A Mild and Regioselective Route to Fluoroalkyl Aromatic Compounds via Directed Cycloaddition Reactions.

The synthesis of perfluoroalkyl-substituted (hetero)arenes by benzannulation strategies is complementary to ring functionalization approaches as it obviates the need for pre-existing functionality and innate regiocontrol. We report a mild and regiospecific boron-directed benzannulation method as a vehicle for accessing a range of perfluoroalkyl-substituted (hetero)aromatic building blocks that can be readily elaborated through established C-B bond functionalization processes.

Introduction

Organofluorine compounds are widely established high-value materials because of their unique chemical and physical properties.[1] For example, perfluoroalkyl chains can impart impressive thermal and chemical stability, and for this reason, such compounds have found application in numerous fields of material science. Of particular importance are CF3-substituted (hetero)aromatic compounds, and these are ubiquitous among marketed medicines because of their favorable physicochemical properties (Figure ).[2]

Figure 1

Prominent bioactive trifluoromethylated aromatic compounds.

Broadly speaking, there are three general approaches to incorporating trifluoromethyl groups into (hetero)aromatic compounds. “Programmed trifluoromethylation” is a popular approach that exploits a pre-existing functional handle, such as a (pseudo)halide or boronate, to deliver the CF3 group to a precise location on the substrate.[3] An alternative strategy is “innate trifluoromethylation” of a C–H group, typically through the reaction of the parent (hetero)arene with a trifluoromethyl radical.[4] A final strategy that has received relatively little recent attention is (hetero)benzannulation using one or more CF3-substituted precursors. Specifically, cycloaddition reactions of this type are complementary to the two strategies outlined earlier because the final position of the CF3 is dictated by neither the presence of existing functional groups nor by the innate preference of the parent (hetero)arene. However, a drawback is that these reactions typically require harsh conditions and deliver products with poor regiocontrol.[5] We report herein that boron-directed cycloadditions[6] allow rapid and regiocontrolled synthesis of fluoroalkyl-substituted (hetero)arenes under mild conditions to deliver products that can be further elaborated through the C–B bond (Scheme ).

Scheme 1

Strategies for the Synthesis of Fluoroalkyl-Substituted Arenes

Results and Discussion

We began our studies by devising an efficient synthesis of the required perfluoroalkyl-substituted alkynyl trifluoroborate salts. We were interested in pursuing a route that avoided the use of glassware-etching substances such as HF or KHF2 and were attracted to the work of Ramachandran[7] that employed the hydrofluorocarbon R-245fa (1,1,1,3,3-pentafluoropropane) as a convenient trifluoromethylacetylide precursor. In addition to reproducing this route, we were able to extend this approach to commercial perfluoroalkyl chain-substituted terminal alkynes to produce a small family of alkyne substrates 1a-c (Scheme ).

Scheme 2

Synthesis of Fluoroalkyl Trifluoroborate Salts

Turning our attention to the arene forming step, we were disappointed to find that subjecting pyridine-substituted 2-pyrone 2a to alkyne 1a in the presence of BF3·OEt2 in CH2Cl2 at 40 °C resulted in very low conversion to the corresponding difluoroborane 3a (Table , entry 1). Upon changing the solvent to 1,2-dichoroethane and heating the reaction at 80 °C, 100% conversion was achieved (Table , entry 2), providing a mixture of products 3a, 4a, and 5 that were characterized by X-ray crystallography. Changing the solvent to toluene provided a marginal improvement in the yield of 3a, but a significant amount of byproduct 4a persisted (entry 3). Attempts to converge this mixture to a single product by disproportionation (treatment with BF3·OEt2 to generate 3a or with a combination of BF3·OEt2 and 1a to generate 5) failed to bring about a change in composition (see the Supporting Information for more details). We next investigated the use of a stronger Lewis acid in BCl3[6d] (Table , entry 4) and were pleased to find that a vigorous reaction took place at room temperature in 30 min to deliver the dichloroborane 3b which was isolated in 92% yield. BBr3 was also successful in promoting the reaction (Table , entry 5), affording the dibromoborane 3c in 60% yield after subsequent purification. We attribute the lower yield in this case to the propensity of this compound to undergo hydrolysis to the corresponding boronic acid.

Table 1

Optimization of the Boron-Directed Cycloaddition

entry	solvent	T (oC)	Lewis acid	Y	3aa	4aa	5a
1	CH2Cl2	40	BF3·OEt2 (3.3 equiv)	F	5%
2	DCE	80	BF3·OEt2 (3.3 equiv)	F	64% (34%)b	27% (29%)b	9% (9%)b
3	toluene	80	BF3·OEt2 (3.3 equiv)	F	63%	29%	6%
4	CH2Cl2	20	BCl3 (1.1 equiv)	Cl	100% (92%)b
5	CH2Cl2	20	BBr3 (1.1 equiv)	Br	82% (60%)b

Yield estimated by 1H NMR spectroscopy.

Yields in parentheses are of isolated products. DCE: 1,2-Dichloroethane.

With this set of results in hand, we set about exploring the scope of the BCl3-promoted process, and our results are summarized in Scheme . The synthesis of 3b could be conducted on gram scale with only a small diminution of yield. Perfluorohexyl-substituted alkynyl trifluoroborate salt 1b was found to undergo the transformation efficiently, providing the expected product 6 in quantitative yield without the need for a subsequent purification step. The corresponding perfluorooctyl-substituted salt 1c also underwent the expected reaction but proceeded to only 85% conversion, affording the product 7 in 57% yield after crystallization. In this instance, the low solubility of 7 in CH2Cl2 resulted in significant precipitation during the reaction, which hampered stirring and probably contributed to the drop in conversion. With respect to the directing group, a selection of substituted pyridines were tolerated in the reaction, providing the products 8–11 in high yield, with the exception of 10 which proceeded in lower conversion, presumably due to the sterically demanding bromide. Thiazole-based analogues 12–15 were also generated in excellent yield, although the reactions to form 12 and 15 were noticeably more sluggish for reasons that are unclear. Likewise, oxazol-4-yl-substituted product 16 was generated in excellent yield after gentle heating. Finally, amides also successfully promoted the arene-forming reaction, although, in this case, products 17 and 18 were not isolated as the expected dichloroboranes, but the corresponding boronic acids.

Scheme 3

Scope of the Boron-Directed Cycloaddition

Reactions carried out on 0.11 mmol of pyrone except where noted. bReaction carried out on 0.66 mmol of pyrone 2a. cReaction carried out on 4.30 mmol of pyrone 2a. dReaction stirred at 40 °C for 16 h. eReaction stirred at 40 °C for 24 h.

Given that these reactions had the potential to deliver a large and complex mixture of arylboranes substituted with combinations of alkyne, F, and Cl, it was gratifying that the reaction mixtures were generally extremely clean. As shown in Scheme , disproportionation experiments revealed that this was due in part to the efficient exchange of F to Cl in the presence of BCl3 (Scheme , 3a to 3b and 4a to 4b), although the alkyne unit resists transfer in this case (Scheme , <2% conversion of 5). Furthermore, these experiments allowed us to put forward a proposed mechanism for the efficient formation of arene dichloroboranes under these conditions. Fluoride abstraction by BCl3 generates an alkynyl-BF2 intermediate that undergoes halide exchange to the corresponding alkynyl dichloroborane,[8] which then participates in a rapid cycloaddition to generate the observed product. The cycloaddition reaction must out-compete alkyne disproportionation (to generate dialkynyl- and trialkynylboranes) as the products formed by these intermediates do not converge to the corresponding dichloroboranes and would therefore be observed in crude reaction mixtures. We cannot rule out cycloaddition via the initially formed alkynyl-BF2 intermediate (Scheme , dashed arrows), but the fact that BCl3-promoted reactions proceed faster than BF3-mediated cycloadditions suggests that, if this is in operation, it is a minor pathway.

Scheme 4

Investigation of Product Disproportionation Using BCl3 and the Proposed Mechanism

We next explored the suitability of this strategy for the synthesis of heteroaromatic compounds by exploring the boron-directed cycloaddition of alternative substituted heterodienes (Scheme ). In the event, the Carboni–Lindsey reaction of tetrazine 19 took place at room temperature in the presence of TMSOTf to afford the expected CF3-substituted pyridazine 20 in 56% yield. BCl3 successfully promoted the cycloaddition of triazine 21 to generate the corresponding pyridine 22. Finally, pyrazole 24 was generated from the boron-directed cycloaddition of sydnone 23. In this case, the alkynylborane was formed instead of the corresponding dihaloborane analogue, in line with previous findings.[6e] Overall, this study confirmed that fluoroalkyl trifluoroborate salts offer a convenient method to generate a range of fluorinated (hetero)arenes under mild conditions and with complete regiocontrol.

Scheme 5

Accessing Fluorinated Heteroarenes

Our final objective was to investigate the reactivity of the boron handle for further elaboration. As shown in Scheme , efficient conditions for Suzuki–Miyaura cross coupling were uncovered using aryl iodides, affording the corresponding CF3-substituted biaryls 25–27 in good yield. 17 was also converted to the phenol 28 in excellent yield after treatment with H2O2 under mild, basic conditions. Finally, the CF3-substituted benzoxaborole 29 was prepared in useful yield by mild reduction of the amide by NaBH4, highlighting the versatility of the intermediate 17 in the synthesis of low-molecular-weight building blocks.

Scheme 6

C–B Bond Functionalization

In summary, we present the boron-directed cycloaddition as a novel entry into the important and rapidly developing field of fluoroalkyl-substituted (hetero)aromatic synthesis. A mild, BCl3-promoted cycloaddition protocol was discovered, allowing convenient access to a range of fluoroalkyl-substituted benzene derivatives in good to excellent yield. The products obtained were amenable to further manipulation at the B center. Moreover, the directed cycloaddition concept was successfully extended to the synthesis of CF3-substituted heteroaromatic compounds.