Induction of cholesteric mesophases by simple cyclic derivatives of p,p'-disubstituted 1,2-diphenylethane-1,2-diols: Importance of shape and polarizability effects

Article abstract of DOI:10.1021/jo990038y

A systematic study of the cholesteric induction in nematic solvents (MBBA and E7) by some cyclic derivatives of unsubstituted and p,p'- disubstituted-1,2-diphenylethane-1,2-diols shows that the values of the twisting power are significantly dependent on t

Full text of DOI:10.1021/jo990038y

4762
J . Org. Chem. 1999, 64, 4762-4767
In d u ction of Ch olester ic Mesop h a ses by Sim p le Cyclic Der iva tives
of p,p′-Disu bstitu ted 1,2-Dip h en yleth a n e-1,2-d iols: Im p or ta n ce of
Sh a p e a n d P ola r iza bility Effects
Stefano Superchi,† Maria Irene Donnoli,† Gloria Proni,‡ Gian Piero Spada,*,‡ and
Carlo Rosini*,†
Dipartimento di Chimica, Universita` della Basilicata, Via N. Sauro 85, 85100 Potenza, Italy, and
Dipartimento di Chimica Organica “A. Mangini”, Universita` di Bologna, Via S. Donato 15,
40127 Bologna, Italy
Received J anuary 8, 1999
A systematic study of the cholesteric induction in nematic solvents (MBBA and E7) by some cyclic
derivatives of unsubstituted and p,p′-disubstituted-1,2-diphenylethane-1,2-diols shows that the
values of the twisting power are significantly dependent on the nature of the link connecting the
two oxygen atoms and on the nature of the p,p′-substituents. This result has been interpreted
considering that the nature of the bridge affects the overall molecular shape and the p,p′-substituents
affect both the molecular polarizability and shape. This investigation points out that the
polarizability of the solute and the solvent is the main parameter in determining the value of the
twisting power while electrostatic arene-arene interactions contribute to a less extent. It has been
also observed that solutes having the same structure and the same absolute configuration can
induce cholesteric helix of opposite sign depending on the substituent on the aromatic ring. This
finding indicates that configurational assignments by cholesteric induction are reliable only if high
values of twisting power are measured.
In tr od u ction
arene-arene interactions,6 a phenomenon that deeply
affects the structures and properties of molecular as-
semblies in biology,7 chemistry,8 and material science.9
We therefore undertook a systematic study aimed at
determining the influence of shape and electronic factors
onto the cholesteric induction. To this end, the skeleton
deriving from 1,2-diphenylethane-1,2-diol has been cho-
sen as chiral template. In fact, (i) 2,2-dimethyl-1,3-
dioxolanes derived from 1,2-diarylethane-1,2-diols are
efficient inducers of cholesteric mesophases;10 (ii) this
template can give rise to different cyclic, conformationally
blocked, derivatives where the shape effects (steric
hindrance and molecular size) due to the cycle can be
studied; and (iii) electron-donating and electron-with-
drawing groups can be easily introduced into the benzene
rings, allowing the study of the role played by the nature
of the substituent.
Induced cholesteric mesophases1 are interesting and
useful optically active materials on account of their
stereochemical1b,2 and technological3 (e.g., displays) ap-
plications. These applications involve induced cholesteric
mesophases with definite properties (sense of twist,
length of the pitch, etc.) that depend on the structure of
the dopant4 and on the strength of the solute/solvent
interactions. A deeper knowledge of the mechanism of
induction and of the interactions that determine it is then
essential to carry out reliable configurational assign-
ments2 and to design materials with suitable properties.
In addition, since both nematic hosts (e.g., MBBA, K15,
etc.) and chiral dopants often bear aromatic rings, a study
of the mechanism of cholesteric induction can provide5 a
useful contribution to the understanding of noncovalent
In this paper, compounds 2-6 (Chart 1) were prepared
from the corresponding diols 1a -e and their twisting
powers (âM) measured in the aromatic nematic solvents
MBBA (a benzylideneaniline derivative), and E7 (a
biphenyl mixture).
* To whom correspondence should be addressed (C.R.). Fax: +39-
0971202223. Tel:+39-0971202241. E-mail: rosini@unibas.it.
† Universita` della Basilicata.
‡ Universita` di Bologna.
(1) (a) A general introduction to liquid crystals, including chiral
mesophases, can be found in: Collings, P. J .; Hird, M. Introduction to
liquid crystalssChemistry and Physics; Taylor & Francis: London,
1997. (b) A more specific treatment of cholesteric systems is given by:
Solladie`, G.; Zimmermann, R. G. Angew. Chem., Int. Ed. Engl. 1984,
23, 348.
(2) Examples of stereochemical applications are reported in: (a)
Gottarelli, G.; Spada, G. P.; Solladie`, G. Nouv. J . Chem. 1986, 10, 692.
(b) Gottarelli, G.; Spada, G. P. Mol. Cryst. Liq. Cryst. 1985, 123, 37.
(c) Spada, G. P.; Proni, G. Enantiomer 1998, 3, 301.
(3) See, for instance: Scheuble, B. S. Kontakte 1989, 1, 34. Liquid
Crystals. Applications and Uses; Bahadur, B., Ed.; World Scientific:
Singapore, 1990 (Vol. 1); 1991 (Vol. 2); 1993 (Vol. 3).
(4) Highly efficient inducers have been recently described in: Park,
J . J .; Sternhell, S.; Vonwiller, S. C. J . Org. Chem. 1998, 63, 6749.
Kuball, H.-G.; Weiss, B.; Beck, A. K.; Seebach, D. Helv. Chim. Acta
1997, 80, 2507. Deussen, H.-J .; Shibaev, P. V.; Vinokur, R.; Bjornholm,
T.; Schaumburg, K.; Bechgaard, K.; Shibaev, V. P. Liq. Cryst. 1996,
21, 327.
(6) (a) Hunter, C. A. Angew. Chem., Int. Ed. Engl., 1993, 32, 1585.
(b) Hunter, C. A. Chem. Soc. Rev. 1994, 101. (c) Hunter, C. A.;
Saunders, J . K. M. J . Am. Chem. Soc. 1990, 112, 5525. (d) Cozzi, F.;
Siegel, J . S. Pure Appl. Chem. 1995, 67, 683. (e) Shetty, A. S.; Zhang,
J .; Moore, J . S. J . Am. Chem. Soc. 1996, 118, 1019. (f) Williams, V. E.;
Lemieux, R. P.; Thatcher, G. R. J . J . Org. Chem. 1996, 61, 1927. (g)
Garc´ıa Mart´ınez, A.; Os´ıo Barcina, J .; de Fresno Cerezo, A.; Gutie´rrez
Rivas, R. J . Am Chem. Soc. 1998, 120, 673.
(7) Dougherty, D. A. Science 1996, 271, 163. Ngola, S. M.; Dougherty,
D. A. J . Org. Chem. 1998, 63, 4566.
(8) The importance of π-π interactions in asymmetric synthesis has
been recently pointed out: J ones, G. B.; Chapman, B. J . Synthesis
1995, 475.
(9) Tamaki, R.; Samura, K.; Chujo, Y. J . Chem. Soc., Chem.
Commun. 1998, 1131.
(5) Williams, V. E.; Lemieux, R. P. J . Chem. Soc., Chem. Commun.
1996, 2259.
(10) Rosini, C.; Spada, G. P.; Proni, G.; Masiero, S.; Scamuzzi, S. J .
Am. Chem. Soc. 1997, 119, 506.
10.1021/jo990038y CCC: $18.00 © 1999 American Chemical Society
Published on Web 06/06/1999

Induction of Cholesteric Mesophases
J . Org. Chem., Vol. 64, No. 13, 1999 4763
a
Ta ble 1. Helica l Tw istin g P ow er s âM of Com p ou n d s
Ch a r t 1
2a -6a a n d 7 in th e Nem a tic Solven ts MBBA a n d E7
â
M (sd)b/µm-1
θc (deg) dc (Å) in MBBAd in E7e
compd
Y
2a
3a
4a
5a
6a
7f
C(CH3)2
CHOCH3
CH2
SO
CO
84
88
79
77
103
143
5.2
5.4
5.4
5.5
5.9
6.4
-5.4 (2.0) -17.8 (3.7)
-11.9 (0.6) -21.0 (1.4)
-17.7 (0.4) -30.6 (2.1)
-16.4 (0.2) -29.9 (0.4)
-22.3 (0.3) -30.3 (1.4)
-37
-32
a
b
See ref 17. Standard deviation. c From MMX calculations.
d
MBBA: N-(4-methoxybenzylidene)-4-butylaniline (from Reidel-
de-Haan). e E7: eutectic mixture (from BDH) of 4-cyano-4′-n-
alkylbiphenyl derivatives. f Data from ref 19.
Ch olester ic In d u ction
Helical twisting powers, âM ) ((pcr)-1 (where p is the
cholesteric pitch, c the molar fraction of the dopant, and
r its enantiomeric excess; the sign is taken positive for a
right-handed, P, cholesteric and negative for a left-
handed, M, one), have been measured by the Grandjean-
Cano method.17
The values of âM of compounds 2-6 in the nematic
solvents MBBA and E7 are collected in Tables 1 and 2.
For the sake of clarity, the discussion will be divided into
two parts concerning, respectively, the influence of the
ring closure and the effect of the p-substituents on the
twisting power.
Syn th esis
(A) Th e Role of th e Rin g Closu r e. To explain the
effects of the structure on the value and sign of âM (i.e.,
on the strength of the solute/solvent interactions), the
conformation and the geometrical features of the dopant
have to be known. The most stable conformations of
compounds (R,R)-2a -6a , as derived from MMX18 calcula-
tions and previous studies,10 are shown in Figure 1. In
this conformation, the two aromatic rings assume a
quasi-gauche relationship when viewed along the C4-
C5 bond (Figure 1a) and each phenyl nearly bisects the
five membered ring (Figure 1b). The nature of the ring
junction determines both the steric hindrance around the
position 2 of the dioxolane ring and the dihedral angle θ
(C1′-C4-C5-C1′′) between the two aromatic rings, thus
affecting the distance d between the centers of the phenyl
rings. In Table 1 are reported the values of θ, d, and âΜ
for compounds (R,R)-2a -6a and for (R,R)-trans-stilbene
oxide (7).19 In these compounds only the type of ring
closure is varied, and this structural feature affects
mainly the molecular shape (vide supra) even if electronic
effects coming from polar moieties (i.e., 5a vs 4a ) have
to be considered. Apparently, the fact that some of these
compounds (i.e., 2a , 4a , and 6a ) have a C2 symmetry does
not have any influence on the values of âΜ measured. The
results in Table 1 indicate that, for these unsubstituted
derivatives, the sign of âΜ is determined only by the
handedness of the dopant while its absolute value is
significantly affected by the structure of the ring. As a
matter of fact, it can be observed that, independently of
the nature of the five-membered ring, R,R derivatives
induce a M cholesteric, both in MBBA and E7. Molecular
Compounds 1a -c were prepared as previously de-
scribed.10 The new compounds 1d and 1e were analo-
gously obtained by catalytic asymmetric dihydroxyl-
ation11 of the corresponding (E)-1,2-diarylethenes derived,
by McMurry reaction,12 from 4-trifluoromethylbenzalde-
hyde or 4-chlorobenzaldehyde. The ee’s of diols 1a -e
(>90%) were determined by HPLC upon the Daicel
Chiralcel OD and OJ columns, and their absolute con-
figurations, assigned as R,R on the basis of the Sharpless
empirical rule,11 were confirmed by the exciton analysis
of the CD spectra of their 2,2-dimethyl-1,3-dioxolanes 2d
and 2e.13 The 2,2-dimethyl-1,3-dioxolanes (R,R)-2a -e
were prepared in high yields (88-97%) by reaction of the
corresponding diols (R,R)-1a -e and 2,2-dimethoxypro-
pane in CHCl3 in the presence of 4 Å molecular sieves
and traces of p-TsOH. When (R,R)-1a was reacted with
methyl orthoformate in DME and catalytic amounts of
p-TsOH, (R,R)-3a was obtained, after usual workup and
crystallization, in 82% yield.14 Compound (R,R)-4a was
obtained in a 70% yield by reacting (R,R)-1a with CH2I2
in CH2Cl2, in the presence of KOH and catalytic amounts
of 18-crown-6. Reaction of (R,R)-1a with SOCl2 in reflux-
ing CCl4 provided, after column chromatography, (R,R)-
5a in 56% yield.15 Finally, the carbonates (R,R)-5a -e
were prepared from the corresponding diols (R,R)-1a -e
by reaction with triphosgene at -70 °C in the presence
of pyridine, in yields ranging from 75 to 85%.16
(11) Kolb, H. C.; Van Nieuwenhze, M. S.; Sharpless, K. B. Chem.
Rev. 1994, 94, 2483.
(12) McMurry, J . E. Chem. Rev. 1989, 89, 1523.
(17) (a) Heppke, G.; Oestreicher, F. Z. Naturforsch. 1977, A32, 899.
(b) Heppke, G.; Oestreicher, F. Mol. Cryst. Liq. Cryst. 1978, 41, 245.
(c) Gottarelli, G.; Samor`ı, B.; Stremmenos, C.; Torre, G. Tetrahedron
1981, 37, 395.
(13) Rosini, C.; Scamuzzi, S.; Pisani Focati, M.; Salvadori, P. J . Org.
Chem. 1995, 60, 8289. Rosini, C.; Scamuzzi, S.; Uccello-Barretta, G.;
Salvadori, P. J . Org. Chem. 1994, 59, 7395.
(14) Machinaga, N.; Kibayashi, H. Tetrahedron Lett. 1989, 31, 4165.
(15) Lohray, B. B. Synthesis 1992, 1035.
(16) Burk, R. M.; Roof, M. B. Tetrahedron Lett. 1993, 34, 395.
(18) MMX, Serena Software, P.O. Box 3076, Bloomington, IN 47402.
(19) Gottarelli, G.; Mariani, P.; Spada, G. P.; Samor`ı, B.; Forni, A.;
Solladie´, G.; Hibert, M. Tetrahedron 1983, 39, 1337.

4764 J . Org. Chem., Vol. 64, No. 13, 1999
Superchi et al.
a
Ta ble 2. Helica l Tw istin g P ow er s âM of Com p ou n d s 2a -e a n d 6a -e in th e Nem a tic Solven ts MBBA a n d E7
ketals âM (sd)b/µm-1
carbonates âM (sd)b/µm-1
c
X
σp
Rd
υe
MBBAf
E7g
MBBAf
E7g
OCH3
CH3
Cl
H
CF3
-0.28
-0.14
0.24
0
1.31
1.23
1.24
1.04
1.16
0.36
0.52
0.55
0
2b
2c
2e
2a
2d
-11.0 (0.1)
-18.2 (0.1)
-6.22 (0.2)
-5.4 (2.0)
+9.4 (1.1)
-26.6 (0.8)
-22.3 (1.7)
-22.0 (0.1)
-17.8 (3.7)
-10.8 (0.2)
6b
6c
6e
6a
6d
-38.7 (1.5)
-29.5 (1.1)
-27 (2)
-22.3 (0.3)
-20.9 (0.2)
-38.4 (0.1)
-26.1 (0.7)
-33.1 (2.3)
-30.3 (1.4)
-20.4 (1.1)
0.53
0.91
a
b
d
See ref 17. Standard deviation. c Data from ref 20. Data (10-23 cm3) calculated from Lorenz-Lorentz equation; see ref 5. e Data
from refs 21 and 22. f MBBA: N-(4-methoxybenzylidene)-4-butylaniline (from Reidel-de-Haan). E7: eutectic mixture (from BDH) of
g
4-cyano-4′-n-alkylbiphenyl derivatives.
shorter biphenyl molecules of E7 (d ) 4.3 Å, Figure 1d)
and why an opposite behavior is shown by the short
molecules of the ketals 2a -6a (d < 6 Å). This interpreta-
tion can also explain the presently observed trend of âΜ
in MBBA; in fact, different types of ring closure can
determine different kinds of steric hindrance as well as
different overall molecular shapes. From Table 1, it is
clearly shown that the trend of âΜ values is related to
the increasing of the interaromatic distances d and on
the steric hindrance at the 2 position. Lower values of
âΜ are obtained with the “shorter” molecule 2a and
medium values with 3a , 4a , and 5a . For these com-
pounds, which have very similar values of d, the differ-
ence in âΜ is then determined by the steric hindrance at
the position 2, and hence, the less hindered 4a and 5a
show larger âΜ than 3a . The highest vales are observed
with the carbonate 6a where the presence of a sp2 CO
carbon not only minimizes the hindrance at C2 but also
causes a widening of the θ angle (Table 1), thus length-
ening the molecule and increasing its fit with the MBBA
molecules. This effect is even more evident with the
epoxide 7 which, having the longer d and devoid of
sterically disturbing groups, shows in MBBA19 the largest
âM (-37).
F igu r e 1. Views of the most stable conformation for deriva-
tives 2-6 along the C4-C5 bond (a) and from the top (b). θ is
the C1′-C4-C5-C1′′ dihedral angle. Distances d between the
aryl cores of a biphenyl (c), taken as a model of E7, and of the
MBBA molecule (d).
mechanics calculations18 indicate that 2a -6a assume10
prevailing conformations where the benzene rings are
M-skewed (Figure 1b), thus justifying why they induce,
both in the E7 and in the MBBA solvent, M-twisted
conformations. An interpretation of the different absolute
values of âΜ observed has, however, to take into account
the influence of the Y group onto the overall molecular
shape. The trend of âΜ values observed in E7 (2a < 3a <
4a ≈ 5a ≈ 6a ) can be explained on the basis of the steric
hindrance exerted by the group Y. As a matter of fact, in
both 2a and 3a there is the presence of relatively bulky
groups on the 2 position that, being placed out of the
plane of the dioxolane ring,10 can create steric hindrance
to the solute/solvent packing. On the other hand, 4a , 5a ,
and 6a bear, respectively, only two hydrogens or an
oxygen on the 2 position, and then steric hindrance to
intermolecular solute/solvent interactions is reduced. In
MBBA, compounds 2a -6a induce lower values of âΜ with
respect to E7, indicating less efficient interaction of these
substrates with MBBA, and the following trend was
observed: 2a < 3a < 4a ≈ 5a < 6a . According to the
accepted model2 of cholesteric induction, a close contact
between the aromatic moieties of the solute and of the
solvent is needed in order to have transmission of
chirality. Such a close contact requires a certain degree
of similarity between solute and solvent, particularly in
shape. This model can then explain10 why the molecules
of trans-stilbene oxide (7), having a distance d between
the aryl cores of 6.4 Å (Table 1), interact stronger with
the “long” MBBA (d ) 6.6 Å, Figure 1c) than with the
As far as the cholesteric induction in E7 is concerned,
the difference between the minimum and the maximum
âM is not as high as in MBBA. This is because the
biphenyl molecules of E7 are “shorter” than those of
MBBA and then there is a better fit with the short
molecules of 2a -6a . Therefore, in E7 the lengthening of
the molecule disfavors the solute/solvent fit, but the
reduction of the steric hindrance still increases the âM
values. The balance of these two factors renders the âM
values less pronounced.
(B) Th e Role of th e p,p′-Su bstitu en ts. In Table 2
are collected the values of the âM of 2,2-dimethyl-1,3-
dioxolanes 2a -e and carbonates 6a -e. The following
comments can be made: (i) For the p,p′-disubstituted
compounds (as for the derivatives 2a -6a discussed
above) the correlation (R,R)-dopant/M-cholesteric also
holds, with the sole exception of 2d in MBBA. (ii) There
is a clear effect of the nature of the p-substituent on the
value of the twisting power. Indeed, electron-donating
substituents (see OMe in 2b and 6b) induce higher âΜ
values than electron-withdrawing groups (CF3), both in
the dioxolane and in the carbonate family. (iii) These
substituents effects are particularly strong considering
that, for instance, in the ketals family passing from 2d
to 2b the âΜ values are more than doubled.
In an attempt to rationalize the experimental facts
observed we shall consider that, as shown in the previous
paragraph, shape (steric) factors have to be taken into
account and that both the electronic nature of the
substituents and the polarizability of the systems can

Induction of Cholesteric Mesophases
J . Org. Chem., Vol. 64, No. 13, 1999 4765
control the âΜ values, as recently shown by Williams and
Lemieux5 in a series of substituted chiral biphenyls. In
Table 2 are collected the values of the Hammett σp
constant20 for the X substituents, which is an index of
their electron-donor or -withdrawing character, the Char-
ton parameter (υ),21 which is related22 to the steric
hindrance of the substituent, and the polarizability
values (R) for the corresponding substituted benzenes.
We assumed that a measure of the polarizability (R) of
2a -e can be given by the polarizability of the corre-
sponding monosubstituted benzenes calculated by the
Lorenz-Lorentz equation.5
A satisfactory interpretation of the âΜ values observed
can be obtained by analyzing their relationship with the
polarizability (R) of the substituted phenyl moieties and
steric hindrance (υ) of the substituent X of the dopants
as reported in Table 2. First, we consider the âΜ values
of compounds 2a -e (ketals) in the biphenyl nematic
solvent E7. The following comments can be easily made:
(i) 2c and 2e show the same âΜ value (within the
experimental error), and this result can be explained
considering that toluene (model of 2c) and chlorobenzene
(model of 2e) have the same R values and that the
parameters υ for Me and chlorine are very similar. (ii)
The highest âΜ value recorded for 2b can be due to the
fact that anisole has by far the highest polarizability and
the methoxy group has a small steric parameter.21 (iii)
The lowest âΜ value is recorded for 2d ; in fact, the
polarizability of trifluoromethylbenzene is the lowest in
the series (apart from benzene) and the steric require-
ment of CF3 is the highest (CF3 has size and steric
parameter υ between a methyl and a tert-butyl group).22,23
(iv) The intermediate âΜ value of 2a is in line with this
interpretation, because the small contribution due to the
polarizability of benzene is not further reduced by effects
due to the size of the substituents.
To further support the above interpretation, it can be
observed that (R,R)-(+)-2,2-dimethyl-4,5-di(4-biphenylyl)-
1,3-dioxolane (8) (Chart 1) shows in E710 a very high âΜ
value (-58): as a matter of fact, the p-phenyl substituent
has a small steric parameter, 0.56 (like a methoxy group),
but shows a very high group polarizability.22
The above interpretation can work also in the case of
MBBA where, as for the unsubstituted derivatives, lower
values of âΜ are obtained (see Table 2). Considering that
MBBA and pentylcyano biphenyl (taken as a model of
the major component of the E7 mixture) have similar
polarizability values (3.15 × 10-23 cm3 vs 2.91 × 10-23
cm3) as calculated by the Lorenz-Lorentz equation, the
difference in âΜ values will depend only on the nature of
the solutes.
F igu r e 2. Electrostatic charge distribution on benzene: face-
to-face (a), offset stacked (b), and edge-to-face (c) disposition
of aromatic moieties.
(as shown above), which can lead to an alternative
mechanism of interaction between solute and solvent,
giving rise to the opposite sense of twisting of the
cholesteric.
The carbonates 6a -e show in E7 higher â values than
the ketals 2a -e, and this observation has been previously
interpreted considering that carbonates do not possess
the steric hindrance of the dimethyl dioxolanes. The
behavior of compounds 6a -e in E7 and MBBA can be
interpreted as for compounds 2a -e. Interestingly, in this
case we do not observe any inversion of the cholesteric
helix: reasonably, compounds (R,R)-6 interact so strongly
with the nematic solvent to give an M cholesteric that
steric effects are not able to overwhelm this tendency.
The analysis reported above agrees with the original
mechanism of cholesteric induction proposed by Gottarelli
et al.,2,24 who, assuming π-facial interactions between the
aromatic moieties of dopant and solvent, state that the
sense of twist of the aromatic moieties of the dopant is
responsible for the sense of the cholesteric induction. We
showed that other factors such as polarizability and
overall molecular shape strongly contribute to the deter-
mination of the value of the twisting power. The role
played by the polarizability of the solute seems to indicate
that attractive dispersive forces may contribute (in
addition to hard-core repulsive interaction)25 to the
cholesteric induction; however, since we are dealing with
aromatic solvents and solutes, arene-arene interactions
cannot be ruled out, and then it is also interesting to
analyze the present experimental data taking into ac-
count the models describing arene-arene interactions.
Some recent experimental6d,e and theoretical6a-c,f studies
have in fact shown that these interactions are mainly of
electrostatic nature and that the benzene ring has a
charge distribution showing a negative potential above
and below the molecular plane and a positive potential
focused on the carbon skeleton and on the hydrogens
(Figure 2).6a,7 According to this electrostatic model, face-
to-face (a) aromatic dispositions are then disfavored,
maximizing π electron repulsion, while shifted, or offset
stacked (b), and edge-to-face (c) orientations, which face
the positively charged σ-framework and the negatively
charged π-electrons are stabilized. By examining the
possibilities of interaction of the aromatic moieties of
solutes 2 and 6 with the phenyl rings of a biphenyl
Unexpectedly, an opposite P sign of the cholesteric is
induced by the 4-CF3-substituted ketal (R,R)-2d in
MBBA. As already shown in Table 1, ketals (being short
molecules) have in MBBA (long molecule) a lower ten-
dency to form M-cholesteric than in E7, and as shown in
Table 2, the lowest âΜ values are obtained with 4-CF3-
substituted compounds. Then it is reasonable that in the
case of (R,R)-2d the low tendency to form M-cholesteric
can be then overwhelmed by the steric hindrance of CF3
(20) March, J . Advanced Organic Chemistry: reactions, mechanism
and structure, 4th ed.; Wiley-Interscience: New York, 1992; p 280.
(21) Hansch, C.; Leo, A. J . Substituent Constants for Correlation
Analysis in Chemistry and Biology; Wiley: New York, 1979.
(22) Charton, M.; Ziffer, H. J . Org. Chem. 1987, 52, 2400.
(23) Schlosser, M.; Michel, D. Tetrahedron 1996, 52, 99.
(24) Gottarelli, G.; Hibert, M.; Samor`ı, B.; Solladie´, G.; Spada, G.
P.; Zimmermann, R. J . Am. Chem. Soc. 1983, 105, 7318.
(25) For
a description of the general theories of the molecular
ordering in liquid crystal phases, see, for example: Madhusudana, N.
V. In Liquid CrystalssApplications and Uses; Bahadur, B., Ed.; World
Scientific: Singapore, 1990; Vol. 1, Chapter 2.

4766 J . Org. Chem., Vol. 64, No. 13, 1999
Superchi et al.
contribution to the molecular polarizability, thus pointing
out the role played by dispersive forces on the chirality
transfer from the solute to the nematic solvent. Taking
into account the aromatic structure of both the solutes
and the solvents tested, it has been also pointed out that
arene-arene electrostatic interactions can contribute,
even if to a lesser extent, to afford an efficient cholesteric
induction. These two types of interactions (i.e., polariza-
tion and electrostatic) are somehow connected as the
substituents, affecting the electron density on the aro-
matic ring, influence both the polarizability and the
charge distribution of the arenes. The importance of the
substituent effects is demonstrated also by the observa-
tion that solutes having the same absolute configuration
but different substitution can induce cholesteric with
opposite sign (see ketals in MBBA). This observation
clearly points out that reliable configurational correla-
tions based on the analysis of the cholesteric induction
are achievable only with solutes affording high values of
âΜ. To obtain a safe configurational correlation, it is then
necessary to choose suitable derivatives of the compounds
studied, similar in shape to the molecules of solvent, so
that high âΜ values result, independently of the nature
of the substituents (the case of the carbonates 6a -e is a
clear example). The findings reported herein can have
important consequences from both theoretical (formula-
tion of quantitative models of cholesteric induction,27
understanding of the factors determining solute/solvent
interactions) and practical points of view (rational design
of chiral liquid crystal materials for technological ap-
plications).
F igu r e 3. Possible dispositions of the aromatic moieties of
solutes 2 and 6 with a biphenyl.
solvent such as E7, the two dispositions in Figure 3 can
be depicted. Orientation (a) in Figure 3, with the biphenyl
molecule shifted in respect to the arenes of the solute, is
compatible with an offset-stacked disposition of the
aromatic rings, while orientation (b) ensures a stabilizing
edge-to-face orientation of the aromatic moieties. It has
to be noted that with both these orientations a R,R
dopant molecule, whose benzene rings define an M
chirality, induce an M-twisted conformation in the bi-
phenyl molecule (compare dispositions (a) and (b) in
Figure 3). This electrostatic model can also explain the
trend of âΜ values reported in Table 2, which increases
with the electron-donor properties of the substituent onto
the aromatic ring of the dopant. As a matter of fact,
according to the edge-to-face or the offset-stacked model,
the electrostatic arene-arene attraction increases with
an increase of the electron density on the arene borne
by the dopant.6c The chlorine-substituted compounds 2e
and 6e constitute a significant exception to this model
because the values of âΜ are higher than those expected
only on the basis of the σp value of chlorine. This result
can be only explained by taking into account also the high
polarizability of Cl (vide supra). This finding, then,
further confirms that the above interpretation based on
polarizability effects and on the intervention of dispersive
forces affords a more complete explanation of the experi-
mental results. The interpretation based on the polariz-
ability effects also agrees with the experimental finding
that ketals 2b, 2d , and 2e show, in the nonaromatic
solvent ZLI2359, âM values of -22.7, -24.9, and -14.8
µm-1, respectively, comparable with those obtained in the
aromatic nematics MBBA and E7. In this case, arene-
arene interactions are clearly not involved, while only
the intervention of dispersive forces can explain the high
âM values, considering the high polarizability (2.06 ×
10-23) showed by dicyclohexyl (i.e., the structural moiety
of the component of ZLI2359).
Exp er im en ta l Section
Gen er a l P r oced u r es. Melting points were determined
with a Kofler hot-stage apparatus and are uncorrected. 1H
NMR (300 MHz) and 13C NMR (75 MHz) spectra were recorded
in CDCl3. Enantiomeric excesses of the diols were determined
by HPLC analysis on a Daicel Chiralcel OD or Chiralcel OJ
column. CCl4 and CH2Cl2 were distilled, respectively, from
CaH2 and P2O5 and stored over activated 4 Å molecular sieves.
Analytical and preparative TLC were performed, respectively,
on 0.2 and 2.0 mm silica gel plates Merck 60 F-254, and column
chromatography was carried out with silica gel Merck 60 (80-
230 mesh). Enantiomerically pure (R,R)-1,2-diarylethane-1,2-
diols 1a -e were prepared by catalytic asymmetric dihydroxyl-
ation11 of the corresponding p,p′-disubstituted-(E)-stilbenes
obtained, in turn, from Wittig reaction or McMurry coupling.12
Dioxolanes 2a -c were prepared as previously described.10
N-(4-Methoxybenzylidene)-4-butylaniline (MBBA) (Reidel-de-
Haan) and E7 mixture (BDH) were used as purchased.
Gen er a l P r oced u r e for th e Syn th esis of 4,5-Dia r yl-2,2-
d im eth yl-1,3-d ioxola n es. To a solution of (R,R)-1,2-diaryl-
ethane-1,2-diol (0.14 mmol) in CHCl3 (5 mL) was added 2,2-
dimethoxypropane (0.172 mL, 1.4 mmol), followed by traces
of p-toluensulfonic acid and activated 4 Å molecular sieves.
The mixture was stirred at room temperature overnight and
then filtered over a short pad of silica gel (eluent CHCl3). The
solvent was evaporated to recover a solid residue of pure
dioxolane.
Con clu sion s
In conclusion, this investigation has confirmed that the
twisting power âM is a function of the overall molecular
shape (the solvent molecules recognize the distortion of
the solute, and they feel whether the solute is “long” or
“short” and if something on the solute surface blocks their
fit) and has demonstrated that a substituent, affecting
the electronic properties (i.e., polarizability26 and electron
density) and the steric requirements of the parent com-
pound, can heavily affect the âΜ values. The effect of
substituents has been interpreted on the basis of their
(R,R)-(+)-4,5-Di(4-tr iflu or om eth ylph en yl)-2,2-dim eth yl-
1,3-d ioxola n e (2d ): yield 95%; mp 74-76 °C; [R]D ) +54.9 (c
1
) 0.65, CHCl3); H NMR δ 1.70 (s, 6H), 4.75 (s, 2H), 7.34 (d,
J ) 8.1 Hz, 4H), 7.60 (d, J ) 8.1 Hz, 4H); 13C NMR δ 27.03,
84.81, 110.33, 125.50, 126.95, 130.60, 131.05, 140.49; MS (EI)
(26) After the submission of this manuscript, we became aware of
the following paper dealing with the role of dispersion and electrostatic
forces on solute/solvent interactions in a nematic liquid crystal phase:
Williams, V. E.; Lemieux, R. P. J . Am. Chem. Soc. 1998, 120, 11311.
(27) Interestingly, theoretical models of cholesteric induction have
been formulated only taking into account molecular shape factors:
Ferrarini, A.; Moro, G. J .; Nordio, P. L. Phys. Rev. E 1996, 53, 681.
Memmer, R.; Kuball, H. G.; Scho¨nhofer, A. Mol. Phys. 1996, 89, 1633.

Induction of Cholesteric Mesophases
J . Org. Chem., Vol. 64, No. 13, 1999 4767
m/z 216 (M+, 100), 159 (67), 147 (79). Anal. Calcd for
SO4. After evaporation of solvent, the collected residue was
purified by column chromatography.
C
19H16F6O2: C, 58.47; H, 4.13. Found: C, 58.23; H, 4.41.
(R,R)-(+)-4,5-Di(4-ch lor oph en yl)-2,2-dim eth yl-1,3-dioxo-
(R,R)-(+)-4,5-Diph en yl-1,3-dioxolan on e (6a):29 yield 85%;
mp 110-112 °C (for (()-6a lit.30 mp 110-111 °C); [R]D ) +66
la n e (2e): yield 85%; mp 101-102 °C; [R]D ) +113.2 (c ) 1.4,
1
1
CHCl3); H NMR δ 1.66 (s, 6H), 4.63 (s, 2H), 7.12 (d, J ) 8.5
(c ) 1.05, CHCl3); H NMR δ 5.45 (s, 2H), 7.4 (m, 10H); 13C
Hz, 4H), 7.30 (d, J ) 8.5 Hz, 4H); 13C NMR δ 27.11, 84.84,
NMR δ 85.34, 126.05, 129.23, 129.78, 134.88, 154.02.
(R,R)-(+)-4,5-Di(4-m eth oxyph en yl)-1,3-dioxolan on e (6b):
yield 78%; [R]D ) +144.7 (c ) 1.05, CHCl3); 1H NMR δ 3.83 (s,
6H), 5.37 (s, 2H), 6.93 (d, J ) 8.7 Hz, 4H), 7.23 (d, J ) 8.7 Hz,
4H); 13C NMR δ 55.28, 85.34, 114.48, 126.33, 127.78, 154.08,
160.63. Anal. Calcd for C17H16O5: C, 67.99; H, 5.37. Found:
C, 68.13; H, 5.28.
109.74, 127.96, 128.73, 134.23, 134.97. Anal. Calcd for C17H16
Cl2O2: C, 63.17; H, 4.99. Found: C, 63.31; H, 4.72.
-
(R,R)-(+)-4,5-Dip h en yl-2-m et h oxy-1,3-d ioxola n e (3a ).
The diol (R,R)-1 (100 mg, 0.47 mmol) was dissolved in DME
(3 mL), and then trimethyl orthoformate (51 µL, 0.47 mmol)
and traces of p-TsOH were added at room temperature. The
mixture was stirred for 1 h and then diluted with Et2O, washed
with saturated NaHCO3 (aq) and brine, and dried over Na2-
SO4. After evaporation of solvent, a solid residue was recovered
that was recrystallized from CH2Cl2-petroleum ether, obtain-
ing 102 mg (85%) of 3a as white crystals: mp 106-107 °C;
(R,R)-(+)-4,5-Di(4-m eth ylp h en yl)-1,3-d ioxola n on e (6c):
yield 82%; [R]D ) +105.7 (c ) 1.2, CHCl3); 1H NMR δ 2.38 (s,
6H), 5.33 (s, 2H), 7.19 (d, J ) 8.3 Hz, 4H), 7.23 (d, J ) 8.3 Hz,
4H); 13C NMR δ 21.22, 85.45, 126.12, 129.81, 131.76, 139.80,
154.22. Anal. Calcd for C17H16O3: C, 76.10; H, 6.01. Found:
C, 75.95; H, 6.23.
1
[R]D ) +76.6 (c ) 2.39, CHCl3); H NMR δ 3.53 (s, 3H), 4.82
(d, J ) 8.6 Hz, 1H), 4.97 (d, J ) 8.6 Hz, 1H), 6.12 (s, 1H), 7.33
(m, 10H). Anal. Calcd for C16H16O3: C, 74.98; H, 6.29. Found:
C, 74.77; H, 6.41.
(R,R)-(+)-4,5-Di(4-t r iflu or om et h ylp h en yl)-1,3-d ioxol-
a n on e (6d ): yield 70%; mp 58-60 °C; [R]D ) +47 (c ) 1.06,
CHCl3); 1H NMR δ 5.45 (s, 2H), 7.45 (d, J ) 8.2 Hz, 4H), 7.74
(d, J ) 8.2 Hz, 4H); 13C NMR δ 84.28, 125.36, 126.44, 132.11,
132.56, 138.16, 153.17. Anal. Calcd for C17H10F6O3: C, 54.27;
H, 2.68. Found: C, 54.43; H, 2.82.
(R,R)-(+)-4,5-Di(4-ch lor op h en yl)-1,3-d ioxola n on e (6e):
yield 77%; mp 70-71 °C; [R]D ) +114.7 (c ) 2.91, CHCl3); 1H
NMR δ 5.36 (s, 2H), 7.24 (d, J ) 8.4 Hz, 4H), 7.42 (d, J ) 8.4
Hz, 4H);13C NMR δ 84.59, 127.46, 129.57, 132.75, 136.07,
(R,R)-(+)-4,5-Dip h en yl-1,3-d ioxola n e (4a ). To a solution
of (R,R)-1 (200 mg, 0.93 mmol) in CH2Cl2 (10 mL) were added
in sequence 18-crown-6 (25 mg, 0.093 mmol), finely ground
KOH (210 mg, 3.72 mmol), and CH2I2 (300 µL, 3.72 mmol).
The mixture was stirred overnight at room temperature and
then diluted with CH2Cl2, washed twice with water and brine,
and dried over anhydrous Na2SO4. After evaporation of solvent,
the residue was purified by column chromatography (3:1
petroleum ether/Et2O) to obtain 168 mg (80% yield) of 4a as a
white solid: mp 51-52 °C; [R]D ) +82.2 (c ) 1.04, CHCl3); 1H
NMR δ 4.71 (s, 2H), 5.47 (s, 2H), 7.33 (m, 10H); 13C NMR δ
85.42, 96.37, 126.49, 128.31, 128.56, 137.52; MS (EI) m/z 226
(M+, 2), 120 (100), 91 (52). Anal. Calcd for C15H14O2: C, 79.62;
H, 6.24. Found: C, 79.58; H, 6.32.
153.42. Anal. Calcd for
Found: C, 58.17; H, 3.41.
C15H10Cl2O3: C, 58.28; H, 3.26.
In d u ced Ch olester ic Mea su r em en ts. Cholesteric pitches
were measured by means of the “lens” version of the Grand-
jean-Cano method, using a standard 16 Zeiss microscope;
helical handedness was obtained from the sign of the rotatory
power and from the sense of the spiral-like disclination
observed under circular boundary conditions.17c
(R,R)-(+)-4,5-Dip h en yl-1,3,2-d ioxa tiola n -2-on e (5a ).15,28
To a suspension of (R,R)-1 (200 mg, 0.93 mmol) in CCl4 (5 mL)
was added SOCl2 (83 µL, 1.12 mmol) under nitrogen atmo-
sphere. The mixture was then refluxed for 1 h and cooled at
room temperature and the solvent evaporated. The solid
residue obtained was purified by column chromatography over
silica gel (petroleum ether/Et2O ) 3:1) to obtain 198 mg (82%)
of 5a as a white solid: mp 98-100 °C; [R]D) +148 (c ) 0.68,
Ack n ow led gm en t. We would like to thank Profes-
sor Giovanni Gottarelli (Bologna) for helpful discussions.
Financial support from MURST (National Project “Liq-
uid Crystals: Structures, Computer Simulations, Prop-
erties and Applications”), Universita` di Bologna, Uni-
versita` della Basilicata and CNR (Roma) is gratefully
acknowledged.
1
CHCl3); H NMR δ 5.22 (d, J ) 9.5 Hz, 1H), 5.70 (d, J ) 9.5
Hz, 1H), 7.45 (m, 10H); 13C NMR δ 86.15, 91.45, 127.40, 127.76,
129.13, 129.54, 129.85, 132.47, 133.53.
Su p p or t in g In for m a t ion Ava ila b le: 1H and 13C NMR
spectra of compounds 2d , 2e, 3a , 4a , 5a , 6a -e. This material
is available free of charge via the Internet at http://pubs.acs.org.
Gen er a l P r oced u r e for th e Syn th esis of 4,5-Dia r yl-1,3-
d ioxola n on es. A solution of 1,2-diarylethane-1,2-diol (0.14
mmol) and pyridine (0.24 mL, 0.84 mmol) in CH2Cl2 (3 mL)
was cooled at -70 °C, and a solution of triphosgene (42 mg,
0.14 mmol, 1 equiv) in anhydrous CH2Cl2 (2 mL) was added.
The mixture was then allowed to warm at room temperature,
stirred overnight, and then washed with saturated NH4Cl (aq)
and the aqueous phase extracted with CH2Cl2. The collected
organic phases were washed with 10% aqueous HCl and with
saturated NaHCO3 (aq) and then dried over anhydrous Na2-
J O990038Y
(28) Murthy, K. S. K.; Dhar, D. N. J . Heterocycl. Chem. 1984, 21,
1721.
(29) (a) Iida, T.; Itaya, T. Tetrahedron 1993, 49, 10511. (b) Nicolaou,
K. C.; Renaud, J .; Nantermet, P. G.; Couladouros, E. A.; Guy, R. K.;
Wrasidlo, W. J . Am. Chem. Soc. 1995, 117, 2409.
(30) Nymann, K.; Svendsen, J . S. Acta Chem. Scand. 1994, 48, 183.