Synthesis of Linear Enamides and Enecarbamates via Photoredox Acceptorless Dehydrogenation

: In recent years, several methods for the direct desaturation of aliphatic compounds have been developed, facilitated by the unique combination of photoredox and transition-metal catalysis. Hereby, alkenes with high functionalization potential can be prepared in a straightforward fashion. We adapted a previously reported system involving tetrabutylammonium decatungstate (TBADT) as hydrogen atom transfer (HAT) agent and a cobalox-ime co-catalyst for dihydrogen evolution for the dehydrogenative preparation of linear enamides and enecarbamates from saturated precursors. The substrate scope includes several natural products and drug derivatives. The reaction does not require noble metal catalysts, exhibits short reaction times compared to previous methods and is suitable for the late-stage functionalization of drug derivatives

The direct dehydrogenation of aliphatic compounds to olefins carried out by enzymes is an essential transformation in living systems, and is involved in several biosynthetic pathways such as the desaturation of fatty acids or carotenoids. [1]The emulation of these reac-tions by chemists has remained unmet, since the activation of two relatively inert C(sp 3 )À H bonds in a single process is kinetically and thermodynamically disfavoured. [2]As a result, direct desaturations often require harsh reaction conditions involving the combination of transition metal catalysts with external oxidants/H-acceptors, [3] or in situ generation of intramolecular radicals as H-acceptors, often at high temperatures. [4]In view of this, the development of alternatives with better atom-and energy efficiency for the conversion of alkanes to alkenes is highly desirable. [5]Over the last decade, photochemistry has emerged as a technique by which transformations under milder reaction conditions can be achieved.[9] Preoxidized or prefunctionalized starting materials [10] can be avoided.7c,8,9] Even though some of these methodologies involve organocatalysts as additional HAT-mediators, [9] most systems rely on a dual catalytic approach by combining a photoexcitable HAT-catalyst, which upon irradiation is capable of generating a C-centered radical, and a transition-metal complex that is required for subse-quent β-hydride elimination and dihydrogen evolution. [7,8]7a] Enamide and enamine scaffolds are valuable synthetic building blocks, which are ubiquitous among agrochemicals, natural products, pharmaceuticals, and as monomers for the preparation of polymeric materials. [11]3d,12] This has partly been circumvented with the development of novel photochemical methodologies relying on dehydrogenation processes.7b] In the same year, the Gevorgyan group reported a site-selective protocol for the desaturation of amines, wherein the position of the double bond is controlled by the effect of a directing group attached to the substrate.4e] More recently, the El-Sepelgy group published an alternative methodology based on the same principle for the desaturation of a variety of amines and amides, using a tin-cobaloxime complex instead (Figure 1b). [13]lthough both protocols give access to a wide array of synthons in good yields, the incorporation of the directing group requires an additional step which limits their application in late-stage functionalization or even for the desaturation of biologically relevant molecules.As an alternative to these approaches, the Huang group reported a versatile protocol for the desaturation of a wide array of aliphatics.7d] This reaction proceeds regioselectively under mild conditions, and with perfect atom economy.
Building on the above-mentioned methodologies developed by the Huang and the König group, [7c,d] we decided to further expand the scope of these protocols.For this purpose, we adapted the synergistic system reported by Sorensen, [7a] for the preparation of challenging linear enamides and enecarbamates (Figure 1e).
7d] Unfortunately, only trace amounts of the desired product 1 b were observed.Inspired by a recent report from the Sorensen group, [7a] we hypothesized that photoexcitable TBADT could act as a HAT agent to form radical intermediate II.The resulting reduced form of TBADT is oxidized by a Co(II) complex, thus closing the catalytic cycle.We found that with 2 mol% of TBADT as the photocatalyst, and 5 mol% of Co(dmgH) 2 (pyr)Cl (COPC) as the co-catalyst in acetonitrile (50 mM) under 395 nm blue LED, the desired dehydrogenated product 1 b was formed in 9% yield after 24 h at 25 °C.Encouraged by this result, we decided to further optimize the reaction conditions.
Different photocatalysts were screened and only decatungstate based photocatalysts were active (Table S1, supporting information).The reaction time was optimized, and 6 h was found to be ideal since extending the reaction time did not further increase the product yield (Table S2, supporting information).Next, the light source, solvent, co-catalyst, and catalyst loading were tested (Table S3-S7, supporting information).A high loading of TBADT was essential for achieving a high product yield in short reaction times.Reducing the catalyst loading leads to a notable increase in reaction time.As the olefinic products slowly degrades when irradiated for longer periods of time (as seen by the decrease in yield after 6 h in Table S2) it is important to complete the reaction in a short time.Gratifyingly, after these optimizations, the yield of the desired dehydrogenated product 1 b increased to 71% yield in presence of 10 mol% TBADT, 5 mol% cobalt benzoate and 25 mol% of dimethylglyoxime (dmgH 2 ) in acetonitrile (50 mM) under 395 nm blue LED for 6 h at 25 °C (Table 1, entry 1).No reaction occurred in the dark either at room temperature or at 50 °C (Table 1, entries 2-3) confirming a visible light mediated transformation.Similarly, no product was observed in the absence of the photo-or cobalt-catalyst (Table 1, entries 4-5).Only trace amounts of the product were obtained when the reaction was carried out without adding dmgH 2 (Table 1, entry 6), suggesting the critical role of the dmgH 2 ligand framework in the hydrogen evolution process. [14]Furthermore, a higher catalyst loading of 15 mol% of TBADT (Table 1, entry 7) did slightly decrease the product yield.
Inspired by our recently published desaturation protocol for N-heterocycles, [7d] several linear structural analogues were prepared for attaining equivalent transformations (Scheme 1).Accordingly, the optimized catalytic system was initially tested with tertiary Nsubstituted scaffolds featuring an electron-withdrawing carbonyl group at the desaturation site.
Hereby, the presented methodology delivered enamide 1 b and N-protected enamines (2 b-8 b) as the corresponding desaturation products in moderate to good yields (30-71%).In addition, the structure of 7 b in the crystal was determined by X-ray analysis. [15]otably, enamide 1 b afforded the best result (71% yield), presumably due to the high electron-withdrawing ability of its aromatic moiety which can stabilize the C-centered radical formed at α-position to the nitrogen atom. [16]Although the aromatic carbamate moiety in 8 b seemed to have a similar effect on the reaction yield (64%), no clear correlation between reactivity and structural or electronic properties could be identified.
With these preliminary results in hand, the methodology was applied to secondary amides, furnishing several to date unreported enamides (9 b-17 b) in satisfactory yields (49-77%).A possible electronic effect on the aromatic amide moiety was investigated.For this, various para-substituted aromatic amides were prepared displaying different EDGs (9 a-12 a).However, no clear trend was observed, as the reaction yield did not show a direct correlation with the variation in the electronic properties of the tested substituents (9 b-12 b).The ether group in compound 16 b, and the β-alanine ester substructure featured by 17 b were well tolerated.Interestingly, while for the previously discussed tertiary N-substituted derivatives the E-isomer was exclusively obtained, secondary moieties besides 16 b and 17 b [17] afforded solely the corresponding Z isomer.Even though the E-configuration of these olefinic products is thermodynamically favoured, [18] the predominance of the Z-configuration among secondary enamides suggests that Z-isomer formation might be favoured due to intramolecular hydrogen bonding between the nitrogen and the carbonyl group.The same argument might be valid for explaining the predominant formation of the Z-isomer of 16 b, as it features a methoxy group which can act as a hydrogen bond acceptor.
Aromatic alkylamides proved suitable too, giving the corresponding vinylamides (18 b-25 b) in moderate yields (30-55%).Herein unreported desaturations of relatively simple molecules could be carried out in a mild and straightforward fashion.These results suggest that apart from nitrogen, an additional EWG attached to the aliphatic desaturation site is not essential for promoting the dehydrogenation process.No electronic effect of the aromatic amide moiety was observed (18 b-22 b).However, a certain steric effect was identified, as the reaction yield dropped notably with increasing substitution on the aliphatic C-end of the substrate (23 b and 24 b).In addition, further investigations revealed that α-substituted amides such as 25 a did not render the expected products (25 b in this case).This is presumably due to the increased steric hindrance, as cobaloxime complexes involving tertiary carbon centers are not common.Contrary to the substitution degree of the carbon atoms attached to the desaturation site, the chain length, as observed for products 18 b, 23 b and 26 b has no relevant impact on the reaction yield.Also, for 26 a only the monodesaturation product 26 b was observed.
The reaction scope for the synthesis of biologically relevant scaffolds was also examined.Alatamide, Table 1.a] Entry Deviation from conditions Yield of 1 b (%) [b] 1 no dmgH 2 n.d.c]   Likewise, the desaturation of the methyl ester of the commercial drug bezafibrate, a potent hypolipidemic agent, [20] gave a novel derivative of this drug slightly lower yield than the small-scale reaction (51%).
Based on the successful desaturation of 17 a to 17 b, two β-alanine derivatives were prepared.Herein, the dehydrogenation of an aminoacid ester (36 a) and a dipeptide (37 a) was accomplished.Regarding the dipeptide, the desaturation proceeded chemoselectively as a single desaturation product was isolated (37 b) while the corresponding two-fold desaturation product was only observed in trace amount by GC-MS.
Lastly, three natural products were successfully prepared by the dehydrogenation of their saturated precursors.Alatamide (38 b) for instance, an amide derived from β-phenylethylamine [18,20] could be prepared through a facile two-step approach in moderate yield as a diastereoisomeric mixture (E/Z = 9:1).21b,23] Although our method is highly efficient for the desaturation of secondary aliphatic carbon centers, tertiary C-centers at α-or β-position to the nitrogen (except for 24 b and 31 b) react less efficient, as sterics seem to prohibit the reaction (Scheme S7).In addition, as observed for failed substrates 4-(N-benzyl-N-methylamino)-2-butanone SM-A and 1-(N-benzyl-N-methylamino)-3-pentanone SM-C (for details, please see Supporting Information) that did not deliver the corresponding desaturated products, the protocol is only suitable for substrates displaying electron-withdrawing N-protecting groups such as amides and carbamates with generally high oxidation potentials.Free amines do not yield the desired products as the excited decatungstate anion with an oxidation potential of around + 2.44 V vs SCE acts as an oxidant via SET.24c] To demonstrate the synthetic utility of the products, a post functionalization reaction was carried out.Herein, the E-isomer of natural product 38 b was used for the preparation of oxazole 41 (70%) via an hypervalent-iodine mediated oxidative cyclization (Scheme 2).
Spectroscopic studies, control experiments, and kinetic isotope effect (KIE) studies were performed to investigate the reaction mechanism.An UV-Vis kinetic study was chosen to gain insight into the combined catalyst system.A solution of substrate 1 a and TBADT in acetonitrile was irradiated (Figure S3).After 2 minutes of irradiation, an immediate growth of a broad band (absorbance maximum at 450 nm, 630 nm, and 780 nm), characteristic for the reduced form of TBADT was observed.Further, no significant change was observed when a solution of 1 a and cobalt complex in acetonitrile, and a solution of cobalt complex and TBADT in acetonitrile was irradiated (Figure S4-S5).Additionally, when we irradiated 1 a, TBADT, and the cobalt complex (Figure S6), a peak at 430 nm was observed which indicated the reduction of Co (III) to Co (II).The results suggested that the oxidation of the reduced photocatalyst takes place by single electron transfer to Co(III).The peak at 780 nm indicated the presence of reduced TBADT in the reaction solution.The hydrogen gas was evolved quantitatively as determined by GC analysis of the crude product mixture (Fig. S7-S8).Light "on-off" experiments suggested that continuous light irradiation is required for the reaction to proceed (Fig. S11).
Kinetic isotope effects (KIE) were calculated by estimating the parallel rates of the reaction for substrates 1 a and 1 a-D 5 (Scheme 3, (1)), as well as from competition experiments between 1 a and 1 a-D 5 (Scheme 3, (2)).Small KIE values of the order 1-1.1 units were observed in both cases.These results exclude the β-hydride elimination step as the rate determining step of the reaction.
Based on the mechanistic routes reported in the literature, [7c,14,24-27] we propose a rational mechanism for the photocatalytic dehydrogenation of aliphatic Nhetero acyclic systems as shown in Scheme 4. Photoexcitation of decatungstate anion [W 10 O 32 ] 4À produces a triplet excited state *[W 10 O 32 ] 4À , which abstracts a hydrogen atom from substrate I, producing alkyl radical II and the protonated reduced decatungstate [W 10 O 32 ] 5À H + . [24]Alkyl radical II is subsequently trapped by the Co(II) complex, yielding an alkyl Scheme 2. Synthetic application of Alatamide (E-isomer) for the preparation of an oxazole.Scheme 3. KIE meassurements.

UPDATES asc.wiley-vch.de
Co(III) intermediate III. [24]7c,25a,c,26] The intermediate V engages with another proton to evolve hydrogen gas and releases a Co(III) complex. [14,27]astly, the oxidation of the protonated reduced photocatalyst [W 10 O 32 ] 5À H + takes place by single electron transfer (SET) from the Co(III) intermediate (E 1/2 Co III /Co II = À 0.68 V vs Ag/Ag + in MeCN, E 1/2 [W 10 O 32 ] 4À /[W 10 O 32 ] 5À = À 0.96 V vs Ag/Ag + in MeCN), regenerating both catalysts in their ground states. [28]n conclusion, we report a light-driven acceptorless cooperative HAT method for the dehydrogenation of aliphatic carbamates and amides to the corresponding enecarbamates and enamides.TBADT and a cobaloxime complex have been used as H-abstractor and βhydride elimination agent respectively.The method features short reaction times and exhibits broad functional group tolerance, making it suitable for late-stage functionalization of derivatives.

Experimental Section
For full Experimental Details see supporting information.
0.015 mmol of TBADT instead of 0.01 mmol.n.d.: product not detected.