Chiral recognition in the solid state: Crystallographically characterized diastereomeric co-crystals between a synthetic chiral selector (Whelk-O1) and a representative chiral analyte

Article abstract of DOI:10.1016/j.tetasy.2005.08.027

Designed to distinguish between the enantiomers of compounds possessing commonly occurring structural features, the chiral selector used in the chiral stationary phase (CSP) 1 (Whelk-O1) is broadly applicable. In an effort to further the understanding of the mechanism of chiral recognition with this chiral selector, both diastereomeric combinations of selector 1 and a representative analyte, the pivalamide of p-bromo-α-phenylethylamine, 2, were successfully co-crystallized and characterized by single crystal X-ray diffraction. The crystal corresponding to the complex that is more stable in solution is consistent with our previously reported chiral recognition model. The aromatic portion of 2 is in the cleft of selector 1, displaying both face-to-face and face-to-edge π-π interactions as well as a hydrogen bond between the benzamide proton of the selector and the carbonyl oxygen of the analyte. For the crystal corresponding to the complex, which is less stable in solution, the aromatic portion of 2 is not in the cleft of selector 1, having approached from the opposite face of the π-acidic dinitrobenzamide moiety so as to undergo face-to-face π-π and hydrogen bonding interactions. Comparisons of these structures and their relevance to enantioselective chromatography are also discussed.

Full text of DOI:10.1016/j.tetasy.2005.08.027

Tetrahedron:
Asymmetry
Tetrahedron: Asymmetry 16 (2005) 3147–3153
Chiral recognition in the solid state: crystallographically
characterized diastereomeric co-crystals between a synthetic
chiral selector (Whelk-O1) and a representative chiral analyte
Michael E. Koscho,* Patrick L. Spence and William H. Pirkle
School of Chemical Sciences, University of Illinois, Urbana, IL 61801, USA
Received 6 June 2005; accepted 12 August 2005
Available online 28 September 2005
Abstract—Designed to distinguish between the enantiomers of compounds possessing commonly occurring structural features, the
chiral selector used in the chiral stationary phase (CSP) 1 (Whelk-O1) is broadly applicable. In an effort to further the understanding
of the mechanism of chiral recognition with this chiral selector, both diastereomeric combinations of selector 1 and a representative
analyte, the pivalamide of p-bromo-a-phenylethylamine, 2, were successfully co-crystallized and characterized by single crystal X-
ray diffraction. The crystal corresponding to the complex that is more stable in solution is consistent with our previously reported
chiral recognition model. The aromatic portion of 2 is in the cleft of selector 1, displaying both face-to-face and face-to-edge p–p
interactions as well as a hydrogen bond between the benzamide proton of the selector and the carbonyl oxygen of the analyte. For
the crystal corresponding to the complex, which is less stable in solution, the aromatic portion of 2 is not in the cleft of selector 1,
having approached from the opposite face of the p-acidic dinitrobenzamide moiety so as to undergo face-to-face p–p and hydrogen
bonding interactions. Comparisons of these structures and their relevance to enantioselective chromatography are also discussed.
ꢀ 2005 Elsevier Ltd. All rights reserved.
1. Introduction
without departing substantially from a low energy con-
formation, can undergo simultaneous face-to-face p–p
interaction with the dinitrobenzamide portion of the
selector as well as a hydrogen bonding interaction with
the amide proton of the selector. Additionally, a face-
to-edge p–p interaction will enhance the affinity of the
selector for the more retained enantiomer of the analyte.
The analyte enantiomer, which is less retained, is unable
to achieve these same analyte–selector interactions with-
out unfavorable torsional and/or steric destabilization.
This model for chiral recognition is consistent with the
chromatographic data and accounts for observed elution
orders for the enantiomers of analytes in cases where the
absolute configuration of the analyte is known. When
applied properly, this model can even allow the assign-
ment of absolute configuration.17 The model is also con-
sistent with NMR data determined for these complexes
using soluble analogues of CSP 1.18,19 Additionally, this
model is consistent with recent computational modeling
studies between an analogue of CSP 1, and derivatives of
N-pivaloyl-a-phenylethylamine.20
We have been concerned with the design, development,
and evaluation of low molecular weight chiral selectors
used in enantioselective chromatography. The selector
used in chiral stationary phase (CSP) 1 was originally
designed to differentiate the enantiomers of naproxen and
other non-steroidal anti-inflammatory drugs.1–3 It has
since been shown that it is capable of resolving the enan-
tiomers of a wide variety of chiral analytes using a range
of chromatographic conditions.4–7 In general, analytes
with an aromatic ring and a hydrogen bond acceptor
near a stereogenic center are candidates for enantiodiffer-
entiation by CSP 1. From systematic chromatographic
studies, a model accounting for the manner in which
the selector interacts with the more retained enantiomer
of numerous analytes has been reported.8–16 The selector
is thought to have a cleft-like binding site formed by the
planes of the naphthyl and dinitrobenzamide rings. CSP
1 preferentially retains the analyte enantiomer, which
Herein, we report the solid state structures of a 1:1
complex for each diastereomeric pair formed from a
CSP 1 analogue compound, 1, and the pivalamide of
*
Corresponding author at present address: Department of Chemistry,
Mississippi State University, MS 39762, USA. Tel.: +1 662 325 9500;
fax: +1 662 325 1618; e-mail: m.koscho@msstate.edu
0957-4166/$ - see front matter ꢀ 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetasy.2005.08.027

3148
M. E. Koscho et al. / Tetrahedron: Asymmetry 16 (2005) 3147–3153
p-bromo-a-phenylethylamine, 2.21–24 Since it is the dif-
ference in energy of the two diastereomeric adsorbates,
which determines the extent of enantioselectivity, struc-
tural modifications, which increase this energy difference
will increase the chromatographic separation factor
whether it is by increasing the affinity of the selector
for the more retained analyte enantiomer or by decreas-
ing the affinity of the selector for the less retained enan-
tiomer of the analyte. A detailed understanding of how
the selector interacts with each analyte enantiomer
should allow design modifications for selectors of im-
proved scope and/or selectivity (Fig. 1).
change in the Cahn–Ingold–Prelog priority at the 3-
position when propyl silyl moiety is exchanged for a
methyl group, which accounts for the stereochemical
descriptor at this position being (R)- for the CSP and
(S)- for the soluble selector when the stereochemical
descriptor at the 4-position is S in both cases.
The chromatographic separation factor (a) is related to
the difference in free energy of the adsorption of the two
enantiomers by the equation: DDG = RTlna. From the
separation of the enantiomers of 2 on CSP 1, the energy
difference between the two diastereomeric adsorbates is
1.44 kcal/mol at 25 ꢁC. From the reciprocal chromato-
graphic experiment, the difference in energy of the two
diastereomeric adsorbates formed with the enantiomers
of 1 and CSP 2 is 1.08 kcal/mol at 25 ꢁC. It should be
borne in mind that these values are weighted averages
influenced by all processes, which retain the enantiomers
on the CSP and are minimum values of the true differ-
ences in energy.
NO2
O
H
N
R
H
N
R
3
4
2
O2N
O
1
Br
The chiral recognition model derived from chromato-
graphic and NMR studies is of value in anticipating
when resolutions can be expected, the order of elution,
and, to some degree, in providing rough estimates of
the separation factors which might be expected. How-
ever, it provides a first approximation of the relative
orientations of the selector and the more retained enan-
tiomer of the analyte. In order to gain a better under-
standing of the manner in which chiral selector 1
interacts with the enantiomers of an analyte, a more
detailed picture of the manner in which selector 1 inter-
acts with each enantiomer is needed.
(3R,4S)-CSP 1 R = (CH2)3Si(CH3)2-(SiO2)n
(3S,4S)-1 R = CH3
(R)-CSP 2 R = NH(CH2)3-(SiO2)n
(R)-2 R = t-Bu
N
O
Si
(SiO2)n
O
2
(S)-CSP 3
Figure 1. Chiral stationary phases and analytes used herein.
2.2. Homochiral complex
2. Results and discussion
The (3R,4R)-enantiomer of selector 1 and the (R)-enan-
tiomer of 2 crystallized as a methylene chloride solvate
in a 2:2:1 ratio, respectively. The complexed pairs are gen-
erally well separated from other interacting pairs in the
solid state with only one notable exception (vide infra).
In the solid state there are actually two selector–analyte
pairs in the unit cell. The measurements (bond lengths,
angles, etc.) for the two pairs are similar but not identi-
cal. Both utilize the same intermolecular interactions in
the complex formation. Measurements are quoted for
both pairs and in a consistent order.
2.1. Chromatographic chiral recognition
From analysis of the proposed chiral recognition model
for selector 1, (3R,4S)-CSP 1 was expected to preferen-
tially retain the (S)-enantiomer of 2. Indeed this was
the case (Table 1). Additionally, a chiral stationary
phase derived from (R)-p-bromo-a-phenylethylamine
(CSP 2) preferentially retains the (3R,4R)-enantiomer
of 1 consistent with the proposed model. In an effort
to avoid confusion, it should be pointed out that the
(3R,4S) CSP 1 and (3S,4S)-1 possess the same relative
stereochemistry about the partially saturated ring of
the selector—in both cases the substituent at the 3-posi-
tion is cis to the dinitrobenzamide moiety. There is a
Depictions of the two components illustrating the con-
formations they adopt in the 1:1 co-crystal, along with
the numbering system employed, are presented in Figure
2. Before considering the structure of the complex, it is
instructive to consider the conformation of each compo-
nent separately. The saturated six-membered ring of
selector 1 adopts a half-chair conformation placing the
methyl group in a pseudo-equatorial position and the
dinitrobenzamide moiety in a pseudo-axial position.
The pseudo-axial disposition of the dinitrobenzamide
moiety sets up the cleft between it and the naphthyl ring,
which is thought to be essential for effective chiral recog-
nition with this selector. The aryl portion of the dinit-
robenzamide moiety shows a slight deviation from
planarity with respect to the amide portion [dihedral
Table 1. Chromatographic data
2 on (3R,4S) CSP 1 1 on (R) CSP 2
Mobile phase
40% Isopropanol/
hexanes
0.59
20% Isopropanol/
hexanes
0.92
Retention factor 1: k1
Retention factor 2: k2
Separation factor: a
More retained
6.91
11.63
(S)
5.85
6.36
(3R,4R)
Conducted at ambient temperature with a nominal flow rate of 2 mL/
min.

M. E. Koscho et al. / Tetrahedron: Asymmetry 16 (2005) 3147–3153
3149
amide nitrogen of 1 and the carbonyl oxygen of 2 is sug-
gestive of a hydrogen bonding interaction (Table 2).
Figure 3. Depiction of the 1:1 homochiral complex in the solid state.
Shown as a ball-and-stick diagram for clarity.
Figure 2. ORTEP diagrams of the components of the homochiral
complex, with numbering. Thirty five percent probability thermal
ellipsoids are shown for non-hydrogen atoms and circles of arbitrary
size for hydrogen atoms.
angles: O5–C16–C17–C18 À16.8ꢁ (13.5ꢁ); N1–C16–
C17–C18 163.8ꢁ (166.6ꢁ)]. This deviation from planarity
directs the benzamide proton toward the cleft of the
selector and is probably a consequence of hydrogen
bonding to 2.
The amide portion of 2 is nearly planar and populates
the (Z)-rotomer about the amide nitrogen–carbon bond
[dihedral angles: O6–C30–N4–C29 À1.0ꢁ (À2.5ꢁ); C31–
C30–N4–C29 179.5ꢁ (177.0ꢁ)]. The methyl group
attached to the stereocenter of 2 is nearly perpendicular
to the plane of the aryl ring, thus leaving the nitrogen
(and the methine hydrogen) ca. 30ꢁ out of the plane of
the aryl ring (dihedral angles: C35–C29–C26–C27
À91.1ꢁ (À94.9ꢁ); N4–C29–C26–C27 31.2ꢁ (31.5ꢁ)).
A depiction of the 1:1 complex is shown in Figure 3. The
aryl portion of the dinitrobenzamide moiety of 1
and the aryl ring of 2 are nearly parallel, with a 3.3ꢁ
(8.1ꢁ) angle between mean planes and a separation of
˚
˚
3.6 A (3.6 A). Projections of the complex orthogonal
to both aryl rings of 1 are shown in Figure 4. The aryl
portion of the dinitrobenzamide moiety of 1 and the aryl
portion of 2 are offset from one another, as is common
to many observed face-to-face p–p interactions between
aryl rings. Additionally, the close approach of the benz-
Figure 4. Projections of the 1:1 homochiral complex orthogonal to the
dinitrobenzoyl aryl portion of 1 and the naphthyl portion of 1,
illustrating the geometry of the p–p interactions.

3150
M. E. Koscho et al. / Tetrahedron: Asymmetry 16 (2005) 3147–3153
Table 2. Hydrogen bonding detailsa
Homochiral
Heterochiral
Intra-complex
Inter-complex
Intra-complex
Inter-complex
N–H bond length
N–O distance
H–O distance
N–H–O angle
0.74 (0.65)
2.90 (2.86)
2.16 (2.26)
168ꢁ (153ꢁ)
0.98 (0.71)
3.02 (2.98)
2.12 (2.35)
153ꢁ (152ꢁ)
0.760
2.916
2.174
166ꢁ
0.841
3.095
2.284
162ꢁ
a Distances are in angstroms.
In addition to the face-to-face p–p and hydrogen bond-
ing interactions, the close approach of the aryl group of
2 to the naphthyl portion of selector 1 is suggestive of
additional stabilizing face-to-edge p–p interaction and
possibly indicates a weak hydrogen bonding interaction
of the methine hydrogen of 2 to the electron-rich naph-
thyl portion of the selector. The angle between the
planes of the naphthyl ring of 1 and the aryl ring of 2
is near perpendicular, 80.1ꢁ (78.0ꢁ). The distance from
the center of the aryl ring of 2 to the naphthyl ring of
the numbering system are shown in Figure 5. The satu-
rated six-membered ring of selector 1 adopts a half-chair
conformation placing the methyl group in a pseudo-
equatorial position and the dinitrobenzamide moiety
in a pseudo-axial position. The aryl portion of the dinit-
robenzamide shows a rather significant deviation from
planarity with respect to the amide portion (dihedral
angles: O5–C16–C17–C22 À32.2ꢁ; N1–C16–C17–C22
148.2ꢁ). This deviation from planarity directs the benz-
amide proton away from the cleft of the selector and
is probably a consequence of hydrogen bonding to 2.
˚
˚
1 is 5.4 A (4.9 A), while the distance of C25 from the
˚
˚
naphthyl ring is 4.1 A (3.6 A). Additionally, the distance
from the carbon at the stereogenic center of 2 (C29) to
˚
˚
the naphthyl ring is 3.6 A (3.7 A).
In addition to the aforementioned interactions, all of
which are consistent with the model for chiral recogni-
tion by selector 1, there is one additional inter-complex
interaction present in the solid state. The close contact
of the amide nitrogen of 2 and the benzamide oxygen
of selector 1 in the adjacent complex in the unit cell is
suggestive of a hydrogen bonding interaction (Table
2). This Ôone-pointÕ interaction likely aids in the crystal-
lization, but is not thought to be important in chiral rec-
ognition in solution, although it could very well occur as
a minor achiral interaction between selectors 1 and 2.
While it has previously been shown that removal of
extraneous sites for achiral interaction generally leads
to greater enantiodifferentiation by a chiral selector,9
in this case, modification of either of these sites would
undoubtedly lead to other undesired consequences.
2.3. Heterochiral complex
The (3S,4S)-enantiomer of selector 1 and the (R)-enan-
tiomer of 2 are crystallized in a 1:1 ratio. It should be
noted that the absolute configuration of the components
is irrelevant; it is the relative configuration of 1 and 2
that is important. The crystallographic data for the com-
plex between (3S,4S)-1 and (R)-2 is merely the mirror
image of the complex between (3R,4R)-1 and (S)-2, both
being diastereomeric to the homochiral complex. The
complexed pairs are generally well separated from other
interacting pairs in the solid state with the exception of
one interaction. As with the homochiral complex, there
is one additional Ôinter-complexÕ interaction. The close
contact of the amide nitrogen of 2 and the benzamide
oxygen of selector 1 is suggestive of a hydrogen bonding
interaction (Table 2).
Figure 5. ORTEP diagrams of the components of the heterochiral
complex, with numbering. Thirty five percent probability thermal
ellipsoids are shown for non-hydrogen atoms and circles of arbitrary
size for hydrogen atoms.
Depictions of the two components illustrating the con-
formation they adopt in the 1:1 co-crystal, along with
The amide portion of 2 is nearly planar and populates
the (Z)-rotomer about the amide nitrogen–carbon bond

M. E. Koscho et al. / Tetrahedron: Asymmetry 16 (2005) 3147–3153
3151
Figure 6. Depiction of the 1:1 heterochiral complex in the solid state.
Shown as a ball-and-stick diagram for clarity.
(dihedral angles: O6–C30–N4–C29 À4.3ꢁ; C31–C30–
N4–C29 174.4ꢁ). The methyl group at the stereocenter
of 2 is nearly perpendicular to the plane of the aryl ring
with the nitrogen ca. 43ꢁ out of the plane of the aryl ring
(dihedral angles: C35–C29–C26–C27 À80.8ꢁ; N4–C29–
C26–C27 43.3ꢁ).
A depiction of the 1:1 complex is shown in Figure 6. The
aryl ring of the dinitrobenzamide moiety of selector 1
and the aryl ring of 2 are in close proximity, suggestive
of a face-to-face p–p interaction. The angle between the
mean planes of the two rings is 15.4ꢁ with a separation
˚
of 3.6 A. A projection of the complex orthogonal to
Figure 7. Projection of the 1:1 heterochiral complex orthogonal to the
dinitrobenzoyl aryl portion of 1 showing the offset geometry of the
face-to-face p–p interaction, and orthogonal to the naphthyl portion of
1.
the aryl ring of 2 is shown in Figure 7. The two rings
are offset from one another, as is commonly observed
for many face-to-face p–p interactions between aryl
rings. Additionally, the close approach of the benzamide
nitrogen of 1 and the carboxyl oxygen of 2 is suggestive
of a hydrogen bonding interaction (Table 2).
Additionally, the methyl group of 2 would be directed
toward the naphthyl ring of selector 1, something which
would tend to sterically destabilize this mode of
complexation.
2.4. Comparison of the solid state structures
For the heterochiral diastereomer, the aromatic portion
of compound 2 is not in the cleft of selector 1. This is
not unexpected from the chiral recognition model for
selector 1. It is evident that in order for the (R)-enantio-
mer of 2 to undergo simultaneous face-to-face p–p inter-
action as well as a hydrogen bonding interaction in the
cleft of the (3S,4S)-enantiomer of selector 1, compound
2 must undergo a conformational change. This confor-
mational change would place the methyl group of 2 near
the plane of the aryl ring and the methine hydrogen near
perpendicular to the aryl plane. This is expected to be
an energetically much less favorable conformation.
3. Conclusions
The two diastereomeric complexes between chiral selec-
tor 1 and compound 2 have been characterized in the
solid state. Both display a face-to-face p–p interaction
between the dinitrobenzoyl aromatic ring of 1 and the
p-bromophenyl ring of 2, as well as a hydrogen bonding
interaction between the benzamide proton of 1 and the
carbonyl oxygen of 2. The relative orientations of 2 with
respect to chiral selector 1 are different for the two

3152
M. E. Koscho et al. / Tetrahedron: Asymmetry 16 (2005) 3147–3153
diastereomeric complexes. The more stable homochiral
diastereomeric pair (as judged from chromatographic
elution order) interacts in such a way that the aryl
moiety of 2 is in the cleft of selector 1. The spatial
complementarity of these two allows complexation in
this cleft without a substantial deviation from a low
energy conformation for either component. Addition-
ally, while in the cleft, the total energy of the complex
is lowered by additional bonding interactions with the
p-electron rich naphthyl ring of selector 1. The less
stable heterochiral diastereomeric pair interacts in a
manner such that the aryl portion of 2 is not within
the cleft of selector 1. Rather, the aryl moiety of 2 is
alongside the Ôback faceÕ of the dinitrobenzamide moiety
and over the saturated portion of the tetrahydrophe-
nanthrene ring system of the selector. The saturated ring
of selector 1 is sterically more demanding than the
planar naphthyl portion of the selector and this tends
to reduce the stability of this complex. Additionally, this
Ôback faceÕ complex cannot enjoy any additional bond-
ing interactions with the naphthyl portion of the selector
as does occur in the homochiral complex.
4.2. Enantiomer resolutions
Selector 1 was resolved using a preparative chiral sta-
tionary phase derived from N,N-diallyl-(S)-naproxen25
(CSP 3: 900 · 25 mm) eluting with 25% THF in hexanes
with a flow rate of 35 mL/min. Injection of 700 mg of
racemate yielded 331 mg of the (3S,4S)-(À)-enantiomer
(>99% ee) followed by 312 mg of the (3R,4R)-(+)-enan-
tiomer (97.5% ee) in a single pass through the column.
Compound 2 (mp 120.0–122.0 ꢁC) was resolved using a
commercial version of (3R,4S)-CSP 1 (250 · 21 mm)
eluting with 40% THF in hexanes with a flow rate of
5 mL/min. Injection of 179 mg of racemate yielded
70 mg of the (R)-(+) enantiomer (mp 130.5–131.5 ꢁC,
>99% ee) followed by 72 mg of the (S)-(À) enantiomer
(mp 132.0–133.0 ꢁC, 98% ee).
4.3. Crystallization
4.3.1. Homochiral complex. The enantiomer of 1
(20 mg, 49 lmol), which is more retained on (S)-CSP 3
and the enantiomer of 2 (14 mg, 49 lmol) first eluted
from (3R,4S)-CSP 1, were dissolved in CH2Cl2 (1 mL).
Diethyl ether vapor was then allowed to diffuse into this
solution, leading to slow crystallization. After 4 days,
the remaining solvents were decanted and a suitable
crystal chosen for X-ray analysis (colorless crystals,
mp >200 ꢁC).
Herein, we have reported two different modes of chiral
recognition observed in the solid state for the diastereo-
meric complexes formed from each of the enantiomers
of chiral selector 1 and the (R)-enantiomer of 2. The
homochiral diastereomeric complex is consistent with
numerous chromatographic studies as well as NMR
chemical shift data of complexes as well as observed
intermolecular nuclear Overhauser enhancements. Addi-
tional evidence is needed to determine the importance of
the mode of chiral recognition displayed by the hetero-
chiral diastereomeric pair. Although, even without any
additional solution state evidence, this complex does
suggest possible modifications to CSP 1, which would
increase enantiodifferentiation by decreasing the extent
of binding in this manner; namely, restricting the analyte
approach from the Ôback faceÕ of the selector.
4.3.2. Heterochiral complex. The enantiomer of 1
(20 mg, 49 lmol) first eluted from (S)-CSP 3 and the
enantiomer of 2 (14 mg, 49 lmol) first eluted from
(3R,4S)-CSP 1 were dissolved in CH2Cl2 (1 mL) and
Et2O (3 mL). The solvents were allowed to slowly evap-
orate. After 21 days the solvent volume had decreased to
ca. 1 mL. The remaining solvents were decanted and a
suitable crystal chosen for X-ray analysis (light yellow
crystals, mp 186.5–187.5 ꢁC).
4.4. X-ray analysis
4. Experimental
4.1. General
A portion of the data crystal was mounted using epoxy
to a thin glass fiber. The data were collected on a Sie-
mens Platform diffractometer at 293 K. Crystal data
are given in Table 3. Structure solution and refinement
were carried out with the use of the SHELXTL family of
programs. Hydrogens thought to undergo hydrogen
bonding interactions were independently refined (H1
and H4). Methyl hydrogen positions were optimized
by rotation about R–C bonds with idealized C–H, R–
H and H–H distances. Remaining hydrogen atoms were
included as fixed idealized contributors. Hydrogen atom
U values were assigned as 1.2 times Ueq of adjacent non-
hydrogen atoms. The maximum shift/error for the final
Analytical chromatography was carried out with a com-
mercial version of (3R,4S)-CSP 1 [250 · 4.6 mm, avail-
able from Regis Technologies under the name (S,S)-
Whelk-O1]. (R)-CSP 2 (250 · 4.6 mm) was available
from a previous study. Compound 1 was prepared as
previously reported.3 Compound 2 was obtained by
acylation of p-bromo-a-phenylethylamine with pivaloyl
chloride as previously reported.11 HPLC grade solvents
were obtained from EM Science. Melting points are
uncorrected. The terms homochiral and heterochiral
are meant only to relate the relative configuration of 1
and 2. The term homochiral is used for the diastereo-
meric combination, in which the Cahn–Ingold–Prelog
stereochemical descriptor is the same for both 1 and 2
[i.e., (R,R)-1 and (R)-2]. Heterochiral refers to the dia-
stereomeric combination in which the stereochemical
descriptors are not the same [i.e., (R,R)-1 and (S)-2].
Crystallographic data (excluding structure factors) for the structures
herein have been deposited with the Cambridge Crystallographic
Data Centre as supplementary publication numbers CCDC 273851
(heterochiral complex), and CCDC 273852 (homochiral complex).
Copies of the data can be obtained, free of charge, on application to
CCDC, 12 Union Road, Cambridge CB2 1EZ, UK [fax: +44 (0)
1223 336033 or e-mail: deposit@ccdc.cam.ac.uk].

M. E. Koscho et al. / Tetrahedron: Asymmetry 16 (2005) 3147–3153
3153
cycle of full-matrix least squares refinement on F2 indi-
cated successful convergence. A final analysis of the
goodness of fit between observed and calculated struc-
ture factors showed no dependence on amplitude or res-
olution. Analysis of the absolute structure parameter
indicated the absolute configuration to be (3R,4R)-1
and (R)-2 for the homochiral diastereomer and
(3S,4S)-1 and (R)-2 for the heterochiral diastereomer.
4. Pirkle, W. H.; Welch, C. J. Tetrahedron: Asymmetry 1994,
5, 777–780.
5. Terfloth, G. J.; Pirkle, W. H.; Lynman, K. G.; Nicolas,
E. C. J. Chromatogr. A 1995, 705, 185–194.
6. Pirkle, W. H.; Brice, J. L.; Terfloth, G. J. J. Chromatogr. A
1996, 753, 109–119.
7. Welch, C. J.; Szczerba, T.; Perrin, S. R. J. Chromatogr. A
1997, 758, 93–98.
8. Pirkle, W. H.; Welch, C. J.; Zych, A. J. J. Chromatogr.
1993, 648, 101–109.
9. Villani, C.; Pirkle, W. H. Tetrahedron: Asymmetry 1995, 6,
27–30.
Table 3. Crystallographic data
10. Villani, C.; Pirkle, W. H. J. Chromatogr. A 1995, 693, 63–
68.
Homochiral
crystal
Heterochiral
crystal
11. Pirkle, W. H.; Gan, K. Z.; Brice, L. J. Tetrahedron:
Asymmetry 1996, 7, 2813–2816.
12. Pirkle, W. H.; Brice, L. J.; Caccamese, S.; Principato, G.;
Failla, S. J. Chromatogr. A 1996, 721, 241–246.
13. Pirkle, W. H.; Koscho, M. E.; Wu, Z. J. Chromatogr. A
1996, 726, 91–97.
14. Pirkle, W. H.; Koscho, M. E. J. Chromatogr. A 1997, 761,
65–70.
15. Pirkle, W. H.; Spence, P. L. J. Chromatogr. A 1997, 775,
81–90.
16. Pirkle, W. H.; Gan, K. Z. J. Chromatogr. A 1997, 790, 65–
71.
17. Wolf, C.; Pranatharthiharan, L.; Volpe, E. C. J. Org.
Chem. 2003, 68, 3287–3290.
18. Pirkle, W. H.; Welch, C. J. J. Chromatogr. A 1994, 683,
347–353.
Formula
C
732.06
35.5H38BrClN4O6
C35H37BrN4O6
689.60
Orthorhombic
P212121
8.884(2)
19.471(4)
20.263(4)
4
1.54178 (Cu Ka)
0.0411
Formula weight
Crystal system
Space group
Orthorhombic
P212121
18.192(4)
19.186(4)
21.085(4)
8
1.54178 (Cu Ka)
0.0810
0.2324
˚
a, A
˚
b, A
˚
c, A
Z
˚
k, A
R1
wR2
0.1175
1.067
À0.02(2)
Goodness of fit
Absolute structure
parameter
1.076
0.01(4)
19. Pirkle, W. H.; Selness, S. R. J. Org. Chem. 1995, 60, 3252–
3256.
Acknowledgments
20. Del Rio, A.; Mayes, J. M.; Stein, M.; Piras, P.; Roussel, C.
Chirality 2004, 16, S1–S11.
21. Pirkle, W. H.; Burke, J. A.; Wilson, S. R. J. Am. Chem.
Soc. 1989, 111, 9222–9223.
This work has been funded by grants from the National
Science Foundation and from Eli Lilly and Company.
22. Dappen, R.; Rihs, G.; Mayer, C. W. Chirality 1990, 2,
185–189.
References
23. Pirkle, W. H.; Murray, P. G.; Wilson, S. R. J. Org. Chem.
1996, 61, 4775–4777.
1. Pirkle, W. H.; Welch, C. J.; Lamm, B. J. Org. Chem. 1992,
24. Okamura, K.; Aoe, K.-I.; Nishimura, N.; Fujiwara, Y.;
Sumida, Y.; Hashimoto, K. Enantiomer 1997, 2, 99–
104.
57, 3854–3860.
2. Pirkle, W. H.; Welch, C. J. J. Liq. Chromatogr. 1992, 15,
1947–1955.
25. Pirkle, W. H.; Spence, P. L.; Lamm, B.; Welch, C. J.
J. Chromatogr. A 1994, 659, 69–74.
3. Pirkle, W. H.; Welch, C. J.; Wilson, S. R. Chirality 1994,
6, 615–622.