Enantiomer separation by enantioselective inclusion complexation-organic solvent nanofiltration

Article abstract of DOI:10.1016/j.tetasy.2006.06.026

A novel chiral separation process, which utilizes a combination of enantioselective inclusion complexation (EIC) and organic solvent nanofiltration (OSN), was developed. Although EIC is an attractive way to resolve racemates, the difficulties associated with enantiomer recovery and chiral host recycle has limited large-scale applications. EIC coupled with OSN replaces distillation for the recovery of enantiomers from enantioenriched solid complex. A decomplexation solvent is employed to dissociate enantiomers from the complex, and subsequent separation of enantiomers from the chiral host is realized using OSN. The new process was investigated using racemic 1-phenylethanol as the guest and (R,R)-TADDOL as the chiral host. This novel technology expands the application of EIC to the resolution of nonvolatile racemates, and enables large-scale application.

Full text of DOI:10.1016/j.tetasy.2006.06.026

Tetrahedron: Asymmetry 17 (2006) 1846–1852
Enantiomer separation by enantioselective inclusion
complexation–organic solvent nanofiltration
Nazlee F. Ghazali,a Frederico C. Ferreira,a Andrew J. P. Whiteb and Andrew G. Livingstona,*
aDepartment of Chemical Engineering, Imperial College, London SW7 2AZ, UK
bDepartment of Chemistry, Imperial College, London SW7 2AZ, UK
Received 18 May 2006; accepted 16 June 2006
Abstract—A novel chiral separation process, which utilizes a combination of enantioselective inclusion complexation (EIC) and organic
solvent nanofiltration (OSN), was developed. Although EIC is an attractive way to resolve racemates, the difficulties associated with
enantiomer recovery and chiral host recycle has limited large-scale applications. EIC coupled with OSN replaces distillation for the
recovery of enantiomers from enantioenriched solid complex. A decomplexation solvent is employed to dissociate enantiomers from
the complex, and subsequent separation of enantiomers from the chiral host is realized using OSN. The new process was investigated
using racemic 1-phenylethanol as the guest and (R,R)-TADDOL as the chiral host. This novel technology expands the application of
EIC to the resolution of nonvolatile racemates, and enables large-scale application.
Ó 2006 Elsevier Ltd. All rights reserved.
1. Introduction
tions. Here we report, for the first time, a novel enantiosep-
aration process, which combines the highly enantioselective
nature of inclusion complexation with the subsequent sep-
aration of enantiomers from a chiral host using solvent
decomplexation and organic solvent nanofiltration
(OSN). In our proposed process (Fig. 1), a racemate is
added to a chiral host suspended in a resolution solvent.
The (S)-enantiomer enantioselectively co-crystallizes with
the chiral host while the (R)-enantiomer remains in the
liquid (Step A). Nanofiltration of the resulting resolution
suspension elutes the (R)-enantiomer (Step B), retaining
the chiral host and the chiral host–(S)-enantiomer com-
plex. A decomplexation solvent is then added to dissolve
and dissociate the complex into (S)-enantiomer and host
(Step C). This solution is subsequently nanofiltered to elute
the (S)-enantiomer, while the soluble host is retained by the
membrane (Step D). Exchanging the decomplexation sol-
vent for the resolution solvent via diafiltration with the res-
olution solvent causes the host to recrystallize (Step E), and
it is returned to the next cycle. The ambient operating tem-
perature and simplicity of OSN separations make com-
bined EIC–OSN ideal for the separation of nonvolatile
and labile enantiomers. Since chiral hosts are used in stoi-
chiometric quantities, their recovery and multiple reuse is a
further key advantage of separation by OSN.
Resolution via diastereomeric salt formation remains com-
mon in large-scale production of optically pure enantio-
mers due to its operational simplicity and reliability.1
However, it is limited to the resolution of acids and bases.
Enantioselective inclusion complexation (EIC) is a promis-
ing technique which, since it is not restricted to proton
transfer interactions, can in principle be used to resolve
compounds with almost any functional group.2 Numerous
successful resolutions have been reported by Toda and
co-workers3 together with other novel developments,4
although we also note that some recent investigations5 re-
port a less optimistic view of the scope of EIC. Neverthe-
less, versatile chiral hosts such as TADDOLs are
excellent in terms of resolving structurally similar alcohols,
ketones, amines, sulfoxides and amino acid esters.6 How-
ever, an outstanding difficulty with EIC is the separation
of enantiomers from the enantioenriched solid complex,
which is typically performed by distillation. This limits
EIC to the separation of enantiomers with appreciable vol-
atility. Even for these it would be difficult to operate at
large scale due to high temperature distillation, with its
concomitant problematic heat transfer and vacuum condi-
*
To demonstrate this process, we used rac-phenylethanol 1
(122 g molꢀ1) as a racemate and (4R,5R)-(ꢀ)-2,2-di-
Corresponding author. Tel.: +44 20 7594 5582; fax: +44 20 7594 5629;
e-mail: a.livingston@ic.ac.uk
0957-4166/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetasy.2006.06.026

N. F. Ghazali et al. / Tetrahedron: Asymmetry 17 (2006) 1846–1852
1847
Figure 1. Process schematic of enantioselective inclusion complexation–organic solvent nanofiltration and structure of model host and enantiomers.
methyl-a,a,a0,a0-tetraphenyl-1,3-dioxolane-4,5-dimethanol
[(R,R)-TADDOL] 2 (466 g molꢀ1) as a chiral host.7 A com-
mercially available polyimide OSN membrane, STAR-
MEMTM 122,8 with a nominal molecular weight cutoff
(MWCO) of 220 g molꢀ1 was used for OSN. These OSN
membranes have been found to be effective for molecular
level separations.9 It was anticipated, based on MWCO,10
that the host would be retained while the enantiomers
would permeate freely. Hexane was used as the resolution
solvent and toluene as the decomplexation solvent.
of 1 possessed a degree of crystallinity, suggesting that a
change of crystal structure had occurred upon the addition
of 1 to crystalline 2.
For the X-ray crystal structure determination, single crys-
tals were obtained by recrystallizing the crude inclusion
complex from a 1:2 (v/v) toluene/hexane mixture. The X-
ray crystal structure of the 2:1 inclusion complex formed
between host 2 and guest 1a is shown in Figure 2. The dom-
inant feature is the formation of a five O–Hꢁ ꢁ ꢁO hydrogen
bond ring formed from two intramolecular linkages be-
tween the hydroxyl groups of each host 2 molecule (inter-
actions a and b), one intermolecular host 2ꢁ ꢁ ꢁhost 2
linkage (interaction c) and two linkages between the in-
cluded guest 1a and host 2 molecules (interactions d and
e). It has previously been shown that in the absence of a
guest molecule, host 2 crystallizes in a somewhat similar
fashion, with two independent molecules forming a four
O–Hꢁ ꢁ ꢁO hydrogen bond ‘square’ (see Fig. 3).12 The addi-
tion of guest 1a effectively results in one side of this square
‘opening up’ to accommodate the extra OH moiety (break-
ing one O–Hꢁ ꢁ ꢁO linkage and forming two new ones), with
one of the host 2 molecules (the one on the right hand side
of both Figs. 2 and 3) taking up a noticeably different
orientation. Interestingly, the inclusion complex formed
between (S)-phenylethylamine (essentially just changing
the OH in guest 1a for an NH2) and host 2 is almost
identical.13
2. Results and discussion
Before performing the process, several fundamental phe-
nomena such as the membrane separation characteristics
of racemate and chiral host, solid-state properties of the
solid complex, kinetics of resolution, effect of solvent com-
position on decomplexation and parameters affecting
resolution were investigated. Racemic 1 was highly perme-
able in the OSN membrane, and had zero rejection11 in
both toluene and hexane. Chiral host 2 was retained effi-
ciently by the OSN membrane (>99% rejection in both sol-
vents). The resolution of 1 with 2 was rapid and reached
equilibrium after 1 h, based on measurements of the enan-
tiomeric excess of the mother liquor. Crystals of 2 initially
present as a suspension in hexane were easily observed un-
der a microscope, but upon the addition of 1 the solid was
converted into a far more finely divided material for which
crystals could not be seen. However, powder X-ray diffrac-
tion patterns showed that the solids formed after addition
From an inspection of Figure 2 it is evident that the
involvement of the hydroxyl moiety of the guest in the

1848
N. F. Ghazali et al. / Tetrahedron: Asymmetry 17 (2006) 1846–1852
in order to explain the selective inclusion of guest 1a with-
in the host 2 matrix, the environment around and con-
tacts formed by the hydride and the methyl must be
considered. In fact both of these substituents form C–
Hꢁ ꢁ ꢁp hydrogen bonds to adjacent phenyl rings, one from
each unique host 2 molecule (interactions f and g, respec-
tively). The positions of the two host 2 molecules are such
that if the hydride and methyl swapped places there
would be a C–Hꢁ ꢁ ꢁp contact from the new methyl group
to an adjacent phenyl ring with an Hꢁ ꢁ ꢁp separation of ca.
˚
1.45 A, clearly too short to be ‘allowable’. Thus, the two
host 2 molecules would have adopted a different mutual
orientation to accommodate guest 1b. We, therefore, sug-
gest that the distortion in the host 2 framework necessary
for the inclusion of guest 1b would severely disrupt the
energetically favourable O–Hꢁ ꢁ ꢁO hydrogen bonding
ring and thus disfavour the formation of an inclusion
complex.
Table 1 shows equilibrium data for a range of resolutions.
All resolutions exhibit selectivity towards (S)-enantiomer
1a; virtually all of the (R)-enantiomer 1b remains in the
liquid phase uncomplexed. At a low concentration of the
feed racemate (entries 1 and 2), the (S)-enantiomer concen-
tration in the liquid remained constant at around 4 mM.
However, as the feed racemate concentration was in-
creased, the concentration of (S)-enantiomer in the liquid
seemed to increase (entries 3–5). For a molar equivalent
P1.0, the resolutions exhibit the same values of ee (solid
and liquid) and concentrations in the liquid (entries 6–9).
At a molar equivalent of host = 0.5, more (S)-enantiomer
was observed in the liquid phase presumably because the
host added was insufficient to complex with the available
(S)-enantiomer. The results of resolution for hexane and
octane were comparable (entries 7 and 17) suggesting that
there is no advantage in performing the resolution in a
slightly more hydrophobic solvent. Practically, there is no
significant improvement in ee by reducing the temperature
(entries 14–16).
Figure 2. The molecular structure of inclusion complex [2]2Æ1a. The O–
˚
Hꢁ ꢁ ꢁO hydrogen bonding geometries, [Oꢁ ꢁ ꢁO], [Hꢁ ꢁ ꢁO] (A), [O–Hꢁ ꢁ ꢁO] (°),
are (a) 2.6747(17), 1.80, 162; (b) 2.6462(18), 1.77, 163; (c) 2.7988(17), 2.02,
144; (d) 2.695(2), 1.81, 170; (e) 2.702(2), 1.86, 156. The geometries of the
˚
C–Hꢁ ꢁ ꢁp contacts [Hꢁ ꢁ ꢁp] (A), [C–Hꢁ ꢁ ꢁp] (°) are (f) 2.67, 142; (g) 2.58, 172.
The decomplexation step is critical, and it is affected by the
solvent mixture composition. Toluene was chosen as the
decomplexation solvent since it is miscible with hexane
and does not form hydrogen bonds with the host. We ex-
plored the possibility of running the resolution in tolu-
ene/hexane mixtures of varying compositions, to avoid
having to revert to pure hexane at the end of each cycle
(Step E in Fig. 1). Reverting to pure hexane would have
required extensive diafiltration. Figure 4b shows that it is
possible to run the resolution anywhere between 0 and
20 vol % toluene without sacrificing the enantiomeric ex-
cess (ee). At a high toluene composition there was insuffi-
cient host remaining as a solid (Fig. 4a) to form a
complex, leading to a lower ee in the liquid. The complex
completely dissociated into free enantiomers and host
above 60 vol % toluene (Fig. 4c). Therefore, from the
process point of view, it is not necessary to use either pure
hexane for resolution (Step A), or pure toluene for
decomplexation (Step D in Fig. 1).
Figure 3. The molecular structure of SEWVUL01. The O–Hꢁ ꢁ ꢁO hydro-
˚
gen bonding geometries, [Oꢁ ꢁ ꢁO], [Hꢁ ꢁ ꢁO] (A), [O–Hꢁ ꢁ ꢁO] (°), are (a)
2.619, 1.73, 171; (b) 2.612, 1.72, 174; (c) 2.74, 2.04, 134; (d) 2.735, 2.05, 132
˚
(O–H distances normalized to 0.90 A).
hydrogen bonding cycle would not be directly affected by
a change in the chirality at C(87) (all that would be re-
quired to change the chirality at this centre is swapping
the locations of the hydride and methyl moieties). Thus,
Figure 5 provides a detailed description of a sequence of
a set of resolution–filtration cycles. All operations were car-

N. F. Ghazali et al. / Tetrahedron: Asymmetry 17 (2006) 1846–1852
Table 1. Effect of several parameters on the resolution of racemate 1 mediated by host 2a
1849
Entry
Initial substrate
concentration (mM)
T (°C)
TADDOL
(equiv)
Solvent
Concn of R in
liquid (mM)
Concn of S
in liquid (mM)
ee (%) of
liquid (R)
ee (%) of
solid (S)
Yieldb of (S)
in solid (%)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
25
50
100
200
400
50
50
50
50
50
22
22
22
22
22
22
22
22
22
22
22
22
22
5
1.0
1.0
1.0
1.0
1.0
0.5
1.0
1.5
2.0
1.0
2.0
1.0
2.0
1.0
1.0
1.0
1.0
2.0
Hexane
Hexane
Hexane
Hexane
Hexane
Hexane
Hexane
Hexane
Hexane
Hexane
Hexane
Hexane
Hexane
Hexane
Hexane
Hexane
Octane
Octane
14
24
51
92
191
25
24
25
25
24
25
45
45
23
23
24
23
23
4
4
7
10
21
14
4
4
4
4
4
5
4
3
3
4
3
4
54
74
75
80
80
30
74
74
73
72
74
80
82
80
81
76
76
73
93
89
85
82
80
93
89
90
86
88
88
82
82
83
85
87
87
88
1
38
74
80
80
9
38
38
41
51
32
69
58
90
90
87
37
29
50
100
100
50
50
50
10
22
22
22
50
50
Note: For entries 14–16, solid ee and yield of S were calculated based on the initial moles fed and the moles in the liquid.
a The duration of resolution was 6 h.
b
S in solid
S fed
Yield ¼
:
60
50
40
30
20
10
0
60
50
40
30
20
10
0
60
50
40
30
20
10
0
Host Solubility
(a)
Complexation
(b)
Decomplexation
(c)
R
S
R
S
Complete Host
Solubilization
Complete Complex
Dissociation
Complete Complex
Dissociation
0
20
40
60
80
100
0
20
40
60
80
100
0
20
40
60
80
100
Toluene in hexane/toluene
mixture / vol %
Toluene in hexane/toluene
mixture / vol %
Toluene in hexane/toluene
mixture / vol %
Figure 4. Solubilities as a function of solvent composition: (a) solubility of host 2; (b) solubility of 1a and 1b with host 2 (complexation) and; (c) solubility
of 1a and 1b when introduced as a solid complex with host 2. Solid complex used in (c) was recovered from a resolution experiment employing 1 equivalent
of host 2.
ried out in a dead end nanofiltration cell. In the first cycle,
racemate 1 (50 mM) was added to the cell containing host
2 (1 equiv) suspended in hexane and stirred. In the course
of resolution, the (S)-enantiomer 1a preferentially co-crys-
tallized with the 2 and this led to an enrichment of (R)-
enantiomer 1b in the mother liquor (Step 1A). After 6 h
of resolution, a pressure of 30 bar was applied and the li-
quid enriched with (R)-enantiomer permeated the mem-
brane (Step 1B1). The solid complex was retained in the
cell. In Steps 1B2 and 1B3, fresh hexane was added to the
cell and two filtrations used to elute R-enantiomer. The
combined permeate stream gave an enantiomeric excess
of 46% (of R) and a yield14 of 89% (of R). After virtually
all of the dissolved (R)-enantiomer had been eluted out,
toluene was added to take the mixture composition above
60 vol % toluene, above which the complex completely dis-
solves as (S)-enantiomer 1a and host 2 (Step 1C). The cell
was then pressurized and the (S)-enantiomer permeated the
membrane, while the soluble host was retained (Step 1D1).
Further decomplexation solvent (60% toluene) was then
added to the cell for elution of the (S)-enantiomer (Step
1D2). The ee and yield of the combined permeate stream
were 80% (of S) and 31% (of S), respectively. After the
(S)-enantiomer had been eluted out, fresh hexane was
added to the retentate to precipitate the host (Step 1E)
and prepare it for the second cycle. To begin the second

1850
N. F. Ghazali et al. / Tetrahedron: Asymmetry 17 (2006) 1846–1852
Figure 5. Process operations of enantioselective inclusion complexation–solvent decomplexation–OSN. Enantiomeric excess and moles of R and S in
aggregated streams are shown.
cycle, fresh racemate was added to the host suspended in
the cell to take the composition to 15 vol % toluene. Similar
(R)- and (S)-elution steps were then executed. For the (R)-
elution steps, the ee and the yield of the combined stream
were 34% (of R) and 80% (of R), respectively. For the
(S)-elution steps, the ee and the yield of the combined
stream were 95% (of S) and 51% (of S), respectively. The
elution profiles are shown in Figure 6. The enantioselectiv-
ity of the resolution was maintained at high ee (80% and
95% of S) in these two cycles, and, in principle, multiple
cycles could be executed.
0.70
0.60
0.50
0.40
0.30
0.20
0.10
0.00
100
80
FIRST CYCLE
SECOND CYCLE
60
R
40
S
20
ee
3. Conclusions
0
Step B
Step D
Step B
Step D
A novel enantioseparation process using combined EIC–
OSN has been demonstrated. We have shown for the first
time that the use of solvent decomplexation and OSN
allows enantiomer purification and the reuse/recycling of
the chiral host. The strength of this process is the direct
use of chiral host (without derivatization or immobiliza-
tion), relatively high operating concentrations and ambient
temperature processing. By selecting appropriate OSN
membranes with a suitable MWCO it is possible to sepa-
rate purified enantiomers and hosts. We regard the method
presented here as having the potential as an alternative
technique for preparative-scale chiral separations, which
could extend the scope of EIC from the laboratory to the
pilot and the industrial scale.
-20
-40
-60
-80
-100
1B1 1B2 1B3 1D1 1D2 2B1 2B2 2B3 2D1 2D2 2D3
Filtration numbers
Figure 6. Elution profile of the resolution–filtration process and ee of the
permeate stream in each filtration (indicated in Fig. 4) in two cycles. Step
B is the elution of R and Step D is the elution of S. A positive ee indicates
an (S)-rich permeate. A negative ee indicates an (R)-rich permeate.

N. F. Ghazali et al. / Tetrahedron: Asymmetry 17 (2006) 1846–1852
1851
4. Experimental
CB21EZ, UK; fax: (+44) 1223-336-033; or deposit@ccdc.
cam.ac.uk).
4.1. General
Acknowledgements
Racemic 1 was obtained from Aldrich and host 2 was
supplied by Fluka. HPLC grade solvents such as hexane
and toluene were supplied by Fisher Scientific UK.
We thank Richard Sweeney (Department of Materials,
Imperial College) for his technical help with the powder
XRD analysis. N.F.G. acknowledges the financial support
from the Ministry of Science, Technology and Innovation
(MOSTI) of Malaysia.
4.2. Analytical techniques
The enantiomers were analyzed by GC (Agilent, US)
equipped with an HP-CHIRAL-20b (Agilent) column.
Host concentration was measured by HPLC (Unicam,
UK) with a Chiralcel OD-H (Daicel) column and using
90:10 (hexane/isopropanol) as the mobile phase.
References
1. (a) Kozma, D. CRC Handbook of Optical Resolution via
Diastereomeric Crystallisation; CRC Press, 2002; (b) Collet,
A. Angew. Chem., Int. Ed. 1998, 37, 3239–3241.
4.3. Resolution–filtration procedure
2. (a) Toda, F. Pure Appl. Chem. 2001, 73, 1137–1145; (b)
Seebach, D.; Beck, A.; Heckel, A. Angew. Chem., Int. Ed.
2001, 40, 92–138; (c) Kaupp, G. Angew. Chem., Int. Ed. Engl.
1994, 33, 728–729; (d) Toda, F.; Tanaka, K. J. Am. Chem.
Soc. 1983, 105, 5151–5152; (e) Weber, E.; Wimmer, C.;
Llamas-Saiz, A. L.; Foces-Foces, C. J. Chem. Soc., Chem.
Commun. 1992, 733–735; (f) Tanaka, K.; Honke, S.; Urban-
czyk-Lipkowska, Z.; Toda, F. Eur. J. Org. Chem. 2000, 3171–
3176; (g) Kassai, C.; Juvancz, Z.; Balint, J.; Fogassy, E.;
Kozma, D. Tetrahedron 2000, 56, 8355–8359; (h) Grandeury,
A.; Petit, S.; Gouhier, G.; Agasse, V.; Coquerel, G. Tetra-
hedron: Asymmetry 2003, 14, 2143–2152.
3. (a) Miyamoto, H.; Sakamoto, M.; Yoshioka, K.; Takaoka,
R.; Toda, F. Tetrahedron: Asymmetry 2000, 11, 3045–3048;
(b) Mori, K.; Toda, F. Tetrahedron: Asymmetry 1990, 1, 281–
282; (c) Toda, F.; Takumi, H.; Tanaka, K. Tetrahedron:
Asymmetry 1995, 6, 1059–1062.
A stainless steel SEPA ST (Osmonics, US) dead end nano-
filtration cell with an effective membrane area of 13.9 cm2
was employed as a resolution–filtration vessel. A mem-
brane disk was clamped in place at the base of the cell.
Membranes were preconditioned with approximately
400–500 ml of toluene in order to remove the lube oil pre-
servative from the polymer and to compress the membrane
at operating conditions. A typical procedure for resolution
filtration was as follows. With a preconditioned membrane
in place, 0.903 g of host 2 (2.047 mmol), hexane (40 ml) and
rac-1 (2.218 mol) were quickly added to the cell and the
suspension formed was agitated. After 6 h, a pressure of
30 bar was applied (supplied by N2 gas) and the permeate
was collected for GC and HPLC analyses. After 75% of
the initial liquid volume had permeated (about 30 ml),
the cell was depressurized. Fresh solvent was added to
the retentate (10 ml), and stirred for 30 min and then fil-
tered again by pressurizing the cell (see Fig. 5). A similar
procedure was used for the decomplexation step, but in-
stead of adding hexane, toluene or toluene/hexane mixture
was added. After elution of R, followed by decomplexation
and elution of S, fresh racemate was added to the cell for
the next cycle of resolution.
4. (a) Legrand, S.; Luukinen, H.; Isaksson, R.; Kilpela¨inen, I.;
Lindstro¨m, M.; Nicholls, I.; Unelius, C. Tetrahedron: Asym-
metry 2005, 16, 635–640; (b) Deng, J.; Chi, Y.; Fu, F.; Cui,
X.; Yu, K.; Zhu, J.; Jiang, Y. Tetrahedron: Asymmetry 2000,
11, 1729–1732; (c) Bortolini, O.; Fantin, G.; Fogagnolo, M.
Chirality 2005, 17, 121–130.
5. (a) Muller, S.; Afraz, M. C.; de Gelder, R.; Ariaans, G. J.;
¨
Kaptein, B.; Broxterman, Q. B.; Bruggink, A. Eur. J. Org.
Chem. 2005, 1082–1096; (b) Muller, S.; Ariaans, G. J.;
¨
Kaptein, B.; Broxterman, Q. B.; Bruggink, A. Tetrahedron:
Asymmetry 2005, 16, 2535–2538.
4.4. X-ray crystal structure analysis
6. (a) Toda, F. Supramol. Sci. 1996, 3, 139–148; (b) Toda, F.;
Tanaka, K. Tetrahedron Lett. 1988, 29, 551–554.
Crystal data for inclusion complex [2]2Æ1a: (C31H30O4)2Æ
C8H10O, M = 1055.26, orthorhombic, P212121 (no. 19),
7. Resolution using chiral host 2 and its derivatives: (a) Toda, F.;
Tohi, Y. J. Chem. Soc., Chem. Commun. 1993, 1238–1240; (b)
Toda, F.; Tanaka, K.; Ootani, M.; Hayashi, A.; Miyahara, I.;
Hirotsu, K. J. Chem. Soc., Chem. Commun. 1993, 1413–1415;
(c) von dem Bussche-Hunnefeld, C.; Beck, A.; Lengweiler, U.;
Seebach, D. Helv. Chim. Acta 1992, 75, 438–441.
8. STARMEMTM 122 (W. R. Grace & Co., US) was kindly
supplied by Membrane Extraction Technology (www.
membrane-extraction-technology.com).
9. (a) Ghazali, N. F.; Patterson, D.; Livingston, A. Chem.
Commun. 2004, 8, 962–963; (b) Luthra, S. S.; Yang, X.; dos
Santos, L. M. F.; White, L. S.; Livingston, A. G. Chem.
Commun. 2001, 16, 1468–1469; (c) Nair, D.; Scarpello, J. T.;
White, L. S.; dos Santos, L. M. F.; Vankelecom, I. F. J.;
Livingston, A. G. Tetrahedron Lett. 2001, 42, 8219–8222.
10. MWCO is obtained from the curve of the rejections (see Ref.
7) of a series of n-alkanes in toluene at 2 wt % versus their
molecular weight, by defining the molecular weight corre-
sponding to 90% rejection as the MWCO.
˚
a = 16.8753(5), b = 16.8943(5), c = 20.2579(6) A, V =
3
˚
5775.5(3) A , Z = 4,
q
calcd = 1.214 g cmꢀ3
,
l(MoKa) =
0.079 mmꢀ1, T = 173 K, colourless prismatic blocks, Ox-
ford Diffraction Xcalibur 3 diffractometer; 19,946 indepen-
dent measured reflections, F2 refinement, R1 = 0.066, wR2 =
0.156, 19,271 independent observed absorption–corrected
reflections [jFoj > 4r(jFoj), 2h max = 65°], 718 parameters.
The absolute structure of inclusion complex [2]2Æ1a could
not be determined by either R-factor tests ½Rþ1 ¼
0:0659; Rꢀ1 ¼ 0:0659ꢂ or by use of the Flack parameter
[x+ = +0.4(6), xꢀ = +0.6(6)] and so was assigned using
the known centre of the guest 1a at C(87). CCDC
277087. The supplementary crystallographic data for this
paper can be obtained free of charge via www.ccdc.cam.
ac.uk/conts/retrieving.html (or from the Cambridge
Crystallographic Data Centre, 12 Union Road, Cambridge

1852
11.
N. F. Ghazali et al. / Tetrahedron: Asymmetry 17 (2006) 1846–1852
13. [LATCOY] Toda, F.; Tanaka, K.; Ootani, M.; Hayashi, A.;
solute in permeate ðmMÞ
Rejection ð%Þ ¼ 1 ꢀ
ꢃ 100:
Miyahara, I.; Hirotsu, K. J. Chem. Soc., Chem. Commun.
1993, 1413–1415.
solute in retentate ðmMÞ
14.
12. [SEWVUL01] Toda, F.; Tanaka, K.; Miyamoto, H.;
Koshima, H.; Miyahara, I.; Hirotsu, K. J. Chem. Soc.,
Perkin Trans. 2 1997, 1877–1886.
enantiomer in permeate ðmolÞ
Yield ð%Þ ¼
ꢃ 100:
enantiomer fed in the cycle ðmolÞ