Stereoselective alkyne semihydrogenations with an air-stable copper(I) catalyst†

Catalytic stereoselective alkyne semihydrogenations are powerful and atom-economic synthetic alternatives to olefination reactions. The resulting alkenes are valuable building blocks especially for diastereoselective follow-up reactions. For Z-selective alkyne semihydrogenations, the Lindlar catalyst has become the first choice, however, it suffers from E/Z-isomerisation processes and overreduction to the corresponding alkanes. While the latter leads to loss of the desired functionality, the former can be problematic with foresight to tedious separations and consecutive diastereoselective transformations. Hydrogenations catalysed by readily available first row transition metals are desirable from an economic point of view. Among them, homogeneous catalysts based on copper(I) have recently emerged as viable alternatives for Z-selective alkyne semihydrogenations. Key reactivity for these catalytic processes is the reported stereoselective insertion of alkynes into copper(I) hydride bonds. While most of the disclosed catalysts allow for good to excellent Z-stereoselectivity in alkyne semihydrogenations, all studied copper(I) complexes need to be prepared in situ as the active catalysts are unstable. This feat can be ascribed to the formation of a Cu–O-bond, which allows for H2 activation 14 but at the same time renders the corresponding complexes sensitive to air and moisture. The need for preactivation hampers the practicability of the overall processes, as can be seen from the studied catalysts so far: a triphenylphosphine/copper(I) complex can be used under an H2 amosphere (5 bar) at elevated temperatures in combination with iso-propanol to transform alkynes into the corresponding Z-alkenes (Scheme 1a). This catalyst has to be activated with an alkoxide at elevated temperatures (100 °C) and is limited to mainly unfunctionalised substrates. We have introduced a highly stereoselective alkyne semihydrogenation based upon copper(I)/N-heterocylic carbene (NHC) complexes bearing an alkoxide tether (Scheme 1b). This system requires high H2 pressure (100 bar) and the catalyst needs to be generated in situ from sensitive mesitylcopper(I). More recently, an NHC/copper(I) complex has been reported which allows for alkyne semihydrogenations at 1 bar H2. The catalyst has to be generated in situ from a copper(I) chloride/ NHC precursor with sodium tert-butanolate and shows somewhat reduced Z/E-selectivity (Scheme 1c).


Introduction
Catalytic stereoselective alkyne semihydrogenations are powerful and atom-economic synthetic alternatives to olefination reactions. 1,2 The resulting alkenes are valuable building blocks especially for diastereoselective follow-up reactions. 3 For Z-selective alkyne semihydrogenations, the Lindlar catalyst 4 has become the first choice, however, it suffers from E/Z-isomerisation processes and overreduction to the corresponding alkanes. 2 While the latter leads to loss of the desired functionality, the former can be problematic with foresight to tedious separations and consecutive diastereoselective transformations.
Hydrogenations catalysed by readily available first row transition metals are desirable from an economic point of view. 5 Among them, homogeneous catalysts based on copper(I) have recently emerged as viable alternatives for Z-selective alkyne semihydrogenations. [6][7][8][9][10] Key reactivity for these catalytic processes is the reported stereoselective insertion of alkynes into copper(I) hydride bonds. [11][12][13] While most of the disclosed catalysts allow for good to excellent Z-stereoselectivity in alkyne semihydrogenations, all studied copper(I) complexes need to be prepared in situ as the active catalysts are unstable. This feat can be ascribed to the formation of a Cu-O-bond, which allows for H 2 activation 14 but at the same time renders the corresponding complexes sensitive to air and moisture. The need for preactivation hampers the practicability of the overall processes, as can be seen from the studied catalysts so far: a triphenylphosphine/copper(I) complex can be used under an H 2 amosphere (5 bar) at elevated temperatures in combination with iso-propanol to transform alkynes into the corresponding Herein, we report on the identification of a preactivated and air-stable NHC/copper(I) hydroxide complex, [IPrCuOH], 16 for highly Z-selective alkyne semihydrogenations. The stability of the precatalyst allows for a more practical effectuation of the semihydrogenation without jeopardizing the stereoselectivity. 17

Results and discussion
Optimisation of the alkyne semihydrogenation The alkyne semihydrogenation has been optimised using pentynol-derived internal alkyne 1 (Table 1)

Substrate scope
With optimised reaction conditions in hand, we set out to investigate the substrate scope of the Z-selective alkyne semihydrogenation with [IPrCuOH] and found generally broad applicability, while Z-stereoselectivity remained high (Scheme 2). A variety of electron-rich and electron-poor aryl/ alkyl-substituted alkynes 3a-3e based upon the pentynolframework gave the corresponding Z-alkenes 4a-4e in high yields. Unlike our previously reported copper(I)/NHC complex (Scheme 1b), 7 [IPrCuOH] does not fully tolerate ketone functional groups, which is showcased by partial overreduction of 4f to the benzylic alcohol (ratio ketone/benzylic alcohol = 88 : 12). We hypothesise that the presence of an intermediate alcohol(ate) disturbs the overall chemoselectivity, as in this case overreduction to the alkane was substantial (15%). This effect of additional alcohol(ate)s mirrors those of our previous study. 7 In contrast, the tolerance of heterocycles differs from our earlier results: thiophene 4g, which was unreactive with the tethered catalyst, 7 can now be obtained in good yield  7 give varying results in terms of overreduction and/or E/Z isomerisation. This displays a vulnerability of the copper(I) catalyst to strongly coordinating substrates. The present protocol is applicable to diaryl-and dialkylalkynes alike: tolane (3k) and its derivatives prove to be suitable precursors for Z-stilbenes 4k-4m under the semihydrogenation conditions. Notably, methoxy-substituted tolane 3l required a somewhat higher H 2 pressure (100 bar). In a similar vein, dialkylalkynes 3n and 3o require slightly more forcing conditions (100 bar H 2 , 60°C) to allow full conversion to the desired Z-alkenes 4n and 4o. The protected allylic alcohol 4p is available from the corresponding propargylic silyl ether in high yield and excellent stereoselectivity using elevated temperature and H 2 pressure. This example marks one of the strong points of our catalytic process, as Z-allylic alcohols are important building blocks for diastereoselective follow-up reactions (see below for a synthetic elaboration of 4p). Finally, methyl-2-nonynoate (3q) shows excellent selectivity towards the corresponding Z-acrylate in our semihydrogenation protocol, albeit with low conversion (20%). Further investigations are needed to identify superior catalysts for this challenging, yet synthetically valuable semihydrogenation of propiolates. When investigating diyne 5, selectively only the E,E-diene 6 was isolated with high yield (87%) with our catalyst (Scheme 3). This is of note, as this class of compounds has been reported to get reduced to a Z-monoalkene by an earlier copper(I) hydrogenation catalyst 6 in alkyne semihydrogenation. To investigate the possible origin of this unexpected E-selectivity, we prepared E-enyne 7, a potential reaction intermediate in a stepwise diyne semihydrogenation from 5 towards 6. From this experiment, we isolated 96% of the E,Z-diene 8, representing only minor loss of stereochemical integrity of the primarily installed E-alkene. With this result, it seems reasonable to conclude that diynes such as 5 do not react step-wise as isolated triple bonds. A potential alkenylcopper(I) intermediate could equilibrate to a butatrienylcopper(I) intermediate 19 (not shown), which accounts for the formation of the thermodynamically more preferred E,E-diene 6 from diyne 5.
Follow-up chemistry of (Z)-allylic alcohols Finally, to demonstrate the usefulness of the present highly Z-selective alkyne semihydrogenation, we further elaborated silyl ether 4p (Scheme 4): after silyl ether deprotection with TBAF to the allylic alcohol 9 (83%), subsequent oxidation with the Dess-Martin periodinane 20 gave Z-acrolein-derivative 10 (88%). A subsequent Horner-Wadsworth-Emmons reaction was carried out to yield E,Z-sorbic acid derivative 11 in 86% yield. This approach underlines that a stereoselective alkyne semihydrogenation with [IPrCuOH] can serve as key step to generate alkene geometries with high selectivity.

Conclusions
In summary, we have developed a highly Z-stereoselective alkyne semihydrogenation protocol, relying on an air-stable and preactivated NHC/copper(I) hydroxide complex, [IPrCuOH]. The practicability of this catalyst circumvents previous shortcomings of other copper(I) complexes as it does not require preactivation. A variety of products are accessible via this protocol in high yields and excellent Z-selectivities. The findings presented here could make a contribution towards widely applicable catalysis with easily accessible first-row transition metals.

Experimental
All reactions were carried out in flame-dried glassware under a nitrogen atmosphere using standard Schlenk techniques. NMR spectra were recorded on AvanceII 400 MHz or AvanceIII 500 MHz or 700 MHz instruments (Bruker). Chemical shifts are reported in parts per million ( ppm) and are referenced to the residual solvent resonance as the internal standard according to literature values. 21   bar, a rubber septum pinched with a needle (0.90 × 50 mm, Braun) in autoclaves Berghof BR-100 or BR-300 equipped with heating blocks. For preparation and characterisation of the starting materials, see the ESI. †

General procedure alkyne semihydrogenation
The reaction vessel was placed in a N 2 -purged autoclave under a counterflow of N 2 . The autoclave was purged with N 2 (3 × 10 bar) and H 2 (3 × 20 bar) before the appropriate H 2 pressure was applied ( pressure is given as initial pressure before heating). The heating block was pre-heated before the autoclave was placed inside. After the reported reaction time the autoclave was allowed to cool to rt and H 2 was released. The autoclave was purged with N 2 (3 × 10 bar) before the reaction vessel was taken out. The reaction mixture was filtered through a small plug of silica (1 mL tert-butyl methyl ether as eluent), and all volatiles were removed under reduced pressure.
Reactions were subsequently analysed either by GC and/or NMR. The crude mixture was then subjected to purification as indicated with the appropriate substrates.