of the EMI + cation on the surface of the SiO2 nanoparticle. Quantum-chemical calculations show a significant binding energy (49.4 kcal mol −1 or 2.14 eV) of two radicals formed during the reduction of EMI + (Figure S15a, ESI). An excess amount of discharge capacity over charging capacity (Figure 14b) occurs at low potentials of 0.7 V. Apparently, here, the activity of an ionic liquid with a carbon material (50 wt.% of cathode) similar to supercapacitors or a dual graphite battery manifests itself [40][41][42][43][44]. The Coulomb efficiency would be 100% if EMIBF4 remained unchanged.
We suppose that during a long cycle (24 h), the next reaction has time to occur. An electron from the carbon material is transferred to the imidazolium cation, and the resulting di-radical ( Figure S15a, ESI) binds to the surface of the SiO2 nanoparticle ( Figure S15b, ESI). This radical then adds the next radical but with a significant gain in energy. This mechanism was confirmed by the presence of changes in the IR spectra of the polymer electrolyte before and after charge-discharge cycling in the battery ( Figure S16, ESI).
The discharge capacity in the first cycle is close to the theoretical one (8e − LiOTAP redox transition) ( Figure 14). Subsequently, it decreases to 150 mAh g −1 , which corresponds to a two-electron transition. Since we consider the active participation of the EMI + cation in these processes, it can be assumed that during the discharge, a partial insertion of the bulky EMI + cation instead of Li + occurs. These bulky cations shield the redox active groups of LiOTAP, and the capacitance efficiency drops in the first 40 charge-discharge cycles. We suppose that during a long cycle (24 h), the next reaction has time to occur. An electron from the carbon material is transferred to the imidazolium cation, and the resulting di-radical ( Figure S15a, ESI) binds to the surface of the SiO 2 nanoparticle (Figure S15b, ESI). This radical then adds the next radical but with a significant gain in energy. This mechanism was confirmed by the presence of changes in the IR spectra of the polymer electrolyte before and after charge-discharge cycling in the battery ( Figure S16, ESI).
The discharge capacity in the first cycle is close to the theoretical one (8e − LiOTAP redox transition) ( Figure 14). Subsequently, it decreases to 150 mAh g −1 , which corresponds to a two-electron transition. Since we consider the active participation of the EMI + cation in these processes, it can be assumed that during the discharge, a partial insertion of the bulky EMI + cation instead of Li + occurs. These bulky cations shield the redox active groups of LiOTAP, and the capacitance efficiency drops in the first 40 charge-discharge cycles.

Quantum-Chemical Modeling
Using quantum-chemical calculations by means of the density functional theory, models of the LiBF 4 salt and the EMIBF 4 ionic liquid were constructed in the form of an associate of two ion pairs, as well as a (LiBF 4 ) 2 (EMIBF 4 ) 2 cluster containing both types of associates ( Figure 15).
3, x FOR PEER REVIEW 17 of 21

Quantum-Chemical Modeling
Using quantum-chemical calculations by means of the density functional theory, models of the LiBF4 salt and the EMIBF4 ionic liquid were constructed in the form of an associate of two ion pairs, as well as a (LiBF4)2(EMIBF4)2 cluster containing both types of associates ( Figure 15). As a model of silicon dioxide nanoparticles, we considered a cluster with a core of 17 bound SiO2 molecules and a hydrated surface as a result of the addition of 6 H2O molecules ( Figure 16). As a model of silicon dioxide nanoparticles, we considered a cluster with a core of 17 bound SiO 2 molecules and a hydrated surface as a result of the addition of 6 H 2 O molecules ( Figure 16). As a model of silicon dioxide nanoparticles, we considered a cluster with a core of 17 bound SiO2 molecules and a hydrated surface as a result of the addition of 6 H2O molecules ( Figure 16).  At the same time, the incorporation of the (EMIBF 4 ) 2 cluster into the surface layer of ions for Model I leads to a greater gain in energy (34.6 and 51.1 kcal mol −1 , respectively). This indicates the advantage of the mixed adsorption process, which results in the formation of a negatively charged surface layer of Li + and BF 4 -ions. This conclusion is consistent with the results of experimental work [15] on the study of the interaction of SiO 2 (Aerosil 200), LiBF 4 and BMIBF 4 nanoparticles using Raman spectroscopy. In the case of our system with SiO 2 nanoparticles (Aerosil 380), due to their highly porous structure, the Raman spectra turned out to be not informative.

Conclusions
New nanocomposite polymer gel electrolytes based on polyethylene glycol diacrylate (PEGDA), LiBF 4 salt and 1-ethyl-3-methylimidazolium tetrafluoroborate (EMIBF 4 ) and SiO 2 nanoparticles (Aerosil 380) were synthesized and studied. The SiO 2 nanoparticles with a highly porous surface are unique in their nature of interaction with ions of lithium salt ions and ionic liquid. In this work, the electro mass transfer in such complex systems was investigated. The NMR method of high resolution and with a pulsed field gradient was used. Together with the electrochemical impedance data, information was obtained on the high conductivity of these systems. The total electrical conductivity was about 10 −4 S cm −1 (−40 • C), 10 −3 S cm −1 (25 • C) and 10 −2 S cm −1 (100 • C). Based on the obtained data, quantum-chemical modeling of the adsorption of ionic complexes on the surface of a SiO 2 nanoparticle was carried out. It was shown that the most energy-efficient process is mixed adsorption, which forms a negatively charged surface layer of Li + BF 4 − ions and then EMI + BF 4 − ions. This is also confirmed by the results of NMR with a pulsed field gradient. These electrolytes are promising for use both in lithium power sources and in supercapacitors. This paper shows the preliminary tests of a lithium cell with an organic electrode based on a pentaazapentacene derivative for 110 charge-discharge cycles.
Supplementary Materials: The following supporting information can be downloaded at https://www.mdpi.com/article/10.3390/membranes13060548/s1. Table of Contents; Table S1: Compositions of the nanocomposite polymer gel electrolytes; Synthesis of the nanocomposite polymer electrolytes, Figure S1: Calculated structure of (a) the initial PEGDA, consisting of 6 (-CH2CH2O-) units; (b) the simplest element of a polymer network, consisting of 4 connected PEGDA fragments, where the dotted line indicates the places of their crosslinking, in which broken C-C bonds are replaced by C-H bonds; and (c) the theoretical IR spectra of these models, Figure S2: The calculated structure of (a) the solvate complex of the Li+ cation with three EC molecules and the BF4counterion; and (b) the theoretical IR spectrum of this solvate complex, Figure S3: FTIR spectra of NPE4 and NPE5 versus EC solvent, PEGDA polymer and the polymer electrolyte without SiO2 and EMIBF4 in a range of 1650÷1900 cm−1, Figure S4 Figure S8: The 13C-1H HSQC spectrum of the polymer electrolyte NPE1, Figure S9: Diffusion decays of NPE1-5 on (a) 7Li and (b) 19F; 1H nuclei for (c) EMIBF4 and (d) EC, Figure S10: Nyquist plots of the SS/NPE/SS cell at room temperature (a) and equivalent scheme (b), where R1 is the resistance of the electrolyte, and R2 is the resistance at the SiO2/electrolyte interface; CPE1 is the capacity of electrical double layer, Table S2: Conductivity of the nanocomposite polymer gel electrolytes, Figure S11: (a) Nyquist plots of the Li/NPE4/Li cells after assembly; (b) the equivalent scheme, where R1 is the electrolyte resistance, and R2 is the resistance of the Li/NPE4 interface; CPE1 is the capacity of electrical double layer; and (c) the visualization of the equivalent circuit parameter calculation performed with the program ZView2, Figure S12: Dependence of the Coulomb efficiency on the cycle number for the Li/NPE3/LiOTAP cells at the C/2 current rate in a voltage range of 0.7-3.5 V, Figure S13: FTIR spectra of the polymer electrolytes NPE3 and NPE3 before and after cycling cells versus LiOTAP, Figure S14: (a) The charge-discharge profiles and (b) dependence of the discharge capacity on the cycle number for the Li/EMIBF4/LiOTAP cells at the C/10 current rate in a voltage range of 0.7-3.5 V, Figure S15: The models of di-radical from imidazolium cation (a) and the adsorption of EMI•-radical on the surface of SiO2 nanoparticle (b), Figure S16: The theoretical IR spectra of dimer EMIBF4 and di-radical EMI (a); the experimental spectra of NPE3 sample before and after cell cycling (b), Table S3: Attachment energy of various ionic complexes to the surface of a SiO 2 nanoparticle. References [45,46] are cited in there.