Synthesis of α-Fluorinated Areneacetates through Photoredox/Copper Dual Catalysis

Abstract

The development of mild and practical conditions for the fluoroalkylation of arenes is an ongoing challenge in chemical organic synthesis. Herein, we report a metallaphotoredox method for the preparation of fluoroalkyl arenes based on the synergistic combination of Ir/Cu dual catalysis from boronic acids. The mild conditions used allows broad functional group tolerance, including substrates containing aldehydes, free phenols, and N-Boc protected amines. Mechanistic investigations support a process proceeding via photoredox/copper dual catalysis.

Graphical Abstract

Taking advantage of the unique characteristics of the fluorine atom and fluorinated groups, the introduction of fluoroalkyl substituents onto aromatic rings is a powerful and widely employed tactic used for the construction of molecules of interest in the pharmaceutical industry to enhance binding selectivity, elevate lipophilicity, and/or circumvent metabolism issues arising from in vivo C-H bond oxidation.^1,2

Despite great achievements in fluorination and trifluoromethylation of organic substrates over the past decade,³ strategies for selective introduction of a difluoromethylene (CF₂) group into organic molecules,⁴ and more particularly, at the benzylic position, have been less explored. Therefore, the development of such methods would represent an important addition to the synthetic toolbox available to practitioners. Indeed, in addition to offering dramatically improved metabolic stability and oral bioavailability of biologically active molecules, these motifs offer several potential downstream transformations, allowing the construction of highly decorated, fluorine-containing molecular scaffolds.

Although significant progress has been made in recent years,⁵ the development of direct (di)fluoroalkylation reactions of arenes is still under-exploited.⁶ Among examples reported recently, many include the use of thermal reactions⁷ catalyzed by transition metals⁸ (palladium, nickel, copper, or even ruthenium), and visible-light-induced C-H activation of electron-rich arenes⁹ (Scheme 1).

Scheme 1.

Fluoroalkylation of Aromatic Compounds

Despite these efficient strategies and promising contributions, many of these examples require expensive transition metal catalysts or fluorination reagents, or sometimes the protocols remain limited in terms of regioselectivity or scope. Thus, it is particularly important to develop an economical, sustainable, and selective method to introduce CF₂ groups into arenes.

To overcome these obstacles, several recent studies have proposed the use of boronic acids as a very accessible and suitable substrate for a copper-mediated arylfluoromethylation of arene derivatives.¹⁰ Specifically, we were interested in the trifluoromethylation protocol of (hetero)arylboronic acids developed by Sanford et al., which merged visible-light photocatalysis and transition metal catalysis (Scheme 1).¹¹ Although this method is effective, it is limited to the use of perfluoroalkyl iodides and para- or meta-substituted arylboronic acids.

Herein, we describe a mild, rapid, and efficient synthesis of aryl(di)fluoroalkylated compounds via photoredox/copper dual catalysis, with broad functional group tolerance. The synthetic utility of this method is demonstrated by providing access to highly attractive and important molecules from commercially available substrates: arylboronic acids as nucleophilic partners, and bromo(di)fluoromethyl derivatives as radical precursors.

Initial efforts exploring this transformation were focused on using copper(II)triflate as a metal source (Table 1). Phenylboronic acid 1a and ethyl bromodifluoroacetate (2a) were selected as model substrates. Despite exploring different photocatalysts, photocatalyst loading, as well as solvent screening, in none of the further explored conditions was it possible to exceed 40% yield of 3a (see Supporting Information). The use of diverse copper(I) sources did not improve the reactivity, except when using copper(I) trifluoromethanesulfonate toluene complex 2:1 [(CuOTf)₂•PhMe, See Supporting Information for additional details]. The use of K₂HPO₄ and 25 mol % of the copper source afforded arene 3a in 55% yield (entry 1 and Supporting Information), with the biaryl dimer 6 from the boronic acid the main by-product. Of note, scaling up the reaction to 0.5 mmole provided 3a in 54% yield (entry 1). Next, control experiments were explored to point out the photoredox/copper dual catalysis nature of this transformation (entries 9 to 12).

Table 1.

Optimization of Reaction Conditions^a

entry	deviation from std condtions	yield of 3a (%)b
1	None	55 (54)c
2	CuBr	5
3	CuCI	n.r.
4	CuOAc	n.r.
5	CuI	n.r.
6	Cu(MeCN)4PF6	15
7	Cu(MeCN)4OTf	40
8	Cu(OTf)2	40
9	no light	n.r.
10	no Ir(ppy)3	n.r.
11	no base	n.r.
12	no (CuOTf)2·C6H5CH3	n.r.

^aReaction conditions: boronic acid 1a (0.1 mmol), 2a (0.2 mmol), K₂HPO₄ (0.2 mmol), Ir(ppy)₃ (2 mol %), (CuOTf)₂·C₆H₅CH₃ (25 mol %) in DMF (1.0 mL, 0.1 M), 16 h irradiation with blue LED strips (λmax = 455 nm).

^bYields were determined by ¹⁹F NMR analysis using (trifluoromethyl)benzene as an internal standard.

^cIsolated yield. std: standard; n.r.: no reaction.

In general, reported syntheses of α-aryl-α,α-difluoroethyl ester 3a require a stoichiometric amount of copper source^10b or high temperatures,^8f while through this photoredox/Cu dual catalysis, 3a could be prepared at room temperature.

Once the optimal conditions were established, a screening of different arylboronic acids was performed using ethyl bromodifluoroacetate as radical precursor (Scheme 2). In general, various aromatic boronic acids bearing either electron-donating groups (EDG: alkyl, ether, sulfide, amide) or electron-withdrawing groups (EWG: methyl ester, cyano, aldehyde, trifluoromethyl) were efficiently incorporated, with protodeboronation a competitive side reaction in the process. As expected, EDGs afforded the desired product if they were located at the para-position (3a-b, 3d-f, and 3h-i), while EWGs must be located at the meta-position (3o-p). Surprisingly, ortho-substituted boronic acids displayed different reactivity, being amenable to both an EDG (3t) and EWGs (3u-v) in moderate yields. Boc-Protected amines were successfully introduced (3g-h, 3n). Various potentially reactive functional groups (free phenol, aldehyde, and ester) were tested, demonstrating the group tolerance and specificity of this metallaphotoredox reaction by exclusively forming the desired difluoro products. Dihalogenated aromatic boronic acids underwent selective difluoroalkylation, with better results being achieved with the meta-regioisomers. Thus, ethyl difluoro acetates 3j-m and 3q-s were formed in moderate to good yields, leaving a halide handle intact for further diversification of their structures by well-established procedures,¹² and showcasing the complementarity of this method to other difluorination procedures.^8–10 Additionally, sterically hindered dibenzo[b,d]furan-4-ylboronic acid was successfully incorporated (3y), as well as multifunctional boronic acids (3w-x). To establish the synthetic utility of this photoredox/copper dual catalysis, ethyl difluoro acetate 3q was synthesized on gram scale from (3-chlorophenyl)boronic acid.

Scheme 2.

Evaluation of Substrate Scope^a

^aGeneral Conditions A. ^bGeneral Conditions B. ^c0.3 mmol scale. ^d0.5 mmol scale. ^e1.5 mmol scale (reaction time 48 h). See Supporting Information for further details.

Related conditions enabled the use of other commercially available, fluorinated analogues [ethyl 2-bromo-2-fluoroacetate and diethyl (bromodifluoromethyl)phosphonate] as substrates. These building blocks are medicinally relevant scaffolds in drug discovery and development,¹³ especially the difluorophosphonate motif, which exhibits better biological properties than its non-fluorinated analog.¹⁴ This photoredox/Cu dual catalysis procedures allows the synthesis of various α-fluoro benzylic ethyl esters (3z-3ac) and α,α-difluoro benzylic phosphonates (3ae-3ag) in moderate yields, as a complement to protocols reported previously.¹⁵ To complete this study, we also tried the experiment in the presence of pinacol ester or potassium trifluoroborate derivatives instead of boronic acid. However, none of these new substrates led to the formation of the desired product.

Based on previous (di)fluorinated- and trifluoromethylated metallaphotoredox transformations,^10d,11 a plausible synergistic dual mechanistic pathway to 3 is displayed in Scheme 3A. Photoexcitation of Ir(ppy)₃ under blue light irradiation generates a potent excited state *[Ir]^III complex (E_1/2 [Ir*^III/Ir^II] = 0.31 V vs SCE).¹⁶ Single-electron transfer (SET) by the Cu^I complex affords a strongly reducing Ir^II and Cu^II complex. Subsequent single-electron reduction of ethyl bromodifluoroacetate (E_1/2^red = −1.60 V vs SCE in MeCN)¹⁷ by [Ir]^II (E_1/2 [Ir^III/Ir^II] = −2.19 V vs SCE)¹⁶ induces formation of C(sp³)-hybridized radical A and restores the ground-state photocatalyst. Radical oxidative addition of radical A followed by base-promoted transmetalation between the corresponding copper complex and arylboronic acid 2 triggers the formation of intermediate Cu^III species C. A subsequent reductive elimination event generates a new C(sp²)-C(sp³) bond, releasing the difluoro compound 3 and closing the copper catalytic cycle.

Scheme 3.

Proposed mechanism

(A) Proposed mechanism to synthesize (di)fluoroacetate arenes 3. (B) Radical trapping experiment with TEMPO. (C) Mechanistic experiment using 4-bromotoluene instead para-tolylboronic acid. ^aYield was determined by ¹⁹F NMR analysis using (trifluoromethyl)benzene as internal standard.

Consistent with the formation of difluoro radical A, upon addition of 2,2,6,6-tetramethyl-1-piperidinyloxy (TEMPO) to the reaction mixture under optimal conditions, adduct 4 was isolated and confirmed via NMR and HRMS analysis. Furthermore, 4-bromotoluene was used as a nucleophilic partner instead para-tolylboronic acid to confirm the proposed mechanism and dismiss 4-bromotoluene as a possible intermediate in the formation of 3. As expected, traces of 3a were observed, confirming the initial hypothesis. On the other hand, the photochemical quantum yield (Φ) value for this transformation is 0.90, indicating the possibility of a combination of catalytic cycle and chain radical pathways (see Supporting Information).¹⁸ At this point, it is thus also feasible to propose that the Ir(ppy)₃ photocatalyst initially promotes the formation of radical A, followed by radical propagation through single electron reduction of ethyl bromodifluoroacetate (2) by Cu^I complex that in situ triggers the generation of radical A and Cu^II species (Scheme 3A).

In summary, the synthesis of a wide range of ethyl aryldifluoromethylacetate derivatives were rapidly generated under mild conditions from commercially available building blocks: arylboronic acids and ethyl bromodifluoroacetate. This protocol proceeds via synergistic combination of an iridium photocatalyst and copper cycles and shows broad functional group tolerance (free phenol, aldehyde or N-Boc protective group). Additionally, the fluoroalkylated conditions described here were successfully adapted to other commercial fluorinated radical precursors, ethyl 2-bromo-2-fluoroacetate and diethyl (bromodifluoromethyl)phosphonate.