Radical-Polar Crossover Reactions of Dienylboronate Complexes: Synthesis of Functionalized Allylboronic Esters.

Radical-polar crossover reactions of dienylboronate complexes are applied to the synthesis of functionalized secondary and tertiary allylboronic esters. The transition-metal-free three-component coupling uses readily accessible dienylboronate esters as substrates in combination with various sp3/sp2 carbon nucleophiles and commercial alkyl iodides as radical precursors. In the visible light-initiated radical chain process, two new C-C bonds are formed, and the E-double bond geometry in the product allylboronic esters is controlled with good to excellent selectivity.

Allylboronic esters are highly useful building blocks in organic synthesis.[1] The great value of these reagents lies in their configurational and chemical stability.[2] They have been successfully applied to cross-coupling reactions,[3] but most often they are used as nucleophiles for the stereoselective synthesis of homoallylic secondary alcohols and amines by an allyl transfer reaction to carbonyls and imines.[4] Notably, functionalized allylboronic esters can also participate in tandem reactions with aldehydes,[5] and it was demonstrated that allylboronate complexes can react with a range of other types of carbon as well as heteroatom electrophiles with high γ-selectivity.[6] Although various direct or indirect methods for the synthesis of allylboronic esters have been introduced,[7] the development of novel strategies for their preparation is still of importance.

Along these lines, Morken and co-workers recently used bis(alkenyl)boronate complexes II in palladium-catalyzed three-component conjunctive cross-coupling reactions.[8] Regioselective electrophilic aryl-palladation of the vinyl moiety induces a 1,2-alkenyl migration to provide, after reductive elimination, α-arylmethyl allylboronic esters of type III (Scheme , a).

Scheme 1

Synthesis of Allylboronic Esters in Three-Component Couplings

Our group and the Aggarwal laboratory have recently shown that electrophilic alkyl radicals add efficiently to in situ generated vinylboronate complexes V and that their corresponding radical anions undergo a radical polar crossover step, inducing a 1,2-R-migration to provide secondary and tertiary alkylboronic esters VI (Scheme , b).[9,10] Motivated by these studies, we considered using dienylboronate complexes in radical-polar crossover reactions for the synthesis of α-alkylated/arylated allylboronic esters. Importantly, the starting dienylboronic esters are readily accessible by hydrozirconation,[11] boron-Wittig reaction[12] or by Heck–Mizoroki coupling.[13] We assumed that a carbon radical would selectively add to the δ-position of the diene moiety in boronate complexes of type VIII (Scheme , c). The resulting allyl radical anion should then undergo a radical polar crossover step with concomitant 1,2 aryl- or alkyl-shift from boron to the α-carbon. During the R-migration, the double bond geometry of the targeted allylboronic ester gets installed. A challenge will therefore lie in the control of the E/Z-selectivity. Moreover, for more complex substituted dienylboronate complexes, regioselective δ-addition of the radical must be ensured. This novel transition-metal-free three-component strategy should enable the efficient construction of highly substituted and functionalized secondary and tertiary allylboronic esters IX.

We commenced our investigations by treating the readily prepared γ-methyl substituted dienylboronic ester 1a (see Supporting Information (SI)) with n-butyllithium at 0 °C in Et2O to generate the corresponding dienylboronate complex. After removal of the solvent, the crude ate complex was redissolved in acetonitrile, and 1.5 equiv of perfluorobutyl iodide was added. Visible light irradiation[14] for 18 h afforded the desired perfluorinated allylboronic ester 2a, resulting from regioselective δ-addition in 63% isolated yield with moderate 4:1 E/Z-selectivity (Table , entry 1). The product derived from β-addition was not identified. Increasing the amount of perfluoroalkyl iodide to 2.0 equiv resulted in a slightly higher yield (67%) and enhanced selectivity (5:1, Table , entry 2). A further increase of the amount of perfluorobutyl iodide significantly affected the double bond selectivity without altering the yield. While using 5.0 equiv of iodide resulted in a 9:1 E/Z-selectivity, 10.0 equiv led to an even further improvement of the stereoselectivity to 12:1 (Table , entries 3 and 4). However, increasing the concentration of the reaction mixture did not affect the selectivity, but led to a slightly decreased yield (Table , entry 5). Note that this positively influenced E/Z-selectivity, as a function of the amount of added radical precursor was found to be general and could be observed also in other transformations. Any time dependence of the E/Z-selectivity could be ruled out by varying the reaction time. Considering reaction economy and selectivity, we regarded 5.0 equiv of the radical precursor R–I as optimal (Table , entry 3). For comparative data on reactions conducted with 2.0 equiv of R–I and varied reaction times, we refer to the SI.

Table 1

Optimization Studies Using 1a as Substrate

entrya	C4F9I (equiv)	solvent (mL)	yield (%)b	E/Z (ratio)c
1	1.5	MeCN (2)	63	4:1
2	2.0	MeCN (2)	67	5:1
3	5.0	MeCN (2)	69	9:1
4	10.0	MeCN (2)	69	12:1
5	5.0	MeCN (1)	66	8:1

Reactions were conducted with 1a (0.3 mmol).

Yields provided represent isolated yields.

Selectivity was determined by 1H NMR spectroscopy.

To document the substrate scope, 1a was reacted under optimized conditions, varying the C-nucleophile and also the radical precursor to give the allylboronic esters 2b–2q (Scheme ). We first tested the ate complex derived from 1a and n-butyllithium in combination with various perfluoroalkyl iodides. With the exception of the volatile trifluoromethyl iodide, all other congeners performed well in this sequence, and the resulting perfluoroalkylated allylboronic esters 2b–2d were isolated in good yields and good E/Z selectivities. α-Iodo ethyl esters are eligible C-radical precursors for this cascade, and higher E/Z selectivity was achieved upon increasing the degree of α-substitution in the ester moiety (2e–2g). As expected, low diastereoselectivity was noted for the reaction with α-iodo ethyl propanoate. Replacing the ethyl by a tert-butyl ester group led to slightly lower E/Z-selectivity (2h). The α-iodo-α-difluorinated ester is a suitable C-radical precursor, and 2i was isolated in good yield and high selectivity. The migrating alkyl group can be readily varied by replacing n-butyllithium with i-butyl-, n-hexyl-, and methyl-lithium (2j–2l). Importantly, the cascade also works efficiently for aryllithium reagents, as documented by the successful preparation of the secondary benzylic allylboronic esters 2m and 2n. We next investigated α-iodo acetonitrile and α-iodo isobutyronitrile as radical precursors. Both nitriles worked well, and in analogy to the reaction with the α-iodo esters, the E/Z selectivity was increased by switching to the bulkier α-dimethyl substituted system (from 9:1 to 26:1). Notably, unprotected α-iodo amides are tolerated in the radical-polar crossover reaction, as documented by the successful preparation of 2q.

Scheme 2

Radical-Polar Crossover Reaction of Various Dienylboronate Complexes

Isolated yields.

10.0 equiv of CF3I were used.

dr = 1.4:1.

Reaction was conducted on a 2.5 mmol scale.

2.0 equiv of R2–I was used.

Yield refers to isolated E isomer.

We next varied the dienylboronic ester moiety using n-butyllithium as the nucleophilic reaction component (Scheme ). Starting with the unsubstituted boronic ester 1b, perfluorobutyl iodide, α-iodo isobutyronitrile, and ethyl dimethyl α-iodo ester provided the targeted allylboronic esters 3a–3c in good yields and excellent E/Z selectivities (up to >99:1). The radical induced 1,2-migration also allows the introduction of secondary alkyl chains, as demonstrated by the successful preparation of the allylboronic ester 3d. Replacing the methyl with a γ-ethyl substituent at the diene acceptor leads to lower E/Z selectivity, as shown for the synthesis of α-iodo ester 3e and the nitrile analogue 3f. Of note, γ-substituted tertiary allylboronic esters also could be accessed (3g–3j). We further demonstrated that the reaction also proceeds on a δ-substituted dienylboronate complex, and the resulting boronic ester 3k was isolated as a diastereomeric mixture in moderate yield and excellent E/Z selectivity. The corresponding product resulting from β-addition was not identified showing the high intrinsic δ-reactivity of these radical diene acceptors.

Scheme 3

Radical-Polar Crossover Reaction of Dienylboronate Complexes: Variation of the Diene Moiety

Isolated yields.

dr = 1.4:1.

2.0 equiv of R2-I was used.

dr = 1.7:1.

The suggested mechanism for the radical-polar crossover reaction is presented in Scheme . The cascade is initiated by light-mediated C–I homolysis. The thus generated electrophilic C-radical selectively adds to the δ-position of the dienylboronate complex A, leading to the corresponding allyl radical anion B, which undergoes single electron oxidation by an alkyl iodide to generate the corresponding zwitterion C. Thereby an alkyl radical is cogenerated, sustaining the radical chain. Hence, the radical cascade belongs to an electron-catalyzed process.[15] The zwitterion C further reacts in an ionic 1,2-R-migration to provide the isolated allylboronic ester D. Currently, we cannot fully exclude that the allyl radical anion B reacts via iodine atom abstraction from the alkyl iodide to generate atom-transfer products Ea and/or Eb, which further react in a Matteson-type rearrangement[16] to D. However, a mechanistic experiment, where the dienylboronate complex derived from 1a and n-butyllithium was treated with the Togni reagent,[17] revealed formation of the trifluoromethylated allylboronic ester 2d. Since this reagent cannot undergo an iodine transfer process, we regard the I atom-transfer/rearrangement sequence as less likely for these transformations.

Scheme 4

Proposed Mechanism

To highlight the synthetic potential of the functionalized allylboron reagents, we conducted a series of allylation reactions (Scheme ). The secondary allylboron reagents 2h and 2a were successfully used in highly diastereoselective benzaldehyde allylations[18] (see 4 and 5). In addition, 2a was applied to a Pd-catalyzed γ-selective cross-coupling with chlorobenzene to give 6.[19] To harvest the potential of the ethyl ester functional group, we used boronic ester 2g as substrate for a one-pot allylation/lactonization sequence[20] and obtained 7 and 8 with excellent E selectivity and very good diastereoselectivities. Hence, our method also represents a conceptually novel approach toward biologically valuable δ-lactones.

Scheme 5

Diverse Functionalization of Allylboron Reagents

In the presence of 10 mol % Sc(OTf)2.

In summary, we have developed a transition-metal-free three-component coupling of dienylboronic esters, alkyl/aryllithium compounds, and R–I radical precursors for the synthesis of functionalized secondary and tertiary allylboronic esters with good to excellent E/Z selectivity. The method is based on a visible light-initiated radical-polar crossover reaction of in situ generated dienylboronate complexes and works with a variety of commercial alkyl radical precursors.