Over the past decade, the strategic development of introducing fluorine-containing groups into organic molecules has emerged as a prominent field of research [
1–
9]. Trifluoromethyl (CF
3), for instance, exhibits a strong electron-withdrawing effect and can significantly enhance key molecular properties, including lipophilicity, metabolic stability, and bioavailability [
10–
12]. Pyridines and their derivatives are prevalent nitrogen-containing heterocycles in pharmaceuticals, materials, and agrochemicals [
13–
18], and also serve as prominent ligands [
19]. Due to the beneficial effects conferred by CF
3 group introduction, CF
3-containing pyridine cores feature in numerous drugs and agrochemicals (
Fig. 1A) [
20–
24]. A prominent example is Tipranavir, a non-peptidic
anti-HIV drug featuring a trifluoromethylated pyridine scaffold, which exhibits approximately 10–fold higher antiviral activity than its phenyl-based counterpart [
25]. Traditional methods for pyridine functionalization rely on prefunctionalized substrates [
26–
35]. However, the step economy and ready availability of C–H bonds as substrates make direct trifluoromethylation of pyridine C–H bonds a highly attractive strategy [
36–
38]. This approach is particularly valuable for the late-stage functionalization of existing bioactive molecules, a capability crucial for accelerating drug candidate screening.
The nitrogen atom in pyridine possesses a lone pair of electrons, enabling protonation or coordination with Lewis acids, which significantly reduces the electron density of the heteroaromatic ring, particularly at the
ortho- and
para-positions [
39]. Consequently, current C-H functionalization of pyridines, including C-H trifluoromethylation, predominantly targets at the
ortho- and
para-positions [
40–
49]. However, the direct
meta-C-H functionalization of pyridines remains highly challenging. In recent years, a number of ingenious strategies have been developed for the
meta-C-H functionalization of pyridines, including the use of directing groups [
50–
52], non-directed metalation [
53–
57], temporary dearomatization strategies [
38,
58–
77] and ring-opening, functionalization, then ring-closing strategies [
78–
82]. Nevertheless, to the best of our knowledge, most of the direct C-H-trifluoromethylation methods have substrate dependence as well as low yields and poor selectivity, Studer and his co-workers conducted a dedicated study on
meta-C-H trifluoromethylation of pyridines [
37]. In 2022, Studer group employed a dearomatized oxazino-pyridine intermediate, which was produced by 1,4-Huisgen cycloaddition with pyridines [
70], for regioselective trifluoromethylation under light conditions, followed by acid-promoted rearomatization to obtain
meta-trifluoromethylated pyridines with high selectivity (
Fig. 1B).
Herein, based on temporary dearomatization strategy, we disclose a method to create a 1,4-dihydropyridinephosphonate that undergoes regioselective trifluoromethylation, followed by base-promoted rearomatization to afford
meta-trifluoromethylated pyridines in one-pot process (
Fig. 1C). This method tolerates a variety of functional groups and is applicable for late-stage
meta-C-H trifluoromethylation of pharmaceutical molecules containing pyridine motifs, while at the same time, it offers a complementary scope to the Studer approach.
The choice of pyridine
N-activating groups and nucleophiles is crucial for the process and directly impacts the selectivity and yield of dearomatization, the efficiency of trifluoromethylation and the ease of rearomatization. Based on Akiba's method [
83] for the synthesis of 1,4-dihydropyridinephosphonates, we implemented several optimizations to the procedure. We began our investigation with 3-(4-fluorophenyl)pyridine (
1a) as the model substrate with three electrophiles potentially suited as
N-activating reagents (
Ⅰ–Ⅲ,
Table 1) and three phosphites as nucleophiles (
Ⅳ–Ⅵ,
Table 1). It indicated that P(OEt)
3 was superior to P(OMe)
3 and P(O
iPr)
3 in forming 1,4-dihydropyridinylphonate from
1a when methyl chloroformate served as
N-activating reagent (entries 1–3). Comparable yields were obtained with ethyl chloroformate as the
N-activating reagent and P(OEt)
3 or P(O
iPr)
3 as the nucleophiles, both of which performed better than P(OMe)
3 (entries 4–6). Notably, N-Tf-dihydropyridinylphosphonate remained inaccessible through this strategy using P(OEt)
3, P(OMe)
3, or P(O
iPr)
3 (entries 7–9). The reaction performed in DCM provided lower yields (entry 10). Only 23% yield of the reaction was observed at 0℃ with methyl chloroformate, leaving substantial unreacted starting material (entry 11), while room-temperature conditions failed to initiate effective pyridine activation (entry 12).
Based on the results in
Table 1, methyl chloroformate was selected as the
N-activating reagent, and the 1,4-dihydropyridinephosphonate
2a derived from triethyl phosphite served as the model substrate for subsequent regioselective CF
3 introduction under copper catalysis, using 1-trifluoromethyl-1,2-benziodoxol-3(1
H)-one (Togni Ⅱ) as the CF
3 source. Through systematic experimentation, we achieved a 71% yield of the trifluoromethylation product
3a using catalytic CuBr(PPh
3)
3 (10 mol%) and 1.2 equiv. Togni Ⅱ (entry 1,
Table 2). Togni Ⅱ was found to be essential for activating substrate
2a to afford
3a. When alternative trifluoromethylation reagents (such as Togni I and Umemoto reagents) were employed, no significant associated trifluoromethylated products were detected (see Supporting information for further details). Given the critical role of copper salts play in enamine trifluoromethylation, systematic screening of copper catalysts was conducted, identifying CuBr(PPh
3)
3 as the optimal catalyst due to its superior efficacy (entries 2 and 3,
Table 2). Notably, no product formation was observed in the absence of copper catalyst (entry 4,
Table 2). In the testing of other solvents including 1,2-dichloroethane (DCE), acetonitrile (ACN), and dimethyl sulfoxide (DMSO), significantly decreased the conversion of
2a (entries 5–7,
Table 2). Subsequently, the rearomatization of
3a was optimized (
Table 3). In the presence of base,
3a was converted to product
4a, with 1,4-diazabicyclo[2.2.2]octane (DABCO) proving more efficient than
t-BuONa or 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU) (entries 2 and 3,
Table 3). The reaction proceeds optimally in DMSO (entries 4 and 5,
Table 3). Additionally, a one-pot, two-step procedure was attempted, affording
4a in 60% yield from
2a (entry 6,
Table 3). This yield proved comparable to that achieved through sequential reactions, while significantly streamlining the synthetic process.
Under the optimized conditions, the scope of the trifluoromethylation reaction was evaluated with various substrates (
Fig. 2). The method proved applicable to a range of functionalized pyridine derivatives. Replacement of the fluorine substituent on the aromatic ring with electron-donating groups (
e.g., methyl, methoxy,
tert-butyl) in mono- and disubstituted derivatives also provided the trifluoromethylated products
4c–
4g and
4m–
4p in moderate yields. Electron-withdrawing groups such as cyano and trifluoromethyl were also compatible, providing products
4h–
4i. Furthermore, carbon–halogen bonds remained intact under the reaction conditions, delivering products
4j–
4l with preserved halogen atoms in moderate yields. These halogenated derivatives serve as versatile synthetic building blocks for further coupling transformations. Notably, when an electron-rich thiophene ring occupied the
meta-position of pyridine, the reaction exclusively delivered
meta-selective trifluoromethylation on the pyridine ring (
4q). Our approach was also successful in achieving
meta-selective trifluoromethylation of pyridine derivatives containing heteroaromatic substituents (
4s–
4z), such as pyrimidine, pyrazine, and pyridazine, despite the significant electron deficiency caused by these heterocycles. Additionally, the protocol also tolerated functional groups such as esters and amides (
4aa–
4ac). Given the multifaceted roles of pyridines in drug development, the late-stage functionalization of pyridines could rapidly optimize a candidate's pharmacokinetic and pharmacodynamic (PK/PD) properties and facilitate the discovery of new drugs. This method was extended to bioactive molecule-derived substrates (
4ad–
4ag), including l-menthol, loratadine, estrone, and vonoprazan, underscoring its utility in medicinal chemistry. Unfortunately, this approach is restricted mostly to
meta-substituted pyridines, the yield of 2,3-disubstituted pyridine derivatives was relatively low (
4ah), sometimes with no product formed (
4ai). For the
ortho-aryl-substituted pyridines, significant steric hindrance prevented dearomatization (
4aj). Finally, a gram-scale reaction of
2b afforded compound
4b in 42% yield.
To gain a deeper understanding of the mechanism, some preliminary studies were conducted. The addition of 2.0 equiv. 2,2,6,6-tetramethyl-1-piperidinyloxy (TEMPO) or 2.0 equiv. butylated hydroxytoluene (BHT) as radical inhibitors completely inhibited the reaction. When TEMPO was reduced to 1.2 equiv., the yield of
3a decreased to 21%, while the TEMPO-CF
3 adduct was detected with a yield of 45% by
19F NMR (
Fig. 3A). Moreover, the structure of intermediate
3ac derived from
2ac was confirmed by single-crystal X-ray diffraction (SC-XRD) analysis (
Fig. 3B). Furthermore, monitoring the rearomatization process by
19F NMR spectroscopy revealed that the concentrations of two intermediates (
δ -70.76 and -60.50) changed as the reaction proceeded (
Fig. 3C). By comparing them with the signals of the starting material
3a (
δ -70.47) and the product
4a (
δ -62.42), in combination with the molecular weights of
5 monitored by HRMS, we surmised that the two intermediates were
D (
δ -70.76) and
E (
δ -60.50), respectively. According to the above experimental results and related literature studies [
84–
94], a plausible reaction mechanism is proposed (
Fig. 3D): (ⅰ) The Togni Ⅱ reagent undergoes copper-catalyzed single-electron transfer (SET) to generate a trifluoromethyl radical and a Cu(Ⅱ) specie
B; (ⅱ) This radical adds to the 1,4-dihydropyridinephosphonate
2a to give the radical intermediate
C; (ⅲ) Intermediate
C reacts with
B to yield intermediate
3a while regenerating the Cu(I) catalyst; (iv) Finally,
3a undergoes DABCO-mediated aromatization
via intermediates
D and
E to afford the
meta-trifluoromethylated pyridine product
4a.
In summary, we developed a novel dearomatization-rearomatization strategy based on 1,4-dihydropyridinephosphonate intermediates, which undergo copper-catalyzed regioselective C–CF3 bond formation with Togni Ⅱ reagent followed by DABCO-promoted rearomatization, achieving meta-C-H trifluoromethylation of pyridine derivatives. A variety of pyridines bearing either electron-donating or electron-withdrawing (hetero)aryl substituents were converted to their corresponding trifluoromethylated products. The reaction proceeded with high regioselectivity, and meta-trifluoromethylated products were obtained as the sole regioisomer. Carbon-halogen bonds, esters, aryl ethers, and amides were tolerated under the reaction conditions. This activation paradigm overcomes the inherent site-selectivity constraints of traditional aromatic systems, demonstrating broad applicability to various pyridine-containing heterocyclic substrates and successful implementation in late-stage trifluoromethylation of pharmaceutical molecules. We anticipate this work will contribute to the development of robust methods for meta-selective transformations of pyridines.
PDF
Full
82
45
关闭全屏
No. 86 Xueyuan South Road, Haidian District, Beijing
010-62662699(Operation and Technical Support)
qkjq@cast.org.cn
