Ferroelectric liquid crystals induced by dopants with axially chiral 2,2′-spirobiindan-1,1′-dione cores

Article abstract of DOI:10.1021/ja054322k

The axially chiral dopants (R)-5,5′-, 5,6′-, and 6,6′-diheptyloxy-2,2′-spirobiindan-1,1′-dione ((R)-2, -3, and -4) were synthesized in optically pure form, and their absolute configurations were assigned by the exciton chirality method using circular dich

Full text of DOI:10.1021/ja054322k

Published on Web 09/08/2005
Ferroelectric Liquid Crystals Induced by Dopants with Axially
Chiral 2,2′-Spirobiindan-1,1′-dione Cores
Christopher J. Boulton,† Jeremy G. Finden,† Eagranie Yuh,† Jeffrey J. Sutherland,†
Michael D. Wand,‡ Gang Wu,† and Robert P. Lemieux*,†
Contribution from the Department of Chemistry, Queen’s UniVersity,
Kingston, Ontario, Canada K7L 3N6, and Displaytech, Inc., 2602 CloVer Basin DriVe,
Longmont, Colorado 80503
Received June 30, 2005; E-mail: lemieux@chem.queensu.ca
Abstract: The axially chiral dopants (R)-5,5′-, 5,6′-, and 6,6′-diheptyloxy-2,2′-spirobiindan-1,1′-dione ((R)-
2, -3, and -4) were synthesized in optically pure form, and their absolute configurations were assigned by
the exciton chirality method using circular dichroism spectroscopy. These new compounds were doped in
four achiral liquid crystal hosts to give chiral smectic C* (SmC*) phases with spontaneous polarizations
(PS) that vary with the core structure of the host. The spontaneous polarization induced by the 5,5′-dialkoxy
derivative (R)-2 is uniformly positive, whereas that induced by the 6,6′-dialkoxy derivative (R)-4 is uniformly
negative and shows a different trend in host dependence. Polarization power (δp) values range from +21
nC/cm2 for (R)-2 in 2′,3′-difluoro-4-heptyl-4′′-nonyl-p-terphenyl (DFT) to -1037 nC/cm2 for (R)-4 in 4-(4′-
heptyl[1,1′-biphen]-4-yl)-1-hexylcyclohexanecarbonitrile (NCB76). The unsymmetrical dopant (R)-3 behaves
like a hybrid of the two symmetrical isomers, with lower absolute values of δp, on average, and varying
signs of PS. 2H NMR spectra of the doped mixtures using racemic mixtures of 2-4 with -OCD2C6H13
side-chains, in combination with phase diagrams, show that relatively minor changes in the dopant structure,
that is, moving the alkoxy side-chains from the 5,5′ to the 6,6′ positions of the spirobiindandione core,
have profound effects on dopant-host compatibility, and on the propensity of the dopant to exert chiral
perturbations in the host environment. The variations in sign and magnitude of δp as a function of alkoxy
group positions are rationalized based on an analysis of zigzag conformations that conform to the binding
site of the SmC host according to the Boulder model.
Introduction
The performance characteristics of FLC devices, including
electrooptical switching time, second-order NLO susceptibility,
Liquid crystals are ordered fluids in which intermolecular
interactions, including molecular recognition interactions be-
tween a mesogenic host and nonmesogenic additive (dopant),
play an important role in defining the physical properties of
the liquid crystal phase. Recent studies have shown that the
induction of macroscopic chiral properties such as the helical
structure of a chiral nematic or smectic C (SmC*) phase, the
electroclinic tilt of a chiral smectic A (SmA*) phase, and the
spontaneous electric polarization (PS) of a SmC* phase in a
surface-stabilized ferroelectric state (SSFLC) can be understood
in terms of molecular recognition between a chiral dopant and
an achiral liquid crystal host.1-3 The design of chiral dopants
based on principles of molecular recognition is of particular
interest in the case of ferroelectric SmC* liquid crystals, which
are used in commercial high-resolution reflective microdisplays,4
and hold significant potential in nonlinear optics5 and photonics
applications.6-8
and photoswitching threshold, often depend on the magnitude
of PS induced by a chiral dopant. Hence, a key aspect of FLC
materials research is to understand the relationship between the
molecular structure of a chiral dopant and the magnitude of the
spontaneous polarization it induces.2,9-11 This structure-
(4) (a) Lagerwall, S. T. Ferroelectric and Antiferroelectric Liquid Crystals;
Wiley-VCH: Weinheim, 1999. (b) Lagerwall, S. T. In Handbook of Liquid
Crystals; Demus, D., Goodby, J. W., Gray, G. W., Spiess, H. W., Vill, V.,
Eds.; Wiley-VCH: Weinheim, 1998; Vol. 2B. (c) Walba, D. M. Science
1995, 270, 250. (d) Clark, N. A.; Lagerwall, S. T. In Ferroelectric Liquid
Crystals: Principles, Properties and Applications; Goodby, J. W., Blinc,
R., Clark, N. A., Lagerwall, S. T., Osipov, M. A., Pikin, S. A., Sakurai,
T., Yoshino, K., Zeks, B., Eds.; Gordon and Breach: Philadelphia, 1991;
pp 409-452. (e) Dijon, J. In Liquid Crystals: Applications and Uses;
Bahadur, B., Ed.; World Scientific: Singapore, 1990; Vol. 1, Chapter 13.
(5) (a) Walba, D. M.; Xiao, L.; Keller, P.; Shao, R.; Link, D.; Clark, N. A.
Pure Appl. Chem. 1999, 71, 2117. (b) Walba, D. M.; Dyer, D. J.; Sierra,
T.; Cobben, P. L.; Shao, R. F.; Clark, N. A. J. Am. Chem. Soc. 1996, 118,
1211. (c) Walba, D. M.; Ros, M. B.; Clark, N. A.; Shao, R. F.; Robinson,
M. G.; Liu, J. Y.; Johnson, K. M.; Doroski, D. J. Am. Chem. Soc. 1991,
113, 5471.
(6) Crossland, W. A.; Wilkinson, T. D. In Handbook of Liquid Crystals; Demus,
D., Goodby, J. W., Gray, G. W., Spiess, H.-W., Vill, V., Eds.; Wiley-
VCH: Weinheim, 1998; Vol. 1, p 763.
† Queen’s University.
‡ Displaytech, Inc.
(1) (a) Proni, G.; Spada, G. P. Enantiomer 2001, 6, 171. (b) di Matteo, A.;
Todd, S. M.; Gottarelli, G.; Solladie´, G.; Williams, V. E.; Lemieux, R. P.;
Ferrarini, A.; Spada, G. P. J. Am. Chem. Soc. 2001, 123, 7842.
(2) Lemieux, R. P. Acc. Chem. Res. 2001, 34, 845.
(3) Hegmann, T.; Meadows, M. R.; Wand, M. D.; Lemieux, R. P. J. Mater.
Chem. 2004, 14, 185.
(7) Ikeda, T.; Kanazawa, A. In Molecular Switches; Feringa, B. L., Ed.; Wiley-
VCH: Weinheim, 2001; p 363.
(8) Lemieux, R. P. Chem. Rec. 2004, 3, 288.
(9) Walba, D. M. In AdVances in the Synthesis and ReactiVity of Solids;
Mallouck, T. E., Ed.; JAI Press Ltd.: Greenwich, CT, 1991; Vol. 1, p
173.
9
13656
J. AM. CHEM. SOC. 2005, 127, 13656-13665
10.1021/ja054322k CCC: $30.25 © 2005 American Chemical Society

Ferroelectric Liquid Crystals Induced by Dopants
A R T I C L E S
property relationship can be expressed in terms of the polariza-
tion power δp according to eq 1,12 where xd is the mole fraction
of chiral dopant and Po is the polarization normalized for
variations in tilt angle θ according to eq 2.13
Less conventional dopants with chiral cores induce spontane-
ous polarizations that tend to vary significantly with the structure
of the SmC host.2,14,15 This host effect may be viewed as a
manifestation of molecular recognition via core-core interac-
tions with the host molecules that cannot be achieved with
conventional dopants due to the higher degree of conformational
disorder among side-chains in the diffuse layer structure of the
SmC phase.16,17 In other words, the assumption of a shape
invariance for the binding site in the original Boulder model
appears to break down in the case of dopants with chiral cores.
For example, we have shown that dopants such as (R)-1 exhibit
remarkably high polarization powers, up to 1738 nC/cm2, in
achiral SmC hosts with a complementary phenylpyrimidine core
structure in which the atropisomeric core can propagate its
chirality through core-core interactions with surrounding host
molecules.18 The resulting chiral perturbations are thought to
cause a chiral distortion of the binding site topography that
enhances the polar order of the dopant as a feedback effect
(chirality transfer feedback, CTF).2 We recently measured the
effect of chiral perturbations exerted by 1 on the polarization
power of a chiral probe molecule (MDW950, vide infra).19 The
results showed that the polarization power of the probe decreases
by a factor of 0.18 in the presence of (R)-1 and increases by a
factor of 1.4 in the presence of (S)-1, which is consistent with
long-range chiral perturbations influencing the conformational
distribution of the probe.
Despite the orientational and conformational ordering imposed
by the binding site of the SmC host, molecular modeling
suggests that the biphenyl core of (R)-1 can rotate almost freely
with respect to the ester side-chains (Figure 1) and that the
intrinsic conformational bias (i.e., neglecting all intermolecular
interactions with host molecules) favoring one orientation of
the core dipole moment along the polar axis (µ ) is very small,
on the order of 1 kcal/mol.21 Hence, we postulated that it may
be possible to achieve even higher polarization powers by
designing dopants with polar chiral cores with high aspect ratio
that are conformationally more restricted when confined to the
zigzag binding site of the SmC host. In this paper, we report
our first implementation of this approach with a series of dopants
with an axially chiral 2,2′-spirobiindan-1,1′-dione core, (R)-2-
4.22 As expected, the polarization power varies with the structure
of the SmC host, and it also varies with the relative positions
of the alkoxy side-chains. These results are rationalized in terms
of an equilibrium between conformations of opposite polarities
which are energetically equivalent in the gas phase, but should
dPo(xd)
δp )
(1)
(2)
(
)
dxd
xdf0
Po ) PS/sin θ
The spontaneous polarization is a chiral bulk property; it is
either left-handed (negative) or right-handed (positive) depend-
ing on the absolute configuration of the chiral dopant.9 At the
microscopic level, the origins of PS can be understood in terms
of asymmetric conformational energy profiles for polar func-
tional groups sterically coupled to a stereogenic center, which
results in the net orientation of transverse molecular dipoles in
one direction along the polar C2 axis of the SmC* phase.
Empirical and semiempirical structure-property relationships
based on conformational analyses of such stereopolar units are
well established for dopants with chiral side-chains, which
represent the vast majority of chiral dopants found in SmC*
formulations.9-11 In general, the polarization power of these
compounds is invariant with respect to the structure of the SmC
liquid crystal host, which is consistent with the Boulder model
for the molecular origins of PS. According to this model, the
SmC phase is considered to be a supramolecular host, and the
conformational and orientational ordering of a chiral dopant is
modeled by a mean field potential which qualitatively behaves
like a binding site analogous to that described in host-guest
chemistry.9,10 The binding site is C2h symmetric and has a zigzag
shape that is assumed to be invariant with respect to the host
structure. To a first approximation, the Boulder model assumes
that a chiral dopant plays the role of a “passive” guest, which
adopts a conformation that best fits the achiral binding site of
the SmC host.
(12) Siemensmeyer, K.; Stegemeyer, H. Chem. Phys. Lett. 1988, 148, 409.
(13) Kuczynski, W.; Stegemeyer, H. Chem. Phys. Lett. 1980, 70, 123.
(14) Osipov, M. A.; Stegemeyer, H.; Sprick, A. Phys. ReV. E 1996, 54, 6387.
(15) Stegemeyer, H.; Meister, R.; Hoffmann, U.; Sprick, A.; Becker, A. J. Mater.
Chem. 1995, 5, 2183.
(16) Tschierske, C. J. Mater. Chem. 1998, 8, 1485.
(17) For a related example of molecular recognition via core-core interactions
in SmC* liquid crystals, see: (a) Yoshizawa, A.; Nishiyama, I. Mol. Cryst.
Liq. Cryst. 1995, 260, 403. (b) Nishiyama, I.; Ishizika, H.; Yoshizawa, A.
Ferroelectrics 1993, 147, 193.
(18) Vizitiu, D.; Lazar, C.; Halden, B. J.; Lemieux, R. P. J. Am. Chem. Soc.
1999, 121, 8229.
(19) Hartley, C. S.; Lazar, C.; Wand, M. D.; Lemieux, R. P. J. Am. Chem. Soc.
2002, 124, 13513.
(10) (a) Glaser, M. A.; Clark, N. A.; Walba, D. M.; Keyes, M. P.; Radcliffe,
M. D.; Snustad, D. C. Liq. Cryst. 2002, 29, 1073. (b) Glaser, M. A. In
AdVances in the Computer Simulations of Liquid Crystals; Zannoni, C.,
Pasini, P., Eds.; Kluwer: Dordrecht, 1999; p 263.
(11) Goodby, J. W. In Ferroelectric Liquid Crystals: Principles, Properties
and Applications; Goodby, J. W., Blinc, R., Clark, N. A., Lagerwall, S.
T., Osipov, M. A., Pikin, S. A., Sakurai, T., Yoshino, K., Zeks, B., Eds.;
Gordon & Breach: Philadelphia, 1991; p 99.
(20) According to the physics convention, the spontaneous polarization points
from the negative to the positive end of a dipole, which is opposite to that
used in chemistry. A positive PS corresponds to the cross product of the
director and the layer normal, n × z, respectively.9
(21) Vizitiu, D.; Lazar, C.; Radke, J. P.; Hartley, C. S.; Glaser, M. A.; Lemieux,
R. P. Chem. Mater. 2001, 13, 1692.
(22) For a preliminary communication, see: Boulton, C. J.; Sutherland, J. J.;
Lemieux, R. P. J. Mater. Chem. 2003, 13, 644.
9
J. AM. CHEM. SOC. VOL. 127, NO. 39, 2005 13657

A R T I C L E S
Scheme 1 a
Boulton et al.
a Reagents and conditions: (a) NaH, DMF, 60 °C; (b) 70% aqueous H2SO4, 90 °C; (c) AlCl3, toluene, reflux; (d) 1-bromoheptane, K2CO3, acetone,
reflux; (e) semiprep HPLC resolution, Daicel Chiralpak AS column, 2% EtOH/hexanes, 3 mL/min.
2-(bromomethyl)benzoate 6 or 11 gave the substitution
products 7, 12, and 13 in 81-99% yields. Acid-cata-
lyzed hydrolysis was followed by decarboxylation and cycliza-
tion in situ to give the racemic dimethoxyspirobiindan-
diones 8, 14, and 16 in 54-82% yields. Demethylation with
AlCl3 and alkylation of the resulting diols with 1-bromoheptane
gave the racemic dopants 2-4 in 46-84% yields. The enantio-
mers of 2-4 were resolved by semiprep chiral stationary phase
HPLC using a Daicel Chiralcel AS column and 2% ethanol/
hexanes as the mobile phase. The resolved compounds were
recrystallized from hexanes prior to doping in the liquid crystal
hosts.
In each case, the absolute configuration of the first HPLC
eluant was assigned as (R) by the exciton chirality method using
circular dichroism (CD) spectroscopy.24 Previous work by
Schlo¨gl has shown that the absolute configurations of 2,2′-
spirobiindan-1,1′-dione and 5,5′-disubstituted 2,2′-spirobiindan
derivatives can be assigned with a reasonable degree of certainty
according to the sign of split Cotton effects corresponding to
1
1La and Lb exciton couplets.25,26 As shown in Figure 2, the
CD spectra of the first eluants of dopants 2-4 feature strong
positive split Cotton effects at wavelengths corresponding to
Figure 1. Rotation of the biphenyl core about the two ester C-O single
bonds of dopant (R)-1 confined to the binding site of the Boulder model in
UV absorption bands at λmax ) 219, 221, and 222 nm,
respectively. Weaker positive split Cotton effects are also
observed at longer wavelengths in the spectra of 2 and 4. The
electronic transitions of the model compounds 5- and 6-meth-
oxyindan-1-one were calculated using the semiempirical ZINDO
method (Figure 3).27 In each case, the calculations predict a
pair of π-π* transitions with excitation wavelengths in the
210-220 nm range, with one of the two transition dipole
an idealized zigzag conformation. The chiral SmC* phase is represented in
a surface-stabilized ferroelectric state (SSFLC); the layer normal z and the
director n are in the plane of the page and form a tilt angle θ. The
spontaneous polarization Ps is coincident with the C2 symmetry axis of the
SSFLC, which is normal to the plane of the page. The sign of PS conforms
to the physics convention.20
have different steric demands in the condensed phase when
confined to the binding site of the SmC host.
(24) (a) Berova, N.; Nakanishi, K. In Exciton Chirality Methods: Principles
and Applications; Berova, N., Nakanishi, K., Woody, R. W., Eds.; Wiley-
VCH: New York, 2000. (b) Harada, N.; Nakanishi, K. Circular Dichroic
Spectroscopy: Exciton Coupling in Organic Photochemistry; University
Science Books: New York, 1983.
Results
Synthesis. The three chiral dopants were synthesized as
racemic mixtures using a strategy reported by Nieman and
Keay for the synthesis of 2,2′-spirobiindan-1,1′-dione.23
As shown in Scheme 1, reactions of the enolate of either
indanonecarboxylate 5 or 10 with the appropriate ethyl
(25) Falk, H.; Fro¨stl, W.; Hofer, O.; Schlo¨gl, K. Monatsh. Chem. 1974, 105,
598.
(26) Langer, E.; Lehner, H.; Neudeck, H.; Schlo¨gl, K. Monatsh. Chem. 1978,
109, 987.
(27) (a) Zerner, M. C. In ReView in Computational Chemistry; Lipkowitz, K.
B., Boyd, D. B., Eds.; VCH: New York, 1991; pp 313-366. (b) Ridley,
J. E.; Zerner, M. C. Theor. Chim. Acta 1973, 32, 111.
(23) Nieman, J. A.; Keay, B. A. Tetrahedron: Asymmetry 1995, 6, 1575.
9
13658 J. AM. CHEM. SOC. VOL. 127, NO. 39, 2005

Ferroelectric Liquid Crystals Induced by Dopants
A R T I C L E S
Figure 3. Transition dipole moments (π-π*) with the corresponding
excitation wavelengths and oscillator strengths (in parentheses) of 5-meth-
oxyindan-1-one (top) and 6-methoxyindan-1-one (bottom) calculated using
the semiempirical ZINDO method.
Scheme 2 a
a Reagents and conditions: (a) LiAlH4, THF, 25 °C; (b) NaH, THF, 25
°C; (c) 1-(4-N,N-dimethylaminobenzoyl)benzotriazole, THF, reflux.
Figure 4. Circular dichroism (-) and UV (- - -) absorption spectra of
compound 19 (2.0 × 10-5 M) in hexanes.
Figure 2. Circular dichroism (-) and UV (- - -) absorption spectra of (a)
dopant (R)-2 (1.7 × 10-5 M), (b) dopant (R)-3 (2.0 × 10-5 M), and (c)
dopant (R)-4 (1.4 × 10-5 M) in hexanes.
which showed two peaks for the methine protons at 4.94
and 5.51 ppm. In this configuration, the transition dipoles of
the two p-dimethylaminobenzoate groups form an angle of ca.
90°, which is optimum for chiral exciton coupling.24,29 As shown
in Figure 4, the CD spectrum of 19 shows a strong negative
split Cotton effect centered at 298 nm, which indicates that the
two benzoate groups describe a left-handed (M) helix corre-
sponding to the (S) absolute configuration in 4. This result
confirms the assignment of the (R) configuration to the
firstHPLC eluant of 4 and strongly supports the same assign-
ments made for 2 and 3.
moments having a large component normal to the helical axis
of the 2,2′-spirobiindan-1,1′-dione chromophore, which is
consistent with the strong split Cotton effects observed in that
spectral region. According to exciton chirality rules, a positive
split Cotton effect results from coupling of transition dipoles
that describe a right-handed (P) helix, which corresponds to
the (R) absolute configuration in 2-4.
To confirm the stereochemical assignments, the second HPLC
eluant of 4 was reduced stereoselectively to the cis/trans-diol
18 in 74% yield using LiAlH4,28 and then converted to the
bis-p-dimethylaminobenzoate derivative 19 in 58% yield
(Scheme 2). The unambiguous assignment of the cis/trans
configuration in 18 was based on the 1H NMR spectrum,
Dopant-Host Compatibility. The compounds (R)-2-4 were
doped in the liquid crystal hosts PhP1, PhB, DFT, and NCB76,
which exhibit I-N-A-C phase sequences. We observed
significant differences in solubility between the three dopants
in the SmC phases of PhP1, PhB, and DFT, which is
(29) This approach was used to determine the absolute configuration of
unsubstituted 2,2′-spirobiindan-1,1′-dione: Harada, N.; Ai, T.; Uda, H. J.
Chem. Soc., Chem. Commun. 1982, 232.
(28) Dynesen, E. Acta Chem. Scand., Ser. B 1975, 29, 77.
9
J. AM. CHEM. SOC. VOL. 127, NO. 39, 2005 13659

A R T I C L E S
Boulton et al.
to measure the order parameters of deuterated mesogens30 and
to characterize the local environment of nonmesogenic dopants
2
in liquid crystal hosts.31,32 As shown in Figure 6, H NMR
spectra of (RS)-2-d4 (10 mol %) in each of the four hosts at
T-TC ) -10 K feature only quadrupolar doublets with ∆νQ
values on the order of 60-70 kHz, which indicates that all of
the dopant molecules in these mixtures reside in a liquid
2
crystalline environment. Remarkably, the H NMR spectra of
(RS)-3-d4 and (RS)-4-d4 in PhP1 (10 and 2.9 mol %, respec-
tively) feature an isotropic singlet together with a weak
quadrupolar doublet, which suggests that the dopant molecules
reside primarily in isotropic microdomains that cannot be
observed by polarized microscopy.32 No isotropic singlet was
observed with the other liquid crystal mixtures containing (RS)-
3-d4 or (RS)-4-d4. These results suggest that all three dopants
form homogeneous mixtures with PhB, NCB76, and DFT
below the solubility limits established by polarized microscopy.
Most of the spectra feature pairs of quadrupolar doublets,
which are well resolved in the spectra of (RS)-4-d4 (∆∆νQ
)
24-30 kHz), but only partially resolved in the spectra of (RS)-
2-d4 (∆∆νQ ) 0-7 kHz). The spectra of (RS)-3-d4 are more
complex due to its unsymmetrical configuration, but appear to
combine the spectral features of the other two dopants. There
are two likely explanations for the presence of quadrupolar
doublet pairs: (i) a partitioning of the dopant between two
different mesophases, or (ii) a chiral perturbation of the local
environment of the dopant causing the methylene deuterons to
become nonequivalent. The first possibility was ruled out based
on observations of doublet pairs with smaller quadrupolar
splittings in the nematic phase of the 10% mixtures in DFT
(Figure 6), as well as in the nematic phase of a 3 mol % mixture
of (RS)-4-d4 in 4-cyano-4′-pentyloxybiphenyl, which forms only
a nematic phase (see Supporting Information). With respect to
the second possibility, Samulski showed that the enantiotopic
methylene deuterons of perdeuterated benzyl alcohol become
diastereotopic when dissolved in a chiral nematic liquid crystal
formed by a solution of poly-γ-benzyl-L-glutamate (PBLG) in
CH2Cl2.33 This has proven to be a general effect for enantiotopic
CD2 groups in molecules such as perdeuterated butanol, octanol,
and 4-cyano-4′-pentylbiphenyl dissolved in PBLG, which was
attributed to a difference in orientational order parameter of the
two C-D bonds due to a reduction in symmetry of the
anisotropic solute-solvent interaction potential.34 In the case
of (RS)-2-4-d4, the methylene deuterons are formally diaster-
eotopic, but they should be undistinguishable in an achiral liquid
crystal host given their distance to the stereogenic center, if the
dopant molecules are passive. However, a chiral perturbation
exerted by the dopant on the liquid crystal host via core-core
interactions could create the local chiral environment necessary
noteworthy in light of molecular modeling predictions that the
three dopants should have similar molecular shapes and con-
formational distributions (vide infra). Solubility limits were
estimated by polarized microscopy based on the observation of
persistent biphasic SmC*/isotropic domains at 10 K below the
Curie point of a given mixture (T-TC ) -10 K). The solubility
limits in PhP1, PhB, and DFT range from ca. 25 to 30 mol %
for the 5,5′-dialkoxy derivative (R)-2, 10 to 15 mol % for the
5,6′-dialkoxy derivative (R)-3, and only 3 to 5 mol % for the
6,6′-dialkoxy derivative (R)-4. The solubility limits are higher
in NCB76: in excess of 25 mol % for both (R)-2 and 3, and 10
mol % for (R)-4. The phase diagrams in Figure 5 show that
addition of (R)-2 causes a severe destabilization of the SmC
phase formed by PhP1, and a more moderate destabilization
of the SmC phase formed by PhB and DFT. Interestingly,
addition of (R)-2 has virtually no effect on the SmA-C phase
transition temperature of NCB76, which suggests that the
binding site of this host provides a good fit for the chiral dopant.
A comparison of phase diagrams for mixtures of (R)-2, -3, and
-4 in NCB76 reveals a growth in biphasic regions and increasing
mesophase destabilization going from the 5,5′- to the 6,6′-
dialkoxy derivative, which are consistent with the trend in
solubility limit established by polarized microscopy.
(30) Dong, R. Y. In Encyclopedia of Nuclear Magnetic Resonance; Grant, D.
M., Harris, R. K., Eds.; John Wiley & Sons: 1996; Vol. 4, pp 2752-
2760.
2
Dopant-host compatibility was also assessed by H NMR
spectroscopy using the racemic dopants (RS)-2-d4, 3-d4, and 4-d4
with dideuterated side-chains (-OCD2C6H13), which were
mixed in the four liquid crystal hosts below the solubility limits
established by polarized microscopy. In an anisotropic liquid
crystal phase, the interaction of the quadrupolar moment of a
deuterium nucleus with the electric field gradient tensor produces
a quadrupolar doublet with a splitting ∆νQ that is directly
proportional to the orientational order parameter of the C-D
bond. Hence, 2H NMR spectroscopy has been used extensively
(31) (a) Workentin, M. S.; Leigh, W. J. J. Phys. Chem. 1992, 96, 9666. (b)
Workentin, M. S.; Fahie, B. J.; Leigh, W. J. Can. J. Chem. 1991, 69, 1435.
(c) Workentin, M. S.; Leigh, W. J.; Jeffrey, K. R. J. Am. Chem. Soc. 1990,
112, 7329. (d) Fahie, B. J.; Mitchell, D. S.; Leigh, W. J. Can. J. Chem.
1989, 67, 148. (e) Fahie, B. J.; Mitchell, D. S.; Workentin, M. S.; Leigh,
W. J. J. Am. Chem. Soc. 1989, 111, 2916.
(32) (a) Vlahakis, J. Z.; Lemieux, R. P. J. Mater. Chem. 2004, 14, 1486. (b)
Maly, K. E.; Wu, G.; Lemieux, R. P. Liq. Cryst. 2001, 28, 457.
(33) Czarniecka, K.; Samulski, E. T. Mol. Cryst. Liq. Cryst. 1981, 63, 205.
(34) (a) Emsley, J. W.; Lesot, P.; Courtieu, J.; Merlet, D. Phys. Chem. Chem.
Phys. 2004, 6, 5331. (b) Merlet, D.; Loewenstein, A.; Smadja, W.; Courtieu,
J.; Lesot, P. J. Am. Chem. Soc. 1998, 120, 963. (c) Meddour, A.; Canet, I.;
Loewenstein, A.; Pe´chine´, J. M.; Courtieu, J. J. Am. Chem. Soc. 1994,
116, 9652.
9
13660 J. AM. CHEM. SOC. VOL. 127, NO. 39, 2005

Ferroelectric Liquid Crystals Induced by Dopants
A R T I C L E S
Figure 5. Partial phase diagrams for mixtures of (a) (R)-2 in PhP1, (b) (R)-2 in PhB, (c) (R)-2 in DFT, (d) (R)-2 in NCB76, (e) (R)-3 in NCB76, and (f)
(R)-4 in NCB76. The phase transition temperatures were measured by polarized microscopy on cooling; xd is the dopant mole fraction.
Figure 6. 2H NMR spectra (92.13 MHz) of (a) 10 mol % mixtures of (RS)-2-d4 in the four liquid crystal hosts, (b) 10 mol % mixtures of (RS)-3-d4 in the
four liquid crystal hosts, and (c) mixtures of (RS)-4-d4 in the four liquid crystal hosts at the following concentrations: 2.6 mol % (PhP1), 3 mol % (PhB),
5 mol % (NCB76), and 3 mol % (DFT). Each spectrum was taken in the SmC* phase at T-TC ) -10 K, except those in the fifth column, which were taken
in the N* phase.
2
to make the pro-R and pro-S deuterons nonequivalent in a H
NMR spectrum. In such cases, the difference in ∆νQ between
two quadrupolar doublets (∆∆νQ) should reflect the degree of
chiral perturbation exerted by the dopant on the liquid crystal
host. In the host DFT, we observed a decrease in ∆∆νQ going
from the SmC* to the N* phase, which is consistent with a
weakening of chiral perturbations by the dopant due to a lack
of positional order (core-core correlation) in the nematic phase.
According to this rationale, the combined results of the 2H NMR
experiments suggest that the 6,6′-dialkoxy derivative (R)-4 exerts
9
J. AM. CHEM. SOC. VOL. 127, NO. 39, 2005 13661

A R T I C L E S
Boulton et al.
Figure 7. Absolute value of reduced polarization |Po| versus dopant mole
fraction xd for dopants (R)-2 (0), (R)-3 (b), and (R)-4 (O) in NCB76 at
T-TC ) -10 K.
Table 1. Polarization Powers δp of (R)-2, -3, and -4 in the Hosts
PhP1, PhB, NCB76, and DFT at T-TC ) -10 K
δ
p (nC/cm2)a,b
dopant
PhP1
PhB
NCB76
DFT
(R)-2
(R)-3
(R)-4
749 ( 35 (+)
>11 (+)
460 ( 13 (+)
130 ( 7 (-)
79 ( 9 (-)
363 ( 30 (+)
144 ( 7 (-)
1037 ( 100 (-)
21 ( 3 (+)
220 ( 5 (-)
285 ( 26 (-)
>160 (-)
a Sign of polarization in parentheses. b Uncertainty is ( standard error
of least-squares fit.
Figure 8. Reduced polarization Po versus mole fraction of MDW950 in
NCB76 at T-TC ) -10 K: (a) in the presence of (R)-2 (b) and (S)-2 (O)
stronger chiral perturbations in the SmC* phase than the 5,5′-
dialkoxy derivative (R)-2.
at x2 ) 0.05 and (b) in the presence of (R)-4 (b) and (S)-4 (O) at x4
)
0.05. The slope of the dashed lines (-214 nC/cm2) corresponds to the
polarization power of MDW950 in the absence of the chiral dopant.
Polarization Power Measurements. Homogeneous mixtures
of (R)-2, -3, and -4 in the four liquid crystal hosts were aligned
as SSFLC films using commercial ITO glass cells with rubbed
polyimide surfaces and a cell gap of 4 µm. Spontaneous
polarizations (PS) and tilt angles (θ) were measured in the SmC*
phase at T-TC ) -10 K by the triangular wave method,35 and
the corresponding Po values were calculated using eq 2. A
minimum of five different mixtures were prepared for each
dopant/host combination (except for (R)-4 in PhP1 due to low
solubility), and the resulting Po values were plotted as a function
of the dopant mole fraction xd. The plots gave good least-squares
fits (R2 ) 0.928-0.998) except for (R)-3 in NCB76, which gave
a plot showing a positive deviation from linearity at xd > 0.10
(Figure 7). The polarization power values were derived from
the Po versus xd plots using eq 1 and are listed in Table 1; δp
values for (R)-3 and -4 in PhP1 are given as lower limits based
Probe Experiments. Previous work has shown that chiral
perturbations exerted by dopants with chiral cores can be
detected in the SmC* phase using the Displaytech dopant
MDW950 as probe.19 These experiments are based on the
assumption that any long-range perturbation exerted by a chiral
dopant should affect the conformational equilibrium of the probe
and, consequently, its polarization power (δprobe). To determine
whether a perturbation is chiral, δprobe is measured in the
presence of each enantiomer of the chiral dopant; if they exert
chiral perturbations, the two enantiomers should have different
effects on δprobe. Hence, to test the hypothesis that (R)-4 may
exert stronger chiral perturbations in the SmC* phase than (R)-
2, we performed a series of probe experiments with MDW950
in the host NCB76, in which the largest difference in ∆∆νQ
between (RS)-2-d4 (0 kHz) and (RS)-4-d4 (30 kHz) was
observed. The reduced polarization induced by MDW950 was
measured as a function of its mole fraction x950 in the presence
of the (R) and (S) enantiomers of each chiral dopant at a constant
mole fraction of 0.05 (5 mol %). As shown in Figure 8, the
resulting plots of Po versus x950 are shifted up and down the
y-axis by values equal to Po induced by the (R) and (S)
enantiomers of (R)-2 and -4 in the absence of the probe. The
2
on the H NMR findings that the dopants reside primarily in
isotropic microdomains. The sign of PS along the polar axis
was assigned from the relative configuration of the electric field
and the switching position of the SSFLC film according to the
established convention.9 As shown in Table 1, the polarization
power of (R)-2 is uniformly positive and varies significantly
with the host structure. Conversely, the polarization power of
(R)-4 is uniformly negative and shows a different trend in host
dependence, with the highest δp value of -1037 nC/cm2
obtained in NCB76. The unsymmetrical dopant (R)-3 seems to
behave as a hybrid of the two symmetrical isomers (R)-2 and
-4. Its polarization power is lower, on average, and the sign of
PS is not the same in all four hosts, which is consistent with
competing conformational biases (vide infra).
results show that the polarization power of MDW950 (δ950
)
increases by the same amount, within error, in the presence of
either (S)-2 or (S)-4 (δ950 ) -355 ( 37 and -386 ( 19 nC/
cm2, respectively). The plots obtained in the presence of (R)-2
and (R)-4 also indicate an increase in δ950, but they both appear
to deviate from linearity. These results suggest that the two
dopants exert similar long-range perturbations in NCB76, which
are, at best, weakly chiral in comparison to the chiral perturba-
(35) Miyasato, K.; Abe, S.; Takezoe, H.; Fukuda, A.; Kuze, E. Jpn. J. Appl.
Phys. 1983, 22, L661.
9
13662 J. AM. CHEM. SOC. VOL. 127, NO. 39, 2005

Ferroelectric Liquid Crystals Induced by Dopants
A R T I C L E S
Figure 9. Space-filling models showing conformational distributions of dopant (R)-2 as side-on views with the spirobiindandione core in planar and C2
orientations. The heats of formation (∆Hf) and dipole moments along the polar axis (µ ) were calculated for each minimized structure at the AM1 level. The
µ
vectors point from negative to positive according to the physics convention.20 The chiral axis of the spirobiindandione core is congruent with the tilt plane
(plane of the page), and the polar axis is normal to the plane of the page.
tions exerted by (R)-1 in PhP1 according to a similar probe
that may be responsible for the induced polarization.36 In each
case, the analysis begins with the minimized conformation that
best “fits” the zigzag binding site. The core is then rotated by
90° increments about its chiral axis, and each structure is
minimized at the AM1 level after the side-chains are rotated to
give the closest fit to the zigzag binding site. As shown in Figure
9, two pairs of degenerate P conformations are obtained, PI/I′
and PII/II′,37 with equal but opposite dipole moments along
the polar axis (µ ). Although the difference in energy between
PI/I′ and PII/II′ is insignificant in the gas phase, the Boulder
model predicts that PII/II′ should be disfavored because its bent
shape does not conform to the zigzag binding site. Similarly,
the conformations C2I and C2III have equal but opposite dipole
experiment (vide supra).19
Conformational Analyses. To understand the molecular
origins of the polarization induced by (R)-2-4 and the
dependence of δp on the structure of the host, conformational
analyses of the three dopants are carried out at the AM1 level
within the framework of the Boulder model; that is, the chiral
dopants were considered to be passive guests that conform to
the achiral binding site of the host in the SmC phase (see Figure
1). In this model, we assume that each dopant adopts a zigzag
conformation with the core more tilted than the side-chains, that
the alkoxy side-chains are fully extended with the methylene
groups in anti conformations, and that the ether linking groups
are coplanar with the aromatic rings. The calculations suggest
that the three dopants are in equilibrium between two conforma-
tions with different core orientations, as shown in Figure 9. In
one conformation (P), the plane of one indanone fragment is
congruent with the tilt plane defined by n and z; in the other
conformation (C2), the C2 axis of the core is coincident with
the polar axis. An analysis of conformational distributions for
the two core orientations of (R)-2 reveals a conformational bias
(36) These gas-phase conformational distributions are oversimplifications of
complex conformational/orientational hypersurfaces and neglect dopant-
host intermolecular interactions, but they do provide a useful basis to
understand the molecular origins of Ps.9 More rigorous analyses based on
atomistic molecular dynamics simulations are currently under develop-
ment: Ghenea, R.; Cann, N. M., unpublished results. A related Monte Carlo
molecular simulation method that calculates helical twisting powers based
on fully atomistic structures of chiral dopants in a Gay-Berne nematic
liquid crystal solvent has recently been reported: Earl, D. J.; Wilson, M.
R. J. Chem. Phys. 2004, 120, 9679.
(37) The two conformations in each degenerate pair can be interconverted by a
180° rotation about the polar axis.
9
J. AM. CHEM. SOC. VOL. 127, NO. 39, 2005 13663

A R T I C L E S
Boulton et al.
Figure 10. Space-filling models showing minimized zigzag conformations of dopants (R)-2, -3, and -4 as end-on views with the spirobiindandione core in
planar and C2 orientations. The heats of formation (∆Hf) and dipole moments along the polar axis (µ ) were calculated for each minimized structure at the
AM1 level. The tilt plane is vertical and normal to the plane of the page, and the polar axis is horizontal and in the plane of the page. The µ vectors point
from negative to positive according to the physics convention.20
moments along the polar axis, but the latter is disfavored because
of its linear shape. The other two degenerate conformations have
no dipole moment along the polar axis and may therefore be
neglected. Overall, the analysis suggests that the spontaneous
polarization induced by (R)-2 depends on the position of the
equilibrium between the zigzag conformations PI/I′ and C2I,
which have dipole moments in opposite directions along the
polar axis.
C2, and that a finer balance between P3 and P3′ may be
responsible for the polarization inversion.38
Results of the probe experiments in NCB76 suggest that
neither (R)-2 nor (R)-4 exert strong, long range chiral perturba-
tions comparable to those exerted by (R)-1 in PhP1 under the
same experimental conditions.19 However, the 2H NMR experi-
ments strongly suggest that both (R)-3 and -4 exert short range
chiral perturbations that might still influence the conformational
distribution of the dopant according to the CTF model. The
positive deviation from linearity observed in the Po versus xd
plot for (R)-3 in NCB76 at relatively high mole fractions (xd >
0.10) suggests a cooperative effect that is consistent with the
CTF model. Unfortunately, we were unable to measure polar-
izations in that higher mole fraction range with (R)-3 in the
other hosts, or with (R)-4 in any host due to solubility limits.
On the other hand, the linearity of the Po versus xd plots for
(R)-2 at mole fractions as high as 0.25, together with the small
Similar conformational analyses were carried out for (R)-3
and -4; the corresponding P and C2 zigzag conformations are
shown in Figure 10 as end-on views along with those of (R)-2.
The conformational distribution of the symmetrical dopant (R)-4
is essentially the same as that of (R)-2, but predicts an induced
polarization of opposite sign. The distribution of the unsym-
metrical dopant (R)-3 is more complex; the C2 core orientation
has two degenerate zigzag conformations, and the planar core
orientation has two nondegenerate zigzag conformations with
µ of opposite signs. In each case, alkoxy and carbonyl groups
contribute to the dipole moment along the polar axis, either
constructively or (partially) destructively, except in the case of
conformation C23, in which the transverse dipole moments of
the two carbonyl groups cancel out.
2
∆∆νQ values observed in the corresponding H NMR spectra,
suggest that (R)-2 exerts very weak, perhaps negligible, chiral
perturbations in the four liquid crystal hosts.
If one assumes that the CTF effect does not contribute to the
polarization induced by (R)-2, the dependence of δp on the
nature of the host may be understood in terms of a shift in the
P/C2 conformational equilibrium that is controlled primarily by
a difference in conformational steric demand in the core region
of the binding site. If one simplifies the binding site of the
Boulder model to three cylindrical sections forming a zigzag
shape, steric demand in the core region may be approximated
by the distance between atomic coordinates where divergence
from the central cylindrical space begins. As shown in Figure
11, this approximation reveals that the core section of the P
conformation is ca. 1.5 Å shorter than that of the C2 conforma-
tion, which suggests that the P/C2 conformational equilibrium
should shift toward the P conformation (positive µ ) as the core
Discussion
The results described herein paint a rather complex picture
of the dependence of δp on the structure of the dopant and the
nature of the liquid crystal host, which includes significant
differences in dopant-host compatibility, and in the degree of
chiral perturbations exerted by the dopants on the achiral SmC
phase. Let us first consider the general trends in sign and
magnitude of induced polarization with respect to dopant
structure. The 5,5′-dialkoxy dopant (R)-2 induces a positive
polarization in all four hosts, whereas the 6,6′-dialkoxy dopant
(R)-4 induces a negative polarization in all four hosts, which is
consistent with the results of the conformational analysis and
suggests that P conformations are generally favored over C2 in
the SmC binding site. The δp data for the “hybrid” dopant (R)-3
is consistent with a more complex conformational distribution,
as predicted by the conformational analysis. The data for (R)-3
may also be rationalized by assuming a predominance of P over
(38) The reason P3 would be favored over P3′ in three of the four hosts remains
unclear. In conformation P3′, the 6-heptyloxyindanone fragment is oriented
perpendicular to the tilt plane, which may be less compatible with the SmC
binding site than the 5-heptyloxyindanone fragment in the same orientation.
Interestingly, the perpendicular orientation of the 6-heptyloxyindanone
fragment is also present in P4, but it cannot be avoided, as in (R)-3, by
switching to another zigzag P conformation. This would be consistent with
the remarkably low miscibility of (R)-4 relative to (R)-3 and (R)-2.
9
13664 J. AM. CHEM. SOC. VOL. 127, NO. 39, 2005

Ferroelectric Liquid Crystals Induced by Dopants
A R T I C L E S
Conclusions
The compounds (R)-2, -3, and -4 represent a new class of
dopants with axially chiral cores for the induction of ferroelectric
liquid crystals, with polarization powers that strongly depend
on the core structure of the liquid crystal host. We have shown
that relatively minor changes in the dopant structure, that is,
moving the alkoxy side-chains from the 5,5′ to the 6,6′ positions
of the spirobiindandione core, have profound effects on the
induced spontaneous polarization (sign and magnitude), on
dopant-host compatibility, and on the propensity of the dopant
to exert chiral perturbations on the host environment. We have
2
also demonstrated the usefulness of H NMR spectroscopy to
probe the local environment of dopant molecules and assess
their miscibility in a liquid crystal host, as well as to detect
short-range chiral perturbations exerted by a dopant on the host
environment. The variations in sign and magnitude of δp as a
function of alkoxy group positions may be rationalized on the
basis of an analysis of zigzag conformations that conform to
the binding site of the SmC phase according to the Boulder
model. The analysis suggests that δp is a function of an
equilibrium between two zigzag conformations with opposite
dipole moments along the polar axis of the SmC* phase and
that the observed host dependence may be understood in terms
of differences in steric demand between the two conformations
in the core region of the binding site, a new form of molecular
recognition in smectic phases. A closer examination of the two
opposite conformations also reveals that it may be possible to
further bias the conformational equilibrium toward the P
conformer, and therefore test our conformational distribution
model, by introducing a substituent at the 6-position of one
indanone fragment in (R)-2, for example. The substituent would
force one alkoxy group in an anti-periplanar conformation
relative to the carbonyl group, which is only present in the P
conformer (Figure 9). The study of such derivatives is in
progress and will be reported in due course.
Figure 11. Space-filling models of the conformations P (left) and C2 (right)
for (R)-2 in relation to a simplified form of the zigzag binding site according
to the Boulder model.
section of the host binding site becomes shorter. The trend in
polarization power versus host structure for (R)-2, which shows
δp increasing from +21 nC/cm2 in the host with the longest
core (DFT, 14.4 Å) to +749 nC/cm2 in the host with the shortest
core (PhP1, 9.7 Å), is consistent with this model.
The trends in polarization power versus host structure are
different in the cases of (R)-3 and (R)-4, which may be due to
the added contribution of the CTF effect. Unfortunately, it is
difficult to quantify the contribution of this effect, and how it
might affect the conformational equilibrium of the chiral dopant,
and the rotational distribution of its transverse dipole moment
relative to the polar axis in a given liquid crystal host.2,14,15 To
rationalize the dependence of δp on the alkoxy group positions,
the data in NCB76 are deemed to be most reliable given the
relatively high compatibility of this host with all three chiral
dopants. The polarization power of (R)-4 is 2.8 times greater
than that of (R)-2 in NCB76, which is consistent, to a first
approximation, with the difference in dipole moments µ
between the corresponding P conformations (-3.5 D vs +1.6
D, respectively). Notwithstanding the series of atropisomeric
dopants that includes (R)-1, the δp value of -1037 nC/cm2 for
(R)-4 in NCB76 is the second highest reported thus far for a
calamitic dopant.39 In the case of (R)-3 in NCB76, the smaller
δp value of -144 nC/cm2 is consistent with the more complex
conformational distribution predicted by the conformational
analysis, which includes competing P conformations of opposite
µ (the dominant P3 conformation has the smaller µ value of
-1.6 D in this case).
Acknowledgment. We are grateful to the Natural Science
and Engineering Research Council of Canada, the Canada
Foundation for Innovation, and the Ontario Challenge Fund for
support of this work. We also thank Profs. Ed Samulski and
Elliott Burnell and Dr. Louis Madsen for useful discussions of
2
the H NMR results.
Supporting Information Available: Full experimental details
including synthetic procedures, physical measurements and
calculation methods, 2H NMR spectrum of (R)-4 in 5OCB, and
Po versus xd plots for (R)-2, -3, and -4 in all four hosts. This
material is available free of charge via the Internet at
http://pubs.acs.org.
(39) A polarization power of 1508 nC/cm2 was reported for a derivative of
2-hydroxy-5,5-dialkyl-δ-valerolactone: Sakashita, K.; Ikemoto, T.; Na-
kaoka, Y.; Terada, F.; Sako, Y.; Kageyama, Y.; Mori, K. Liq. Cryst. 1993,
13, 71.
JA054322K
9
J. AM. CHEM. SOC. VOL. 127, NO. 39, 2005 13665