Controlled Selectivity through Reversible Inhibition of the Catalyst: Stereodivergent Semihydrogenation of Alkynes

Catalytic semihydrogenation of internal alkynes using H2 is an attractive atom-economical route to various alkenes, and its stereocontrol has received widespread attention, both in homogeneous and heterogeneous catalyses. Herein, a novel strategy is introduced, whereby a poisoning catalytic thiol is employed as a reversible inhibitor of a ruthenium catalyst, resulting in a controllable H2-based semihydrogenation of internal alkynes. Both (E)- and (Z)-alkenes were obtained efficiently and highly selectively, under very mild conditions, using a single homogeneous acridine-based ruthenium pincer catalyst. Mechanistic studies indicate that the (Z)-alkene is the reaction intermediate leading to the (E)-alkene and that the addition of a catalytic amount of bidentate thiol impedes the Z/E isomerization step by forming stable ruthenium thiol(ate) complexes, while still allowing the main hydrogenation reaction to proceed. Thus, the absence or presence of catalytic thiol controls the stereoselectivity of this alkyne semihydrogenation, affording either the (E)-isomer as the final product or halting the reaction at the (Z)-intermediate. The developed system, which is also applied to the controllable isomerization of a terminal alkene, demonstrates how metal catalysis with switchable selectivity can be achieved by reversible inhibition of the catalyst with a simple auxiliary additive.


Supplemental Notes
Note S1. Different rates of Z-E isomerization Figure S1. Different rates of Z-E isomerization process. Different Z-E isomerization rates were observed under different conditions without thiol. The whole actual catalytic reaction, which contains the Z-E isomerization process, only lasted <15 min (Figure S1, eq. a) but the independent test of Z-E isomerization starting from (Z)-2a took more than 1 h under the similar conditions (Figure S1, eq. b). In addition, decreasing the catalyst loading to 1/5 (0.08 mol%) greatly slowed down the reaction (more than 80 times), mainly because of the slower isomerization process ( Figure S1, eq. c).
It is proposed that the concentration of fac-Ru-1 in the system, which is the actual catalytic species, causes the different isomerization rates. In the actual catalysis, the catalyst Ru-1 (mer) fully converted into ruthenium alkenyl species Ru-2 (fac) promoted by the large excess amounts of alkyne substrate, which underwent further hydrogenolysis to generate fac-Ru-1. Such high concentration of fac-Ru-1 quickly catalyzed the isomerization process of the alkene before it flipped back to Ru-1 (mer).
In the control experiment of the isomerization test ( Figure S1, eq. b), only a small S3 percentage of fac-Ru-1 was generated from the flip of Ru-1 (mer) (ΔG mer to fac in Tol = 10.1 kcal/mol), thus the isomerization rate decreased compared to the actual catalysis involving alkynes.
To verify this proposal, another control experiment of Z-E isomerization was carried out but in the presence of a catalytic amount of 1a ( Figure S1, eq. d, premix with Ru-1 for 4 h). Expectedly, the catalysis was complete within 5 min (color change to yellow occurred within <10 s). These results support the proposal that the available concentration of fac-Ru-1 affected the isomerization rate.   Figure S4. 13 C-DEPTQ NMR spectrum of Ru-2 in C 6 D 6 .

VT-NMR of alkenyl species Ru-2 and its reaction with NACET
To a J. Young NMR tube was added complex Ru-1 (11.4 mg, 0.02 mmol), alkyne 1a (3.6 mg, 0.02 mmol) and toluene-d 8 (0.5 mL) in a N 2 glovebox. The NMR tube was rotated at room temperature for 6 h, and then measured at rt, 0 o C, -20 o C, -40 o C, and rt again. Two different phosphorus peaks were observed upon lowering the measuring temperature, which supported the fac-conformation of the complex. After the measurement, the J. Young NMR tube was charged with NACET (0.8 equiv) and the resulting solution was analyzed by NMR, indicating that the majority of Ru-2 transformed into Ru-6 along with the accumulation of (Z)-2a.

Preparation of Ru-4 and Ru-6
To a J. Young NMR tube was added complex Ru-1 (11.4 mg, 0.02 mmol), NACET (3.8 mg, 0.02 mmol) and MeOH-d 4 (0.5 mL) in a N 2 glovebox. The NMR tube was rotated at room temperature for 2 h (it takes time to dissolve Ru-1 in MeOH), with the color changing from orange to light green. NMR analysis indicated the generation of Ru-4, in accordance with our previous observations. 2 Due to the existence of the chiral center in the complex, multiple couplings were observed.  Figure S14. 1 Figure S18. COSY spectrum of Ru-4 in CD 3 OD. S18 To a J. Young NMR tube was added complex Ru-1 (11.4 mg, 0.02 mmol), NACET (3.8 mg, 0.02 mmol) and benzene-d 6 (0.5 mL) in a N 2 glovebox. The NMR tube was heated at 100 o C for 10 min with occasional release of the generated H 2 . The color of the solution changed from orange to dark green upon heating, and finally turned to brown when the solution was cooled down to room temperature. NMR analysis indicated the generation of Ru-6 as the major species, with a bidentate thiolate binding to the ruthenium center. The generation of Ru-6 by heating Ru-1 and NACET is not a clean reaction, with around 10% Ru-4 and some unidentified species from the thermal decomposition of Ru-6 in the solution. Due to the existence of chiral centers in the complex, multiple couplings were observed.

Ru
During the heating the color of the solution was dark green, which is in accordance to the color of a ruthenium thiolate complex (monodentate Ru-5). 2 The color of the solution quickly converted into brown upon cooling, implying the coordination of an additional group to the ruthenium center. Similar color change was also observed employing butyl 3-mercaptopropionate as the additive.

Control experiments
To a J. Young NMR tube was added the carboxylate complex 6 (3.5 mg, 0.005 mmol) and toluene (0.5 mL) in a N 2 glovebox. The NMR tube was taken out of the box and pressurized with 5 bar of H 2 , and allowed to rotate at room temperature for 5 min.     To a J. Young NMR tube was added Ru-1 (2.8 mg, 0.005 mmol), hexanethiol (1.8 μL, 2.5 equiv), 1a (44.5 mg, 50 equiv) and toluene-d 8 (0.5 mL) in a N 2 glovebox. The NMR tube was taken out of the box and measured by NMR, indicating the generation of Ru-4 (HexSH) and Ru-5 (HexSH). The tube was pressurized with 5 bar of H 2 and allowed to rotate once. Immediate color change from green to red brown was observed. After standing for 5 min, the red brown color changed back to light green gradually, with the top solution still keeping red brown ( Figure S28). 31 P NMR measurement of the sample only detected Ru-4 (HexSH) and Ru-5 (HexSH) during the reaction ( Figure S29). However, from the 1 H NMR we could observe the vinylic proton of Ru-2, although in a small amount ( Figure S30). In addition, the red brown colour observed during the reaction, which is characteristic of Ru-2, also supports that the alkenyl species appeared during the reaction. These results imply an inner-sphere hydrogenation of alkynes in the presence of thiol.
Note: As mentioned above, Ru-2 has a very broad 31 P signal, thus it's difficult to detect, especially in a small amount, by 31 P NMR. Note: It was found a little higher TOF (>1150 h -1 ) could be achieved using a bigger tube with 10 mL volume. In addition, the rotation of the NMR tube is essential to the conversion of the NMR experiments of the hydrogenation reactions, possibly due to the solvation rate of H 2 into the solution. Thus, the abscissa of the kinetic profiles of the reaction only counts the rotation time (except Figure 1f). In addition, a little longer reaction period was required for the monitoring than the actual catalysis, possibly because of the disruption of the rotation of the NMR tube (especially the step of isomerization, see Note S1). NMR experiments in Figure 1f (independent Z-E isomerization tests) were carried out without the rotation of the tube, and measured by in-situ NMR.    Then the crude mixture was dissolved in hexane and passed through a silica column, after which the solvent was removed to afford a mixture of alkenes and the unreacted alkyne (84% NMR yield according to the weight and integrals).

Single crystal X-ray diffraction analysis
Single crystal XRD were measured by a sealed tube Rigaku Synergy-S dual source equipped with Dectris Pilatus3 R CdTe 300K detector and microfocus diffractometer, with MoK (=0.71073 Å). Data collection was performed in low temperature under LN. Data were processed with CrysAlis PRO (Rigaku). Structure was solved using SHELXT 7 and refinement performed based on F 8 with SHELXL 8 and OLEX2 9 with full matrix least-squares. All non-hydrogen atoms were refined aniostropically.
Hydrogens were placed at calculated positions and refined using a riding model.   where E ωB97M-V Toluene is the single point energy; and where corr M06-L freq is the thermal correction to the Gibbs free energy from the frequency calculation.
Free energy values (Gº ) were then corrected to account for changes in standard states (Gº  G).
Standard state corrections 22 were employed such that all species are treated as 1M (using an ideal gas approximation), with the exception of H 2 (maintained as 1 atm). [23][24][25] Other than these standard state corrections, the transformation of hydrogen from the condensed phase to the gas phase is not additionally corrected in the free energy quantities provided.

S68
Ethanethiol were studied as minimal models for hexanethiol in the system.
Directionality of ∆G and ∆G TS values are indicated by the ordering of X,Y and all energies are reported in kcal/mol. Mass balance is ensured throughout.