.
Angewandte
Communications
A base elimination strategy would allow for direct access
from numerous commercially available, substituted anthra-
nilic acid derivatives in 2–3 steps (see the Supporting
Information for details). Ultimately, the success of our
approach would rely on the concomitant generation of the
Ao-QM and the NHC catalyst using a single Brønsted base
along with balancing the concentrations of each reactive
intermediate during the course of the reaction (i.e., 1b and
1c).[19] Additionally, it was undetermined if the Lewis-basic
NHC catalyst would directly trap a reactive electrophile 1c.
Initial experiments with benzyl chloride 2 and benzaldehyde
in the presence of precatalysts A or B with DBU (1,8-
diazabicyclo[5.4.0]undec-7-ene) yielded the desired ketone 3
in moderate to low yield (Table 1, entries 1 and 2). Under
with benzoin D furnished the ketone in a similar yield,
providing evidence for the reversibility of the benzoin
condensation under these conditions (entry 5). Importantly,
the ability to cycle this undesired Umpolung product back
into the desired reaction allows for less aldehyde substrate
(i.e., 1.2 equiv) than typical intermolecular acyl anion reac-
tions.[22,23] Further improvement in yield was achieved by
reducing the equivalents of benzaldehyde and cesium car-
bonate (entries 6 and 7).
With optimized conditions for the NHC/Ao-QM synthesis
of ketone 3 in hand, the in situ dehydration for the prepara-
tion of 2-aryl-indole 4 was explored. The exposure of the
isolated ketone 3 to TFA in dichloromethane delivered the
desired N-Boc indole in nearly quantitative yield (5 min,
96%). Unfortunately, the direct addition of TFA or other
organic acids to the reaction mixture did not produce the
desired results. Based on the results with TFA, it was clear
that the Lewis basicity of the solvent played a key role in the
dehydration step. To streamline the indole synthesis to
a single flask operation, many ethereal solvents and organic
acids were evaluated. We found that 1,4-dioxane mediated
the NHC-catalyzed ketone synthesis as well as the acid
promoted dehydration. The addition of methanesulfonic acid
promoted the in situ dehydration of ketone 3 to the desired
indole 4 (entry 8). Attempts to further optimize the con-
ditions by reducing the precatalyst loading or by increasing
the reaction temperature were unsuccessful and rather
reduced the overall yield (entries 9 and 10). The use of
other bases, such as potassium carbonate, showed minimal
conversion (entry 11).
Table 1: Optimization of the reaction conditions.
Entry NHC Base (equiv) PhCHO Solvent
(equiv)
Yield
Yield
3 [%][a] 4 [%][a]
1
2
A
B
B
C
C
C
C
C
C
C
C
DBU (1.5)
DBU (1.5)
1.5
1.5
THF
THF
CHCl3
THF
THF
THF
THF
19[b]
35[b]
60
82
80
86
88
–
–
–
–
–
–
–
–
82
56
61
–
3[c]
4
Cs2CO3 (2.5) 10
Cs2CO3 (2.5) 1.5
Cs2CO3 (2.5) 1.0[d]
Cs2CO3 (2.5) 1.2
Cs2CO3 (1.2) 1.2
Cs2CO3 (1.2) 1.2
Cs2CO3 (1.2) 1.2
Cs2CO3 (1.2) 1.2
5
6
7
8[e]
9[e,f]
10[e,g]
11
1,4-dioxane
1,4-dioxane
1,4-dioxane
–
–
With efficient single-flask conditions identified for the
synthesis of 2-aryl-indoles, we surveyed the scope of this
transformation (Table 2). Aryl aldehydes with electron-with-
drawing or -donating groups in the para-position were
tolerated, giving rise to the corresponding indoles in good
to excellent yield (4a–4g). Furthermore, meta-substituted
aryl aldehydes with electron-withdrawing and -donating
groups were also accommodated under the reaction condi-
tions, affording indoles 4h–4k in high yield.
K2CO3 (3.0)
1.2
1,4-dioxane 31[h]
[a] Yield of the isolated product. [b] Determined by 1H NMR spectros-
copy (500 MHz) with 1,3,5-trimethoxybenzene as internal standard.
[c] 30 mol% azolium. [d] Benzoin D used instead of benzaldehyde.
[e] After 36 h, 6.5 equiv MsOH was added. [f] 10 mol% azolium. [g] The
reaction was conducted at 508C. [h] Conversion after 36 h.
Indoles 4c and 4i were isolated in higher yield when the
reaction was performed at an elevated temperature (508C).
The synthesis of indole 4i is noteworthy, because it has never
been prepared to date (N-Boc or N-H) and transition-metal-
catalyzed strategies should prove difficult due to competing
insertion reactions.[24] The yield of indole 4h was significantly
higher when the amount of aryl aldehyde (2.2 equiv) was
increased. At this time, reactions with ortho-substituted aryl
aldehydes usually lead to the recovery of unreacted starting
materials or low conversion to the intermediate ketone.[22]
Either the production of the nucleophilic Breslow intermedi-
ate (1b) is slow and/or disfavored, because of destabilizing
interactions, or, once formed, the engendered strain promotes
a conformational change, placing the aryl ring orthogonal to
the enol thiazolium system, thus sterically encumbering the
nucleophilic acyl anion carbon [Eq. (1)]. Investigation of
heteroaryl aldehydes produced indoles with 2-naphthyl (4l),
pyridyl (4m), and thiophenyl (4n) substituents at C-2.
Aliphatic aldehydes (4o) showed only minimal activity with
azolium B and no activity with C.
these reaction conditions, no other triazolium or imidazolium
NHC precursor gave appreciable yield (> 10%) of the
desired ketone 3. Thorough investigations revealed the
formation of unproductive adducts between nucleophilic
bases (e.g., DBU, Et3N) and chloride 2. In turn, an exhaustive
screen of azolium precatalysts and solvents with non-nucle-
ophilic cesium carbonate as base was undertaken. These
experiments showed that 30 mol% of azolium B in chloro-
form afforded the highest yield of ketone 3, however, 10 equiv
of aldehyde was required (entry 3).
The promising yields obtained with azoliums A and B
prompted us to explore several new N-aryl thiazoliums
prepared with the 2,6-diethylphenyl moiety, which has been
shown to increase catalyst performance.[20,21] Gratifyingly,
thiazolium C in THF afforded the isolated ketone 3 in 82%
yield (entry 4). Notably, the replacement of benzaldehyde
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 9603 –9607