Organic Letters
Letter
Scheme 1. Kinetic Resolution of rac-2a in a Multigram-Scale
Reaction and Deprotection of Recovered 2a for
Determination of Absolute Configuration
significantly (2a vs 2o, from s = 41 to 25) while maintaining
its reactivity (∼50% conv within 5 h) (Figure 4). On the other
Figure 3. Kinetic resolution of various 3-hydroxy-3-substituted 2-
oxindoles with 1 mol % of catalyst 1f.15
aryl group did not significantly affect the s-factor (2h [s = 63, 5
h] vs 2i [s = 60, 5 h]). A polar functional group within the
substrate could be tolerated. The use of 2k having a silyl ether
showed an excellent s-factor (s = 60), and 2l having a β-
hydroxy ketone (aldol adduct) showed a good s-factor (s =
30). In the latter case, we expected dynamic kinetic resolution
of 2l via retro-aldol reaction, but kinetic resolution only
proceeded (s = 30). 5-Cl- and 5-MeO-substituted substrates
2m and 2n also showed good to excellent s-factors (s = 60 and
31, respectively). According to these experiments, the current
kinetic resolution process allowed us to access various
enantioenriched 3-hydroxy-3-substituted 2-oxindoles having
an N-methoxycarbonyl group with s = 17−60 within 3−9 h.
We also conducted a multigram-scale reaction with 2a to
test the scalability of this reaction (Scheme 1). With only 19.2
mg (1 mol %) of catalyst 1f, 10.0 g of rac-2a was smoothly
resolved to give monoacylate 3a (6.7 g, 93:7 er) and recovered
2a (4.7 g, 95:5 er) with excellent selectivity (51% conv, s = 43)
at 0 °C for 5 h. The N-methoxycarbonyl group of recovered 2a
was successfully removed by Krapcho-type decarboxylation
under mild conditions (NaCl and H2O in DMSO) to give 4a,
the absolute configuration of which was determined to be R by
comparison to the reported value for the specific optical
rotation of (S)-4a.16 Other recovered alcohols 2b−n were also
assigned R by analogy.
Figure 4. Kinetic resolution of various N-protected 3-hydroxy-3-
substituted 2-oxindoles 2o−q with 1 mol % of catalyst 1f.
hand, the use of N-alkyl substrates 2p and 2q, an N-methyl or
benzyl group, required a longer reaction time (30 h) for
achieving ∼50% conv and resulted in a significant decrease in
the s-factor (s = 16 for 2p, s = 21 for 2q, respectively). These
results suggested that the electron-withdrawing group attached
to the nitrogen atom is required for acceleration of the
reaction.
One possible explanation for this effect is that the electron-
withdrawing group at the N-position might increase the acidity
of a hydroxy group, which was easily deprotonated by
carboxylate ion (−OCOi-Pr) fixed by a hydrogen-bonding
network (see Figure 6), which accelerates the rate of the
reaction. Furthermore, 3-hydroxy-3-substituted 2-benzofura-
none was also subjected to kinetic resolution [eq 1]. Thus, the
kinetic resolution of 2r showed a reasonably good s-factor (s =
28 within 9 h). Compared to N-methoxycarbonyl variant 2b,
almost the same reactivity was observed (48% conv for 9 h
with 2b in Figure 3, and 46% conv for 9 h with 2r). The
relatively strong electronegativity of oxygen atom maintains the
acidity of the hydroxyl group, which is important for
acceleration of the reaction. Finally, the importance of the
five-membered moiety was also confirmed (eq 2). The reaction
of 2s, where the five-membered ring is deconstructed, resulted
in poor conversion and a low s-factor within 26 h (5% conv, s =
2) compared to 2a having a five-membered structure. Overall,
Next, several control experiments regarding substrate
structure were carried out to obtain further insight into these
reactions. Change of the N-protecting group from methox-
ycarbonyl to tert-butoxycarbonyl decreased the s-factor
1171
Org. Lett. 2021, 23, 1169−1174