Organic Letters
Letter
or at the ortho position when the para position is blocked (3m,
48%); interestingly, 3l was obtained in comparable yield under
Brønsted acid catalysis (71%). An N-monosubstituted aniline
also reacted at the para position to give the double substitution
product 4n, but reacted through the nitrogen atom when the
para position was masked (single addition, 3o). Finally, the
parent aniline failed to produce any desired product, with only
yndiamide decomposition observed.
Related methoxylated arenes displayed similar behavior:
while phenol and anisole proved unsuccessful, the more
electron-rich 2,6-dimethoxyphenol gave 3p in good yield
(72%), albeit requiring an extended reaction time. Heightened
reactivity was observed using 1,3,5-trimethoxybenzene, which
delivered 3q in excellent yield (91%). Two non-aryl carbon
nucleophiles were also tested; while a silyl enol ether produced
the desired product 3r (in low yield), decomposition was
observed using allyltrimethylsilane (3s). To our surprise, a
significant increase in the yield of 3r was observed using 10
mol % HNTf2 as catalyst (53%). The observation of double
substitution under Brønsted acid catalyzed conditions (Figure
1, 4b) offers potential for further functionalization of the
monosubstituted aminoamide products. Indeed, a rapid and
quantitative acid-catalyzed substitution of the α-sulfonamide
group in 3a using indole as nucleophile was achieved with just
5 mol % HNTf2 promoter, giving bis-indole acetamide 4a in
quantitative yield (Scheme 2a). This transformation was also
successful using pyrrole as the nucleophile, giving the bis-
heteroarylated product 4c in 83% yield. The two nitrogen
atoms in yndiamide 3l could also be further differentiated by
selective detosylation12 (Mg/NH4Cl, Scheme 2b), affording a
high yield of acetamide 5l from aminoamide 3l (84%). A
robustness test13 was also performed to assess functional group
tolerance (Scheme 2c); pleasingly, the reaction tolerated 10
out of 12 external additives covering a range of functional
groups, the exceptions being a thiol (73%) and 1,3-diol (46%).
The latter induced the formation of unidentified side products,
likely due to competing nucleophilic attack by the diol.
As noted, ynamide functionalizations typically benefit from
high regioselectivity due to the inherent polarization of the
alkyne. Achieving regiocontrol in yndiamide functionalization
was likely to be more challenging but could arise through
tuning of the two electron-withdrawing groups to render one
end of the alkyne more nucleophilic. As such, the unsym-
metrical yndiamides 2b and 2c (Scheme 3a) were subjected to
Scheme 3. Regioselectivity in Gold-Catalyzed Oxidative
a
Functionalization of Unsymmetrical Yndiamides
Scheme 2. Further Transformations of α-Aminoamides, and
a
Robustness Test
a
Reactions conducted with 0.1 mmol of yndiamide 2b−e. Standard
conditions: 2-chloropyridine N-oxide (2.0 equiv), indole (3.0 equiv),
IPrAuNTf2 (5 mol %), rt, 2 h. Yields in a refer to combined isolated
yield of both isomers. Yields in b refer to isolated yield of the major
isomer, apart from 3x/3x′ which were not separable. Regioisomer
1
ratios (r.r.) were determined by H NMR spectroscopic analysis of
of assignment of regioisomers.
the oxidative functionalization using indole. However, only
moderate regioselectivity was achieved on replacing Ts with
either a more (p-F3CC6H4SO2, 3t/3t′, r.r. = 3.7:1) or less
(phosphoryl, 3u/3u′, r.r. = 1:4.3) electron-withdrawing group.
We hypothesized that the use of other pyridine N-oxides might
affect this ratio based on altered nucleophilicity. The use of
pyridine N-oxide itself led to a significant reduction in yield,
while 4-fluoropyridine N-oxide maintained similar reaction
efficiency; however, neither improved the regioselectivity of
the reaction.
An alternative solution to this problem is steric differ-
entiation of the two nitrogen atoms (Scheme 3b). Accordingly,
yndiamide 2d, featuring cyclohexyl and benzyl substituents,
delivered a single regioisomer of product 3v. The extent of this
a
Reactions in Scheme 2c were performed using 0.05 mol of 2a in 0.1
mL of anhydrous DCE; NMR yields of 3a are stated, which were
determined by quantitative H NMR experiments with dimethylsul-
fone as the internal standard.
1
4890
Org. Lett. 2021, 23, 4888−4892