Communication
methacrylate 5b was used. We also investigated other
electron-withdrawing alkenes such as acrylonitrile, cinna-
mate ester, maleate ester, fumarate ester, and b-nitrostyr-
ene. Unfortunately, all above examples were unsuitable,
especially the 1,2-disubstituted ones. Hence this photoca-
talytic reaction was very sensitive to the steric environ-
ment at the aryl radical addition site.
Table 3. Scope of aryl hydrazine derivatives.[a,b]
The structure of the hydroperoxyarylated product was
unambiguously determined by X-ray crystallography (see
the Supporting Information). We also used one-dimension-
al selective NOESY NMR experiments to confirm the regio-
selectivity of hydroperoxyarylated product (see the Sup-
porting Information). Last but not least, hydroperoxide 4o
was reduced to the corresponding alcohol in 97% yield by
using 10% Na2S2O3 solution and PPh3 (1 equiv) as reduc-
tants [Supporting Information, Eq. (S1)].
Phenyldiazene was proposed to be a key intermediate
in the reaction.[7a] Recently, Heinrich and co-workers
showed the trapping of phenyldiazenes in cycloaddition
reactions.[16] To understand the reaction mechanism, sever-
al experiments were performed. The hydroperoxyarylation
of olefins was expected to proceed through a radical reac-
tion. When phenylhydrazine 2a was subjected to 2,2,6,6-
tetramethylpiperidinoxyl (TEMPO), compound S2 was ob-
tained [Eq. (S2)], which was similar to previous observation
with aryl diazonium salts.[4a] When TEMPO was added to the re-
action under standard conditions, the TEMPO-trapped inter-
mediate S3 along with the hydroperoxyarylated product were
obtained [Eq. (S3)]. This experimental evidence pointed to-
wards a radical reaction pathway.
[a] Unless otherwise noted, the reaction conditions were as followed: Aryl hy-
drazine 2 (0.10 mmol), photocatalyst 1e (1 mol%), olefin 3l (1 mmol), MeCN
(1.0 mL), 23 W CFL, 12 h. [b] Yield of isolated product. See the Supporting Infor-
mation for details.
such as 4-vinylpyridine or 2-vinylthiophene failed to react. In
addition, the reactivities of other electronically unbiased olefins
such as b-pinene and 1,2-disubstituted styrene derivatives,
turned out to be low.
Next, several aryl hydrazine derivatives 2a–i reacted with
olefin 3l and the results are summarized (Table 3). Generally,
the aryl hydrazines containing electron-donating and electron-
withdrawing groups (4m–p) at the para-position gave lower
yields than phenylhydrazine (4l). At the meta-position, the
electron-donating substituent appeared to perform better than
its electron-withdrawing counterpart (4q, r). This was also the
similar case for the ortho-substituents (4s, t). All the aryl hydra-
zines used were in the base form. We attempted to use its hy-
drochloride salt form with K2CO3 but our efforts proved to be
futile. The aryl hydrazines containing electron-withdrawing
substituents were stable but the electron-donating substitu-
ents were freshly basified just before the reaction was carried
out. We also experimented with alkyl hydrazines containing
tert-butyl or 2-hydroxyethyl groups but these were unsuccess-
ful. Scaling-up the model reaction (Table 1) to 0.3 mmol result-
ed in 55% yield. Further scaling-up led to diminishing yields.
We extended our method to methyl acrylate derivatives 5,
which contained an electron-withdrawing group (Scheme 1).
We were pleased to find that the a-perhydroxylated ester
product 6a was formed in 64% yield when methyl acrylate 5a
was used. The yield even increased to 80% when methyl
Hydrazine was oxidized into diimide during the transfer hy-
drogenation of olefins.[12a] We conducted the hydroperoxyaryla-
tion reaction in CD3CN and did not observe any reduction of
alkene to the saturated alkane using 1H NMR spectroscopy
[Eq. (S4)]. We conducted fluorescence quenching experiments
with catalyst 1e and hydrazine 2a but no quenching effect
was observed. As this reaction also did not proceed under
dark conditions (Table 1, entry 12), catalyst 1e was likely to be
a singlet oxygen photosensitizer.[17] When aqueous H2O2 was
added, the hydroperoxyarylation reaction proceeded sluggishly
with 28% yield in the absence of catalyst and visible light at
358C [Eq. (S5)]. Then the same reaction was repeated but with
added catalyst 1e. This time, the yield went up to 64%
[Eq. (S6)]. It was a well-established fact that diazenes decom-
posed under acidic or basic conditions (e.g. in the Wolff–Kish-
ner reaction).[18] Therefore, basic catalyst 1e might possibly be
catalyzing the decomposition of aryldiazene 7.
Based on the above-mentioned observations and precedent
literature, we outlined a plausible mechanism (Scheme 2). Ini-
tially photocatalyst 1e acted as a photosensitizer and generat-
ed singlet oxygen.[17] Then the resultant singlet oxygen oxi-
dized aryl hydrazine 2 to aryl diazene 7.[19] Next diazene 7 un-
derwent further oxidation by dioxygen to form radical 8 and
hydroperoxide radical. Then radical 8 decomposed thermally
to give aryl radicals. Subsequently the aryl radicals were
trapped by olefins to give radical intermediate 9. Then radical
9 would be terminated by the hydroperoxide radical, which
Scheme 1. Hydroperoxyarylation of acrylate derivatives.
Chem. Asian J. 2015, 10, 1618 – 1621
1620
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim