1. Introduction
Halogenated borane derivatives are successfully used as Lewis acids [1,2], as Brønsted superacids [3,4,5], as weakly coordinating anions [6,7], and as components of ionic liquids [8] and solid-state electrolytes [9,10]. Radio-halogenated boron clusters are used in tumor imaging and therapy [11].
Di- and tri-substituted halogenated derivatives of BH4− (BH2X2− and BHX3−) can be obtained through MBH4–MBX4 (X = F, Cl) exchange [12,13,14]. However, this reaction always leads to a mixture of products (due to disproportionation and low selectivity of the process), as well as to the formation of by-products as a result of the interaction of substitution products with the initial reagents [12]. Another possible approach to the synthesis of halogenated derivatives is direct halogenation; however, the reported halogenation of BH4− by the reaction with HF has resulted in the formation of a mixture of substitution products [14]. However, as exemplified in [15], the use of 1H NMR monitoring during the bubbling of tBuNH2·BH3 with anhydrous hydrogen halides HX (X = F, Cl, Br) allows for the reaction to be stopped when the desired halogenation product is formed.
The halogenated derivatives of [B6H6]2− are typically obtained by the reaction in highly alkaline aqueous solutions in the presence of free halogens X2 (X = Cl, Br, I) [16,17]. However, this approach always results in mixtures of several components. Moreover, due to the fast rate of halogenation, the mono- and dihalogen derivatives are available as minor products, while tri- and penta-substituted derivatives are major products. Therefore, the use of preparative scale ion exchange column chromatography is essential for the isolation of pure compounds. However, for geometrical isomers with the same degree of halogenation, the difference in phase mobility is small (5–15%); thus, their separation could be achieved only by repeated ion exchange column chromatography [18].
Due to these factors, there are difficulties in the synthesis of halogen-derived borohydrides. These reactions are not selective; in a general case, a mixture of various substitution products is obtained, and to isolate the desired substitution product, additional separation and purification are necessary. To overcome these difficulties and develop an approach for the targeted synthesis of these compounds, we undertook DFT calculations for the halogenation of tetrahydroborate-anion (BH4−, which is the smallest available boron hydride model) and hexahydro-closo-hexaborate dianion ([B6H6]2−, which is the smallest stable closo-borane anion) with hydrogen halides.
Previously, a DFT investigation of halogenation was performed, albeit only for [B6H6]2− on the example of interaction with HOX (X = Cl, Br) and only for the first two steps [17]. In the case of BH4−, the DFT study of the stepwise hydrolysis [19] and alcoholysis [20] were reported. Its interaction with strong acids was investigated only for the first reaction step — and only for CF3OH [21]. From one side, hydrogen halides (such as HCl and HBr) are strong inorganic acids which violently react with BH4− and cause halogenation of closo-boranes via the electrophile-induced nucleophilic substitution (EINS) mechanism [22,23]. From the other side, the halogen is the electron-withdrawing group (EWG), which should increase electron deficiency at the boron atom, altering the reactivity of the B–H bond in halogen-substituted derivatives. In this regard, it was interesting to investigate the direct halogenation of BH4− and [B6H6]2− by hydrogen halides.
In the course of our work, we began with the DFT study of the mechanism of direct successive halogenation using the example of the reaction of BH4− with HX (X = F, Cl, Br) in acetonitrile (MeCN). Then, using the knowledge gained, we studied halogenation of [B6H6]2− with HCl in MeCN, which proceeded as an electrophile-induced nucleophilic substitution (EINS) process. Thus, in this work, a principal mechanism of direct halogenation was established for BH4− and [B6H6]2−. The changes in the reactivity of the B–H bond entailed by halogens’ introduction were evaluated via calculations of their hydride donor ability (HDA) and proton donor ability (PDA). Hydride donor ability (HDA)—i.e., thermodynamic hydridicity, defined as the Gibbs free energy (∆G°[H−]) value for the reaction of hydride ion (H−) detachment—is an important reactivity characteristic in transition metals [24,25] and the main group hydrides [20,26,27,28,29]. Similarly, the proton donor ability (PDA), i.e., thermodynamic acidity, was calculated as the Gibbs free energy (∆G°[H+]) for the reaction of proton (H+) detachment. In the case of anionic boron hydrides, both HDA and PDA parameters have been estimated by only a few researchers [30,31].
2. Results and Discussion
2.1. Interaction Tetrahydroborate-Anion with Hydrogen Halides
In our previous work [20], we showed (by DFT investigation of stepwise alcoholysis of tetrahydroborate-anion BH4−) that first proton-hydride transfer is the rate-limiting stage of the reaction, and that the energy profile takes the form of a cascade, since the energy of each next transition state is lower than the one previous. That, in turn, is associated with a decrease in Lewis acidity of the corresponding neutral borane (RO)xBH(3−x) and a weakening of B–H bond in [(RO)xBH(4−x)]− (x = 0−3) anion entailed by the successive substitution of H− by RO− at the boron atom. Thus, the B–H bond activation of BH4− and its ability to hydride transfer appeared to be crucial to this process.
It is known that alkaline tetrahydroborates react violently with concentrated mineral acids (such as H2SO4, HF, etc.) [14,32], yielding the substituted products and diborane, which is formed as a result of interaction between the reaction products. However, as we demonstrated in [30], an introduction of EWG (X−, X = F, Cl, Br) to boron atoms increases the Lewis acidity of the parent borane BH(3−x)Xx and reduces the ability of B–H bonds in [BH(4−x)Xx]− (x = 0−3) to hydride transfer. In this regard, it was interesting to investigate, in the current work, the reaction of BH4− with strong mineral acids such as hydrogen halides. For these purposes, we investigated this stepwise process of halogenation of BH4− with HX (where X = F, Cl, Br) in acetonitrile (MeCN) at M06/6-311++G(d,p) and wB97XD/6-311++G(d,p) levels of theory.
The dihydrogen-bonded (DHB) complexes are active intermediates of the BH4− reaction with HX [33] and HOR [20,21,34,35]. DHB complexes with hydrogen halides HX (X = F, Cl, Br) found in this work (Table S1) were monodentate (except for H3BH−···HF) with H···H distances of 1.425–1.850 Å and angles ∠X–H···H = 159–178°. Their structures were similar to those found in BH4− alcoholysis [20,21], with the geometric parameters typical of DHB complexes [36,37,38]. It should be noted, that for [H2XBH]− and [HX2BH]− (X = F, Cl, Br) molecules, two coordination modes were found, where halogen atom X was located in trans- and cis-position to the dihydrogen-bonded HX molecules. The formation energies of DHB complexes with a halogen atom in trans-position are slightly less favorable than those of DHB complexes with X in cis-position; the difference in the Gibbs free energy of these two complexes, ∆∆Gf°MeCN = ∆Gf°MeCN(INTnDHBtrans)– ∆Gf°MeCN(INTnDHBcis), reaches 1.8 kcal/mol. Moreover, the transition states of H2 elimination (TSnELIM and TSnCONC, see below) are energetically destabilized when X is located in trans-position, compared to those where X is located in cis-position. The difference ∆∆G°MeCN ‡ = ∆G°MeCN ‡(TSntrans) − ∆G°MeCN ‡(TSncis) can reach up to 2.2 kcal/mol. Thus, herein, we considered only the configurations with X located in cis-position to HX.
Upon going from HF to HBr, the enthalpy of DHB complexes’ formation (∆Hf°MeCN, Table S1) and other parameters characterizing the H···H interactions strength (EH···H and E2 are given in Table S1, and Wiberg bond indices (WBI) and QTAIM delocalization indices (DI)—in Table S3) decreased, in agreement with the HX acidity and hydrogen bond donor strength [39].
Substitution of one H− by X− (X = F, Cl, Br) in BH4− led to an increase of s-character of the B–H bond hybrid orbital in the resulting [BH3X]−, from 25.0% up to 31.9% (from sp3 to sp2, Table S3). In general, an increase of s-character of a B–H bond, as in the case of BH4− alcoholysis [20], is associated with strengthening of B–H bonds (according to QTAIM delocalization index, DI, Table S4) and weakening of DHB bonds. However, in the case of the BH4− reaction with HF for the H2FBH−···HF and HF2BH−···HF complexes (Figure 1), there was an increase of DHB interaction strength (E2 = 19.2 kcal/mol and 18.6 kcal/mol, respectively, Table S1). In fluoroborohydrides, the strong electron-withdrawing inductive σ-effect—due to the large difference in boron and fluorine electronegativity (∆χA-R = 2.1) [40]—led to the large positive charge on the boron atom (1.53 for BF3) [41]. On the other hand, this effect was compensated by the strong back donation from nonbonding 2pπ electron pairs of F to vacant 2pπ (B) orbital of the same symmetry [42,43]. For BCl3 and BBr3, due to a larger size of 2pπ orbitals, their effective overlap with vacant boron 2pπ orbital decreased; therefore, back-donation for BF3 was stronger than for BCl3 and BBr3, and Lewis acidity decreased in the row BF3 > BH3 > BCl3 > BBr3 (as can be seen from their hydride affinity values, HA298K, Table S3).
The reaction of BH4− with HX (where X = F, Cl, Br) in acetonitrile (MeCN) occurs as a successive proton-hydride transfer and H2 elimination, yielding [BH(4−x)Xx]− (x = 1−3) halogenated products (Figure 2, Figures S1 and S2). In the case of weak acids like HF (pKaMeCN(estim) = 25.2 [44]) only one product-like concerted transition state (TS) of proton-hydride transfer and H2 elimination was found (TS1CONCHF, Figure 3). In the case of stronger acids HCl (pKaMeCN= 10.3 [45]) or HBr (pKaMeCN = 5.4 [45]), two TS’s were observed: a reagent-like transition state of proton transfer (TS1PTHCl) and a product-like transition state of H2 elimination (TS1ELIMHCl). Both concerted TS of proton-hydride transfer/H2 elimination TSnCONCHX and TS of H2 elimination TSnELIMHX (preceded by proton transfer step TSnPTHX) exhibited a similar triangle-shaped form of front-side nucleophilic substitution of H2 by halide-ion X−. The B···H2 and H···H distances (Tables S5 and S6) in TSnELIMHX could be considered indirect indicators of activation barrier height. With the increase of the activation energy, the B···H2 distances decreased for the borohydrides with EWG substituents, while at the same time, H···H distances shortened for weak proton donors (such as HF, ROH, Table S7). Similar transition states were described previously in a theoretical study of BH4− alcoholysis by weak (MeOH, CF3CH2OH) and strong (CF3OH) proton donors [21].
In the case of strong acids (like HCl, HBr), the formation of a metastable H2 intermediate X−(η2-H2)BH3 (INTnH2) was observed (Figure 3). This resembled the intermediates of transition metal hydrides’ protonation [46], which have also been proposed for the protonation of main group element hydrides [21,47,48,49,50,51]. It should be noted that INTnH2 was formed only at the first and second reaction step for HCl, and at the first step for HBr. This was due to an increase in electron deficiency of BH2Cl, BH2Br, and BHCl2 fragments, strengthening of B–H and X···H(H) (X = Cl, Br) bonds, and a decrease of the donation from σ(B–R) (where R = H, Cl, Br) to σ*(H–H–B), which ultimately destabilized the H2 intermediate (Table S2). It should be noted that σ(B–R) to σ*(H–H–B) donation from R group in trans position to H2 complex was the greatest (E21 = 8.8–13.3 kcal/mol) and had a key effect on the stabilization of (η2-H2) complex. This is also in line with the less energy-demanding attack of HX in the case of [H2XBH]− and [HX2BH]− bearing X in trans-position to HX, in comparison to complexes with X in cis-position.
Thus, for the appearance of the transition state of proton transfer, which is absent in the case of weak acids and leads to the (η2-H2) complex formation, two conditions must apparently be satisfied. First, the pKa of the acid should be greater than the pKa of H3B(η2-H2) and, second, R3B borane should be a good Lewis acid; at that, σ(B–R) orbital (such as R = H, SiH3, SiF3 [51]) should donate to σ*(B–H–H) orbital and R should not be able to π-back-donate to a vacant 2pπ(B) orbital.
In addition, a strong acid can significantly lower the activation energy of the 2nd reaction stage—that is, the H2 elimination. However, the process of H2 elimination will still be rate-limiting and will always have the highest value of the activation barrier, because of the relatively high B–H bond strength and its hydride transfer ability. In this regard, the reactivity of substituted borohydrides in reactions involving hydride transfer can be evaluated via their hydride donor ability (HDAMeCN, Table S3). In the case of BH4− reaction with HF (Figures S1 and S2), the substitution led to HDAMeCN values of [BH3F]−, [BH2F2]− and [BHF3]−, reduced in comparison with BH4− (54.64 kcal/mol, 53.12 kcal/mol and 61.08 kcal/mol vs. 64.86 kcal/mol, respectively, for their Li salts; Table S3). As a result, the activation energy decreased by 7.6 and 2.4 kcal/mol for the reaction with the 2nd and 3rd molecule of HF. However, it increased by 6.3 kcal/mol at the last stage (Table 1). Thus, the reaction of BH4− with HF followed a trend similar to that of THF·BH3 and Me2NH·BH3 alcoholysis by (CF3)2CHOH [20,52].
In the case of HCl (Figure 2, Figure S3) and HBr (Figures S4 and S5), the reaction featured two distinct TS’s, i.e., for proton transfer (TSnPT) and H2 elimination (TSnELIM). The activation barriers of TSnPT were small (0.5–4.1 kcal/mol in the case of M06/6-311++G(d,p) and 3.2–7.1 kcal/mol in the case of wB97XD/6-311++G(d,p), see Figures S3 and S5), comparable to the first step of the gas phase reaction between BH4− and CF3OH (4.2 kcal/mol) [21]. The case of HCl demonstrates how the energy of the complex with molecular hydrogen Cl− [(η2-H2)BH(3−x)Clx] (x = 0−3) changes during the reaction (Figure 2). At the first step (n = 1, Figure 2), η2-H2 complex was a local minimum with ∆G°MeCN = 3.8 kcal/mol. Then, at the second step (n = 2, Figure 2), the energy rose, and it became a quasi-stationary intermediate at 11.7 kcal/mol. Over the next two steps, it disappeared, and the transition state became concerted. These changes were not surprising; during the reaction, as H− was replaced by X− (X = Cl, Br), the electron deficiency of the boron atom increased and, on the other hand, π-back-donation from occupied 2pπ(X) orbital to vacant 2pπ(B) orbital decreased. This led to an increase in the HDAMeCN of B–H bond in substitution products, and hence to a decrease of the B–H bond’s capability to hydride transfer. As a result, the hydrohalogenation process became concerted, and the activation barriers (∆G°MeCN‡(span), Tables S5 and S6) increased (from 9.1 to 26.1 kcal/mol for HCl and from 14.6 to 29.5 kcal/mol for HBr; and, in the case of wB97XD/6-311++G(d,p), from 10.1 to 28.6 kcal/mol for HCl and from 13.1 to 29.2 kcal/mol for HBr). In addition, due to the high electrophilicity of boron atoms in chloro- and bromoborohydrides the propensity of B–H bonds to proton and hydride transfer become comparable in magnitude (PDAMeCN = 85.5 and 83.5 kcal/mol vs. HDAMeCN = 73.0 and 77.6 kcal/mol for Li[BHCl3] and Li[BHBr3] respectively).
Thus, the value of the activation energy was influenced by two factors: the strength of the acid, which significantly lowered the activation barrier by increasing the polarization of the B–H bond in DHB complexes, and the ability of the occupied 2pπ(X) orbital to effectively back-donate to vacant 2pπ(B). In view of both of these factors, hydrogen chloride was the golden reactant, and the observed activation energy values the reaction of BH4− with HCl were the lowest (Table 1). Taking into account these findings, the DFT investigation of the reactivity of [B6H6]2− in electrophile-induced nucleophilic substitution (EINS) was carried out only on an example of the interaction with HCl in acetonitrile (MeCN).
2.2. Interaction of Hexahydro-Closo-Hexaborate Dianion with Hydrogen Chloride
Polyhedral closo-boranes are characterized by rather high values of hydride donor ability (e.g., HDAMeCN = 71.9 kcal/mol for [B6H6]2− and 79.6 kcal/mol for [B12H12]2−), which indicates the weak ability of the nonactivated B–H bond in closo-boranes to hydride transfer [29]. It is not surprising, therefore, that the most common method of functionalization of polyhedral boron hydrides follows the mechanism of electrophile-induced nucleophilic substitution (EINS) (Scheme 1). At the first stage, an electrophile promoter E+ is attached to the boron skeleton of a closo-borane. Then, the hydride bound to the electrophile is eliminated, with the formation of a quasi-borinium cation [53]. Finally, at the last stage, the nucleophile attachment occurs. At that point, if E+ = H+, the protonated closo-borane intermediate should convert into a η2-H2 complex—which is generally unstable and readily loses H2, making the addition of the nucleophile easier.
However, the main problems of functionalization of polyhedral boron hydrides through this mechanism are low specificity of nucleophilic substitution and difficulties in controlling the degree of substitution [54]. The reactivity of the remaining BH groups in nucleophilic substitution products can be estimated using the hydride donor ability (HDAMeCN). Thus, as discussed in this section, applying this approach, we explored the EINS successive mechanism on the example of the reaction of [B6H6]2− with HCl in acetonitrile.
The [B6H6]2− is the smallest stable closo-borane dianion [29,55,56], where the bulk of electron density is located on the boron atoms of the octahedral skeleton. The high nucleophilicity of boron atoms [57] in [B6H6]2− leads to its easy protonation. However, an attempt to isolate protonated species from Cs2[B6H6] in acidic aqueous solutions was unsuccessful due to irreversible hydrolysis of [B6H6]2− to borates and boric acid under these conditions [58]. Nevertheless, [facHB6H6]− (pKaH2O = 7.00) [18,59], in the form of Bu4N+ and Ph4P+ salts, was isolated from weakly acidic (pH = 4.5) H2O-HCl solutions and characterized by X-ray diffraction [58,60]. It was found that the facH+ proton was located above the B3 triangle facets of the octahedron and coordinated through the 4c–-3e bond [18,58,60]. The tendency of [B6H6]2− to hydrolyze in an acidic environment can be explained by the instability of the hypothetical doubly-protonated B6H8 closo-borane [18,61,62]. However, [B6H6]2− could be stable in acidic solutions if the protonation were stopped at the stage of [facHB6H6]− formation [61].
It should be noted that the mechanisms of [B6H6]2− consecutive halogenation remain practically unexplored. Only the addition of the second halogen to [B6H5X]2− in the presence of X2 and H2O (X = Cl, Br) has been explored by DFT, which proceeds with the partial opening of the boron cluster [17].
Thus, herein, we focused on the investigation of the EINS successive mechanism, simulated on the example of a model reaction in acetonitrile (MeCN) at DFT/M06/6-311++G(d,p) (Figure 4) and wB97XD/6-311++G(d,p) (Figure S7, Table S12) theory levels. It was found that the reaction of [B6H6]2− with HCl proceeded via preliminary formation of hydrogen-bonded complex on B3 triangle facets (INTnHB3) in three stages (Figure 5). These stages were: (1) protonation of the closo-borane cluster (TSnPT) on one of the B3 triangle facets with the formation of protonated form Cl− [facHB6H(6−x)Clx]− (x = 0–5) (INTnPT); (2) structural isomerization with H+ migration from B3 triangle facet to B apex (TSnISOfac-apx) with the formation of complexes with molecular hydrogen Cl− [(η2-H2)B6H(5−x)Clx]− (η2-H2-intermediate, INTnH2); and (3) H2 elimination followed by frontal nucleophilic attack on boron atoms by Cl− (TSnELIM), which is the last—and rate-limiting—stage. As can be seen (Figure 4), the activation barrier of the protonation stage was very low for all dianions [B6H(6−x)Clx]2− (x = 0−5) (∆G‡ of TSnPT = 2.5–10.7 kcal/mol); therefore, it did not matter which type of acid we chose to use for protonation. The only important limitation is that the pKa of the proton donor used should be higher than the pKa of Cl− [facHB6H(6−x)Clx]− (x = 0−5). After protonation, the reaction turned into a classical one—isomerization of Cl− [facHB6H(6−x)Clx]− to Cl− [(η2-H2)B6H(5−x)Clx]− (x = 0–5) (TSnISOfac-apx) and H2 elimination (TSnELIM) in the presence of a nucleophile (Cl−). As can be seen from the reaction profile (Figure 4), the H2 elimination was the rate-limiting stage and, therefore, the activation barrier of this stage and the selectivity of this process should be determined by the metastable η2-H2 reaction intermediate.
The first reaction intermediate was the hydrogen-bonded complex (INTnHB3) with HCl coordination on π-density of the B3 triangle facet of the [B6H(6−x)Clx]2− (x = 0–5) polyhedron [63]. It should be noted that the electron density of hydride atoms in [B6H6]2− is much lower than the electron density of B3 triangle facets; therefore, in contrast to BH4− [34,35,64], [B10H10]2− [49,65], and [B12H12]2− [65], the complexation of proton donors occurred exclusively on B3 triangle facets [66,67]. Upon the successive replacement of H− with Cl− in the closo-borane structure, gradual increases in the electron deficiency of the skeleton of closo-borane and the electrophilicity of boron atoms were observed. As the result, the values of hydride donor ability (HDAMeCN, Table S8) increase, whereas the values of proton donor ability (PDAMeCN) decrease (Table S8). This indicates a change in the character of boron atoms in closo-borane from nucleophilic to highly electrophilic and an increase of the Lewis acidity of the corresponding quasi-borinium cations [B6H(5−x)Clx]− formed as the result of H− transfer from [B6H(6−x)Clx]2−. Thus, for example, for [B6HCl5]2−, the proton transfer becomes more favorable than hydride transfer, which correlates with better hydrolytic stability of tetra- and penta-halogen derivatives of [B6H6]2− in acidic aqueous solutions and increased acidity of protonated monosubstituted halogenated derivatives [facHB6H5X]− (pKaH2O for X = Cl, Br, I were 5.35, 5.00 and 4.65, respectively; in comparison, pKaH2O = 7.00 for [facHB6H6]−) [42,43]. The study of the electronic structure of polyhedral closo-borane dianions [B6H(6−x)Clx]2− (x = 1−5) utilizing the NBO method showed that, as the substitution proceeded, the polarization of the hydrogen atom occurred. which It became less hydridic (the charge of the hydrogen atom varied from −0.027 to 0.002). These changes were manifested in a weakening of the hydrogen bond (decrease of EH···B(B2) and E2, Table S8), an elongation of H···B3 distances (Table S9), and a decrease in the enthalpy of H···B3 (INTnHB3) complexes’ formation (Table S9).
The energy barrier of the proton transfer reaction was very low (2.5–7.5 kcal/mol); thus, in the presence of acids, the protonated reaction intermediates Cl− [facHB6H(6−x)Clx]− (x= 0−5) (INTnPT) were immediately formed (Figure S3). At the early stages of the reaction, these were characterized by significant energy formation (∆Gf°MeCN from −8.5 kcal/mol for Cl− [facHB6H6]− to −0.4 kcal/mol for Cl− [facHB6H2Cl4]−). Previously, instability of [B6H6]2− in acidic water solutions was explained by possible further protonation and formation of doubly protonated products [18,61,62]. The calculated energy barrier of Cl− [facHB6H(6−x)Clx]− (x = 0−5) protonation by an equivalent of HCl was higher by 8.8–16.0 kcal/mol (∆G°MeCN‡ = 9.5–23.0 kcal/mol for TSnPT2nd, Table S10) than protonation of [B6H(6−x)Clx]2−, and its energy increased over the course of halogenation. It should be noted that the formation of doubly-protonated intermediates Cl− (facHB6H(6−x)ClxfacH)·Cl− (x = 0−5) is endoergic (∆Gf°MeCN values ranged from 9.7 kcal/mol for Cl− (facH2B6H6facH)·Cl− to 29.3 kcal/mol for Cl− (facHB6HCl5facH)·Cl−). The activation energy of the reverse reaction (TSPT2nd(rev), Table S10) was less than 3.8 kcal/mol. All this indicated that it was possible to stop the protonation reaction at the stage of the monoprotonated product, in the presence of equimolar amounts of the proton donor.
It should be noted that the structure of [facHB6H(6−x)Clx]− (x = 0−5) anions is nonrigid and, in solutions, they are prone to the fluctuation of facH+ proton between octahedral edges [18]. During [B6H6]2− halogenation, the Cl− [facHB6H(6−x)Clx]− to Cl− [(η2-H2)B6H(5−x)Clx]− (x = 0−5) isomerization (TSnISOfac-apx, Figure 5) was observed, with energy barriers decreasing from 20.1 kcal/mol at the first step to 13.8 kcal/mol at the sixth (Table 2)—correlating with an increase in the electrophilic nature of boron atoms during halogenation.
There was also a competitive Cl− [facHB6H(6−x)Clx]− to [B6H(6−x)ClxHfac*]−·Cl− (x = 0−5) isomerization process (TSnISOfac-fac*, Table S9), which corresponded to facH/Hterm exchange. As is known, [B6H6]2− can be completely deuterated in D2O solution—and activation energy of the deuteration (18.0 kcal/mol) [18,59] was close to the calculated activation energy of 17.3 kcal/mol for the TSnISOfac-fac* isomerization of [facHB6H6]−. Thus, the exchange between facH and Hterm ligands (TSnISOfac-fac*, Scheme 2) featuring a lower barrier than TSnELIM did not result in H2 elimination—which requires η2-H2 complex formation.
A third reaction intermediate, which is Cl− [(η2-H2)B6H(6−x)Clx]− complex (INTnH2) (Table S10) was observed for almost all stages of reaction of [B6H(6−x)Clx]2− (x = 0−5) with HCl (except [B6HCl5]2−). This complex is a key intermediate of the process which determines the selectivity of nucleophilic substitution, since stabilization of H2 at one of the vertices of the closo-borane octahedron determines the position of nucleophilic substitution at this vertex. Notably, this complex was not revealed at the gas phase calculations of Cl− [facHB6H6]− [68]. Similarly to the Cl− [(η2-H2)B10H9]− [49] it could be stabilized only in the presence of HX acid residue X−.
As the halogenation of [B6H6]2− progressed, the formation energy of the H2 intermediate decreased, and B–H and Cl···H bonds became stronger while H···H interaction weakened. This coincided with a trend toward an increase in the electrophilicity of the B–H bond and an increase in its acidity. Thus, the established regularities allowed for the use of the stabilization of H2 complex to control the process of nucleophilic substitution and lower the activation energy of this process.
The same trend was observed for the rate-limiting stage of H2 elimination, for which the energy barrier increased from 19.6 to 42.3 kcal/mol (∆G°MeCN‡(TSnPT), shown in Table 2, and from 24.6 to 44.6 kcal/mol for wB97XD/6-311++G(d,p), shown in Table S12. It should be noted that nucleophilic substitution by Cl− and solvation contributed greatly to the stabilization of the transition state—which was noticeable, as the activation energies obtained were compared with gas-phase activation energies for H2 elimination from [facHB6H6]− (∆E = 45.9 kcal/mol) [69].
In the reaction of [B6H6]2− with HCl there were three branch points at the second, third, and fourth steps (Scheme 3) of halogenation when several geometries of halogenated derivatives were possible (Table 2). However, the optimal reaction pathway (with the lowest activation energy of TSnELIM at 2nd and 3rd steps) was [B6H6]2− → 1-[B6H5Cl]2− → 1,2-[B6H4Cl2]2− (cis) → 1,2,3-[B6H3Cl3]2− (fac) → 1,2,3,4-[B6H2Cl4]2− (cis) → 1,2,3,4,6-[B6HCl5]2−. The existence of such branch points was found for 1-[B6H5Cl]2−, 1,2-[B6H4Cl2]2− (cis) and 1,2,6-[B6H3Cl3]2− (mer)-halogenated derivatives, in which there were two types of hydrides with different reactivities (as can be seen from HDAMeCN values, Table S8. As follows from Table 2, the smallest activation barrier corresponded to nucleophilic substitution at the boron atom, the bond of which was characterized by the smallest HDAMeCN value.
It should be noted that under real conditions of halogenation of [B6H6]2− by X2 (X = Cl, Br) in strong alkaline solutions, the reaction is nonspecific. In the second, third, and fourth steps of halogenation, the geometric isomers are formed in different ratios [18]. In this case, for X = I pure geometric isomers, 1,6-[B6H4I2] (trans), 1,6,2-[B6H3I3] (mer) and 1,6,2,4-[B6H2I4] (trans) were observed [70,71]. This was due to the less effective back-donation of larger-sized 2pπ iodine orbitals to vacant 2pπ(B). That, in turn, led to the favorable formation of the protonated complex [facHB6H5X]−, where facH+ was located in antipodal-sphere of the [B6H6]2− octahedron [18,72]. Our calculations showed that, for [facHB6H5Cl]−, the formation energy of the complex with facH+ in the ipso-sphere of the [B6H6]2− octahedron was higher than that of the complex with facH+ in the antipodal-sphere (∆∆Gf°MeCN = 3.0 kcal/mol). Moreover, ipso-[facHB6H5Cl]− was the active intermediate of [B6H5Cl]2− halogenation to 1,2-[B6H4Cl2]2− in our mechanism, as well as in that reported in [17]. Apparently, the larger electron-withdrawing effect of iodine in [B6H(6−x)Ix]2− (x = 0–5) derivatives increases the difference between ipso/antipodal complexes.
Thus, the above-described reaction mechanism on the example of a model reaction of [B6H6]2− with HCl suggests the possibility of controlling the selectivity and the depth of the nucleophilic substitution processes by increasing the stabilization of the H2 complex on the desired position of the closo-borane cluster. Additionally, at the same time, HDAMeCN values for the BH groups would permit the carrying out of a nucleophilic substitution of B–H bonds activated by Lewis acids (BL3).
3. Materials and Methods
Computational Details
In the present manuscript, the calculations were performed at the DFT/M06/6-311++G(d,p) level for comparison to the hydride donor ability (HDA) values of boron hydrides and other main group hydrides, as previously calculated by the same method [28,30,73]. Additionally, the values obtained can be used as a reference for assessing the effectiveness of the B–H bond activation by transition metals. In this regard, M06 is a more versatile method than M06-2X (generally recommended for thermochemistry calculations of main group elements) and can be used in cases where multi-reference systems are or might be involved, since it has been parametrized for both main group elements and transition metals [74]. The continuum solute model based on density (SMD) was used for nonspecific solvation, since this model was parametrized to work with the Minnesota functionals family (such as M06, M06-2X, etc.) and has proven to be effective for use in both charged and uncharged systems [75]. The nature of the stationary points on potential energy surfaces was proved by reoptimization of M06 geometries at wB97XD/6-311++G(d,p) level of theory.
All calculations were performed without symmetry constraints using the M06 hybrid functional [74] and wB97XD dispersion-corrected, range-separated hybrid functional [76] implemented in the Gaussian09 (Revision D.01) [77] software package, using 6-311++G(d,p) [78] basis set. Vibrational frequencies were calculated for all optimized complexes at the same level of theory to confirm a character of local minima on the potential energy surface. Visualization-optimized geometries were realized using the Chemcraft 1.8 graphical visualization program [79].
The inclusion of nonspecific solvent effects in the calculations was performed using the SMD method [75]. Acetonitrile (MeCN, ε = 35.7) was chosen as a solvent for optimization because of significant data, specific to MeCN, already existed on reduction potentials, pKa values, and experimental values of hydride donor ability (HDA) of transition metal hydrides [24,25].
Calculations were carried out with an ultrafine integration grid and a very tight SCF option to improve the accuracy of the optimization procedure.
Hydride donor ability (HDAMeCN) and proton donor ability (PDAMeCN) of E–H bond in MeCN were calculated as Gibbs free energy of hydride or proton transfer:
HDAMeCN = ΔG°[H−] MeCN = G°MeCN (E+) + G°MeCN (H−) − G°MeCN (E−H)(1)
PDAMeCN = ΔG°[H+] MeCN = G°MeCN (E−) + G°MeCN (H+) − G°MeCN (E−H)(2)
The value of G°MeCN (H−) = −413.9 kcal/mol was calculated by M06/6-311++G(d,p) for the H− in MeCN. The values of G°MeCN (H+) = −260.2 kcal/mol were taken from [80].
From data obtained during the geometry optimization for each molecule, the most stable configuration was chosen. To find the most stable configuration of cationic boranes, the terminal hydrogen atoms (B–Hterm) were torn off as H− or as H+ from each vertex in optimized molecules. In the case of polyhedral closo-boranes, due to the rigid frame of the boron cluster [49,50], several quasi-borinium cations were observed, localized on vertices from which hydride was torn off.
The Wiberg bond indices [81] (WBI)—as a measure of electron distribution between two atoms [82]—and natural atomic charges were calculated using the natural-bond orbital (NBO) analysis [83] implemented in GAUSSIAN09 (Revision D.01) [77]. Topological analysis of the electron-density distribution function ρ(r) was performed using the AIMALL [84] program package based on the wave function obtained by M06 calculations. The delocalization index (DI) [85,86,87], a measure of electrons that are shared or exchanged between two atoms or basins obtained from the integration of the Fermi hole density, was directly related to the bond order [87,88]. The X–Y bond ellipticity, εX–Y, was defined as ε = (λ1/λ2 − 1), where λ1 and λ2 are the negative eigen values of the Hessian of the electron density at the bond critical point, ordered such that λ1 < λ2 < 0 [89,90,91]. The energies of H···H or H···B3 interactions were calculated as EH···H = 0.5·V(r), according to [92,93].
4. Conclusions
The mechanism of the sequential halogenation of tetrahydroborate anions [BH4]− by hydrogen halides (HX, X = F, Cl, Br) and halogenation of hexahydro-closo-hexaborate dianions [B6H6]2− was established by ab initio DFT calculations (M06/6-311++G(d,p) and wB97XD/6-311++G(d,p)) in acetonitrile (MeCN), taking into account nonspecific solvent effects (SMD model). The active intermediates of the reaction—XH···HB dihydrogen-bonded (DHB) complexes for [BH4]− and ClH···B3 complexes for [B6H6]2−—were identified, and their geometric parameters, energetic characteristics, and electron structures were deduced. In the case of [B6H6]2−, an electrophile-induced nucleophilic substitution (EINS) mechanism of halogenation was established. The obtained values of activation energies of H2 elimination were relatively high. Nonetheless, the identified pattern in the changes of electronic characteristics, the reactivity of B–H bonds (such as thermodynamic hydridicity and acidity of halogenated products) and the structures of H2 complexes suggests that it should be possible to control the degree of halogenation and carry out the targeted synthesis of products with a low degree of substitution using a wide range of synthetic approaches.
Supplementary Materials
The following are available online, Table S1. M06-computed geometric parameters (distances r(X–Y) in Å, angles in °) of dihydrogen-bonded (DHB) complexes (INTnDHB) of [BH(4−x)Xx] − (x = 0–3) with HX (X = F, Cl, Br) and their energies. Table S2. M06-computed geometric parameters (distances r(X–Y) in Å, angles in °) of H2 complexes of X−[(H2)BH(4−x)Xx]− (x = 1–2, X = Cl, Br) (INTH2) and their energies. Table S3. Hydride affinity (HA298K in kcal/mol) of [BH(4−x)Xx]− (x = 0–3) (X = F, Cl, Br). Hybridization of boron orbitals in B–H moieties, QTAIM delocalization indices (DI) of BH bond in of [BH(4−x)Xx]− (x = 0–3, X = F, Cl, Br). Hydride (HDAMeCN in kcal/mol) and proton (PDAMeCN in kcal/mol) donor abilities for Li[BH(4−x)Xx] in MeCN. Table S4. QTAIM delocalization indices (DI) and Wiberg bond indices (WBI) for H···H contacts, their differences for XH and HB bonds, and hybridization of boron orbitals in BH moieties of M06-computed geometries of DHB complexes for [BH(4−x)Xx]− (x = 0–3) with HX (X = F, Cl, Br) (INTnDHB). Table S5. Free Gibbs activation energy in acetonitrile (∆G°MeCN‡ in kcal/mol) of the transition state of H2 elimination (TSnELIMHX), and concerted transition state of proton-hydride transfer/H2 elimination (TSnCONCHX) for the reaction of BH4− with HX (X = F, Cl, Br)—and their M06-optimized geometric parameters. Table S6. Free Gibbs activation energy in acetonitrile (∆G°MeCN‡ in kcal/mol) of the transition state of H2 elimination (TSnELIMHX) and concerted transition state of proton-hydride transfer/H2 elimination (TSnCONCHX) for the reaction of BH4− with HX (X = F, Cl, Br) and their wB97XD-optimized geometric parameters. Table S7. Free Gibbs activation energy in dichloromethane (∆G°DCM‡ in kcal/mol) of the concerted transition state of proton-hydride transfer/H2 elimination (TSCONCROH) for the reaction of BH4− with HOR [ROH = MeOH, CF3CH2OH (TFE), (CF3)2CHOH (HFIP)] and their M06-optimized geometric parameters. Table S8. Hydride (HDAMeCN in kcal/mol) and proton (PDAMeCN in kcal/mol) donor abilities for [B6H(6−x)Clx]2− (x = 0−5) in MeCN. Table S9. M06-computed geometric (distances r(X–Y) in Å, angles in degrees) parameters of hydrogen-bonded complexes of [B6H(6−x)Clx]2− (x = 0–5) with HCl (INTnHB) and their energies. Table S10. M06-computed geometric (distances r(X–Y) in Å, angles in degrees) parameters of hydrogen-bonded complexes of Cl− [facHB6H(6−x)Clx]− (x = 0–5) with HCl and their energies. Table S11. M06-computed geometric parameters (distances r(X–Y) in Å, angles in °) of H2 complexes of Cl−[(H2)B6H(5−x)Clx]− (n = 0–5) (INTnH2) and their energies. Table S12. Activation barriers (∆G°MeCN‡, in kcal/mol) for TSs for the reaction of [B6H6]2− with HCl (pKaMeCN = 10.3) in MeCN, computed at the wB97XD/6-311++G(d,p) theory level. Figure S1. M06-calculated energy profile (∆G°MeCN in kcal/mol) of the reaction of BH4− with HF in MeCN. Figure S2. wB97XD-calculated energy profile (∆G°MeCN in kcal/mol) of the reaction of BH4− with HF in MeCN. Figure S3. wB97XD-calculated energy profile (∆G°MeCN in kcal/mol) of the reaction of BH4− with HCl in MeCN. Figure S4. M06-calculated energy profile (∆G°MeCN in kcal/mol) of the reaction of BH4− with HBr in MeCN. Figure S5. wB97XD-calculated energy profile (∆G°MeCN in kcal/mol) of the reaction of BH4− with HBr in MeCN. Figure S6. M06-optimized geometries of the protonated forms [facHB6H(6−x)Clx]− (n = 0−4) and their free Gibbs formation energies (∆Gf°MeCN in kcal/mol). Figure S7. wB97XD-calculated energy profile of halogenation of [B6H(6−x)Clx]2− (x = 0−5) by HCl for the pathway: [B6H6]2− → 1-[B6H5Cl]2− → 1,2-[B6H4Cl2]2− (cis) → 1,2,3-[B6H3Cl3]2− (fac) → 1,2,3,4-[B6H2Cl4]2− (cis) → 1,2,3,4,6-[B6HCl5]2−. Animation of the vibration motion at the imaginary frequency for the TSs (GIF). Cartesian coordinates and electronic energies (in Hartrees) for the calculated species (XYZ).
Author Contributions
Funding acquisition, I.E.G.; Investigation, I.E.G.; Resources, O.A.F.; Supervision, L.M.E. and E.S.S.; Visualization, I.E.G.; Writing—original draft, I.E.G.; Writing—review & editing, O.A.F. and N.V.B. All authors have read and agreed to the published version of the manuscript.
Funding
This research was financially supported by the Russian Science Foundation (RSF grant number 19-73-00309). L.M.E. and E.S.S. thank the Ministry of Science and Higher Education of the Russian Federation for the partial support of this research.
Institutional Review Board Statement
Not applicable.
Informed Consent Statement
Not applicable.
Data Availability Statement
Not applicable.
Conflicts of Interest
The authors declare no conflict of interest.
Sample Availability
Not available.
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Figures, Schemes and Tables
Figure 1. M06-optimized geometries of DHB complexes of [BH(4−x)Fx]− (x = 0–3) with HF (INTnDHBHF). The r(X···Y) bond distances and ∠H···HF angles shown are in Å and °, respectively.
Figure 2. M06-calculated energy profile (∆G°MeCN in kcal/mol) of the stepwise reaction of BH4− with HCl in MeCN.
Figure 3. M06-optimized geometries of transition states (TS) and H2 reaction intermediates (INTH2) for the first stage of reaction between BH4− with HX (X = F, Cl): concerted transition state of proton-hydride transfer and H2 elimination for HF (TS1ELIMHF) and transition states of proton transfer (TS1PTHCl) and H2 elimination (TS1ELIMHCl) and H2 intermediate (Cl−(η2-H2)BH3, INT1H2HCl) for HCl. The r(X···Y) bond distances and ∠H···HF angles shown are in Å and in °, respectively.
Figure 4. M06-calculated energy profile (∆G°MeCN in kcal/mol) of halogenation of [B6H(6−x)Clx]2−(x = 0−5) by HCl for the pathway: [B6H6]2− → 1-[B6H5Cl]2− → 1,2-[B6H4Cl2]2− (cis) → 1,2,3-[B6H3Cl3]2− (fac) → 1,2,3,4-[B6H2Cl4]2− (cis) → 1,2,3,4,6-[B6HCl5]2−.
Figure 5. M06-optimized transition states of reaction [B6H6]2− with HCl. The r(X···Y) bond distances shown are in Å.
Scheme 2. The Hfac/Hfac exchange mechanism in Cl− [facH*B6H6]−. ∆G°MeCN‡(span) shown in bold numbers are in kcal/mol.
Scheme 3. Pathways of [B6H6]2− halogenation by HCl. Free Gibbs energy span (∆G°MeCN‡(span) in kcal/mol) indicated as black bold numbers. H-atoms are omitted for clarity. Cl atoms added at each particular step are shown in magenta.
Activation barriers (∆G‡MeCN, in kcal/mol) for TSnCONCHX (or TSnPTHX/TSnELIMHX) in the step-wise reaction of BH4− with hydrogen halides in MeCN, computed at the M06/6-311++G(d,p) theory level relative to the corresponding DHB complexes.
n | HF | HCl | HBr |
---|---|---|---|
pKaMeCN = 25.2 a | pKaMeCN = 10.3 b | pKaMeCN = 5.4 b | |
1 | 27.4 | 2.3/5.2 c | 4.1/6.5 c |
2 | 19.8 | 0.5/7.3 c | 13.0 |
3 | 17.5 | 14.8 | 21.0 |
4 | 23.8 | 21.0 | 25.3 |
a An estimated value is taken from [44]; b Experimental values are taken from [45]; c The values for TSnPTHX/TSnELIMHX are given.
Table 2Hydride donor ability (HDAMeCN) of the B–H center in [B6H(6−x)Clx]2− (x = 0−5), involved in halogenation reactions, and activation barriers (∆G°MeCN‡, in kcal/mol) for TSs for the reaction of [B6H6]2− with HCl (pKaMeCN= 10.3 [42]) in MeCN, computed at the M06/6-311++G(d,p) theory level.
n | N of Position | HDAMeCN,a | ∆G°MeCN‡ |
∆G°MeCN‡ |
∆G°MeCN‡ |
∆G°MeCN‡(span) d |
---|---|---|---|---|---|---|
1 | 1 | 71.9 | 0.7 | 20.1 | 19.6 | 39.3 |
2 | 1,2 (cis) | 76.6 | 1.3 | 19.8 | 21.9 | 38.2 |
1,6 (trans) | 80.0 | 2.3 | 19.0 | 26.0 | 44.3 | |
3 | 1,2,3 (fac) | 81.0 | 2.1 | 19.4 | 23.0 | 39.2 |
1,2,6 (mer) | 84.7 | 2.2 | 17.9 | 26.0 | 41.5 | |
1,6,2 (mer) | 81.1 | 3.6 | 16.7 | 23.4 | 40.2 | |
4 | 1,2,3,4 (cis) | 88.6 | 2.1 | 17.4 | 26.6 | 42.3 |
1,6,2,3 (cis) | 84.9 | 4.2 | 16.0 | 24.4 | 42.0 | |
1,6,2,4 (trans) | 88.6 | 3.9 | 15.0 | 26.4 | 40.2 | |
5 | 1,2,3,4,6 | 92.5 | 4.8 | 13.8 | 26.8 | 43.7 |
1,6,2,4,3 | 88.4 | 7.2 | 37.4 * | 42.4 | ||
6 | 1,2,3,4,6,5 | 95.6 | 8.0 | 37.6 * | 44.9 |
a HDAMeCN are given for the corresponding reactive B–H centre in [B6H(6−x)Clx]2− (x = 0−5) molecule; b Calculated as ∆G°MeCN‡ (TSnPT) = G°MeCN(TSnPT) − G°MeCN(INTnHB3) b Calculated as ∆G°MeCN‡ (TSnISOfac-apx) = G°MeCN(TSnISOfac-apx) − G°MeCN(INTnPT); c Calculated as ∆G°MeCN‡(TSnELIM) = G°MeCN(TSnELIM) − G°MeCN(INTnH2); d Calculated as ∆G°MeCN‡(span) = G°MeCN(TSnELIM) − G°MeCN(INTnH2) for n = 1−4 and as ∆G°MeCN‡(span) = G°MeCN(TSnELIM) − G°MeCN([B6H(6−x)Clx]2−) − G°MeCN(HCl) for n = 5−6; * TSnCONC for 5th and 6th halogenation steps.
You have requested "on-the-fly" machine translation of selected content from our databases. This functionality is provided solely for your convenience and is in no way intended to replace human translation. Show full disclaimer
Neither ProQuest nor its licensors make any representations or warranties with respect to the translations. The translations are automatically generated "AS IS" and "AS AVAILABLE" and are not retained in our systems. PROQUEST AND ITS LICENSORS SPECIFICALLY DISCLAIM ANY AND ALL EXPRESS OR IMPLIED WARRANTIES, INCLUDING WITHOUT LIMITATION, ANY WARRANTIES FOR AVAILABILITY, ACCURACY, TIMELINESS, COMPLETENESS, NON-INFRINGMENT, MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. Your use of the translations is subject to all use restrictions contained in your Electronic Products License Agreement and by using the translation functionality you agree to forgo any and all claims against ProQuest or its licensors for your use of the translation functionality and any output derived there from. Hide full disclaimer
© 2021 by the authors.
Abstract
The mechanism of the consecutive halogenation of the tetrahydroborate anion [BH4]− by hydrogen halides (HX, X = F, Cl, Br) and hexahydro-closo-hexaborate dianion [B6H6]2− by HCl via electrophile-induced nucleophilic substitution (EINS) was established by ab initio DFT calculations [M06/6-311++G(d,p) and wB97XD/6-311++G(d,p)] in acetonitrile (MeCN), taking into account non-specific solvent effects (SMD model). Successive substitution of H− by X− resulted in increased electron deficiency of borohydrides and changes in the character of boron atoms from nucleophilic to highly electrophilic. This, in turn, increased the tendency of the B–H bond to transfer a proton rather than a hydride ion. Thus, the regularities established suggested that it should be possible to carry out halogenation more selectively with the targeted synthesis of halogen derivatives with a low degree of substitution, by stabilization of H2 complex, or by carrying out a nucleophilic substitution of B–H bonds activated by interaction with Lewis acids (BL3).
You have requested "on-the-fly" machine translation of selected content from our databases. This functionality is provided solely for your convenience and is in no way intended to replace human translation. Show full disclaimer
Neither ProQuest nor its licensors make any representations or warranties with respect to the translations. The translations are automatically generated "AS IS" and "AS AVAILABLE" and are not retained in our systems. PROQUEST AND ITS LICENSORS SPECIFICALLY DISCLAIM ANY AND ALL EXPRESS OR IMPLIED WARRANTIES, INCLUDING WITHOUT LIMITATION, ANY WARRANTIES FOR AVAILABILITY, ACCURACY, TIMELINESS, COMPLETENESS, NON-INFRINGMENT, MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. Your use of the translations is subject to all use restrictions contained in your Electronic Products License Agreement and by using the translation functionality you agree to forgo any and all claims against ProQuest or its licensors for your use of the translation functionality and any output derived there from. Hide full disclaimer