Calcium Hydride Cation Dimer Catalyzed Hydrogenation of Unactivated 1‐Alkenes and H2 Isotope Exchange: Competitive Ca−H−Ca Bridges and Terminal Ca−H Bonds

Abstract Recently, it was shown that the double Ca−H−Ca bridged calcium hydride cation dimer complex [LCaH2CaL]2+ (macrocyclic ligand L=NNNN‐tetradentate Me4TACD) exhibited remarkable activity in catalyzing the hydrogenation of unactivated 1‐alkenes as well as the H2 isotope exchange under mild conditions, tentatively via the terminal Ca−H bond of cation monomer LCaH+. In this DFT mechanistic work, a novel substrate‐dependent catalytic mechanism is disclosed involving cooperative Ca−H−Ca bridges for H2 isotope exchange, competitive Ca−H−Ca bridges and terminal Ca−H bonds for anti‐Markovnikov addition of unactivated 1‐alkenes, and terminal Ca−H bonds for Markovnikov addition of conjugation‐activated styrene. THF‐coordination plays a key role in favoring the anti‐Markovnikov addition while strong cation‐π interactions direct the Markovnikov addition to terminal Ca−H bonds.


Results and Discussion
To gain deep mechanistic insight into the cation dimer 1 2 + ·THF catalyzed hydrogenation of 1-alkenes and H 2 isotope exchange reactions, state-of-the-art dispersion-corrected DFT calculations are performed at the PW6B95-D3/def2-QZVP + COSMO-RS// TPSS-D3/def2-TZVP + COSMO level in THF solution (see below for computational details), and final free energies (at 298.15 K and 1 M concentration) are used in our discussion unless specified otherwise. In THF solution, the weakly (or non-) coordinating counter-anions BAr 4 À should be solvated into separated ions, and thus not considered further. Consistent with experiment, [7] our DFT calculations show that the THF coordination to 1 2 + is indeed À 2.8 kcal/mol exergonic, with a decisive stabilizing dispersion contribution of À 7.6 kcal/mol otherwise the coordination should be 4.8 kcal/mol endergonic and thus thermodynamically unstable. For comparison, the THF coordination to the cation monomer 1 m + is only À 1.5 kcal/mol exergonic with a smaller dispersion contribution of À 4.2 kcal/ mol. The formation of double CaÀ HÀ Ca bridged 1 2 + from two 1 m + monomers is À 18.4 kcal/mol exergonic again with a sizable dispersion contribution of À 6.2 kcal/mol. The THF coordination to 1 2 + ·THF smoothly leads to two 1 m + ·THF cation monomers that is 15.3 kcal/mol endergonic, indicating a rapid dimer-to-monomer equilibrium even at room temperature (see Supporting Information Table S1).
In two related mechanistic studies on 1-alkene hydrogenation reactions, the THF-coordinated [(THF)(BDI)CaH] 2 [9] and the non-THF-coordinated [(BDI)CaH 2 Ca(BDI)] [6] calcium hydride dimers were used as catalyst. Dispersion-uncorrected DFT calculations suggested that the dimer [(THF)(BDI)CaH] 2 is about 7.7 kcal/mol higher in free energy than two (THF)(BDI)CaH monomers in benzene solution, [9] in contrast to previous experiment [3b] and our dispersion-corrected DFT results (À 18.2 kcal/mol lower than two monomers). Moreover, dispersion-uncorrected DFT calculations suggested that the dimer [(BDI)CaH 2 Ca(BDI)] is 40.4 kcal/mol lower in enthalpy than two (BDI)CaH monomers in gas-phase, [6] which however should be further enhanced by 17.5 kcal/mol and decreased by 26.8 kcal/ mol due to dispersion interactions and solvation in benzene, respectively, according to our dispersion-corrected DFT calculations. It is thus crucial to include both dispersion corrections and suitable solvation model in modeling such catalytic reactions in solution.

Chemistry-A European Journal
Research Article doi.org/10.1002/chem.202202602 (alkyl R' = CH 2 CH 2 R) bridge. The alternative Markovnikov addition with the hydride added to the terminal alkene carbon is kinetically 7.6 kcal/mol less favorable (via TS1a 2 + , see Supporting Information Figure S1 and Table S1). The subsequent H 2 addition over the CaÀ R'À Ca bridge of A 2 + is highly exergonic and kinetically 5.8 kcal/mol more favorable (via TS2 2 + ) to form the hydrogenated product CH 3 CH 2 R along with regenerated 1 2 + ·THF after THF coordination. Such a catalytic 1-alkene hydrogenation is thus À 23.7 kcal/mol exergonic over a sizable barrier of 23.8 kcal/mol involving anti-Markovnikov alkene addition to cooperative CaÀ HÀ Ca bridges.
On the other hand, via the THF-coordinated cation monomer 1 m + ·THF that is 7.8 kcal/mol higher in free energy than 1 2 + ·THF, the anti-Markovnikov addition of unactivated CH 2 = CHR to the terminal CaÀ H bond is still 4.0 kcal/mol endergonic but over nearly the same free energy barrier of 23.8 kcal/mol (via mTS1 + ) to selectively form the calcium alkyl complex mA + [LCaHR'·THF] + , with the alternative Markovnikov addition being kinetically 6.0 kcal/mol less favorable (via mTS1a + , see Supporting Information Figure S1 and Table S1). The subsequent hydrogenolysis of the CaÀ C bond of mA + with H 2 is highly exergonic and kinetically 2.9 kcal/mol more favorable (via mTS2 + ) to form the hydrogenated product CH 3 CH 2 R along with regenerated 1 m + ·THF. When cyclohexene is used as an internal alkene substrate in 1 m + ·THF catalyzed hydrogenation, a 3.3 kcal/mol higher barrier (via cTS1 + , see Supporting Information Table S1) is found for the anti-Markovnikov addition to the terminal CaÀ H bond, consistent with the experimentally observed selective hydrogenation of terminal alkene. [8] Interestingly, after THF elimination, the terminal CaÀ H bond of solvent-free 1 m + is kinetically 0.7 kcal/mol less reactive (via mTS10 + , see Supporting Information Figure S1 and Table S1) than that of 1 m + ·THF towards CH 2 = CHR addition, mainly due to a less Lewis-basic hydride (Mulliken atomic charges: À 0.16 e in 1 m + vs. À 0.53 e in 1 m + ·THF). Together with the positive THF affinity of 1 m + , this leads to a 2.1 kcal/mol higher barrier of 25.9 kcal/mol for CH 2 = CHR addition without the coordinating THF. Interestingly, despite nearly the same free energy barriers (23.8 kcal/mol) at 298 K for the competitive mechanisms via TS1 2 + and mTS1 + , our DFT calculations show that the former via the solvent-free dimer 1 2 + becomes kinetically 1.4 kcal/mol more favorable than the latter via the THFcoordinated 1 m + ·THF (barriers: 24.2 versus 25.6 kcal/mol) upon heating at 60°C (333 K) in experiment, [8] mainly due to entropyfavored THF-elimination from 1 2 + ·THF to reach the solvent-free TS1 2 + . In contrary to the usual intuition that the dimer-tomonomer conversion should be favored upon heating due to favorable entropy effects, our DFT calculations show that the equilibrium of 1 2 + ·THF + THF!1 m + ·THF + 1 m + ·THF in solution is hardly affected by such temperature change due to negligible entropy difference with an unchanged number of molecules on both sides (see Supporting Information Table S1).
As seen in Figure 2, very facile HÀ HÀ H type H 2 isotope exchange may occur via one of two cooperative CaÀ HÀ Ca bridges of the solvent-free cation dimer 1 2 + over a low barrier of 16.9 kcal/mol (via TS3 2 + ) after THF-elimination from stable 1 2 + ·THF catalyst. Interestingly, three exchanging hydrogen atoms are placed evenly between two calcium ions and perpendicular to the remaining CaÀ HÀ Ca bridge, suggesting potentially strong cooperative effects. With the coordinating THF within 1 2 + ·THF, similar H 2 isotope exchange (via TS3a 2 + ) becomes kinetically 4.5 kcal/mol less favorable. On the other hand, facile H 2 isotope exchange may also occur via terminal CaÀ H bonds that are more reactive but higher in free energy than CaÀ HÀ Ca bridges. Indeed, the THF-coordinated cation monomer 1 m + ·THF is intrinsically 3.9 kcal/mol (via mTS3 + ) more reactive than 1 2 + for H 2 isotope exchange but 7.8 kcal/ mol higher in free energy, eventually leading to a 3.9 kcal/mol higher barrier. Interestingly, the solvent-free cation monomer 1 m + with a less Lewis-basic hydride turns out to be intrinsically 3.9 kcal/mol less reactive than 1 m + ·THF in mediating H 2 isotope exchange, which is further disfavored by its positive THF affinity of 1.4 kcal/mol. It is thus clear that very efficient H 2 isotope exchange observed at room temperature [8] is actually catalyzed by cooperative CaÀ HÀ Ca bridges of solvent-free 1 2 + over a low barrier of 16.9 kcal/mol (via TS3 2 + ) rather than recently proposed terminal CaÀ H bonds.
As shown in Figure 3, starting from the stable cation dimer catalyst 1 2 + ·THF, the Markovnikov addition of styrene (CH 2 = CHPh) as a typical π-conjugation-activated 1-alkene to a cooperative CaÀ HÀ Ca bridge of 1 2 + after THF-elimination is 0.2 kcal/mol endergonic over a sizable barrier of 24.3 kcal/mol (via TS4 2 + ), selectively leading to the dinuclear calcium-complex B 2 + [LCaHR"CaL] 2 + containing both a CaÀ HÀ Ca and a CaÀ R"À Ca (alkyl R" = CH(CH 3 )Ph) bridge. Note that the hydride is now added to the terminal rather than the inner alkene carbon as directed by strong Ca + …Ph cation-π interactions; the anti-Markovnikov addition encounters a 2.7 kcal/mol higher barrier (via TS4a 2 + ; see Supporting Information Figure S2 and Table S1) and thus is kinetically disfavored. The subsequent hydrogenolysis of the CaÀ R"À Ca bridge of B 2 + with H 2 is highly exergonic and kinetically 7.7 kcal/mol more favorable (via TS5 2 + ) to form the hydrogenated product CH 3 CH 2 Ph along with regenerated catalyst 1 2 + ·THF after THF-coordination. Such dimeric catalytic
Interestingly, starting from the stable complex 1 2 + ·THF, the Markovnikov styrene addition to the terminal CaÀ H bond of the solvent-free cation monomer 1 m + (via mTS4 + ) is now kinetically very efficient and again directed by strong Ca + …Ph ion-π interaction, which is now À 7.7 kcal/mol exergonic over a moderate barrier of 20.6 kcal/mol to selectively form the calcium benzyl complex mB + LCaR" + . The alternative anti-Markovnikov styrene addition is 6.4 kcal/mol less favorable (via mTS4a + , see Supporting Information Figure S2 and Table S1) and is kinetically highly disfavored. The subsequent H 2 hydrogenolysis of the CaÀ C bond of mB + is À 14.6 kcal/mol exergonic over a moderate barrier of 19.0 kcal/mol (via mTS5 + ) to form the hydrogenated product CH 3 CH 2 Ph along with regenerated catalyst 1 2 + ·THF after 1 m + dimerization and THF-coordination, which is kinetically 1.6 kcal/mol more favorable than the preceding Markovnikov styrene addition. It is thus clear that the catalytic styrene hydrogenation via the terminal CaÀ H bond of the solvent-free cation monomer 1 m + encounters only a moderate barrier of 20.6 kcal/mol, which is kinetically 3.7 kcal/ mol more favorable than that via cooperative CaÀ HÀ Ca bridges of the solvent-free cation dimer 1 2 + and reasonably accounts for the efficient styrene hydrogenation observed at room temperature. [8]

Conclusion
Extensive dispersion-corrected DFT calculations disclose a novel substrate-dependent catalytic mechanism of 1-alkene hydrogenation and H 2 isotope exchange reactions with the same calcium hydride cation dimer catalyst 1 2 + ·THF. It is shown that cooperative CaÀ HÀ Ca bridges of solvent-free dimer 1 2 + are favored for H 2 isotope exchange, competitive CaÀ HÀ Ca bridges of 1 2 + and the terminal CaÀ H bond of cation monomer 1 + ·THF are involved in the anti-Markovnikov addition of unactivated 1alkenes, while the terminal CaÀ H bond of solvent-free monomer 1 + is preferred for the Markovnikov addition of conjugationactivated styrene as directed by strong cation-π interactions. The novel mechanism can reasonably explain the known experimental observations and may be a useful guide for rational catalyst design.

Computational Methods
All DFT calculations were performed with the TURBOMOLE 7.4 suite of programs. [11] The structures were fully optimized at the TPSSÀ D3/def2-TZVP + COSMO level in THF solution, which combines the TPSS meta-GGA density functional [12] with the BJdamped DFTÀ D3 dispersion correction [13] and the def2-TZVP basis set, [14] using the Conductor-like Screening Model (COSMO) [15] for THF solvent (dielectric constant ɛ = 7.58 and diameter R solv = 3.18 Å). The density-fitting RIÀ J approach [16] was used to accelerate the calculations. The optimized structures were characterized by frequency analysis (no imaginary frequency for true minima and only one imaginary frequency for transition states) to provide thermal free-energy corrections (at 298.15 K and 1 atm) according to the modified ideal gas-rigid rotor-harmonic oscillator model. [17] More accurate solvation free energies in THF solution were computed with the COSMO-RS model [18] (parameter file: BP_ TZVP_C30_1601.ctd) using the COSMOtherm package [19] based on the TPSSÀ D3 optimized structures, corrected by + 1.89 kcal/ mol to account for the 1 mol/L reference concentration in solution. To check the effects of the chosen DFT functional on the reaction energies and barriers, single-point calculations at both TPSSÀ D3 [12] and hybrid-meta-GGA PW6B95-D3 [20] levels were performed using the larger def2-QZVP [14] basis set. Final reaction free energies (ΔG) were determined from the electronic single-point energies plus TPSSÀ D3 thermal corrections and COSMO-RS solvation free energies. As noted previously for similar hydrogenation reactions, [10,21] the reaction energies from both DFT functionals are in very good mutual agreement of 0.0 � 1.4 kcal/mol (mean � standard deviation) though as expected 0.2 � 1.6 kcal/mol higher barriers were found at the PW6B95-D3 level. In our discussion, the more reliable PW6B95-D3 + COSMO-RS free energies (in kcal/mol, at 298.15 K and 1 mol/L concentration) were used unless specified otherwise. The applied DFT methods along with the large AO basis set provide usually accurate electronic energies leading to errors for chemical energies (including barriers) on the order of typically 1-2 kcal/mol. This has been thoroughly tested for the huge data base GMTKN55 [22] which is the common standard in the field of DFT benchmarking. For general recommendations on DFT based computational chemistry studies see Ref. [23].