Access to a new class of synthetic building blocks via trifluoromethoxylation of pyridines and pyrimidines

One-pot trifluoromethoxylation of functionalized pyridines and pyrimidines via OCF3-migraion.


Introduction
The triuoromethoxy (OCF 3 ) group has made a signicant impact in medicinal, agrochemical, life-and materials science research 1-5 since Booth and Burcheld reported the rst synthesis of triuoromethyl ethers in 1935. 6 The increasing importance of the OCF 3 group can be attributed to its unique structural and electronic properties. First of all, in aryl tri-uoromethyl ethers the OCF 3 moiety lies in the plane orthogonal to arene ring (Fig. 1a) 7 and studies have shown that this unusual orientation may be benecial for providing additional binding affinity in drug-target complexes. 8 In addition, the OCF 3 group is among the most electronegative groups (c(F) ¼ 4.0, c(OCF 3 ) ¼ 3.7). 9 Molecules bearing an electron-withdrawing group have better metabolic stability. Moreover, the OCF 3 group has an excellent lipophilicity (p x (SCF 3 ) ¼ +1. 44, p x (SF 5 ) ¼ +1. 23, p x (OCF 3 ) ¼ +1.04, p x (CF 3 ) ¼ +0.88, p x (OCH 3 ) ¼ À0.02); 10 compounds with higher lipophilicity show enhancement in their in vivo uptake and transport in biological systems. Therefore, the OCF 3 group is introduced into biologically active molecules to improve their efficacy and minimize their side effects (Fig. 1b). 1,2,5 Furthermore, incorporation of the OCF 3 group into organic molecules can increase their melting point and boiling point difference under ambient pressure, and lower their surface tension, dielectric constant, and pour point. 1,11,12 These properties are particularly useful in designing electronic devices and materials; as a result, the OCF 3 -containing molecules can be found in electro-optical materials used for the development of liquid crystal displays, 13 soluble organic semiconductor, 14 and melt-processable uoropolymers such as per-uoroalkoxy alkanes. 12 Given the unique properties of the OCF 3 group and the ubiquity of pyridines and pyrimidines in biologically active molecules and functional materials, triuoromethoxylated pyridines and pyrimidines could serve as valuable synthetic building blocks for the discovery and development of new drugs, agrochemicals, and functional materials. However, synthesis of OCF 3 containing heteroarenes through either O-CF 3 or C-OCF 3 bond formation remains a formidable challenge in organic synthesis (Fig. 1c). [1][2][3][4][5]15 Unlike its analogous methoxy (OCH 3 ) group, the OCF 3 group cannot be formed via tri-uoromethylation of hard nucleophiles such as phenoxides with CF 3 I through S N 2 type mechanism. 11,16,17 This is due to (i) strong electron repulsion between three uorine atoms and an incoming nucleophile; (ii) formation of energetically disfavoured CF 3 carbocation transition state structure (TS); and (iii) competing iodination of nucleophiles due to the reversed electron density. In addition, the thermal instability of transition metal-OCF 3 complexes (they readily decompose to form uorophosgene and metal uoride) 18 and the poor nucleophilicity of the OCF 3 anion (a reactive electrophile is needed for the C-OCF 3 bond formation) 19 have hampered the development of the C-OCF 3 bond formation through either transition metal-catalysed C-O bond formation or nucleophilic substitution. Strategies for the synthesis of triuoromethoxylated heteroaromatic compounds are very rare. [20][21][22][23][24][25] Leroux and co-workers reported a detailed examination of several different approaches and concluded that the presence of a chlorine atom at the a-, and/or a 0 -position of hydroxy-pyridines is critical (Fig. 1d). 20 Without it, little or no desired product was isolated. This requirement greatly limited its application. Recently, Qing and co-workers reported a novel, direct synthesis of pyridyl triuoromethyl ethers from unprotected hydroxypyridines. 25 However, excess amounts of reagents and oxidants were required. In addition, only two examples with moderate yield were reported. Due to the lack of a general synthetic method for the synthesis of tri-uoromethoxylated pyridines and pyrimidines, their full potential has not been fully exploited in pharmaceutical, agrochemical, and materials applications.
Herein, we report a scalable and operationally simple protocol for regioselective synthesis of triuoromethoxylated functionalized pyridines and pyrimidines. Several unique features distinguish our strategy from the existing approaches: (i) many substrates with complex skeletons are tri-uoromethoxylated at or below room temperature (17 out of the 30 examples); (ii) a wide range of functional groups and substitution patterns are tolerated; (iii) this transformation is amenable to gram-scale synthesis; (iv) halogen or amino group is used as synthetic handles for further elaborations, and (v) the operational simplicity of our protocol would render tri-uoromethoxylation available to broader synthetic community. More importantly, this strategy allows access to a new class of synthetic building blocks to aid the discovery and development of new functional molecules.

Results and discussion
It is known that the N-O bond is relatively weak (bond dissociation energy ¼ $57 kcal mol À1 ) due to the lone-pair electron repulsion between the nitrogen and the oxygen atoms. In addition, electron withdrawing O-substituent and/or electron donating N-substituent could promote heterolytic cleavage of the N-O bond to form nitrenium ion and oxy-anion. 26 Recently, we took advantage of these properties and successfully synthesized triuoromethoxylated aromatic compounds through O-triuoromethylation of N-aryl-N-hydroxylamine derivatives to form N-OCF 3 compounds followed by thermally induced OCF 3migration. 27 However, to apply this strategy to the synthesis of triuoromethoxylated heteroaromatic compounds such as pyridine and pyrimidine, two challenges had to be addressed.
First of all, reaction conditions for the synthesis of N-acetyl/ methoxycarbonyl-N-pyridinylhydroxylamine precursors are very limited, [28][29][30] because the presence of heteroarenes complicates their synthesis. For instance, reduction of nitro-pyridines to the corresponding N-pyridinyl-N-hydroxylamines oen accompanied by over reduction side-products (i.e. formation of aminopyridines). In addition, formation of undesired bis-protected side products and pyridinium salts during the protection of N-pyridinyl-N-hydroxylamines lowered the yields of precursors. Aer extensive optimization, we were able to obtain pure precursors in good to excellent yields. We have identied a robust catalytic hydrazine reduction protocol (using 5% rhodium on carbon as a catalyst) for converting nitropyridines and pyrimidines to Nheteroaryl-N-hydroxylamines. This method is general and high yielding; most of the reduction products can be used directly without further purication (only ltration through celite to remove Rh/C followed by the removal of the solvent is required). We have also observed that isolation of the intermediate hydroxylamines is not necessary; subsequent N-protection can be performed one-pot and the selectivity can be controlled by the rate of addition of acyl chloride or methylchloroformate at appropriate temperature (see ESI † for detail experimental procedures).
Several features of the reaction are noteworthy. First of all, the reaction is sensitive to the electronic properties of substituents on pyridine. Substrates with an electron donating substituent para to the protected N-hydroxylamine readily undergo rearrangement to yield the desired products of tri-uoromethoxylation at or below room temperature (2a-b, 2i, 2kn, 2p, 2r-2s, 2v-y). In the absence of such substituents, higher reaction temperatures are required for the OCF 3 -migration step (2c-h, 2o, 2q, 2t-u). These observations are consistent with the formation of nitrenium ion through heterolytic cleavage of N-O bond (vide infra). 26 Secondly, for the reactions that take place at or below room temperature, the OCF 3 group is introduced exclusively to the a 0 -position. [37][38][39] Since aand a 0 -carbon of pyridines are metabolically labile sites, incorporation of an electron withdrawing OCF 3 group to the a 0 -position could improve their metabolic stability. 40,41 If the a 0 -position is blocked, product of g-OCF 3 pyridine is formed instead (2g and 2h). Interestingly, atropisomers are obtained in these cases. This is because the OCF 3 group lies in the plane orthogonal to the pyridine ring (Fig. 1a), which prevents the free rotation of the adjacent amide or carbamate group (see ESI †). 7,[42][43][44][45] Finally, the regioselectivity erodes as the reaction temperature increases (2d-f, 2o, 2u).
To probe the applicability of the triuoromethoxylation reaction to other heteroarenes, pyrimidines substituted with benzimidazolyl (3a), indolyl (3b), methoxy (3c), phenoxy (3d), or estronyl (3e) groups were examined (Table 3). To our delight, these substrates were triuoromethoxylated to afford the corresponding desired products (4a-4e) in good yields. Notably, none of the triuoromethoxylated pyridines and pyrimidines reported here has ever been prepared before.
To ensure that our products can serve as useful building blocks for molecular screening, our protocol must be scalable and further functionalization of the triuoromethoxylated products must be possible. To evaluate the reaction efficacy on preparative scale, a gram-scale reaction of 1a (1.39 g, 5.00 mmol) was performed (Scheme 1a) and the efficiency of the small-scale reaction was retained upon scale-up. Our tri-uoromethoxylated products also proved to be versatile (Scheme 1b). For instance, 2a could be further elaborated through palladium-catalysed Suzuki and Sonogashira couplings to afford the desired products (6a, 8a) in good yields. In addition, deprotected amino-pyridine (2a 0 ) could be efficiently Table 2 Selected examples of trifluoromethoxylation of pyridines a Cited yields and isomeric ratios are of isolated material by column chromatography. RT then 50 C in CH 2 Cl 2 . b 4 C in CH 2 Cl 2 (0.01 M). c Following the O-triuoromethylation reaction in CH 2 Cl 2 at RT, the reaction mixture was concentrated, the residue was dissolved in MeNO 2 , and the resulting mixture was heated. d 120 C. e 80 C. f 60 C. g Atropisomeric ratio. Table 3 Selected examples of trifluoromethoxylation of pyrimidines a Cited yields are of isolated material by column chromatography. Following the O-triuoromethylation reaction in CH 2 Cl 2 at RT, the reaction mixture was concentrated, the residue was dissolved in MeNO 2 , and the resulting mixture was heated at 80 C. b CH 2 Cl 2 (0.01 M). c RT then 50 C in CH 2 Cl 2 (0.03 M). d RT then 50 C in CH 2 Cl 2 . incorporated into other molecules through amidation and palladium-catalysed Buchwald-Hartwig coupling (5a, 7a).
Although our strategy is operationally simple and scalable, has broad substrate scope, and tolerates a wide range of functional groups, this procedure, much like other methods, is not without limitations. First of all, substrates with the protected N-hydroxylamino-group at a-, g-, or a 0 -position do not give the product of triuoromethoxylation. Presumably, the formation of nitrenium ion is energetically disfavoured in these cases, because it involves placing the positive charge on the endocyclic nitrogen atom. 46 In addition, preparation of N-heteroaryl-N-hydroxylamine precursors is required for this transformation. Furthermore, Togni reagent I is relatively expensive, and thus large-scale synthesis would be costly. Therefore, more improvements are needed for the development of a truly general and industrially practical tri-uoromethoxylation reaction. Nevertheless, with the method accessing unprecedented and versatile synthetic building blocks in hand, the discovery and development of new pharmaceuticals, agrochemicals, and functional materials can be expected.
To gain some insight into the reaction mechanism, we performed reactions in the presence of radical trap butylated hydroxytoluene (BHT) (Scheme 2a). We chose to use substrate 1d because we could isolate the O-triuoromethylated intermediate 1d 0 and study each step (i.e. O-triuoromethylation and OCF 3migration) separately. Addition of BHT (1 equiv.) to a reaction mixture of 1d and Togni reagent I had detrimental effect to the formation of O-triuoromethylated N-hydroxylamine intermediate 1d 0 . This result strongly suggests the involvement of radical species in the reaction pathway, which is in agreement with literature precedents. 47,48 On the other hand, BHT did not affect the reaction yield for the OCF 3 -migration process (step 2, Scheme 2a). These experiments argue against the presence of long-lived radical species in the OCF 3 -migration process and are consistent with our previous nding. 27 Moreover, introduction of electron rich substituent para to the N-OCF 3 group facilitates the OCF 3migration process. These observations support the formation of nitrenium ion and triuoromethoxide. 26,27 On the basis of these results, a plausible mechanism for the triuoromethoxylation reaction is illustrated in Scheme 2b. Deprotonation of 1d forms N-hydroxyl anion 9, which undergoes single-electron transfer (SET) with Togni reagent I to generate N-hydroxyl radical 10, tri-uoromethyl radical, and alkoxide 11. 48 Reaction of N-hydroxyl radical and triuoromethyl radical affords the O-tri-uoromethylated hydroxylamine 1d 0 , which could be isolated and characterized. This intermediate will then undergo thermally induced heterolytic cleavage of the N-O bond to form a tight ion pair of nitrenium ion 12 and triuoromethoxide. Rapid recombination of this ion pair gives 13, which upon tautomerization (i.e. migration of proton) yields the desired product 2d.

Conclusions
In summary, we reported an operationally simple protocol for the regioselective triuoromethoxylation of functionalized pyridines and pyrimidines. The strategy uses commercially available and bench stable Togni reagent I, features a broad substrate scope, tolerates a wide range of functional groups, and is amenable to gram scale synthesis. With this procedure, a variety of highly functionalized pyridines and pyrimidines could be triuoromethoxylated under mild reaction conditions. Since heteroarenes are ubiquitous in biologically active natural products, pharmaceuticals, and agrochemicals, we expect that our work will provide valuable OCF 3 -containing heteroaromatic building blocks for the discovery and development of new drugs, agrochemicals, and functional materials.
from Stony Brook University in support of this work. K.N.L. is grateful for the Chemistry Graduate Research Fellowship provided by the Department of Chemistry. We thank TOSOH F-Tech, Inc. for their gi of TMSCF 3 for the preparation of Togni reagent I. Fuhua Zhao is acknowledged for his contribution to the preparation of 2-substituted nitropyridine precursors. We express our special thanks to the American Chemical Society Division of Organic Chemistry for providing summer undergraduate research fellowship (SURF) for Fuhua Zhao.