Protonation of Borylated Carboxonium Derivative [2,6-B10H8O2CCH3]−: Theoretical and Experimental Investigation

The process of protonation of [2,6-B10H8O2CCH3]− was investigated both theoretically and experimentally. The most suitable conditions for protonation of the derivative [2,6-B10H8O2CCH3]− were found. The process of protonation was carried out in the presence of an excess of trifluoromethanesulfonic acid CF3SO3H at room temperature in dichloromethane solution. The structure of the resulting complex [2,6-B10H8O2CCH3*Hfac]0 was established using NMR data and the results of DFT calculations. An additional proton atom Hfac was found to be localized on one of the facets that was opposite the boron atom in a substituted position, and which bonded mainly with one apical boron atom. The main descriptors of the B-Hfac bond were established theoretically using QTAIM and NBO approaches. In addition, the mechanism of [2,6-B10H8O2CCH3]− protonation was investigated.


Introduction
The investigation of covalent and noncovalent interactions is one of the main tasks of modern inorganic chemistry [1,2]. Such studies provide an opportunity to better understand the structures of chemical substances and their properties. By examining chemical bonds, it is possible to establish factors that influence their breaking and formation [3,4]. This fact can be used to produce new substances with given properties. A combined theoretical and experimental approach is the best way to investigate such phenomena [5,6].
There are several driving forces behind bonding formation: orbital, electrostatic, and the van der Waals interactions [7][8][9]. Information about the energy characteristics of such interactions is one of the most important descriptors. Theoretical methods allow exploration of the nature of chemical interactions and estimation of their energies in a simple and intuitive way. The application of such popular and well-established methods as QTAIM (Quantum Theory of Atoms in Molecules), ELF (Electron Localization Function), and NBO (Natural Bond Orbitals) analysis provides the opportunity to find essential information about the chemical structure and bonding of target compounds [10][11][12][13][14].

Experimental Protonation
First, the protonation process of [2,6-B 10 H 8 O 2 CCH 3 ] − was conducted experimentally. As described previously [63], systems of the general form [B 10 H 9 L] − have a lesser possibility of forming a protonated complex than the [B 10 H 10 ] 2− anion. A possible reason for this is that the introduction of a positively charged group reduces the ability of the cluster cage to donate its electron density. Anion [B 10 H 10 ] 2− can be easily transferred into [B 10 H 11 ] − in the presence of trifluoroacetic acid CF 3 COOH. For the preparation of a protonated complex of the general form [B 10 H 9 OR 1 R 2 *H fac ] 0 , however, trifluoromethanesulfonic acid CF 3 SO 3 H was used. In the present work, trifluoroacetic acid CF 3 COOH was also used as a proton donor, but the protonation process did not take place. As in the case of oxonium derivatives [B 10 H 9 OR 1 R 2 ] − , the protonation of the derivative [2,6-B 10 H 8 O 2 CCH 3 ] − was carried out in the presence of CF 3 SO 3 H at room temperature (Scheme 1). Dichloromethane CH 2 Cl 2 was used as a solvent. The application of another solvent was impossible due to the interaction of [2,6-B 10 H 8 O 2 CCH 3 *H fac ] 0 with the molecules of the solvents. [2,6-B 10 H 8 O 2 CCH 3 *H fac ] 0 was formed immediately after the addition of the acid. The distinctive feature of this process is that the protonation of the cluster anion requires an excess of CF 3 SO 3 H acid. For an investigation of these phenomena, the experiment with sequential addition of CF 3 SO 3 H acid to a dichloromethane solution of the initial [2,6-B 10 H 8 O 2 CCH 3 ] − derivative was conducted. The process was monitored by 11 B-NMR spectroscopy ( Figure S1). Detailed information about the NMR spectra of the initial and target substances and the correlation of all signals are given below in the NMR spectra analysis section. When one equivalent of acid was added to the system, the overall structure of the spectrum characteristic of the original derivative was preserved, but one of the signals was broadened. This signal corresponds to boron atoms in apical positions. The given pattern indicates the possibility of the formation of intermolecular contacts between the carboxonium derivative and the trifluoromethanesulfonic acid molecule. When one more acid equivalent is added to the system, a more significant change in the spectrum was observed, which may indicate the formation of molecular complexes between the [2,6-B 10 H 8 O 2 CCH 3 ] − derivative and the acid molecule. The formation of the target substance was also observed, as indicated by the presence of a broadened signal in the region of 20 ppm. However, the yield of the target product did not exceed approximately 10 percent. A more accurate estimation of the target product yield could not be made due to the overlapping signals of the target product and the cluster-acid complex. At ratios [2,6-B 10  presence of trifluoroacetic acid CF3COOH. For the preparation of a protonated complex of the general form [B10H9OR1R2*H fac ] 0 , however, trifluoromethanesulfonic acid CF3SO3H was used. In the present work, trifluoroacetic acid CF3COOH was also used as a proton donor, but the protonation process did not take place. As in the case of oxonium derivatives [B10H9OR1R2] − , the protonation of the derivative [2,6-B10H8O2CCH3] − was carried out in the presence of CF3SO3H at room temperature (Scheme 1). Dichloromethane CH2Cl2 was used as a solvent. The application of another solvent was impossible due to the interaction of [2,6-B10H8O2CCH3*H fac ] 0 with the molecules of the solvents. [2,6-B10H8O2CCH3*H fac ] 0 was formed immediately after the addition of the acid. The distinctive feature of this process is that the protonation of the cluster anion requires an excess of CF3SO3H acid. For an investigation of these phenomena, the experiment with sequential addition of CF3SO3H acid to a dichloromethane solution of the initial [2,6-B10H8O2CCH3] − derivative was conducted. The process was monitored by 11 B-NMR spectroscopy ( Figure  S1). Detailed information about the NMR spectra of the initial and target substances and the correlation of all signals are given below in the NMR spectra analysis section. When one equivalent of acid was added to the system, the overall structure of the spectrum characteristic of the original derivative was preserved, but one of the signals was broadened. This signal corresponds to boron atoms in apical positions. The given pattern indicates the possibility of the formation of intermolecular contacts between the carboxonium derivative and the trifluoromethanesulfonic acid molecule. When one more acid equivalent is added to the system, a more significant change in the spectrum was observed, which may indicate the formation of molecular complexes between the [2,6-B10H8O2CCH3] − derivative and the acid molecule. The formation of the target substance was also observed, as indicated by the presence of a broadened signal in the region of 20 ppm. However, the yield of the target product did not exceed approximately 10 percent. A more accurate estimation of the target product yield could not be made due to the overlapping signals of the target product and the cluster-acid complex. At ratios [2,6-B10H8O2CCH3] − :CF3SO3H equal to 1:3 and 1:4, an increase in the yield of the target product was observed. In the case of the 1:4 ratio, the yield of the target product reached approximately 40 percent. Finally, in the presence of 5 equivalents of CF3SO3H, complete protonation of [2,6-B10H8O2CCH3] − with formation of [2,6-B10H8O2CCH3*H fac ] 0 occurred. Further addition of the acid excess did not change the form of the spectrum, indicating the completeness of the protonation process. It is noteworthy that [2,6-B10H8O2CCH3*H fac ] 0 is stable only as a solvent at room temperature and without air.

Fukui Function
In the next step, the potential position in the boron cluster for the coordination of an additional proton atom was investigated. The protonation process of [2,6-B10H8O2CCH3] − can be considered to be an electrophilic attack on the boron cluster. The Fukui functions approach is the best tool for investigation of the most likely position in the molecule for an electrophilic or nucleophilic attack [64][65][66]. Previously, it has been shown, with the help of Fukui functions, that electrophilic attack on closo-borate anions is performed

Fukui Function
In the next step, the potential position in the boron cluster for the coordination of an additional proton atom was investigated. The protonation process of [2,6-B 10 H 8 O 2 CCH 3 ] − can be considered to be an electrophilic attack on the boron cluster. The Fukui functions approach is the best tool for investigation of the most likely position in the molecule for an electrophilic or nucleophilic attack [64][65][66]. Previously, it has been shown, with the help of Fukui functions, that electrophilic attack on closo-borate anions is performed predominantly in the apical position [67,68]. In the present investigation, Fukui functions were used with different approaches employed for their calculation (Table S2). As in the case of [B n H n ] 2− systems [68], the Hirshfeld approach is the best for calculating the electrophilic attack positions. In the present case, electrophilic attack on closo-borate anions was performed predominantly in the apical position (the optimized structures of [2,6-B 10 Figure 1). Thus, the data from the Fukui functions analysis indicated that the H fac was localized near apical positions (Table S3).
predominantly in the apical position [67,68]. In the present investigation, Fukui functions were used with different approaches employed for their calculation (Table S2). As in the case of [BnHn] 2-systems [68], the Hirshfeld approach is the best for calculating the electrophilic attack positions. In the present case, electrophilic attack on closo-borate anions was performed predominantly in the apical position (the optimized structures of [2,6-B10H8O2CCH3] − and [2,6-B10H8O2CCH3*H fac ] 0 are shown in Figure 1). Thus, the data from the Fukui functions analysis indicated that the H fac was localized near apical positions (Table S3).

Protonation Mechanism
The mechanism of [2,6-B10H8O2CCH3] − protonation was investigated. Initially, complex [2,6-B10H8O2CCH3 * CF3SO3H] − (Comp) formed endoergonically (by 23 kJ*mol −1 , in terms of Gibbs free energies). The main driving force behind the formation of this complex is the dihydrogen bond between the proton atom from CF3SO3H and the equatorial hydrogen atom from the boron polyhedron. The contact length of Heq-H was equal to 1.60 Å. The presence of the dihydrogen bond was proved with the help of QTAIM analysis of the Comp structure ( Figure 2). In addition, the C-F···H-B contact was detected. According to the main topological descriptors of electron density C-F···H-B, contact can be characterized as being very weak.

Protonation Mechanism
The mechanism of [2,6-B 10 formed endoergonically (by 23 kJ·mol −1 , in terms of Gibbs free energies). The main driving force behind the formation of this complex is the dihydrogen bond between the proton atom from CF 3 SO 3 H and the equatorial hydrogen atom from the boron polyhedron. The contact length of H eq -H was equal to 1.60 Å. The presence of the dihydrogen bond was proved with the help of QTAIM analysis of the Comp structure ( Figure 2). In addition, the C-F· · · H-B contact was detected. According to the main topological descriptors of electron density C-F· · · H-B, contact can be characterized as being very weak.
case of [BnHn] 2systems [68], the Hirshfeld approach is the best for calculating the electrophilic attack positions. In the present case, electrophilic attack on closo-borate anions was performed predominantly in the apical position (the optimized structures of [2,6-B10H8O2CCH3] − and [2,6-B10H8O2CCH3*H fac ] 0 are shown in Figure 1). Thus, the data from the Fukui functions analysis indicated that the H fac was localized near apical positions (Table S3).

Protonation Mechanism
The mechanism of [2,6-B10H8O2CCH3] − protonation was investigated. Initially, complex [2,6-B10H8O2CCH3 * CF3SO3H] − (Comp) formed endoergonically (by 23 kJ*mol −1 , in terms of Gibbs free energies). The main driving force behind the formation of this complex is the dihydrogen bond between the proton atom from CF3SO3H and the equatorial hydrogen atom from the boron polyhedron. The contact length of Heq-H was equal to 1.60 Å. The presence of the dihydrogen bond was proved with the help of QTAIM analysis of the Comp structure ( Figure 2). In addition, the C-F···H-B contact was detected. According to the main topological descriptors of electron density C-F···H-B, contact can be characterized as being very weak. In the next step, the proton migrated to a boron polyhedral through the formation of the transition state (TS). The preferred location for the proton attack was the face of the polyhedron opposite the substituted position. The proton from CF 3 SO 3 H acid in TS connected mainly with the boron atom in the apical position, which was demonstrated with the help of QTAIM analysis ( Figure 3). In the case of TS, the contact length between the apical boron atom and the proton atom from CF 3 SO 3 H was equal to 1.48 Å. The overall energy barrier of proton migration was 69 kJ·mol −1 .
the transition state (TS). The preferred location for the proton attack was the face of the polyhedron opposite the substituted position. The proton from CF3SO3H acid in TS connected mainly with the boron atom in the apical position, which was demonstrated with the help of QTAIM analysis (Figure 3). In the case of TS, the contact length between the apical boron atom and the proton atom from CF3SO3H was equal to 1.48 Å. The overall energy barrier of proton migration was 69 kJ · mol −1 . Finally, [B10H8O2CCH3*H fac ] 0 formed. The overall process of boron cluster protonation is endoergonical (by 18.5 kJ · mol −1 , in terms of Gibbs free energies) ( Figure 4). Thus, due to ΔG > 0 overall, a reaction did not occur in the presence of 1 equivalent of CF3SO3H. This finding correlates well with experimental data and the main reason it occurred is the weak stabilization of CF3SO3 − in the dichloromethane solution. Additional proton donors are required to stabilize this anion. The use of an excess of trifluoromethanesulfonic acid can provide this. The CF3SO3 − can be stabilized by hydrogen bonds between oxygen atoms of this anion and hydrogen atoms of CF3SO3H. For example, complex between CF3SO3 -anion and CF3SO3H was observed ( Figure S8). This complex possesses strong hydrogen bonds and the formation of a given system is an exergonic process. It is worth noting that the addition of the acid excess also significantly changed the properties of the medium, which in turn can change the solvation energy of the model system under consideration. This aspect may also contribute to the fact that the protonation process becomes exergonic when an excess of the acid is supplied. Finally, [B 10 H 8 O 2 CCH 3 *H fac ] 0 formed. The overall process of boron cluster protonation is endoergonical (by 18.5 kJ·mol −1 , in terms of Gibbs free energies) ( Figure 4). Thus, due to ∆G > 0 overall, a reaction did not occur in the presence of 1 equivalent of CF 3 SO 3 H. This finding correlates well with experimental data and the main reason it occurred is the weak stabilization of CF 3 SO 3 − in the dichloromethane solution. Additional proton donors are required to stabilize this anion. The use of an excess of trifluoromethanesulfonic acid can provide this. The CF 3 SO 3 − can be stabilized by hydrogen bonds between oxygen atoms of this anion and hydrogen atoms of CF 3 SO 3 H. For example, complex between CF 3 SO 3 − anion and CF 3 SO 3 H was observed ( Figure S8). This complex possesses strong hydrogen bonds and the formation of a given system is an exergonic process. It is worth noting that the addition of the acid excess also significantly changed the properties of the medium, which in turn can change the solvation energy of the model system under consideration. This aspect may also contribute to the fact that the protonation process becomes exergonic when an excess of the acid is supplied.   11 B NMR spectra were used for the confirmation suggestion based on the Fukui index results and to establish the structure of the resulting complex. It is useful to consider the spectra of the [2,6-B10H8O2CCH3] − anion. Due to the high symmetry of the given system, the two boron atoms in the substituted positions and the two apical atoms are pairwise The deciphering of other abbreviations is given in the main text. 11 B NMR spectra were used for the confirmation suggestion based on the Fukui index results and to establish the structure of the resulting complex. It is useful to consider the spectra of the [2,6-B 10 H 8 O 2 CCH 3 ] − anion. Due to the high symmetry of the given system, the two boron atoms in the substituted positions and the two apical atoms are pairwise equivalent. Atoms in equatorial positions are split into two groups of signals ( Figure 5). Thus, as described previously, the 11 B NMR spectra of the initial anion were characterized by four signals, as follows: a signal at 0.0 ppm with integral intensities I = 2, corresponding to the boron atoms from equivalent substituted positions. This signal did not split in the absence of broadband decoupling; a signal at −7.1 with integral intensities I = 2, corresponding to boron atoms from equivalent apical positions; and signals from equatorial positions appeared at −17.6 (I = 2) and −30.0 (I = 4). This interpretation of the spectrum is based on previously obtained spectral data for several closo-decaborate derivatives with B-O exo-polyhedral bonds. This approach does not, however, enable a complete correlation of signals and leaves some room for speculation. The theoretical calculation of chemical shielding enabled the full correlation of the signals in the 11 B NMR spectra with the boron positions in a cluster cage. Several computational levels for calculating NMR shifts were tested: wB97X-D3, B3LYP, and B97 DFTfunctionals with 6-31++G(d,p), IGLO-III, and EPR-III basis sets (Table S5). To evaluate the difference between the theoretical and experimental data, the values of the root-meansquare deviation (RMSD) were used. The wB97X-D3 level of theory provided the worst results, with all basis sets. The values of RMSD were quite large (the range of RMSD values lay in the range [46][47] and applying this method did not enable the deciphering of the 11 B spectra of given compounds. Improved results were obtained with the application of the hybrid functionals B3LYP and B97. In the case of the B97 functional, the best values of RMSD were obtained (Table S5). The RMSD value for the B97/IGLO-III level of theory was equal to 3.7. Thus, applying the results of the theoretical calculation, the assumption about the chemical shifts of atoms in the substituted and apical positions proved to be correct. In addition, the signal at −17.6 ppm corresponded to boron atoms from the equatorial belt in the B3 and B9 positions. The signal from −30.0 ppm corresponded to atoms in the B4, B5, B7, and B8 positions.

NMR Spectra Analysis
The 1 H NMR spectra of [2,6-B10H8O2CCH3] -were considered. For a simple view of the spectrum and the correct correlation of signals, it is best to use 1 H- 11 B NMR spectra ( Figure S2). Some of the signals appeared, however, as broadened lines, making it difficult to identify them. Signals from the methyl substituent appeared at 2.18 ppm, and signals from apical positions at 3.16 ppm were well represented. Signals from equatorial positions appeared at 1.79, 0.39 and 0.12 ppm. The theoretical calculation of chemical shielding enabled the full correlation of the signals in the 11 B NMR spectra with the boron positions in a cluster cage. Several computational levels for calculating NMR shifts were tested: wB97X-D3, B3LYP, and B97 DFT-functionals with 6-31++G(d,p), IGLO-III, and EPR-III basis sets (Table S5). To evaluate the difference between the theoretical and experimental data, the values of the root-meansquare deviation (RMSD) were used. The wB97X-D3 level of theory provided the worst results, with all basis sets. The values of RMSD were quite large (the range of RMSD values lay in the range [46][47] and applying this method did not enable the deciphering of the 11 B spectra of given compounds. Improved results were obtained with the application of the hybrid functionals B3LYP and B97. In the case of the B97 functional, the best values of RMSD were obtained (Table S5). The RMSD value for the B97/IGLO-III level of theory was equal to 3.7. Thus, applying the results of the theoretical calculation, the assumption about the chemical shifts of atoms in the substituted and apical positions proved to be correct. In addition, the signal at −17.6 ppm corresponded to boron atoms from the equatorial belt in the B3 and B9 positions. The signal from −30.0 ppm corresponded to atoms in the B4, B5, B7, and B8 positions.
The 1 H NMR spectra of [2,6-B 10 H 8 O 2 CCH 3 ] − were considered. For a simple view of the spectrum and the correct correlation of signals, it is best to use 1 H- 11 B NMR spectra ( Figure S2). Some of the signals appeared, however, as broadened lines, making it difficult to identify them. Signals from the methyl substituent appeared at 2.18 ppm, and signals from apical positions at 3.16 ppm were well represented. Signals from equatorial positions appeared at 1.79, 0.39 and 0.12 ppm.
The addition of a proton reduced the symmetry of the resulting system and the spectra of [2,6-B 10 H 8 O2CCH 3 *H fac ] 0 were more complicated than those of [2,6-B 10 H 8 O 2 CCH 3 ] − . Boron atoms were split into separate signals. Thus, it was impossible to elucidate the structure of [2,6-B 10 H 8 O2CCH 3 *H fac ] 0 by only applying the 1D 11 B spectra. To decipher these spectra, the 11 B-11 B COSY spectra and theoretical calculations were used ( Figure S5). This type of spectrum indicates that, contrary to the case of the [B 10 H 11 ] − anion, the proton in this anion does not migrate along one of the equatorial belts but is localized on one of the edges. Based on the calculated data, the most stable isomer Iso1 was found. In this isomer, the proton was localized on the edge opposite the substituted position. The structure of this isomer is described in more detail below. Similar to the initial anion [2,6-B 10 H 8 O 2 CCH 3 ] − , the B97 functional worked well for the prediction of NMR spectra. The RMSD value for the B97/IGLO-III level of theory was equal to 3.5 (Table S5).
As in the case of [B 10  In the 1 H NMR spectra of [2,6-B 10 H 8 O 2 CCH 3 *H fac ] 0 , the signal of the methyl group was observed at 2.45 ppm ( Figure S3). Signals from the exo-polyhedral cluster were in the range 5.16-0.33 ppm. It is noteworthy that the signal at 5.16 ppm corresponded to a hydrogen atom connected with a boron atom in the B10 position. This assumption is based on the result of 1 H NMR spectra modeling. According to the results of the calculations, the proton in the downfield region belonged to the given atom. The signal from H fac , according to theoretical calculations, was in the downfield region at 0.00 ppm. In the experimental spectra, the most upfield signal was at 0.33 ppm. In the 13 C NMR spectra of [2,6-B 10 H 8 O 2 CCH 3 *H fac ] 0 , the signal of the carbonyl group carbon atom was observed at 194.0 ppm and the signal of the methyl group carbon atom was observed at 20.10 ppm ( Figure S4). In the IR-spectra, the B-H bands were shifted from 2490 cm −1 in the initial [2,6-B 10 Figure S6).

Structure Elucidation
The structure of [2,6-B 10 H 8 O 2 CCH 3 *H fac ] 0 was examined using DFT calculation. Several possible isomers were calculated (Figures 6 and S7, Table S4). Calculations were performed in the gas phase and considering the solvation effect. In both cases, the isomers differed significantly in terms of energy. The Iso1 and Iso2 isomers were the most thermodynamically stable ( Figure 6). In these isomers, an extra proton, in addition to the equatorial boron atoms, was localized on the apical boron atom. The Iso1 in which the proton was localized on the edge opposite the substituted position had the highest negative Gibbs energy among all isomers. The energetic barrier for the migration of the proton atom from Iso1 to Iso2 was equal to 30 kJ·mol −1 in the gas phase and 25 kJ·mol −1 in dichloromethane solution. Isomers with the localized proton only on equatorial boron atoms were the least thermodynamically stable and had the smallest negative Gibbs energy values. The energy barrier between Iso1 and Iso3 was 50 kJ·mol −1 for both the gas and dichloromethane phase. This energy barrier was less than that of the anion [B 10  The geometric parameters of this isomer were discussed in detail (Table S1). The bond length between H fac and the apical boron atom was 1.28 Å, whereas the analogous parameter for the bond between H fac and the equatorial boron atom was 1.42 Å. The bond length between equatorial boron atoms and the apical boron atom that bonded with the proton increased compared to the same parameter in the original cluster. This parameter increased from 1.72 to 1.84 Å. The phenomena of B-H fac contacts were investigated using Wiberg bond order indices and QTAIM analysis (Figure 7). Applying the QTAIM method, a molecular graph of the electron density distribution in [2,6-B10H8O2CCH3*H fac ] 0 was obtained. The molecular graph indicated that H fac has interactions with apical and equatorial boron atoms. The Bap-H fac interaction was characterized by a greater value of the Wiberg bond order index compared to that of Beq-H fac . The Bap-H fac interaction was also characterized by greater values of ρ(r), total energy at the bond critical point, and the delocali-  The geometric parameters of this isomer were discussed in detail (Table S1). The bond length between H fac and the apical boron atom was 1.28 Å, whereas the analogous parameter for the bond between H fac and the equatorial boron atom was 1.42 Å. The bond length between equatorial boron atoms and the apical boron atom that bonded with the proton increased compared to the same parameter in the original cluster. This parameter increased from 1.72 to 1.84 Å. The phenomena of B-H fac contacts were investigated using Wiberg bond order indices and QTAIM analysis (Figure 7). Applying the QTAIM method, a molecular graph of the electron density distribution in [2,6-B 10 H 8 O 2 CCH 3 *H fac ] 0 was obtained. The molecular graph indicated that H fac has interactions with apical and equatorial boron atoms. The B ap -H fac interaction was characterized by a greater value of the Wiberg bond order index compared to that of B eq -H fac . The B ap -H fac interaction was also characterized by greater values of ρ(r), total energy at the bond critical point, and the delocalization index. These data indicate that H fac is predominantly bonded with the boron atom in the apical position. An analogous investigation was carried out for the case of the [B 10

Computational Details
The full geometry optimization of all model structures was carried out at the ωB97X-D3/6-31++G(d,p) level of theory [69,70] with the help of the ORCA 4.2.1 program package (Mülheim an der Ruhr, Germany) [71] (the atom-pairwise dispersion correction with the zero-damping scheme was utilized [70]). The convergence tolerances for the geometry optimization procedure were as follows: energy change = 5.0 × 10 −6 Eh, maximal gradient = 3.0 × 10 −4 Eh/Bohr, RMS gradient = 1.0 × 10 −4 Eh/Bohr, maximal displacement = 4.0 × 10 −3 Bohr, and RMS displacement = 2.0 × 10 −3 Bohr. Spin-restricted approximation for the model structures with closed electron shells was used. Symmetry operations were not applied during the geometry optimization procedure for any of the model structures. The Hessian matrices were calculated numerically for all optimized model structures in order to prove the location of correct minima on the potential energy surfaces (no imaginary frequencies for all reactants, intermediates, and final products; only one imaginary frequency for transition states). The connectivity of each reaction step was also confirmed using the intrinsic reaction coordinate (IRC) calculation from the transition states [72][73][74]. Solvent effects were considered using the Solvation Model based on Density (SMD) [75]. The natural bond orbital (NBO) method was emplyed, using the NBO7 program package (Madison, WI, USA) [76]. Topological analysis of the electron density distribution, using the Quantum Theory of Atoms in Molecules (QTAIM) formalism developed by Bader [12], was employed with the Multiwfn program,version 3.7 (Beijing, China) [77]. The Cartesian atomic coordinates for all optimized equilibrium model structures are presented in the Supplementary Materials. Visualization of the optimized structures was carried out with the help of ChemCraft program version 1.7 (Ivanovo, Russia) [78]. In the case of the molecular graph showing the results of the topological analysis of the electron density distribution visualization, the Multiwfn program (version 3.7) was employed [77]. A possible reason for these phenomena may be the reduction in the total charge in the system compared to that of the initial anion [2,6-B 10 H 8 O 2 CCH 3 ] − . The atomic charges on the boron atoms are reduced and the electron repulsion in the exo-polyhedral bonds is also reduced.

Computational Details
The full geometry optimization of all model structures was carried out at the ωB97X-D3/6-31++G(d,p) level of theory [69,70] with the help of the ORCA 4.2.1 program package (Mülheim an der Ruhr, Germany) [71] (the atom-pairwise dispersion correction with the zero-damping scheme was utilized [70]). The convergence tolerances for the geometry optimization procedure were as follows: energy change = 5.0 × 10 −6 Eh, maximal gradient = 3.0 × 10 −4 Eh/Bohr, RMS gradient = 1.0 × 10 −4 Eh/Bohr, maximal displacement = 4.0 × 10 −3 Bohr, and RMS displacement = 2.0 × 10 −3 Bohr. Spin-restricted approximation for the model structures with closed electron shells was used. Symmetry operations were not applied during the geometry optimization procedure for any of the model structures. The Hessian matrices were calculated numerically for all optimized model structures in order to prove the location of correct minima on the potential energy surfaces (no imaginary frequencies for all reactants, intermediates, and final products; only one imaginary frequency for transition states). The connectivity of each reaction step was also confirmed using the intrinsic reaction coordinate (IRC) calculation from the transition states [72][73][74]. Solvent effects were considered using the Solvation Model based on Density (SMD) [75]. The natural bond orbital (NBO) method was emplyed, using the NBO7 program package (Madison, WI, USA) [76]. Topological analysis of the electron density distribution, using the Quantum Theory of Atoms in Molecules (QTAIM) formalism developed by Bader [12], was employed with the Multiwfn program, version 3.7 (Beijing, China) [77]. The Cartesian atomic coordinates for all optimized equilibrium model structures are presented in the Supplementary Materials. Visualization of the optimized structures was carried out with the help of ChemCraft program version 1.7 (Ivanovo, Russia) [78]. In the case of the molecular graph showing the results of the topological analysis of the electron density distribution visualization, the Multiwfn program (version 3.7) was employed [77].

IR Spectra
The IR spectra of the prepared compounds were recorded on an Infralyum FT 02 Fourier transform spectrometer (Lumex Instruments Research and Production Company, Fraserview Place, Vancouver, BC, Canada) in the region of 300-4000 cm -1 and with a resolution of 1 cm -1 . Samples were prepared as dichloromethane CH 2 Cl 2 solution.

NMR Spectra
The NMR ( 1 H, 11 B, 13 C) spectra of the solutions of the studied compounds in CD3CN were recorded on a Bruker Avance II 300 spectrometer (Ettlingen, Germany) operating at 300.3, 96.32, and 75.49 MHz, respectively, using an internal deuterium lock. Tetramethylsilane and boron trifluoride etherate were used as external references.

Conclusions
The process of protonation of the carboxonium derivative [2,6-B 10 H 8 O 2 CCH 3 ] − was investigated. In the case of the carboxonium derivatives of closo-decaborate anions, trifluoromethanesulfonic acid CF 3 SO 3 H was used as a proton donor. By considering the reaction mechanisms, the reason for the excess of CF 3 SO 3 H was established. The excess acid was required to stabilize the anion CF 3 SO 3 − , which in turn caused the total protonation process to become exergonic. In contrast to the anion [B 10 H 11 ] − , in the case of [2,6-B 10 H 8 O 2 CCH 3 *H fac ] 0 , the proton did not migrate along the equatorial belt and was localized on a facet opposite the B atoms bonded with the exo-polyhedral substituent. Based on theoretical modeling data, it was shown that the proton was predominantly bound to the apical boron atom. In addition, the 11 B-1 H NMR spectra of the final compound were quite complicated and difficult to decipher. Using theoretical modeling data, all signals were correlated with the positions of the boron atoms in the cluster. Theoretical calculations at the B97/IGLO-III level of theory corresponded perfectly with the experimental 11