Non-superimposable mirror image crystals of enantiomers by spontaneous resolution and the chiral discrimination mechanism

Article abstract of DOI:10.1016/j.cclet.2013.04.045

Non-superimposable mirror image crystals of both enantiomers (S/R) of cyclic γ-alkenyl alcohol (2) have been recognized and remarkably identified by the naked eye. More interestingly, both crystals are an outcome of most astonishingly H-bond and intermole

Full text of DOI:10.1016/j.cclet.2013.04.045

Chinese Chemical Letters 24 (2013) 695–698
Contents lists available at SciVerse ScienceDirect
Chinese Chemical Letters
journal homepage: www.elsevier.com/locate/cclet
Original article
Non-superimposable mirror image crystals of enantiomers by spontaneous
resolution and the chiral discrimination mechanism
*
Muhammad Sohail, Yao-Feng Wang, Shao-Xiang Wu, Wei Zeng, Ji-Yi Guo, Fu-Xue Chen
School of Chemical Engineering and the Environment, Beijing Institute of Technology, Beijing 100081, China
A R T I C L E I N F O
A B S T R A C T
Article history:
Non-superimposable mirror image crystals of both enantiomers (S/R) of cyclic
g-alkenyl alcohol (2) have
Received 2 March 2013
Received in revised form 25 March 2013
Accepted 12 April 2013
Available online 5 June 2013
been recognized and remarkably identified by the naked eye. More interestingly, both crystals are an
outcome of most astonishingly H-bond and intermolecular interactions. They accounted for the
s/p–p
relatively rare and less predictable spontaneous resolution with optical purity >99% ee from the racemic
mixture. The chiral discrimination mechanism of this spontaneous resolution has also been proposed.
ß 2013 Fu-Xue Chen. Published by Elsevier B.V. on behalf of Chinese Chemical Society. All rights
reserved.
Keywords:
Spontaneous resolution
Preferential crystallization
Mirror images
Octahydrobenzofuran
1. Introduction
recognition by the naked eye of both enantiomers (S,R) of cyclic g-
alkenyl alcohol (2) from solution and separation on the basis of
Asymmetric synthesis and chiral resolution are the two major
tactics to achieve desired chiral compounds [1]. Both strategies
always involve the use of discrete enantiomers, except with
spontaneous resolution [2]. However, this phenomenon is very
rare, poorly documented [3] and generally involves tedious
preferential crystallization [4]. Homochiral crystallization is the
simplest way to prepare enantiomers directly from the racemic
mixture [5]. Should spontaneous resolution happen, it is a more
convenient and economical approach [6], receiving huge interest.
However, it is less predictable as the conglomeration processes
have not yet been explored comprehensively [7]. It is well-known
that supramolecular interactions, such as hydrogen bonding and
coordination bonding have specific directions that can sometimes
effectively transmit stereochemical information between adjacent
homochiral molecules and play an imperative role in spontaneous
resolution [8]. In 1848, Louis Pasteur, discovered the first, simplest
separation of enantiomers by homochiral crystallization of a
conglomerate. Two crystals were visually distinctive as hemihe-
dral forms [9] and this uncommon and rare wonder was selected as
one of the most memorable discoveries in chemistry by readers
[10]. A few years ago, Lahav described a method for the direct
determination of absolute configuration of chiral polar crystals
based on the changes in crystal habit induced by tailor-made
impurities [11]. In the history of chemistry and after the historic
achievement of Pasteur, it seems miraculous that, herein, the
physical appearance as a mirror image, has been achieved.
2. Experimental
We have targeted the chiral center at C-10 position of cyclic
g-
alkenyl ester 1 and alcohol 2 because of its importance in the
syntheses of enantiopure octahydrobenzofuran derivatives. In this
regard, the racemic 2 was synthesized (Scheme 1) using the
literature procedure with modifications [12].
2.1. Methyl 2-(cyclohex-2-enyl)-2,2-diphenylacetate (1)
Lithium di-isopropylamide (LDA) was generated in situ by slow
addition of n-BuLi (1.65 mL, 1.6 mol/L, 2.65 mmol, 1.3 equiv.) to
the solution of di-isopropylamine (371 mg, 2.87 mmol, 1.2 equiv.)
in THF (4 mL) and stirred for 0.5 h at ꢀ78 8C. Methyl diphenyla-
cetate (500 mg, 2.21 mmol) in THF (5 mL) was added slowly over
1 h at the same temperature and stirred for an additional 1 h. 3-
Bromocyclohexene (391.3 mg, 2.431 mmol, 1.1 equiv.) was added
slowly to the reaction mixture, allowed to warm up at room
temperature freely and stirred for 24 h. Quenched with aqueous
HCl (5 mL, 1 mol/L), stirred and partitioned between H2O (10 mL)
and CH2Cl2 (3 ꢁ20 mL), the organic layer was combined, dried over
anhydrous Na2SO4, filtered, concentrated under reduced pressure
and purified by column chromatography with EtOAc/n-hexane (1/
20, v/v) afforded 1 as a clear colorless oil (500 mg, 86%) which was
crystallized from n-hexane. 1H NMR (400 MHz, CDCl3):
d 7.29–7.26
(m, 10 H, 2Ph-H), 5.67–5.62 (m, 2H, 20,30-H), 3.81 (m, 1H, 2-H), 3.62
(s, 3H, Me), 1.88–1.01 (m, 6H, 40,50,60-H).
* Corresponding author.
E-mail address: fuxue.chen@bit.edu.cn (F.-X. Chen).
1001-8417/$ – see front matter ß 2013 Fu-Xue Chen. Published by Elsevier B.V. on behalf of Chinese Chemical Society. All rights reserved.
http://dx.doi.org/10.1016/j.cclet.2013.04.045

[
M. Sohail et al. / Chinese Chemical
$
Letters 24 (2013) 695–698
Br
Ph
LDA
LiAlH
4
OMe
+
Ph
Ph
Ph
OMe
Ph
Ph
OH
O
O
(
)-2
1
Scheme 1. Synthetic route of alcohol 2.
2.2. 2-(Cyclohex-2-enyl)-2,2-diphenylethanol (2)
The solution of methyl ester 1 (891 mg, 2.30 mmol) in THF
(15 mL) was added dropwise to a stirred suspension of LiAlH4
(122 mg, 3.22 mmol, 1.4 equiv.) in THF (10 mL) at 0 8C. The reaction
mixture was stirred at room temperature for 24 h. Quenched with
NaOH (3 mL, 1 mol/L) and H2O (5 mL) at 0 8C, the suspension was
filtered through celite and washed with Et2O (30 mL). The filtrate
and organic layer were combined, dried over anhydrous Na2SO4,
filtered, concentrated under reduced pressure, and purified by
column chromatography. The elution of 4% EtOAc in n-hexane
afforded 2 as a white solid (500 mg, 86%), mp 126–128 8C (from n-
hexane); 1H NMR (400 MHz, CDCl3):
d 7.23–7.21 (m, 10H, 2Ph-H),
5.81 (d, 1H, J = 10.4 Hz, 20-H), 5.68–5.64 (m, 1H, 30-H), 4.2 (s, 2H, 1-
H), 3.3–3.2 (m, 1H, 10-H), 1.9–1.0 (m, 6H, 40,50,60-H). Resolution was
accomplishedandbothenantiomerswererecognizedbychiralHPLC
on Chiralpak AD-H (n-hexane/i-propanol = 99.9/0.1, v/v, 1.0 mL/
min, 254 nm), tS = 16.9 min, tR = 18.3 min; ½a D28
ꢀ1.9 (c 0.254,
ꢂ
CH2Cl2, for (S)-2), ½a D28
ꢂ
+1.6 (c 0.254, CH2Cl2, for (R)-2).
Fig. 1. Resolution of (ꢃ)-2. Step 1: spontaneous resolution; Step 2: preferential
2.3. 7-Bromo-3,3-diphenyl-octahydrobenzofuran (3)
crystallization; and Step 3: simple recrystallization.
N-Bromosuccinimide (NBS, 10.5 mg, 0.06 mmol, 1.2 equiv.) was
added to the solution of alcohol (+)-2 (13.9 mg, 0.05 mmol,
1.2 equiv.) in CH2Cl2 (0.5 mL) at ꢀ78 8C (Scheme 2). The reaction
mixture was stirred at ꢀ78 8C for 2.5 h. Upon completion as
monitored by thin layer chromatography, the crude product was
directly loaded on column and purified by flash silica gel column
chromatography (Et2O/petroleum ether (1/40, v/v) afforded 3,
white crystals (15 mg, 84%). Mp 87–89 8C (from n-Hexane), 1H-
spontaneous resolution occurred and yielded bona fide crystal
flowers. Upon further scrutinizing, each crystal flower consisted
of numerous well defined crystal petals with purity ranging from
50% ee to 89% ee (Fig. 1, Step 1). Thus, these crystal petals were
separated and used as seed crystals in the subsequent preferential
crystallization (Fig. 1, Step 2). With these rectangular shape
crystal seeds, more crystal flowers were grown with the same
absolute configuration as the seed from a hot (50 8C), diluted
solution of racemic 2 in n-hexane (resolution efficiency [13],
E = 45.9). The crystals of the other enantiomer were obtained
either from mother liquor (E = 84.3), or by the same procedure,
using seed crystals with the opposite handedness (Fig. 1, Step 2).
Further, simple recrystallization (Fig. 1, Step 3) was carried out 2
or 3 times leading to successful resolution of both enantiomers of
2 in >99.9% ee. The absolute configuration was established based
on the single crystal of halogenation derivative (+)-(3aR,7aS,7S)-3
(Scheme 2) [14].
NMR (600 MHz, CDCl3):
d 7.37 (d, 2H, J = 7.5 Hz, Ph), 7.29 (d, 2H,
J = 7.5 Hz, Ph), 7.27–7.23 (m, 2H, Ph), 7.17–7.13 (m, 4H, Ph), 4.73 (d,
1H, J = 8.2 Hz, 2-Ha), 4.65 (d, 1H, J = 8.2 Hz, 2-Hb), 4.42 (t, 1H,
J = 4.1 Hz, 7-H), 4.38 (dd, 1H, J = 7.5 Hz, 4.14 Hz, 7a-H), 3.31–3.27
(m, 1H, 3a-H), 1.95–1.93 (m, 1H, 6a-H), 1.89–1.88 (m, 1H, 6b-H),
1.73–1.69 (m, 2H, 5-H), 1.32–1.29 (m, 1H, 4a-H), 1.14–1.12 (m, 1H,
4b-H). 13C NMR (100 MHz, CDCl3):
d 145.7, 143.3, 128.8, 128.4,
128.2, 126.8, 126.5, 126.3, 81.2, 76.1, 58.8, 52.3, 41.5, 29.4, 24.9,
19.7. HRMS calcd. for C20H21BrO: 379.0668 [M+Na]+, found:
379.0678. ½a 2D8
ꢂ
+133 (c 0.218, CH2Cl2 for (3aR,7aS,7S)-3).
3. Results and discussion
Scrutinizing the crystal flowers provided a foundation to
develop a highly efficient and clean spontaneous resolution.
Further optimization was attempted to get pure crystals or crystal
flowers of each enantiomer by spontaneous resolution directly
from a racemic solution. Low temperature (ꢀ15 8C) gave the
crystal of low purity. Interestingly, by changing the volume and
polarity of the solvent, slow crystallization affected the ratio of
crystals and crystal flowers, as well as the size and optical purity
(see Table S1 in Supporting information,). Finally, after great
laborious work, crystal flowers consisting of well-defined crystal
petals of each enantiomer were separated with a purity >99% ee
from the solution of 2 (300 mg) consisting of a mixture of EtOAc/n-
hexane (33 mL, 1/10, v/v) (Fig. 2). All these crystals were identified
by the naked eye (S/R) and were separated on the basis of physical
appearance as a mirror image (Fig. 3, also see Fig. S2 in Supporting
information).
3.1. The spontaneous resolution
Through free evaporation of the standing solution of 2 (700 mg)
iS n mixed ethyl acetate and n-hexane (50 mL, 1:40, v/v) at 25 8C,
[(chme_2)TD$FIG]
Scheme 2. Synthetic route of 3.

[
M. Sohail et al. / Chinese Chemical
T
Letters 24 (2013) 695–698
697
Fig. 2. Graphic representation of spontaneous resolution of (ꢃ)-2.
Fig. 4. Molecular packing in unit cell of (ꢀ)-2.
G
[(ig._5)TD$FIG]
Fig. 5.
s/p
–p
and allylic H-bond interactions. (a) (ꢀ)-2 and (b) (+)-2.
[i(g._6)TD$FIG]
Fig. 3. SEM mirror images and structure of 2. (a) (S)-(ꢀ)-2 and (b) (R)-(+)-2.
3.2. The mechanism of the spontaneous resolution
Fig. 6. s–p and H-bond interactions in 1.
Once a clean spontaneous resolution procedure was estab-
lished, which interestingly led to non-superimposable mirror
image crystals, the discrimination mechanism of this spontaneous
resolution was elucidated by the non-covalent interactions during
the crystallization. As we know, the characteristics of solid-state
materials are inherently linked to the order in which the atoms are
arranged [15]. During crystallization, molecules are arranged in an
orderly manner and packed as tightly as possible through H-bond
chiral resolution. The good quality single crystal of (ꢀ)-2 was
obtained from n-hexane for X-ray single crystallographic analysis
[17]. The colorless, hexagonal crystal belongs to the monoclinic
space group P2(1) and the unit cell contains two molecules (Fig. 4).
Crystallographic lattice parameters of all related compounds are
summarized in Table 1. Scrutiny of these single crystal data helps
us to elucidate the mechanism of this spontaneous resolution.
network,
p
–p
stacking, van der Waals interaction, and supramo-
First, p–p stacking and special hydrogen bonds are recognized,
lecular packing mode to realize the minimum energy in the crystal
lattice [16]. These numerous weak intermolecular interactions
between homo- and hetero-chiral enantiomers are important for
in the chiral discrimination [18] and for growth of both (ꢀ)-2 and
˚
(+)-2 single crystals. The distance of about 7.65 A clearly
demonstrates the absence of ordinary hydrogen bonds between
Table 1
Single crystal structure data of (ꢃ)-1, (ꢀ)-2 and (+)-2.
(ꢀ)-2
(+)-2
(
ꢃ)-1
Empirical formula
Formula weight
C20H22
O
C20H22
O
C21H22O2
306.39
278.38
278.38
Temperature
153(2) K
143(2) K
153 (2) K
˚
˚
˚
Wavelength
0.71073 A
0.71073 A
0.71073 A
Crystal system, space group
Unit cell dimensions
Monoclinic, P2(1)
Monoclinic, (1)
Triclinic, P-1
˚
˚
˚
˚
a = 8.947(2) A, alpha = 908
a = 8.981(3) A, alpha = 908
a = 8.866(3) A, alpha = 96.248(2)8 b = 9.638(3) A,
˚
˚
˚
b = 7.1033(19) A, beta = 95.199(4)8
b = 7.106(2) A, beta = 95.604(5)8
beta = 102.875(4)8 c =11.666(4) A, gamma = 116.890(4)8
˚
˚
c = 11.670(3) A, gamma = 908
c = 11.773(4) A, gamma = 908
3
3
3
˚
˚
˚
Volume
738.6(3) A
747.7(4 A
841.2(5) A
Z, calculated density
Absorption coefficient
Crystal size
2, 1.252 Mg/m3
2, 1.236 Mg/m3
2, 1.210 Mg/m3
0.076 mm–1
0.075 mm–1
0.074 mm–1
0.40 ꢁ 0.38 ꢁ 0.36 mm
7328/2611 [R(int) = 0.0251]
31.50 99.0%
0.19 ꢁ 0.19 ꢁ 0.16 mm
6325/2145 [R(int) = 0.0316]
29.15 98.6%
0.40 ꢁ 0.39 ꢁ 0.35 mm
9589/3808 [R(int) = 0.0261]
27.50 98.4%
Reflections collected/unique
Completeness to theta
Final R indices [I > 2 sigma (I)]
R indices (all data)
R1 = 0.0357, wR2 = 0.0853
R1 = 0.0375, wR2 = 0.0865
R1 = 0.0545, wR2 = 0.1275
R1 = 0.0684, wR2 = 0.1383
R1 = 0.0591, wR2 = 0.1417
R1 = 0.0777, wR2 = 0.1549

[
M. Sohail et al. / Chinese Chemical Letters 24 (2013) 695–698
Fig. 7.
s
/p
–p
and H-bond interactions between and within layers in (ꢀ)-2.
free OH groups. Contrary to our anticipation, in the single crystal of
Chinese Scholarship Council (M. Sohail) are acknowledged. We
thank Dr. Kai-Bei Yu at Beijing Institute of Technology for the single
crystal analysis.
(ꢀ)-2 (Fig. 5a), hydrogen (H10) of the hydroxyl group in one
molecule has strong interaction with
p electrons of two carbons
(C12 and C13) of the phenyl ring of the neighboring molecule at a
˚
˚
distance of 2.78 A and 2.86 A, respectively, which is not observed in
the single crystal of (+)-2 (Fig. 5b). And in turn, the phenyl ring via
Appendix A. Supplementary data
hydrogen (H12) of the latter molecule shows vertical
p–p
interaction with the phenyl ring of the former molecule (C15 and
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.cclet.2013.04.045.
˚
˚
C16) at a distance ranging from 2.83 A to 2.74 A (Fig. 5). Most
astonishingly, oxygen (O1) of the OH group exerts a strong
connection with allylic hydrogen (H3b) at a distance ranging from
References
˚
˚
2.54 A to 2.58 A.
Comparing with 2, spontaneous resolution of its precursor
methyl ester 1 failed under the same conditions, but yielded crystal
aggregates or crystals with optical purity about 2% ee or 4% ee.
Closely examining the single crystal structure of (ꢃ)-1 [17] reveals
that lone electron pair of the carbonyl oxygen (O2) connects to the
hydrogen (H17) attached to the phenyl ring of the opposite
[1] M. Gruttandauria, F. Giacalone, M. Raj, V.K. Singh, Catalytic Methods in Asym-
metric Synthesis, Wiley, Hoboken, NJ, 2011, pp. 1–82.
[2] H. Pellissier, Chirality from Dynamic Kinetic Resolution, Royal Society of Chem-
istry, Aix-Marseille, 2011, pp. 1–45.
[3] A. Lennartson, M. Hakansson, Total spontaneous resolution of five-coordinate
complexes, Angew. Chem. Int. Ed. 48 (2009) 5869–5871.
[4] (a) F. Yagishita, H. Ishikawa, T. Onuki, et al., Total spontaneous resolution by
deracemization of isoindolinones, Angew. Chem. Int. Ed. 51 (2012) 13023–13025;
(b) K. Harada, Total optical resolution of free a-amino acids by the inoculation
method, Nature 206 (1965) 1354–1355.
˚
enantiomeric molecule at a distance of 2.47 A, as well as the carbon
˚
(C17) to the hydrogen (H21) of methyl ester at a distance of 2.87 A
(Fig. 6). It indicates the directions of s/p–p interactions and an
uncommon H-bond involving allylic hydrogen (Fig. 5) and provides
the reason of the spontaneous resolution of (ꢃ)-2.
[5] A. Lennartson, A. Hedstrom, M. Hakansson, Toward total spontaneous resolution
of sec-butylzinc complexes, Organometallics 29 (2010) 177–183.
[6] C.J. Eckhardt, N.M. Peachey, D.R. Swanson, et al., Separation of chiral phases
in monolayer crystals of racemic amphiphiles, Nature 362 (1993)
614–616.
[7] E. D’Oria, P.G. Karamertzanis, S.L. Price, Spontaneous resolution of enantiomers by
crystallization, Cryst. Growth Des. 10 (2010) 1749–1756.
[8] R. Fasel, M. Parschaun, K.H. Ernst, Amplification of chirality in two-dimensional
enantiomorphous lattices, Nature 439 (2006) 449–452.
Further, in the single crystal of (ꢀ)-2, layers of molecules are
well stacked over each other and stabilized by vertical
s/p–p
interaction [19]. In addition, as shown in Fig. 7, the series of layers
are packed by an additional interaction between H5A and C10.
s–p
Close observation of these layers indicates the benzene ring carbon
(C10) has strong interaction with hydrogen C60-H5A at a distance
[9] (a) H.D. Flack, Louis Pasteur’s discovery of molecular chirality and spontaneous
resolution in 1848, together with a complete review of his crystallographic and
chemical work, Acta Cryst. A65 (2009) 371–389;
˚
of 2.76 A. Similarly the phenyl carbon of this molecule (C10)
(b) R.G. Kostyanovsky, Louis Pasteur did it for us especially Mendeleev Commun
Electronic Version, 2003, 1–6.
interacts with the sp3 hydrogen H5A of another molecule and thus
[10] T. Fumiao, Enantiomer Separation Fundamentals and Practical Methods, Spring-
er-Verlag, Berlin, 2011, pp. 167–168.
[11] Z.B. Yellin, L. Addadi, M. Idelson, L. Leiserwitz, M. Lahav, The absolute configura-
tion of chiral polar crystals, Nature 296 (1982) 27–34.
[12] R.E.M. Brooner, R.A. Widenhoefer, Stereochemistry and mechanism of the
Brønsted acid catalyzed intramolecular hydrofunctionalization of an unactivated
cyclic alkene, Chem. Eur. J. 17 (2011) 6170–6178.
[13] T. Fumiao, Enantiomer Separation Fundamentals and Practical Methods, Spring-
er-Verlag, Berlin, 2011, pp. 177–178.
extends the stacked layers (Fig. 7). Consequently, the concerted
s–
p
interactions within layers and network between layers
s/p–p
exist in the homochiral single crystal, which should be the most
expected reason of this spontaneous resolution.
4. Conclusion
In conclusion, after the discovery of Pasteur, macro enantio-
merically pure, non-superimposable mirror image crystals have
been recognized and efficient spontaneous resolution has been
[14] CCDC 927205 contains detail of the X-ray analysis of single crystal of (+)-
(3aR,7aS,7S)-3.
[15] S. Haq, N. Liu, V. Humblot, A.P.J. Jansen, R. Raval, Drastic symmetry breaking in
supramolecular organization of enantiomerically unbalanced monolayers at
surfaces, Nature Chem. 1 (2009) 409–414.
established as a result of intermolecular s/p–p stacking and allylic
H-bond interaction. This investigation afforded not only a new
example of spontaneous resolution, but also an excellent
preparative route for the synthesis of chiral heterocycles without
any external chiral sources.
[16] J.D. Dunitz, A. Gavezzotti, How molecules stick together in organic crystals: weak
intermolecular interactions, Chem. Soc. Rev. 38 (2009) 2622–6233.
[17] X-ray analysis of single crystal of 1, (S)-2 and (R)-2 were deposited at Cambridge
crystallographic data centre with CCDC 908421, 908683 and 915108,
respectively.
[18] Y. Kobayashi, H. Hiroaki, J. Maeda, Factors determining the pattern of a hydrogen-
bonding network in the diastereomeric salts of 1-arylethylamines with enantio-
pure P-chiral acids, Chirality 20 (2008) 577–584.
Acknowledgments
[19] F. Cherif, J.H.T. Matta, H.T. Ting, F.W.B. Richard, Hydrogen–hydrogen bonding: a
Financial support from Beijing Institute of Technology (No.
2011CX01008 and others), partially from NSFC (No. 20972016) and
stabilizing interaction in molecules and crystals, Chem. Eur. J.
1940–1951.
9 (2003)