Cobaloxime Catalysis: A General Platform for Mizoroki-Heck-Type Reactions and Catalytic Dehydrohalogenations of Unactivated Alkyl Chlorides

Abstract

Unactivated alkyl chlorides are attractive substrates for chemical synthesis that remain underutilized in many fundamental metal-catalyzed transformations owing to their relative stability compared to alkyl bromides and iodides. Herein, we report the surprising ability of a readily accessible, simple cobaloxime to catalyze visible light-promoted Mizoroki-Heck-type reactions and catalytic dehydrohalogenations of unactivated alkyl chlorides. These air and water tolerant reactions proceed efficiently with a range of unactivated alkyl chlorides under mild conditions with excellent functional group compatibility. Our studies highlight the critical role of the electrolyte in facilitating this new reaction manifold of cobaloxime catalysis.

Graphical Abstract

The catalytic constructions of carbon-carbon bonds using unactivated alkyl halides are valuable in chemical synthesis, as exemplified by the many useful cross-couplings with these widely available substrates.¹ Alkyl chlorides are particularly attractive organohalide building blocks owing to several factors including their superior availability,² reduced toxicity,³ and enhanced stability.⁴ Despite these advantages, the low reactivity of unactivated alkyl chlorides is a well-documented challenge in catalysis, and the development of catalytic carbon-carbon bond constructions using these substrates has significantly lagged those involving alkyl iodides and bromides (Figure 1).^2,5,6

Figure 1.

Cobaloxime catalysis for the Mizoroki-Heck-type reactions and dehydrohalogenations of unactivated alkyl chlorides.

A representative challenge in alkyl chloride activation is the development of general Mizoroki-Heck-type reactions using these substrates.⁷ Despite many successful efforts targeting alkyl Mizoroki-Heck-type transformations with a range of catalysts, few systems are successful with unactivated alkyl chlorides.⁸ Early efforts demonstrated the styrylation of alkyl chlorides using either titanium^8a or cobalt^8b catalysis, but required the addition of highly reactive Grignard reducing agents; recently, a similar transformation was formally achieved via a Pd-catalyzed chlorine atom transfer radical addition followed by dehydrohalogenation to enable styrylation.^8c A pioneering early study of palladium-catalyzed Mizoroki-Heck-type cyclizations included unactivated alkyl chlorides, but the scope was strictly limited to primary alkyl chlorides and terminal alkene partners.^8d

We were inspired by the unique reactivity of cobaloxime catalysts to develop a general Mizoroki-Heck-type reaction of unactivated alkyl chlorides (Figure 1).^9,10 Vitamin B₁₂ and related complexes are well known to engage alkyl electrophiles via both S_N2 and single-electron modes of activation.^2,11,12 Moreover, alkyl cobaloximes are established precursors of carbon-centered radicals for alkene additions, and can also furnish alkenes via metal hydrogen atom transfer (MHAT) elimination.¹³ While Mizoroki-Heck-type reactivity involving cobaloxime catalysis is precedented with alkyl iodides⁹ and bromides¹³, there is a conspicuous absence of unactivated alkyl chlorides in prior studies. Herein, we report the successful development of the first general system for the Mizoroki-Heck-type carbocyclization of unactivated alkyl chlorides using simple cobaloxime catalysis promoted by light. These reactions use an inexpensive, commercially available catalyst, and proceed efficiently even in the presence of air and water. The cobaloxime system also enables the catalytic dehydrohalogenation of unactivated alkyl chlorides under mild conditions. Interestingly, our studies indicate that the added electrolyte is essential in facilitating this novel reaction mode of cobaloxime catalysis.

Our studies commenced with the Mizoroki-Heck-type carbocyclization of primary alkyl chloride 1 (Table 1). A simple, inexpensive catalytic system comprised of 5 mol % dichloro(dimethylglyoxime)(dimethylglyoximato)cobalt(III) [CoCl₂(dmgH)₂] was chosen along with zinc as the reductant in DMF. We were particularly mindful of the role that electrolytes can play in generating anionic Co(I),¹⁴ and therefore added 5 equiv of TBACl to the reaction. This catalytic system delivered the carbocyclization product 2 in good yield upon blue LED irradiation (Table 1, entry 1). We were pleased to find that the cobaloxime catalyst outperformed vitamin B₁₂ (cyanocobalamin), which is significantly more expensive and must be stored in the dark (entry 2). Interestingly, the use of LiCl as electrolyte led to a marked decrease in yield, demonstrating that the choice of electrolyte is critical to the catalysis (entry 3). The use of Mn as reductant slightly reduced yield (entry 4). Importantly, a reaction under an air atmosphere also proceeded in moderate yield, highlighting the practicality of the catalytic system (entry 5). A reaction performed at 90 °C in the dark also delivered the product albeit in slightly lower yield, suggesting that irradiation assists, but is not essential to the reaction (entry 6). Likewise, a reaction at ambient temperature was also successful although slightly less efficient (entry 7). Adding pyridine as a potential ligand led to a slight decrease in yield (entry 8). Control experiments performed in the absence of either cobaloxime, TBACl, or reductant resulted in no product formation or conversion of starting material (entries 9–12).

Table 1.

Catalyst system development for the carbocyclizations of alkyl chlorides.^a

entry	variation from standard conditions above	combined yield (%)b
1	none	75
2	Cyanocobalamin instead of [Co(dmgH)2(Cl)2]	23
3	LiCl instead of TBACl	36
4	Mn instead of Zn	68
5	air atmosphere instead of Ar	52
6	no 440 nm LEDs, 90 °C	69
7	ambient temperature instead of 90 °C	66
8	with 30 mol % pyr	69
9	no [Co(dmgH)2(Cl)2]	0
10	no TBACl	0
11	no Zn	0

^aReactions performed with [1]₀ = 0.25 M and run for 16.5 hr.

^bYield determined by ¹H NMR spectroscopy of crude reaction mixture using an internal standard.

Upon identifying a suitable catalytic system for alkyl chloride activation, we surveyed the carbocyclizations of a range of substrates (Table 2). Reactions of primary alkyl chlorides were successful with unactivated di- and trisubstituted alkenes (entries 1–4) to afford five-membered carbo- and heterocyclic products. Notably, the Thorpe-Ingold effect was not required for efficient cyclization (entry 4). Cyclizations of substrates 9 and 11 generated spirocyclic and fused bicyclic products, respectively (entries 5–6). The reaction is not limited to five-membered ring synthesis; primary chloride 13 participated in a 6-exo cyclization to afford piperidine derivative 14 (entry 7). Secondary alkyl chlorides are also useful substrates, as fused oxabicycle 16 was formed with excellent diastereoselectivity (entry 8). Additionally, 7-endo-trig cyclization of chloromethylsilyl ether 17 delivered a seven-membered silyloxycycle (entry 9). Previous studies accessing 18 via a silyl methyl alkyl-Mizoroki-Heck-type cyclization required the preparation of a significantly more reactive iodomethylsilyl ether.^9b,16 We note that an attempted 5-exo cyclization involving a terminal alkene proceeded in low conversion, with reductive cyclization as the only significant product (see Supporting Information for details).

Table 2.

Substrate scope of the cobaloxime-catalyzed Mizoroki-Heck-type carbocyclizations of unactivated alkyl chlorides.^a

^aFor reactions conditions, see Table 1.

^bYield determined by ¹H NMR spectroscopy of crude reaction mixture using an internal standard, with isolated yields given in parentheses.

We next investigated the potential for intermolecular couplings of unactivated alkyl chlorides using the cobaloxime system (Figure 2). The styrylations of both 1-chlorooctane and 8-chlorooctan-1-ol were both successful in moderate yield, delivering styrenes 19 and 20, respectively. The alkyl-Mizoroki-Heck-type cross-coupling is also applicable to secondary alkyl chlorides, as demonstrated by the reaction of N-Boc 4-chloropiperidine affording product 21. These styrylations proceeded with high E:Z selectivities, which have also been observed in other radical additions involving cobaloxime catalysis.^13h,i Overall, these carbocyclizations and cross-couplings clearly indicate the capability of a simple cobaloxime to catalytically activate challenging alkyl chlorides. Interestingly, a recent study involving the electroreductive activation of alkyl chlorides for reductive Giese-type additions to electron-poor alkenes reported a chloro(pyridine)cobaloxime to be ineffective for alkyl chloride activation, requiring a switch to a more reactive cobalt(salen) complex.²

Figure 2.

Mizoroki-Heck-type styrylations of unactivated alkyl chlorides. Yield and selectivity determined by ¹H NMR spectroscopy of crude reaction mixtures using an internal standard and are reported as an average over three reactions.

We hypothesized that the cobaloxime catalytic system would likewise enable facile dehydrochlorinations. While previous catalytic dehydrohalogenations have required more reactive alkyl bromides (and tosylates),¹⁷ West has reported a useful approach to dehydrohalogenation of unactivated alkyl electrophiles using vitamin B₁₂ which extended to several primary alkyl chlorides.¹⁸ Undesired reductions to alkane byproducts were significant (>30%) in these studies, however, limiting overall utility. Moreover, the catalytic system required NaBH₄ as a reductant, which limits substrate scope owing to chemoselectivity issues.

The cobaloxime system indeed unlocks a practical approach to the catalytic dehydrohalogenations of unactivated alkyl chlorides under mild conditions (Table 3). Primary, secondary, and tertiary alkyl chlorides were all excellent substrates in the dehydrohalogenation. Importantly, reductive byproducts were not significant with any substrate examined (<15%); for example, the dehydrohalogenation of 1-chlorooctane (22) delivered 1-octene in a 10:1 alkene:alkane ratio. Notably, the system tolerates the presence of polar functionality, including unprotected alcohols (entry 2). In contrast to prior efforts in cobalt-catalyzed dehydrohalogenation, the reaction proceeds chemoselectively in the presence of carbonyl functionality (entries 3–4); the strong reductants used in alternative systems would pose significant challenges with such substrates.

Table 3.

Cobaloxime-catalyzed dehydrohalogenations of unactivated alkyl chlorides.^a

^aReactions performed with [substrate]₀ = 0.25 M in 10:1 tBuCN:H₂O, 10 mol % [CoCl₂(dmgH)₂], 30 mol % pyr, 2 equiv Zn, TBACl (1.25 M) under air atmosphere and 440 nm irradiation.

^bYields determined by GC unless otherwise noted. Minor amounts of reduction byproducts were observed in a subset of cases, see Supporting Information for details.

^cIsolated yield.

^d90 °C, 0.25M tBuCN, no pyr.

^e60 °C, no pyr.

^fYield determined by ¹H NMR spectroscopy of crude reaction mixture using an internal standard.

We propose that the exceptional ability of this simple cobaloxime system to activate alkyl chlorides under mild conditions is due to an electrolyte-mediated production of a Co(I) nucleophile. We postulate that TBACl functions to facilitate the Co(II)/Co(I) reduction (−1.13 V vs. SCE in MeCN)¹⁹, which does not occur in the absence of electrolyte. A recent report by Stahl and co-workers indicated the critical role electrolyte can play in modulating the thermodynamic potential of Zn as a reductant.²⁰ The effect of the TBACl electrolyte on the catalytic system is clear from the dehydrohalogenation results in Figure 3. In the absence of electrolyte, dehydrohalogenations of primary (22) or secondary (30) chlorides do not proceed. The use of LiCl as electrolyte also led to significantly decreased reactivity, which we hypothesize may be the result of decreased solubility as compared to TBACl. Additionally, it is likely that the nucleophilicity of the cobaloxime is enhanced with the significantly less coordinating [nBu₄]⁺ cation present. The use of non-chloride electrolyte TBAPF₆ was also unsuccessful, consistent with its inferior ability to modulate the Zn reductant.²⁰ In contrast to the primary and secondary alkyl chlorides, some conversion of the tertiary alkyl chloride 32 was observed in all cases.

Figure 3.

Effect of electrolyte on dehydrohalogenations involving primary, secondary, and tertiary alkyl chlorides. See Table 3 for standard conditions.

A plausible mechanism for the Mizoroki-Heck-type cyclization consistent with our observations is shown in Scheme 1. Activation of the primary alkyl chloride substrate proceeds via S_N2 reaction with the Co(I) anion, which following bond homolysis of the alkylcobaloxime generates a carbon-centered radical primed for cyclization. We note that alkylcobaloxime homolysis may be promoted by either light or heat.^10a,13c This was observed during our optimization studies of the cyclization of primary alkyl chloride 1, which does proceed without irradiation (Table 1, entry 6). Alternatively, the intermolecular couplings and dehydrohalogenations of primary alkyl chlorides did not proceed efficiently without irradiation. The dependance of the reaction on light is likely dependent upon the steric environment of the alkylcobaloxime and any carbon-centered radicals formed in the reaction.^12c Upon cyclization, H-atom abstraction by Co(II) delivers the alkyl-Mizoroki-Heck-type product and Co(III)–H. The cobalt hydride can undergo standard hydrogen evolution followed by reduction to regenerate Co(I).²¹

Scheme 1.

Plausible catalytic cycle for the cobaloxime-catalyzed Mizoroki-Heck-type reaction of alkyl chlorides.

The precise mode of activation likely depends upon the identity of the alkyl chloride. We hypothesize that activations of both primary and secondary alkyl chlorides involve the Co(I) anion, as no reactivity of primary or secondary alkyl chlorides is observed in the absence of TBACl (Figure 3). Primary alkyl chlorides likely participate in S_N2-type activation, while activations of secondary alkyl chlorides could proceed via S_N2-type or atom-abstraction pathways.² The dehydrohalogenations would involve very similar reactivity, with the exception of the tertiary alkyl chloride 32, which owing to its reactivity even in the absence of electrolyte suggests a potential alternative activation pathway. A possible mechanism in this case is atom-abstraction involving a Co(II) intermediate, an established reaction mode of cobaloximes.^10a,12c,22

In conclusion, we have developed a general platform for the catalytic activation of alkyl chlorides using a readily available, shelf-stable, and inexpensive cobalt catalyst. The simple cobaloxime unlocks the first general system using a first row metal for Mizoroki-Heck-type carbocyclizations of unactivated alkyl chlorides, and extends to a chemoselective, catalytic dehydrochlorination which proceeds under mild conditions. Key to the present studies was the identification of TBACl as a crucial electrolyte, facilitating the reactivity of a Co(I) nucleophile even in the presence of air and water. We anticipate that this unique catalytic system will prove broadly useful in the catalytic activation of challenging substrates in chemical synthesis.

Supplementary Material

ASSOCIATED CONTENT

Supporting Information. Experimental procedures and spectral data for all new compounds. This material is available free of charge via the Internet at http://pubs.acs.org.