External Photocatalyst-Free Visible Light-Promoted 1,3-Addition of Perfluoroalkyl Iodides to Vinyldiazoacetates

Abstract

Open AccessCCS ChemistryRESEARCH ARTICLE1 Feb 2022External Photocatalyst-Free Visible Light-Promoted 1,3-Addition of Perfluoroalkyl Iodides to Vinyldiazoacetates Weiyu Li†, Xiaoyu Zhou†, Tiebo Xiao, Zhuofeng Ke and Lei Zhou Weiyu Li† School of Chemistry, School of Materials Science and Engineering, Sun Yat-sen University, Guangzhou 510006 †W. Li and X. Zhou contributed equally to this work.Google Scholar More articles by this author , Xiaoyu Zhou† School of Chemistry, School of Materials Science and Engineering, Sun Yat-sen University, Guangzhou 510006 †W. Li and X. Zhou contributed equally to this work.Google Scholar More articles by this author , Tiebo Xiao School of Chemistry, School of Materials Science and Engineering, Sun Yat-sen University, Guangzhou 510006 Google Scholar More articles by this author , Zhuofeng Ke *Corresponding authors: E-mail Address: [email protected] E-mail Address: [email protected] School of Chemistry, School of Materials Science and Engineering, Sun Yat-sen University, Guangzhou 510006 Google Scholar More articles by this author and Lei Zhou *Corresponding authors: E-mail Address: [email protected] E-mail Address: [email protected] School of Chemistry, School of Materials Science and Engineering, Sun Yat-sen University, Guangzhou 510006 Google Scholar More articles by this author https://doi.org/10.31635/ccschem.021.202000713 SectionsSupplemental MaterialAboutAbstractPDF ToolsAdd to favoritesDownload CitationsTrack Citations ShareFacebookTwitterLinked InEmail Alkenes and alkynes are typical substrates used in atom transfer radical addition (ATRA) reactions, resulting in the formation of vicinal disubstituted products. In this report, vinyldiazoacetates were first developed as the radical acceptors in an ATRA reaction of RfI, leading to 1,3-difunctional adducts. The reaction is driven by visible light, without the need for external photocatalysts and additives. A series of control experiments and density functional theory (DFT) calculations indicate the 1,3-addition proceeds via a radical-chain process and the initial Rf radical was generated by homolytic dissociation of a radical pair complex of triplet free vinylcarbene and perfluoroalkyl iodide. After postreaction isomerization, various 1-iodo-3-perfluoroalkyl-alkenes were obtained in good yields with high Z selectivity. The synthetic utility of the ATRA 1,3-adduct was demonstrated by cross-coupling reactions and defluorination of perfluoroalkyl groups. Download figure Download PowerPoint Introduction Perfluoroalkyl groups, especially the trifluoromethyl group, are the most popular fluorinated groups in the area of biological research due to their strong electron-withdrawing nature and large hydrophobicity area.1–4 Among the diverse perfluoroalkylation methods,5–9 the atom transfer radical addition (ATRA) reaction of RfX (X = I and Br) has attracted continuous attention because it can introduce a perfluoroalkyl group and a halide onto unsaturated C–C bonds in one step.10–18 Typical substrates for ATRA reactions are alkenes and alkynes, leading to vicinal disubstituted products (Scheme 1a). However, radical additions for the formation of 1,3-difunctionalized products are far more difficult to achieve, presumably owing to the lack of suitable radical acceptors. Radical ring-opening of cyclopropanes would be an obvious method, but only reactive Br-radical is effective for σ-bond cleavage.19–21 To make the process energetically feasible, ATRA reactions of activated alkyl halides with highly strained [1.1.1]propellanes have been developed (Scheme 1b-1).22–26 Recently, the Studer group27,28 disclosed an elegant radical 1,3-trifluoromethylation/alkynylation of allylboronic esters via 1,2-boron migration (Scheme 1b-2). Wu et al.29 showed a cobalt-catalyzed one-pot, two-step, three-component radical reaction between RfI, TMSCHN2, and alkynes to give 1,3-difunctionalized compounds through the combination of 1,1-ATRA to a carbenoid and 1,2-ATRA to an alkyne (Scheme 1b-3). Scheme 1 | (a–c) Radical 1,2 and 1,3-difunctionalization. Download figure Download PowerPoint Vinyldiazo compounds have a rich and varied reactivity due to the presence of two valuable conjugated functional groups (alkene and diazo group) in their structure.30,31 A number of powerful synthetic methods have been developed based on these versatile reagents, especially their use as three-carbon building blocks to access various carbo- or heterocyclic frameworks via metal-catalyzed [3C + n] cycloadditions (n = 1–5).32–36 Most methods rely on the initial formation of electrophilic metal carbene intermediates through a diazo decomposition. By contrast, the development of new reactions triggered by their alkene motif is still challenging because many of the vinyldiazo reagents are prone to 1,5-electrocyclization to form pyrazoles if the diazo functionality cannot be rapidly decomposed.37,38 This limitation also restricts their use in a radical reaction, even though vinyldiazo compounds can be considered as special types of alkenes. Although radical addition to vinyl azides and nitrogen extrusion to generate iminyl radicals is a well-known process,39–42 surprisingly, an analogous method for the generation of vinyl radicals through radical addition to vinyldiazo reagents has been unexplored. The noncovalent interaction between perfluoroalkyl halides and Lewis bases has been developed as a new protocol for the generation of perfluoroalkyl radicals.43 The success of these radical perfluoroalkylation reactions depends on the formation of photon-absorbing electron donor–acceptor (EDA) complexes, allowing an intermolecular electron transfer to deliver radical species without external photocatalysts.44 A variety of Lewis bases, such as amines,45,46 phosphines,47 and phenolates,48 have been reported as electron donors. Despite these significant advances, radical perfluoroalkylation based on the interactions between reactive intermediates and perfluoroalkyl halides remains scarce. Recently, Jurberg and Davies,49 Koenigs et al.,50–54 Zhou et al.,55 and others56–58 have reported the use of low-energy blue light to liberate free carbenes from donor–acceptor diazo compounds. We wonder whether it is possible to generate Rf radicals by forming a transient intermediate between perfluoroalkyl iodides and free carbene (either singlet or triplet), thus initiating the 1,3-ATRA addition of RfI to vinyldiazoacetates. Herein, we report a visible light-promoted 1,3-addition of perfluoroalkyl iodides to vinyldiazoacetates without external photocatalyst and additives (Scheme 1c). The control experiments and density functional theory (DFT) calculations indicate the reaction proceeds via a radical-chain process, and the initial Rf radical is generated by homolytic dissociation of a radical pair complex formed from triplet free vinylcarbene and perfluoroalkyl iodide. To the best of our knowledge, this is the first report of a vinyldiazo reagent acting as the radical acceptor, which might lead to the discovery of new methods via fundamentally distinct routes. Experimental Methods To a solution of vinyldiazoacetate 1 (0.4 mmol) in dichloromethane (DCM) (2.0 mL), RfI 2 (1.2 mmol) was bubbled using a gastight syringe. The solution was stirred at room temperature upon the irradiation of a 5 W blue light-emitting diode (LED) for 0.5–5 h. Subsequently, excess amounts of volatile RfI as well as the solvent were removed in vacuo. Mn2(CO)10 (15 mol %) and DCM (2.0 mL) were added to the resulting residue. The solution was irradiated with a 5 W blue LED for an additional 5 h at room temperature. After the removal of the solvent in vacuo, the residue was purified by column chromatography on silica gel to give product 3. Computational Methods All calculations were accomplished using Gaussian 09 package. DFT with M06-2X functional was performed in all the geometry calculations. The Stuttgart/Dresden effective core potential (SDD) effective core potential (ECP) basis was carried out to describe the iodine, and the 6-311++G(d,p) basis set for other atoms (H, C, N, O, and F) was used in the geometry optimization. Calculations of vibrational frequency and the intrinsic reaction coordinate (IRC) were prepared at the same level as above. The vibrational frequency calculation was given to examine the stable point on the potential energy surface. There was not any imaginary frequency for the intermediates but only one imaginary frequency for the transition states at 298.15 K and 1 atm. Solvation effects were performed to all geometry optimizations and frequency analysis in CH2Cl2 solvent with the solvation model based on density (SMD) method. Results and Discussion Optimization studies Recently, Ferreira et al.59 reported a [3 + 2] cycloaddition between alkenes and vinyldiazo reagents using a Cr or Ru complex as the photocatalyst, in which vinyldiazo compounds act as the nucleophiles. Encouraged by this report, we decided to use Ru(bpy)3Cl2 as the photocatalyst to initiate the generation of perfluoroalkyl radical via photoredox catalysis. Initially, a MeCN solution of n-butyl vinyldiazoacetate ( 1a) and CF3I ( 2a) was irradiated with a 5 W blue LED at room temperature in the presence of Ru(bpy)3Cl2 (1 mol %). To our delight, the desired 1,3-bifunctional adduct 3aa was obtained in 55% yield accompanying the pyrazole formation at a yield of 21% (Table 1, entry 1). Several photocatalysts, such as Ru(phen)3Cl2, fac-Ir(ppy)3, and eosin B, were then examined (Table 1, entries 2–4). Among them, the use of Ru(phen)3Cl2 can improve the yield of 3aa to 76% and minimize the yield of pyrazole 4 to 10%. The reaction was completely shut down by adding organic base N,N,N′,N′-tetraethylethylenediamine (TEEDA) or inorganic base Cs2CO3 (Table 1, entries 5 and 6). Interestingly, we found that the reaction can proceed well in the absence of photocatalyst, providing 3aa in 95% yield without any side products (Table 1, entry 7). A set of solvents were also examined, and DCM gave the best result, in which 3aa was obtained in 97% yield (Table 1, entries 8–10). Oxygen has no essential effects on the reaction as nitrogen protection provides the identical yield of 3aa (Table 1, entry 11). Further control experiments indicated that the irradiation of blue light was necessary. No product was detected when the reaction was carried out in the dark (Table 1, entry 12).a Under the above reaction conditions, the Z/E selectivity of 3aa is poor (about 1.6:1). However, it can be improved to <30:1 by a Mn-catalyzed isomerization using Mn2(CO)10 as the catalyst under the irradiation of blue light (see Supporting Information Table S1).60 Due to the simple and clean reaction conditions for 1,3-addition (only solvent and visible light), the resultant Z/E mixture of 3aa can apply to the isomerization without purification.b Table 1 | Visible-Light-Promoted 1,3-Iodotrifluoromethylation of 1aa Entry Catalyst (mol %) Base (equiv) Solvent 4 (%) 3aa (%) Z/Eb 1 Ru(bpy)3Cl2 (1) MeCN 21 55 1.6∶1 2 Ru(phen)3Cl2 (1) MeCN 10 76 1.6∶1 3 fac-Ir(ppy)3 (1) MeCN 16 62 1.7∶1 4 Eosin B (1) MeCN 13 43 1.5∶1 5 Ru(phen)3Cl2 (1) TEEDA (2) MeCN 26 23 1.6∶1 6 Ru(phen)3Cl2 (1) Cs2CO3 (2) MeCN 17 24 1.6∶1 7 — — MeCN 0 95 1.6∶1 8 — — DCM 0 97 1.6∶1 9 — — EtOAc 9 41 1.5∶1 10 Tetrahydrofuran (THF) 8 44 1.3∶1 11c — — DCM 0 95 1.6∶1 12d — — DCM 0 0 — aReaction conditions: 1a (0.4 mmol), 2a (1.2 mmol), solvent (2 mL), photocatalyst (1 mol %), 5 W blue LED, 30 min at room temperature. Yields were determined by 1H NMR using 1,3,5-trimethoxybenzene as the internal standard. bZ/E selectivity was measured by 1H NMR. cThe reaction was carried out under the protection of N2. dIn the dark; HOAc (1 M) was added to quench the reaction before exposure the reaction mixture to the ambient light. Scope of the reaction Having established the optimal reaction conditions, we explored the scope of the reaction using n-butyl vinyldiazoacetate 1a and a series of perfluoroalkyl iodides. As shown in Table 2, good yields of 3aa– 3af were obtained using n-CmF2m+1I (m = 1, 3, 4, 6, and 10) as well as i-C3F7I. Replacing a fluorine atom in CF3I by other electron-withdrawing groups, such as CO2Et ( 3ag), PhSO2 ( 3ah), and benzoyl ( 3ai), the yields were also satisfied. Less-activated 1,1,1-trifluoro-2-iodoethane was consistent with the present conditions, providing the 1,3-difunctional adduct 3aj in 71% yield (Z/E = 13∶1). When 1,2-diiodotetrafluoroethane was employed, only one C–I bond was cleaved ( 3ak). Unfortunately, the 1,3-addition of bromodifluoroacetate to 1a under the present conditions was unsuccessful due to the higher bond dissociation energy (BDE) of the C–Br bond. Table 2 | Substrate Scope of Perfluoroalkyl Iodidesa aReaction conditions: 1a (0.4 mmol), RfI (1.2 mmol), DCM (2.0 mL), blue LED, 0.5–1 h (as indicated in the table) at room temperature; then the solution was concentrated in vacuo, DCM (2.0 mL) and Mn2(CO)10 (15 mol %) were added, irradiating with blue LED for additional 5 h. Isolated yields based on 1a. bAt 5 mmol scale. cRfI was removed by flash column chromatography due to its high boiling point (b.p.). Next, we used CF3I 2a to examine its 1,3-addition to various alkenyldiazoacetates (Table 3). As expected, the nature of the ester substituent has little impact on the reaction outcome. Thus, 1,3-iodotrifluoromethylation of vinyldiazoacetates 1b– 1i provides the corresponding alkenyl iodides 3ba– 3ja in 45–85% yields and <30:1 Z/E selectivity. Several observations are noteworthy: (1) Inactive C(sp3)–Cl could survive in this ATRA-type reaction. 1,3-Addition of CF3I to 6-cholorohexyl vinyldiazoacetate 1c gives the desired product 3bc in 78% yield with high Z selectivity. (2) For substrates 1d, 1i, and 1j bearing an additional C–C double bond, as well as 1e possessing a terminal C–C triple bond in the ester motifs, these unsaturated groups remain intact. Due to the partial negative charge of diazo carbon, the vinyl group attached to the diazo unit is more electron-rich and easier to accept the addition of electrophilic fluoroalkyl radicals. (3) For allyl vinyldiazoacetate 1d and propargyl vinyldiazoacetate 1e, competitive intramolecular radical cyclization was not observed, indicating that the iodine trapping of vinyl radical is rapid.c (4) When vinyldiazoacetates bearing derivatives of Vitamin E ( 1g), glucofuranose ( 1h), cholesterol ( 1i), and oleanic ester ( 1j) were employed, a mixture of DCM/MeCN (1∶1) was the better solvent system because of the poor solubility of these substrates in DCM. Table 3 | Substrate Scope of Alkenyldiazoacetatesa aReaction conditions: 1b–1t (0.4 mmol), CF3I (1.2 mmol), DCM (2.0 mL), blue LED, 0.5–5 h (as indicated in the table) at room temperature; then the solution was concentrated in vacuo, DCM (2.0 mL) and Mn2(CO)10 (15 mol %) were added, irradiating with blue LED for additional 5 h. Isolated yields were based on alkenyldiazoacetates. bDCM/MeCN (1:1) as the solvents for both steps. c5 mol % of Mn2(CO)10 was used in the first step. 1,3-Addition of CF3I to vinyldiazoacetates substituted at the alkenyl moiety also worked well. Substrates with n-hexyl ( 1k) and N-Boc piperidinyl ( 1l) substitution at C-α provided the corresponding alkenyl iodides 1ka and 1la in yields of 66% and 71%, respectively. Treatment of C-β methyl vinyldiazo compound 1m with CF3I led to 1,3-difunctional adduct 3ma in 71% yield. However, only 3:1 Z/E selectivity was finally obtained after the Mn-catalyzed isomerization. Replacing the β-methyl with phenyl group ( 1n), the reaction appeared to be less efficient due to the fast intramolecular cyclization to form pyrazole. Gratifyingly, alkenyldiazo compounds 1p and 1q, derived from cyclohexanone and 2-indanone, were suitable substrates for the reaction, affording double bond migration exocyclic alkenyl iodides 3pa and 3qa in good yields. The reaction was also applied to the late-stage modification of estradiene derivative, which not only introduces a CF3 group to its skeleton but also leaves a vinyl C–I bond for further functionalization ( 3oa). Finally, we found that the reaction of CF3I with C-α aryl vinyldiazoacetates provided the 1,3-difunctional products 3ra– 3ta in good yields with <30∶1 Z/E selectivity, despite the radial addition to the C-β position of C–C double bond to give a more stable benzyl radical was supposed to occur. The regioselectivity is also attributed to the partial negative charge of the diazo carbon, which makes C-α more electron-rich than C-β. Thus, the addition of electron-deficient CF3 radical to C-α site is more favorable. To further explore the potential of this 1,3-difunctionalization reaction, cyclic vinyldiazo ester 1u was examined, which was easily prepared by the treatment of commercially available 3,5-dihydropyran-2-one with 4-acetamidobenzenesulfonyl azide ( p-ABSA) in the presence of 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU).61 As shown in Scheme 2, the reaction of 1u with CF3I, C4F9I, and ICF2CO2Et gave the corresponding 3-iodo-5-fluoroalkyl 2H-pyran-2-ones in 73–75% yields. Postreaction isomerization is unnecessary for these products. 2-Pyrones are an important class of lactones present in many natural products and pharmaceuticals.62 Since 2H-pyran-2-one can be easily oxidized to 2-pyrone,63 the reaction provides a facile method to introduce both iodo and fluoroalkyl group into 2-pyrones with high regioselectivity. Scheme 2 | Reaction of cyclic vinyldiazo ester 1u with RfI. Download figure Download PowerPoint Mechanistic studies A series of experiments were performed to gain insight into the mechanism of this ATRA-type 1,3-addition of RfI to alkenyldiazoacetates. We found the reaction was completely suppressed when 1 equiv of 2,2,6,6-tetramethyl-1-piperidinyloxy (TEMPO) was added. An adduct of CF3 radical and TEMPO could be detected by gas chromatography–mass spectrometry (GC-MS; see Supporting Information Scheme S1). The quantum yield (Φ) measured for the reaction of 1u and C4F9I was found to be 48 (λ = 436 nm in DCM, using potassium ferrioxalate as the actinometer), suggesting a radical-chain pathway. In most cases, we observed the rapid product formation within 30 min with poor Z/E selectivity. To figure out the original source of Rf radical, we used UV–vis absorption spectroscopy to investigate whether an EDA complex can be directly formed from vinyldiazoacetate 1u and C4F9I without the irradiation of visible light. However, neither a dramatic color change nor significant bathochromic shift in the absorption spectrum was observed (Figure 1a and Supporting Information Figure S1). Heating the DCM solution of 1u and C4F9I at 60 °C in the dark, no desired 1,3-adduct 3uc was detected (see Supporting Information Scheme S2). In addition, we did not find any new signals in 19F NMR by mixing 1u and C4F9I in CDCl3 (see Supporting Information Figure S3). These results indicated that both photon absorbing and thermally sensitive EDA complexes cannot be formed between vinyldiazoacetates and RfI. On the contrary, typical vinyldiazo compounds, such as 1a, 1p, and 1u, present absorbance spectra in the visible region (Figure 1b and Supporting Information Figure S2). When vinyldiazoacetate 1a and benzyl alcohol were irradiated with blue light, the O–H insertion product 5 was detected in 11% yield, suggesting the generation of a free vinyl carbene (Figure 1c and Supporting Information Scheme S3). Figure 1 | Mechanistic studies: (a) UV–vis absorption spectra of 1u and C4F9I. (b) UV–vis absorption spectra of 1a, 1p, and 1u. (c) The use of benzyl alcohol to trap free vinyl carbene. Download figure Download PowerPoint To better understand the mechanism of the present radical 1,3-addition, DFT was performed based on experimental observations. Control experiments have excluded the electron transfer from 1a to CF3I via EDA complex. Actually, the DFT calculation indicates the formation of 1a˙ +, CF3 radical, and iodine anion is an endothermic process (Figure 2, path a, ΔGS1 = 63.8 kcal/mol and ΔGT1 = 84.8 kcal/mol). Glorius and co-workers64 have demonstrated a triplet–triplet energy transfer activation of S–S σ-bond of disulfides. In this reaction, the direct energy transfer between excited 1a* with CF3I is less likely (Figure 2, path b, ΔGS1 = 1.2 kcal/mol and ΔGT1 = 22.3 kcal/mol). On the contrary, the decomposition of 1a* to free vinyl carbene 1 A can be more easily occur (Figure 2, path c) according to the calculated energies (ΔGS1 = −34.9 kcal/mol and ΔGT1 = −13.8 kcal/mol). Figure 2 | A possible mechanism for the initial radical generation. Download figure Download PowerPoint After confirming the generation of free vinylcarbene A, the free energies of the 1,3-addition via the interaction of A and CF3I were calculated (Figure 3). Concerning intersystem crossing (ISC) converting higher energy singlet state 1 A to the triplet state 3 A is generally a fast process,65,66 they are likely to produce two possible intermediates C: iodonium ylide complex 1 C (singlet) and radical pair 3 C (triplet).53,67 The iodonium ylide 1 C (ΔG = −0.3 kcal/mol) can be generated by the singlet species 1 B (ΔG = 6.1 kcal/mol) via the transition-state 1 TS-BC (ΔG = 13.0 kcal/mol), while the iodine transfer from the CF3I to 3 A through transition-state 3 TS-BC leads to the radical pair intermediate 3 C (ΔG = −2.5 kcal/mol). The structure of iodonium ylide 1 C is significantly different from radical pair 3 C. In iodonium ylide 1 C, the distance between iodine atom and free carbene is about 2.04 Å, which is shorter than that between iodine atom and CF3 fragment (2.26 Å), and the angle of C–I–C is about 99°. However, the iodine atom in radical pair 3 C is far away from the CF3 fragment (3.31 Å) and interacting with the free carbene center with the C–I–C angle of 177°. These are consistent with the calculated results of iodonium ylide and radical pair structures in previous studies.68–71 Meanwhile, the activation free energy of the radical pair formation pathway via transition-state 3 TS-BC is about 9.3 kcal/mol, which is lower than that of the iodonium ylide formation pathway by about 3.7 kcal/mol. The calculated results indicated that the radical pair formation pathway with the triplet state is favored in kinetics. In addition, the iodine atom is crucial for the formation of radical pair 3 C. The free-energy barrier of Cl- 3 TS-BC and Br- 3 TS-BC is higher than 3 TS-BC by 13.1 and 4.8 kcal/mol, which is in good agreement with the inefficiency of bromodifluoroacetate as a reactant observed experimentally (see Supporting Information Figure S4). It should be noted that the calculated energy of iodonium ylide 1 C is slightly lower than that of the radical pair intermediate 3 C by 1.5 kcal/mol, despite the free energy of 1 C with entropy correction is higher than that of 3 C by 2.2 kcal/mol, which suggests both the iodonium ylide 1 C and the radical pair 3 C might be the possible intermediates. However, the activation of the iodonium ylide 1 C via CF3 transfer must overcome an activation free energy of 14.8 kcal/mol ( 1 TS-C3aa). In contrast, the dissociation of radical pair 3 C spontaneously generates free CF3 radical and allylic radical D (ΔG = −10.0 kcal/mol). The allylic radical D involves two resonance forms Da (38.5%) and Db (61.5%) describing the radical position. The resonance form of Db is preferred, due to the strong electron-withdrawing conjugation property of the CO2nBu group. Nevertheless, a radical–radical combination of CF3 radical with Da leading to 3aa is more favorable (ΔG = −82.8 kcal/mol), while another isomer 3aa′ was not observed (ΔG = −70.5 kcal/mol). All these results support that the initial CF3 radical was generated by the formation of a radical pair complex. Figure 3 | Free-energy profiles for 1,3-addition of CF3I to 1a via radical–radical coupling pathway (in kcal/mol). Download figure Download PowerPoint The high quantum yield (Φ = 48) indicates a radical-chain process might be the major pathway to produce 3aa, and their potential energy profiles are shown in Figure 4. The addition of CF3 radical to the alkene motif of 1a through transition-state TS-1aE (ΔG = −7.5 kcal/mol) gives the stable intermediate E (ΔG = −51.6 kcal/mol). Nitrogen extrusion from E via transition-state TS-EF (ΔG = −47.6 kcal/mol) yields the vinyl radical F (ΔG = −74.3 kcal/mol). Finally, the vinyl radical F grabs the iodide from CF3I, affording 3aa (ΔG = −82.8 kcal/mol) via transition-state TS-F3aa (ΔG = −62.1 kcal/mol) concomitant with the regeneration of a CF3 radical. Scheme 3 | Suggested pathway for radical 1,3-addition. Download figure Download PowerPoint Based on the control experiments and DFT calculations, a plausible mechanism for the overall transformation is proposed in Scheme 3. The first step is the unimolecular loss of nitrogen from 1a upon blue-light irradiation, generating the singlet free carbene 1 A. Fast ISC of 1 A leads to triplet state 3 A, which subsequently forms the radical pair C by the interaction with CF3I. Then, the dissociation of radical pair C delivers CF3 radical and allylic radical D. In the propagation stage, the addition of CF3 radical to the C–C double of 1a forms intermediate E, followed by nitrogen extrusion to generate vinyl radical F. The reaction propagates by abstracting iodide from CF3I to give product 3aa and regenerates CF3 radical. The DFT calculated free-energy profiles for 1,3-addition of CF3I to 1a via radical propagation suggest that this process is kinetically and thermodynamically feasible, as shown in Figure 4. In the termination step, the combination of CF3• and allylic radical Da also delivers vinyl iodide 3aa, which is expected to be diffusion controlled. Figure 4 | Free-energy profiles for 1,3-addition of CF3I to 1a via radical propagation pathway (in kcal/mol). Download figure Download PowerPoint Synthetic utility The synthetic utility of the 1,3-difunctional adducts has also been demonstrated (Scheme 4). The Suzuki reaction of 3aa with phenylboronic acid proceeded smoothly, affording the coupling product 6 in 79% yield. No defluorinative products were detected even in the presence of Cs2CO3. (Z)-1,3-enyne 7 can be easily obtained in 71% yield by treating 3aa with phenylacetylene. Cu-catalyzed Ullmann-type coupling of 3aa with p-toluenethiol and NaN3 led to the corresponding heterosubstituted alkenes 8 and 9 in the yields of 73% and 55%, respectively. Interestingly, when 1-iodo-3-perfluoropropyl alkene 3ab was treated with 1,4-diazabicyclo[2.2.2]octane (DABCO), defluorinative product 1-iodo-4-fluorobutadiene 10 was isolated in a yield of 86% with high Z selectivity. Scheme 4 | Synthetic utility of the products. Download figure Download PowerPoint Conclusion Vinyldiazo compounds have been developed as radical acceptors in an ATRA reaction of perfluoroalkyl iodides, leading to 1,3-difunctional adducts. The reaction proceeds under mild, catalyst- and additive-free conditions, promoted by the irradiation of visible light. Key to the success of the reaction was the formation of a radical pair complex between perfluoroalkyl iodide and free triplet vinylcarbene, which delivers Rf radicals by homolytic dissociation of C–I bond to initiate the ATRA process. The Z/E selectivity of the resultant 1-iodo-3-perfluoroalkyl-alkenes could be improved up to <30:1 after a Mn-catalyzed postreaction isomerization. This study further demonstrates the possibility to use vinyldiazo reagents as three-carbon building blocks in a radical-based reaction. Further investigations are ongoing in our lab. Footnotes a HOAc (1M) was added to quench the unreacted 1a before exposure of the reaction solution to ambient light. b The excess CF3I should be removed before Mn2(CO)10 was added because the C–I bond of CF3I is weaker than that of the products. c Selective vinyl C–I termination was reported by Nagib in a ATRA reaction of in situ generated α-acetoxy iodides with allyl and propargyl propiolates, see ref 60. Supporting Information Supporting information is available, including the general information, preparation of vinyl diazoesters, optimization of reaction conditions, experimental procedures, characterization data, copies of NMR spectra for products, and Cartesian coordinates of intermediates and transition states. Conflict of Interest There is no conflict of interest to report. Acknowledgments The authors are grateful to the funds from the National Basic Research Program of China (no. 2016YFA0602900), the National Natural Science Foundation of China (nos. 21871300, 21673301, and 21973113), and