Merging Visible Light Photocatalysis and Transition Metal Catalysis in the Copper-Catalyzed Trifluoromethylation of Boronic Acids with CF₃I

Abstract

This communication describes the development of a mild method for the cross-coupling of arylboronic acids with CF₃I via the merger of photoredox and Cu catalysis. This method has been applied to the trifluoromethylation of electronically diverse aromatic and heteroaromatic substrates and tolerates many common functional groups.

Trifluoromethyl substituents are widely prevalent in pharmaceuticals and agrochemicals.¹ As such, the development of mild and versatile synthetic methods for generating carbon–CF₃ bonds has become a field of intense research effort. Over the past 3 years, a variety of Pd^2,3 and Cu^4,5-based cross-coupling protocols have been developed for the trifluoromethylation of aryl halides, aryl boronic acids, and aromatic carbon– hydrogen bonds. As exemplified in Scheme 1a/b for the Cu-promoted trifluoromethylation of boronic acids, these transformations typically involve “CF₃⁻” ^4c,4e,4l,5i or “CF₃⁺” ^4f,5d,e reagents and are believed to proceed via nucleophilic or electrophilic transfer of the CF₃ group to the Cu center, respectively. Many of these new methods represent significant progress in comparison to the traditional Swarts reaction⁶ (which requires highly reactive fluorinating reagents and harsh conditions).

Scheme 1.

Cu-Mediated/Catalyzed Trifluoromethylation of Boronic Acids

Despite these important advances, most current strategies for aryl–CF₃ cross-coupling suffer from one or more limitations. In some cases, temperatures greater than 100 °C^2b,3a,b,4d and/or strong acids or bases (TFA^3a or ^tBuOK^4j) are necessary. Other methods require expensive trifluoromethylating reagents (e.g., S-(trifluoromethyl)thiophenium salts,^3a,4f,5d Togni’s reagent,^5b,5e,5g or TESCF₃).^3b,5a Finally, many protocols exhibit limited substrate scope/generality.^3c,5a,b

One attractive approach to begin to address these limitations would be to access alternative and potentially complementary mechanistic manifolds. We reasoned that a radical pathway (Scheme 1c) would be particularly interesting, since CF₃• can be generated under mild, neutral conditions from commercially available and relatively inexpensive CF₃I.⁷ In particular, we noted a recent report by MacMillan demonstrating the conversion of CF₃I to CF₃• at room temperature in the presence of a photocatalyst, visible light, and a reductant (Scheme 2).⁸ On the basis of this work, we hypothesized that the merger of visible light photocatalysis (to generate CF₃•) with Cu catalysis (to generate reactive Cu–aryl species) (Scheme 2) might provide a mild and general method for the trifluoromethylation of boronic acid derivatives.

Scheme 2.

Proposed New Pathway for Radical Trifluoromethylation of Boronic Acids via Cu/Ru Photocatalysis

Our initial investigations focused on the Cu and Ru-photocatalyzed trifluoromethylation of 1,1’-biphenyl-4-ylboronic acid with CF₃I to form 4-(trifluoromethyl)-1,1’-biphenyl (1a). We were delighted to find that Cu/Ru photocatalysis provided product 1a in modest to excellent yield under a number of conditions (Table 1). A variety of different bases (to promote transmetalation) and reaction solvents were screened for this reaction (see Table S1) and the use of K₂CO₃ in DMF proved optimal. Copper(I) catalysts generally performed better than Cu^II salts, and the highest yield of 1a (76%) was obtained with CuOAc. The optimal conditions were as follows: 1 equiv of boronic acid 1, 5 equiv of CF₃I, 1 equiv of K₂CO₃, 20 mol % of CuOAc, and 1 mol % of Ru(bpy)₃Cl₂·6H₂O with irradiation from two 26 W household light bulbs. The major side product was 4-iodo-1,1’-biphenyl (formed in 9% yield under the optimal conditions).

Table 1.

Optimization of Reaction Between 1 and CF₃I^[a]

Entry	[Cu]	Base	Yield
1	Cu(OTf)2	K2CO3	14%
2	[Cu(OTf)]2•C6H6	K2CO3	28%
3	Cul	K2CO3	34%
4	Cu	K2CO3	40%
5	Cu(OAc)2	K2CO3	68%
6	CuOAC	K2CO3	76%
7	CuOAC	NaOAc	34%
8	CuOAC	KF	50%
9	CuOAC	none	6%
10[b]	CuOAC	K2CO3	1%
11[c]	none	K2CO3	3%
12[d]	CuOAC	K2CO3	3%

^[a]General conditions: substrate (0.05 mmol, 1 equiv), CF₃I (5 equiv), [Cu] (0.2 equiv), Ru(bpy)₃Cl₂•6H₂O (0.01 equiv), base (1 equiv), DMF (0.17 M in substrate), 60 °C, 12 h, 26 W compact fluorescent light bulb. ¹⁹F NMR yield.

^[b]General conditions, but with no light.

^[c]General conditions, but with no CuOAc.

^[d]General conditions, but with no Ru(bpy)₃Cl₂•6H₂O.

This Cu/Ru-catalyzed coupling between 1 and CF₃I is practical and easily scalable. The reaction in Table 1, entry 6 was performed on a 0.05 mmol scale and provided 76% yield as determined by ¹⁹F NMR spectroscopy and GCMS. Nearly identical isolated yields were obtained on both 1 and 5 mmol scales (72% and 70% isolated yield, respectively).

A variety of control reactions were conducted to establish the role of each component of the reaction mixture. As shown in Table 1, entries 10–12, when light, CuOAc or Ru(bpy)₃Cl₂·6H₂O were excluded under otherwise identical conditions, ≤3% yield of 1a was obtained.⁹ These results clearly indicate the necessity of all three components to achieve high yields under these conditions, consistent with the major pathway to 1a proceeding via dual Ru/Cu catalysis (vide infra). The iodinated side product 4-iodo-1,1’-biphenyl was also subjected to the reaction conditions to establish whether it is an intermediate in the boronic acid trifluoromethylation process. Less than 2% of the aryl–CF₃ product was formed, strongly suggesting that the major pathway to 1a does not involve an iodinated intermediate. Finally, the reactivity of boronic acid substrate 1 was investigated under conditions reported by Baran and MacMillan to promote C–H trifluoromethylation reactions via in situ generation of CF₃•.¹⁰ In both cases, <2% of 1a was observed. These results indicate that 1a is not formed by the direct reaction of CF₃• with the boronic acid.

This transformation was next applied to a variety of different aryl- and heteroaryl boronic acid derivatives. Representative examples are shown in Scheme 3 and were selected to highlight not only the broad scope but also the limitations of this method.¹¹ Aromatic boronic acids bearing both electron-donating (t-butyl, methoxy) as well as electron-withdrawing (cyano, trifluoromethyl, fluoro, methyl ester) substituents underwent trifluoromethylation in high yield. A variety of different potentially reactive functional groups (aromatic alcohols, ketones, aldehydes, esters, and amides) were quite well-tolerated. A boronic acid embedded in the estrone framework underwent trifluoromethylation to generate 22a in 80% isolated yield. Most remarkably, 4-iodo-phenylboronic acid underwent selective trifluoromethylation to form 7a, leaving the aryl iodide intact for subsequent functionalization. This demonstrates the complementarity of this method to many other Cu-catalyzed trifluoromethylation protocols.^5a,5c,5f Furthermore, it provides additional evidence against the possibility of aryl iodide intermediates in this transformation.

Scheme 3.

Substrate Scope for Cu/Ru-Catalyzed Trifluoromethylation of Boronic Acids[a]

^[a] General conditions: substrate (1 equiv), CF₃I (5 equiv), [Cu] (0.2 equiv), Ru(bpy)₃Cl₂•6H₂O (0.01 equiv), K₂CO₃ (1 equiv), DMF (0.17 M in substrate), 60 °C, 12 h, 26 W compact fluorescent light bulb; Isolated yield (≥95% purity). ^{[b] 19}F NMR yield. ^[c] 0.5 equiv of CuOAc. ^[d] Isolated as a 1:1 mixture with inseparable protodeboronated product. ^[e] 0.1 equiv of CuOAc. ^[f] Isolated as a 10:1 mixture with inseparable protodeboronated product. ^[g] 3 equiv of CF₃I. ^[h] Reaction run at 70 °C. ^[i] Reaction run at 40 °C. ^[j] 0.05 equiv of CuOAc.

The use of sterically hindered substrates such as 1-naphthyl and 2,4,6-trimethylphenyl boronic acid is typically challenging for copper-mediated cross-coupling reactions.¹² As shown in Scheme 2, similar effects were seen in the current transformation, with products 9a and 10a being formed in modest yields (42% and 39%, respectively). In these cases, competing protodeboronation was problematic, and the major side product was naphthalene or mesitylene, respectively.

Heteroaromatic substrates are of particular relevance to the pharmaceutical and agrochemical industries due to the prevalence of heteroarenes in biologically active compounds.¹³ Boronic acids derived from pyridine, quinoline, furan and thiophene all underwent trifluoromethylation in modest to good yield.¹⁰ In some of these cases, modification of the catalyst loading and/or reaction temperature was required to achieve optimal yield. Importantly, with all of these substrates, trifluoromethylation of the boronic acid moiety out-competed uncatalyzed C–H trifluoromethylation of the heterocycle with CF₃•. Thus, this method provides an attractive route for the site-selective installation of CF₃ substituents into these scaffolds.

Related conditions could also be applied to analogous perfluoroalkylation reactions. This is a significant advantage of the current method, since perfluoroalkyl analogues of other common trifluoromethylating reagents (e.g., R₃SiCF₃, S-(trifluoromethyl)thiophenium salts, or Togni’s reagent) are expensive and/or not commercially available. As shown in Scheme 4, perfluorobutyl and perfluorodecyl iodides reacted with 1 to afford products 1b and 1c in good yields under the Cu/Ru catalyzed conditions.

Scheme 4.

Perfluoroalkylation of 1 Catalyzed by Cu/Ru[a]

^[a] General conditions: substrate (1 equiv), CuOAc (0.5–1 equiv) Ru(bpy)₃Cl₂•6H₂O (0.01 equiv), K₂CO₃ (1 equiv), DMF (0.17 M in substrate), 60 °C, 12 h, 26 W compact fluorescent light bulb; Isolated yield. ^[b] 5 equiv of C₄F₉I, 0.5 equiv of CuOAc. ^[c] 1.2 equiv of C₁₀F₂₁I, 1 equiv of CuOAc.

While a detailed mechanistic picture of this transformation remains to be elucidated, a possible set of catalytic cycles is shown in Scheme 5. In this sequence, photoexcitation of Ru(bpy)³²⁺ to Ru(bpy)₃²⁺* is followed by 1e⁻ reduction by Cu^I to generate Ru(bpy)₃⁺ and Cu^II.¹⁴ Reduction of CF₃I by Ru(bpy)₃⁺ then affords CF₃• and I⁻. Notably, literature reduction potential data indicates that both of these reactions should be thermodynamically favorable (see Figure S1). The CF₃• could then react with Cu^II to generate a Cu^III(CF₃) intermediate. Subsequent base-promoted transmetalation between Cu^III and the aryl boronic acid would afford Cu^III(aryl)(CF₃), which could undergo aryl–CF₃ bond-forming reductive elimination to release the organic product and regenerate the Cu^I catalyst.¹⁵

Scheme 5.

Possible Mechanism for Cu/Ru-Catalyzed Trifluoromethylation of Boronic Acids

In summary, this communication describes a mild and general approach for the Cu/Ru-catalyzed trifluoromethylation/perfluoroalkylation of aryl boronic acids. This method takes advantage of visible light photoredox catalysis to generate R_F• under mild conditions and merges it with copper-catalyzed aryl boronic acid functionalization. The combination has enabled a method for the trifluoromethylation of a wide variety of aromatic and heteroaromatic substrates bearing many common functional groups. This transformation demonstrates the feasibility of achieving Cu-catalyzed trifluoromethylation via a radical pathway. Furthermore, it represents a new example of combining organometallic and photoredox catalysis to achieve synthetically useful organic transformations.¹⁶