Full Paper
Table 1. Optimization of the reaction conditions.[a]
Entry[a]
C=O source (equiv)
acetone ([d]
Base (equiv)
t [h]
Yield [%][b]
1[c]
2[c]
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19[f]
20[g]
)
KOAc (2.0)
KOAc (2.0)
KOAc (2.0)
KOAc (2.0)
KOAc (2.0)
KOAc (2.0)
KOAc (2.0)
KOAc (2.0)
KOAc (2.0)
KOAc (2.0)
NaOAc (2.0)
CsOAc (2.0)
CsOPiv (2.0)
KOAc (1.5)
KOAc (1.1)
KOAc (1.1)
KOAc (1.1)
KOAc (1.1)
KOAc (1.1)
KOAc (1.1)
24
24
24
18
18
18
18
18
18
18
18
18
18
18
18
12
12
12
12
12
11
19
28
40
51
64
73 (68)
24
0
70
72
70
69
78
85
87 (83)
70
0
85
acetone (2.0)
acetone (2.0)
cyclohexanone (2.0)
cyclopentanone (2.0)
nPrCHO (2.0)
iPrCHO (2.0)
tBuCHO (2.0)
PhCHO (2.0)
CyCHO[e] (2.0)
iPrCHO (2.0)
iPrCHO (2.0)
iPrCHO (2.0)
iPrCHO (2.0)
iPrCHO (2.0)
iPrCHO (2.0)
iPrCHO (1.2)
-
Scheme 2. RhIII-Catalyzed alkyne annulation by CÀH activation.
ketones. With this and our previous one-pot isoquinoline syn-
thesis[16] in mind, we envisioned that a hydrazone could be
formed in situ from hydrazines and serve as a DGox for RhIII-cat-
alyzed CÀH activation and alkyne annulation (Scheme 2). In
this way, the hydrazone group could be an auto-formed and
auto-cleavable directing group, which we have termed as an
auto-directing group (DGauto),[17–19] leading to a one-pot synthe-
sis with a high step economy. Herein, we disclose our design
and development of a hydrazone DGauto strategy for the rapid
assembly of unprotected indoles in a practical and external-
oxidant-free way.
iPrCHO (2.0)
iPrCHO (2.0)
<5
[a] Reaction conditions: 1a (0.20 mmol), 2a (0.24 mmol), KOAc, and
[(Cp*RhCl2)2] (2.5 mol%) in 1.0 mL MeOH at 808C under N2 atmosphere.
[b] Yield of 3aa is based on GC by using n-C24H50 as internal standard;
value in parentheses is isolated yield at 0.5 mmol scale. [c] 608C.
[d] MeOH–acetone (4:1) as solvent. [e] Cy=cyclohexyl. [f] 5 mol% [Cp*Rh-
(MeCN)3][SbF6]2 was used. [g] 2.5 mol% [{Ru(p-cymene)Cl2}2] was used.
yields (Table 1, entries 11–13). Reaction in other solvents such
as MeCN, DCE, THF, and t-AmOH led low conversion and de-
clined yield.
Results and Discussion
We initiated our study employing phenylhydrazine hydrochlo-
ride (1a) and diphenylacetylene (2a) as substrates, and
[(Cp*RhCl2)2] as a precursor of the RhIII catalyst (Table 1). Ace-
tone was used as the C=O source to generate corresponding
hydrazone from 1a. Excess KOAc was used as base to neutral-
ize the hydrochloride and facilitate the generation of [Cp*Rh-
(OAc)n] species as the active catalyst.[11,16] As expected, reaction
at 608C for 24 h gave full conversion of 1a to corresponding
acetone hydrazone, but only 11% indole product 3aa was ob-
tained (Table 1, entry 1). Reaction with 2.0 equivalents of ace-
tone in MeOH at 808C gave improved yield (Table 1, entries 2
and 3).
To our delight, simply reducing the amount of KOAc to
1.1 equiv gave a satisfactory yield (87% GC, 83% isolated)
within a shorter reaction time (Table 1, entry 16). Reducing the
amount of iPrCHO to 1.2 equiv gave a reduced yield (Table 1,
entry 17) and reducing it to a catalytic amount led to much
lower conversions. Further optimization focused on the devel-
opment of a special C=O source for metal-organic cooperative
catalysis[17] is ongoing. No 3aa was detected without a C=O
source (Table 1, entry 18), and no hydrohydrazination product
of 2a was detected in either entry, indicating the cyclization
did not proceed by an alkyne hydrohydrazination followed by
a Fischer indole synthesis.[20] The cationic Cp*RhIII complex
showed similar catalytic activity with [(Cp*RhCl2)2] (Table 1,
entry 19). The structurally analogous RuII complex,[12] which is
active in some cases with NÀO containing DGox, showed little
catalytic activity in this reaction, and only a trace of 3aa was
detected by GC-MS (Table 1, entry 20).
Considering that structure of the hydrazone group may
have a great influence on the generation of the rhodacycle in-
termediate and reactivity of the NÀN bond, we then focused
on screening various C=O sources. Cyclic ketones gave im-
proved yields (Table 1, entries 4 and 5), whereas benzaldehyde
gave no desired product (Table 1, entry 9). For alkyl aldehydes
RnCH3-nCHO, those with a 28 carbon connected to CHO (n=2,
Table 1, entries 7 and 10) proved to be better C=O sources
than those with a 18 carbon (n=1, Table 1, entry 6) or a 38
carbon (n=3, Table 1, entry 8).With isobutyraldehyde (iPrCHO)
as a C=O source, other carboxylates such as NaOAc, CsOAc
and CsOPiv were tested and gave similar or slightly lower
With the optimized reaction conditions in hand, the general-
ity of the one-pot synthesis of indoles was examined. By em-
ploying 2a as the alkyne partner, various substituted aryl hy-
drazines were surveyed (Table 2). o-Methylphenylhydrazine
gave an improved yield (3ba, 90%) and m-methylphenylhydra-
zine afforded 3ca as a single regioisomer. Aryl hydrazines with
an electron-donating group (EDG) as para-substituent showed
Chem. Eur. J. 2014, 20, 2352 – 2356
2353
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim