Enantiomer separation of rac-2,2′-dihydroxy-1,1′-binaphthyl (BNO) by inclusion complexation with racemic or achiral ammonium salts and a novel transformation of a 1:1:1 racemic complex of BNO, Me4N +·Cl- and MeOH into a conglomerate complex in the solid state

Article abstract of DOI:10.1016/j.tet.2004.05.122

The complete simultaneous and mutual enantiomer resolution of 2,2′-dihydroxy-1,1′-binaphthyl (BNO) and N-(3-chloro-2- hydroxypropyl)-N,N,N-trimethylammonium chloride, Me₃N ⁺CH₂CH(OH)CH₂Cl·Cl^- into their enantiomers by inclusion complexation between their racemates in EtOH in the presence of a chiral seed crystal is reported. The enantiomer resolution of the rac-BNO was also accomplished easily by inclusion complexation with achiral ammonium salts, N-(2-hydroxyethyl)-N,N,N-trimethylammonium chloride, Me ₃N⁺CH₂CH₂OH·Cl^- and tetramethylammonium chloride, Me₄N⁺·Cl ^-. Inclusion complexation of the rac-BNO with Me₃N ⁺ CH₂CH₂OH·Cl^- gave only a 1:1 conglomerate inclusion complex but not a racemic complex. Recrystallization of the rac-BNO and an equimolar amount of Me₄N⁺· Cl^- from MeOH (7 ml) and MeOH (15 ml) gave a 1:1:1 racemic complex, BNO·Me₄N⁺·Cl^-·MeOH and a 1:1 conglomerate complex, BNO·Me₄N⁺·Cl ^-, respectively. Novel transformation of the former racemate into the latter conglomerate occurred by heating or by exposure to MeOH vapor in the solid state.

Full text of DOI:10.1016/j.tet.2004.05.122

Tetrahedron 60 (2004) 7767–7774
0
0
Enantiomer separation of rac-2,2 -dihydroxy-1,1 -binaphthyl
(BNO) by inclusion complexation with racemic or achiral
ammonium salts and a novel transformation of a 1:1:1 racemic
1
2
complex of BNO, Me N ·Cl and MeOH into a conglomerate
4
complex in the solid state
*
Kazuhiro Yoshizawa, Shinji Toyota and Fumio Toda
Department of Chemistry, Faculty of Science, Okayama University of Science, Ridai-cho 1-1, Okayama 700-0005, Japan
Received 24 March 2004; revised 19 May 2004; accepted 21 May 2004
Available online 26 June 2004
0
0
Abstract—The complete simultaneous and mutual enantiomer resolution of 2,2 -dihydroxy-1,1 -binaphthyl (BNO) and N-(3-chloro-2-
þ
2
hydroxypropyl)-N,N,N-trimethylammonium chloride, Me N CH CH(OH)CH Cl·Cl into their enantiomers by inclusion complexation
3
2
2
between their racemates in EtOH in the presence of a chiral seed crystal is reported. The enantiomer resolution of the rac-BNO was also
accomplished easily by inclusion complexation with achiral ammonium salts, N-(2-hydroxyethyl)-N,N,N-trimethylammonium chloride,
þ
2
þ
2
þ
Me N CH CH OH·Cl and tetramethylammonium chloride, Me N ·Cl . Inclusion complexation of the rac-BNO with Me N
3
3
2
2
2
4
CH
equimolar amount of Me
conglomerate complex, BNO·Me
2 2
CH OH·Cl gave only a 1:1 conglomerate inclusion complex but not a racemic complex. Recrystallization of the rac-BNO and an
þ
2
þ 2
N ·Cl ·MeOH and a 1:1
4
N ·Cl from MeOH (7 ml) and MeOH (15 ml) gave a 1:1:1 racemic complex, BNO·Me
4
þ
2
N ·Cl , respectively. Novel transformation of the former racemate into the latter conglomerate occurred
4
by heating or by exposure to MeOH vapor in the solid state.
q 2004 Elsevier Ltd. All rights reserved.
1. Introduction
0
0
Enantiomer separation of rac-2,2 -dihydroxy-1,1 -binaph-
thyl (rac-BNO, 1a) is of considerable interest, since
optically active BNO (1b, 1c) is important as a chiral
1
ligand for catalysis in asymmetric synthesis, as a chiral host
2
for enantiomer separation of rac-guests, and as a chiral
shift reagent for determination of optical purity of
1
3
enantiomers using H NMR spectral methods. For
example, enantiomer separation of 1a by inclusion com-
plexation with the chiral hosts, (R,R)-(þ)-2,3-dimethoxy-
0
0
4
N,N,N ,N -tetraalkylsuccinamides (2a, 2b), (þ)-N-benzyl-
5
cinchonidinium chloride (3), optically active N,N,N-tri-
methyl-N-(2-hydroxy-1-alkylethyl)ammonium bromide
6
(
4a-d), and (S)-(2)-N-(3-chloro-2-hydroxypropyl)-N,N,N-
7
trimethylammonium chloride (5b), has been reported.
A commercially available separation method of 1a has
8
also been reported.
We found, however, enantiomer separation of 1a can easily
be accomplished by that inclusion complexation with
conglomeratic ammonium salt (5a) or achiral ammonium
salts such as N,N,N-tetramethyl-N-(2-hydroxyethyl)ammo-
nium chloride (6) and tetramethylammonium chloride (9).
Keywords: Enantiomer separation; Trimethylammonium chloride; Racemic
complex.
*
Corresponding author. Tel./fax: þ81-86-256-9604;
e-mail address: toda@chem.ous.ac.jp
0040–4020/$ - see front matter q 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2004.05.122

7768
K. Yoshizawa et al. / Tetrahedron 60 (2004) 7767–7774
In these enantiomer separations, no chiral source is
necessary. Furthermore, through inclusion complexation
between 1a and 5a, simultaneous and mutual resolution of
both their components was accomplished easily. During the
course of the studies on enantiomer separation of 1a by
complexation with 9, the novel transformation of 1:1:1
racemic inclusion crystals of 1a, 9 and MeOH into
conglomerate crystals in the solid state was found.
Scheme 1. Simultaneous and mutual resolution of 1a and 5a by preferential
crystallization in the presence of chiral seed crystals.
2
. Results and discussion
resolution of 1a was quite good, although that of 5a was
moderate. Nevertheless, the efficient resolution of 1a by
using only a small amount of the seed crystals of 1c·5b and
1b·5c complexes is valuable.
Previously, an efficient resolution of 1a by complexation
with 5b in EtOH, which gives 1b of 99.5% ee and 1c of 99%
ee in 69 and 63% yields, respectively, has been reported.
7
Recently, however, we found that the optically activity of
the ammonium salt is not necessary for the efficient
resolution of 1a.
In order to learn the reason why chiral inclusion complex
1c·5b or 1b·5c crystallizes out preferentially (rather than the
racemic one 1a·5a) from a solution of 1a and 5a in MeOH,
the difference in the stability of these complexes was
estimated by comparison of their melting points. Although
rac-BNO 1a (mp 220 8C) melts at a higher temperature than
does chiral BNO 1b or 1c (mp 211–212 8C), the 1:1 chiral
complexes 1b·5b or 1c·5c melt at much higher temperature
(mp 224–225 8C) than that of racemic complex 1a·5a (mp
187 8C). These data suggest that the chiral complex is more
stable than the racemic one. The fact that 5a exists only as a
conglomerate and not as a racemate also makes easier the
formation of the chiral inclusion crystals 1b·5b or 1b·5c.
Formation of the relatively unstable racemate complex
1a·5a seems rather difficult.
In other words, the enantiomer resolution of 1a can be
accomplished without using any optically active source.
When a complexation of 1a with the conglomeratic salt 5a
was carried out by seeding of 1:1 complexes of 1b with 5b
and of 1c with 5c, efficient simultaneous and mutual optical
resolutions of 1a and 5a were accomplished. According to
the experimental sequences shown in Scheme 1, to a
solution of 1a (7.15 g, 25 mmol) and 5a (3.30 g, 17.5 mmol)
in EtOH (90 ml), was added powdered complex of 1c and
5
1
b (3 mg) and the solution was kept at room temperature for
2 h to give a 1:1 complex of 1c and 5b (0.77 g, 19%). The
complex was dissolved in a mixture of ether and water.
From the ether phase, 1c of 96% ee was obtained, and from
aqueous phase, 5b of 60% ee was obtained. To the EtOH
solution left after separation of the complex of 1c and 5b by
filtration, was added powdered complex of 1b and 5c (3 mg)
and the solution was kept at room temperature for 14 h to
give a 1:1 complex of 1b and 5c (0.80 g, 19%). The complex
was dissolved in a mixture of ether and water. From ether
solution, 1b of 97.5% ee was obtained, and from aqueous
solution, 5c of 66% ee was obtained. By repeating the
seeding experiments a total of eight times as shown in
Scheme 1, 1c of 96% ee (34% yield) and 5b of 59% ee (48%
yield), 1b of 96.5% ee (31% yield), and 5c of 62.5% ee
Easy formation of the complex of the chiral BNO 1b or 1c
with the ammonium salt 5 as a conglomerate rather than a
racemate prompted us to develop further the efficient
enantiomer separation method of 1a by complexation with
a much simpler ammonium salt. N-(2-Hydroxyethyl)-
N,N,N-trimethylammonium chloride (6) was found to
form a 1:1 inclusion complex with 1b or 1c. CD spectra
of these chiral complexes in the solid state are shown in
Figure 1. The racemic complex of 1a and 6 was not formed.
Recrystallization of 1a and 6 from MeOH formed a 1:1
complex 1b·6 or 1c·6 as conglomerate crystals of about
2£7£4 mm in size and average 60 mg in weight for
one piece of crystal. By using this easy formation of
the conglomerate, enantiomer separation of 1a was
(44% yield) were obtained in the yields indicated. In the
mutual resolution between 1a and 5a, the efficiency for the

K. Yoshizawa et al. / Tetrahedron 60 (2004) 7767–7774
7769
99.5% ee (1.85 g, 64% yield). From the aqueous solutions
obtained from both experiments, and from the residue left
after the resolution experiment, 6 (2.92 g, 87% yield) was
recovered unchanged in the total yield indicated. Since 6 is
very cheap, $70/kg, the resolution of 1a with 6 is
economical. In order to clarify the mechanism of the
efficient chiral recognition process between 6 and 1b or 1c,
X-ray structure of the 1:1 complex of 6 and 1b was analyzed
(
lattice is constructed by a hydrogen bond network between
Table 1). As shown in Figure 2, the inclusion crystalline
2
the OH groups of 1c and the Cl of 6. The intramolecular
hydrogen bond between the OH group and Cl2 in 6 also
plays an important role in construction of the inclusion
crystalline lattice. These data suggest that similar ammo-
nium chlorides substituted with trimethyl and hydroxyalkyl
Table 1. Crystallographic data for the inclusion complexes of 1c·6 and 1b·9
Figure 1. CD spectra of 1c·6 (A) and 1b·6 (B) as Nujol mulls.
1
c·6
1b·9
accomplished by complexation with 6 in the presence of the
chiral seed crystals.
Formula
Crystal system
Space group
C 28ClNO
Orthorhombic
25
H
3
C 26ClNO
2
Orthorhombic
1 1 1
P2 2 2 (no. 19)
9.0002(1)
14.1519(3)
16.6811(3)
4
2124.67(6)
1.238
0.0434
24
H
To a solution of 1a (5.72 g, 20 mmol) and 6 (2.79 g,
0 mmol) in MeOH (120 ml), was added powdered 1b·6
P2 2 2 (no. 19)
1
1
1
˚
2
a (A)
9.1385(2)
13.9422(3)
16.9315(3)
4
2157.26(7)
1.311
0.0238
˚
b (A)
complex (3 mg) and the solution was kept at room
temperature for 15 h to give 1b·6 complex (0.68 g, 16%
yield). To the MeOH solution left after separation of the
complex 1b·6 by filtration, was added powdered 1c·6
complex (3 mg) and the solution was kept at room
temperature for 15 h to give 1c·6 complex (1.31 g, 31%
yield). By repeating the same operations a further six times
˚
c (A)
Z
V (A )
Dcalc
3
˚
R
Temperature (8C)
2150
2150
(
as shown in Scheme 2) the 1b·6 complex (3.19 g, 76%
yield) and 1c·6 complex (3.28 g, 77% yield) were obtained
in the total yields indicated. The 1b·6 complex (3.19 g) was
dissolved in a mixture of ether and water. From the ether
solution, 1b of 99% ee was obtained (1.87 g, 66%). The
same treatment of the 1c·6 complex (3.28 g) gave 1c of
Scheme 2. Resolution of 1a by inclusion complexation with 6 in the
presence of chiral seed crystals.
Figure 2. X-ray structure of 1c·6.

7770
K. Yoshizawa et al. / Tetrahedron 60 (2004) 7767–7774
groups such as N-(3-hydroxypropyl)-N,N,N-trimethyl-
ammoniuum chloride (7) and N-(4-hydroxybutyl)-N,N,N-
trimethylammonium chloride (8) could also be used instead
of 6 for the enantiomer separation of 1a. Since 7 formed
chiral crystals with 1b or 1c, 7 can probably be used for the
enantiomer separation of 1a. However, 8 formed only a
racemic complex with 1a, rather than a conglomerate
complex. The alkyl chain length of the hydroxyalkyl group
in 6-8 would be important for the construction of the
inclusion complex with 1. It has been reported that N-hexyl-
N,N,N-trimethylammonium bromide forms only a racemic
which gave 10 in the former experiment, by occasional
shaking of the flask or by seeding with conglomerate
crystals, only 11 was produced. However, the reverse did
not occur.
Trimethylammonium chloride (13) also formed a conglom-
erate complex with 1. However, dimethyl (14) and
methylammonium chloride (15) did not form any complex
with 1. These results are contrasted with that of tetra-
methylammonium chloride (9) that forms both racemic and
conglomerate complexes with 1. The N-hexyl-N,N,N-tri-
methylammmonium bromide, which has a longer alkyl
chain and does not form conglomerate complex, is also
9
complex with 1a. This also supports that the idea that the
alkyl chain length of the ammonium salt is important for
the formation of conglomerate complexes.
9
contrasted with 9.
During the course of the study of the enantiomer separation
of 1a by complexation with the simpler ammonium salt
tetramethylammonium chloride (9), some interesting new
results were found. In the inclusion complexation of 1a
with 6, only the conglomerate complex 1b·6 and 1c·6 was
formed, and the racemic complex 1a·6 was not formed. In
the inclusion complexation of 1a with 9, however, both
racemic (1a·9) and the conglomerate complex (1b·9 and
By using this preferential crystallization of the conglomer-
ates, an efficient enantiomer separation of 1a was accom-
plished. For example, to a solution of 1a (5.7 g, 20 mmol)
and 9 (2.2 g, 20 mmol) in MeOH (70 ml), was added
powdered 1c·9 (3 mg) as seed crystals and the solution was
kept at room temperature for 16 h to give the chiral complex
1c·9 (0.83 g, 21% yield). To the filtrate left after separation
of 1c·9 by filtration, was added powdered 1b·9 as seed
crystals and the solution was kept at room temperature for
16 h to give the chiral complex 1b·9 (1.18 g, 30% yield). By
repeating the preferential crystallization procedure a further
six times as shown in Scheme 3, the two chiral complexes
1c·9 (2.5 g, 63.5% yield) and 1b·9 (2.76 g, 69% yield) were
obtained in the total yields indicated. By treatment with an
ether–water mixture by the same procedure applied for the
resolution of 1a with 5 or 6, 1c of 99% ee (1.5 g, 53% yield)
and 1b of 99% ee (1.44 g, 50% yield) were finally obtained
in the total yields indicated. In order to clarify the precise
chiral recognition mechanism between 1 and 9, the X-ray
crystal structure of the 1b·9 complex was studied (Fig. 4).
As shown in Table 1, crystallographic data of the two
1c·9) was formed. When 1a (1.43 g, 5 mmol) and 9 (0.55 g,
5 mmol) were recrystallized from MeOH (7 ml), a racemic
1:1:1 complex, 1a·9·MeOH (10) crystallized out as colorless
needles. However, when 1a (1.43 g, 5 mmol) and 9 (0.55 g,
mmol) were recrystallized from MeOH (15 ml), a 1:1
mixture of the conglomerate complex 1b·9 and 1c·9 (11)
5
crystallized out as colorless prisms. The racemic structure of
1
0 was elucidated by HPLC analysis and CD spectral
measurement for one piece of single crystal. The IR
spectrum (Fig. 6) and melting point of the conglomerate
mixture 11 (mp 312 8C) are comparable to those of an
authentic sample of the chiral 1b·9 or 1c·9 (mp 313–315 8C)
which are prepared by complexation of 1b or 1c with 9 in
MeOH. These two enantiomeric complexes gave the CD
spectra in the solid state as nice mirror images (Fig. 3). The
difference between the preferential formation of 10 and 11
in the complexation is very dependent on the concentration
of the MeOH solution. However, conglomerate 11 is formed
much more easily. For example, when recrystallization of
1
a and 9 from MeOH was carried out under the conditions
Scheme 3. Resolution of 1a by inclusion complexation with 9 in the
presence of chiral seed crystals.
Figure 3. CD spectra of 1c·9 (A) and 1b·9 (B) as Nujol mulls.

K. Yoshizawa et al. / Tetrahedron 60 (2004) 7767–7774
7771
1
0 into 11 in the solid state were also found (Scheme 4). One
is a thermal transformation. TG and DTA measurements of
0 showed the evaporation of MeOH at 120 8C and phase-
1
transfer from racemate 10 to conglomerate 11 at 183 8C, and
finally showed melting of 11 at 310 8C (Fig. 5).
Figure 5. DTA measurement of racemic complex 1a·9·MeOH (10).
The thermal transformation of 10 into 11 was monitored by
continuous measurements of IR spectra in the solid state by
using the ATR method (Fig. 6). By heating 10 up to 100 8C,
the MeOH absorption of 10 at around 1000 cm2 gradually
decreased as MeOH evaporated from the complex and gave
finally a MeOH-free racemic complex, 1a·9 (12). By further
heating up to 160 8C, the racemic complex intermediate
1
12 turned into the conglomerate 1b·9 and 1c·9 (Scheme 4,
Figure 4. X-ray structure of 1b·9.
Fig. 6). The spectrum of 11 (C in Fig. 6) is identical to that
of an authentic sample of the pure chiral complex 1b·9 (D in
Fig. 6). Finally, the thermal transformation from 10 into 11
was clarified to proceed via the meta-stable racemic
complex intermediate 12. What happens at the molecular
level at the phase-transfer stage from the racemate 10 to the
conglomerate 11? Thermal reaction of one piece of single
crystal of 10 at 160 8C was observed by microscope, but no
special observation was noted, just final conversion into 11.
Heating of fine single needle crystals of 10 at 100 8C for
inclusion complexes 1c·6 and 1b·9 are very comparable.
The mechanism of chiral recognition in these complexes
would also be very similar.
Similar enantiomer separations by the formation of
conglomerate inclusion complexes of 2,7-dimethyltri-
3
,8
cyclo[4.3.1.1 ]undecane-syn-2,syn-7-diol with dimethyl-
0
1
0
0
sulfoxide and of 1,1 -binaphthyl-2,2 -dicarboxylic acid
1
1
with pyrazoles have been reported. The mechanism of
these chiral recognitions using X-ray data analysis has also
been well studied.
3
1
0 min gave 12 as needle crystals. Further heating of 12 at
60 8C for 5 min finally gave 11 as similar needle crystals
(
Fig. 7). However, neither 11 nor 12 seem to be single
crystals. The crystalline shape of 11 is different from that of
the pure chiral crystals 1c·9 (Fig. 7).
3
. Transformation of racemic complex into a
conglomerate complex
In order to certify that 11 is a conglomerate and not a
racemate, the X-ray powder diffraction pattern of 11 was
compared with that of an authentic sample of 1b·9. As
shown in Figure 8, the spectrum of 10 (A) is similar to that
of 12 (B), since both are racemates. However, those of 10
and 12 are completely different from that of 11 (C). The
spectrum of 11 is more similar to that of an authentic sample
of the chiral crystal 1b·9 (D). Slight differences between
the spectra of 11 and that of the authentic sample are
probably due to a difference of molecular packing in the
conglomerate crystals.
During the course of the study of the racemate 10 and
conglomerate 11, two interesting transformation courses of
Very recently, a transformation of the racemic 2,3:6,7-di-
benzo[3.3.1]-nona-2,6-diene-4,8-dione (16) into the con-
glomerate at 130 8C was found as the first example of the
thermal transformation of a racemate to conglomerate in
Scheme 4. Formation of racemic complex 1a·9·MeOH and conglomerate
complexes 1b·9 and 1c·9, and transformation of the former into the latter.

7772
K. Yoshizawa et al. / Tetrahedron 60 (2004) 7767–7774
Figure 6. Monitoring by IR spectral measurements with using the ATR method of the thermal transformation of racemic complex 1a·9·MeOH (10) into
conglomerate complex (11) via racemic intermediate (12): (A) racemic complex (10), (B) racemic intermediate (12), (C) conglomerate complex (11), (D)
authentic sample 1b·9.
Similar conglomerate complex formation by contact with
MeOH vapor was observed for a combination of 1a and 6.
When a 1:1 mixture of 1a and 6 is exposed to MeOH vapor
for 1 h, a 1:1 mixture of the chiral complexes 1b·6 and 1b·6
was obtained, although formation of this conglomerates by
simple mixing of the components using a mortar and pestle
took at least 7 days. The enantioselective movement and
arrangement of BNO molecules in the solid state so as to
form conglomerate complexes of 1 with 6 and 9 is very
interesting. The reason for the interesting role of MeOH
vapor in the formation of these conglomerate complexes in
the solid state is not clear. However, it is known that solid
state reactions can be accelerated in the presence of a small
1
3
amount of a solvent molecule. Similarly, the important
role of guest compound vapor in the formation of
conglomerate complexes has been reported. Generation of
chirality in an inclusion complex of the achiral host, tetra-
(p-bromophenyl)ethylene (17) and an achiral guest occurs
by twisting of the p-bromophenyl groups from the
molecular plane in a clockwise or counterclockwise
direction. When crystalline 17 is exposed to THF and
b-picoline vapor for 24 h, conglomerate complexes of 17
with THF (1:2) and b-picoline (1:1) were produced in the
ratio indicated, although recrystallization of 17 from these
Figure 7. Observation of thermal transformation of racemic complex
a·9·MeOH (10) into conglomerate complex (11) via the racemic
intermediate (12).
1
1
2
the solid state. However, this transformation which occurs
for the crystal of the compound 16 itself is slightly different
from that occurs for the host–guest inclusion crystals of 1
and 9
1
4
neat guest gave the corresponding racemic complexes.
The transformation of 10 into 11 also occurred just by
contact of 10 with MeOH vapor in the solid state (Scheme
4). For example, when powdered or crystalline 10 (30 mg)
was kept in a flask filled with MeOH vapor for 30 min, 11
was obtained in a quantitative yield. The same treatment of
1
2 also gave 11. Very interestingly, contact of a mixture of
powdered 1a and 9 with MeOH vapor for 30 min also gave
1 (Scheme 4), although simple mixing for 30 min of
1
powdered 1a with an equimolar amount of 9 using a mortar
and pestle did not give any complex. However, contact of
these compounds with CHCl , Et O, toluene, or hexane
In the case of BNO, however, MeOH vapor which plays an
important role in the construction of the conglomerate
crystalline lattice is not included in the complex formed.
3
2
vapor did not cause any change.

K. Yoshizawa et al. / Tetrahedron 60 (2004) 7767–7774
7773
Figure 9. Observation of transformation of racemic complex 1a·9·MeOH
10) into conglomerate complex (11) by exposure to MeOH vapor.
(
diffraction analysis. By exposure to MeOH vapor, the
racemic complex crystal (A in Fig. 8) changed to the
conglomerate complex 11 (A in Fig. 10). The spectrum of
11 (A in Fig. 10) is comparable to that of the authentic
sample of the chiral complex 1b·9 (B in Fig. 10).
Figure 10. X-ray powder diffraction pattern of (A) conglomerate complex
(
11) formed by exposure of 10 to MeOH vapor and (B) authentic sample
b·9 (11).
1
Unfortunately, however, each piece of the polycrystallite is
too tiny in size to determine its conglomerate structure by
spectral measurements just for one piece of single crystal.
Figure 8. Monitoring of the thermal transformation of racemic complex
a·9·MeOH (10) into conglomerate complex (11) via racemic intermediate
12) by X-ray powder diffraction analyses: (A) racemic complex (10), (B)
1
(
racemic intermediate (12), (C) conglomerate complex (11), (D) authentic
sample 1b·9.
4. Experimental
4.1. CD and IR spectra and X-ray analysis
Furthermore, the transformation of racemate into conglom-
erate occurs just by enantioselective molecular movement
but not by change of molecular structure of BNO. Finally,
this is the first example of the transformation of racemate
into conglomerate which occurs by contact with molecular
vapor in the solid state.
CD spectra were measured as Nujol mulls according to the
1
5
reported procedure. IR spectra were measured by the ATR
(Attenuated Total Reflection) method. X-ray analysis was
carried out on a Rigaku RAXIS-IV apparatus.
4.1.1. Inclusion complexes 1b·6. Mp 281–283 8C, nOH,
3337 and 3139 cm . Found: C, 70.49; H, 6.63; N, 3.29%.
2
1
In the case of the transformation of 10 into 11 by exposure to
MeOH vapor, a significant change in crystalline shape was
observed. On exposure to MeOH vapor, a single crystal of
racemic complex 1a·9·MeOH (10) was changed into the
polycrystalline conglomerate 1b·9 and 1c·9 (11) (Fig. 9).
This transformation was also monitored by X-ray powder
Calcd for C H ClNO : C, 70.35; H, 6.58; N, 3.10%. 1b·9,
3
2
5
28
mp 300–302 8C, nOH, 3052 cm2 . Found: C, 72.81; H,
1
6.62; N, 3.54%. Calcd for C H ClNO : C, 72.62; H, 6.59;
2
4
26
2
N, 3.71%. 1a·9·MeOH (10), mp was not determined, and did
not give correct elemental analytical results, since MeOH

7774
K. Yoshizawa et al. / Tetrahedron 60 (2004) 7767–7774
1
evaporates easily. The ratio was determined by TG and H
NMR measurements. Racemic structure was elucidated by
HPLC analysis of one piece of single crystal. Conglomerate
References and notes
1. (a) Noyori, R.; Tomino, I.; Tanimoto, Y.; Nishizawa, M. J. Am.
Chem. Soc. 1984, 106, 6709. (b) Seebach, D.; Beck, A. K.;
Roggo, S.; Wonnacott, A. Chem. Ber. 1985, 118, 3673.
(c) Trost, B. M.; Murphy, D. J. Organometallics 1985, 4, 1143.
(d) Wang, J.-T.; Fan, X.; Feng, X.; Qian, Y. M. Synthesis 1989,
291.
mixture of 1b·7 and 1c·7, mp 243–245 8C, nOH,
1
3
275 cm2 . Found: C, 70.77; H, 6.71; N, 3.34%. Calcd
for C H ClNO : C, 70.98; H, 6.87; N, %. 1a·8, mp 188 8C,
2
6
30
3
nOH, 3173 cm2 . Found: C, 71.52; H, 7.39; N, 3.17%.
1
Calcd for C H ClNO : C, 71.43; H, 7.10; N, %.
2
7
32
3
Conglomerate mixture of 1b·13 and 1c·13, mp 229–
2
2. (a) Toda, F.; Mori, K.; Stein, Z.; Goldberg, I. Tetrahedron Lett.
1
2
30 8C, nOH, 3178 cm . Found: C, 72.54; H, 6.43; N,
.76%. Calcd for C H ClNO : C, 72.34; H, 3.67; N, %.
1
989, 30, 1841. (b) Mori, K.; Toda, F. Bull. Chem. Soc. Jpn
3
2
3
24
2
1990, 63, 2127. (c) Toda, F.; Mori, K. J. Chem. Soc., Chem.
Commun. 1986, 1059. (d) Toda, F.; Mori, K. J. Chem. Soc.,
Chem. Commun. 1988, 1357. (e) Toda, F.; Mori, K.; Stein, Z.;
Goldberg, I. J. Org. Chem. 1988, 53, 308.
4
.1.2. Dissociation of inclusion complexes to the com-
ponents by dissolving in a mixture of ether and water.
For example, a 1:1 complex of 1b and 6 (3.19 g) was
dissolved in a 1:1 mixture of ether and water (200 ml). The
ether layer was dried over MgSO4 and kept at room
temperature until 1b crystallized out to give 1b of 99% ee
3
4
5
. Toda, F.; Mori, K.; Okada, J. Chem. Lett. 1988, 131.
. Toda, F.; Tanaka, K. J. Org. Chem. 1988, 3607.
. Tanaka, K.; Okada, T.; Toda, F. Angew. Chem., Int. Ed. Engl.
1
993, 32, 1147.
(
(
(
(
(
1.87 g, 87% yield). Evaporation of aqueous solution gave 6
0.98 g, 93% yield). When a 1:1 complex of 1c and 9
2.50 g) was dissolved in a 1:1 mixture of ether and water
200 ml), from the ether and aqueous solution, 1c of 99% ee
1.52 g, 84% yield) and 9 (0.6 g, 86% yield) were obtained,
6
. Toda, F.; Tanaka, K. Chem. Commun. 1997, 1087.
. Toda, F.; Yoshizawa, K.; Toyota, S.; Chatzziethimiou, S.;
Mavridis, I. M. Org. Biomol. Chem. 2004, 2, 449.
. K. Reuter, WO 9912623, 1999 to Reuter Chemische
Apparatenbau (CA 130:211335).
7
8
9
respectively.
. Marfo-Owusu, E.; Okuyama, K.; Noguchi, K. Mol. Cryst. Liq.
Cryst. 2000, 339, 73.
0. Bishop, R.; Downing, G. A.; Craig, D. C.; Scudder, M. L.
4
1
.1.3. Determination of optical purity. Optical purity of
b and 1c was determined by HPLC using the column
1
1
1
J. Incl. Phenom. Mol. Recog. Chem. 1998, 31, 145.
containing Chiralpak AD (1:1 i-PrOH–hexane). Optical
purity of 5b and 5c was determined by comparison of its
[
1. Hager, O.; Llamas-Saiz, A. L.; Foces-Foces, C.; Claramunt,
R. M.; Lopez, C.; Elguero, J. Helv. Chim. Acta 1999, 82, 2213.
2. Levkin, P. A.; Strelemko, Y. A.; Lyssenko, K. A.; Schurig, V.;
Kostyanovsky, R. G. Tetrahedron: Asymmetry 2004, in press.
3. Shan, N.; Toda, F.; Jones, W. Chem. Commun. 2002, 2372.
4. Tanaka, K.; Fujimoto, D.; Oeser, T.; Irngartinger, H.; Toda, F.
Chem. Commun. 2000, 413.
a] value with that of an authentic sample.
D
1
1
Acknowledgements
We thank Professor Tsukio Ohtani of Okayama University
of Science for his helpful discussion and PhD students
Shinya Hirano and Kazuaki Ikeda of the same University for
their technical assistance on X-ray analysis and X-ray
powder diffraction analysis.
1
5. Toda, F.; Miyamoto, H.; Kikuchi, S.; Kuroda, R.; Nagami, F.
J. Am. Chem. Soc. 1996, 118, 11315.