Olefin Metathesis Catalyzed by a Hoveyda–Grubbs-like Complex Chelated to Bis(2-mercaptoimidazolyl) Methane: A Predictive DFT Study

Although highly selective complexes for the cross-metathesis of olefins, particularly oriented toward the productive metathesis of Z-olefins, have been reported in recent years, there is a constant need to design and prepare new and improved catalysts for this challenging reaction. In this work, guided by density functional theory (DFT) calculations, the performance of a Ru-based catalyst chelated to a sulfurated pincer in the olefin metathesis was computationally assessed. The catalyst was designed based on the Hoveyda–Grubbs catalyst (SIMes)Cl2Ru(=CH–o–OiPrC6H4) through the substitution of chlorides with the chelator bis(2-mercaptoimidazolyl)methane. The obtained thermodynamic and kinetic data of the initiation phase through side- and bottom-bound mechanisms suggest that this system is a versatile catalyst for olefin metathesis, as DFT predicts the highest energy barrier of the catalytic cycle of ca. 20 kcal/mol, which is comparable to those corresponding to the Hoveyda–Grubbs-type catalysts. Moreover, in terms of the stereoselectivity evaluated through the propagation phase in the metathesis of propene–propene to 2-butene, our study reveals that the Z isomer can be formed under a kinetic control. We believe that this is an interesting outcome in the context of future exploration of Ru-based catalysts with sulfurated chelates in the search for high stereoselectivity in selected reactions.


Fig. S1
Alternative C=C coordination to the metal. The dissociative path reported in Figure 5 in the main manuscript corresponds to φ < 0, which may be counterintuitive because the isopropoxy group hinders the styrene rotation in this direction. However, the energy barrier associated with this step is only 11.8 kcal/mol, which is calculated via 2d ‡ (φ = -69.6°, dRu-O = 3.75 Å). A continued styrene rotation leads to the structure 3d' localized at 9.9 kcal/mol (φ = -113.5°, dRu-O = 4.21 Å), followed by 3d (φ = 176.8°, dRu-O = 4.66 Å). In the case of incomplete rotation of the styrene (these are not shown in Figure 5 and are denoted as xd', x = 3 to 6), the isopropoxy group is located below the ruthenium atom (in standard orientation with NHC located above the ruthenium atom) in 4d and, based on linear transit calculations, we observed that the side-and bottom-bound mechanisms depend on the ethylene carbon that reacts first. In the case of 5d' ‡ cis, the isopropoxy group was spontaneously linked to ruthenium again (φ = 15.9°, dRu-O = 2.48 Å). On the other hand, the analogous transition state 5d ‡ trans prevents the styrene back-rotation (φ = 167.9°, dRu-O = 4.70 Å). The structural descriptors d1, d2, and d3 are similar for both transition states, but the energy barrier from 4d to 5d' ‡ cis is slightly lower (12.5 compared to 14.8 kcal/mol with 5d ‡ trans, see Figure S2). On the other hand, 6d'cis is highly stabilized by 14.1 kcal/mol compared to 1d. Attempts to localize 7d' ‡ from 6d'cis were unsuccessful, since the Ru-O bond dissociates during the linear-transit calculations. The high stabilization of 6d'cis represents an important drawback from a practical point of view, since the course of the reaction may fall into a potential well: reversion via side 5d' ‡ cis requires 24.9 kcal/mol, and progression to active catalyst 9d would require at least 31 kcal/mol. In contrast, even though MCB 6dtrans is moderately stabilized by 6.8 kcal/mol compared to 1d, the 2,2cycloreversion is blocked by a large energy barrier of 41.7 kcal/mol, according to 7d ‡ . Furthermore, we investigated an alternative pathway starting from the species 3dr, where the subscript "r" stands for rotated styrene at φ = -140.8° and dRu-O = 4.38 Å. Although 5dr ‡ trans is 4.7 kcal/mol lower than the analogous 5dr ‡ cis (details in Figure S2), the resulting MCB 6dr is even more stabilized by 18.6 kcal/mol as compared to 1d. Therefore, these results suggest that productive olefin metathesis is hindered by path d.

Pathways via 1e and 1f
In the case of complex 1e, styrene rotation occurs via 2e (φ = 69.7°, dRu-O = 3.69 Å), which is characterized as a local minimum at an energy cost of 10.3 kcal/mol (see Fig. S3). During styrene rotation towards 3e, both thiones were bonded to Ru and both Ru-S bonds were kept until the MCB. Ethylene coordination releases 1.3 kcal/mol via 4ecis, and the energy barrier calculated with 5e ‡ cis to form the MCB is 17.2 kcal/mol. Even though 6ecis is less stabilized than side-bound 6a-c, the product release is hampered by 31.7 kcal/mol at the 2,2-cycloreversion step, despite 7e ‡ is only 16.1 kcal/mol above the precatalyst 1e. Geometry optimizations of 7e ‡ resulted in the rupture of one Ru-S bond. We assumed the reaction from 6ecis may proceed with only one Ru-S bond; but it would be probably reverted through the dissociative step before MCB formation. Additionally, considering the reaction mechanisms formulated by Houk et al. (J. Am. Chem. Soc. 2012, 134, 1464 for the Grubbscarboxylate catalyst shown in Scheme 2c (analogous to nitrate), some intermediate species resulted in only one Ru-O bond, which suggests the chelating agent links the metal centre depending on the electronic environment.

Fig. S3
Gibbs free energies profiles (kcal/mol) of the initiation phase for complexes 1e and 1f. Energy differences are relative to 1a.
In the case of 1f, the higher energy structure though the dissociative path is 3f' (φ = -118.0°, dRu-O = 4.13 Å) instead of 2f (φ = -67.4°, dRu-O = 3.12 Å), both are local minima, and the resulting energy barrier is therefore only 5.6 kcal/mol. Continued rotation is an exergonic process leading to intermediate 3f (φ = 121.8°, dRu-O = 4.14 Å), and releasing 4.9 kcal/mol. We additionally localized a transition state related to the η 2 -coordination of ethylene to form 4fcis, which adds a second barrier of 8.2 kcal/mol via 4f ‡ as compared to 3f. The formation of 4fcis releases 12.0 kcal/mol and the energy cost associated to the formation of a MCB is 21.5 by means of 5f ‡ cis, yet the MCB 6fcis is highly stabilized by 20.2 kcal/mol. Nonetheless, we observed that reversion of the reaction from the olefin coordination step towards 1f occurs at a lower energy cost. We conclude therefore that olefin metathesis across complex 1f is not viable since the reaction will be probably reverted before reaching the respective MCB. * Strain is evaluated considering distortion of precatalyst 1.

Fig. S4
Coordination of the thione C=S bond to Ru followed by catalyst decomposition.

Fig. S5
3D representation of DFT-optimized geometries of a) active catalyst 9, and b) olefin coordination corresponding to the propagation phase. Gibbs energy comparisons are given in kcal/mol for each case. Structures to the left are the ones discussed in the main manuscript. Hydrogen atoms are hidden for the sake of clarity.

Table S3
Total energy values (E) and Gibbs free energy values (G; as defined in the manuscript) for structures reported in Table S1 in the same order, along with Cartesian coordinates and its corresponding 3D view of optimized geometries. Species used for the analysis of stereoselectivity are also included consecutively.