Organic Letters
Letter
pH, including a particularly challenging formation of
potentially antiviral gel materials.
To obtain more effective catalysts, we investigated the
condensation between 4-nitrobenzaldehyde and benzoylhy-
drazine as a model reaction (Scheme 1). All reactions were
and 19.10 times rate enhancement over aniline (Scheme 1,
entry 9 vs entry 1). We first tested our secondary amine
hypothesis with N-methylaniline. Unfortunately, N-methylani-
line and its derivatives bearing either para-methoxy or ortho-
carboxylic acid displayed negligible catalytic effects. Other
cyclic aromatic secondary amines were examined. To our
delight, we found that indoline (pKa = 5.6)14 showed moderate
catalytic activity for the model reaction, with a 5.0-fold rate
enhancement (entry 10) over that of aniline. In contrast,
1,2,3,4-tetrahydro-quinoline(pKa = 4.7)14 showed poor cata-
lytic activity (not shown). Pyrrolidine (pKa = 11.2) was also
inactive under the present conditions, which may suggest that
the benzene ring decreases the basicity of nitrogen in indoline
and stabilizes the iminium intermediate under aqueous
solution, which pinpoints the importance of aromatic amino-
catalysis.15 All these results confirm the concept that the pKa of
catalysts closest to the neutral buffer solution provide more
rate enhancements for acylhydrazone/hydrazine formatio-
n.3a,10a,16
a
Scheme 1. Screening of Catalysts
Finding the indoline scaffold prompted us to improve its
activity, and indolines with different substitutions were
investigated next. Most substituted indolines are commercially
available or easy to prepare in high yields (for details, see
indoline derivatives with carboxylic acid groups in the 2- and
7-positions were first explored. Unexpectedly, indoline-2-
carboxylic acid (entry 11) shows the kapp was 0.057 M−1 s−1,
only 1.90 times faster than aniline, while indoline-7-carboxylic
acid was virtually inactive. Compared with indoline, 2-
methylindoline (entry 12) showed a substantial reduction in
catalytic efficiency, likely a result of a steric effect on iminium
formation. However, 7-hydroxyindoline (entry 13) showed
slightly higher activity. 7-aza-Indoline endowed with a fused
pyridinyl ring had no catalytic activity at all.
Next the substitution effects on the benzene ring of indoline
were examined (entries 14−18). Indolines with electron-
donating groups showed enhanced reaction rates. For example,
5-methylindoline (entry 15) showed a 11-fold rate enhance-
ment, and 5-methoxylindoline (entry 14) showed enhance-
ment up to 15.50-fold, comparable to 5-methyl-2-amino-
benzene phosphonic acid (entry 9). With increasing electron-
withdrawing ability, the reaction rate dramatically decreased
(entries 16−18) and 5-nitroindoline (not shown) showed no
activity. A good linear relationship was obtained when the
logarithm of rate constants log kx/kH was plotted against
Hammett substitute constant σ (Figure 2, black line). A large
negative slope (ρ = −1.804) reveals the electron-deficient
nature of the key reaction species in the rate-limiting step,
a
Apparent second-order rate constants, kapp (M−1 s−1), are given as
mean values standard deviations based on triplicate measurements
or more. Conditions: 123 mM NaCl, 2.4 mM KCl, and 10.7 mM
sodium and potassium phosphate (pH 7.4) with 10% DMF as
cosolvent.
performed under pseudo-first-order conditions in phosphate
buffered saline (PBS) at pH 7.4 with 10% DMF as cosolvent.
Apparent second-order rate constant kapp (M−1 s−1) values
were obtained by monitoring UV−vis absorption at 329 nm of
the hydrazone product. To calculate the rate constants we used
nonlinear regression by linear least-squares fits, following the
tion).6b,13 The observed rate increases linearly for both 4-
nitrobenzaldehyde and catalyst under excess concentration,
indicating the reaction is first order for each substrates and
acylhydrazone formation is more challenging than hydrazone
formation under neutral aqueous conditions. Without catalysis,
the model reaction is too slow to reach equilibrium in pH = 7.4
within 24 h.
Selected screening results are shown in Scheme 1. Aniline
(pKa = 4.6 in H2O)3a,7a was selected as the reference (entry 1).
A range of primary amine catalysts including the well-explored
aniline analogue catalysts were also investigated for compar-
ison.9,10 Among all the primary amine catalysts examined
(Scheme 1, entries 1−9), 5-methyl-2-aminobenzene phos-
phonic acid, previously reported by Kool,10a showed good
catalytic effects, with a 0.573 M−1 s−1 apparent rate constant
Figure 2. Hammett plot of substitution effects of indoline (black) and
benzaldehydes (red).
B
Org. Lett. XXXX, XXX, XXX−XXX