Role of Alkaline-Earth Metal-Catalyst: A Theoretical Study of Pyridines Hydroboration

Density functional theory (DFT) calculations have been performed to investigate the mechanism of alkaline-earth-metal-catalyzed hydroboration of pyridines with borane. In this reaction, the active catalytic species is considered to be an alkaline earth metal hydride complex when the corresponding alkaline earth metal is used as the catalyst. The theoretical results reveal that initiation of the catalytic cycle is hydride transfer to generate a magnesium hydride complex when β-diimine alkylmagnesium is used as a pre-catalyst. The magnesium hydride complex can undergo coordination of the pyridine reactant followed by hydride transfer to form a dearomatized magnesium pyridine intermediate. Coordination of borane and hydride transfer from borohydride to magnesium then give the hydroboration product and regenerate the active magnesium hydride catalyst. The rate-determining step of the catalytic cycle is hydride transfer to pyridine with a free energy barrier of 29.7 kcal/mol. Other alkaline earth metal complexes, including calcium and strontium complexes, were also considered. The DFT calculations show that the corresponding activation free energies for the rate-determining step of this reaction with calcium and strontium catalysts are much lower than with the magnesium catalyst. Therefore, calcium and strontium complexes can be used as the catalyst for the reaction, which could allow mild reaction conditions.

Alkaline-earth (Ae) metals, which are located in group IIA of the periodic table of elements, have attracted wide attention because of their low cost, availability, and abundant reserves in nature (Green et al., 2007). Ae metals, such as magnesium, calcium, strontium, and barium, are commonly used as homogeneous catalysts. Because the d 0 valence electron configuration at the center of the divalent cation (Ae 2+ ) in the Ae metal catalyst shows partial "lanthanide" characteristics, a similar catalytic cycle can be constructed in homogeneous catalysis (Li and Marks, 1996;Westerhausen, 2008;Krieck et al., 2009;Hill et al., 2016;Ma et al., 2016;Rochat et al., 2016;Rossin and Peruzzini, 2016;Xu et al., 2017a).
Hill and co-workers (Arrowsmith et al., 2011) first reported the magnesium-catalyzed hydroboration of Pyridines in 2011. In 2014, Harder's group (Intemann et al., 2014) reported hydroboration of pyridine catalyzed by the Ae metal magnesium and obtained the 1,2-selective addition compound 4 as the major product (Scheme 1). Harder et al. suggested that magnesium hydride species 6 acts as a catalyst in the catalytic cycle. Based on the experimental observations and our previous theoretical studies of Ae catalysis, there are two possible reaction pathways. Raw material pyridine 1 coordinates with magnesium in active catalytic species 6 to give intermediate 7 (Scheme 2). The C = N double bond in pyridine then inserts into the Mg-H bond to give dearomatized magnesium amino intermediate 8, which could coordinate with pinacolborane 2 to give nitrogen-boron compound 9. Finally, magnesium hydride species 6 is regenerated by hydride transfer with the release of 1,2-borohydropyridine derivative 4. Alternatively, intermediate 8 could be isomerized by 1,3-hydrogen migration to give intermediate 10, which then coordinates to pinacolborane 2 to give nitrogen-boron complex 11. After the corresponding hydride transfer, 1,4-borohydropyridine product 5 is obtained and active magnesium hydride species 6 is regenerated.
Here, we performed a theoretical mechanistic study of Ae catalysis and determined the trend of the reactivates of Aecatalyzed hydrogenation reactions. The pyridine hydroboration reaction catalyzed by Ae metals reported by Harder's group was used as the template reaction for the theoretical calculations. The calculations were performed based on density functional theory (DFT) and the reaction mechanism is discussed to better understand the reaction process and provide theoretical guidance for subsequent organic reactions catalyzed by Ae metals.

COMPUTATIONAL METHODS
All of the DFT calculations were performed with the Gaussian 09 software package (Frisch et al., 2013). The B3-LYP (Lee et al., 1988;Becke, 1993;Stephens et al., 1994;Adamo and Barone, 1999) functional with the standard 6-31G(d) (Hehre et al., 1972;Francl et al., 1982) basis set (SDD Dolg et al., 1987 basis set for strontium atoms) was used for the geometry optimizations. Harmonic vibrational frequency calculations were performed for all of the stationary points to determine whether they are local minima or transition structures and to derive the thermochemical corrections for the enthalpies and free energies. The M11 (Peverati and Truhlar, 2011) functional with the 6-311+G(d) basis set (SDD for strontium atoms) was used to calculate the single-point energies, because it is expected that this strategy will provide more accuracy with regard to the energetic information Zhao et al., 2012;Yu and Lan, 2013;Cui et al., 2019). The solvent effect of benzene was considered by single-point calculations of the gasphase stationary points with the solvation model based on the density (SMD), which is a continuous model (Cances et al., 1997;Marenich et al., 2009). The reported free energies are the M11calculated Gibbs free energies in benzene solvent based on the B3-LYP-calculated geometries with thermodynamic corrections. The geometric configurations of the key reaction intermediates and transition states were generated with CYLview software (Legault, 2009). The total energies for all of the calculated structures are listed in Supplementary Material. SCHEME 1 | Magnesium-catalyzed hydroboration of pyridines with borane. SCHEME 2 | Plausible mechanism for magnesium-catalyzed hydroboration of pyridines with borane.

Mechanism of Pyridine Hydroboration Catalyzed by Magnesium-Hydrogen Species
DFT calculations were performed for hydroboration of pyridines catalyzed by magnesium (Scheme 2). We selected β-diimine magnesium hydride 6 as the relative zero point of the Gibbs free energy (Figure 1). Pyridine 12 coordinates with active magnesium hydride species 6 to form intermediate 13. This process is exergonic by 7.9 kcal/mol, indicating that the magnesium hydride species can be stabilized as the Lewis acid by the corresponding Lewis base. After pyridine is activated by the magnesium hydride species, the C=N double bond in pyridine 12 inserts into the Mg-H bond via transition state 14ts. the energy barrier of insertion step is 29.7 kcal/mol, indicating that this step is the rate-determining step of the catalytic cycle. The geometric structure of transition state 14-ts is shown in Figure 2. The length of the C-H bond to be formed is 1.58 Å. The length of the Mg-H bond to be ruptured is 1.91 Å. The   length of the C-N bond in pyridine involved in the reaction is 1.40 Å. The above data of the geometrical structure show that formation of the C-H bond and rupture of the Mg-H bond simultaneously occur in this reaction. Intermediate 15 is obtained when C=N double insertion occurs. Because the aromaticity of pyridine in intermediate 15 is destroyed, its relative energy is 9.9 kcal/mol higher than that of intermediate 13. In addition, we also considered another route to obtain intermediate 15.
With pinacolborane 2 acts as a bridge, the hydrogen atom migrates from magnesium to boron via hexatomic ring transition state 20-ts and then transfers from boron to the ortho carbon of nitrogen to give intermediate 15. The activation energy of this process is 33.5 kcal/mol, which indicates that it is a possible alternative pathway to generate intermediate 15.  calculations show that this process proceeds via transition state 18-ts with an activation free energy of only 5.3 kcal/mol. After this step, 1,2-borohydropyridine product 19 is released and active catalytic species 6 is regenerated. The complete catalytic cycle is only exothermic by 6.2 kcal/mol. This indicates a weak thermodynamic driving force, which can be attributed to the aromaticity of pyridine being destroyed. Nevertheless, a stable N-B covalent bond is generated to drive the reaction.

Mechanism of Pyridine Hydroboration Catalyzed by Other Ae Metals
Harder's group only determined the catalytic activities of magnesium species. Followed these results, Harder and co-workers (Intemann et al., 2015) reported the calcium catalyzed hydrosilylation of pyridine, demonstrating that a beta-deketiminate calcium hydride is capable of faster selective 1,2-dearomatisation in 2015. According to the basic rule of the periodic law of elements and our previous theoretical studies (Gao et al., 2016;Qi et al., 2017;Xu et al., 2017a,b; elements of the same main group often have similar chemical properties (Cotton et al., 1988). After investigating the mechanism of pyridine hydroboration catalyzed by the magnesium catalyst, we performed M11 calculations to predict the catalytic activities of other Ae metals in the group. In the pyridine hydroboration reaction, the rate-determining step is considered to be pyridine C = N bond insertion with Mg-H bond cleavage. Therefore, a weak metal-H covalent bond is favorable for this reaction. The calculated bond dissociation energies (BDEs) of Ae-H bonds are shown in Figure 3. The BDEs of Be-H, Mg-H, Ca-H, and Sr-H are 96.3, 75.5, 64.6, and 62.2 kcal/mol, respectively. Therefore, beryllium, which is in the same group as magnesium, has a small atomic radius, large steric hindrance, low metallic property, and strong Be-H bond. We believe that hydroboration cannot be catalyzed by beryllium hydride species. Furthermore, calcium and strontium could have higher catalytic activity than magnesium because of the weak metal-H covalent bonds owing to their larger atomic radii and lower electronegativities.
We also calculated the potential energy surfaces of the pyridine hydroboration reaction catalyzed by calcium and strontium hydride (Figure 4). In Figure 4A, active calcium hydride catalyst 6a is selected as the relative zero of the Gibbs free energy. First, calcium hydride species 6a coordinates with pyridine 12 to give intermediate 13a with a relative Gibbs free energy of 1.0 kcal/mol. The pyridine C=N bond then inserts into the Ca-H covalent bond via transition state 14a-ts. The energy barrier of this process is 17.4 kcal/mol, indicating that this step is the rate-determining step of the catalytic cycle. Compared with the corresponding reaction catalyzed by magnesium, the activation energy required for the rate-determining step decreases by 12.3 kcal/mol. Thus, we can conclude that the activity of the calcium catalyst is FIGURE 5 | Geometry of transition states 14a-ts, 16a-ts, 18a-ts, 14b-ts, 16b-ts, and 18b-ts. The bond lengths are given in ångstroms.
Frontiers in Chemistry | www.frontiersin.org higher than that of the magnesium catalyst. Intermediate 15a coordinates with pinacolborane 2 via the corresponding lowenergy transition state 16a-ts to give boron-nitrogen Lewis acidbase complex 17a. Finally, 1,2-borohydropyridine product 19 is released by hydride transfer with regeneration of active catalytic species 6a. The geometric structures of 14a-ts, 16a-ts, and 18ats are shown in Figure 5. The corresponding Ae metal transition states have similar structures.
Strontium hydride species 6b can also coordinate with pyridine 12 to give intermediate 13b, which is endergonic by 2.1 kcal/mol ( Figure 4B). Intermediate 13b is unstable and prone to react. The activation energy required for C=N bond insertion into the Sr-H bond via transition state 14b-ts is 14.4 kcal/mol. However, intermediate 17b generated by coordination to pinacolborane 2 is very stable, and its free energy is 9.5 kcal/mol lower than that of the 1,2-borohydropyridine product 19 and regenerated active catalyst 6b. Therefore, the complete activation free energy for strontium catalysis of this reaction is 23.9 kcal/mol, which is 5.8 kcal/mol lower than that for magnesium catalysis, but 5.8 kcal/mol higher than that for calcium catalysis. Therefore, our theoretical calculations predict that the catalytic activities of the Ae metal complexes are calcium > strontium > magnesium.

CONCLUSION
The mechanism of pyridine hydroboration catalyzed by a series of Ae metals has been systematically investigated by DFT calculations. When a magnesium complex is used as the catalyst, the magnesium hydride intermediate is the active catalytic species for this reaction. According to the theoretical calculations, the reaction mechanism of this reaction is as follows. Pyridine first coordinates with the magnesium atom in the active catalyst, followed by the pyridine C=N double bond inserting into the Mg-H bond, resulting in dearomatization. The magnesium amino intermediate coordinates to pinacolborane and the magnesium hydride intermediate is regenerated by hydride transfer to release the product. The catalytic activities of various Ae metals in the same group were predicted by DFT calculations. The trend of the Ae-H BDEs is Mg-H > Ca-H > Sr-H. The theoretical calculations indicate that when a calcium catalyst is used, there is a much lower activation free energy for C=N double bond insertion into the Ca-H bond, which is the rate-determining step of the catalytic cycle. However, because the boron-nitrogen Lewis acid-base complex with the strontium catalyst is very stable, the apparent activation free energy for the strontium catalyst is higher than that for the calcium catalyst, but lower than that for the magnesium catalyst. Therefore, we believe that milder reaction conditions could be used if a calcium or strontium complex is used as the catalyst.

AUTHOR CONTRIBUTIONS
The work was completed by cooperation of all authors. YyL, RB, and YL were responsible for the study of concept and design of the project. MW and HC searched the intermediates and transition states, JZ and DX analyzed the data and drew energy profiles. YyL, LQ, and YL drafted and revised the manuscript.