Room Temperature Photochemical Copper-Mediated Fluorination of Aryl Iodides

Abstract

This report describes a method for the photochemical Cu-mediated fluorination of aryl iodides with AgF via putative aryl radical (Ar•) intermediates. It involves irradiating an aryl iodide with UVB light (λ_max = 313 nm) in the presence of a mixture of Cu^I and Cu^II salts and AgF. Under these conditions, fluorination proceeds at room temperature for substrates containing diverse substituents, including alkoxy and alkyl groups, ketones, esters, sulfonate esters, sulfonamides, and protected amines. This method has been translated to radiofluorination using a combination of K¹⁸F, K₃PO₄, and AgOTf.

Graphical Abstract

(Hetero)aryl fluorides (ArF) are prevalent in agrochemicals, active pharmaceutical ingredients, and positron emission tomography (PET) tracers due to their unique biological and imaging properties.^1–5 As such, mild and efficient synthetic methods to form C(sp²)–F bonds are in high demand.^6–8 The ideal Ar–F bond-forming reaction would couple a nucleophilic fluoride salt (MF) with a stable and abundant (hetero)aryl precursor (ArX) under mild conditions. A variety of ArX have been leveraged in nucleophilic fluorination reactions, including with X = halide,^9–12 nitro,¹³ diazonium,¹⁴ phenol,¹⁵ carboxylic acid,¹⁶ boronic acid/ester,¹⁷ stannane,¹⁸ and siloxane.¹⁹ Among these, aryl halides (X = Cl, Br, I) stand out as particularly attractive precursors as they are broadly available, inexpensive, stable, and have orthogonal reactivity to most common functional groups.

Previous work on the nucleophilic fluorination of aryl halides has focused on three approaches: (i) nucleophilic aromatic substitution (S_NAr),⁹ (ii) Pd-catalyzed cross-coupling,¹⁰ and (iii) Cu-mediated reactions.^11,12 While the first two have been well studied, they have key limitations, including forcing reaction conditions, narrow substrate scope, and/or the formation of isomeric products. For instance, S_NAr fluorination is only effective for aryl halides bearing electron withdrawing substituents; furthermore, these reactions typically require temperatures of ≥130 °C.⁹ Pd-catalyzed fluorination exhibits a much broader scope but involves specialized ligands and high temperatures (≥130 °C).¹⁰ Cu-mediated fluorination is arguably the most promising but least developed approach. Seminal work by Fier and Hartwig demonstrated the (^tBuCN)₂Cu(OTf)-mediated conversion of electronically diverse aryl iodides to aryl fluorides using AgF (Scheme 1a).¹¹ However, high temperatures (≥140 °C) were required to overcome the barrier for Ar–I oxidative addition at [Cu^I]–F (TS¹) to generate the Cu^III(Ar)(F) intermediate A. This kinetic challenge can be addressed using directing groups.¹² However, the non-directed Cu-mediated fluorination of aryl halides under mild conditions has remained elusive.

Scheme 1.

(a) Hartwig and Fier: Thermal Cu^I-mediated fluorination of ArI via concerted oxidative addition pathway.¹¹ (b) This work: Photochemical Cu-mediated fluorination of ArI via radical addition pathway at room temperature.

We reasoned that photochemical conversion of ArX to the corresponding aryl radical (Ar•) and subsequent Ar• addition to a Cu^II–fluoride could form A under mild conditions (Scheme 1b).^20–23 Notably, Ritter recently demonstrated the Cu-mediated fluorination of aryl thianthrenium salts via an analogous radical addition pathway.²⁴ In this system, Ar• was accessed by using an Ir photoredox catalyst to reductively cleave the Ar–S bond. A key challenge for applying this approach to aryl iodides is that reductive cleavage requires more strongly reducing conditions, which are unlikely to be compatible with the Cu^II mediators.²³ As such, we pursued direct photolysis with UV light as an alternative approach for generating Ar• from aryl iodides.

Aryl iodides are known to undergo homolytic cleavage to form Ar• upon UV excitation (Scheme 2a, step i).²⁵ Since these conditions are not strongly reducing, we anticipated that they could be compatible with the Cu^II mediators required for fluorination (Scheme 2a). Indeed, while this work was in progress, Ritter leveraged a similar approach for the photochemical Cu-mediated etherification of aryl halides, using either energy transfer (with 1,8-diazabicyclo[5.4.0]undec-7-ene) or direct photolysis to convert ArX to Ar•.^26,27 We selected 4-iodoanisole (1a) as a model substrate based on its low reactivity/selectivity towards fluorination via other pathways.²⁸ As such, a method for converting 1a to the corresponding aryl fluoride 2a would complement the current state-of-the-art.

Scheme 2.

(a) Proposed mechanism for UVB homolysis of ArI for Cu-mediated fluorination. (b) Optimized conditions. RT = room temperature (30–35 °C in photobox with fan cooling).

Our initial investigations evaluated the reaction of 1a with various Cu^II mediators and fluoride sources in MeCN at room temperature upon irradiation with UVB light (λ_max = 313 nm). As predicted, these conditions yielded varying amounts of 4-fluoroanisole (2a) accompanied by the protodehalogenation product anisole (3a). Cu(OTf)₂ and AgF were identified as the optimal Cu^II and fluoride source, respectively, and following optimization of stoichiometries (see Tables S1–S18 for complete details on reaction optimization), this combination afforded 96% conversion of 1a, 56% yield of 2a, and 8% of 3a (Scheme 2b). The protodehalogenation product is likely formed by the reaction of Ar• with solvent,²⁹ and Ritter showed that this side reaction can be suppressed by the inclusion of a Cu^I salt.^16,30 Similarly, in our system the addition of 1 equiv of Cu(MeCN)₄PF₆ under otherwise identical conditions led to an enhanced yield of 2a (66%) and a decrease in 3a (to 2%).

The scope of this reaction was explored under the optimized conditions (Table 1). A range of aryl iodides bearing electron-donating (2a-e), -neutral (2f), and -withdrawing (2g-m) substituents reacted to afford aryl fluoride products. Moderate to high yield was observed with substituents at the ortho- or para-positions, and no aryl fluoride isomers were detected. Aryl sulfonate esters (2s) and sulfonamides (2t), which are prevalent in pharmaceuticals, proved compatible. Nitrogen-containing molecules, including 4-iodophenylalanine, were also effective substrates, so long as the basic nitrogen was protected as an amide or carbamate (2u, 2v). In all cases, the major by-product was the arene derived from protodeiodination, although this was typically formed in ≤10% yield, (see p. S15–16 for details with each substrate). It could typically be removed by careful chromatography on silica gel to afford the fluorinated products in ≥95% purity. A list of unsuccessful substrates is provided on p. S23. A common theme is that strong Lewis bases (amines, nitrogen heterocycles, and sulfur heterocycles) are incompatible, possibly because they disrupt the reaction by ligating to the copper mediator.

Table 1.

Substrate scope of photochemical Cu-mediated fluorination of aryl iodides using UVB irradiation.

Aryl iodide was the limiting reagent, using 100 μmol in degassed dry MeCN (0.5 mL). RT = room temperature; 30–35 °C in photobox with fan cooling. The NMR yield of protodehalogenated by-product for each substrate that was not isolated is provided in the SI (p. S16–S17) and was ≤10% in all cases. For these same substrates, conversion of ArI was ≥90% except for 2j, where 12% of the aryl iodide remained.

^aDetermined by GC-FID with naphthalene as standard.

^bDetermined by ¹⁹F NMR spectroscopy with α,α,α-trifluorotoluene as standard.

^cReaction time 36 h.

^dIsolated yield in ≥95% purity.

^e≥92% purity.

^fIsolated as HPF₆ salt.

A key advantage of aryl iodide substrates is that numerous derivatives are commercially available. Furthermore, electron-rich and -neutral derivatives can be synthesized from abundant Ar–H starting materials by electrophilic aromatic substitution.^31–33 As one example, the fluorination of naproxamide was achieved by sequential electrophilic iodination with N-iodosuccinimide followed by photochemical Cu-mediated fluorination to form 2w in 74% NMR yield (57% isolated; eq. 1).

Finally, we leveraged this transformation to achieve the ¹⁸F radiolabeling of aryl iodides.³⁴ Since the most readily available source of fluorine-18 is [¹⁸F]KF, we sought to use this in combination with a non-nucleophilic silver salt (e.g., AgOTf) to generate Ag¹⁸F in situ. As shown in Table 2, entry 1, this approach proved effective for the ¹⁹F fluorination of 1k, affording a 73% yield of product 2k. In contrast, with [¹⁸F]KF only traces (4%) of 2k-¹⁸F were obtained (entry 2). The key difference between the non-radioactive ¹⁹F and radioactive ¹⁸F reactions is the stoichiometry. In the former, 1k is the limiting reagent, and 4 equiv of [¹⁹F]KF is added. However, in the latter [¹⁸F]KF is the limiting reagent, with 1k in >1000-fold excess. This change in stoichiometry dramatically impact reaction rates as well as the speciation/ligation environment of the Cu mediator. We next added 2 equiv of [¹⁹F]KF to the radiochemical reactions, hypothesizing that these conditions would better simulate the ¹⁹F fluorination. Consistent with this hypothesis, the reaction afforded 84% radiochemical conversion to 2k-¹⁸F (entry 3).

Table 2.

entry	modification	2k or 2k-18F (%)
1	2 equiv KF	73
2	[18F]KF	4a
3	[18F]KF, 2 equiv KF	84a
4	[18F]KF, 2 equiv K2CO3	41a
5	[18F]KF, 2 equiv KH2PO4	7a
6	[18F]KF, 2 equiv K3PO4	47a
7	[18F]KF, 1 equiv K3PO4 b	53 ± 13a

^aRadiochemical conversion determined by radioTLC.

^bConditions: 1 equiv Cu(OTf)₂, 1 equiv Cu(MeCN)₄PF₆, 2 equiv AgOTf, 1 equiv K₃PO₄, in MeCN, 1 h, UVB.

While the carrier-added approach provides proof-of-principle for the feasibility of radiofluorination, it affords product with low molar activity, which is undesirable for imaging applications. As such, we next explored substituting the [¹⁹F]KF carrier with alternative salts that are (i) basic (analogous to anhydrous fluoride) and (ii) weakly nucleophilic (such that they do not promote competitive functionalization). Carbonate and phosphate salts proved effective (entries 4–6), with 2 equiv of K₃PO₄ affording 47% radiochemical conversion to 2k-¹⁸F. This was increased to 53% through further optimization (entry 7 and Table S20). As summarized in Scheme 3, the optimal conditions proved effective for the photochemical Cu-mediated radiofluorination of electron rich and electron deficient aryl iodide substrates, in radiochemical yields ranging from 23–66%. Overall, this method offers a mild approach to the radiofluorination of aryl iodide substrates, including electron-neutral/rich substrates that would be unreactive in traditional S_NAr pathways.

Scheme 3.

Radiofluorination of aryl iodides. Reaction set up under N₂, and [¹⁸F]KF was added through septum cap as a solution in MeCN (0.1 mL). RT = room temperature; 30–35 °C in photobox with fan cooling. All yields are radiochemical yields as determined by radioTLC and radioHPLC. In all cases, radiochemical purity (as measured by radioHPLC) was ≥86%.

In conclusion, this report describes a photochemical method for the room temperature nucleophilic fluorination of aryl iodides using a Cu^II mediator, a Cu^I additive, and AgF as the fluoride source. Mechanistically, this transformation is proposed to proceed via aryl radical intermediates (Ar•) that react with Cu^II–F to yield ArF products. Detailed mechanistic details of the Ar• formation and radical addition steps are ongoing. This method affords modest to good yields with a range of aryl iodide substrates, including derivatives bearing both electron-donating and electron-withdrawing substituents. It was translated to radiofluorination using a combination of [¹⁸F]KF, AgOTf, and K₃PO₄. Overall, this method provides a complementary approach to the current state-of-the-art for accessing these important products.

Data Availability Statement

The data underlying this study are available in the published article and its Supporting Information.