Introduction

Polyfluoroalkyl substituents impart unique physicochemical properties—high thermal and chemical stability, hydrophobicity, and distinctive interfacial behavior—that underpin their broad applications in pharmaceuticals, agrochemicals, and advanced materials1,2,3,4. These features originate from the exceptional strength and polarity of C(sp3)–F bonds, which simultaneously confer robustness and impede selective activation. The high bond dissociation energies and strong electron-withdrawing effects of these bonds frequently lead to unselective pathways and competing multi-bond cleavage5,6,7,8, underscoring the continued need for catalytic strategies capable of controlling C–F functionalization with precision.

Considerable attention has been devoted to the selective activation of trifluoromethyl groups and their derivatives9,10,11,12,13, and Lewis acid-promoted difluorocarbocation formation14,15,16, radical spin–center shifts17, and photoredox-mediated single-electron processes18,19,20,21,22,23,24 have enabled diverse C–F transformations of CF3-containing substrates. In contrast, the selective functionalization of longer perfluoroalkyl chains (CnF2n + 1, n ≥ 2) remains far less developed. Representative examples include the electroreductive α-C–F cleavage of pentafluoroethylarenes followed by CO2 carboxylation25, Mg/TMSCl- or Cu–Mg-mediated reductive activation, which predominantly affords silylated or reduced products with limited control over site selectivity26, and Ni(0)-enabled sequential double C–F activation of pentafluoroethyl alkenes in [3 + 2] cycloadditions with alkynes27. Photoredox-mediated benzylic defluoroallylation of Ar–nC4F9, Ar–nC6F13, and Ar–CF(CF3)2 scaffolds has also been demonstrated using allylic stannanes as coupling partners28, while sequential organophotocatalytic/Lewis acidic protocols enable benzylic C–F substitution with carbon nucleophiles such as allylsilanes, stannanes, and silyl enol ethers29. Although these studies demonstrate that longer perfluoroalkyl groups can undergo C–F activation under tailored conditions, they uniformly rely on prefunctionalized coupling partners and often require strongly reducing or Lewis acidic media, and none provides a general method for direct C–C bond formation with simple, unfunctionalized alkenes. Beyond these methodological constraints, the strong electron-withdrawing nature of perfluoroalkyl groups destabilizes the radical intermediates formed upon C–F cleavage, making them highly susceptible to competing pathways. These challenges—including steric hindrance, radical destabilization, and reverse processes such as back-electron transfer (BET) and fluoride readdition—often suppress productive reactivity30,31,32,33,34 (Fig. 1a). These intrinsic features complicate the selective incorporation of perfluoroalkyl motifs into pharmacophores and complex molecular scaffolds, underscoring the need for broadly applicable catalytic strategies.

Fig. 1: Challenges in C(sp3)–F activation of polyfluoroalkyl groups and our defluoroallylation strategy.
Fig. 1: Challenges in C(sp3)–F activation of polyfluoroalkyl groups and our defluoroallylation strategy.
Full size image

a Recent advances in C(sp3)–F activation and challenges in activating longer perfluoroalkyl groups. b Limitations of prefunctionalized reagents in catalytic defluoroallylation. c Site-selectivity challenge in defluoroallylation of alkenes with multiple similar allylic C–H bonds. Color-coded spheres indicate distinct allylic C–H sites with comparable electronic and steric environments. d This work: synergistic photoredox/cobalt catalysis enables radical defluoroallylation of polyfluoroalkyl compounds with alkenes.

Alkenes are privileged building blocks in synthesis, offering diverse modes of reactivity for the rapid construction and late-stage modification of complex molecules35,36. In particular, allylic C–H functionalization has attracted significant attention for its ability to introduce multifunctional motifs while preserving the integrity of the olefinic bond37,38,39,40. Despite advances in catalytic defluoroallylation, current methodologies predominantly rely on prefunctionalized reagents such as allylsilanes or allylstannanes11,16,28,41, which exhibit limited substrate scope (Fig. 1b). A more streamlined strategy would enable the direct defluoroallylation of simple alkenes using perfluoroalkyl compounds without prefunctionalization. Such a transformation would be especially valuable in complex settings—such as natural products bearing multiple, electronically and sterically similar allylic C–H sites—where achieving high levels of site selectivity remains difficult (Fig. 1c).

To surmount the challenge of direct C(sp3)–F activation in perfluoroalkylarenes and their subsequent coupling with alkenes, we propose a synergistic radical strategy that combines photoredox and cobalt catalysis (Fig. 1d). We hypothesize that a photoredox catalyst can reduce perfluoroalkyl compounds to generate unstable radical anions, which then eliminate fluoride to give perfluoroalkyl radicals. Steric hindrance may further impede bond formation, potentially leading to reversibility via back-electron transfer or fluoride readdition. To address these issues, we introduce Lewis acidic fluorine scavengers to accelerate radical generation and suppress undesired electron transfer42,43, thereby enhancing overall reaction efficiency. The perfluoroalkyl radicals thus formed are expected to add smoothly to alkenes, with subsequent cobalt-catalyzed hydrogen atom transfer44,45,46 (HAT) facilitating the regioselective formation of the desired products. In this work, we demonstrate a dual photoredox/cobalt catalytic system that enables redox-driven, site-selective allylation of perfluoroalkylarenes and polyfluorinated aliphatic amides. This methodology provides an efficient and complementary strategy for the selective transformation of perfluoroalkyl and polyfluorinated compounds, enabling the precise assembly of complex molecular scaffolds.

Results

In our initial investigations, we selected methyl 4-(perfluorobutyl)benzoate (1a) and 2-methylpent-1-ene (2af) as model substrates. Systematic optimization (Table 1) established the optimal conditions (Condition A): 0.5 mol% fac-Ir(ppy)3 as the photocatalyst, 3 mol% methylcobalamin as the HAT catalyst, 0.25 equivalents of DIPEA, and 0.3 equivalents of HBpin, which functions as both base and fluoride scavenger. Reactions conducted in 2-ethoxyethanol under blue LED irradiation for 36 h afforded the desired product 3af in 85% yield by GC analysis (83% isolated yield, Table 1, entry 1). Replacing fac-Ir(ppy)3 with [Ir(dtbbpy)(ppy)2]PF6, Ir[FCF3(CF3)ppy]2(4,4’-dCF3bpy)PF6, or Mes-Acr+BF4− resulted in only trace or no product (entries 2–4). Similarly, methylcobalamin proved essential for high yield and exclusive site selectivity, as alternative cobalt catalysts (B12, Co-TPP, CoPc, or cobaloxime complexes) significantly reduced the reaction outcome (entries 5–10). Reducing the equivalents of olefin 2af decreased the yield to 77% (entry 11), while substituting DIPEA with K2CO3 or 2-ethoxyethanol with DMSO led to diminished yields of 68% and 32%, respectively (entries 12–13). Moreover, alternative boron reagents (B2pin2, Et3B) provided unsatisfactory results (entries 14–15). Control experiments showed that Ir(ppy)3, methylcobalamin, and visible light are essential for the reaction (entries 16–18), while DIPEA and HBpin improve the reaction efficiency (entries 19–20).

Table 1 Reaction optimization
Full size table

With the optimized conditions established, we next explored the alkene substrate scope with a particular focus on site selectivity. As summarized in Fig. 2, the reactivity of various natural terpenes, such as (+)-limonene, β-pinene, nootkatone, dihydrocarvone, and (+)-carvone (3aa–3ae), was explored. These substrates exhibited excellent site selectivity (>20:1), underscoring the efficiency and practicality of our method for modifying complex natural molecules. We subsequently evaluated 1,1-disubstituted alkenes bearing methyl and primary alkyl substituents (3af–3ak). Under the optimized conditions, our strategy effectively differentiated between primary and secondary allylic C–H bonds, selectively cleaving the primary site to furnish the desired products exclusively. For 1,1-disubstituted alkenes containing both methyl and secondary alkyl groups (3al), the reaction predominantly proceeded via cleavage of the methyl C–H bond, affording the corresponding products. Additionally, a diverse array of disubstituted olefins—including simple alkyl, aryl, and heteroaromatic alkenes incorporating pyridine or benzodioxole motifs (3am–3au)—served effectively as fluoroalkyl radical acceptors, delivering the target products in moderate yields. In all cases, selective activation was confined to a single C–F bond; however, tetrasubstituted alkenes failed to react due to steric encumbrance, while monosubstituted alkenes exhibited low conversion and poor selectivity.

Fig. 2: Scope of alkene substrates.
Fig. 2: Scope of alkene substrates.
Full size image

Reaction conditions: 1a (0.2 mmol), 2 (6 equiv.), fac-Ir(ppy)3 (0.5 mol%), methylcobalamin (3 mol%), HBpin (0.3 equiv.), and DIPEA (0.25 equiv.) in degassed 6 mL 2-ethoxyethanol, blue LEDs, 36 h, under nitrogen. a1a (0.2 mmol), alkene 2 (6 equiv.), fac-Ir(ppy)3 (1 mol%), methylcobalamin (3 mol%), DIPEA (0.5 equiv.), NaOAc (0.5 equiv.), LiF (0.5 equiv.) in degassed 6 mL 2-ethoxyethanol, blue LEDs, 48 h, under nitrogen. b1a (0.2 mmol), 2 (6 equiv.), fac-Ir(ppy)3 (1 mol%), Co-TPP (3 mol%), HBpin (0.3 equiv.) and DIPEA (0.25 equiv.) in degassed 6 mL 2-ethoxyethanol, blue LEDs, 48 h, under nitrogen. c1b (0.2 mmol), 2 (6 equiv.), fac-Ir(ppy)3 (3 mol%), methylcobalamin (3 mol%), Cs2CO3 (1 equiv.) HBpin (1.2 equiv.) and DIPEA (3 equiv.) in degassed 6 mL DMF, 1 atm CO2 (headspace volume: 14 mL), blue LEDs, 36 h. All yields are isolated yields. All regioselectivity >20:1.

To overcome the limitations encountered with monosubstituted alkenes, we optimized the reaction parameters (see Supplementary Table 4). By substituting methylcobalamin with Co-TPP, we achieved effective allylic fluoroalkylation of monosubstituted alkenes (3ba–3bi), exhibiting both exceptional site-selectivity and remarkable stereoselectivity (E/Z > 20:1). Our catalytic system also accommodates trifluoroacetamide substrates; employing Ir(ppy)3 to reduce CO2 generates the key intermediate CO2•–, which in turn facilitates the selective defluoroalkylation of trifluoroacetamides with excellent chemoselectivity and regioselectivity (4a–4j). Furthermore, trisubstituted alkenes—whether simple or derived from natural terpenes—were well tolerated, affording the desired products in moderate to good yields. Collectively, these findings underscore the versatility and robustness of our method, establishing a valuable platform for the selective conversion of diverse olefin substrates into complex fluorinated architectures.

The scope of perfluoroalkylarenes and trifluoroacetamides was further examined using 2,3-dimethylbut-1-ene (2al) under the standard conditions (Fig. 3). In addition to the C4F9 group, perfluoroalkyl moieties including C3F7, C6F13, C8F17, and C10F21 underwent selective allylic functionalization via benzylic C(sp3)–F bond cleavage (5a–5d). Furthermore, arenes bearing additional electron-withdrawing groups, such as cyano and alkynyl substituents, participated effectively in the reaction (5e–5f). N-Aryl trifluoroacetamides featuring a diverse range of functional groups were converted selectively to the corresponding products with high chemoselectivity (5g–5l). For instance, substituents such as methyl, methoxy, 3-fluoro, and 4-chloro were well tolerated, with selective reduction occurring exclusively at the C–F bonds adjacent to the carbonyl group, while the CF3 moiety remained intact (5j), suggesting that C–F bond cleavage is facilitated by the proximal carbonyl. In addition, sensitive functionalities such as alkynes were compatible under the reaction conditions (5n), whereas tertiary amides (5o–5q) exhibited only trace reactivity (<5%). Next, the scope of biologically derived molecules was examined under the standard conditions. Derivatives of D-allofuranose, citronellol, D-menthol, pregnenolone, and cholesterol (5r–5v) underwent the defluoroallylation, affording the corresponding products in moderate to good yields.

Fig. 3: Scope of perfluoroalkylarenes and polyfluorinated aliphatic amides for allylic fluoroalkylation.
Fig. 3: Scope of perfluoroalkylarenes and polyfluorinated aliphatic amides for allylic fluoroalkylation.
Full size image

Reaction conditions: 1 (0.2 mmol), 2 (6 equiv.), fac-Ir(ppy)3 (0.5 mol%), methylcobalamin (3 mol%), HBpin (0.3 equiv.), and DIPEA (0.25 equiv.) in degassed 6 mL 2-ethoxyethanol, blue LEDs, 36 h, under nitrogen. a1 (0.2 mmol), 2 (6 equiv.), fac-Ir(ppy)3 (3 mol%), methylcobalamin (3 mol%), Cs2CO3 (1 equiv.) HBpin (2 equiv.) and DIPEA (3 equiv.) in degassed 6 mL DMF, 1 atm CO2 (headspace volume: 14 mL), blue LEDs, 36 h. All yields are isolated yields. All regioselectivity >20:1.

To further demonstrate the synthetic utility of this methodology, representative products 3bb and 4g were subjected to downstream transformations (Fig. 4). Bromination of 3bb with Br2 afforded dibromide 6, while epoxidation of 3bb yielded the corresponding epoxide 7. Catalytic hydrogenation of 4g with Pd/C gave product 8. In addition, treatment of 4g with a BH3–THF complex reduced both the carbonyl group and the C=C bond, affording β-fluoroamine 9.

Fig. 4: Synthetic applications.
Fig. 4: Synthetic applications.
Full size image

a Bromination of 3bb with Br2 in DCM. b NBS-mediated oxidative functionalization of 3bb in tBuOH/H2O followed by NaOH. c Hydrogenation of 4g over Pd/C under H2. d Reduction of 4g with BH3–THF.

To gain more insights into the reaction mechanism, we conducted a series of mechanistic experiments (Fig. 5a–e). Radical trapping experiments revealed that the addition of 2 equivalents of either TEMPO or BHT significantly suppressed the transformation, and HR-MS analysis confirmed the formation of the corresponding fluoroalkyl adducts, supporting a radical pathway (Fig. 5a and Supplementary Figs. 1–3). Furthermore, radical-clock experiments employing (–)-β-pinene with both substrate classes (methyl 4-(perfluorobutyl)benzoate 1a and 2,2,2-trifluoro-N-phenylacetamide 1b), afforded the corresponding ring-opened adducts in 85% and 55% yield, respectively, under the standard conditions, further corroborating the involvement of fluoroalkyl radicals (Supplementary Figs. 4 and 5). Monitoring of the reaction revealed the formation of formate, supporting the involvement of the CO2•− intermediate generated via single-electron reduction of CO247 (Fig. 5b and Supplementary Figs. 6–9). EPR spectroscopy further confirmed this hypothesis by directly detecting the characteristic signal of CO2•− under photocatalytic conditions containing only fac-Ir(ppy)3 and CO2, thus providing unequivocal evidence for the single-electron reduction of CO2 (Fig. 5c and Supplementary Fig. 10). Stern–Volmer luminescence quenching studies revealed that the excited state of fac-Ir(ppy)3 was quenched by perfluoroalkylarene 1a, whereas 2al and DIPEA exhibited only minor effects. These results support that the photocatalytic cycle is initiated by single-electron transfer (SET) from the excited photocatalyst to 1a. (Fig. 5d and Supplementary Figs. 11–14).

Fig. 5: Mechanistic studies.
Fig. 5: Mechanistic studies.
Full size image

a Radical inhibition experiments. N.d. not detected. b Detection of HCO2− and (C2O4)2− formation. c EPR spectroscopy of CO2•− formation. d Stern−Volmer luminescence quenching experiments. e Time-course 1B/19F NMR study of [BF2(pin)]− formation in the presence of HBpin.

In the trifluoroacetamide system, neither 2,2,2-trifluoro-N-phenylacetamide 1b, 2al, nor DIPEA quenched the excited state of fac-Ir(ppy)3 (Supplementary Figs. 15–17). These observations collectively suggest that the photocatalytic cycle in this system is not initiated by SET from the excited photocatalyst to the amide substrate. Instead, based on the EPR evidence for CO2•− formation under conditions containing only fac-Ir(ppy)3 and CO2, it is proposed that the excited photocatalyst transfers an electron to CO2, generating the CO2•− intermediate that subsequently engages in the catalytic cycle. Finally, time- course 11B and 19F NMR analyses revealed the progressive emergence of a 11B triplet at δ = +5.7 ppm, accompanied by a 19F multiplet at δ ≈ −140 ppm, corresponding to the literature-reported signals for the [BF2(pin)]− anion48,49. These resonances were absent in the absence of HBpin, indicating that HBpin functions as a fluoride scavenger in the reaction (Fig. 5e and Supplementary Figs. 18–21).

Density functional theory (DFT) calculations were performed to further clarify the reaction mechanism and site-selectivity control (Fig. 6). The computational findings indicate that the fluorobenzyl radical 2R1 initially adds to alkene via 2TS-1 with an energy barrier of 18.1 kcal/mol, resulting in the formation of a stable alkyl radical intermediate 2R2. As documented in prior studies45,46, this radical intermediate shows β-hydrogen atom transfer (β-HAT) reactivity in the presence of Co-TPP(II) ([2CoII]). The Gibbs free energy differences for the HAT processes at the Z/E configurations of the radical center were calculated to be 8.0 kcal/mol (via OSSTS-2) and 13.1 kcal/mol (via OSSTS-3), respectively. Product 3bb is formed via the cobaloxime(II)-assisted HAT transition state OSSTS-2. Subsequently, the exergonic proton transfer between the reactive [1CoIII]-H and base, such as DIPEA, generates [1CoI] via 1TS-5. This process is exergonic by 2.9 kcal/mol. Finally, through oxidation promoted by the 2IrIV compound, the [1CoI] intermediate is converted back to the active [2CoII] catalyst, facilitating catalytic turnover, thus enabling catalyst regeneration. Moreover, there might be another adjacent position for the β-HAT process in the radical intermediate 2R2. The Gibbs free energy barrier of OSSTS-4 is 12.7 kcal/mol, which is 4.7 kcal/mol higher than that of OSSTS-2. The higher energy barrier can be attributed to the increased ligand-radical steric repulsion (see Fig. S25 for details).

Fig. 6: DFT calculations and proposed mechanism.
Fig. 6: DFT calculations and proposed mechanism.
Full size image

a DFT study of 3bb formation through the Giese addition of radical Rf and the Co-TPP (II)-assisted HAT process. Free energy profiles were calculated at the M06/6-311+G(d,p)-SDD/SMD(2-Ethoxyethanol, eps = 13.38)//B3LYP-D3(BJ)/6-31G(d)-LANL2DZ level of theory. b Quantitative steric-electronic effects dissection (QSED) of transition states TS-2 and TS-3 through the combination of distortion-interaction analysis and energy decomposition analysis (EDA). The EDA calculation was performed at M06-L/6-311+G(d,p)-SDD/SMD(2-Methoxyethanol) level of theory based on structures optimized using B3LYP-D3(BJ)/6-31G(d)-LANL2DZ. c Proposed mechanism.

To further clarify the E/Z-selectivity of product and determine the influencing factors, the quantitative steric-electronic effects dissection (QSED) model developed in our group50 is utilized to quantify the energies of steric effect (ΔEsteric), electronic effect (ΔEelec), and dispersion interaction (ΔEdisp). The distortion-interaction analysis suggests that the difference in interaction energies (ΔΔEint = 6.2 kcal/mol) is critical for the E/Z-selectivity. Subsequently, the detailed dissection results were illustrated in a pie chart (Fig. 4b) to present a better visual representation. Because the electronic effect is decomposed into charge transfer (ΔEct), polarization (ΔEpol), and electrostatics (ΔEelstat), the pie chart consists of five sectors in total. Each sector represents the comparison of each energy term between OSSTS-3 and OSSTS-2 (ΔΔE). The positive value indicates that this term favors OSSTS-2, i.e., the E-selective HAT. According to the proportions of different sectors, the dispersion interaction is the dominant factor that promotes E-selective HAT transition state OSSTS-2. This quantitative dissection results account for the origin of E/Z-selectivity.

Based on the foregoing results, a plausible mechanism was proposed in Fig. 6c. Upon excitation, the Ir(III) complex initiates a SET to the polyfluoroalkyl compounds, yielding a radical anion alongside an Ir(IV) species. The resulting radical anion undergoes C–F bond fragmentation to generate a perfluoroalkyl radical and fluoride, with HBpin serving primarily as a fluoride scavenger that captures the liberated fluoride and suppresses back-electron transfer. The perfluoroalkyl radical then adds to the olefin in an anti-Markovnikov manner to form a carbon-centered radical, which is subsequently converted to the final product via a Co(II)-mediated selective hydrogen atom transfer (HAT) process that produces a Co(III)–H species. Deprotonation of the Co(III)–H intermediate regenerates the Co(I) complex51, which in turn reduces the Ir(IV) species back to Ir(III) via a second SET, thereby completing the catalytic cycle.

Discussion

In conclusion, we have developed a photoredox/methylcobalamin-catalyzed radical defluoroallylation method for polyfluoroalkyl compounds. This method selectively cleaves C(sp3)–F bonds under mild conditions, enabling site-selective allylic fluoroalkylation using readily available alkenes or natural terpenes. Mechanistic investigations reveal that the transformation proceeds through a HAT pathway mediated by methylcobalamin, which preferentially cleaves robust primary allylic C–H bonds. This precise reorganization of the olefin framework circumvents the formation of isomeric mixtures, thereby ensuring remarkable site control. Collectively, our findings establish a robust platform for the selective transformation of polyfluoroalkyl compounds and open new avenues for the streamlined assembly of complex fluorinated architectures.

Methods

General Information

All reactions were conducted under an inert nitrogen atmosphere unless otherwise stated. Solvents were degassed prior to use. Photochemical reactions were irradiated using 3 W blue LEDs. After completion, solvents were removed under reduced pressure, and crude mixtures were purified by column chromatography on neutral alumina or silica gel as specified. TLC was used to monitor reaction progress. All commercially available reagents were used without further purification unless noted.

Synthesis of fluoroalkyl radical precursors

Copper-mediated coupling procedure

A solution of copper powder (11.3 mmol, 0.723 g), 2,2′-bipyridine (0.434 mmol, 0.0677 g) and aryl iodide (ArI, 5.15 mmol) in DMSO (8 mL) was prepared, and perfluoroalkyl iodide (6.12 mmol) was added. The reaction mixture was stirred at 80 °C for 48 h. After cooling to room temperature, the mixture was extracted with chloroform (10 mL × 3). The combined organic layers were washed with aqueous NH3 (20 mL × 2) and brine (20 mL), dried over MgSO4, filtered, and concentrated under reduced pressure. The crude residue was purified by column chromatography to afford the corresponding fluoroalkyl radical precursor.

Synthesis of trifluoroacetamide derivatives

To a solution of aniline (5 mmol, 1 equiv.) and triethylamine (1.5 equiv.) in DCM (30 mL) at 0 °C, trifluoroacetic anhydride (1.2 equiv.) was added dropwise. The reaction mixture was allowed to warm to room temperature and stirred until TLC indicated full consumption of aniline. The solvent was removed under reduced pressure, and the resulting crude material was purified by flash chromatography on silica gel to give the desired trifluoroacetamide precursor.

Esterification procedure for preparing fluoroalkyl-containing precursors

A solution of alcohol (10 mmol) and 4-(perfluorobutyl)benzoic acid (2.83 g, 10 mmol) in DCM (30 mL) was cooled to 0 °C. 1-(3-Dimethylaminopropyl)-3-ethylcarbodiimide hydrochloride (12 mmol) and DMAP (1.2 mmol) were added. The mixture was stirred at room temperature until TLC indicated complete consumption of the alcohol. The reaction was quenched with water and extracted with DCM (3×). The combined organic layers were washed with brine, dried over Na2SO4, filtered, and evaporated. The residue was purified by flash chromatography (petroleum ether/ethyl acetate = 20:1) to afford the desired ester precursor.

Defluoroallylation of perfluoroalkylarenes with olefins

A solution of perfluoroalkylarenes 1 (0.2 mmol), olefins (6 equiv.), fac-Ir(ppy)3 (0.5 mol%), methylcobalamin (3 mol%), HBpin (0.3 equiv.), N, N-diisopropylethylamine (0.25 equiv.) in degassed 2-ethoxyethanol (6.0 mL) was stirred under nitrogen atmosphere and irradiated by 3 W blue LEDs for 36 h. After completion of the reaction, the solvent was removed under reduced pressure by rotary evaporation, and then purified by column chromatography on neutral alumina.

Defluoroallylation of perfluoroalkylarenes with terpenes

A solution of perfluoroalkylarenes 1 (0.2 mmol), olefins (6 equiv.), fac-Ir(ppy)3 (1 mol%), methylcobalamin (3 mol%), NaOAc (0.5 equiv.), LiF (0.5 equiv.), N, N-diisopropylethylamine (0.5 equiv.) in degassed 2-ethoxyethanol (6.0 mL) was stirred under nitrogen atmosphere and irradiated by 3 W blue LEDs for 36 h. After completion of the reaction, the solvent was removed under reduced pressure by rotary evaporation. Then purified by column chromatography on neutral alumina.

Defluoroallylation of perfluoroalkylarenes with monosubstituted olefins

A solution of perfluoroalkylarenes 1 (0.2 mmol), olefins (6 equiv.), fac-Ir(ppy)3 (1 mol%), Co-TPP (3 mol%), HBpin (0.3 equiv.), N, N-diisopropylethylamine (0.25 equiv.) in degassed 2-ethoxyethanol (6.0 mL) were stirred under nitrogen atmosphere and irradiated by 3 W blue LEDs for 48 h. After completion of the reaction, the solvent was removed under reduced pressure by rotary evaporation, and then purified by column chromatography on neutral alumina.

Defluoroallylation of polyfluorinated aliphatic amides with olefins

A solution of perfluoroalkyl reagent 1, olefin 2 (6.0 equiv.), fac-Ir(ppy)3 (3 mol%, 3.9 mg), methylcobalamin (3 mol%, 8.0 mg), HBpin (30.7 mg, 1.2 equiv.), N,N-diisopropylethylamine (77.4 mg, 3 equiv.), and Cs2CO3 (65.2 mg, 1 equiv.) in degassed DMF (6.0 mL) was stirred under CO2 atmosphere (headspace volume: 14 mL) and irradiated by 3 W blue LEDs for 36 h. After completion of the reaction, the reaction mixture was extracted with ethyl acetate (15 mL × 3). The combined organic layers were washed with saturated brine (20 mL), then dried over anhydrous Na2SO4. The solvent was removed under reduced pressure using rotary evaporation. The crude product was then purified by column chromatography on neutral alumina.