Borane catalysed cyclopropenation of arylacetylenes

Triarylboranes have gained substantial attention as catalysts for C–C bond forming reactions due to their remarkable catalytic activities. Herein, we report B(C 6 F 5 ) 3 catalysed cyclopropenation of a wide variety of arylacetylenes using donor–acceptor diazoesters. A mild reaction protocol has been developed for the synthesis of functionalised cyclopropenes (33 examples) in good to excellent yields.

Transition metal catalysed C-C bond forming reactions overwhelm the chemical literature. 1 Although the use of precious transition metal catalysts has achieved immense success, metal impurities in the final compounds are often unavoidable. This is particularly significant when considering products taken into the body where toxic metal contamination must be kept to a minimum. Over the last few years, main group-based catalysts have been extensively investigated as substantial alternative to the precious transition metals. 2 More precisely, the Lewis acidic triarylboranes 3 have found multitude applications towards C-C bond forming reactions. 4,5 The presence of empty d-orbitals in transition metals allows them to lend or remove electrons from the coupling partners, and thus can be employed as a catalyst for wide variety of reactions. 6 Likewise, the empty p-orbital of the central boron atom of Lewis acids renders them strongly electrophilic in nature and therefore they can readily react with Lewis bases by accepting a lone pair of electrons. 7 Relating to this, an important initial contribution made by Zhang in 2016, 8 showed that B(C 6 F 5 ) 3 could act as a catalyst for the orthoselective C-H alkylation of unprotected phenols with a-aryl a-diazoesters. The mechanism for this reaction was revealed computationally to be the activation of the diazoester through O -B adduct formation to generate carbenes. 9 Therefore, by using diazoester precursors, a carbene transfer reactions can be carried out using B(C 6 F 5 ) 3 as a catalyst. 10 Carbene transfer reactions are one of the most fundamental reactions in organic synthesis and widespread studies have been conducted to explore the synthetic utility of carbenes for making a variety of novel compounds. 11 Recently, we 12 and Wilkerson-Hill 13 observed that catalytic amounts of B(C 6 F 5 ) 3 enable the cyclopropanation reactions of styrenes (Scheme 1A) using a-aryl a-diazoesters. This exciting outcome motivated us to investigate this reactivity further to see if arylacetylenes could also be used as substrates in reactions with a-aryl a-diazoesters using B(C 6 F 5 ) 3 as a catalyst. This reaction, cyclopropenation, has been largely investigated using precious transition metals, such as Rh, 14 Ir, 15 Ag, 16 and Au. 17 Nonetheless, the use of non-precious transition metals, including Cu 18 and Co 19 (Scheme 1B), have also been reported. Typically, in the presence of a metal catalyst, diazoesters afford a metal-carbenoid species which then readily undergoes a [2+1] cycloaddition with an arylacetylene to form the 3-membered carbocycle. Recently, Koenigs et al. revealed that cyclopropenation of arylacetylenes using a-aryl a-diazoesters is also possible by employing visible light (blue light; 470 nm). 20 We initiated our studies into the cyclopropenation reaction using phenylacetylene (1.3 equiv.) and a-aryl a-diazoester 1a (1 equiv.) as model substrates ( Table 1). The reaction between 1a and phenylacetylene showed no formation of the cyclopropene product (3i) in absence of any catalyst at both ambient temperature and at reflux in CH 2 Cl 2 ( Table 1, entries 1 and 2). Addition of BF 3 ÁOEt 2 as a Lewis acid catalyst (20 mol%) also showed no formation of the desired carbocycle (3i) and only decomposition of the diazo compound into the homocoupled product was observed ( Table 1, entry 3). 40 mol% of the Brønsted acid (TfOH, triflic acid) also failed to promote the reaction ( Table 1, entry 4). When 20 mol% B(C 6 F 5 ) 3 was employed for the reaction, no product formation was observed at ambient temperature, however reaction at 45 1C afforded 3i in 48% yield after 24 h ( Table 1, entries 5 and 6). Switching the solvent from CH 2 Cl 2 to C 2 H 4 Cl 2 slightly improved the yield of 3i to 57% but still did not give satisfactory results (Table 1, entry 7). Increasing the reaction temperature further to 70 1C however was detrimental to the reaction leading to the formation of a complicated reaction mixture and the isolation of the desired product 3i failed ( Table 1, entry 8). Reducing the catalytic loading of B(C 6 F 5 ) 3 from 20 mol% to 10 mol% showed improvement in the yield of 3i to 65%. However, reducing catalytic loadings further to 5 mol% gave lower yields of the desired carbocycle 3i of 32% (Table 1, entries 9 and 10). Additionally, we tested other triarylfluoroborane catalysts for the cyclopropenation reaction and we observed that although 10 mol% (2,4,6-F 3 C 6 H 2 ) 3 B [(2,4,6-Ar F ) 3 [11][12][13].
Interestingly, the yield of 3i was further improved to 75% when we used a slight excess of 1a (1.3 equiv.) ( Table 1, entry  14). Thus our optimised reaction conditions were a 1 : 1.3 stoichiometric ratio of phenylacetylene : 1a and carrying out the reaction in C 2 H 4 Cl 2 at 50 1C using 10 mol% B(C 6 F 5 ) 3 .
Unfortunately, attempts to synthesise 3-membered heterocycles from the insertion of the carbene into CQO, CQN or CRN bonds failed. Using the optimised reaction conditions, the reaction between benzaldehyde and 1c was examined with the goal to produce the corresponding epoxide. However, multinuclear spectroscopic data of the isolated compound confirmed the formation of methyl 3-oxo-2,3-diphenylpropanoate (see ESI †) formed from the homologation of benzaldehyde with the diazo compound (Roskamp-Feng reaction). 21 We propose the mechanism of the cyclopropenation reaction to proceed in a similar manner to that for the cyclopropanation reaction reported in our previous studies 12 (Fig. 2). Initially, the Lewis acidic B(C 6 F 5 ) 3 binds effectively with the ester functionality of the a-aryl a-diazoester 1. This facilitates loss of N 2 forming the highly electron deficient intermediate I and its resonance form I 0 . Subsequently the reactive carbene intermediate can then react with the nucleophilic arylacetylene forming intermediate II. The generation of the carbocationic centre in II explains the need for arylacetylenes in the reaction to stabilise this intermediate. Finally, attack of the boron enolate onto the carbocation in II then generates the product and regenerates the catalyst.
In conclusion, a metal-free mild reaction protocol has been developed for the cyclopropenation of alkynes using diazo compounds. Our studies demonstrate that catalytic amounts of B(C 6 F 5 ) 3 readily react with a-aryl a-diazoesters to promote the carbene transfer reaction when reacted with arylacetylenes generating the desired 3-membered carbocycle. A wide range of substrates were investigated and good to excellent yields of the cyclopropenated products were obtained. This methodology adds to the ever-increasing range of reactions that the Lewis acid B(C 6 F 5 ) 3 can catalyse.
AD and RLM are grateful to the EPSRC for funding and the awarding of an EPSRC Early Career Fellowship (EP/R026912/1). Information about the data that underpins the results presented in this article, including how to access them, can be found in the Cardiff University data catalogue at http://doi.org/ 10.17035/d.2021.0136187455. Fig. 1 Solid-state structure of compound 3n (left) and 3o (right). Thermal ellipsoids drawn at 50% probability. Carbon: black; oxygen: red. H atoms omitted for clarity. Fig. 2 Proposed reaction mechanism for the cyclopropenation reaction.