Organic Letters
Letter
ethane solvent, we got a poor conversion, resulting mainly with
the formation of the anthranil 3a in small amounts.
Gratifyingly, when we switched the solvent to toluene, the
uncharacterized product 5aa that was encountered with AuBr3
was obtained in 54% yield along with the anthranil 3a (12%).
Gratifyingly, the spectral data and single crystal X-ray crystal
structure analysis of compound 5aa revealed that it was the
desired indazole and, importantly, it proved the point that the
intermolecular trapping of gold carbene that results during the
nitroalkyne cycloisomerization was possible. With these results,
we next proceeded further to improve the yield of the reaction.
As shown in Table 1, our initial attempts with changing the
solvent were not encouraging (entries 5−8). This prompted us
to explore the other gold(III) and cationic gold(I)-complexes
in this pursuit. As shown, in Table 1 (entry 9), with
dichloro(2-pyridinecarboxylato)gold(III) (PicAuCl2, entry 9)
the requisite indazole was obtained as a major product in
moderate yields. Similarly, with AuCl and with other cationic
gold(I)-complexes, the results are not encouraging (entries
10−13). In the majority of the cases, the cycloisomerization
leading to anthranil 3a was the major event, along with the
formation of the requisite 5aa. Gratifyingly, with AuCl3, in
toluene, when the reaction was heated to 80 °C, the yield of
5aa was improved to 57% and the anthranil 3a was also
obtained in 10% yield (entry 14). Finally, when the reaction
was carried out by slow addition of 1a to a solution of 2a and
AuCl3 catalyst at rt, the yield of 5aa was improved to 69%, and
the formation of the cycloisomerization product 3a was also
minimized (entry 15).
Scheme 1. Nitroalkyne Cycloisomerization and Associated
Nitroso Stabilized α-Oxo Metal Carbene Intermediate and
Possible Interruption with an External N-Based Nucleophile
Leading to Indazoles
To examine these hypothetical possibilities, initially, employ-
ing 1-(but-1-yn-1-yl)-2-nitrobenzene (1a) as a substrate, we
examined its cycloisomerization in the presence of anthranil 4a
following reported conditions using AuBr3 (in toluene) and
Pd(CH3CN)2Cl2 (in acetonitrile) complexes. As shown in
Table 1 (entries 1 and 2), with the palladium complex,
a
Table 1. Optimization of the Reaction Conditions
As illustrated in Scheme 2, a wide range of substituted
nitroalkynes have been employed to examine the scope of the
reaction and to understand how the substituents influence the
outcome. Initially, we employed the substrates having different
pendant substituents on the alkyne unit. As shown, changing
the length of the side chain did not alter the reaction outcome,
and it was also found that the protecting groups such as O-
acetyl and O-TBS placed on this alkyl chain were intact.
Interestingly, with the cyclopropyl substituted alkyne, the
intramolecular cyclization was successfully completed and
resulted in the requisite indazole 5ha as a minor product and
the corresponding anthranil 4h as the major product. For the
substrates 5ia−5la having different substituents placed para to
the alkyne unit, the outcome of the reaction was influenced.
For example, when a carboxylate as well as cyano was present,
the yield was 72% (5ka) and 77% (5la) revealing that the
presence of an electron-withdrawing group (EWG) at this
position had a stabilizing effect on the intermediate gold
carbene. In contrast, when the substituent was fluorine, the
yield was reduced to 47% (5ia), which indicates that σ-
acceptors at this position are not compatible. When a methyl
group 5ja was present at this position, the yield was only 56%.
It is noteworthy that the good yield was retained even when
the reaction was carried out on 500 mg scale (5aa, 65% yield).
Next, we varied the substituents on the anthranil ring. When
halogen-substituted anthranils were employed as the sub-
strates, the reactions in general resulted in the corresponding
indazoles 5ab−5ad in low to moderate yields depending upon
the position of the halogen atom. The yields are poor when the
halogen is placed para to the nitrogen. On the other hand, with
5-methyl and 5-carboxymethyl anthranils 4e and 4f respec-
tively, the corresponding indazoles were obtained in good to
moderate yields. Interestingly, when a methoxy group is placed
on the anthranil ring, there was no interception of the original
Yield (%)
b
b
Entry
Catalyst
Solvent
PhMe
MeCN
ClCH2CH2Cl
PhMe
MeCN
1,4-Dioxane
PhCF3
PhCl
PhMe
PhMe
PhMe
PhMe
5aa
2a or 3a
1
2
3
4
5
6
7
8
AuBr3
trace
−
64 (3a)
69 (2a)
trace (3a)
12 (3a)
−
Pd(CH3CN)2Cl2
AuCl3
−
54
−
−
44
49
42
35
26
20
27
57
69
AuCl3
AuCl3
AuCl3
AuCl3
−
13 (3a)
10 (3a)
10 (3a)
24 (3a)
38 (3a)
46 (3a)
42 (3a)
10 (3a)
trace (3a)
AuCl3
9
PicAuCl2
AuCl
PPh3AuCl/AgSbF6
IPrAuCl/AgOTf
(ArO)3PAuCl/AgNTf2
AuCl3
10
11
12
13
14
15
PhMe
PhMe
PhMe
c
d
AuCl3
a
In general, the reactions were carried out with 0.2 mmol of 1a and
0.22 mmol of 4a in 2 mL of solvent and 5 mol % of catalyst at rt with
a reaction time of 3−4 h. Isolated yield. Reaction was carried at 80
°C. Slow addition of 1a through syringe pump at rt for 4 h.
b
c
d
isatogen 2a was obtained exclusively.14 On the other hand,
with the AuBr3,7 the corresponding anthranil 3a was mainly
obtained with trace amounts of a new product, with the
expected mass corresponding to the desired indazole 5aa.
Next, we examined the reactions with AuCl3, which had
shown superior reactivity over AuBr3 in nitroalkyne cyclo-
isomerizations.15 However, in the prescribed 1,2-dichloro-
2633
Org. Lett. 2021, 23, 2632−2637