Mechanistic Insights into Selective Hydrogenation of C=C Bonds Catalyzed by CCC Cobalt Pincer Complexes: A DFT Study

: The mechanistic insights into hydrogenations of hex-5-en-2-one, isoprene, and 4-vinylcyclohex-1-ene catalyzed by pincer ( Mes CCC)Co (Mes = bis(mesityl-benzimidazol-2-ylidene)phenyl) complexes are computationally investigated by using the density functional theory. Different from a previously proposed mechanism with a cobalt dihydrogen complex ( Mes CCC)Co-H 2 as the catalyst, we found that its less stable dihydride isomer, ( Mes CCC)Co(H) 2 , is the real catalyst in those catalytic cycles. The generations of final products with H 2 cleavages for the formations of C − H bonds are the turnover-limiting steps in all three hydrogenation reactions. We found that the hydrogenation selectivity of different C=C bonds in the same compound is dominated by the steric effects, while the hydrogenation selectivity of C=C and C=O bonds in the same compound could be primarily influenced by the electronic effects. In addition, the observed inhabition of the hydrogenation reactions by excessive addition of PPh 3 could be explained by a 15.8 kcal/mol free energy barrier for the dissociation of PPh 3 from the precatalyst.

In 2004, Budzelaar and co-workers [21] developed Co pincer complexes LCoR (L = 2,6-[RN=CMe] 2 C 5 H 3 N; R = n-C 6 H 13 or 2,6-(i-Pr) 2 C 6 H 3 ) for the catalytic hydrogenation of monosubstituted olefins under 50 • C and 20 atm pressure. They found that reducing the steric bulk at the imine positions and changing the metal from cobalt to rhodium did not change catalytic activities much. Such results indicated that cobalt complexes could be "rhodium-like" catalysts with proper ligands. In 2012, Hanson and co-workers [22] reported a versatile Co(II) alkyl complex for the catalytic hydrogenation of olefins, ketones, aldehydes, and imines with yields of up to 90% under mild conditions. They also developed a cationic Co(II) alkyl complex as an effective precatalyst for the dehydrogenation of alcohols and hydrogenation of olefins and ketones with high yields (>95%) under mild conditions (25 • C, 1 atm H 2 ) [23]. Their experimental studies suggested that the olefin hydrogenation reaction underwent an insertion mechanism with a Co(II) hydride complex as the catalyst, while the alcohol dehydrogenation reaction proceeded through a Co(I)/(III) redox catalytic cycle. Later on, Peters and co-workers [24,25] found bis-(phosphino)boryl cobalt complexes as catalysts for C=C bond hydrogenations with high yields (>95%) at room temperature and 1 atm H 2 pressure. Their kinetic studies indicated that the turnoverlimiting step involved a binuclear cobalt complex. In 2015, Chirik and co-workers [26] developed a bis(imino)pyridine cobalt complex for the catalytic hydrogenation of substituted benzofused five-, six-, and seven-membered alkenes with high yields (>95%) and enantioselectivities (>95% ee) under mild conditions (25 • C and 4 atm H 2 ). They also found that both the ring size and exo/endo disposition affected the stereochemistry.
Fout and co-workers [27] recently developed a series of cobalt catalysts with electronrich monoanionic bis(carbene) ligands, Mes CCC (bis(mesityl-benzimidazol-2-ylidene)phenyl) and DIPP CCC (bis(diisopropylphenyl-benzimidazol-2-ylidene)phenyl), for the rapid and highly chemoselective hydrogenation of olefins (Scheme 1). They found the viability of Co I /Co III redox cycles in such olefin hydrogenation reactions and proposed a plausible mechanism with a cobalt dihydrogen complex ( Mes CCC)Co-H 2 as the catalyst (Scheme 2) [27]. In the proposed mechanism, a reactant molecule fills the vacant position in 4 and forms intermediate 5. The oxidative addition of H 2 in 5 forms a dihydride complex 6. When another H 2 approaches 6, 8 and 8 could be formed through H 2 cleavage and C−H bond formation. Then, the product is formed and the catalyst 4 is regenerated with the formation of another C−H bond. Their following experimental studies indicated the existence of cobalt-alkyl hydride complex generating (5 → 6) and β-H elimination (8 → 6) steps [28]. They also extended the application of ( Mes CCC)Co complexes for the catalytic semihydrogenation of alkynes and found the generation of E-selective products from a wide range of alkynes with yields of up to 80% and a E/Z selectivity over 99% [29]. Furthermore, Fout and co-workers [30] studied the electronic modification effect of Mes CCC R pincer ligands and found that the tert-butyl group did not affect the reactivity, while the CF 3 group changed the product ratios. Although a plausible catalytic cycle has been proposed, the mechanistic insights into the above ( Mes CCC)Co catalyzed hydrogenation reactions, especially the origin of high chemo-selectivities, still remain unclear. Herein, we computationally investigated detailed mechanisms of the hydrogenation reactions of hex-5-en-2-one, isoprene, and 4-vinylcyclohex-1-ene catalyzed by ( Mes CCC)Co using the density functional theory (DFT), analyzed the causes of selectivities in hydrogenations of C=C and C=O bonds, as well as the C=C bonds in the same compounds, and explained why the addition of excessive PPh 3 ligands inhibited the reactions.
Catalysts 2021, 11, x FOR PEER REVIEW 2 of 16 "rhodium-like" catalysts with proper ligands. In 2012, Hanson and co-workers [22] reported a versatile Co(II) alkyl complex for the catalytic hydrogenation of olefins, ketones, aldehydes, and imines with yields of up to 90% under mild conditions. They also developed a cationic Co(II) alkyl complex as an effective precatalyst for the dehydrogenation of alcohols and hydrogenation of olefins and ketones with high yields (>95%) under mild conditions (25 °C, 1 atm H2) [23]. Their experimental studies suggested that the olefin hydrogenation reaction underwent an insertion mechanism with a Co(II) hydride complex as the catalyst, while the alcohol dehydrogenation reaction proceeded through a Co(I)/(III) redox catalytic cycle. Later on, Peters and co-workers [24,25] found bis-(phosphino)boryl cobalt complexes as catalysts for C=C bond hydrogenations with high yields (>95%) at room temperature and 1 atm H2 pressure. Their kinetic studies indicated that the turnover-limiting step involved a binuclear cobalt complex. In 2015, Chirik and co-workers [26] developed a bis(imino)pyridine cobalt complex for the catalytic hydrogenation of substituted benzofused five-, six-, and seven-membered alkenes with high yields (>95%) and enantioselectivities (>95% ee) under mild conditions (25 °C and 4 atm H2). They also found that both the ring size and exo/endo disposition affected the stereochemistry. Fout and co-workers [27] recently developed a series of cobalt catalysts with electronrich monoanionic bis(carbene) ligands, Mes CCC (bis(mesityl-benzimidazol-2-ylidene)phenyl) and DIPP CCC (bis(diisopropylphenyl-benzimidazol-2-ylidene)phenyl), for the rapid and highly chemoselective hydrogenation of olefins (Scheme 1). They found the viability of Co I /Co III redox cycles in such olefin hydrogenation reactions and proposed a plausible mechanism with a cobalt dihydrogen complex ( Mes CCC)Co-H2 as the catalyst (Scheme 2) [27]. In the proposed mechanism, a reactant molecule fills the vacant position in 4 and forms intermediate 5. The oxidative addition of H2 in 5 forms a dihydride complex 6. When another H2 approaches 6, 8 and 8′ could be formed through H2 cleavage and C−H bond formation. Then, the product is formed and the catalyst 4 is regenerated with the formation of another C−H bond. Their following experimental studies indicated the existence of cobalt-alkyl hydride complex generating (5 → 6) and β-H elimination (8 → 6) steps [28]. They also extended the application of ( Mes CCC)Co complexes for the catalytic semihydrogenation of alkynes and found the generation of E-selective products from a wide range of alkynes with yields of up to 80% and a E/Z selectivity over 99% [29]. Furthermore, Fout and co-workers [30] studied the electronic modification effect of Mes CCC R pincer ligands and found that the tert-butyl group did not affect the reactivity, while the CF3 group changed the product ratios. Although a plausible catalytic cycle has been proposed, the mechanistic insights into the above ( Mes CCC)Co catalyzed hydrogenation reactions, especially the origin of high chemo-selectivities, still remain unclear. Herein, we computationally investigated detailed mechanisms of the hydrogenation reactions of hex-5-en-2-one, isoprene, and 4-vinylcyclohex-1-ene catalyzed by ( Mes CCC)Co using the density functional theory (DFT), analyzed the causes of selectivities in hydrogenations of C=C and C=O bonds, as well as the C=C bonds in the same compounds, and explained why the addition of excessive PPh3 ligands inhibited the reactions.

Hydrogenation of C=C Bond
The predicted catalytic cycle and the corresponding free energy profile for the hydrogenation of hex-5-en-2-one to hexan-2-one are shown in Scheme 3 and Figure 1, respectively. The optimized structures of key transition states in this reaction are displayed in Figure 2.  The predicted catalytic cycle and the corresponding free energy profile for the hydrogenation of hex-5-en-2-one to hexan-2-one are shown in Scheme 3 and Figure 1, respectively. The optimized structures of key transition states in this reaction are displayed in Figure 2.

Hydrogenation of C=C Bond
The predicted catalytic cycle and the corresponding free energy profile for the hydrogenation of hex-5-en-2-one to hexan-2-one are shown in Scheme 3 and Figure 1, respectively. The optimized structures of key transition states in this reaction are displayed in Figure 2.  At the beginning of the reaction, a H2 molecule replaces the N2 in 1 and forms an 8.0kcal/mol more stable intermediate 3. The dissociation of PPh3 from 3 is a 15.8-kcal/mol uphill step. The dihydrogen complex ( Mes CCC)Co-H2 (4) was considered as the catalyst in a previous study [27]. Once 4 is formed, a hex-5-en-2-one molecule can easily coordinate to 4 with its C=C bond and form a 5.  At the beginning of the reaction, a H2 molecule replaces the N2 in 1 and forms an 8.0kcal/mol more stable intermediate 3. The dissociation of PPh3 from 3 is a 15.8-kcal/mol uphill step. The dihydrogen complex ( Mes CCC)Co-H2 (4) was considered as the catalyst in a previous study [27]. Once 4 is formed, a hex-5-en-2-one molecule can easily coordinate to 4 with its C=C bond and form a 5.  At the beginning of the reaction, a H 2 molecule replaces the N 2 in 1 and forms an 8.0-kcal/mol more stable intermediate 3. The dissociation of PPh 3 from 3 is a 15.8-kcal/mol uphill step. The dihydrogen complex ( Mes CCC)Co-H 2 (4) was considered as the catalyst in a previous study [27]. Once 4 is formed, a hex-5-en-2-one molecule can easily coordinate to 4 with its C=C bond and form a 5.1-kcal/mol more stable intermediate 5. The oxidative addition of H 2 in 5 has a very low barrier of 1.9 kcal/mol (TS 5,6 ). Such a low barrier indicates that the transformation between 5 and 6 is reversible, which corresponds to the experiments. Then, a hydride in 6 can easily transfer from cobalt to the end carbon atom in the coordinated C=C bond via TS 6,7 and form a more stable intermediate 7. Another H 2 molecule can coordinate to 7 and form a 5.5-kcal/mol less stable complex 8.
There are two ways for the cleavage of H 2 in 8 to occur. One is a proton transfer from H 2 to the carbon bonding to Co via TS 8,4 with a free energy barrier of 19.2 kcal/mol. A stable dihydride complex 4 is formed with the dissociation of hex-2-one. 4 could attract a hex-5-en-2-one molecule and complete a catalytic cycle with the formation of 6. The other way for H 2 cleavage to occur is a proton transfer from H 2 to the hydride bonding to Co for the formation of 8 , which is an 11.1-kcal/mol less stable isomer of 8 with rearranged hydrogen atoms. Then, a hex-2-one molecule is formed through reductive elimination (TS 8 ,4 ), which is 9.1-kcal/mol higher than TS 8,4 in free energy and unlikely to happen in the reaction. Therefore, the dihydride complex 4 is believed to be the more reasonable catalyst for the hydrogenation of hex-5-en-2-one, with a total free energy barrier of 19.2 kcal/mol (3 → TS 8,4 ). It is worth noting that the slightly lower free energy of TS 8,8 than 8 is caused by thermal corrections. We can consider that this does not practically exist at the experimental temperature.

Hydrogenation of C=O Bond
The reaction cycle and corresponding free energy profile for the hydrogenation of hex-5-en-2-one to hex-5-en-2-ol are shown in Scheme 4 and Figure 3, respectively. The optimized structures of key intermediates and transition states in this cycle are displayed in Figure 4. After the formation of 4, the coordination of the C=O bond in hex-5-en-2-one to Co forms a 7.1-kcal/mol less stable intermediate 9. We believe that this is primarily caused by the methyl and butene groups on carbonyl, which prevent the end-on bonding of carbonyl to Co and make the coordination of C=O much weaker than the Dewar-Chatt-Duncanson (DCD) model bonding between C=C and Co. In addition, the methyl group on carbonyl also increases the difficulty of C=O bonding to Co. After the formation of 9, the oxygen atom in hex-5-en-2-one could assist H 2 splitting for the formation of an O−H bond in 10 with a free energy barrier of 41.3 kcal mol −1 (TS 9,10 ). Then, the complex 11 is formed with the coordination of another H 2 molecule. Like 8 in the C=C bond hydrogenation mechanism shown in Scheme 3, 11 is the bifurcating point in the reaction cycle for C=O bond hydrogenation. The free energy profile in Figure 3 indicates that the formation of hex-5-en-2-ol by simultaneous H 2 cleavage and C−H bond formation via TS 11,4 is 8.8 kcal/mol lower than TS 11 ,4 . Therefore, TS 11,4 is the rate-determining step in the reaction with a total free energy barrier of 44.1 kcal/mol (3 → TS 11,4 ) for the formation of hex-5-en-2-ol. Such a high barrier indicates that ( Mes CCC)Co cannot catalyze the hydrogenation of the C=O bond in hex-5-en-2-one. The oxygen atom in TS 11,4 decreases the electron density of carbon atom bonding to Co. The low electron density of the carbon atom makes it hard for the TS 11,4 step to happen. Such a reason may explain the selectivity of the hydrogenation of C=C and C=O bonds in a compound.          The predicted catalytic cycle and the corresponding free energy profile for the hydrogenation of isoprene to 2-methylbut-1-ene are shown in Scheme 5 and Figure 5, respectively. The optimized structures of key intermediates and transition states in this reaction are displayed in Figure 6. The predicted catalytic cycle and the corresponding free energy profile for the hydrogenation of isoprene to 2-methylbut-1-ene are shown in Scheme 5 and Figure 5, respectively. The optimized structures of key intermediates and transition states in this reaction are displayed in Figure 6.   The predicted catalytic cycle and the corresponding free energy profile for the hydrogenation of isoprene to 2-methylbut-1-ene are shown in Scheme 5 and Figure 5, respectively. The optimized structures of key intermediates and transition states in this reaction are displayed in Figure 6.  Once 4 is formed, a singly substituted C=C bond in an isoprene molecule can coordinate to 4 and form a 2.8-kcal/mol more stable intermediate 12 with a DCD model bonding. The oxidative addition of H2 in 12 for the formation of the dihydride complex 13 has a rather low barrier of 2.7 kcal/mol (TS12,13). Such a low barrier indicates that the transformation between 12 and 13 is reversible, which corresponds to the experiments. After the H2 cleavage, a hydrogen can easily transfer from cobalt to the end carbon atom in the coordinated C=C bond via TS13,14 and form the intermediate 14, which is 9.6 kcal/mol more stable than 4. Then, another H2 molecule comes in for the formation of 15, which is the bifurcating point in this catalytic reaction. The formation of 2-methylbut-1-ene with the transfer of a proton from H2 to the coordinated carbon atom via TS15,4′ is 9.2-kcal/mol more favorable than TS15′,4. Therefore, we believe that TS15,4′ is the rate-determining step for the hydrogenation of isoprene with a total free energy barrier of 21.3 kcal/mol (3 → TS15,4′). Once 4 is formed, a singly substituted C=C bond in an isoprene molecule can coordinate to 4 and form a 2.8-kcal/mol more stable intermediate 12 with a DCD model bonding. The oxidative addition of H 2 in 12 for the formation of the dihydride complex 13 has a rather low barrier of 2.7 kcal/mol (TS 12,13 ). Such a low barrier indicates that the transformation between 12 and 13 is reversible, which corresponds to the experiments. After the H 2 cleavage, a hydrogen can easily transfer from cobalt to the end carbon atom in the coordinated C=C bond via TS 13,14 and form the intermediate 14, which is 9.6 kcal/mol more stable than 4. Then, another H 2 molecule comes in for the formation of 15, which is the bifurcating point in this catalytic reaction. The formation of 2-methylbut-1-ene with the transfer of a proton from H 2 to the coordinated carbon atom via TS 15,4 is 9.2-kcal/mol more favorable than TS 15 ,4 . Therefore, we believe that TS 15,4 is the rate-determining step for the hydrogenation of isoprene with a total free energy barrier of 21.3 kcal/mol (3 → TS 15,4 ).

Hydrogenation of the Doubly Substituted C=C Double Bond
In order to find out the key factors that influence the selectivity of different C=C bonds, we also studied the mechanism for the hydrogenation of the doubly substituted C=C bond in isoprene. The reaction cycle and corresponding free energy profile for the hydrogenation of isoprene to 3-methylbut-1-ene are shown in Scheme 6 and Figure 7, respectively. The optimized structures of key intermediates and transition states in this cycle are displayed in Figure 8.

Hydrogenation of the Doubly Substituted C=C Double Bond
In order to find out the key factors that influence the selectivity of different C=C bonds, we also studied the mechanism for the hydrogenation of the doubly substituted C=C bond in isoprene. The reaction cycle and corresponding free energy profile for the hydrogenation of isoprene to 3-methylbut-1-ene are shown in Scheme 6 and Figure 7, respectively. The optimized structures of key intermediates and transition states in this cycle are displayed in Figure 8.   In order to find out the key factors that influence the selectivity of different C=C bonds, we also studied the mechanism for the hydrogenation of the doubly substituted C=C bond in isoprene. The reaction cycle and corresponding free energy profile for the hydrogenation of isoprene to 3-methylbut-1-ene are shown in Scheme 6 and Figure 7, respectively. The optimized structures of key intermediates and transition states in this cycle are displayed in Figure 8.  The formation of 3-methylbut-1-ene has a similar pathway as the hydrogenations of hex-5-en-2-one and isoprene, but slightly different relative free energies. Because of the bulky structures of mesitylene groups in the CCC ligand, the coordination of the doubly substituted C=C bond to Co is 1.1-kcal/mol less favorable than the coordination of the singly substituted C=C bond in isoprene. Complex 19 is the bifurcating point in this reaction, while the H2 cleavage for the formation of 3-methylbut-1-ene is the turnover-limiting step with a total free energy barrier of 24.6 kcal/mol (3 → TS19,4′), which is 3.3 kcal/mol higher than the barrier for the formation of 2-methylbut-1-ene ( Figure 5). Such a high barrier indicates that ( Mes CCC)Co cannot catalyze the hydrogenation of the doubly substituted C=C bond in isoprene because the steric effect between the doubly substituted C=C bond and ( Mes CCC)Co is larger than that of singly substituted C=C bonds. The distance between C(CH3)2-CH=CH2 and H2 bonding to Co in TS19,4′ is larger than that between CH(CH3)-C(CH3)=CH2 and H2 bonding to Co in TS15,4′; the large distance makes it hard for the TS19,4′ step to happen. Such a reason may explain the selectivity of the hydrogenation.

Hydrogenation of the Exocycle C=C Bond
In order to find out the key factors that influence the selectivity in the hydrogenation of the C=C bond in cycloalkene derivatives, we further explored the mechanism for the hydrogenation of 4-vinylcyclohex-1-ene to 4-ethylcyclohex-1-ene. The predicted catalytic cycle and the corresponding free energy profile are shown in Scheme 7 and Figure 9, respectively. The optimized structures of key intermediates and transition states are displayed in Figure 10. The formation of 3-methylbut-1-ene has a similar pathway as the hydrogenations of hex-5-en-2-one and isoprene, but slightly different relative free energies. Because of the bulky structures of mesitylene groups in the CCC ligand, the coordination of the doubly substituted C=C bond to Co is 1.1-kcal/mol less favorable than the coordination of the singly substituted C=C bond in isoprene. Complex 19 is the bifurcating point in this reaction, while the H 2 cleavage for the formation of 3-methylbut-1-ene is the turnoverlimiting step with a total free energy barrier of 24.6 kcal/mol (3 → TS 19,4 ), which is 3.3 kcal/mol higher than the barrier for the formation of 2-methylbut-1-ene ( Figure 5). Such a high barrier indicates that ( Mes CCC)Co cannot catalyze the hydrogenation of the doubly substituted C=C bond in isoprene because the steric effect between the doubly substituted C=C bond and ( Mes CCC)Co is larger than that of singly substituted C=C bonds. The distance between C(CH 3 ) 2 -CH=CH 2 and H 2 bonding to Co in TS 19,4 is larger than that between CH(CH 3 )-C(CH 3 )=CH 2 and H 2 bonding to Co in TS 15,4 ; the large distance makes it hard for the TS 19,4 step to happen. Such a reason may explain the selectivity of the hydrogenation.

Hydrogenation of the Exocycle C=C Bond
In order to find out the key factors that influence the selectivity in the hydrogenation of the C=C bond in cycloalkene derivatives, we further explored the mechanism for the hydrogenation of 4-vinylcyclohex-1-ene to 4-ethylcyclohex-1-ene. The predicted catalytic cycle and the corresponding free energy profile are shown in Scheme 7 and Figure 9, respectively. The optimized structures of key intermediates and transition states are displayed in Figure 10.   After the formation of 4, a 4-vinylcyclohex-1-ene molecule coordinates to 4 with its C=C double bond at the exocycle and forms a 7.5-kcal/mol more stable intermediate 20.
The following H2 cleavage and hydrogenation process for the formation of 4-ethylcyclohex-1-ene are similar to the above pathways in the hydrogenations of hex-5-en-2-one and isoprene. 22 is the bifurcating point of the reaction, like 8 in the hydrogenation of the C=C bond in hex-5-en-2-one. The calculated free energy profile indicates that the formation of 4-ethylcyclohex-1-ene with the cleavage of H2 (TS22,4′) is the turnover-limiting step in this reaction. 4′ is believed to be the more reasonable catalyst for the hydrogenation of 4-vinylcyclohex-1-ene with a total free energy barrier of 21.4 kcal/mol (3 → TS22,4′).

Hydrogenation of the C=C Bond in Cycle
The reaction cycle and corresponding free energy profile for the hydrogenation of 4vinylcyclohex-1-ene to vinylcyclohexane are shown in Scheme 8 and Figure 11, respectively. The optimized structures of key intermediates and transition states are displayed in Figure 12. After the formation of 4, a 4-vinylcyclohex-1-ene molecule coordinates to 4 with its C=C double bond at the exocycle and forms a 7.5-kcal/mol more stable intermediate 20.
The following H 2 cleavage and hydrogenation process for the formation of 4-ethylcyclohex-1-ene are similar to the above pathways in the hydrogenations of hex-5-en-2-one and isoprene. 22 is the bifurcating point of the reaction, like 8 in the hydrogenation of the C=C bond in hex-5-en-2-one. The calculated free energy profile indicates that the formation of 4-ethylcyclohex-1-ene with the cleavage of H 2 (TS 22,4 ) is the turnover-limiting step in this reaction. 4 is believed to be the more reasonable catalyst for the hydrogenation of 4-vinylcyclohex-1-ene with a total free energy barrier of 21.4 kcal/mol (3 → TS 22,4 ).

Hydrogenation of the C=C Bond in Cycle
The reaction cycle and corresponding free energy profile for the hydrogenation of 4-vinylcyclohex-1-ene to vinylcyclohexane are shown in Scheme 8 and Figure 11, respectively. The optimized structures of key intermediates and transition states are displayed in Figure 12.
Because of the bulky structures of cyclohexene in the 4-vinylcyclohex-1-ene and mesitylene groups in the CCC ligand, the coordination of the C=C bond in cyclohexene to Co forms a 4.7-kcal/mol less stable intermediate 23.
Although the formations of vinylcyclohexane and 4-ethylcyclohex-1-ene have similar reaction pathways, their energy profiles are different. As shown in Figure 11, the turnover-limiting step TS 26,4 has a total free energy barrier of 29.9 kcal/mol (3 → TS 26,4 ), which is 8.5 kcal/mol higher than the barrier for the formation of 4-ethylcyclohex-1-ene. Such a high barrier indicates that ( Mes CCC)Co cannot catalyze the hydrogenation of the C=C bond in the cyclohexene of 4-vinylcyclohex-1-ene because the steric effect between the C=C bond in cyclohexene and ( Mes CCC)Co is larger than that of the exocycle C=C bond. The distance between the C atom in cyclohexene and H 2 bonding to Co in TS 26,4 is larger than that between the C atom in chain and H 2 bonding to Co in TS 22,4 ; the large distance makes it hard for the TS 26,4 step to happen. Such a reason may explain the selectivity of the hydrogenation.   Because of the bulky structures of cyclohexene in the 4-vinylcyclohex-1-ene and mesitylene groups in the CCC ligand, the coordination of the C=C bond in cyclohexene to Co forms a 4.7-kcal/mol less stable intermediate 23.
Although the formations of vinylcyclohexane and 4-ethylcyclohex-1-ene have similar reaction pathways, their energy profiles are different. As shown in Figure 11, the turnover-limiting step TS26,4′ has a total free energy barrier of 29.9 kcal/mol (3 → TS26,4′), which is 8.5 kcal/mol higher than the barrier for the formation of 4-ethylcyclohex-1-ene. Such a high barrier indicates that ( Mes CCC)Co cannot catalyze the hydrogenation of the C=C bond in the cyclohexene of 4-vinylcyclohex-1ene because the steric effect between the C=C bond in cyclohexene and ( Mes CCC)Co is larger than that of the exocycle C=C bond. The distance between the C atom in cyclohexene and H2 bonding to Co in TS26,4′ is larger than that between the C atom in chain and H2 bonding to Co in TS22,4′; the large distance makes it hard for the TS26,4′ step to happen. Such a reason may explain the selectivity of the hydrogenation.

Computational Details
All DFT calculations in this study were executed using the Gaussian 09 programs package [31] for the ωB97X-D functional [32]. The all-electron 6-31G(d,p) basis set was used for H, C, O, N, and P atoms [33,34], while the Stuttgart relativistic effective core potential basis set (ECP10MDF) was used for Co [35]. All structures were optimized with solvent effect corrections using the integral equation formalism polarizable continuum model (IEFPCM) [36] with the SMD (solvation model based on the quantum mechanical charge density) [37] variation for benzene. Thermal corrections were calculated within the harmonic potential approximation under T = 298.15 K and 1 atm pressure. The number of imaginary frequencies (IFs) obtained from frequency calculations confirmed the nature of all intermediates (no IF) and transition states (only one IF). All transition states were confirmed to connect corresponding reactants and products by intrinsic reaction coordinate calculations. The 3D molecular structures shown in this paper were drawn using the JIMP2 molecular visualizing and manipulating program [38]. We also evaluated the reliability of the ωB97X-D functional for this cobalt catalytic system, as well as the spin states of the structures in the reaction coordinates. The results are provided in the Supplementary Materials as Table S1−S4.

Conclusions
In summary, our DFT study of the mechanistic insights into the hydrogenations of C=C bonds in hex-5-en-2-one, isoprene, and 4-vinylcyclohex-1-ene catalyzed by ( Mes-CCC)Co complexes reveals that a Co dihydride complex 4′ is the real catalyst in the catalytic cycles. In all three hydrogenation reactions, the complex 3 with a PPh3 ligand coordinated to Co is the resting state. The H2 cleavages for the formations of C−H bonds in the

Computational Details
All DFT calculations in this study were executed using the Gaussian 09 programs package [31] for the ωB97X-D functional [32]. The all-electron 6-31G(d,p) basis set was used for H, C, O, N, and P atoms [33,34], while the Stuttgart relativistic effective core potential basis set (ECP10MDF) was used for Co [35]. All structures were optimized with solvent effect corrections using the integral equation formalism polarizable continuum model (IEFPCM) [36] with the SMD (solvation model based on the quantum mechanical charge density) [37] variation for benzene. Thermal corrections were calculated within the harmonic potential approximation under T = 298.15 K and 1 atm pressure. The number of imaginary frequencies (IFs) obtained from frequency calculations confirmed the nature of all intermediates (no IF) and transition states (only one IF). All transition states were confirmed to connect corresponding reactants and products by intrinsic reaction coordinate calculations. The 3D molecular structures shown in this paper were drawn using the JIMP2 molecular visualizing and manipulating program [38]. We also evaluated the reliability of the ωB97X-D functional for this cobalt catalytic system, as well as the spin states of the structures in the reaction coordinates. The results are provided in the Supplementary Materials as Tables S1−S4.

Conclusions
In summary, our DFT study of the mechanistic insights into the hydrogenations of C=C bonds in hex-5-en-2-one, isoprene, and 4-vinylcyclohex-1-ene catalyzed by ( Mes CCC)Co complexes reveals that a Co dihydride complex 4 is the real catalyst in the catalytic cycles. In all three hydrogenation reactions, the complex 3 with a PPh 3 ligand coordinated to Co is the resting state. The H 2 cleavages for the formations of C−H bonds in the final products are the turnover-limiting steps, with total free energy barriers of 19.2 (3 → TS 8,4 ), 21.3 (3 → TS 15,4 ), and 21.4 kcal/mol (3 → TS 22,4 ) in the hydrogenations of hex-5-en-2-one, isoprene, and 4-vinylcyclohex-1-ene, respectively. Our calculation results also indicate that the hydrogenation selectivity of different C=C bonds is dominated by the steric effect, while the hydrogenation selectivity of C=C and C=O bonds in the same compound could primarily be influenced by the electronic effect. In addition, the observed inhibition of the hydrogenation reactions by the excessive addition of PPh 3 could be explained by a free energy barrier of 15.8 kcal/mol for the dissociation of PPh 3 from 3.
Author Contributions: Conceptualization, Z.Z. and X.Y.; Methodology, Z.Z.; Writing-original draft, Z.Z.; Writing-review and editing, X.Y. All authors have read and agreed to the published version of the manuscript.
Funding: This work is supported by the National Natural Science Foundation of China (21873107, 21703256).