Base-Promoted Radical Azofluoromethylation of Unactivated Alkenes

Abstract

The base induced reaction of aryl diazonium salts with commercially available CF₃SO₂Na/CF₂HSO₂Na allows for the generation of the corresponding diazene radicals along with fluoromethyl radicals. The addition of fluoromethyl radicals to alkenes with subsequent diazene trapping provides the azofluoromethylation products in good to excellent yields. This metal-free method under mild reaction conditions has broad functional group compatibility and is applicable in the late-stage modification of various natural products and bioactive molecules.

Graphical Abstract

The selective installation of fluoroalkyl groups (such as CF₃ and CF₂H) into pharmaceutical and agrochemical compounds, as well as functional organic materials, often delivers compounds with unique physicochemical and biological properties.^1–8 Therefore, new methodologies for the efficient selective incorporation of these substituents into diverse molecular structures have received increasing attention.^9–15 Notably, radical fluoromethylation of alkenes using fluoromethyl radical precursors (electrophilic or nucleophilic) has been extensively studied and proven highly useful.¹⁶ Alkenes are privileged chemicals due to their feedstock accessibility and versatile carbon-carbon double bonds.^17,18 Arguably, fluoromethylative difunctionalization of alkenes is one of the most straightforward strategies for the synthesis of the fluoromethylated building blocks.^19,20 Furthermore, the simultaneous introduction of the fluoromethyl group and another functional group is step-economical.

There are two popular strategies to generate fluoromethyl radicals. One strategy uses electrophilic fluoromethylating reagents such as Umemoto’s reagent, Togni’s reagents, or triflyl chloride, that generate the CF₃ radical from SET reduction processes in the presence of photoredox catalysts^21,22 or copper catalysts^23–28 (Scheme 1a). The other strategy employs chemical/electrochemical oxidation^29–38 of radical/nucleophilic trifluoromethylation reagents (e.g., Langlois or Baran reagents). (Scheme 1b) Despite their synthetic utility, those methods have drawbacks, chiefly among them, expensive noble metal catalysts, complicated reaction setups, electrolytes, or the use of an excessive amount of oxidants, which inevitably generate environmentally hazardous waste or significantly reduce functional group compatibility. What is more, the majority of these studies have focused on styrene and other activated alkenes. Developing a sustainable and practical approach for the fluoroalkylation functionalization of unactivated alkenes is still an unmet challenge. We are glad to report a base-induced three-component radical azo-tri(di)fluoromethylation of alkenes under mild reaction conditions using easily accessible diazonium salts and commercially available, inexpensive CF₃SO₂Na/CF₂HSO₂Na as CF₃/CF₂H sources (Scheme 1c).

Scheme 1.

Previous work on radical fluoromethylative difunctionalization of alkenes

Our method does not require an external oxidant because the diazonium salt plays a dual role as a diazene source and oxidant. It is worth noting that diazenes are used as dyes,³⁹ pharmaceuticals,^40–42 and photoswitches.^43,44 The catalytic synthesis of diazenes by diazonium salt trapping of alkyl radicals is underexplored.^45–48 Our mild protocol offers the first metal-free azotrifluoromethylation of olefins without the use of stoichiometric oxidants, with broad functional group compatibility.

Several reports have indicated that diazonium salt could generate aryl or diazene radicals with a base.^49–52 Since this process involves an electron absorption by a diazonium salt, we speculated that the Langlois’ reagent could serve as an electron donor that can be oxidized to a trifluoromethyl radical, which may further react with an alkene and produce an azotrifluoromethylated product. To prove our hypothesis, we chose allylbenzene 1a, 4-bromophenyldiazonium tetrafluoroborate 2a, and CF₃SO₂Na (Langlois’ reagent) to test a base-catalyzed three-component azotrifluoromethylation. The results are summarized in Table 1. We screened different organic and inorganic bases and found that the reaction indeed generated the azotrifluoromethylation product and that tetrabutylammonium acetate (TBAOAc) gave the highest yield (entries 1–9), and that acetonitrile or DMF were superior solvents. The addition of water was conducive to the reaction (entries 9–17). The amount of TBAOAc was also crucial because the reaction gave lower yields with either higher or lower molar equivalent of TBAOAc (entries 11–13).

Table 1.

Reaction Conditions for the Optimization of the Azotrifluoromethylation of Alkenes^a

entry	solvent	base	conc.	yieldb
1	DMF/H2O = 3/1	LiOAc	0.2 M	35%
2	DMF/H2O = 3/1	kOAc	0.2 M	46%
3	DMF/H2O = 3/1	CsOAc	0.2 M	34%
4	DMF/H2O = 3/1	K2CO3	0.2 M	37%
5	DMF/H2O = 3/1	Py	0.2 M	21%
6	DMF/H2O = 3/1	Et3N	0.2 M	54%
7	DMF/H2O = 3/1	DIPEA	0.2 M	20%
8	DMF/H2O = 3/1	DBU	0.2 M	48%
9	DMF/H2O = 3/1	TBAOAc	0.2 M	61%
10	DMF	TBAOAc	0.2 M	55%
11	MeCN/H2O = 3/1	TBAOAc	0.2 M	58%
12c	MeCN/H2O = 3/1	TBAOAc	0.2 M	31%
13d	MeCN/H2O = 3/1	TBAOAc	0.2 M	44%
14	MeCN	TBAOAc	0.2 M	54%
15	Acetone/H2O=3/1	TBAOAc	0.2 M	54%
16	Acetone	TBAOAc	0.2 M	46%
17	DCM/H2O = 3/1	TBAOAc	0.2 M	30%
18e	DMF/H2O = 3/1	TBAOAc	0.2 M	28%
19f	DMF/H2O = 3/1	TBAOAc	0.2 M	17%
20g	DMF/H2O = 3/1	TBAOAc	0.2 M	26%
21	DMF/H2O = 3/1	TBAOAc	0.1 M	14%
22	DMF/H2O = 3/1	TBAOAc	0.5 M	78%

^aUnless otherwise noted, reactions were conducted with: a solution of 2a (2 equiv) in 200 μL mixed solvent (DMF/H₂O = 3/1) was added dropwise to the mixture 1a (0.2 mmol), base (0.5 equiv) and 3 (2 equiv) in 800 uL mixed solvent (DMF/H₂O = 3/1) at −10 °C, then the reaction stirred at rt for 1h.

^bYields were determined by ¹⁹F NMR using 4-fluoroaniline as an internal standard.

^c0.1 equiv of TBAOAc.

^d1 equiv of TBAOAc.

^eRoom temperature.

^f3 (1.5 equiv) and 2a (1.5 equiv) were used.

^g3 (2 equiv) and 2a (1.5 equiv) were used.

The temperature during the diazonium salt addition also impacted the reaction efficiency. At room temperature, the reaction produced fierce bubbling and lower yields. In contrast, if the reaction mixture was kept at −10 °C during the diazonium addition, the reaction was smoother and the yield higher (entry 18 vs. entry 9). Further, two equivalents of both diazonium salt and Langlois reagent were required (entry 9 vs. entries 18, 19). Remarkably, increasing the reaction concentration to 0.5 M increased the reaction yield to 78% (entry 22).

With optimized reaction conditions in hand, we explored the reaction scope. As shown in Scheme 2, all the monosubstituted, disubstituted, and trisubstituted alkenes showed good to excellent yields of the corresponding products. Linear (4d, 4t) and cyclic (4u, 4v) internal alkenes also displayed good yields of the corresponding azo compounds, albeit with varying degrees of diastereoselectivity. A wide range of functional groups such as esters (4b, 4f-4i, 4q, 4r, 4z), ethers (4f, 4j-4n, 4p, 4s, 4y), nitro (4g, 4k), nitriles (4c, 4j), aldehydes (4f, 4m), ketones (4n, 4x), alcohol (4w), sulfonate (4e, 4o, 4z) were tolerated in this protocol. Acceptable to good yields were also obtained with heterocyclic substrates like thiophene (4q) and furan (4h). We then explored late-stage azotrifluoromethylation of natural products (4w, 4x) and biologically active molecule derivatives (4y, 4z). We found that the natural products (−)-ß-citronellol (4w) and nootkatone gave excellent yields. Methyl eugenol (4y), an active natural ingredient pollinator attractant, furnished the azotrifluoromethylation product in an acceptable yield. Our protocol provides an easy-to-use synthetic tool for the modification of drug molecules. For example, we obtained a derivative of Probenecid (4z) in 85% yield. Probenecid is a prototypical uricosuric agent used to treat patients with renal impairment. These examples further demonstrate that our azotrifluoromethylation protocol is suitable for the late-stage, protecting-group-free modification of biologically interesting molecules.

Scheme 2.

Substrate Scope of Alkenes^{a,b a}Reaction Conditions: A solution of diazonium salt 2a (2 equiv) in 200 uL mixed solvent (DMF/H₂O = 3/1) was added dropwise to the mixture of alkene 1 (0.2 mmol), TBAOAc (0.5 equiv) and 3 (2 equiv) in 200 uL mixed solvent (DMF/H₂O = 3/1) at −10 °C, then the reaction stirred at rt for 1h.

^bIsolated yield; ^c Diastereomeric ratios were determined by ¹⁹F NMR analysis of the crude mixture.

Encouraged by these results, we explored the scope of diazonium salts under the standard conditions (Scheme 3). It was found that aryldiazonium salts 2b-k bearing electron-withdrawing or -donating groups at the para-, meta-, or ortho-positions were well tolerated, affording the corresponding products 5a–h in moderate to good yields. However, electron-rich diazonium salt gave a lower yield than the electron-deficient diazonium salts (5g vs 5d-f). The reason for this result could be attributed to electron-rich diazonium salts being less stable under basic conditions.

Scheme 3.

Substrates Scope of Diazonium Salts^{a,b a}Reaction conditions: A solution of diazonium salt 2 (2 equiv) in 200 uL mixed solvent (DMF/H₂O = 3/1) was added dropwise to the mixture of alkene 1b (0.2 mmol), TBAOAc (0.5 equiv) and 3 (2 equiv) in 200 uL mixed solvent (DMF/H₂O = 3/1) at −10 °C, then the mixture was stirred at rt for 1h. ^bIsolated yield.

In contrast to various methods for the synthesis of trifluoromethylated organic substrates, direct difluoromethylation is still underdeveloped,^53–63 albeit the difluoromethyl group (CF₂H) is an intriguing structural motif in drug design.^64,65 We are glad to find that our protocol can also be applied to the azodifluoromethylation of alkenes by just switching the Langlois’ reagent with the commercially available CF₂HSO₂Na. A simple reaction condition optimization showed that using CF₂HSO₂Na as CF₂H source, MeCN/H₂O = 3/1 as the solvent, and KOAc as the base at 0.5 M concentration delivered the azodifluoromethylated product at an acceptable 53% yield (Table 2).

Table 2.

Reaction Conditions for the Optimizations of the Azodifluoromethylation of Alkenes^a

Entry	Base	Solvent	Conc.	Yield (%)b
1	TBAOAc	CH3NO2 / H2O = 3/1	0.5 M	27
2	TBAOAc	Acetone / H2O = 3/1	0.5 M	45
3	TBAOAc	DMF/ H2O = 3/1	0.5 M	33
4	TBAOAc	MeCN / H2O = 3/1	0.5 M	45
5	TBAOAc	MeCN / H2O = 3/1	0.1 M	51
6	TBAOAc	MeCN / H2O = 3/1	0.2 M	36
7	KOAc	MeCN / H2O = 3/1	0.5 M	53
8	LiOAc	MeCN / H2O = 3/1	0.5 M	34
9	Et3N	MeCN / H2O = 3/1	0.5 M	30
10	DBU	MeCN / H2O = 3/1	0.5 M	41

^aUnless otherwise noted, reactions were conducted with: a solution of 2a (2 equiv) in 200 uL mixed solvent (MeCN/H₂O = 3/1) was added dropwise to the mixture of 1q (0.2 mmol), base (0.5 equiv) and 6 (2 equiv) in 200 uL mixed solvent (MeCN/H₂O = 3/1) at −10 °C, then the reaction was stirred at rt for 1h.

^bYields were determined by ¹⁹F NMR using 4-fluoroaniline as an internal standard.

The azodifluoromethylation protocol was less efficient than azotrifluoromethylation, but still showed good substrate scope and functional group compatibility. As illustrated in Scheme 4, monosubstituted (7a, 7l), disubstituted (7b-7j) and trisubstituted alkenes (7k) were amenable substrates. A wide range of functional groups such as esters (7a-7b, 7j, 7l), ethers (7c-7h), nitro (7e), nitrile (7d), aldehyde (7h), ketone (7f), sulfonamide (7i), alcohol (7k), and alkyne (7j) were compatible with our protocol. We also applied this protocol to the late-stage azodifluoromethylation of a natural product (7k) and a biologically active molecule derivative (7l). Both compounds afforded the corresponding azodifluoromethylation products in acceptable yields.

Scheme 4.

Substrates Scope of Azodifluoromethylation of Alkene Using Various Alkenes^{a,b a}Reaction Conditions: A solution of diazonium salt (2 equiv) in 200 uL mixed solvent (MeCN/H₂O = 3/1) was added dropwise to the mixture of alkene (0.2 mmol), KOAc (0.5 equiv) and 6 (2 equiv) in 200 uL mixed solvent (MeCN/H₂O = 3/1) at −10 °C, then the reaction was stirred at rt for 1h. ^bIsolated yield. ^c Diastereomeric ratios were determined by ¹⁹F NMR analysis of the crude mixture.

To gain more insight into the reaction mechanism, we added TEMPO, a radical scavenger, to the reaction mixture (Scheme 5). As expected, the reaction led to a trace amount of the product (<2%), and TEMPO-CF₃ was observed (8%), which indicated a radical pathway. We postulated a mechanism initiated by the homolysis of the initially formed diazoacetate A (Scheme 6).^50,52 The resulting acetyloxy radical oxidizes the Langlois’ reagent and delivers the trifluoromethyl radical and acetate. The addition of the CF₃ radical to an alkene affords the β-CF_3-substituted radical intermediate C, SET reduction of cation radical D by NaSO₂CF₃ affords product 4 and the CF₃ radical, which could initiate the radical chain propagation (pathway I).^46,52,66 Alternatively, the trapping of the intermediate C by diazoacetate A (pathway II) could provide the azofluoromethylation product 4 and the acetyloxy radical. The resulting acetyloxy radical oxidizes the Langlois’ reagent and delivers the trifluoromethyl radical and acetate. The highly regioselective formation of the terminal fluoromethylated product implied that the addition of the CF₃ radical occurred first as it delivered a more stable β-CF₃-substituted radical intermediate.

Scheme 5.

Control Experiments

Scheme 6.

Plausible Mechanism for the Azotrifluoromethylation of Alkenes

In summary, we have developed a base-induced three-component radical azo-tri(di)-fluoromethylation of alkenes under mild conditions, using easily accessible diazonium salts and commercially available, inexpensive CF₃SO₂Na/CF₂HSO₂Na. This transition metal-free protocol is operationally simple, avoids the use of external stoichiometric chemical oxidants, and exhibits a broad substrate scope for the difunctionalization of a broad range of alkenes.