CHEMCATCHEM
COMMUNICATIONS
(entry 1). Alternatively, the presence of an array of copper salts
afforded the mono ortho-brominated product 2a in poor yield
(entries 2–6). The best yield was only 33% obtained with the
presence of CuBr (entry 6). The di-brominated product 4a was
not detected. With these unsuccessful outcomes, we turned to
search a more powerful halogenating reagent. Hinted by some
early reports[10] that acyl hypobromite BrOAc, generated in situ
from MOAc (M=H, Ag) and Br2, was appreciably a reactive
brominating reagents in EAS, as well as N-bromination, we
conceived that acyl hypohalites XOC(O)R may also be a more
powerful reagent than N-halosuccinimides (NXS; X=Cl, Br, I) in
metal-catalyzed direct CÀH halogenation of arenes. Moreover,
we are also interested in whether such reagents could be
formed through the combination of carboxylic acid and NXS
since NXS is safer than X2 for handling.
Table 2. Selective mono-bromination and chlorination of various 2-aryl
pyridines.[a]
Accordingly, the effect of carboxylic acid on the reaction was
examined (Table 1). Delightedly, we found that the addition of
2.0 equiv of HOAc in the reaction system increased dramatical-
ly the yield of the mono-brominated product 2a to 77% ac-
companied by the formation of 9% of di-brominated 4a
(entry 7). However, the excessive use of HOAc as solvent
showed a significant detrimental effect on the reaction, afford-
ing 2a in only 26% yield (entry 8). These results indicate that
the amount of carboxylic acids is of crucial importance for the
reaction. A careful optimization showed that 0.5 equiv of HOAc
was optimal in terms of yield and the ratio of 2a to 4a
(entry 10, 85% of 2a; 2a:4a>14:1). Our control experiment
revealed that the reaction did not proceed when NBS was
omitted from the reaction system (entry 10 vs. 12). Thus, the
above results clearly exemplified that an appropriate combina-
tion of NBS and carboxylic acid is essential both for obtaining
high yield and selectivity. In addition, the yield and selectivity
were slightly decreased when the reaction was performed
under oxygen instead of nitrogen (entry 10 vs. 13), indicating
that nitrogen atmosphere is a better option for the transforma-
tion. A screening of various carboxylic acids (entries 14–19)
showed that benzoic acid, CF3COOH, n-butyric acid, and iso-
butyric acids were also promising additives although their
overall efficacy was somewhat lowered to compare with HOAc
(entries 14–17). Notably, PTSA has previously been demonstrat-
ed as a powerful additive for the Pd-catalyzed CÀH mono-halo-
genation of anilides,[5e] but it proved to be ineffective for this
reaction in our system (entry 20). Finally, we also inspected the
effect of solvents. However, most of them such as dioxane,
DMSO, DMF were ineffective except for dichloroethylene (DCE)
and ethyl acetate (entries 21 and 22).
[a] Unless otherwise noted, the reaction conditions were: 2-aryl pyridine
1 (0.5 mmol), CuX (1.0 equiv, CuBr was used for bromination and CuCl
was used for chlorination), NXS (2.0 equiv, NBS was used for bromination
and NCS was used for chlorination), HOAc (0.5 equiv), CH3CN, 1008C, N2,
24 h (the reaction time was not optimized); isolated yield of mono-halo-
genated products. [b] iso-Butyric acid (0.5 equiv) was used as additive in-
stead of HOAc. [c] 1.5 equiv of NBS was used. [d] 1,2-Dichloroethylene
was used as solvent and benzoic acid was used as additives. [e] Benzoic
acid was used as additives. [f] m/d was the ratio of 2 f and 2 f’ (or 3 f and
3 f’) to the corresponding di-halogenated product.
(2b–2 f, and 3b–3 f). Interestingly, for the meta-substituted
substrates, e.g., the CF3 group substituted substrate produced
only the mono-halogenated 2e and 3e, respectively, in high
yields. However, the Br group modified substrate afforded the
regioisomers of 2 f and 2 f’, and 3 f and 3 f’. In addition,
a small amount of di-halogenated derivatives were also
formed. This is attributed possibly to a stronger electron-with-
drawing, as well as a relatively larger steric effect of CF3 as
compared with Br group. Finally, the reaction was somewhat
sluggish for a substrate modified by a strong electron-deficient
NO2 group (2g and 3g).
The method also exhibited good compatibility for the sub-
strates whose aryl ring was decorated by an electron-donating
group (2h–2j, and 3h–3j). In addition, a comparison study by
introducing a substituent at the 4- or 3-position of the pyridyl
ring revealed that the mono-selectivity for the 3-substituted
substrate could be increased dramatically, affording only the
mono-halogenated products (2k vs. 2l and 3k vs. 3l). A similar
substitution effect has also been observed for Pd-catalyzed hal-
ogenation.[4] This is attributed to the steric interaction between
the 3-methyl group in the pyridyl ring and the 2’-halogen
group in the aryl ring incorporated in the first halogenation,
which reduces the coplanarity between the pyridyl and phenyl
Through the extensive investigation we have established
a promising approach for the Cu-mediated mono-selective
bromination. The generality of this protocol was examined by
employing the conditions shown in entry 10 in Table 1 as gen-
eral conditions. Both the bromination and chlorination (CuCl
and NCS were used for chlorination) were inspected. The
yields and selectivity of various 2-arylpyridine derivatives were
listed in Table 2. In general, the mono-halogenated products
could be obtained in good isolated yields and high selectivity
for a range of substrates whose phenyl ring was unsubstituted
(2a and 3a), or was modified by electron-withdrawing groups
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 123 – 126 124
Du, Zhi-Jun
