COMMUNICATION Direct Csp-Csp cross-coupling of diarylzinc reagents with benzylic, 1°, 2° and 3° alkyl halides**

In this article, we report on the direct Csp-Csp crosscoupling of diarylzinc reagents with benzylic, 1°, 2° and 3° alkyl halides. Under etherate-free conditions, cross-coupling proceeds rapidly at ambient temperature with benzylic, 2° and 3° alkyl halide substrates without addition of an external catalyst. The Csp-Csp cross-coupling has excellent functional group tolerance and products are isolated in high yields, generally without the requirement for purification by chromatography. This process represents an expedient, operationally simple method for the construction of new Csp-Csp bonds. The efficient construction of new carbon-carbon bonds is of the utmost importance to contemporary molecular synthesis. In particular, the generation of new Csp-Csp bonds represents a significant enabler for the preparation of complex molecules prevalent in both pharmaceutical agents and natural products. Current state-of-the-art methods for the generation of Csp-Csp bonds typically employ alkyl electrophiles, organometallic nucleophiles (based upon organotin, organoboron, organozinc or organomagnesium reagents) and palladium or nickel based (pre)catalysts. Complementary methodologies have also emerged that employ organolithium reagents in palladium catalyzed or transition metal-free approaches to Csp-Csp bond formation. Despite these notable advances, challenges in cross-coupling chemistry still remain including the development of a single system applicable to the cross-coupling of benzylic, 1°, 2° and 3° alkyl halides with electron rich, electron deficient and sterically hindered aryl nucleophiles. In recent years there has been a conscious move towards the development of new cross-coupling methodologies that employ cheaper and more environmentally benign (pre)catalysts. Iron catalysts have achieved impressive advances in this area, particularly in the cross-coupling of alkyl electrophiles with organomagnesium reagents and, to a lesser extent organozinc reagents. Zinc systems able to generate Csp-Csp bonds from alkyl halides and aryl nucleophiles would also be particularly attractive due to the low cost, low toxicity and environmentally benign nature of zinc. Whilst the utility of organozinc reagents in certain catalytic and stoichiometric carbon-carbon bond forming reactions is well established (e.g., as catalysts for the Aldol reaction and in the direct addition of R2Zn to carbonyls, [22] respectively), zinc systems that can activate alkyl halide electrophiles for subsequent cross-coupling are extremely limited. Recent reports have shown that zinc compounds are indeed viable for the cross-coupling of alkyl halides with activated diboron reagents in the absence of additional catalysts in a radical mediated process. However, the direct, or zinc catalysed, coupling of aryl zinc nucleophiles with alkyl halides to form new Csp-Csp bonds is to the best of our knowledge unknown beyond examples using allyl and propargyl halides. Herein, we report on the solvent dependent direct Csp-Csp cross-coupling of diarylzinc reagents with a wide range of benzylic, 1°, 2° and 3° alkyl halides in an operationally simple manner which allows expedient access to a diverse range of carbon based structural motifs. During ongoing investigations into iron catalyzed Csp-Csp cross-coupling transformations and in particular the hydrocarbyl transmetallation step employing arylboronate nucleophiles we became interested in the roles of additives such as diarylzinc reagents in these transformations. Control experiments examining the influence of diarylzinc additives included the stoichiometric addition of diphenylzinc, 1, to a prototypical electrophile, 3-methoxybenzyl bromide, 2, in benzene-d6 without the transition metal catalyst. This produced an unexpected outcome in that within 5 minutes at ambient temperature, a colourless precipitate had separated from solution (presumably PhZnBr) and analysis of the soluble components by H and C{H} NMR spectroscopy revealed the complete consumption of 2 and formation of a single new species, consistent with the Csp-Csp crosscoupled product, 3a (Table 1, entry 1). The assignment of perprotio3a formed from 1 and 2 was also confirmed by GC-MS, ruling out any Friedel-Crafts benzylation of the benzene-d6 solvent. [29] This was an extremely surprising outcome considering previous investigations concerning the stoichiometric arylation of 3-methoxybenzyl bromide with 1 in dioxane afforded essentially no cross-coupled product (3%) after 16 h at 60°C. Table 1. Direct arylation of 2 with 1 in the presence / absence of additives [a] Entry Solvent T(°C) Time Additive (Equiv.) 2 3a 4 1 C6D6 20 5min 0 >99 0 2 C6D6 20 20h THF(10) 78 19 3 3 C6D6 60 20h THF(10) 11 65 24 4 C6D6 20 20h Et2O(10) 50 50 0 5 C6D6 20 20h MTBE(10) 21 79 0 6 C6D6 20 1h THF(1) 85 14 1 7 C6H5Cl 20 20min 0 >99 0 [a] Standard reaction conditions: diphenylzinc (99%) (22.0 mg, 0.1 mmol); 3methoxybenzyl bromide (14.0 μL, 0.1 mmol); solvent (0.8 mL); +/additive (1 or 10 equiv.). The influence of solvent, particularly an excess of coordinating ethereal solvents was subsequently investigated as the likely cause of the disparity between outcomes. The reaction between 1 and 2 performed in benzene-d6 with the addition of THF (10 equiv.), resulted in the drastic suppression of reactivity (entry 2). Even after 20h, only 19% of the cross-coupled product, 3a, was observed along with 2 (78%) and 1,2-bis(3-methoxyphenyl)ethane 4 (3%). Increasing the reaction temperature to 60°C did afford an increase in both cross-coupled and homo-coupled products 3a and 4 (65% and [*] Dr J. J. Dunsford, Dr E. R. Clark, Dr M. J. Ingleson School of Chemistry, University of Manchester Oxford Road, Manchester M13 9PL (UK) E-mail: [email protected] [**] The authors gratefully acknowledge the Royal Society (M.J.I. for the award of a University Research Fellowship) the European Research Council (J.J.D. Grant number 305868) and the Leverhulme trust (E.R.C.). We also acknowledge the Engineering and Physical Sciences Research Council (grant number EP/K039547/1) for financial support and Dr. Alessandro Del Grosso for helpful discussions relating to this project. Supporting information for this article is given via a link at the end of the document.