Lewis Acid Catalyzed Transfer Hydromethallylation for the Construction of Quaternary Carbon Centers.

The design and gram-scale synthesis of a cyclohexa-1,4-diene-based surrogate of isobutene gas is reported. Using the highly electron-deficient Lewis acid B(C6 F5 )3 , application of this surrogate in the hydromethallylation of electron-rich styrene derivatives provided sterically congested quaternary carbon centers. The reaction proceeds by C(sp3 )-C(sp3 ) bond formation at a tertiary carbenium ion that is generated by alkene protonation. The possibility of two concurrent mechanisms is proposed on the basis of mechanistic experiments using a deuterated surrogate.

The catalytic preparation of quaternary carbon centers using C(sp3)−C(sp3) bond‐forming reactions has been identified as a key challenge in organic synthesis.1 Recent progress in this area has mainly centered on transition‐metal‐catalyzed transformations.2, 3, 4 Metal‐free approaches are less common.5, 6 One strategy to achieve C(sp3)−C(sp3) bond formation is the attack of a carbon nucleophile onto a tertiary carbenium ion.7, 8 Although this has been exploited in the context of carbenium ion generation by dehydration of tertiary alcohols with catalytic amounts of strong acid,5, 6, 9 the complementary process by catalytic protonation of alkenes is far less well explored.10, 11

Our laboratory has been investigating cyclohexa‐1,4‐diene‐based surrogates of difficult‐to‐handle compounds for metal‐free transfer reactions.12, 13, 14, 15 As part of this program we previously developed cyclohexa‐1,4‐dienes 5 and 6 as surrogates of isobutane gas and reported their use in the transfer hydro‐tert‐butylation of alkenes using the strong boron Lewis acid B(C6F5)3 (Scheme 1, top).14, 16 Hydride abstraction from the bisallylic position of the surrogates led to formation of the tert‐butyl‐substituted Wheland intermediates 7 + and transfer of the electrofugal tert‐butyl group to the terminus of 1,1‐diaryl‐substituted alkenes. Borohydride addition to the resulting benzylic tertiary carbenium ion delivered formally anti‐Markovnikov alkylation products such as 2. However, this process was hampered by side reactions to give 3 and 4, and the substrate scope was quite limited.

Scheme 1

Cyclohexa‐1,4‐diene‐based surrogates of isobutane and isobutene gas for metal‐free transfer hydroalkylation and ‐allylation, respectively.

We wondered whether an appropriately substituted cyclohexa‐1,4‐diene would facilitate abstraction of a nucleofugal hydrocarbon group and then lead to the complementary Markovnikov hydroalkylation.17 In seeking a plausible hydrocarbon unit, we were interested in a report by Ménard und Stephan who had achieved the stoichiometric C−H activation of isobutene with the frustrated Lewis pair tBu3P/B(C6F5)3 (Scheme 1, middle).18 The methallyl borate 8 − formed represents a potential nucleophilic source of the methallyl group,19, 20 and we hoped that a cyclohexa‐1,4‐diene surrogate of isobutene gas could be developed to allow for the abstraction of the methallyl nucleofuge by B(C6F5)3 to give 8 − and a Brønsted acidic Wheland intermediate. Interception of this complex by an alkene might then allow for transfer hydromethallylation and the formation of a quaternary carbon center by way of C(sp3)−C(sp3) bond formation (Scheme 1, bottom).21, 22, 23

The required surrogate would likely necessitate a quaternary center adjacent to the methallyl group to prevent undesired side reactions such as transfer hydrogenation,24 as well as to help stabilize the positively charged Wheland intermediate. Surrogates 9 and 10 fulfill this requirement and were easily accessible by Birch alkylation from benzoic acid and biphenyl, respectively (see the Supporting Information for experimental details). We then subjected the surrogates to a model reaction with 1,1‐diphenylethene (1 a) and 10 mol % B(C6F5)3 in CH2Cl2 (Table 1, entry 1). Benzyl ether based surrogate 9 did not give any of the desired hydromethallylation product 11 a, instead providing indane 12 a in 6 % yield; 12 a presumably results from the intramolecular Friedel–Crafts alkylation of 11 a.25 Biphenyl‐based surrogate 10 was more reactive, forming the intended product 11 a in 3 % and indane 12 a in 40 % yield (entry 2). Switching the solvent to toluene improved the selectivity for 11 a (entry 3). We eventually established that electron‐rich styrene derivatives allowed the reaction to proceed to full conversion, and we could suppress cyclization to indanes 12 by modifying the steric environment of the alkene. Using para‐anisyl‐substituted alkene 1 b, we were initially able to form 11 b in 82 % yield with no indane 12 b observed (entry 4). The loading of both surrogate 10 and B(C6F5)3 could be reduced with no detrimental impact on the yield, and 11 b was then isolated in 82 % yield (entry 5). The reaction could also be scaled up to 1.00 mmol (entry 6). No reaction occurred in the absence of the catalyst (entry 7), but interestingly, B(C6F5)3 was able to partially decompose surrogate 10 to biphenyl in the absence of the alkene starting material (entry 8).

Table 1

Optimization of the B(C6F5)3‐catalyzed transfer hydromethallylation.[a]

Entry	Alkene	Surrogate	B(C6F5)3 [mol %]	Solv.	Yield of 11 [%][b]	Yield of 12 [%][b]
1	1 a	9	10	CH2Cl2	<1	6
2	1 a	10	10	CH2Cl2	3	40
3	1 a	10	10	PhMe	32	11
4	1 b	10 [c]	10	PhMe	82	<1
5	1 b	10	5.0	PhMe	85 (82)[d]	<1
6[e]	1 b	10	5.0	PhMe	74[d]	<1
7	1 b	10	–	PhMe	<1	<1
8[f]	1 b	10	5.0	PhMe	<1[f]	<1

[a] Unless otherwise noted, reactions were performed on a 0.10 mmol scale with 1.3 equiv 9 or 10 in 0.25 mL (0.4 m) of the indicated solvent. [b] Determined by 1H NMR spectroscopy by the addition of 1,2‐dibromomethane as an internal standard. [c] 1.5 equiv surrogate 10 used. [d] Yield of isolated product. [e] On a 1.0 mmol scale. [f] Alkene 1 b was not added. 22 % conversion of surrogate 10 to biphenyl.

–

With optimized conditions in hand, we proceeded to investigate the substrate scope (Scheme 2). A range of electron‐rich α‐substituted styrenes could be used, with benzyl ethers as in 11 c and cyclic ethers as in 11 d both tolerated. The indane‐based alkene 11 e could be prepared in 92 % yield and, using this framework, an exocyclic trisubstituted alkene could also be reacted to give 11 f in 92 % yield. As previously observed in other transfer reactions, tetrasubstituted alkenes were unreactive.12, 13 Products bearing two contiguous quaternary carbon centers could be accessed, with 11 g formed in 94 % yield. Here, slightly higher loadings of B(C6F5)3 and surrogate 10 were required to push the reaction to completion. These conditions found further application with other less reactive substrates. Increased substitution at the α‐position of the styrene was also tolerated, as in 11 h–j, although in these cases cyclization to indanes 12 h–j could not be completely suppressed; the proportion of indanes 12 h–j increased with the size of the α‐substituent. Performing the reaction with two equivalents of surrogate 10 did help to reduce the extent of cyclization; 11 h and 12 h were, however, formed in a 3:1 mixture with 1.3 equiv surrogate 10. A substrate bearing the bulky tert‐butyl group in the α‐position was unreactive. Cyclic alkanes containing quaternary carbon centers could also be synthesized using this methodology, with cyclohexane 11 k and cycloheptane 11 l being formed in 89 % and 95 % yield, respectively. Finally, a series of compounds bearing pendant aryl groups was prepared. Phenyl‐substituted 11 m was formed in 85 % yield, and products with halogen substituents, as in 11 n and 11 o, as well as an electron‐donating methoxy substituent, as in 11 p, could also be accessed. To illustrate the utility of these products, a selection of these compounds were further derivatized (see the Supporting Information for details).

Scheme 2

Scope of the B(C6F5)3‐catalyzed transfer hydromethallylation of alkenes. Method A: B(C6F5)3 (5.0 mol %), surrogate 10 (1.3 equiv); Method B: B(C6F5)3 (5.0 mol %), surrogate 10 (2.0 equiv); Method C: B(C6F5)3 (7.5 mol %), surrogate 10 (2.0 equiv).

To gain insight into the reaction mechanism, we prepared deuterated surrogate α‐10‐d 2 (Scheme 3; see the Supporting Information for experimental details). Although this surrogate was deuterated solely at the allyl terminus (γ‐position), its reaction with alkene 1 b produced 11 b‐d 2 with deuterium incorporation in both the α‐ and γ‐positions. This suggests the presence of at least two concurrent mechanisms (see the Supporting Information for more elaborate catalytic cycles). First, based on the work of Ménard and Stephan,18 we propose that alkene attack from surrogate α‐10‐d 2 to B(C6F5)3 results in abstraction of the methallyl group and formation of α‐deuterated methallyl borate complex [γ‐8‐d 2]− along with protonated biphenyl [H⋅C12H10]+ (Scheme 3, Pathway 1). Protonation of alkene 1 b by the Brønsted acidic Wheland complex yields the tertiary carbenium ion 13 + with concomitant aromatization of the surrogate core to biphenyl. Transfer of the methallyl group from methallyl borate [γ‐8‐d 2]− to 13 + then provides γ‐deuterated alkene α‐11 b‐d 2 with formation of a C(sp3)−C(sp3) bond. Alternatively, direct transfer of the methallyl fragment from α‐10‐d 2 to carbenium ion 13 + would result in the formation of the regioisomer γ‐11 b‐d 2 with deuteration in the α‐position (Pathway 2). In this scenario, B(C6F5)3 operates as an initiator and methallyl borate [γ‐8‐d 2]− as a spectator counteranion. The dual role of B(C6F5)3 as catalyst and initiator in transfer chemistry has previously been discussed on the basis of computational calculations.26 Another pathway to arrive at γ‐11 b‐d 2 is the transfer of the methallyl group between two boron centers (see Pathway 3 in the Supporting Information). This would lead to γ‐deuterated α‐8‐d 2, and attack of the methallyl group onto carbenium ion 13 + then provides γ‐11 b‐d 2. It has not yet been possible to distinguish between the two possible pathways leading to γ‐11 b‐d 2.

Scheme 3

Mechanistic experiment with deuterated surrogate and proposed mechanism.

To summarize, a cyclohexa‐1,4‐diene‐based surrogate of isobutene gas has been developed and utilized in the transfer hydromethallylation of electron‐rich styrene derivatives. The method enables the catalytic formation of sterically congested quaternary carbon atoms and represents a rare example of the formation of C(sp3)−C(sp3) bonds from carbenium ions that have been formed by the protonation of an alkene. A range of different scaffolds could be incorporated, and the utility of the products was demonstrated by their derivatization.

Conflict of interest

The authors declare no conflict of interest.

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer reviewed and may be re‐organized for online delivery, but are not copy‐edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

Supplementary

Click here for additional data file.