Semihydrogenation of Alkynes Catalyzed by a Pyridone Borane Complex: Frustrated Lewis Pair Reactivity and Boron–Ligand Cooperation in Concert

Abstract The metal‐free cis selective hydrogenation of alkynes catalyzed by a boroxypyridine is reported. A variety of internal alkynes are hydrogenated at 80 °C under 5 bar H2 with good yields and stereoselectivity. Furthermore, the catalyst described herein enables the first metal‐free semihydrogenation of terminal alkynes. Mechanistic investigations, substantiated by DFT computations, reveal that the mode of action by which the boroxypyridine activates H2 is reminiscent of the reactivity of an intramolecular frustrated Lewis pair. However, it is the change in the coordination mode of the boroxypyridine upon H2 activation that allows the dissociation of the formed pyridone borane complex and subsequent hydroboration of an alkyne. This change in the coordination mode upon bond activation is described by the term boron‐ligand cooperation.


Introduction
The seminalf inding that specific combinations of sterically en-cumberedL ewis bases and Lewis acids, named "frustrated Lewis pairs" (FLPs), can activate hydrogen, stimulated the development of catalytic metal-free hydrogenations. [1] Early examplesi ncluded the hydrogenation of (di)imines, nitriles, aziridines, silyl enol ethers, and enamines, but the scope of FLP catalyzed hydrogenations was extendedt oh eterocycles, alkenes, allenes, and aromatic hydrocarbons. [2,3] The heterolytic hydrogen cleavage by the FLP yields at etravalent borohydride species. Therefore, hydrogenations by FLPs consist of ahydride and as ubsequent protont ransfer step (or vice versa)a nd require activated alkenes. [3] An otable exception is the semihydrogenation of alkynes catalyzed by an intramolecular FLP that was reported by Repo et al. [4] In that case, mechanistic investigationss howed that the protolysis of the FLP under the reaction conditions yields an amine-hydroborane that initiates the catalytic cycle by hydroboration of the alkyne. [5] Ap rotodeborylation of the alkenylborane yields then, in ah ighly stereoselective reaction, the cis-alkene. [6] We recently reported reversible H 2 activation by the boroxypyridine 3. [7] Ad istinguishing feature of this system is that the H 2 activationi sa ssociated with at ransition of the covalently bound oxypyridine substituent to an eutral pyridone donor ligand (Scheme 1). This mode of action was, in analogy to the concept of metal-ligand cooperation, termed boron-ligand cooperation. The change in the coordination mode of the pyridone substituent might enable the dissociation of the pyridone borane complex 4 in the ligand6 -tert-butylpyridone 5 and Piers borane 6.P iers borane has been shown to display the typical reactivity of at rivalent borane, for example,i te ffects the hydroboration of alkenes and alkynes.S uch dissociation is not possible for classic FLPs that, as aforementioned, therefore rather display borohydride reactivity upon H 2 activation (Scheme1).

Results and Discussion
We envisioned the hydroborationo fa na lkene to be av alid test reaction to elucidate whether 3 displays borane reactivity upon hydrogen activation, since hydroborationr equires the presenceo fatrivalent borane. Indeed,w hen 3 was reacted with one equivalent of styrene under moderate H 2 -pressure at RT,t he formation of the alkyl borane 7 was observed (Scheme 2). The alkylborane 7 is also formed when styrene is reactedw ith the pyridone borane 4,w hich supports the assumption that 4 is an intermediate in the formation of 7 starting from 3.
The alkylborane 7 does not undergo ap rotodeborylation. However,w ee nvisioned that an analogous alkenylborane, originating from ar eactions equence consisting of H 2 activation and hydroboration of an alkyne might succumb to protonolysis. This reactionw ould regenerate the boroxypyridine 3 and close ac atalytic cycle for the hydrogenation of alkynes that consists of H 2 activation by 3,h ydroboration of an alkyne and protonolysis of the alkenylborane (Scheme 3).
Indeed, 2-hexynew as stereoselectively converted to cis-2hexene in 87 %y ield in the presence of catalytic amountso f4 at 80 8Cu nder 5bar H 2 pressure (Scheme 4). The catalyst 4 was generated in situ by coordination of 5 to Piers borane 6.Ani nitial screening of reaction conditions showedt hat as light excess of Piers borane 6 (1.3 equivalents with respectt o5)i s beneficial to obtain reproducible good yields. Under the same conditions, cis-2-octene is obtained in very good yields from the hydrogenation of 2-octyne. Likewise, cis-3-hexene is formed upon hydrogenation of 3-hexyne in excellent yield after only 8h reaction time. The hydrogenation of 4-methyl-2pentyne leads to the corresponding cis alkene in av ery good yield after 16 hr eaction time. Upon hydrogenation of the respectivea lkyne, 1-phenyl-1-propenei sobtained in an excellent yield of 93 %. Ethers are suitable substrates, as proven by the successful hydrogenation of 1-(para-methoxyphenyl)-propyne.
While 3-hexyne is obtained after 8h exclusively as cis isomer,aprolonged reaction time of 16 hl ed to a1 :1 mixture of the cis and the trans isomer (Scheme 5). After 20 h, the trans isomer is the major product. Liu et al. reported that Piers borane can isomerize cis-alkenes via reversible hydroboration. [5] We, therefore, assume that the catalytic reaction yields first cis-3-hexene that is then subsequently isomerized by the Piers borane 6 that is present in the reaction mixture. Thus, both stereoisomersa re accessible with the catalytic protocol described herein.
Scheme3.Envisionedmechanism of the hydrogenation of alkynes catalyzed by 3:H 2 activationyields the pyridone borane complex 4 that undergoes a dissociation. Piers borane 6 hydroboratesa nalkyne, formation of the pyridone alkenylborane complex and its protolysisare closing the catalytic cycle.
Scheme4.Substrate scope of the semihydrogenation of internal alkynes. Yields were determined by 1 HNMR with trimethoxybenzene as internal standard and are given as the average of two runsa)8hreaction time; b) 16 h reactiont ime.
Scheme5.Stereoselectivity of the hydrogenation of 3-hexyne in dependence of the reaction time.
The known metal-free protocols for the hydrogenation of alkynes are limited to internal alkynes. We were pleased to find that the catalyst described herein is capable to hydrogenate1octyne in good yield with ac atalyst loading of 10 mol % (Scheme 6). The catalytic protocol can also be used for the hydrogenation of other aliphatic alkynes such as cyclohexyl-and adamantly acetylene. While aromatic rings are tolerated, the hydrogenation of phenylacetylene and para-(trifluoromethyl)phenylacetyleney ielded the corresponding alkenes in lower yields. Again, ethers are suitable substrates, as demonstrated by the hydrogenationo f6 -methoxy-1-hexylacetylene.
With these resultsi nh and, we aimed foramechanistic understanding of the catalytic reaction. To verify that the pyridone 5 is indeed vital for the reaction, we attempted the hydrogenation of 2-hexyne only with Piers borane 6 as catalyst (Scheme7). Less than 1% product was formed under reaction conditions that are identical to those reported in Scheme 4, clearly indicatingt hat the presence of the pyridone 5 is essential for the reaction outcome.
We then focusedo nt he identification of the resting state of the catalytic reaction. For this purpose, the catalytic hydrogenation of 3-hexyne was monitored by NMR (Scheme 8). Under 4bar H 2 -pressure, rapid formation of cis-3-hexene was observed at 70 8Ci n[ D 6 ]benzene, which impliest hat the observations made by this experiment are meaningful regarding the catalytic transformation.
The bispyridone complex 8 that was previously described and characterizedi nd etail was observed by 1 HNMR as the resting state of the catalytic reaction ( Figure 1). [8] Furthermore, 1 Ha nd 11 BNMR proved formation of boroxypyridine 3 with progressing reaction and hydrogen consumption.T his finding strongly supports the assumption that 3 is part of the catalytic cycle. [9] To elucidate whether the envisioned protonolysiso ft he alkenylborane can be assumed to be part of the catalytic reaction, 5 was added to the borane 9,d erived from the reaction of Piers borane 6 and 3-hexyne. The reactionp rogress at RT was monitored by NMR spectroscopy (Scheme 9). Within 30 minutes, the formation of the expected pyridone alkenylborane complex 10 was observed. Furthermore, signals that were assigned to cis-3-hexene, the product of the protonolysis, were detected. The presence of cis-3-hexenei mplies that boroxypyridine 3,o riginating from the protonolysis must be present. Indeed, the formation of the bispyridone complex 8 that containsone equivalent of 3 was observed.
EXSY NMR spectroscopy shows an exchange of the pyridone 5 between 10 and 8 at RT,w hich furthers upports that 8 is not an unreactive, irreversibly formed species but rather a resting state. The mechanism of the catalytic reaction was further investigated computationally at revDSD-PBEP86-D4/def2-QZVPP//PBEh-3c (Figure2). [10,11] The SMD model for n-hexane was used to implicitly account for solvent effects. [12] The hydro-Scheme6.Substrate scope of the semihydrogenation of terminal alkynes. Yields were determined by 1 HNMR with trimethoxybenzene as internal standard and are given as the average of two runs.
Scheme8.NMR monitoring of the catalytic hydrogenationo f3-hexyne. Scheme9.Stoichiometric reaction of the alkenylborane 9 with the tert-butylpyridone 5. gen activation by 3 requires af ree activation energy of 19.4 kcal mol À1 .T his elementary step is according to our computations thermoneutral, which agreesw ith the previously observed facile reversibility of the hydrogen activation. [7] The free energy change that is associated with the dissociation of 4 into Piers borane 6 and the pyridone 5 is 16.8 kcal mol À1 .R elaxed potential energy surface scans indicatet hat the dissociation is barrierless.A st he experimental results indicate that the bispyridonec omplex 8 is the resting state of the transformation, we consideredt he coordinationo ft he free pyridone 5 to the boroxypyridine 3.I ndeed,t he formationo f8 is according to the computations exergonic. The hydroboration of the model substrate 2-butyne requires am oderate activation energy of 4.9 kcal mol À1 and yields the alkenylborane 11.T he bispyridonec omplex 8 together with 11 is the restings tate of the catalytic transformation. [13] The pyridone 5,t hat is bound in complex 8,c oordinates than to 11 forming the pyridone alkenylborane complex 12.
Note that pyridone exchange between 8 and the pyridone alkenylborane complex 10 was observed experimentally by EXSY NMR. The activation barrierf or the protodeborylation is 22.2 kcal mol À1 ,w hich corresponds to ah alf-life time of 12 of 35.8 minutes at 25 8C. [14] This agreesw ith the experimental observation that the protodeborylation takes place at RT (Scheme 9). The "Energetic Span", that is the kinetic barriero f the catalytic transformation,i sb etween the resting state (8 and 11)a nd the transition state of the protodeborylation. [15] ClassicF LP type catalysts are not suitable fort he hydrogenation of terminal alkynes, presumably because they are deactivated by an irreversible C sp ÀHc leavage. [3] To understand why the catalysts ystem described herein tolerates terminal alkynes, 3 was reactedw ith cyclohexylacetylene at RT.A sp reviously reported, this reaction led to the formation of the alkynylborane complex 13 (Scheme 10). [16] Upon additiono fp henylacetylene Figure 2. Gibbs free energy profile for the hydrogen activation by 3 computed at revDSD-PBEP86-D4/def2-QZVPP//PBEh-3c. Bulk solvation was considered implicitlywith the SMD model for hexane.
After 1hat 80 8C, the ratio of 14 to 13 was 4:1. This experiment indicates that the C sp ÀHc leavage is reversible under the reactionc onditions. The assumption that the formation of the alkynylborane is reversible is further supported by DFT computations ( Figure 3). According to the computations, the liberation of cyclohexyacetylene from 13 requires af ree Gibbs activation energy of 24.1 kcal mol À1 ,w hich corresponds to ah alflife time of 79 seconds at 80 8C. The formation of the phenyl alkynyl borane complex 14 is kinetically and thermodynamically favored.
The computed Gibbs free energy differenceo f0 .4 kcal mol À1 corresponds to aratio of 2:1, whichisinr easonable agreement with the experimentally observed proportion of the two alkynyl borane complexes.I ti sc ertainly the reversibility of the C sp À Hc leavaget hat allows H 2 activation in the presence of terminal alkynes and thus the first metal-free hydrogenation of terminal alkynes.

Conclusions
We have documented the efficient semihydrogenation of internal and terminal alkynes by aboroxypyridine that displays frustrated Lewis pair reactivity and is, therefore, able to activate hydrogen. However,t he change in the coordinationm ode of the pyridonate substituent enablesh ydroboration as the initial step of the hydrogenation and is thus vital for the catalytic reaction. We expect this finding to pavet he way for novel metal-free catalytic reactions that rely on this mode of action.

Experimental Section
General Procedure for hydrogenation of alkynes:P iers borane 6 (13.5 mg, 0.039 mmol) and 6-tert-butyl-2-pyridone 5 (4.5 mg, 0.030 mmol) were dissolved in n-hexane (5 mL) in aF isher-Porter type 150 mL reaction vessel equipped with as tirring bar.T he respective alkyne (0.60 mmol or 0.30 mmol) was added. The reaction vessel was closed and connected to an H 2 bomb with ag as hose. The hose was rinsed with H 2 several times and the reaction vessel pressurized with H 2 (5 bar). The reaction vessel was placed inside an 80 8Cp reheated oil bath and stirred at 1000 rpm. After 20 h, the reaction mixture was cooled to room temperature and the excess H 2 gas was released. An aliquot was taken, and the yield determined by 1 HNMR using 1,3,5-trimethoxybenzene as internal standard.