À
Pd-Catalyzed Multiple C H Functionalization
COMMUNICATION
ing group appear compatible with the present catalytic reac-
tions. Thus, the reaction of 4-chloro-, 3-chloro-, 2-chloro-, 3-
bromo-, and 4-trifluoromethyl-benzaldehyde oxime ethers
1e–j with 2a afforded the corresponding fluorenones 3e–3j
in 82–74% yields (Table 1, entries 4, 10, 11, 12, and 6). 4-
Phenyl and 4-nitro benzaldehyde oximes 1k–l were also effi-
ciently arylated under the reaction conditions (Table 1, en-
tries 7 and 8) to give the expected products 3k–l in good
yields. The reaction of 1-naphthyl and 2-naphthyl oxime
ethers with 2a also proceeded smoothly to give benzoan-
thracenone and benzofluorenone, in moderate yields
via PdIV.[13] In the second cycle, 7a is converted to fluore-
À
none oxime ether 3A catalyzed by the Pd complex with C
H bond activation and an insertion reaction as the key steps.
During the reaction, the Pd0 that is generated is reoxidized
to the active PdII by potassium persulfate.
The proposed mechanism is strongly supported by the iso-
lation of a 5-membered palladacycle 4 (Scheme 2) and its re-
action with arenes under various conditions. Complex 4 was
prepared by the cyclometallation of 1a with PdACHTUNGTRENNUNG(OAc)2
(1 equiv) in TFA at 508C in 67% yield. The structure of 4
was confirmed by the single-crystal X-ray crystallographic
analysis. Notably, the Pd complex is dimeric in nature with
trifluoroacetate ligands bridging the two palladium metals.
À
(Table 1, entries 13 and 14). There are two possible C H
bond functionalization sites at C2 and C8 for substrate 1m,
but the C8-functionalized product 3m was observed exclu-
sively.
In addition, there is a weak interaction between the two pal-
[11f,g]
À
ladium nuclei with a Pd Pd distance of 2.882 ꢁ.
Next, we investigated the arylation of benzaldehyde
oxime ether 1a with various arenes. Benzaldehyde oxime
ether 1a reacted nicely with toluene and anisole as the cou-
pling partner to afford the regioselective fluoren-9-ones 3o
and 3p in good yields (Table 1, entries 15 and 16). The
transformation is highly site-selective at the 4-position and
then at 3-position of arenes. No other regioisomeric product
of 3o and 3p was observed indicating that the arylation step
does not occur at C2- or C3-position of the substituted
arene. The reaction of fluorobenzene 2d having an electron-
withdrawing fluoro group with 1a was less efficient and
gave fluorenone 3q in a lower yield (Table 1, entry 17). Fi-
nally, the coupling of aldoxime ether 1e having an electron-
withdrawing 4-chloro group with 2b and 2c afforded the
functionalized fluorenones 3r–3v in excellent yields
(Table 1, entries 18–22).
When complex 4 was treated with benzene and K2S2O8
(2 equiv) in TFA at 1208C for 15 h, the cyclization product
3a was produced in 83% yield (Scheme 3). If the same reac-
Scheme 3.
tion was carried out in the absence of K2S2O8 and TFA, no
desired product was observed. Similarly, no product 3A was
observed when 4 was treated with benzene and K2S2O8 in
the absence of TFA. These results show that both K2S2O8
and TFA are required for the transformations of complex 4
to product 3A via intermediate 6a. In the above reaction
(Scheme 3) or in the catalytic reactions (Table 1), no inter-
mediate product 6 was found in the product mixture. How-
ever, if the catalytic reaction of 1a with 2a was allowed for
only 25 min, product 6a was isolated in 7%. Whereas, with
a longer reaction time, this intermediate (6a) disappeared
completely. These observations, coupled with the fact that 4
can be easily generated at low temperature, suggest that the
arylation step, that is, the reaction of 4 with arene, is likely
to be the rate limiting step in the present catalytic reaction.
To further understand the nature of the catalytic reaction,
we next examined competition experiments using 4 as the
substrate. Treatment of 4 with 1:1 mixture of benzene and
anisole afforded products 3 A and 3P in a 1:2 ratio, indicat-
ing that the arylation is faster with electron-rich arene
(Scheme 4). This is in agreement with the results shown in
Table 1 that electron-rich arenes gave high yields of product
3 (Table 1, entries 15–17). Indeed, a catalytic competition
experiment of 1a with benzene and chlorobenzene
(Scheme 5) also supports the relative trends.
A plausible mechanism for the catalytic reaction of aldox-
ime ether 1a with arene 2a is proposed as shown in
Scheme 2, which is based on previously reported chelation-
À
assisted cross-couplings through C H bond functionaliza-
tion.[8,13] The catalytic reaction likely consists of two catalytic
cycles. The first one involves the formation of palladacycle
4, the oxidation of 4 by persulfate to give a PdIV intermedi-
ate 5 and arylation of 5 by benzene to afford 6. Then, reduc-
tive elimination leads to ortho-arylated product 7a and PdII.
It is known that acetoxylation and arylation of palladacycle
in the presence of strong oxidant were proposed to proceed
In conclusion, we have developed a useful and convenient
method for the synthesis of fluoren-9-one derivatives from
Scheme 2. Proposed mechanism (For clarity, some ligands on palladium
intermediates are omitted). X=OCOCF3.
Chem. Eur. J. 2011, 17, 14723 – 14726
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
14725