BCl3‐Induced Annulative Oxo‐ and Thioboration for the Formation of C3‐Borylated Benzofurans and Benzothiophenes

Abstract BCl3‐induced borylative cyclization of aryl‐alkynes possessing ortho‐EMe (E=S, O) groups represents a simple, metal‐free method for the formation of C3‐borylated benzothiophenes and benzofurans. The dichloro(heteroaryl)borane primary products can be protected to form synthetically ubiquitous pinacol boronate esters or used in situ in Suzuki–Miyaura cross couplings to generate 2,3‐disubstituted heteroarenes from simple alkyne precursors in one pot. In a number of cases alkyne trans‐haloboration occurs alongside, or instead of, borylative cyclization and the factors controlling the reaction outcome are determined.

Benzofurans and benzothiophenes are important structures found in pharmaceutical targets (e.g., desketoraloxifene) and organic materials. [1,2] Theb oronic acid derivatives of these heteroaromatics are desirable as they are bench-stable,h ave low toxicity and are effective in many functional group transformations,i ncluding the ubiquitous Suzuki-Miyaura cross coupling reaction. [3] Ty pically,t he formation of these borylated compounds is achieved via the C À Ho rC À X borylation of the pre-formed heteroaromatic. [3b, 4] An alternative more efficient approach is to form the heteroaromatic scaffold and the CÀBb ond in one pot via the borylative cyclization of alkynes.T his can be mediated by transition metal catalysts [5] or in the absence of ametal catalyst by using strong boron electrophiles. [6] Thel atter approach was pioneered using B(C 6 F 5 ) 3 which on addition to appropriately substituted alkynes led to ar ange of borylated heterocycles, including products derived from aminoboration [7] and oxoboration (Scheme 1). [8] Other catalyst-free cyclitive elementoborations have been reported, albeit to alesser extent, [5,6] with reports of cyclitive thioboration particularly rare. [9] Whilst B(C 6 F 5 ) 3 was crucial in developing metal-free alkyne borylative cyclization it leads to zwitterionic products such as A (Scheme 1). Theuse of these species in subsequent functional group transformations is not established, currently limiting their synthetic utility. [10] Using alternative boron Lewis acids such as BCl 3 to effect borylative cyclization enables the formation of organo-boronic acid derivatives on work-up, [11] and consequentially access to the myriad of already proven transformations.H owever,t his is an underdeveloped approach with demonstrated, modular protocols scarce.T wo notable exceptions are 1) the BCl 3 -induced alkyne borylative cyclization where a( hetero)aromatic moiety is the nucleophile attacking the BCl 3 -activated alkyne (Scheme 2, top left), [12] and 2) the use of B-chlorocatecholborane to produce borylated lactones via cyclitive alkyne oxoboration (Scheme 2, bottom left). [13] Both protocols generate desirable boronic acid derivatives on (trans)esterification, and are complementary to electrophilic iodinative cyclization (which generates organic electrophiles). [14] From these studies key requirements enabling borylative cyclization without metal catalysts can be identified, including that the boron electrophile must:a )bind reversibly to the heteroatomic moiety,a nd b) induce borylative cyclization preferentially to dealkylation reactions (e.g., cyclization occurs prior to OÀRc leavage). Guided by these herein we report our studies into the reaction of BCl 3 with 2-alkynylanilines,a nisoles and thioanisoles,w hich led to the development of as imple new route to important boronic acid derivatives of benzothiophenes and benzofurans.T his route (Scheme 1, bottom right) is catalyst-free and thus distinct to ar ecent cyclitive alkyne oxo-boration report which required Au catalysts (Scheme 1, top right). [15] Our studies commenced by combining equimolar BCl 3 and N,N-dimethyl-2-(phenylethynyl)aniline (1)f or compar- ison with B(C 6 F 5 ) 3 which formed zwitterion A. [7] In contrast to B(C 6 F 5 ) 3 addition of BCl 3 did not lead to ab orylated indole with X-ray diffraction studies revealing it had instead formed 2 (Figure 1), the product from alkyne trans-haloboration. The reactivity disparity between BCl 3 and B(C 6 F 5 ) 3 is attributed to stronger N!Bcoordination with BCl 3 due to the lower steric crowding around boron. Notably, 2 is not the expected product from the direct haloboration of an alkyne with BCl 3 , which would proceed by syn-addition of Cl 2 BÀCl, [16] suggesting 2 is formed by ad ifferent mechanism. Precedence for alkyne trans-haloboration is extremely limited, with compound C (Figure 1), the trans-haloboration/demethylation product from the addition of BBr 3 to o-alkynyl-anisole B anotable exception. [17] With direct haloboration precluded it is possible that the reaction proceeds from the (N,N-Me 2aniline)-BCl 3 adduct by chloride transfer from boron to carbon, related to that calculated for intramolecular alkyne trans-hydroboration. [18] With the formation of C3-borylated indoles disfavored under these conditions due to trans-haloboration the propensity of o-alkynyl anisoles to undergo borylative cyclization was explored. Ther apid formation of C from B clearly indicates that trans-haloboration also is viable with o-alkynylanisoles,h owever, this reaction was proposed to proceed via initial ether demethylation then haloboration. [17] While ether cleavage of anisoles with BBr 3 is well documented, detailed studies into the mechanism are rare, [19] but one recent report calculated that PhOÀMe cleavage is ab imolecular process involving two Me(Ph)OÀBBr 3 moieties. [19a] Thus B may be prearranged to undergo rapid ether cleavage and other oalkynyl anisoles may be less prone to ether cleavage, particularly using BCl 3 instead of BBr 3 .C onsistent with this the combination of equimolar anisole and BCl 3 in DCM at 20 8 8Cr esulted in the formation of as ingle 11 Br esonance at 32 ppm with minimal ether cleavage observed even after 30 h at 20 8 8C( only ca. 2.5 %C H 3 Cl was formed by 1 HNMR spectroscopy). The3 2ppm 11 Bc hemical shift is consistent with an equilibrium between the Lewis adduct and free BCl 3 and anisole.T hus anisole binding to BCl 3 is reversible and ether cleavage is not significant at 20 8 8C, suggesting that alkyne borylative cyclization using BCl 3 is viable.
1-Methoxy-2-(phenylethynyl)benzene (3a)w as cyclized in DCM using BCl 3 (Scheme 3). Ther eaction was rapid (< 5min at 20 8 8C), as indicated by the consumption of 3a along with the generation of CH 3 Cl (d 1H 3.02 ppm) and anew major resonance centered at 51 ppm in the 11 BNMR spectrum, consistent with ah eteroaryl-BCl 2 species.Aminor broad resonance at 14.2 ppm in the 11 BNMR spectrum was also observed. Esterification with pinacol/NEt 3 enabled the isolation of 4a in 56 %yield without column chromatography. No intermediates are observed so detailed discussion of the mechanism is not warranted, although alkyne activation by BCl 3 and cyclization presumably occurs prior to demethylation based on the slow ether cleavage observed on combining anisole and BCl 3 .Itisnoteworthy that anon-linked analogue of B,1,2-bis(2-methoxyphenyl)ethyne,undergoes rapid transhaloboration and demethylation with both BCl 3 and BBr 3 , thus the reactivity disparity between 3a and B is not due to the use of different boron trihalides.
Exploration of the substrate scope revealed that electrondonating and -withdrawing groups on the anisole ring are compatible in certain positions (4b-e). Furthermore,b orylative cyclization is not limited to diarylalkynes with benzyland methyl-substituted alkynes converted to the benzofurans 4fand 4g in good yield, with the structure of 4g confirmed by X-ray crystallography. 4g was also accessible on agram scale and using non-purified solvents under ambient conditions in good yield. Whilst ap henyl group substituted with an electron-withdrawing group para to the alkyne led to the borylated benzofuran in moderate isolated yield (4h), when ester and nitro groups were incorporated into the anisole ring para to the alkyne this led to low conversions to the benzofuran-BCl 2 species (the d 11B 51 ppm is the minor component). Instead a d 11B 15 ppm resonance was the major product with 3i (Scheme 4), whilst for 3j (Scheme 4), where an aphthyl group has been incorporated resulting in an increase in the steric environment around the alkyne,t he major d 11B resonance is centered at 14 ppm. With both these substrates after the addition of BCl 3 the 1 HNMR spectra revealed that minimal CH 3 Cl had formed (consistent with d 11B 51 ppm being am inor resonance). Instead as inglet was observed at 4.56 and 4.49 ppm, respectively from 3i and 3j, more consistent with an intact ArylOMe unit coordinated to aLewis acid. Attempts to isolate the product derived from 3i after esterification with Et 3 N/pinacol led to isolation of the starting alkyne,p resumably due to E2 elimination. The naphthyl derivative 5 was formed as the major product post esterification, with 1 H, 13 C{ 1 H}, 11 BNMR spectroscopy fully consistent with haloboration, af ormulation supported by mass spectroscopy.T herefore to form borylated benzofurans in acceptable isolated yields by BCl 3 -induced borylative cyclization significant bulk around the alkyne and strong EWG in the para position (to the alkyne) of the anisole moiety have to be avoided.
With the substituent effects probed the functional group tolerance of BCl 3 -induced borylative cyclization was further explored using the "robustness screen" methodology; [20] specifically,m onitoring the cyclization of 3b in the presence of various additives.T his revealed that borylative cyclization was not affected by additives containing nitro,v inyl or CF 3 groups (in each case > 80 %ofthe borylated benzofuran was formed with the additive not consumed). However,b enzaldehyde and acetone were not compatible,with the addition of BCl 3 to separate reactions containing these additives and 3b leading to additive consumption and significantly reduced benzofuran formation. Other Lewis basic groups were compatible with borylative cyclization provided that > 2equivalents of BCl 3 was used, with the first equivalent of BCl 3 coordinating to the Lewis basic group (in each case > 70 %conversion to the borylated benzofuran was observed in the presence of atertiary amine,atertiary amide,apyridine and an itrile). Established routes to 3-borylated-2-organobenzofurans generally proceed from 3-halo-2-organo-benzofurans by metallation/quenching with B(OR) 3 ,o rb yP dcatalyzed Miyaura borylation. [3b] Notably these routes are not compatible with some of the functional groups tolerated by BCl 3 -induced borylative cyclization (e.g.,amide/nitrile groups are generally incompatible with metallation). Furthermore, this methodology is complementary to iridium-catalyzed C À Hb orylation which provides C2-or C7-borylated benzofurans. [4] Finally,i tw orth emphasizing that 4a-h are formed at ambient temperature without ac atalyst using inexpensive BCl 3 ,i nc ontrast the previous borylative cyclization route to C3-borylated benzofurans required pre-installation of the borane (using NaH/CatBCl), Au catalysis,r aised temperatures and ! 20 h. [15] Multiple borylative cyclizations also proceed with appropriately substituted diynes,w ith 6 converted to 7,adiborylated diaryl-benzo[1,2-b:4,5-b']difuran, in excellent yield using BCl 3 (Scheme 5). 7 represents av ersatile precursor to 2,3,6,7-tetraarylbenzo[1,2-b:4,5-b']difurans which are of interest as hole transport materials. [2] To the best of our knowledge 3,7-diborylated benzodifurans have not been previously reported.
While the purified borylated benzofurans reported herein are effective in Suzuki-Miyaura cross couplings (e.g., 4g with 4-bromo-toluene) to enhance the utility of this methodology ao ne-pot borylative cyclization/Suzuki-Miyaura cross coupling procedure was developed (Scheme 6). This does not require isolation of the borylated benzofuran, instead the benzofuran-BCl 2 product is hydrolyzed in situ to the boronic acid and then subjected to conventional Suzuki-Miyaura cross coupling conditions.T his one-pot procedure is asimple and rapid way to generate 2,3-disubstituted benzofurans from simple alkynyl precursors in good yield (72 %isolated yield of 8).
o-Alkynyl-thioanisoles and BCl 3 were explored next to assess if BCl 3 induced borylative cyclization was possible via alkyne thio-boration. Firstly,equimolar thioanisole and BCl 3 were combined which led to as pecies with d 11B 7.9 ppm, indicating significant adduct formation, but importantly no SÀ Me cleavage was observed. Furthermore,p revious work has shown that thioanisole-(BH x Cl 3Àx )(x = 1or2)compounds are effective hydroborating agents at 20 8 8Ci ndicating that an electrophilic borane is accessible from these Lewis adducts. [21] Therefore BCl 3 was added to methyl(2-(phenylethynyl)phenyl)sulfane (9a)i nD CM with in situ 11 BNMR spectroscopy revealing one major product had formed with ab road resonance centered at 4ppm, which does not correspond to a3 -BCl 2 -benzothiophene species (expected d 11B ca. 52 ppm). [22] This is consistent with no chloromethane being observed in the 1 HNMR spectrum. Methylsulfonium cations are significantly weaker methylating agents (less prone to Me + transfer to nucleophiles) than methyloxonium cations, [23] therefore we surmised that the major compound is the zwitterion 10 a analogous to A (Scheme 7). In our hands crystalline material of 10 could not be isolated therefore support for this assignment was provided by combining 9a with BCl 3 (to form 10 a)a nd then adding Et 3 Na sas tronger nucleophile to induce demethylation. This led to formation of [Et 3 NMe] + (by 1 HNMR spectroscopy) and an ew major broad 11 Bresonance at 6.3 ppm attributed to the product from demethylation of 10 a by Et 3 N. On addition of one equivalent of AlCl 3 this compound was then converted to an ew major species displaying ab road 11 Br esonance at 52.9 ppm fully consistent with ab enzothiophene-BCl 2 compound. [22] The same boron species is formed by initial addition of AlCl 3 to 10 a followed by Et 3 N. Esterification of the d 11B 52.9 ppm species with excess pinacol/Et 3 Nl ed to the desired product 11 a in good isolated yield (68 %), unequivocally confirming that borylative cyclization has taken place.T his reaction is notable as arare example of cyclitive alkyne thioboration. [9c] It should be noted that attempts to directly esterify the zwitterion 10 a led to significantly lower isolated yields of 11 a (38 %). This is attributed to 10 a having agreater propensity to undergo protodeboronation due to the more nucleophilic anionic benzothienyl-BCl 3 moiety (relative to benzothienyl-BCl 2 ).
With the functional group tolerance already assessed in benzofuran formation other thioanisole substrates were selected to assess if alkyne haloboration was ac ompetitive pathway.A st here was no evidence (in situ or post work-up) for haloboration with 9a bulkier substituents,n aphthyl and mesityl, 9b and 9c,r espectively,w ere incorporated into the alkyne.A ddition of BCl 3 to these alkynes resulted in similar outcomes to that observed with 9a with no evidence for haloboration in either case,suggesting it is not acompetitive reaction with thioanisoles.A gain, the isolated yield of the benzothiophene pinacol boronate ester is higher on addition of Et 3 N/AlCl 3 prior to esterification (e.g., for producing 11 b yield = 48 %direct from the zwitterion 10 b whereas it is 73 % on esterification after addition of Et 3 N/AlCl 3 ). To demonstrate further that this methodology allows access to otherwise challenging to synthesize boronic acid derivatives 11 d was produced in 55 %i solated yield. Compound 11 d is not readily accessible by established borylation routes commencing from 2-(thiophen-3-yl)benzo[b]thiophene (e.g., Ir-catalyzed borylation and halogenation/lithiation approaches would all proceed at the thienyl alpha position). [3b,4] In conclusion, two distinct reaction pathways operate on addition of BCl 3 to arylalkynes possessing ortho EÀMe (E = NMe,Oor S) moieties,specifically borylative cyclization and trans-haloboration. Thel atter occurs with N,N-dimethyl-2-(phenylethynyl)aniline whilst all the o-alkynyl-thioanisoles studied react selectively by borylative cyclization. For oalkynyl-anisoles both pathways are observed, with borylative cyclization dominating provided strong electron-withdrawing groups on the anisole moiety para to the alkyne,orsignificant steric bulk are absent. This methodology is asimple,scalable and metal-free route to useful benzofuran and benzothiophene boronic acid derivatives,m any of which would be challenging to access by other established borylation methodologies.