Palladium(III)-Catalyzed Fluorination of Arylboronic Acid Derivatives
Anthony R. Mazzotti‡, Michael G. Campbell‡, Pingping Tang, Jennifer M. Murphy, and Tobias Ritter*
Department of Chemistry and Chemical Biology, Harvard University, 12 Oxford Street, Cambridge, Massachusetts 02138, United States

Supporting Information Placeholder
ABSTRACT: A practical, palladium-catalyzed synthesis of aryl fluorides from arylboronic acid derivatives is presented. The reaction is operationally simple and amenable to multi-gram-scale synthesis. Evaluation of the reaction mechanism suggests a single-electrontransfer pathway, involving a Pd(III) intermediate that has been isolated and characterized. iodides was reported by Hartwig,6 and silver-mediated fluorination reactions have been developed for aryl stannanes, aryl silanes, and arylboronic acids.7 There are several other metal-mediated fluorination reactions of arylboronic acid derivatives, using either palladium8a or copper.8b,c Further development of the reported metal-mediated reactions to use only catalytic quantities of the transition metal has remained difficult. For arylboronic acid derivatives, slow transmetallation of the arene from boron to the transition metal complex is frequently a hurdle to achieving C–F bond-forming catalysis.7c,8a

In the past decade there has been an increase in the number of available methods for the installation of fluorine and fluorine-containing functional groups into organic molecules.1 However, the development of practical carbon–fluorine bond forming reactions to provide aryl fluorides still remains as one of the most challenging transformations in the field of fluorination.1 In this communication, we report a palladium-catalyzed fluorination of arylboronic acid derivatives, which allows for an operationally simple, multi-gram-scale synthesis of functionalized aryl fluorides. A metal-catalyzed fluorination of arylboronic acid derivatives has not previously been reported. Kinetic studies suggest a mechanism distinct from other known arene fluorination reactions, which proceeds through a single-electron-transfer (S.E.T.) pathway without the formation of organopalladium species, and involving an unusual Pd(III) intermediate (2) that has been isolated and characterized (Scheme 1). To date, only two catalytic reactions have been reported that provide a general route to functionalized aryl fluorides: Buchwald’s palladium-catalyzed fluorination of aryl triflates,2 and our group’s silvercatalyzed fluorination of aryl stannanes.3 Our silver-catalyzed reaction requires the preparation and use of toxic aryl stannanes. The palladium-catalyzed nucleophilic fluorination uses more readily available aryl triflates; it currently requires dried fluoride salts and can give mixtures of constitutional isomers for some substrates due to competing pathways in addition to C–F reductive elimination. Work toward catalytic C–H fluorination has been reported by the groups of Groves,4a Sanford,4b,c and Yu.4d,e Direct C–H fluorination is ideal from a perspective of step- and atom-economy, but the development of catalysts that provide selectivity for a broad range of substrates remains challenging. There are a handful of modern stoichiometric arene fluorination reactions. On gram- and smaller scale, deoxyfluorination with PhenoFluor5 is in our opinion currently the most practical method to obtain a large variety of aryl fluorides, but it requires stoichiometric amounts of PhenoFluor. Metal-mediated procedures have been developed for a variety of arene precursors, but frequently require superstoichiometric amounts of transition metal. A copper-mediated fluorination of aryl

Scheme 1. Catalytic Fluorination of Aryl Trifluoroborates, and Isolated Pd(III) Intermediate 2a
1 mol% 1 2 mol% terpy 1.1 equiv Selectfluor EtO O 1.0 equiv NaF MeCN, 40 °C, 15 h open flask 2 N N PdII N 2 BF4 N N N PdIII N N N 2 Isolated Pd(III) Intermediate O 88% Yield 10.6 g Isolated 3 3 BF4

CN EtO

BF3K

CN

F

MeCN 1 Pre-Catalyst

a

terpy = 2,2′:6′,2′′-terpyridine.

Herein we describe a palladium-catalyzed fluorination of arylboronic acid derivatives, using terpyridyl Pd(II) complex 1 as a pre-catalyst (Scheme 1). We propose a mechanism that proceeds without the formation of organopalladium intermediates, which circumvents the problem of transmetallation from the arylboron reagent. Complex 1 has been prepared in one step from Pd(OAc)2, terpyridine (terpy), and HBF4 on decagram scale, and all reagents used in the catalytic fluorination reaction including 1 are stable to air and moisture. The reaction can be performed in an open flask, and is effective for milligram to at least multi-gram scale synthesis of aryl fluorides, which are readily isolated. Inseparable side products from protodeborylation were not observed for the majority of substrates, which may also be due in part to a mechanism that does not involve organopalladium intermediates. Protodeborylation is a common problem for fluorination reactions of arylboronic acid derivatives.8b,c

As shown in Table 1, a wide variety of aryl trifluoroborates can be fluorinated, including both electron-rich and electron-poor arenes. DMF was found to be the optimal solvent for most electron-rich and electron-neutral arenes, while acetonitrile typically provided higher yields for arenes with electron-withdrawing substituents. Ketones, primary amides, carboxylic acids, esters, alcohols, basic heterocycles, aryl bromides, and ortho, ortho’-disubstitution are tolerated in the reaction.

substituents gave significant amounts of side products resulting from demethylation. A variety of commercially available Pd(II) salts can be used in the fluorination reaction, as shown in Table 2. In general, palladium salts with less coordinating anions gave the highest yields. Anion metathesis using NaBF4 as an additive resulted in higher yields for pre-catalysts with coordinating anions such as acetate. Palladium salts with halide anions were not suitable pre-catalysts for the fluorination reaction. Ultimately, we found 1 to be the most convenient pre-catalyst because no additive was needed, and due to its robust stability toward air and moisture as compared to [Pd(MeCN)4][BF4]2. Complex 1 can be stored on the benchtop under ambient conditions without observable decomposition or decrease in catalytic reactivity for at least six months.

Table 1. Fluorination of Aryl Trifluoroboratesa
2 mol% 1 4 mol% terpy 1.0 equiv NaF R 3 1.0–1.2 equiv Selectfluor DMF or MeCN 4–40 °C, 15 h F Me Me PhO 4b DMF, 99% Me F OMe 4d DMF, 75%b F Ph 4g DMF, 73%b F H2N O 4k MeCN, 81% (+9% isomers)d O Me 4n DMF, 70%
a

BF3K R

F

Table 2. Effectiveness of Other Palladium Pre-Catalystsa
CN F EtO O [Pd] 1 2 [Pd(MeCN)4][BF4]2 Pd(OAc)2 Pd(O2CCF3)2
a

4 Solvent, Yield CN EtO O

BF3K

[Pd], terpy 1.0 equiv NaF EtO O

CN

F

F

1.1 equiv Selectfluor MeCN, 40 °C, 15 h Additive none none none NaBF4 (2.0 equiv) NaBF4 (1.0 equiv)

Me

4a DMF, 98%

4c MeCN, 96% F

Yield 96% 95% 95% 91%b 91%

Me F

Me Me 4e DMF, 63%

Me Me

Ph 4f DMF, 85% (+4% biphenyl)

HO

F HO 4h DMF, 71% F

O 4i DMF, 74%c

F

2 mol% [Pd] and 4 mol% terpy were used. Yields refer to isolated, purified material. b 5 mol% Pd(OAc)2 and 10 mol% terpy were used. To highlight the reaction’s practical utility, we have demonstrated that other common arylboron reagents are viable substrates. In situ formation of aryl trifluoroborates via addition of a mixture of NaF and KHF2 allowed for efficient fluorination of pinacol boronic esters and arylboronic acids (eq 1–2). The ability to directly use a variety of arylboronic acid derivatives, without the need for prior isolation of the aryl trifluoroborate, allows for fluorination of a greater range of starting materials.
O B 2 mol% 1 4 mol% terpy 1.1 equiv Selectfluor EtO 2.0 equiv KHF2 1.0 equiv NaF MeCN, 40 °C, 15 h 2 mol% 1 4 mol% terpy B(OH)2 1.1 equiv Selectfluor EtO O 2.0 equiv KHF2 3.0 equiv NaF MeCN, 40 °C, 15 h Me N B O O PhO O O 2 mol% 1 4 mol% terpy 1.5 equiv Selectfluor DMF, 23 °C, 15 h PhO 70% O 86% O 95% F

F

Br 4l MeCN, 96% BocO BocO 4o O MeCN, 74% F N F

N 4m MeCN, 86% O O

F

CN

O

CN

(1)

O 4p MeCN, 83%

EtO O

Yields refer to isolated material of ≥98% purity unless otherwise noted. b No NaF was used. c 5 mol% 1 and 10 mol% terpy were used. d 9% of a mixture of constitutional isomers (ortho- and metafluorobenzamide). Competing formation of constitutional isomers is a challenge for metal-catalyzed fluorination reactions.2 Such impurities are usually difficult to separate from the desired aryl fluoride product. The majority of the aryl fluorides shown in Table 1 were isolated cleanly, with ≥98% purity. In particular, electron-rich aryl trifluoroborates generally did not react to form inseparable side products. Substrates with electron-withdrawing substituents are more likely to give constitutional isomers and diflurorinated products along with the expected aryl fluoride product (typically ≤10%): substrate 3k is a representative example, which reacts to form 4k along with 9% of ortho- and metafluorobenzamide. Other electron-poor substrates such as 3c provided clean isolated product. The Pd-catalyzed fluorination reaction is ineffective for fluorination of heterocycles, and arenes bearing methoxy

CN EtO

CN

F

(2)

F

(3)

We observed that MIDA esters of electron-rich arylboronic acids can also undergo Pd-catalyzed fluorination, albeit in lower yield and requiring a larger amount of Selectfluor (eq 3). No aryl fluoride product was obtained when either NaF or KHF2 was added, suggesting that fluorination proceeds without formation of

Scheme 2. Proposed Mechanism for Pd-Catalyzed Fluorination, Synthesis and X-ray Structure of Pd(III) Intermediate 2, and DFT Calculated Structure of Pd(II) Intermediate 5a
[(terpy)PdII(Solv)]2+ + terpy
1 N F N PdII N MeCN, 23 °C 62% Cl 2 BF4 2 2 BF4 1) terpy 2) Selectfluor N N N PdIII N N N 2 3 3 BF4

[(terpy)2PdII]2+
5 turnoverlimiting oxidation F

N N

MeCN 1

+ BF3 S.E.T.

H

[PdIII]3+
2

F BF3K

+

BF4 F

N N

Cl +

[PdIII]3+
2

F• transfer Cl BF4 BF3K Calculated HOMO of 5 X-ray of Pd(III) Cation of 2

N N

a

terpy = 2,2′:6′,2′′-terpyridine; Solv = solvent (DMF or MeCN); See Supporting Information for details of DFT calculations; X-ray structure of 2 shown with 50% probability ellipsoids; H-atoms, counteranions, and solvent molecules omitted for clarity. the aryl trifluoroborate. MIDA esters of electron-poor arylboronic acids did not afford product. The direct fluorination of MIDA boronates, in the absence of exogenous fluoride anion, indicates a mechanism in which the fluorine atom involved in C–F bond formation is derived from Selectfluor, rather than added fluoride anion. We propose that the Pd-catalyzed fluorination reaction proceeds via an outer-sphere pathway, involving an unusual mononuclear Pd(III) intermediate. A mechanism that is consistent with the experimental data, as described below, is shown in Scheme 2: first, turnover-limiting oxidation of a bis-terpyridyl Pd(II) complex (5) by Selectfluor affords Pd(III) 2 and a Selectfluor radical cation; F• transfer from the Selectfluor radical cation to an aryl trifluoroborate forms the C–F bond and generates a delocalized radical; finally, S.E.T. from the radical to 2 regenerates 5, and provides a delocalized cation, which converts to the aryl fluoride with loss of BF3. The generated BF3 can react with fluoride anion or adventitious water, which may be why the addition of one equivalent of NaF typically increases the yield of aryl fluoride. Complexes of palladium in the +III oxidation state are uncommon, and have only recently been identified as relevant intermediates in organic and organometallic reactions.9 Aryl fluoride formation displayed a well-behaved kinetic profile throughout the course of the catalytic reaction, and no induction period was observed. Therefore, we were able to experimentally determine the rate law using initial rate kinetics, by monitoring aryl fluoride formation via 19F NMR spectroscopy. The reaction displays first-order kinetic dependence on the palladium catalyst, saturation kinetics with respect to terpyridine, zero-order dependence on aryl trifluoroborate, and a non-integer kinetic order of 1.4 with respect to Selectfluor. The saturation behavior observed for terpyridine, along with in situ 1H NMR spectroscopy of the reaction mixture, indicates a catalyst resting state consisting of an off-cycle equilibrium between bis-terpyridyl Pd(II) complex 5 and a terpyridyl Pd(II) solvento complex (e.g. 1). In DMF, the equilibrium of 5 with 1 and free terpyridine is rapid, with a measured binding constant of Ka = 3×103. The non-integer kinetic order experimentally determined for Selectfluor suggests that Selectfluor also participates in a rapid equilibrium with 5, prior to turnoverlimiting oxidation (vide infra). No reaction was observed between pre-catalyst 1 and aryl trifluoroborates in the presence or absence of exogenous terpyridine ligand, and less than 5% background reaction was observed between Selectfluor and the evaluated aryl trifluoroborates. When 1 was treated with one equivalent of terpyridine, followed by one equivalent of Selectfluor, a color change from orange to deep red occurred. The color persisted in MeCN, and crystallization afforded red needles of Pd(III) complex 2 in 62% yield. The structure of 2 was determined by X-ray crystallography (Scheme 2) and exhibits a Jahn-Teller distorted octahedral geometry. The identity of 2 in MeCN solutions was confirmed by EPR spectroscopy (g = 2.09 at 77 K), magnetic susceptibility, and UV-vis/NIR spectroscopy. The Jahn-Teller distorted octahedral geometry and the metric parameters of the terpyridine ligands are consistent with a d7 configuration at Pd with an unpaired electron in a dz2based orbital, rather than a ligand-centered radical, which is also supported by the UV-vis/NIR data and DFT calculations. In the solid state, 2 is stable for months under ambient conditions. Pd(III) complex 2 is a chemically competent catalyst in the fluorination reaction, and was not observed to react with aryl trifluoroborates in the absence of Selectfluor, consistent with the mechanism shown in Scheme 2. Additionally, 2 was not observed to react further when treated with additional Selectfluor, suggesting that a Pd(IV) intermediate is not accessible under the reaction conditions. The structure of the initial complex formed by 1 and terpyridine, Pd(II) 5, is predicted to have a pseudo-octahedral geometry, which is supported by DFT calculations. The calculated HOMO of 5, shown in Scheme 2, is primarily of dz2 parentage with respect to palladium and antibonding between palladium and the apical pyridyl ligands. Removal of one electron from the HOMO results in the Jahn-Teller distorted structure displayed by 2. Selectfluor’s ability to act as a singleelectron oxidant has been supported through a combination of experimental and theoretical studies.10 Electrochemical measurements show that oxidation of Pd(II) 5 to Pd(III) 2 does not proceed by outersphere S.E.T., which suggests the formation of an intermediate adduct between 5 and Selectfluor. The formation of such an adduct is also consistent with the non-integer kinetic order measured for Selectfluor (1.4). The specific mode of interaction between the palladium catalyst

and Selectfluor is unclear at this point, but is likely critical to the success of the fluorination reaction; we speculate that the fluxional binding of terpyridine in 5 is important to the observed reactivity (see Supporting Information). The observation of turnover-limiting oxidation during catalysis prevents us from studying the C–F bond forming step via kinetic analysis. We postulate that C–F bond formation occurs via one of two pathways after initial oxidation of 5 by Selectfluor: (1) direct F• transfer to the aryl trifluoroborate; or, (2) S.E.T. from the aryl trifluoborate to the Selectfluor radical cation, to afford a radical cation, followed by nucleophilic attack of fluoride. In both cases, one-electron oxidation of the resulting radical by Pd(III) 2, as shown in Scheme 2, would afford product and regenerate Pd(II) 5. We carried out an isotopic labeling experiment to distinguish between the two pathways, in which the fluorination reaction was performed in the presence of exogenous [18F]fluoride. Aryl fluoride formation proceeded in 72% yield, but no incorporation of the 18F label was observed (Scheme 3). While the S.E.T./fluoride attack pathway via a tight solvent cage mechanism cannot be rigorously excluded, the absence of 18F incorporation suggests the F• transfer pathway for C–F bond formation.

‡These authors contributed equally.

A CKNOWLEDGMENT
We thank J.R. Brandt for helpful discussions and assistance with binding constant analysis, C.N. Neumann for performing 18F experiments, NIH-NIGMS (GM088237) and NSF (CHE-0952753) for funding, the NSF for a graduate fellowship for A.R.M., and the DOE SCGF for a graduate fellowship for M.G.C.

REFERENCES
(1) (a) Brown, J. M.; Gouverneur, V. Angew. Chem. Int. Ed. 2009, 48, 8610– 8614. (b) Furuya, T.; Klein, J. E. M. N..; Ritter, T. Synthesis 2010, 2010, 1804– 1821. (c) Furuya, T.; Kamlet, A. S.; Ritter, T. Nature 2011, 473, 470–477. (d) Liang, T.; Neumann, C. N.; Ritter, T. Angew. Chem., Int. Ed. 2013, 52, DOI: 10.1002/anie.201206566. (2) (a) Watson, D. A.; Su, M.; Teverovskiy, G.; Zhang, Y.; Garcia-Fortanet, J.; Kinzel, T.; Buchwald, S. L. Science 2009, 325, 1661–1664. (b) Noël, T.; Maimone, T. J.; Buchwald, S. L. Angew. Chem. Int. Ed. 2011, 50, 8900–8903. (c) Maimone, T. J.; Milner, P. J.; Kinzel, T.; Zhang, Y.; Takase, M. K.; Buchwald, S. L. J. Am. Chem. Soc. 2011, 133, 18106–18109. (3) Tang, P.; Furuya, T.; Ritter, T. J. Am. Chem. Soc. 2010, 132, 12150– 12154. (4) (a) Liu, W.; Huang, X.; Cheng, M. J.; Nielsen, R. J.; Goddard, W. A.; Groves, J. T. Science 2012, 337, 1322–1325. (b) Hull, K. L.; Anani, W. Q.; Sanford, M. S. J. Am. Chem. Soc. 2006, 128, 7134–7135. (c) McMurtrey, K. B.; Racowski, J. M.; Sanford, M. S. Org. Lett., 2012, 14, 4094–4097. (d) Wang, X.; Mei, T. S.; Yu, J.-Q. J. Am. Chem. Soc. 2009, 131, 7520–7521. (e) Chan, K. S. L; Wasa M.; Wang X.; Yu J.-Q. Angew. Chem. Int. Ed. 2011, 50, 9081–9084. (5) (a) Tang, P.; Wang, W.; Ritter, T. J. Am. Chem. Soc. 2011, 133, 11482– 11484. (b) Sladojevich, F.; Arlow, S. I.; Tang, P.; Ritter, T. J. Am. Chem. Soc. 2013, 135, 2470–2473. (6) Fier, P. S.; Hartwig, J. F. J. Am. Chem. Soc. 2012, 134, 10795–10798. (7) (a) Furuya, T.; Strom, A. E.; Ritter, T. J. Am. Chem. Soc. 2009, 131, 1662–1663. (b) Furuya, T.; Ritter, T. Org. Lett. 2009, 11, 2860–2863. (c) Tang, P.; Ritter, T. Tetrahedron 2011, 67, 4449–4454. (8) (a) Furuya, T.; Kaiser, H. M.; Ritter, T. Angew. Chem. Int. Ed. 2008, 47, 5993–5996. (b) Fier, P. S.; Luo, J.; Hartwig, J. F. J. Am. Chem. Soc. 2013, 135, 2552–2559. (c) Ye, Y.; Sanford, M. S. J. Am. Chem. Soc. 2013, 135, 4648–4651. (9) For recent reviews of Pd(III) complexes: (a) Powers, D. C.; Ritter, T. Top. Organomet. Chem. 2011, 503, 129–156. (b) Mirica, L. M.; Khusnutdinova, J. R. Coord. Chem. Rev. 2012, 257, 299–314. For references describing mononuclear Pd(III) complexes: (c) Lanci, M. P.; Remy, M. S.; Kaminsky, W.; Mayer, J. M.; Sanford, M. S. J. Am. Chem. Soc. 2009, 131, 15618–15620. (d) Khusnutdinova, J. R.; Rath, N. P.; Mirica, L. M. J. Am. Chem. Soc. 2012, 134, 2414–2422. For references describing dinuclear Pd(III) complexes: (e) Powers, D. C.; Ritter, T. Nat. Chem. 2009, 1, 302–309. (f) Powers, D. C.; Geibel, M. A. L.; Klein, J. E. M. N.; Ritter, T. J. Am. Chem. Soc. 2009, 131, 17050–17051. (g) Powers, D. C.; Benitez, D.; Tkatchouk, E.; Goddard, W. A., III; Ritter, T. J. Am. Chem. Soc. 2010, 132, 14092–14103. (h) Powers, D. C.; Xiao, D. Y.; Geibel, M. A. L.; Ritter, T. J. Am. Chem. Soc. 2010, 132, 14530–14536. (i) Chuang, G. J.; Wang, W.; Lee, E.; Ritter, T. J. Am. Chem. Soc. 2011, 133, 1760– 1762. (j) Campbell, M. G.; Powers, D. C.; Raynaud, J.; Graham, M. J.; Xie, P.; Lee, E.; Ritter, T. Nat. Chem. 2011, 3, 949–953. (k) Powers, D. C.; Lee, E.; Ariafard, A.; Sanford, M. S.; Yates, B. F.; Canty, A. J.; Ritter, T. J. Am. Chem. Soc. 2012, 134, 12002–12009. (l) Powers, D. C.; Ritter, T. Acc. Chem. Res. 2012, 45, 840–850. (10) (a) Serguchev, Y. A.; Ponomarenko, M. V.; Lourie, L. F.; Fokin, A. A. J. Phys. Org. Chem. 2010, 24, 407–413. (b) Nesterenko, A. M.; Ponomarenko, M. V.; Lur'e, L. F.; Serguchev, Y. A. Theor. Exp. Chem. 2002, 38, 156–161. (c) Zhang, X.; Wang, H.; Guo, Y. Rapid Commun. Mass Spectrom. 2006, 20, 1877– 1882. (d) Zhang, X.; Liao, Y.; Qian, R.; Wang, H.; Guo, Y. Org. Lett. 2005, 7, 3877–3880. (e) Nyffeler, P. T.; Durón, S. G.; Burkart, M. D.; Vincent, S. P. P.; Wong, C.-H. Angew. Chem. Int. Ed. 2005, 44, 192–212. (11) (a) Furuya, T.; Ritter, T. J. Am. Chem. Soc. 2008, 130, 10060–10061. (b) Furuya, T.; Benitez, D.; Tkatchouk, E.; Strom, A. E.; Tang, P.; Goddard, W. A., III; Ritter, T. J. Am. Chem. Soc. 2010, 132, 3793–3807.

Scheme 3. Isotopic Labeling Experiment
1, terpy BF3K PhO

[18F]KF
N N Cl 2 BF4 PhO 72%

19F

19F

• No Isotopic Substitution • 19F Incorporation Only

In previously reported metal-mediated or -catalyzed arene fluorination reactions, including our group’s palladium- and silver-mediated fluorination of arylboronic acids, carbon–fluorine bond formation is proposed to occur via reductive elimination from an aryl–metal fluoride complex.11 The palladium-catalyzed fluorination reaction presented here is unusual in that it seems to proceed without the formation of organopalladium intermediates, yet provides high levels of selectivity. In conclusion, we have reported the first metal-catalyzed fluorination of arylboronic acid derivatives. The reaction proceeds under mild conditions, is tolerant towards moisture and air, and is amenable to multi-gram-scale synthesis of functionalized aryl fluorides. We propose a single-electron-transfer mechanism involving a well-defined Pd(III) intermediate. This reaction provides a level of practicality and operational simplicity not previously achieved by metal-catalyzed or mediated arene fluorination reactions, and does not generally afford side products from protodemetallation, a common problem for the synthesis of aryl fluorides. Drawbacks of the reaction include the inability to fluorinate heterocycles, and the formation of constitutional isomers for some electron-poor substrates.

ASSOCIATED CONTENT
Supporting Information
Detailed experimental procedures, spectroscopic data for all new compounds, details of DFT calculations, crystallographic data for 1, 2, S1, and S2 (CIF). This material is available free of charge via the Internet at http://pubs.acs.org.

AUTHOR INFORMATION
Corresponding Author
ritter@chemistry.harvard.edu

Author Contributions

Insert Table of Contents artwork here

BF3K R

Pd(III)
R

F

Catalysis

• Up to 99% Isolated Yield • 10 Gram Scale • Open Flask

5