Crystal Growth & Design
Article
finally obtained by filtration in a 75% yield with a 99% ee
(Table 3, entry 3, and Figure 5). In the reaction using C2 and
DBU, crystals were filtered and washed with a small amount of
cold ethanol to give a 50% yield of 1 with a 94% ee (Table 3,
entry 4, and Figure 5).
Scheme 3. One-Pot Asymmetric Synthesis of 1 By
Combining Benzoin Condensation Using Achiral NHC
Catalysts with Dynamic Crystallization
In the reaction using catalyst C3, it was also possible to carry
out the reaction simultaneously with the asymmetric
amplification. The use of NaOH as a base gave an 88% yield
of 1 in a 99% ee (Table 3, entry 5, and Figure 5). When using
DBU as a base, we could not obtain the crystalline 1 (Table 3,
entry 6).
As described above, the formation of racemic 1 by the
benzoin condensation of prochiral p-anisaldehyde was followed
by deracemization via Viedma ripening. It is plausible that
racemization is promoted through the enolate ion such as in
the case of the deracemization of anisoin shown in Table 2. On
the other hand, benzoin condensation is a well-established
reversible reaction, and it has also been suggested that
racemization possibly occurs through the reverse reaction.
Therefore, this reaction system is suitable for dynamic
crystallization using a conglomerate crystal.
It was possible to isolate a 99% ee of 1 by simple filtration
from the reaction mixture without using an external
asymmetric source. These crystals could be easily purified
into 100% ee by a recrystallization from chloroform/hexane. It
was possible to isolate 100% ee p-anisoin from prochiral p-
anisaldehyde simply by filtering the crystals twice.
Next, the control of the handedness of the chirality was
examined. In dynamic optical resolution by spontaneous
crystallization, it was impossible to predict which enantiomer
crystal would converge (Figure S1). In this case, it was possible
to control the handedness by starting from slightly enriched
crystals or adding enantiomer crystals as a seed; however, we
examined another approach to control the handedness of the
deracemization using optically active amino acids.
A catalytic amount of L- or D-valine crystals was added
during the Viedma ripening of racemic 1. In a sealed glass tube,
1 (272 mg, 1.0 mmol), EtOH (0.5 mL), DBU (15 mg, 0.10
mmol), glass beads, and L- or D-valine (12 mg, 0.1 mmol) were
suspended with stirring at 600 rpm using a cross-shaped
stirring bar at 60 °C. Deracemization began immediately and
reached an enantiomeric excess of 99% in just five days (Table
4, entries 1 and 2, and Figure S2). The configuration of the
obtained 1 matched that of the added valine. When L-valine
was used, (R)-1 was obtained, while the addition of D-valine
caused the handedness of the deracemization to converge to
(S)-1. The experiments were repeated five times each.
The same asymmetric control was observed when valine
coexisted in the benzoin condensation from p-anisaldehyde.
Deracemization proceeded efficiently and faster than that
without valine. (R)-Anisoin 1 was obtained in 68−70% yield
with a >99% ee by adding L-valine (Table 4, entry 3), while the
reaction in the presence of D-valine gave (S)-1 in all five
experiments (Table 4, entry 4).
of cold ethanol to give a 70% yield of p-anisoin with a 93% ee
(Table 3, entry 3, and Figure 5). When using DBU instead of
NaOH, the chemical yield of 1 was low, and solids were not
obtained for suspension (Table 3, entry 2).
Table 3. Asymmetric Synthesis of p-Anisoin from p-
Anisaldehyde Involving Benzoin Condensation and Viedma
a
Ripening
b
c
entry
catalyst
solvent
base
yield (%)
70
n.a.
ee (%)
93
d
1
C1
C1
C2
C2
C3
C3
EtOH/H2O
EtOH/H2O
EtOH
EtOH
EtOH
NaOH
DBU
NaOH
DBU
NaOH
DBU
e
2
3
4
5
6
75
50
88
99
94
99
e
EtOH
n.a.
a
All experiments were performed in the presence of p-anisaldehyde
(2.0 mmol), 0.5 mL of solvent, catalyst (10 mol %), and base (15 mol
b
%). The solution was suspended with stirring. Yield of crystallized 1
c
obtained by filtration. One of the two enantiomers predominantly
d
amplified was random. After the reaction at rt for 24 h, the solvent
was removed under reduced pressure. Next, EtOH (0.5 mL) was
e
added and suspended with stirring at 60 °C. Crystalline 1 was not
obtained.
Figure 5. Time course of the asymmetric synthesis of 1 involving
deracemization by Viedma ripening. Entry numbers correspond to
those in Table 3.
To clarify this phenomenon, we first investigated the crystal
growth of 1 in the presence of optically active valine. However,
a clear difference was not observed in the growth rate between
(R)-1 and (S)-1 in the presence of L-valine, and both
enantiomer crystals were grown.34 Additionally, to investigate
the possibility of forming mixed crystals with optically active
valine, the crystals obtained from recrystallization in the
presence of valine were analyzed by NMR spectroscopy. It was
found that valine was not included in the crystals of 1.35
In the case of the reaction using catalyst C2, we could use
EtOH as a solvent without water, and it was possible to carry
out the reaction and the asymmetric amplification at the same
time. In a sealed glass tube, p-anisaldehyde (272 mg, 2.0
mmol), EtOH (0.5 mL), C2 (0.2 mmol), powdered NaOH
(12 mg, 3.0 mmol), and glass beads were reacted At 60 °C.
Crystals of 1 immediately precipitated and were suspended
with stirring at the same temperature. The changing ee value of
the crystals was analyzed using the above method, and 1 was
2426
Cryst. Growth Des. 2021, 21, 2423−2428