Allylic Functionalization of Unactivated Olefins with Grignard Reagents

Abstract

Allylic functionalization of unactivated olefins with carbon nucleophiles is a powerful strategy for converting olefins into complex products. We report a general method for functionalizing olefins with aromatic, aliphatic, and vinyl Grignard reagents. In a one-pot process, olefins are oxidized by a commercially available reagent to allylic electrophiles, which undergo selective Cu-catalyzed allylic alkylation with Grignard reagents. Products are formed in high yield and regioselectivity.

The functionalization of unactivated olefins with carbon nucleophiles is an attractive strategy for synthesizing value-added small molecules from inexpensive and abundant components of petrochemical feedstock (Scheme 1).^[1,2] It is quickly becoming an alternative approach to the Heck reaction for the conversion of simple terminal olefins into more complex products.^[3] Elegant examples of allylic alkylations have been reported with stabilized carbon nucleophiles,^[4] tethered carbon nucleophiles,^[5] diazoesters,^[6] and trifluoromethane donors.^[7] To complement these studies, we were interested in functionalizing unactivated olefins with Grignard reagents, which are a readily accessible and affordable class of diverse carbon nucleophiles.

Scheme 1.

Allylic Functionalization of Unactivated Olefins

Herein, we describe a general method for functionalizing unactivated olefins with aromatic, aliphatic, and vinyl Grignard reagents.^[8] This one-pot process is performed by the sequential addition of a stoichiometric oxidant, copper catalyst, and a Grignard reagent to unactivated terminal olefins, which are converted to internal olefins with high E-selectivity. In contrast to most Heck reaction protocols, our strategy exhibits broader scope in the coupling partner and does not require activated olefins for high levels of reactivity and selectivity in the formation of a single olefin isomer.^[9]

Allylic functionalization with carbon nucleophiles usually requires a stoichiometric oxidant to mediate the oxidation state of the metal in the catalytic cycle. We envisioned a conceptually distinct approach, in which the stoichiometric oxidant could oxidize the unactivated olefin in the absence of a catalyst to a reactive intermediate, which would be susceptible to metal-catlayzed allylic subsititution with Grignard reagents in the same reaction flask (Scheme 2). We were drawn to the metal-free allylic oxidation of unactivated olefins 1 with commercially available sulfurdiimide reagent 2. This chemistry was initially explored for the allylic amination of unactivated olefins through the [2,3]-rearrangement of allylic sulfinamide 3.^[10,11] Motivated by early examples of metal-catalyzed couplings of allylic sulfides and sulfoxides with Grignard reagents,^[12] we were interested in exploiting sulfinamide 3 as a new class of electrophiles for metal-catalyzed allylic substitution. We now report the broad substrate scope and unique product selectivity exhibited by this novel strategy in allylic alkylation.

Scheme 2.

General Approach to the Allylic Functionalization of Unactivated Olefins with Grignard Reagents

Our initial challenge in functionalizing olefins with Grignard reagents via sulfinamide 3 was the suppression of the facile [2,3]-rearrangement to allylic amine 5 in favor of the desired coupling with Grignard reagent 6 (Table 1). Sulfinamide 3a was generated upon mixing olefin 1a and sulfurdiimide reagent 2 for 12 h at 4 °C in various solvents. This oxidized intermediate was subsequently cooled to −20 °C and treated with several low valent metals that typically generate electrophilic metal allyl intermediates 7 from allylic acetates, carbonates, and halides. Palladium, iridium, rhodium, and molybdenum catalysts did not yield the internal olefin 4a in the presence or absence of various classes of ligands (entries 1–6). Gratifyingly, copper(I) salts emerged as promising catalysts for this process (entries 7–11).^[13] In the presence of copper(I) thiophene-2-carboxylate (CuTc), internal olefin 4a was formed in 22% isolated yield (entry 7). The source of copper impacted the isolated yield of the desired product (entries 7–9). When CuTc was replaced with CuBr•SMe₂, olefin 4a was produced in slightly greater yield (entry 9). Given the propensity for sulfinamide 3a to undergo a facile [2,3]-rearrangement to allylic amine 5, we hypothesized that the proper selection of solvent for the oxidation step was essential for the efficiency of the overall process. In Et₂O, sulfurdiimide 2 and sulfinamide 3a were largely insoluble, resulting in an inefficient oxidation reaction and subsequent allylic substitution with Grignard reagent 6 (entry 9). In DME, sulfurdiimide 2 and sulfinamide 3a were completely soluble, which caused significant [2,3]-rearrangement to allylic amine 5 before the reaction mixture was treated with copper and Grignard reagent (entry 10). A 1:1 mixture of DME:Et₂O produced the ideal environment for the formation and stability of sulfinamide 3a, which was smoothly converted in situ to internal olefin 4a (entry 11). Under these conditions, the desired product was isolated in 72% yield with high E-olefin selectivity (15:1) and complete regioselectivity.

Table 1.

Optimization of Allylic Functionalization with Grignard Reagents

Entry	MLn	Solvent	Yield (%)
1	Pd2dba3•CHCl3a	Et2O	< 5
2	[Pd(C3H5)Cl]2a	Et2O	< 5
3	[Ir(cod)Cl]2a	Et2O	< 5
4	[Ir(cod)BF4a	Et2O	< 5
5	[Rh(cod)Cl]2a	Et2O	< 5
6	C7H8Mo(CO)3a,b	Et2O	< 5
7	CuTc	Et2O	22
8	CuCl	Et2O	6
9	CuBr•SMe2	Et2O	25
10	CuBr•SMe2	DME	53
11	CuBr•SMe2	DME:Et2O (1:1)	72

Reaction conditions: (Step 1) Olefin 1a (1 equiv), sulfur diimide 2 (1.2 equiv), solvent (0.3 M). (Step 2) DME (0.2 M), ML_n (5 mol%), Grignard reagent 6 (4 equiv).

^[a]Metal complex was pretreated with (10 mol%) phosphine ligands, pyridine ligands, phosphoramidite ligands, or no ligand.

^[b]C₇H₈Mo(CO)₃ = cycloheptatriene molybdenum tricarbonyl.

With optimal reaction conditions in hand for the functionalization of unactivated olefins with Grignard reagents, we explored the substrate scope of this process (Table 2). A diverse range of aliphatic Grignard reagents could be utilized in this transformation without affecting the overall efficiency of the process (Table 2a, 4a–f), including Grignard reagents with trimethylsilyl groups (4e) and remote stereocenters (4f). A hindered t-butyl Grignard reagent furnished the neopentyl internal olefin 4d. When aryl Grignard reagents were employed in the reaction, the in situ conversion of sulfinamide 3 resulted in considerably lower yields, presumably because of side reactions between the highly reactive aryl Grignard reagents and byproducts from the oxidation reaction. Optimal yields were obtained for the coupling of these aryl Grignard reagents after isolating sulfinamide 3 and adding catalytic amounts of radical scavenger TEMPO in the second copper-catalyzed step (4g–i). Notably, products 4g–i formed with high E-allylic selectivity, which is distinct from the mixture of E-allylic and E-styrenyl products typically obtained through indiscriminate β-hydride eliminations under traditional Heck conditions,^[3] or the high E-styrenyl selectivity observed under modified Heck conditions recently reported by the Sigman group.^[14] Regardless of the Grignard reagent utilized in the reaction, we always isolated E-olefins as the major products (>20:1 in most cases).

Table 2.

Substrate Scope

Reaction conditions: (1) Olefin 1 (1 equiv), sulfur diimide 2 (1.2 equiv), DME:Et₂O 1:1 (0.3 M). (2) DME (0.2 M), CuBr•SMe₂ (5 mol%), Grignard reagent (4 equiv).

^[a]Intermediary adduct was filtered and isolated before treating with copper catalyst and Grignard reagent. In addition, 8 mol% TEMPO was utilized in the second step.

We also examined an array of terminal olefins as substrates for coupling with i-butyl Grignard reagent (Table 2b). Unsaturated hydrocarbons furnished internal E-olefins in good yields (4j–l). This reaction was tolerant of several functional groups in the olefin, including alkyl chlorides (4m), silyl ethers (4n), carbonates (4o), and esters (4p). When terminal 1,1-disubstituted exocyclic olefins were employed, we obtained trisubstituted olefins (4q–r).

To highlight the synthetic potential of the functionalization of olefins with Grignard reagents, we synthesized a series of skipped dienes (Scheme 3). Although methods have been developed for the assembly of specific classes of skipped dienes,^[15] the allylic alkylation of unactivated olefins is not considered a general strategy for accessing these structures. This may be due to the difficulty in controlling the reactivity of multiple olefins in the skipped diene product for further undesired allylic alkylations. We recognized that our strategy for functionalizing olefins is uniquely suited for synthesizing skipped dienes, because we can control the reactivity of a single olefin in the presence of other unsaturation through a selective oxidation reaction with sulfurdiimide 2. We coupled terminal olefin 1a with unsaturated Grignard reagents 8 and 10 to furnish skipped dienes 9 and 11, respectively (Scheme 3a). We also converted various dienes 12a–d to mono-alkylation products 14a–d by selectively controlling the mono-oxidation reaction between one olefin in the substrates with one equivalent of sulfurdiimide 2 (Scheme 3b). Finally, we controlled the mono- and di-alkylation between olefin 12c and i-butyl Grignard reagent 15 by altering the relative equivalents of the two substrates, which resulted in the selective synthesis of skipped dienes 16 and 17 (Scheme 3c).

Scheme 3.

Synthesis of Skipped Dienes

[a] Yield calculated relative to sulfurdiimide 2 as the limiting reagent.

We have commenced a mechanistic examination of the copper-catalyzed functionalization of allylic sulfinamide 3 with Grignard reagents due to the distinct reactivity and product selectivity exhibited by this novel class of electrophiles. For example, while most copper-catalyzed Grignard additions to allylic electrophiles favor the S_N2’ pathway to yield branched products,^[16] the allylic alkylation of sulfinamide 3 favored the S_N2 product under all the conditions we examined. In addition, the use of excess Grignard reagent was necessary to obtain a high yield of product (Scheme 4a). Most notably, in the presence of 2 equivalents of Grignard reagent 6, we did not observe product formation, which suggests that Grignard reagents have a multifunctional role in the overall allylic substitution process. For example, one equivalent of Grignard reagent may initially react with allylic sulfinamide 3 prior to allylic substitution.^[12] Another equivalent of Grignard reagent may also react with the sulfonamide-containing byproduct after displacement with the Grignard reagent.^[17]

Scheme 4.

Mechanistic Studies for the Copper-Catalyzed Coupling of Grignard Reagents and Allylic Sulfinamide 3

To probe the presence of free radical intermediates in the copper-catalyzed step, we isolated sulfinamide 3a and subjected this intermediate to the copper-catalyzed allylic substitution conditions with varying amounts of TEMPO. When sulfinamide 3a was coupled with phenylmagnesium bromide, olefin 4g was isolated in 78% yield even in the presence of 2 equivalents of TEMPO, with no trace of adduct 18. This observation suggests that the productive pathway may not proceed by free radical intermediates (Scheme 4b).^[18]

In conclusion, we have developed a one-pot functionalization of unactivated olefins with Grignard reagents. We demonstrated the conversion of terminal olefins to sulfinamide intermediates that are susceptible to highly selective nucleophilic substitutions with Grignard reagents. Overall, the process exhibited a broad substrate scope for unactivated olefins as well as Grignard reagents. The internal olefin products were formed in high yield as predominantly single olefin regioisomers and stereoisomers. Allylic sulfinamide 3 was unique in its ability generate predominantly S_N2 products under all the conditions examined so far, presumably via dialkyl π-allylcopper(III) complexes, which are known to favor the formation of S_N2 products over S_N2’ products.^{[13, 19]} We exploited the controlled activation of olefins in polyunsaturated structures to synthesize a variety of skipped dienes, which are prevalent in natural products and are difficult to synthesize by known methods for allylic alkylation. We are currently expanding this strategy to include allylic functionalization with other nucleophiles to develop a diverse array of chemical transformations.