3.1. Molecular Analysis
According to the simulation and optimization of the analyzed graphene structure carried out in the SCIGRESS v.FJ 2.7 program, the C-C bond length is 1.42 Å, while the C-C-C torsion angle in the undefected area is 120°. For such an energetically optimized structure, the system energy simulations were carried out depending on the place of attachment of the -BF
2 group or the formation of -BH
2 groups. Designations of the analyzed graphene systems depending on the place of attachment of -BF
2 groups or formation of -BH
2 groups are summarized in
Table 2.
First, the analysis covered the interaction between the -BF3- ion and a model graphene flake composed of 149 carbon atoms and 46 hydrogen atoms with a structural defect in the form of a vacancy of one C atom. The interaction energy of the -BF3- ion as a function of the distance from the surface of the unmodified, defective graphene flake (Gdef.) decreases from 406 to 230 kcal/mol.
Then, the energy values of the most probable systems for which the addition of BF
2 group or the formation of -BH
2 groups may take place were analyzed. Based on the analysis of the energy values of graphene systems depending on the mentioned parameters, it was determined that the most thermodynamic stable structures are the ones in which, in the reaction of BF
3·THF with graphene, -BF
2 groups are formed on its surface (
Table 3,
Figure 1,
Figure 2). On the other hand, energetically, there are no privileged reactions leading to the formation of in-atomic systems in relation to which the carbon atoms of the graphene structure are substituted with boron (the energy of the system is 413 kcal/mol). The energy of the analyzed defective graphene (vacancy of 1 C atom), which is the reference for the calculations, is 465 kcal/mol. Attachment of the -BF
2 groups both to the edge of the flake (system energy equals 283 kcal/mol) and to carbon with hybridization of sp
3 or sp
2 in the area of graphene defect (system energy equals 296 kcal/mol), and also adatomically (to carbon from graphene structure with sp
2 hybridization, with system energy equalling 297 kcal/mol) is practically on the same level.
In the areas of graphene where the reaction with the BF
3- ion produces the -BF
2 group, there is a growth of negative zones. Additionally, there is a clearly marked tendency to occupy the central zones of the graphene flake by a negative charge. A similar relationship was found by X. Duan, K. O’Donnell [
30], who analyzed graphene systems doped with sulfur and/or nitrogen.
Then, for the most thermodynamically stable system (G def.-Bad-atom-F-described in
Table 2), taking into account the electrostatic relations, the interaction energy with the lithium ion as a function of its distance from the graphene structure was examined (
Figure 3). It was assumed that the lithium ion would only interact with the structure of functionalized graphene. In this case, the transport of the shielded lithium through the electrolyte and its interaction with the SEI layer of the lithium-ion battery was omitted.
For such assumptions, it was found that the change in the electrostatic potential of the functionalized graphene flake (Cgraphene-BF2 chemical bond) causes the physical interaction of graphene with the lithium cation. The consequence of this phenomenon is the reduction in the system energy as a result of the electrostatic attraction of Li+ by the areas of electrostatic interaction of fluorine. This phenomenon explains why it is possible to achieve a relatively thermodynamically stable physical lithium bond in modified graphene systems—as opposed to unmodified graphene systems. Although the energy of the system decreases as Li+ approaches the graphene surface, the minimum value of the energy of the system is close to 670 kcal/mol compared to Gdef.-Bad-atom-F, whose energy is about 503 kcal/mol.
3.2. FTIR
In order to identify functional groups present in the research material after the functionalization process, FTIR analysis was performed.
Figure 4 shows the FTIR spectra of rGO without modification as well as after functionalization with BF
3·THF. In the spectrum of pure, unmodified rGO there is a broad band with a maximum at the wavenumber of 3440 cm
−1, which comes from the stretching vibrations of the -OH groups. This bandwidth for the modified rGO samples is slightly lower. The deformation vibrations of the -OH groups are also visible at the wave number of 1450 cm
−1. Another clear peak in the unmodified rGO spectrum is the peak at 1720 cm
−1, which comes from the stretching vibrations of the C=O bond, which may indicate that, despite the GO reduction process, some carbon–oxygen bonds still remained in its structure [
31].
The peak characteristic for the graphene structure is located at 1580 cm
−1, which comes from the stretching vibrations of C=C bonds [
32]. Increasing the intensity of this absorption maximum for rGO after BF
3·THF functionalization may result from the overlapping of π–π bonds in the so-called stacks. Then in the spectrum of unmodified rGO there is a wide, flat band in the range of 1400–950 cm
−1. In this case, the components are mainly -OH and C=C bonds, characteristic for a material of this type. In turn, for this range, the most significant differences in the course of the spectra were observed for samples after BF
3·THF functionalization. Additional, distinct peaks appeared at 1310–1280, 1126, 1085, 1055, 1035 and 1018 cm
−1.
The band in the 1310–1280 cm
−1 range comes from the stretching vibrations of the bonds of the CF-CH
3 terminal group [
33] and the stretching vibrations of the asymmetric C-F bonds in the CF
2 group [
34]. The wide band with a maximum of 1120 cm
−1 comes from symmetrical stretching vibrations of the C-F bonds also belonging to the CF
2 group [
34,
35]. Another confirmation of the presence of C-F bonds can be found in the peaks located at 1085 and 1035 cm
−1, which come from the fragment of the structure in which the carbon atom is connected to only one fluorine atom [
36]. It should be remembered that with the wavenumber value of 1085 cm
−1, the B-OH stretching vibrations may also have their maximum absorption [
37]. In the spectra of some rGO samples there is a weak peak at 1055 cm
−1, which comes from the ether bond fragment belonging to the not fully dissociated tetrahydrofuran [
36].
Table 4 shows the areas of the above-discussed peaks. The obtained data show that the highest intensity of the peaks coming from vibrations of CF bonds adjacent to the methyl group have the samples modified with BF
3·THF with a concentration of 1.5% for the 24- and 36-h bath times, and with a higher concentration of 3%, the stronger effect was obtained after 16 h. In this range of wavenumbers there may also be vibrations of B-F bonds constituting “contamination” after the used modifier [
38]. In the case of the total amount of C-F bonds contained in CF
2 and CF groups, for a modifier concentration of 1.5%, the times in the range of 16–32 h are promising. A further increase in the functionalization time does not cause a further increase in the number of groups containing fluorine atoms in their structure. A slightly different relationship was obtained for the samples functionalized with a concentration of 3%. In this case, extending the duration of this process led to a gradual increase in the amount of attached fluorine. The content of impurities in the form of C-O-C groups changes randomly. In this case, no dependence on the concentration of BF
3·THF or the time of modification was observed. With the wavelength value of 1018 cm
−1, another peak appeared, which confirms that the graphene structure was not only doped with fluorine atoms, but also an addition of boron atoms. The peak next to this wavenumber value clearly proves that there are also C-B bonds [
39]. Its amount in each sample undergoes similar changes as the content of fluorine. Both spectra also distinguish a broad band with a maximum of 720 cm
−1, which additionally confirms the presence of the -CF
2 group. In this case, the deformation vibrations of the C-F bond are present.
3.3. Raman Spectroscopy
Further characterization of the structure of the obtained materials was performed using Raman spectroscopy. On the spectrum obtained for the starting material, which was rGO, peaks characteristic of graphene structures were observed: peaks D (1342 cm
−1), G (1580 cm
−1) and 2D (2698 cm
−1) (
Figure 5).
The D-band corresponds to the disorder of the structure due to the disturbance of the symmetry of the graphene lattice, the presence of sp
3 hybridization-based defects, vacancies, grain boundaries and even edges. It comes from the secondary Raman scattering process involving one iTO phonon and one defect [
40]. In the case of the parent graphite, it has a relatively low intensity, which proves the highly crystalline structure of the tested sample [
41].
On the other hand, the shape and intensity of the G-band result from the vibration of carbon atoms with sp
2 hybridization in a two-dimensional hexagonal lattice. The G applies to all carbon structures with sp
2 hybridization, including amorphous carbon, carbon nanotubes, graphite, etc. [
42]. It is the result of photon scattering on the optical phonon and comes from the main, primary Raman scattering [
43].
The 2D-band comes from the secondary Raman scattering process involving two iTO phonons near the K point [
40]. On the Raman spectra of the functionalized structures, the delamination of the G peak (D’) and the appearance of the D + G peak can also be observed.
The process of rGO functionalization with the use of BF
3·THF significantly influenced the changes in individual bands. The G peak was observed at a wavelength of 1566 cm
−1; there was a shift towards lower wavenumbers relative to the peak position for the unmodified sample. We observe an increase in its frequency with a reduction in the FWHM (full width at half maximum). It comes from the fact that the doping of the graphene structure causes the Fermi energy to move away from the Dirac point, and the plasma–phonon coupling effect weakens, which is manifested by an increase in the G phonon energy [
40].
The number of defects in the graphene structure, increasing with the progress of the functionalization process, was manifested by an increase in the intensity of the D, D’ and D+G peaks, observed at 1339, 1596 and 2461 cm−1.
The D ’peak becomes visible on the Raman spectrum if the graphene material contains randomly distributed impurities or surface charges. This is due to the fact that the localized vibration modes of contaminants can interact with the extended phonon modes of graphene, which results in the observed division [
44].
The presence of the D peak is related to the disorder and defect of the graphene layer. In the starting material, it is the result of the presence of oxygen functional groups. The functionalization process leads to an increase in the intensity of the D band, which is related to the reactions taking place and the formation of sp3 hybridization bonds.
The D and D ’bands are created as a result of the photon scattering on iTO phonons from the vicinity of the K point of the Brillouin zone and the iLO from the vicinity of the point Γ Brillouin zone. In order to fulfill the principle of conservation of momentum, the proportion of defects that take over this excess momentum without changing the energy is necessary. Thus, the analysis of the intensity of the D and D’ bands allows the determination of the level of damage to the carbon layer [
45].
All types of sp
2 hybridization carbon materials show a strong peak in the range 2500–2800 cm
−1 in the Raman spectrum. In combination with the G-band, this spectrum is the Raman signature of graphite materials and is called 2D [
46]. The 2D band is a second-order two-phonon process and shows a strong dependence of frequency on the energy of the excitation laser. It is also used to identify a single layer of graphene by examining the half-width and the ratio of the intensity of 2D and G bands. For a single layer of high-quality (defect-free) graphene, it should equal 2 and the half-width should be close to ~30 cm
−1. In the case of a greater number of layers, the 2D band becomes wider, while the ratio of the intensity of the 2D band to G is lower than one [
47].
The I2D/IG ratio for the starting material is 0.62 and for modified samples it ranges from 0.56 to 0.69. The half-width for the rGO before the modification process is 203.6 cm−1, and for successively functionalized materials 127.93 cm−1 (A1), 135.84 cm−1 (A2), 114.73 cm−1 (A3), 123.89 cm−1 (A4), 146.59 cm−1 (A5), which confirms the fact that the tested material consists of more layers.
Based on the presented data in
Table 5, we can observe that the 2D shifts towards lower wavenumbers, which indicates the presence of tensile stresses [
44].
Figure 5 shows the Raman spectra for the rGO material before modification and the obtained end products (A1–A5).
Table 5 shows the position of the individual bands and the intensity ratios of the D and G peaks (I
D/I
G).
By analyzing the ratio of the IDA/IGA peak intensities read from the Raman spectrum, the level of the defect in the graphene structure can be characterized. The ID/IG intensity ratio for rGO is 0.89, which may indicate that, despite the GO reduction process, some carbon–oxygen bonds still remained in its structure.
For the samples subjected to functionalization, the ID/IG ratio increases with the extension of the functionalization time, which may indicate a progressive disturbance of the graphene lattice and an increase in the volume of defects with sp3 hybridization.
3.6. Determination of Electrochemical Properties by Cyclic Voltammetry
The ability of electric charge storage was determined on the basis of cyclic voltammetry. The material tested showed a capacity between 450 and 550 mAh * g−1 (working up to 2.5 V). In practice, the anode materials for lithium-ion cells do not work in the range greater than 1.5 V. In this case, the capacity in this range is less than 300 mAh * g−1, which is less than the value assigned to commercial graphite anodes (max. 372 mAh * g−1).
Figure 8 presents the graphs for the first and fifth cycles of cyclic voltammetry for the tested materials, and comparatively for unmodified material (marked 12MN). As can be seen, the nature of the processes (as indicated by the shape and course of the curves) is very similar for all three samples. Comparing the tested samples with the starting material, it can be seen that the structure of the material has changed during the functionalization. The obtained curves indicate a material with a disordered structure, as opposed to graphite or base material. The synthesized materials show capacitive properties resulting from the expansion of the specific surface area.
The peaks visible in the graphs (0.005–0.3 V vs. Li/Li+) indicate reactions of attached functional groups or structures formed with lithium. These peaks repeat in each cycle, which indicates that the reaction is reversible.
We also observe a large loss of capacity during the first cycle of operation. This is due to the reaction of some attached functional groups (e.g., fluorine) with lithium during the insertion process (reactions taking place in primary cells) and due to the formation of a solid electrolyte interphase (so called SEI) layer (peaks at c.a. 0.55 V vs. Li/Li
+). The decrease in the capacity associated with the SEI formation is mainly caused by the expansion of the specific surface area of the material after the functionalization process [
48,
49].