Influence of flexible spacers on liquid-crystalline self-assembly of T-shaped bolaamphiphiles

Article abstract of DOI:10.1021/ja200822z

T-shaped bolaamphiphiles composed of a biphenyl rigid core, a semiperfluorinated lateral chain, two polar 1,2-diol groups in the terminal positions and flexible alkyl spacers connecting the polar groups with the biphenyl core have been synthesized and investigated by polarizing microscopy, DSC and X-ray scattering. The influence of spacer length and position of the spacer on the self-assembly in liquid-crystalline phases was studied. A series of four different columnar phases (Col_hex/p6mm, Col _rec/p2gg, Col_squ/p4gm and Col_squ/p4mm), representing liquid-crystalline honeycomb structures composed of cylinders having hexagonal, pentagonal, and square cross section, were found on increasing the spacer length. It is also shown that introduction of aliphatic spacers in the backbone of the T-shaped bolaamphiphiles replaces the Col_rec/c2mm phase made up of rhombic cylinders with the Col_squ/p4mm phase composed of square cylinders. It also causes the 2d lattice of pentagonal cylinders to increase the symmetry from Col_rec/p2gg to Col _squ/p4gm. A temperature-dependent second-order phase transition between these two pentagonal cylinder structures was observed for the first time. Beside these effects on cylinder shape and phase symmetry the flexible spacer units also lead to reduced phase transition temperatures and allow adjustment of cylinder side length to envelop a wider range of side-chain sizes. Electron density maps suggest that this may involve sacrificing some of the hydrogen bonds.

Full text of DOI:10.1021/ja200822z

ARTICLE
pubs.acs.org/JACS
Influence of Flexible Spacers on Liquid-Crystalline
Self-Assembly of T-Shaped Bolaamphiphiles
Xiaohong Cheng,*,†,‡ Feng Liu,§ Xiangbing Zeng,§ Goran Ungar,*,§,^ Jens Kain,|| Siegmar Diele,||
||
Marko Prehm,‡, and Carsten Tschierske*,‡
†Department of Chemistry, Key Laboratory of Medicinal Chemistry for Natural Resources (Ministry of Education),
Yunnan University, Kunming, Yunnan 650091, P. R. China
‡Organic Chemistry and Physical Chemistry, Institute of Chemistry, Martin-Luther University Halle-Wittenberg, D-06120 Halle,
Germany
§Department of Materials Science and Engineering, University of Sheffield, Mappin Street, Sheffield S13JD, U.K.
^WCU program Chemical Convergence for Energy and Environment, School of Chemical and Biological Engineering,
Seoul National University, Seoul, Korea
S Supporting Information
b
ABSTRACT: T-shaped bolaamphiphiles composed of a biphenyl
rigid core, a semiperfluorinated lateral chain, two polar 1,2-diol
groups in the terminal positions and flexible alkyl spacers
connecting the polar groups with the biphenyl core have been
synthesized and investigated by polarizing microscopy, DSC
and X-ray scattering. The influence of spacer length and
position of the spacer on the self-assembly in liquid-crystalline
phases was studied. A series of four different columnar phases
(Colhex/p6mm, Colrec/p2gg, Colsqu/p4gm and Colsqu/p4mm),
representing liquid-crystalline honeycomb structures composed
of cylinders having hexagonal, pentagonal, and square cross section, were found on increasing the spacer length. It is also shown that
introduction of aliphatic spacers in the backbone of the T-shaped bolaamphiphiles replaces the Colrec/c2mm phase made up of
rhombic cylinders with the Colsqu/p4mm phase composed of square cylinders. It also causes the 2d lattice of pentagonal cylinders to
increase the symmetry from Colrec/p2gg to Colsqu/p4gm. A temperature-dependent second-order phase transition between these
two pentagonal cylinder structures was observed for the first time. Beside these effects on cylinder shape and phase symmetry the
flexible spacer units also lead to reduced phase transition temperatures and allow adjustment of cylinder side length to envelop a
wider range of side-chain sizes. Electron density maps suggest that this may involve sacrificing some of the hydrogen bonds.
polygonal cylinder phases. In these soft honeycomb structures
the rodlike cores organize into cylinder frameworks, held together
by attractive interactions between the terminal groups located at
the edges of the cylinder walls and the resulting cells are filled with
the disordered lateral chains.15 In these structures the length of
the rigid core with respect to the volume of the lateral chains
determines the number of rodlike cores organized in the circum-
ference of the cylinders. In this way, increasing the volume of the
lateral chains at constant core length leads to a series of cylinder
structures starting with rhombuses, followed by pentagons, regular
and stretched hexagons,19 and ending with giant pentagons
(Figure 1cꢀg),12,15 followed by a bursting of these cylinders17b with
formation of a new kind of lamellar phases (Lam, Figure 1hꢀk).20
For the series of bolaamphiphilic biphenyl derivatives with
lateral alkyl (Hm-1/1) or semiperfluoroalkyl substituents
1. INTRODUCTION
The investigation of molecular self-assembly is one of the most
exciting areas of contemporary chemical research.1 Liquid-crystal-
line (LC) systems2,3 combine order and mobility and therefore
can be regarded as simple model systems in which the driving
forces of self-assembly at the nanoscale level can be studied in a
systematic and controlled manner.4 LC dendrimers5,6 and
supermolecules,7 chirality-induced LC superstructures,8 formation
of ferroelectric organic materials and spontaneous achiral symme-
try breaking in mesophases formed by bent-core molecules,9 as
well as LC magnets10 and liquid quasicrystals11 are selected topics
of contemporary research in this rapidly developing field.
Based on the concept of competitive polyphilicity, it was also
possible to increase the complexity of LC phases considerably.12ꢀ16
This concept was successfully applied to two types of T-shaped
ternary amphiphiles incorporating a rigid rodlike core, namely
bolaamphiphiles12,13,15,17 and facial amphiphiles.15,18 A major
group of LC phases formed by these polyphilic tectons represent
Received: January 27, 2011
Published: May 02, 2011
r
2011 American Chemical Society
7872
dx.doi.org/10.1021/ja200822z J. Am. Chem. Soc. 2011, 133, 7872–7881
|

Journal of the American Chemical Society
ARTICLE
Figure 1. Mesophases reported for calamitic biphenyl-based bolaamphiphiles with nonpolar (alkyl or semiperfluoroalkyl) lateral chains in the
peripheral 3-position; compounds Fm-1/1: R = CmF2mþ1; compounds Hm-1/1: R = CmH2mþ1; abbreviations: SmA = smectic A phase; Colrec
=
rectangular columnar phase; Colsqu = square columnar phase; Colhex = hexagonal columnar phase; LamIso, LamN, LamSm, laminated phases with
different orders of the rodlike units in the layers.15.
Chart 1. Structure and Designation of the Molecules under
Investigation
Figure 2. Molecular structure, transition temperatures (T/ꢀC), and
molecular self-assembly of compound F04-1/1 in its Colsqu/p4mm type
liquid-crystalline phase; the brackets indicate that the Colsqu/p4mm
phase is monotropic and can only be observed on cooling from the
isotropic liquid state.13
Furthermore, the influence of spacer length and position on the
mode of LC self-assembly is discussed. In the compounds denoted
Fm-n1/n2, m indicates the number of CF2 groups in the lateral
chain and n1/n2 give the spacer lengths used to couple the 1,2-diol
units to the two different ends of the biphenyl core (see Chart 1).
It is shown that with increasing spacer length (n1, n2), rhombic
cylinders are replaced by square cylinders, resulting in the
following sequence of four distinct honeycomb LC phases:
(Fm-1/1), shown in Figure 1, a transition from the SmA phase
via a c2mm-type rectangular columnar phase composed of
rhombic cylinders (Figure 1c) to a pentagonal cylinder phase
with p2gg symmetry (Figure 1d) was observed as the first steps
upon elongation of the lateral chain (increasing m).12,13 The
c2mm phase (Figure 1c) can be regarded as a slightly distorted
version of the square cylinder phase (Colsqu/p4mm, see Figure 2,
right) with angles unequal to 90ꢀ. The nondistorted p4mm phase
was observed for ternary T-shaped polyphiles with longer aro-
matic cores, such as p-terphenyls,18,21 diphenylthiophenes22 and
even longer rodlike cores.23,24 For the biphenyl derivatives, only
in one exceptional case a short monotropic (metastable) region
of the square cylinder phase (Colsqu/p4mm) could be achieved
after shifting the position of the lateral chain from the peripheral
3-position to the more central 2-position (compound F04-1/1 in
Figure 2).13,25 Hence, it appears that the distance between lateral
chain and polar groups decides whether cylinders with a square
or a rhombic cross section are formed.
Colhex/p6mm (hexagons) ꢀ Colrec/p2gg (pentagons), Colsqu
/
p4gm(pentagons) ꢀ Colsqu/p4mm (squares). Moreover, for the
first time a temperature-induced transition between two penta-
gonal cylinder phases with symmetries p2gg and p4gm was
observed in one of the compounds.
2. RESULTS AND DISCUSSION
2.1. Synthesis. Bolaamphiphiles Fm-n1/n2 were prepared as
shown in Scheme 1 using Suzuki coupling reactions26 as key
steps. The 4-bromophenols 3-Fm (m = 6,8) with perfluorinated
lateral chains attached via a propylene spacer to the 2-position
were synthesized by Pd0-catalyzed addition of 1-iodoperfluor-
oalkanes to 2-allylphenol,27 followed by para-selective bromina-
tion using HBr/AcOH/DMSO.28 Starting with compounds
3-Fm the building blocks 8-Fm-n2 (m = 6, 8 and n2 = 1, 4, 9)
were produced in three different ways. Etherification of the
Here we report new T-shaped bolaamphiphiles with a semi-
perfluorinated lateral chain but having an additional aliphatic
spacer between the rigid aromatic core and the polar groups.
7873
dx.doi.org/10.1021/ja200822z |J. Am. Chem. Soc. 2011, 133, 7872–7881

Journal of the American Chemical Society
ARTICLE
a
Scheme 1. Synthesis of Bolaamphiphiles Fm-n1/n2
by quenching with trimethylborate at ꢀ90 ꢀC. Hydrolysis to the
boronic acids 10-n1 was carried out at 0 ꢀC using phosphate
buffer adjusted to pH = 4.5ꢀ5.12 The coupling reactions of the
boronic acids 10-n1 with the appropriate aryl bromides 8-Fm-n2
were carried out using standard conditions with Pd(PPh3)4, in
aqueous glyme and NaHCO3 as base. In the final step, the
acetonide groups were cleaved (HCl in EtOH) and the bolaam-
phiphiles Fm-n1/n2 were purified by repeated crystallization.
Detailed synthetic procedures and analytical data are given in the
Supporting Information (SI).
2.2. Methods of Investigation and General Trends. Com-
pounds Fm-n1/n2 were investigated by polarized optical micro-
scopy (POM; Optiphot 2, Nikon in conjunction with a heating
stage FP 82 HT, Mettler), differential scanning calorimetry
(DSC; DSC-7, Perkin-Elmer), and X-ray diffraction (XRD) of
powder samples and surface aligned samples. Investigations of
surface aligned samples were performed using a 2D detector
(HI-Star, Siemens). Uniform orientation was achieved by slowly
cooling a drop of the compound on a glass surface. The X-ray
beamwas appliedparalleltothesubstrate (exposuretime:between
30 min and 3 h).
Two series of compounds were prepared and investigated,
series F6-n1/n2 with a C6F13 perfluorinated segment, and series
F8-n1/n2 with a C8F17 segment, attached to the lateral chain. The
phase transition temperatures and corresponding enthalpy
values for both series are collated in Table 1, together with other
data. All compounds form broad regions of enantiontropic
(thermodynamically stable) liquid-crystalline phases. In the
mesomorphic temperature ranges the X-ray diffraction patterns
show a broad diffuse wide angle scattering with maxima at
d = 0.52 nm for compounds F6-n1/n2 (Figure 3g) and at
d = 0.54 nm for compounds F8-n1/n2 (Figure S6e, SI) confirm-
ing the LC nature of all mesophases. This diffuse scattering has a
ring-like shape and the rather large d-value of the maxima indi-
cate that the RF segments significantly contribute to this scattering.
Indexing of the small-angle Bragg reflections (see Table S1 in SI)
leads to the phase assignments with plane groups shown in Table 1.
2.3. Effect of Spacer Length on Mesomorphic Self-Assem-
bly in the Series of Compounds F6-n1/n2. For the series of
compounds F6-n1/n2 with the shorter fluorinated segment four
distinct types of columnar phases were observed depending on
the spacer length, the distribution of the spacers along the
lipophilic core moiety and temperature. The columnar phase
of compound F6-1/1 with two glycerol-type polar end groups
(n1, n2 = 1) was previously reported as a hexagonal columnar
phase composed of an array of hexagonal cylinders (see Table 1
and Figure 1e).13
Compound F6-1/4 is obtained by increasing the length (n2) of
the spacer adjacent to the benzene ring carrying the RF chain to
n2 = 4. For this compound a birefringent texture appears at
T = 143 ꢀC on cooling from the isotropic liquid. This texture, shown
in Figure 3a, is typical of a columnar phase, characterized by
birefringent domains and homeotropically aligned areas which
appear completely dark, indicating that this mesophase is opti-
cally uniaxial. Investigation with a λ-retarder plate (Figure 3c)
leads to the conclusion that this mesophase is optically
negative,34 like those of all Fm-n1/n2 compounds. This means
that the major intramolecular π-conjugation pathway along the
long axis of the biphenyl cores is perpendicular to the column
long axis. This mesogen orientation is common to most poly-
gonal cylinder phases. On further cooling (at T = 129 ꢀC) weak
birefringence appears in the homeotropically aligned regions.
a Reagents and conditions: i) K2CO3, CH3CN, reflux, 6 h (80ꢀ93%). ii)
OsO4, N-methylmorpholine-N-oxide, H2O, acetone, 25 ꢀC (52-66%).
iii) Me2C(OMe)2, Py,TosOH, CH2Cl2, 25 ꢀC, 24 h (73ꢀ93%). iv) 1. n-
BuLi, ꢀ100 ꢀC; 2. B(OMe)3; 3. phosphate buffer, 0 ꢀC (43ꢀ60%). v)
Pd(PPh3)4, NaHCO3, glyme, H2O, reflux, 6 h (25ꢀ92%). vi) 10% HCl,
EtOH, 25 ꢀC, 24 h (90ꢀ95% crude products, 12ꢀ22%, after all purification
steps).
4-bromophenols 3-Fm with 4-(4-bromobutyl)-2,2-dimethyl-1,
3-dioxolane 529 afforded compounds 8-Fm-4 with buty-
lene spacer. Etherification of the 4-bromophenols 3-Fm
with 11-bromo-undecane-1,2-diol 7 gave the substituted
11-(4-bromophenoxy)-undecane-1,2-diols 6-Fm-9 in the first
step. Protection of the diol groups of the obtained compounds
6-Fm-9 using 2,2-dimethoxypropane and catalytic amounts of
pyridinium tosylate produced the bromoarenes 8-Fm-9 with the
longest spacer unit (n2 = 9).30 For the preparation of the
bromoarenes 8-Fm-1 with the shortest spacers, we used the
etherification of 4-bromophenols 3-Fm first with allylbromide,
followed by dihydroxylation of the double bond with osmium-
tetroxide/NMMNO (N-methylmorpholine-N-oxide) in an acet-
one/water mixture,31 and protection of the diol group as the final
step. The glycerol functionalized benzene boronic acids 10-n1
(n1 = 1, 4) were prepared by halogen/metal exchange reaction of
the arylbromides 9-n132 with n-butyllithium at ꢀ100 ꢀC followed
7874
dx.doi.org/10.1021/ja200822z |J. Am. Chem. Soc. 2011, 133, 7872–7881

Journal of the American Chemical Society
ARTICLE
Table 1. Transition Temperatures, Associated Enthalpy Values (lower lines, in italics) and Lattice Parameter and Other Data of
the T-Shaped Bolaamphiphiles Fm-n1/n2.a
a Transition temperatures were determined by DSC (peak temperatures, first heating scan, 10 K minꢀ1); Lmol = molecular length as measured between
the ends of the primary OH groups at both ends in the most extended conformation of the molecules (CPK models); ncell = number of molecules in the
unit cell, calculated using the crystal volume increments of Immirzi33 as described in the SI (Table S2); “polygon” describes the shape of the cylinder
cross section; nwall = average thickness of the cylinder walls given by the average number of biphenyl units arranged side by side in the walls separating two
adjacent cylinder cores, calculated by dividing ncell by the number of cylinder walls in the unit cell, i.e. for Colsqu/p4mm, ncell is divided by 2, for Colhex
/
p6mm by 3, and for Colrec/p2gg/Colsqu/p4gm by 10; abbreviations: Cr = crystalline solid, Colsqu/p4mm = square columnar phase with p4mm lattice
(square honeycomb), Colhex/p6mm = hexagonal columnar phase with p6mm lattice (hexagonal honeycomb), Colsqu/p4gm = square columnar phase
with p4gm lattice, Colrec/p2gg = rectangular columnar mesophase with p2gg lattice (pentagonal honeycombs), M = unknown mesophase, Iso = isotropic
liquid. b Values for Colsqu/p4gm phase at T = 130 ꢀC . c Values for Colrec/p2gg phase at T = 120 ꢀC.
The birefringence increases continuously upon further cooling,
and a birefringent stripy pattern develops (Figure 3b). At this
transition the orientation of the columns does not change
(compare c and d of Figure 3) and there is no associated enthalpy
change (Figure 3h), which suggests a second-order or only
weakly first-order phase transition.
Small angle XRD patterns of an aligned sample recorded at
T = 130ꢀ135 ꢀC can be indexed on a square lattice with p4gm
symmetry (see Figure 3e; asqu = 6.2 nm at T = 130 ꢀC). This
changes below 130 ꢀC to a rectangular lattice with p2gg sym-
metry and parameters a = 6.5 nm and b = 6.0 nm at T = 120 ꢀC
(see Figure 3f and Tables S1, S2 in SI). In previously reported
T-shaped amphiphiles the columnar phases with p4gm lattice
were found to display two different honeycomb structures, one
combining square and triangular cylinders in a 1: 2 ratio, and the
other consisting of pentagonal cylinders only.18c The triangle-
square structure can be excluded for the present case since the
p4gm phase of F6-1/4 changes into a hexagonal cylinder phase if
the length of the bolaamphiphilic core is reduced by decreasing
the spacer length (compound F6-1/1) and to a square cylinder
phase if the bolaamphiphilic core is enlarged by increasing the
spacer length (compounds F6-4/4 and F6-1/9, see Table 1). It
follows that the cylinders forming the p4gm phase of compound
F6-1/4 must have a shape between square and hexagon but not
triangle, which therefore excludes the triangle-square tiling.
Hence, the present p4gm phase is likely to represent an array
of cylinders with pentagonal cross section made up of five
aromatic cores in circumference. The cylinders are grouped in
pairs and arranged in a herringbone pattern. The electron density
map (Figure 3i) reconstructed from the diffraction intensities
(Figure 3e) shows a cut through this cylinder structure. The high
electron density areas (blue, purple) indicate the position of the
columns incorporating the RF segments of the lateral chains and
in the green shells the alkyl parts of the lateral chains are partly
mixed with the RF chains. These high electron density areas are
centered in the pentagonal cylinder cells; the low electron
density areas (red, yellow) mark the positions of the polar
(diol) columns and spacer units. The superimposed black lines
indicate the average positions of the biphenyls interconnecting
these polar columns. The pentagons are not regular, as regular
pentagons cannot tile a plane. The p2gg phases of T-shaped
bolaamphiphiles are also known to consist of pentagonal
7875
dx.doi.org/10.1021/ja200822z |J. Am. Chem. Soc. 2011, 133, 7872–7881

Journal of the American Chemical Society
ARTICLE
cylinders, but the pentagons are even more distorted than in the
p4gm phase (see Figure 3j).13 An estimate of the number ncell of
molecules in each hypothetical unit cell with an assumed height
of h = 0.45 nm was obtained by comparing the unit cell
volume with the molecular volume as calculated from volume
increments33 (the reason for choosing h = 0.45 nm is explained in
a footnote35 and for details of the calculations, see Table S2 in SI).
The resulting values are ncell =19ꢀ20 and ncell =20forthep4gm and
p2gg phases, respectively. These values are in good agreement with
those previously reported for pentagonal cylinder phases of other
compounds.13,18c,18d,22 Considering there are 10 cylinder walls in a
unit cell the above ncell values indicate that the walls are formed by
two aromatic cores arranged side-by-side (nwall = 1.9ꢀ2.0).
It is a characteristic of second-order phase transitions that
there are sizable fluctuations on both sides of the transition
temperature Tc. Thus just above Tc the square phase is expected
to contain local domains of the rectangular p2gg phase with equal
probability of the long rectangle axis being aligned along x1 and
x2 axes of the p4gm square lattice. Similarly, just below Tc there
are local domains of the rectangular p2gg phase oriented per-
pendicular to the dominant direction elsewhere in the lattice. As a
result, upon cooling a uniformly aligned sample of the p4gm
phase (Figure 3e), the p2gg phase forms as a domain structure
composed of two nearly equally populated domains aligned in
mutually perpendicular directions (Figure 3f). Ease of deforma-
tion of the pentagonal cylinders should lower the energy of the
domain boundaries and thus facilitate the formation of the
entropically favored square phase at elevated temperature. It is
therefore not surprising that the transition between the p2gg and
p4gm pentagonal cylinder phases is observed here for the first
time; the long spacer as part of the molecular backbone allows
flexibility in wall length, which lowers the energy of the domain
boundaries. In contrast, the related compounds Fm-1/1 without
long spacers form exclusively the p2gg pentagonal honeycomb.
Hence, it seems that increasing the distance between aromatic
core and polar diol group provides flexibility necessary to achieve
high-entropy high-symmetry phases.22 It is also worth pointing
out that in facial amphiphiles, which have a polar oligo-
(oxyethylene) side-chain and two flexible terminal alkyl chains,
pentagonal cylinders were only ever found to form the p4gm
phase, and never the p2gg.18a,c,d
This, however, does not mean that the backbone flexibility is
the only factor bringing about the p4gm phase. There are several
other factors. The first is the tendency of the linear RF chain
segments for parallel alignment which favors stretched penta-
gons, destabilizing p4gm and favoring p2gg symmetry at lower
temperatures. The more flexible oligo(oxyethylene) side chain of
the above-mentioned facial amphiphiles18c,d easily fill the “p4gm-
pentagons” and therefore do not require cell deformation.
Additional factors favoring the p4gm phase are a more central
position of the lateral chain which allows an easier filling of the
more circular area inside “p4gm-pentagons” and the correct ratio
of cube root of side-chain volume to backbone length.
Figure 3. Compound F6-1/4. (a) Texture as seen between crossed
polarizers in the Colsqu/p4gm phase at T = 132 ꢀC; (b) in the Colrec/p2gg
phase T = 60 ꢀC; (c) texture of the Colsqu/p4gm and (d) Colrec/p2gg
phase with λ-retarder plate (same areas and temperatures; the indicatrix
orientation inthe compensatorisshown in theinset); (e) X-raydiffraction
pattern of an aligned sample in the Colsqu/p4gm phase at T = 132 ꢀC and
(f) of the Colrec/p2gg phase at T = 60 ꢀC; (g) scan over the wide angle
region at T = 120 ꢀC; (h) DSC heating and cooling scans (10 K minꢀ1) in
the temperature range of the p2gg-p4gm-Iso phase transition; (i) electron
density map as calculated for the p4gm phase from the diffraction pattern
at T = 132 ꢀC; for color code, see bar on the right; the positions of the
bolaamphiphilic moieties are indicated by the dark lines (relative Bragg
diffraction intensities used: I11 = 4.2, I20 = 68, I21 = 100, respective phase
angles π, π, 0); and (j) comparison of the p4gm and p2gg lattices.
Compound F6-1/9 with the longest spacer between biphenyl
core and the diol group (n2 = 9) displays a square columnar phase
as indicated by the ratio of sinθ of the small angle reflections
√
being 1: 2 (see Table 1; Table S1 in the SI). On the basis of the
2d XRD pattern obtained for an aligned sample this square phase
can be assigned to a simple square lattice with plane group p4mm
(Figure S3aꢀc, SI). The lattice parameter asqu = 2.94 nm is close
to the molecular length (Lmol = 3.2 nm). This is in agreement
with a square honeycomb structure (see Figure 2, right), which is
7876
dx.doi.org/10.1021/ja200822z |J. Am. Chem. Soc. 2011, 133, 7872–7881

Journal of the American Chemical Society
ARTICLE
Figure 4. Texture as seen for Colsqu/p4mm phase of compound F6-1/9
(a) between crossed polarizers at T = 115 ꢀC, the dark areas are
homeotropically aligned regions; (b) texture with λ-plate (same area),
the indicatrix orientation in the compensator is the same as shown in the
inset in Figure 3c,d.
also in line with the lack of birefringence for light traveling along
the columns in the homeotropic areas (Figure 4a) and the
negative birefringence in other areas (see Figure 4b).
The phase sequence Colhex/p6mm(hexagons) ꢀ Colrec/p2gg/
Colsqu/p4gm(pentagons) ꢀ Colsqu/p4mm(squares) as observed
upon increasing the spacer length n2 indicates that extending the
spacer has the same effect as reducing the length of the lateral
chains (see Figure 1). In the series of compounds F6-1/n2, by
increasing n2 from 1 via 4 to 9, the length of the side of the
polygon increases while the volume (area) required by the lateral
chains remains constant. Therefore, fewer cores are required to
enclose the semiperfluorinated lateral chains, and the number of
sides of the polygon decreases (Table 1).
F6-4/4 is different from the compounds discussed so far, as in
this compound two equally long spacers separate the biphenyl
core from the terminal polar groups. XRD patterns of com-
pounds F6-4/4 and F6-1/9 are both indexed on a square lattice
with p4mm symmetry (see Figures 5a and S4f,g, SI). For F6-4/4,
the phase structure was further corroborated by electron density
reconstruction. The electron density map obtained from Bragg
peak intensities of the small angle synchrotron powder pattern
(Figure S4g, Table S3a in SI) is shown in Figure 5c. The
maximum (A) is due to the perfluorinated lateral chains in the
center of the honeycomb cell. Perfluorinated alkyl is known to
have a far higher electron density than the rest of the molecule;
for the molecular model of this phase see Figure 5e.
The lattice parameter asqu is slightly larger for F6-4/4
(asqu = 3.0 nm) than for F6-1/9 (asqu = 2.9 nm), although F6-4/4
(Lmol = 3.1 nm) has a total spacer length two CH2 units shorter
that F6-1/9 (Lmol = 3.2 nm). This suggests that a longer and thus
more flexible alkylene spacer at only one end (compound F6-1/9)
allows more efficient space filling. In principle, there are three
possible arrangements of the bolamphiphilic cores in the cylinder
walls, either with segregation of the biphenyls from the alkylene
spacers (parallel alignment as in Figure 5f), a nonsegregated
structure with antiparallel side-by-side organization of the bolaam-
phiphilic cores or a randomly mixed structure. It is reasonable to
assume that the flexible alkylene spacers segregate from the rigid
biphenyls14 as in Figure 5f. However, there is no indication of a
long-range ordered super lattice with reduced symmetry and long-
range order of the segregated aromatic and aliphatic regions of
compound F6-1/9 in the cylinder walls. Therefore, it appears
Figure 5. Colsqu/p4mm phases of compounds F6-4/4 (left) and F8-1/9
(right): (a) X-ray diffraction pattern of an aligned sample (small angle
region) of compound F6-4/4 at 90 ꢀC and (b) F8-1/9 at 97 ꢀC; (c,d)
electron density maps (surface and contour plots, four unit cells) as
calculated from the synchrotron powder SAXS data for (c) compound
F6-4/4 (see Figure S4g in SI) and (d) compound F8-1/9 (see Figure
S7d in SI); see SI for details on map construction and related 2d
presentations (Figure S1in SI); (e,f) arrangement of molecules in the
Colsqu/p4mm phases of(e) compound F6-4/4and (f) compoundF8-1/9
after molecular dynamic simulation (for details see SI).
reasonable that core orientation in the cylinder walls is on a larger
scale averaged, resulting in overall p4mm symmetry.36 To distin-
guish these possibilities UV-investigations of dilute solutions as
well as thin films of compounds F8-1/9 and F6-4/4 were
performed. In dilute solution (10ꢀ4 mol/L in CH2Cl2) there is
only one peak with a maximum at 287 nm for F8-1/9 and at
288 nm for F6-4/4. In the thin films obtained after spin-coating
from THF solution (thickness about 1 μm) the absorption is blue-
shiftedto266 nmforF6-4/4andto269 nmfor compoundF8-1/9
and for both compounds the peak is broadened (Figure S8 in SI).
This hypsochromic shift indicates the formation of H-aggregates
(there is nearly no fluorescence which is in line with formation of
H-aggregates) in which the biphenyls are packed in a molecular
arrangement with the transition moments aligned parallel to each
other but perpendicular to the line joining their centers (face-to-
face arrangement). This confirms the stacking of the aromatics
along the cylinder walls in both cases. For F6-4/4 the hypsochro-
mic shift is stronger (Δλ = 22 nm) than for F8-1/9 (Δλ = 18 nm),
which indicates that there are larger aggregates of the aromatics in
the case of F6-4/4 than for F8-1/9. This is in accordance with the
7877
dx.doi.org/10.1021/ja200822z |J. Am. Chem. Soc. 2011, 133, 7872–7881

Journal of the American Chemical Society
ARTICLE
proposed models; for F6-4/4 with symmetric distribution of the
spacers there is a nearly uninterrupted packing of the biphenyls
along the cylinder walls, whereas for compounds with a nonsymmetric
distribution of the spacers (F8-1/9 and F6-1/9) there are smaller
side-by-side aggregates of the biphenyls along the cylinder walls,
and the aggregates of the aromatics are interrupted by parts of the
walls formed by the aliphatic spacers. This leads to an averaging
of core orientation which is in line with the p4mm symmetry
recorded for the mesophases of compounds F8-1/9 and F6-1/9.
In line with this model based on local segregation and long-range
averaging, also mesophase stability of F6-1/9, as judged by the
isotropization temperature, is not reduced if compared with
compound F6-4/4. Remarkably, compound F6-4/4, symmetric
regarding the spacers (see Figure 5e) and therefore not having
this kind of packing issue, even has a significantly reduced
mesophase stability compared to F6-1/9 with the same degree
of fluorination. The increased stability of the Colsqu/p4mm phase
of F6-1/9 is in line with its more efficient molecular packing and
smaller cell size. Thus, it seems that it is favorable for the
molecules to have one long rather than two shorter spacers of
equal combined length. This effect of the distribution of the
spacer on phase stability could to some extend also be due to the
different chemical structures of the polar groups in these two
compounds. Compound F6-1/9 has a glycerol group at one end
providing a higher polarity and one more hydrogen bonding
acceptor site than the diol group which is attached to a long
aliphatic spacer. The symmetric compound F6-4/4 has both diol
groups attached to aliphatic spacers, leading to weaker segrega-
tion and reduced hydrogen bonding density which reduces the
phase stability.
2.4. Compounds F8-n1/n2. The effect of enlargement of the
fluorinated segment in the lateral chain is 2-fold: (i) the stability
of the Colsqu/p4mm phase is increased, as judged by the
increased temperature of isotropization (compare F6-1/9 and
F8-1/9 in Table 1); (ii) if the phase changes, the new phase
comprises cylinders with a larger number of sides (compare
F6-1/4 and F8-1/4 in Table 1). The increasing stability of the
p4mm phase can be attributed to the enhanced fluorophobic
effect which strengthens segregation. However, it is also infor-
mative to compare the electron density maps of F6-m/n and
F8-m/n compounds that form the common p4mm phase.37 The
map of compound F6-4/4 displays only one central maximum;
see Figures 5c and S1a (SI). On the other hand, the map for
F8-1/9 has a second maximum at the corners of the square cylinder
(marked C in Figure 5d; see also Figure S1b in SI). This second
maximum is related to the strong (11) Bragg reflections (see
Figure 5b and Figure S7c,d in SI). It arises from clustering of the
relatively electron-dense glycerol and diol groups at the apexes of
the square section of the column. Molecular simulation shows
these clusters clearly in compound F8-1/9 (Figure 5f—note the
clusters of red oxygens) but not so clearly in compound F6-4/4
(see Figure 5e). Inspection of the molecular models suggests that
the bolaamphiphilic molecular backbones in compound F6-4/4
are crumpled, resulting in many broken H-bonds. Most likely this
is in order to partially compensate for the deficit in volume of the
relatively short lateral semiperfluorinated chain. Thus, it would
appear that some molecules forego H-bonding with adjacent
head groups and instead have their backbones participate in
partially filling the excess free volume inside the cylinder. This
option is facilitated by the opportunity to form hydrogen bonds
with the ether oxygens at the aromatic cores. This reduces the
polar group clustering (nanosegregation) and hence the phase
Figure 6. (a) X-ray diffraction pattern of an aligned sample of com-
pound F8-1/4 at T = 110 ꢀC (for more details see Table S1 and Figure
S6 in SI); (b) model showing molecular organization in the hexagonal
honeycomb LC phase (the biphenyl core together with the spacer units
is shown as a gray rod).
stability.14 By contrast, in F8-1/9 that has a larger side chain, the
cell interior of the p4mm honeycomb is filled more densely;
hence, the molecular cores can stretch, allowing their polar end
groups to cluster and increase the phase stability (see Figure 5f).
Consistent with the above is the significant increase in lattice
parameter from asqu = 2.9 nm for F6-1/9 to asqu = 3.2 nm for F8-1/9.
Such variation in cell parameter is quite unusual for polygonal
cylinder phases, where the cell size is fixed by the length of the
aromatic core.15
It is interesting that the diffraction pattern of the p4mm phase
formed by compound F6-1/9 has the same weak (11) Bragg
reflection (Figure S3d in SI), as observed for compound F6-4/4.
This indicates a uniform electron density at the corners of the
square cylinder and hence a lack of clustering of the polar groups.
Thus, it appears that the RF chain length, i.e. the space filling
inside the cells, is more important for segregation of the polar
groups than the kind of distribution of the aliphatic spacers along
the backbone.
Increasing the ratio of the lateral chain volume to the length of
the molecular backbone beyond that fitting a square honeycomb
results in cylinders with a larger number of sides. The columnar
mesophases of compounds F8-4/1 and F8-1/4 were confirmed
√
as hexagonal by the observed 1: 3:2 ratio of reciprocal X-ray
d-spacings (Tables S1 in SI). The XRD pattern of an aligned
sample of compound F8-1/4, shown as an example in Figure 6a,
confirms hexagonal symmetry. The hexagonal lattice parameters
for compounds F8-4/1 and F8-1/4 (ahex = 4.1 nm) agree well
√
with the expected value ahex
=
3Lmol =4.3nm,whereLmol =2.5nm
is the molecular length. Together with the negative birefringence
of the Colhex phase and its optical uniaxiality (see Figure S6c,d in
SI), this confirms the hexagonal honeycomb structures for the
Colhex phase of compounds F8-4/1 and F8-1/4 (Figure 6b). The
number of molecules within a unit cell with an assumed height
of h = 0.45 nm is around 6; hence, the calculated wall thickness
is about two molecules, again in agreement with values found
for other cylinder phases of T-shaped amphiphiles. The present
is thus the same hexagonal honeycomb phase as previously re-
ported for compounds F6-1/1 and F8-1/1 that contained no
distinct spacer units, as the single CH2 groups (n1, n2 = 1) can be
considered as belonging to the polar glycerol groups.13
As expected, the hexagonal lattice parameters of the two
isomeric compounds F8-4/1 and F8-1/4, differing in the posi-
tion of the spacer unit relative to the side chain, are very similar.
However, there is a relatively large difference in mesophase
stability between these two compounds, for F8-1/4 being 21 K
7878
dx.doi.org/10.1021/ja200822z |J. Am. Chem. Soc. 2011, 133, 7872–7881

Journal of the American Chemical Society
ARTICLE
to adopt the lattice with the highest possible symmetry. Alter-
natively, the higher symmetry may merely be due to increasing
entropy-driven disorder, facilitated by higher molecular flexibil-
ity, as discussed above.
Interestingly, even rather long spacer units can retain the
molecular self-assembly in honeycomblike structures, although
the cylinder walls become more flexible. These aliphatic spacers
provide additional flexibility to these structures which allows a
specific type of cylinder structure to tolerate a broader range of
lateral chain volumes by enabling some expansion/compression
of the honeycombs. The electron density maps suggest that such
adjustment may occur at the expense of some hydrogen bonding.
Formation of certain cylinder walls by aliphatic chains instead of
aromatic cores was previously reported for the trapezoid cylinder
phase formed by some T-shaped facial amphiphiles.38 However,
in that case the alkyl chains form walls of their own, free of the
aromatic moiety.
Although the change of the cylinder shape is principally the
same as observed by elongation of the π-conjugated rigid core,
the effect of spacer elongation on the phase transition tempera-
tures is reversed. Whereas elongation of a linear and rigid moiety
dramatically increases both melting and clearing temperatures,21
backbone extension by alkyl spacers reduces the clearing tem-
peratures without influencing the melting temperatures signifi-
cantly. Thus, replacing parts of the aromatic cores by flexible
units is an appropriate tool for increasing the size of the
polygonal honeycombs without moving the transitions to un-
favorably high temperature ranges.
Figure 7. Phase sequence of distinct polygonal cylinder phases,
depending on the length of the lateral chains and the total length of
the space units n1 þ n2.
higher than for F8-4/1, see Table 1. It seems that placing the
longer spacer beside the lateral chain is favorable for mesophase
formation. This effect of the spacer position might also be compared
to the reduced mesophase stability of the Colsqu phase of compound
F6-4/4 compared to that of compound F6-1/9, as discussed in the
previous section. Hence, for the self-assembly in LC cylinder phases
it seems to be advantageous to have molecules with the longer
spacer adjacent to the lateral chain. It is probable that the effect is
due to the compatibility between the alkyl segments of the lateral
chains and the adjacent backbone alkyl spacers.
In summary, extension of the bolaamphiphilic core by flexible
spacer units can be used to increase the size of the polygonal
cylinders, to achieve larger periodicities and to increase the
symmetry of the self-assembled liquid-crystalline honeycombs.
3. SUMMARY
T-shaped bolaamphiphiles composed of a biphenyl core, a
semiperfluorinated lateral chain, and aliphatic spacer units
(n1, n2) introduced between the rigid core and the polar groups
have been synthesized and investigated by DSC, polarizing
microscopy, and XRD. Suggested phase structures are based
on XRD patterns, electron density reconstruction, simulations,
and investigation of the influence of structural modifications
on observed phase types. Four different columnar phases
(Colhex/p6mm, Colrec/p2gg, Colsqu/p4gm, and Colsqu/p4mm)
were found in T-shaped bolaamphiphiles Fm-n1/n2 by increasing
the length of aliphatic spacer units n1, n2. These LC phases
represent polygonal cylinder phases where the aromatic cores
together with the aliphatic spacers make up the walls of polygonal
honeycombs fused at the edges by the hydrogen-bonding diol
groups and enclosing cells filled by the semiperfluorinated lateral
chains. In the observed phase sequence, with increasing overall
length of the backbone spacers the cross section of the polygonal
honeycomb cells changes from hexagon via pentagon to square
(Figure 7). This sequence is in line with the decreasing area-to-
circumference ratio achieved by extending the bolaamphiphilic cores.
By introduction of the aliphatic spacers the phase symmetry is
enhanced; for example the rhombic cylinder phase, typically
observed for related biphenyl-based compounds without spacers
and with appropriate side-chain length (compounds Fm-1/1,
Figure 1c), is replaced by the square cylinder phase with higher
symmetry. Likewise, there is a tendency to form the square p4gm
lattice rather than the less symmetric rectangular p2gg lattice for
pentagonal honeycombs, as indicated by the p2gg ꢀ p4gm
transition in compound F6-1/4. It is most likely that the more
central position of the alkyl chains at the bolaamphiphilic core
reduces restrictions in chain packing, thereby allowing the system
’ ASSOCIATED CONTENT
S
Supporting Information. Synthesis procedures and ana-
b
lytical data, additional DSCs, textures and XRD data. This material
is available free of charge via the Internet at http://pubs.acs.org.
’ AUTHOR INFORMATION
Corresponding Author
xhcheng@ynu.edu.cn; carsten.tschierske@chemie.uni-halle.de;
g.ungar@sheffield.ac.uk
’ ACKNOWLEDGMENT
This work was supported by the National Natural Science
Foundation of China (No. 21074105 and No. 20973133), the
Yunnan Science Foundation (2010CD018). M.P. acknowledges
the support by the Cluster of Excellence “Nanostructured
Materials” and DFG (FG 1145). G.U. also acknowledges the
support from the WCU program through the National Research
Foundation of Korea funded by the Ministry of Education,
Science and Technology (R31-10013). For help with the syn-
chrotron experiments we thank Drs. Nick Terrill and Jen Hiller
(beamline I22) of the Diamond Light Source, U.K.
’ REFERENCES
(1) (a) Steed, J. W.; Atwood, J. L. Encyclopedia of Supramolecular
Chemistry; Marcel Dekker: New York, 2004. (b) Lehn, J. M. Supramo-
lecular Chemistry; Wiley-VCH: Weinheim, Germany, 1995.
7879
dx.doi.org/10.1021/ja200822z |J. Am. Chem. Soc. 2011, 133, 7872–7881

Journal of the American Chemical Society
ARTICLE
(2) (a) Collings, P. J.; Hird, M. Introduction to Liquid Crystals; Taylor
& Francis: London, 1997. (b) Demus, D.; Goodby, J.; Gray, G. W.;
Spiess, H.-W.; Vill, V. Handbook of Liquid Crystals; Wiley-VCH:
Weinheim, Germany, 1998, Vol. 1ꢀ3. (c) Pauluth, D.; Tarumi, K.
J. Mater. Chem. 2004, 14, 1219–1227.
(3) (a) Kato, T.; Mizoshita, N.; Kishimoto, K. Angew. Chem., Int. Ed.
2006, 45, 38–68. (b) Saez, I. M.; Goodby, J. W. J. Mater. Chem. 2005,
15, 26–40. (c) Ungar, G.; Zeng, X. Soft Matter. 2005, 1, 95–106.
(d) Laschat, S.; Baro, A.; Steinke, N.; Giesselmann, F.; H€agele, C.;
Scalia, G.; Judele, R.; Kapatsina, E.; Sauer, S.; Schreivogel, A.; Tosoni, M.
Angew. Chem., Int. Ed. 2007, 46, 4832–4887. (e) Lee, M.; Cho, B.-K.; Zin,
W.-C. Chem. Rev. 2001, 101, 3869–3892.
(4) (a) Ringsdorf, H.; Schlarb, B.; Venzmer, J. Angew. Chem., Int. Ed.
1988, 27, 113–158. (b) Greig, L. M.; Philp, D. Chem. Soc. Rev. 2001,
30, 287–302. (c) Lehn, J.-M. Proc. Natl. Acad. Sci. U.S.A. 2002,
99, 4763–4768. (d) Hoeben, F. J. M.; Jonkheijm, P.; Meijer, E. W.;
Schenning, A. P. H. J. Chem. Rev. 2005, 105, 1491–1546. (e) Keizer,
H. M.; Sijbesma, R. P. Chem. Soc. Rev. 2005, 34, 226–234.
(5) Rosen, B. M.; Wilson, C. J.; Wilson, D. A.; Peterca, M.; Imam,
M. R.; Percec, V. Chem. Rev. 2009, 109, 6275–6540.
(6) (a)Donnio, B.; Guillon, D. Adv. Polym. Sci. 2006, 201, 45–155. (b)
Lenoble, J.; Campidelli, S.; Maringa, N.; Donnio, B.; Guillon, D.;
Yevlampieva, N.; Deschenaux, R. J. Am. Chem. Soc. 2007, 129, 9941–9952.
(7) (a) Baranoff, E. D.; Voignier, J.; Yasuda, T.; Heitz, V.; Sauvage,
J.-P.; Kato, T. Angew. Chem., Int. Ed. 2007, 46, 4680–4683. (b)
AprahamianI.; Yasuda, T.; Ikeda, T.; Saha, S.; Dichtel, W. R.; Isoda,
K.; Kato, T.; Stoddart, J. F. Angew. Chem., Int. Ed. 2007, 46, 4675–
4679.
(8) Kitzerow, H.-S.; Bahr, C. Chirality in Liquid Crystals; Springer:
New York, 2001; Pansu, B. Mod. Phys. Lett. B 1999, 13, 769–782.
(9) (a) Reddy, R. A.; Tschierske, C. J. Mater. Chem. 2006,
16, 907–961. (b) Takezoe, H.; Takanishi, Y. Jpn. J. Appl. Phys. 2006,
45, 597–625.
(10) (a) Ikuma, N.; Tamura, R.; Shimono, S.; Kawame, N.; Tamada,
O.; Sakai, N.; Yamauchi, J.; Yamamoto, Y. Angew. Chem., Int. Ed. 2004,
43, 3767–3682. (b) Zienkiewicz, J.; Fryszkowska, A.; Zienkiewicz, K.;
Guo, F.; Kaszynski, P.; Januszko, A.; Jones, D. J. Org. Chem. 2007,
72, 3510–3520. (c) Terazzi, E.; Bourgogne, C.; Welter, R.; Gallani, J.-L.;
Guillon, D.; Rogez, G.; Donnio, B. Angew. Chem., Int. Ed. 2008, 47,
490–495.
(19) (a) Prehm, M.; Liu, F.; Zeng, X.-B.; Ungar, G.; Tschierske, C.
Angew. Chem., Int. Ed. 2007, 46, 7972–7975. (b) Prehm, M.; Enders, C.;
Anzahaee, M. Y.; Glettner, B.; Baumeister, U.; Tschierske, C. Chem.—Eur.
J. 2008, 14, 6352–6368. (c) Kieffer, R.; Prehm, M.; Pelz, K.; Baumeister,
U.; Liu, F.;Hahn, H.; Lang, H.; Ungar, G.; Tschierske, C. SoftMatter2009,
5, 1214–1227.
(20) (a) Cheng, X. H.; Das, M. K.; Diele, S.; Tschierske, C. Angew.
Chem., Int. Ed. 2002, 41, 4031–4035. (b) Prehm, M.; Cheng, X. H.;
Diele, S.; Das, M. K.; Tschierske, C. J. Am. Chem. Soc. 2002,
124, 12072–12073. (c) Prehm, M.; Diele, S.; Das, M. K.; Tschierske,
C. J. Am. Chem. Soc. 2003, 125, 614–615. (d) Patel, N. M.; Dodge, M. R.;
Zhu, M. H.; Petschek, R. G.; Rosenblatt, C.; Prehm, M.; Tschierske, C.
Phys. Rev. Lett. 2004, 92, 015501. (e) Patel, N. M.; Syed, I. M.;
Rosenblatt, C.; Prehm, M.; Tschierske, C. Liq. Cryst. 2005, 32, 55–61.
(21) See, for example, compound D in: Kieffer, R.; Prehm, M.;
Glettner, B.; Pelz, K.; Baumeister, U.; Liu, F.; Zeng, X.-B.; Ungar, G.;
Tschierske, C. Chem. Commun. 2008, 3861–3863.
(22) Cheng, X. H.; Dong, X.; Huang, R.; Zeng, X.-B.; Ungar, G.;
Prehm, M.; Tschierske, C. Chem. Mater. 2008, 20, 4729–4738.
(23) Cheng, X. H.; Dong, X.; Wei, G.; Prehm, M.; Tschierske, C.
Angew. Chem., Int. Ed. 2009, 48, 8014–8017.
(24) Prehm, M.; G€otz, G.; B€auerle, P.; Liu, F.; Ungar, G.; Tschierske,
C. Angew. Chem., Int. Ed. 2007, 46, 7856–7859.
(25) A monotropic phase assigned as Colsqu/p4mm in ref 13
(compound F4-1/1) turned out to be actually a 3D-hexagonal phase:
Liu. F.; Zeng, X.; Ungar, G.; Tschierske, C. unpublished results.
(26) (a) Miyaura, N.; Yanagi, T.; Suzuki, A. Synth. Commun. 1981,
11, 513–519. (b) Hird, M.; Gray, G. W.; Toyne, K. J. Mol. Cryst. Liq.
Cryst. 1991, 206, 187–204. (c) Miyaura, N.; Suzuki, A. Chem. Rev. 1995,
95, 2457–2483.
(27) Chen, Q.-Y.; Yang, Z. Y.; Zhao, C.-X.; Qiu, Z. M. J. Chem. Soc.,
Perkin Trans. 1 1988, 563–567.
(28) Majetch, G.; Hicks, R.; Reister, S. J. Org. Chem. 1997, 62,
4321–4326.
(29) Yamoto, Y.; Hatsuya, S.; Yamada, J. J. Org. Chem. 1990, 55,
3118–3128.
(30) Kitamura, M.; Isobe, M.; Ichikawa, Y.; Goto, T. J. Am. Chem.
Soc. 1984, 106, 3252–3257.
(31) Van Rheenen, V.; Cha, D. Y.; Hartley, W. M. Org. Synth. 1979,
58, 43–51.
(32) K€olbel, M.; Beyersdorff, T.; Tschierske, C.; Diele, S.; Kain, J.
Chem.—Eur. J. 2000, 6, 3821–3837.
(11) Zeng, X.; Ungar, G.; Liu, Y.; Percec, V.; Dulcey, A. E.; Hobbs,
J. K. Nature 2004, 428, 157–160.
(12) K€olbel, M.; Beyersdorff, T.; Cheng, X. H.; Tschierske, C.; Kain,
J.; Diele, S. J. Am. Chem. Soc. 2001, 123, 6809–6818.
(33) Immirzi, A.; Perini, B. Acta Crystallogr., Sect. A. 1977, 33,
216–218.
(13) Cheng, X. H.; Prehm, M.; Das, M. K.; Kain, J.; Baumeister, U.;
Diele, S.; Leine, D.; Blume, A.; Tschierske, C. J. Am. Chem. Soc. 2003,
125, 10977–10996.
(14) (a) Tschierske, C. J. Mater. Chem. 1998, 8, 1485–1508. (b)
Tschierske, C. J. Mater. Chem. 2001, 11, 2647–2671. (c) Tschierske, C.
Ann. Rep. Progr. Chem. Ser. C. 2001, 97, 168–191. (d) Tschierske, C.
Curr. Opin. Colloid Interface Sci. 2002, 7, 69–80.
(15) Tschierske, C. Chem. Soc. Rev. 2007, 36, 1930–1970.
(16) Glettner, B.; Liu, F.; Zeng, X.; Prehm, M.; Baumeister, U.;
Bates, M. A.; Walker, M.; Boesecke, P.; Ungar, G.; Tschierske, C. Angew.
Chem., Int. Ed. 2008, 47, 9063–9066.
(17) (a) Cheng, X. H.; Das, M. K.; Baumeister, U.; Diele, S.;
Tschierske, C. J. Am. Chem. Soc. 2004, 126, 12930–12940. (b) Prehm,
M.; Liu, F.; Baumeister, U.; Zeng, X. ꢀB; Ungar, G.; Tschierske, C.
Angew. Chem., Int. Ed. 2007, 46, 7972–7975.
(18) (a) Chen, B.; Baumeister, U.; Diele, S.; Das, M. K.; Zeng, X.-B.;
Ungar, G.; Tschierske, C. J. Am. Chem. Soc. 2004, 126, 8608–8609. (b)
Chen, B.; Zeng, X.-B.; Baumeister, U.; Diele, S.; Ungar, G.; Tschierske,
C. Angew. Chem., Int. Ed. 2004, 43, 4621–4625. (c) Chen, B.; Zeng,
X.-B.; Baumeister, U.; Ungar, G.; Tschierske, C. Science 2005,
307, 96–99. (d) Chen, B.; Baumeister, U.; Pelzl, G.; Das, M. K.; Zeng,
X.-B.; Diele, S.; Ungar, G.; Tschierske, C. J. Am. Chem. Soc. 2005,
127, 16578–16591. (e) Liu, F.; Chen, B.; Baumeister, U.; Zeng, X.;
Ungar, G.; Tschierske, C. J. Am. Chem. Soc. 2007, 129, 9578–9579.
(34) That the biphenyl cores are oriented perpendicular to the
column axis is shown by the color of the fans in the optical micrographs
(Figures 3c,d, and 4b, Figures S4d, S6d, S7g in SI) taken with a λ-plate.
The yellow and blue colors define the orientation of the high-index axis
as radial rather than tangential. Since the columns are tangential in the
fans, and the high-index axis is known to be parallel to the biphenyl long
axis, it follows that the biphenyls are perpendicular to the columns;
(n < n^ = optically negative).
(35) The value h = 0.45 nm was used in all calculations, independent
of the actual position of the maximum of the wide angle scattering in the
diffraction pattern. The reason is that the bolaamphiphilic units (their
average diameter is 0.45 nm) are responsible for the lattice of the cylinder
shells, whereas the semiperfluoralkyl chains only fill the channels within
this structure. Because the cross section of the perfluorinated segments is
larger than the rest of the molecules, there is a drift of the maximum of
the diffuse scattering in the X-ray diffraction pattern to larger d-values
with increasing degree of fluorination. However, this shift is due to a
change of a molecular parameter rather than due to a change of the
structure. Because the diameter of the fluorinated segments is larger than
that of the nonfluorinated bolaamphiphilic moieties, which form the
cylinder walls, there is a mixing of the chains of adjacent (hypothetical)
unit cells along the cylinders, i.e. the perfluorinated chains in one
(hypothetical) unit cell also contribute to the space filling in the adjacent
unit cell.
7880
dx.doi.org/10.1021/ja200822z |J. Am. Chem. Soc. 2011, 133, 7872–7881

Journal of the American Chemical Society
ARTICLE
(36) A nonsegregated antiparallel packing of the bolaamphiphilic
cores in the cylinder walls could become possible if steric and packing
effects are dominant over the segregation between biphenyl and alkylene
spacers (see: Lose, D.; Diele, S.; Pelzl, G.; Dietzmann, E.; Weissflog, W.
Liq. Cryst. 1998, 24, 707–717. Pensec, S.; Tournilhac, F.-G.; Bassoul, P.;
Durliat, C. J. Phys. Chem. B 1998, 102, 52–60). However, as the cross-
sectional area of alkyl chains and biphenyl cores is nearly equal, a
randomly mixed structure becomes unlikely for compounds 1/nꢀ1/9,
as confirmed by the investigation of UV absorption; alternation of blocks
with different core orientations along the cylinder walls could remove
small differences in packing of the biphenyls and alkylene spacers.
(37) Increasing the length of the high electron density RF chains
from n = 6 to n = 8 (compounds F6-4/4, F8-1/9) by retaining the p4mm
lattice leads to an expansion of the high electron density regions in the
centers of the low electron density continuum in the electron density
maps (see Figure 5c,d); hence, the phase choice must be correct, as
expansion of the high electron density units leads to an increase of the
high electron density areas. In this way inverted phase structures could
be excluded.
(38) Liu, F.; Chen, B.; Glettner, B.; Prehm, M.; Das, M. K.;
Baumeister, U.; Zeng, X. ꢀB; Ungar, G.; Tschierske, C. J. Am. Chem.
Soc. 2008, 130, 9666–9667.
7881
dx.doi.org/10.1021/ja200822z |J. Am. Chem. Soc. 2011, 133, 7872–7881