The effect of the shape of the mesogenic group on the structure and phase behavior of 2,3,4-tris(dodecyloxy)benzenesulfonates with alkaline cations

Soft Matter

View Article Online

View Journal | View Issue

PAPER

The eﬀect of the shape of the mesogenic

group on the structure and phase behavior

of 2,3,4-tris(dodecyloxy)benzenesulfonates

with alkaline cations†

Cite this: Soft Matter, 2014, 10, 1746

ab

Artem V. Bakirov,^acAndrei N. Yakunin,^dUwe Beginn,^e

*

Maxim A. Shcherbina,

Linglong Yan,^fMartin Moller^fand Sergei N. Chvalun^ac

¨

Synthesis and mesophase structure characterisation are reported for a group of alkali salts of 2,3,4-

tris(dodecyloxy)benzenesulfonic acid. As revealed by a combination of polarizing optical microscopy,

diﬀerential scanning calorimetry and X-ray scattering, variation of the eﬀective mesogen shape due to

changes of the cation size leads to systematic transformation of the materials' phase behaviour.

Thermotropic mesophases of diﬀerent types and dimensionalities were observed: 1D (smectic bilayers),

2D (ordered and disordered columnar phases), and 3D (high-temperature micellar mesomorphic phase,

low-temperature crystalline one). Cubic packing prevails when the cation size is small and, thus, the

eﬀective mesogen shape is close to the conic one. With increasing ion size, the mesogen shape

becomes more tapered, and columnar mesophases appear to be more stable.

Received 3rd July 2013

Accepted 2nd December 2013

DOI: 10.1039/c3sm51821c

www.rsc.org/softmatter

Recently a new class of “cunitic”¹molecules, 2,3,4- and 3,4,5-

tris(alkoxy)benzenesulfonates, became of paramount interest

Introduction

for materials design due to an unique richness of their phase

behaviour, i.e. exhibiting thermotropic and lyotropic smectic,

columnar and cubic phases.^2–7Such compounds are important

examples of self-assembling systems which are extremely

sensitive to the state of the environment, because even a slight

modication of the shape of a molecule, caused by a change in

the functional of its free energy,⁸may signicantly alter the

shapes of its supramolecular aggregates, which is the reason

why the structure of the material can be controlled precisely.

For instance, self-assembling under an electric eld makes it

possible to orient nano-conducting channels on the substrate

surface.⁹Electric elds have also been used to orient structural

elements of block-copolymer lms.¹⁰Changes in the structure

of supramolecular aggregates containing magnetic nano-

particles and forming photonic crystals were also observed

under a magnetic eld.¹¹

Supramolecular materials chemistry oﬀers the possibility to

generate functional materials containing highly complex

structures by exploiting the self-assembling abilities of matter.

Instead of forcing the required molecular arrangement and

macroscopic morphologies by macroscopic physical experi-

mental techniques (“top down” approach), chemical synthesis

is used to design molecular shapes and mutual interactions

allowing for the spontaneous formation of the desired struc-

tures and functionalities (“bottom up” approach).

^aEnikolopov Institute of Synthetic Polymer Materials RAS, 70, Profsoyuznaya street,

Moscow, 117393, Russian Federation. E-mail: shcherbina@ispm.ru; Fax: +7 495

7183404; Tel: +7 495 3325815

^bMocsow Institute of Physics and Technology, 9, Institutsky per., Dolgoprudny, Moscow

region, 141700, Russian Federation

Benzenesulfonates were also shown to form stable columnar

aggregates with temperature-switchable ion channels, orga-

nized in a two-dimensional hexagonal lattice stabilized by

chemical cross-linking of their aliphatic tails.¹²Such cross-

linking was achieved by the introduction of methacryloyl groups

near the ends of the aliphatic tails. Subsequent X-ray irradiation

leads to the formation of a uniform paraﬃn matrix with regu-

larly embedded aromatic columns.

^cKurchatov Institute, 1, Kurchatov sq., Moscow, 123098, Russian Federation. E-mail:

serge@ispm.ru; Fax: +7 495 7183404; Tel: +7 495 3325815

^dKarpov Institute of Physical Chemistry, 10, Vorontsovo Pole, Moscow, 105064,

Russian Federation. E-mail: andreiyakunin@yandex.ru; Fax: +7 495 9172490; Tel:

+7 495 917430

e

¨

Universitat Osnabruck, Institut fur Chemie, OMC, Barbarastraße 7, D-49076

¨

Osnabruck, Germany. E-mail: ubeginn@uni-osnabrueck.de; Fax: +495419692370;

Tel: +495419692790

^fDWI/RWTH

Aachen

University,

Building

5293/DWI

Erweiterungsbau,

Furthermore, these compounds can even form isolated

elongated supramolecular structures and organogels from

diluted solutions. 2,3,4-Tris(alkoxy)benzenesulfonates are

Forckenbeckstraße 50, 52056 Aachen, Germany. E-mail: moeller@dwi.rwth-aachen.

de; Fax: +492418023301; Tel: +492418023300

† Electronic supplementary information (ESI) available. See DOI:

10.1039/c3sm51821c

1746 | Soft Matter, 2014, 10, 1746–1757

This journal is © The Royal Society of Chemistry 2014

View Article Online

Paper

Soft Matter

among the most interesting representatives of cunitic mole- would form a lattice of particular symmetry. Imagine a sphere of

cules.¹³They are related in shape to the well-known tris(alkoxy) variable radius r with a center coinciding with the centre of each

benzoates,¹⁴though, being stronger acids, they carry stronger Voronoi polyhedron. Let v_i(r) be the volume of the part of the i^th

negative charges at the focal point of the mesogenic group.

crystalline cell that lies within the sphere. At small r the sphere

Moreover, the sulfur atom of the sulfonate tip possesses sp³grows freely; therefore, dV/dr ꢀ r². As the sphere approaches the

hybridization, and, correspondingly, is of spherical shape, Voronoi polyhedron boundaries, the volume growth rate

while carboxylates contain a at sp²hybridized carbon centre. decreases as the radius increases. To understand the relation-

For these reasons, 2,3,4-tris(alkoxy)benzenesulfonates allow the ship between the calculated volume distributions and the

study of the eﬀect of charge density as well as the shape of the dendron shape, we will consider a conical molecule projected

cunitic tip, or the type, of the used cation on the self-organi- onto a plane which is perpendicular to the cone axis. Then the

zation, and phase behaviour of the compounds by (i) elucida- ideal envelope of this pattern is described by the distribution

tion of the structure and shape of their respective dV/dr.

supramolecular aggregates and (ii) by comparison of such

This paper also aims to extend an analysis of the self-assem-

aggregates to those formed by well characterized carboxylate bling of tapered and conical dendrons on a new class of cunitic

cunitic molecules. These goals will be achieved in the present molecules: 2,3,4-tris(alkoxy)benzenesulfonate derivatives.

paper by a comparative analysis of thermotropic liquid crystal-

line phases formed by four alkali salts of 2,3,4-tris(alkoxy)-

benzene-sulfonic acid with substantially diﬀerent cation sizes,

Experimental methods

that is lithium- (compound I), sodium- (II), potassium- (III) and X-ray scattering

caesium- (IV) 2,3,4-tris(alkoxy)benzenesulfonate.

The structures of the compounds I–IV were investigated by

means of small- and wide-angle X-ray diﬀraction methods. High

resolution 2-D SAXS patterns have been obtained, using a

NanoStar Bruker© AXS diﬀractometer (CuK_a-radiation, l ¼

˚

1.5406 A) with a proportional Hi-Star detector and point colli-

¨

mation systems on the basis of Gobel mirror (forming slit

diameters are 0.75, 0.4 and 1.0 mm respectively). The device

setup allowed us to record scattering patterns in the range of

wave vector amplitudes, corresponding to d-spacings from 0.8

nm to 17.5 nm. The exposure time varied from 400 to 2000 s,

depending on the intensity of scattered radiation.

X-Ray diﬀraction patterns in small and wide angle scattering

regions were also recorded using an S3-Micropix system (Hecus

˚

company), CuK_a-radiation, l ¼ 1.5406 A with a Xenocs Genix

source (voltage and current are 50 kV and 1 mA respectively). A

Pilatus 100K detector was employed, as well as a linear PSD 50M

gas detector (Ar–Me mixture at 8 ꢁ 10⁵Pa). The collimation

system Fox 3D with Kratky collimation slits of 0.1 mm and

As a tentative hypothesis, complex behaviour of the cunitic

sulfonate amphiphiles described below may be interpreted in

terms of a microphase – segregation model. The large, non-polar

alkoxy-benzene-part tends to segregate from the ions, which

hence form an ion-cluster phase, conned by the non-polar

phase. Since the anion is covalently bound to the non-polar

molecule fraction, the two phases cannot move independently.

Although it is not required (and, of course improbable) that the

cation is permanently xed in a dened spatial position to a

certain sulfonate anion, the macroscopic behaviour is deter-

mined by the cation-phase/anion-phase volume ratio. It is hence

hypothesized that either Izraelachvili's model of micelle forma-

tion¹⁵or the theories of block-copolymer mesophases^16,17may be

transferred to the sulfonate amphiphile system. Thus, it may be

allowed to discuss the eﬀect of the cation size as if the cation was

virtually xed to the tip of the cunitic anion eﬀective shape

approach.

0.2 mm width was used, allowing reliable measurements in a

ꢂ1

wave vector range from q ¼ 0.003 A to q ¼ 1.9 A^ꢂ1where q ¼

˚

4p sin q/l and 2q is the scattering angle. To get rid of the

¨

scattering of X-rays on air molecules, Gobel mirrors and scat-

tering path had been vacuumed (residual air pressure is about

2.6 O 5.0 Pa). The exposure time varied from 600 to 5000 s. The

thermal behavior of samples was studied using Joule attach-

ꢃ

ment in the temperature range from 23 C to 300 C.

Nascent powder samples were put between two ultrathin

polyimide lms and installed into the holder of the vacuumed

X-ray camera. This design was chosen to ensure that the samples

remained in vacuum in the course of the X-ray experiment.

Domain sizes L and paracrystallinity parameter g were esti-

mated, using the Hoseman–Bonart method¹⁹from integral half-

widths Ds of several (usually two) reection orders:

A theoretical approach to an analysis of the eﬀect of the

shape of the mesogenic group on materials' phase behaviour

was developed by the calculation of the mean radial distribution

functions of volume dV/dr for diﬀerent 3D structures.¹⁸These

functions describe in fact the ideal shape of the dendron that

!

2

ꢀ

ꢁ

²ꢂ

1

Dq

2p

pg²n²q_hk

ðnÞ

¼

(1)

2

L_n

This journal is © The Royal Society of Chemistry 2014

Soft Matter, 2014, 10, 1746–1757 | 1747

View Article Online

Soft Matter

Paper

where V is the unit cell volume, and

Electron density maps

Electron density maps were calculated, using structural factors

F_hkl, obtained from experimental intensities of X-ray reection

maxima I(s_hkl):

rðrÞ ꢂ hri

qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ

DrðrÞ ¼

(8)

2

r²ðrÞ ꢂ hri

is the dimensionless normalized density uctuation (map of

electron density distribution).

I(s_hkl) f F_hkl²mLP

(2)

Another approach to the choice of correct phase combina-

tion includes the construction of electron density histograms,

and their comparison with theoretical ones. An application of

this technique to the structural studies of cubic mesophases

formed by dendrons on the basis of gallic acid can be found in

ref. 22. It should be noted also that electron density recon-

struction appears to be a quite reliable technique, when more

than three or four X-ray reections are used. Hence, it is very

important to compare the electron density maps obtained with

the results of molecular modelling, showing the principal

possibility of one or another packing of the matter into

supramolecular aggregates of known macroscopic density and

shape.

Corrections on reection repetition m and Lorentz factor L ꢀ

ꢂ2

q_hkl(ref. 20) have been made. Polarization factor P equals to

unity for the geometrical setups used.

According to the general theory of diﬀraction,²⁰periodic

functions of electron density r(r) can be represented as Fourier

series whose coeﬃcients are structural factors of individual

reections F_hkl

:

X

Dr h rðrÞ ꢂ r₀f

F_hklexp½ ꢂ 2piðhx þ ky þ lzÞꢄ

(3)

hkl

where r₀is an average electron density and x, y, z are partial

coordinates of unit cell nodes. Phase coeﬃcients of every term

in this sum are imaginary numbers, thus a reconstruction of

electron density is impossible for arbitrary crystal cells.

However, there exist several important cases, in which the

imaginary part of every term in eqn (3) is equal to zero. For such

symmetries, a choice of the phase factor for every reection is

reduced to a choice of its sign: “+” for a phase of (hkl) which is

equal to zero, and “ꢂ” for a phase angle of p. For instance, in

cubic centrosymmetric crystal lattices:²¹

Thermal analysis

Thermal analysis was carried out, using a “Perkin Elmer DSC-_ꢂ7₁”

ꢃ

calorimeter, heating–cooling rates being 10 and 20 C min

,

respectively. Phase transition temperatures were dened as the

maxima of the corresponding endo- and exo-thermal peaks

appearing in the thermograms.

X

NMR-spectroscopy

Drf

ꢅ jF_hklj cos½2pðhx þ ky þ lzÞꢄ

(4)

hkl

¹H-NMR (300 MHz) and ¹³C-NMR (75 MHz) spectra were

recorded on a Bruker DPX-300 spectrometer with tetrame-

thylsilane (TMS) as an internal standard in deuterated chloro-

Another important case of such possibilities is cylinder

symmetry of the system. The electron density distribution can

here be rewritten in cylindrical coordinates as:

ꢃ

form at 20 C.

ð

N

2p

N

Elemental analysis

rðr; j; zÞ ¼

FðR; j; ZÞ

0

The carbon and hydrogen elemental composition was deter-

mined in the Microanalytical Laboratory Kolbe, Muhlausen,

ꢁ exp ½ ꢂ 2pðRr cosðj ꢂ JÞ þ ZzÞꢄR dR dJ dZ (5)

¨

Germany.

Taking into account cylindrical symmetry of the system, one

can rewrite eqn (5), as:

Molecular modeling

X

rðrÞ ¼ 2p

ꢅjFðR_iÞjJ₀ð2pR_irÞR_iDR_i

(6)

The Accelrys Materials Studio program set was employed

for molecular modelling of compounds studied. We used

two sets of potentials which allow us to take into account the

non-covalent interactions of the mesogenic groups inside

the cylinders of the columnar mesophases: COMPASS

(Condensed-phase Optimized Molecular Potentials for Atom-

istic Simulation Studies) and UFF (Universal Force Field). The

COMPASS set is suitable for modelling of isolated molecules

and condensed phases of mainly organic, polymeric and of

some inorganic compounds,^23–25it also allows us to parame-

terise partial charges and valences ab initio with subsequent

system optimization. To prove the results of modelling, we

applied UFF potentials, used for calculation of geometry of

organic molecules containing metal–organic complexes. UFF

does not have any limitation on the chemistry of compounds

involved.^26–28

i

where J₀represents the Bessel function of 0^thorder, i is the

sequence number of the reection, and DR_iis its integral half-

width.

For f reections experimentally observed, there are 2^f

permutations of phase signs. Accurate determination of phase

signs is a particular important structural problem. Its unam-

biguous solution requires the use of a number of additional

approaches and techniques. The most widely used criterion for

the choice of phase signs is that their correct combination

causes a minimum value of the fourth moment of electron

density uctuations h(Dr)⁴i:²²

ð

D

E

1

4

ðDrðrÞÞ

f

ðDrðrÞÞ dV

(7)

V

1748 | Soft Matter, 2014, 10, 1746–1757

This journal is © The Royal Society of Chemistry 2014

View Article Online

Paper

Soft Matter

29.39, 29.72, 30.00, 30.29, 31.94, 73.96, 75.77, 107.40, 123.20,

124.78, 141.89, 149.73, 157.79; IR (KBr, cm^ꢂ1): 3435.11 (s,

broad), 2956.86 (m), 2917.63 (vs), 2851.29 (s), 1732.56 (w),

1631.30 (w), 1582.85 (w), 1486.22 (w), 1466.57 (m), 1441.26 (w),

1380.14 (w), 1304.58 (w), 1281.34 (w), 1226.56 (m), 1162.99 (w),

1087.42 (s), 1041.04 (w), 885.31 (w), 802.42 (w), 787.36 (w),

719.54 (w), 709.03 (w), 691.90 (w), 634.69 (w), 606.35 (w), 578.87

Synthesis

Synthesis of the studied compounds is described in detail

elsewhere.¹¹H and ¹³C NMR spectra of compounds I–IV can be

found in the ESI.†

Materials

1,2,3-Trihydroxy-benzene (99+%, Aldrich), 1-bromododecane (w), 531.66 (w); elemental analysis calculated (%): C 71.04, H

(99+%, Aldrich), potassium carbonate (97%, Aldrich), sulfuric 10.93; found (%): C 69.13, H 11.32.

acid (95–97%, Merck), methyllithium (Aldrich), sodium meth-

Lithium 2,3,4-tris(dodecyloxy)benzenesulfonate (I). In a

anolate (ACS reagent), potassium methanolate (ACS reagent), nitrogen atmosphere, 0.2 ml (0.32 mmol) ether solution of

caesium hydroxide (97%, Aldrich), and potassium carbonate methyl lithium (1.6 mol L^ꢂ1) was added into 5 ml of freshly

(ACS reagent) were used as received. Ethanol, methanol, dried ether solution of compound 2 (227.6 mg, 0.32 mmol) at

acetone, diethylether, n-hexane, toluene, ethyl acetate, tetrahy- room temperature. The reaction mixture was stirred for 2 hours,

drofurane, dichloromethane, isopropyl ether, and all ACS subsequently the solvent was removed under vacuum_ꢃusing a

reagents were used as received. DMF (ACS reagent grade) was membrane pump. The solid obtained was dried at 40 C in an

dried over calcium hydride (CaH₂) for 12 h and distilled in a oil-pump vacuum to yield a white powder of I. Yield: 0.2 g

ꢃ

vacuum. All deuterated solvents for NMR studies were received (z87% of theory); m.p. ꢂ11.1 C (by DSC); TLC (EtOAc–MeOH

1

ꢃ

from Chemotrade (Leipzig).

¼ 6 : 1): R_f¼ 0.45; H NMR (300 MHz, CDCl₃, 20 C, TMS, d,

1,2,3-Tris(dodecyloxy)benzene (1). 10.1 g (0.08 mol) 1,2,3- ppm) (Fig. S1 of ESI†): 0.88 (t, 9H, CH₃, J ¼ 6.23 Hz), 1.26

trihydroxy-benzene and 55.5 g (0.4 mol) K₂CO₃were mixed with (overlapped peaks, 54H, CH₃(CH2)₉), 1.77 (m, 6H, O(CH₂)CH₂),

125 ml of freshly dried DMF in a 250 ml three-necked ask with 3.84 (broad, 4H, 3- and 4-OCH₂), 4.08 (broad, 2H, 2-OCH₂), 6.34

a magnetic stirrer under a nitrogen atmosphere. At 60 ^ꢃC, (broad, 1H, 5-benzene-H), 7.34 (d, 1H, 6-benzene-H, J ¼ 8.28

13

ꢃ

55.2 ml (0.23 mol) 1-bromodecane was added dropwise. The Hz); C NMR (75 MHz, CDCl₃, 20 C, TMS, d, ppm) (Fig. S2 of

reaction mixture was stirred for 5 hours at 60 ^ꢃC. Aer reaction, ESI†): 14.12, 22.72, 25.75, 26.33, 29.44, 29.55, 29.81, 30.53,

the mixture was slowly poured into 600 ml ice water. The 31.99, 32.04, 68.62, 73.67, 106.85, 123.74, 128.96, 141.82,

precipitate was isolated by ltration and then dried in vacuo. 150.02, 155.73; elemental analysis calculated (%): C 70.35, H

The crude product was recrystallized from 300 ml acetone three 10.85; found (%): C 68.28, H 10.97.

times to yield a white powder. Yield: 38.3 g (76% of theory); m.p.

Sodium 2,3,4-tris(dodecyloxy)benzenesulfonate (II). In a

1

39–40 ^ꢃC; TLC (n-hexane–EtOAc ¼ 20 : 1): R_f¼ 0.68; H NMR nitrogen atmosphere, 19.0 mg (0.35 mmol) sodium methoxide

(300 MHz, CDCl₃, 20 ^ꢃC, TMS, d, ppm): 0.88 (t, 9H, CH₃, J ¼ was dissolved in 5 ml of freshly dried methanol. Subsequently,

6.6 Hz), 1.26 (overlapped peaks, 48H, CH₃(CH₂)₈), 1.47 (m, 6H, 250 mg (0.35 mmol) of compound 2 was added, and the reaction

O(CH₂)₂CH₂), 1.78 (m, 6H, O(CH₂)CH₂), 3.90 (overlapped t, 6H, mixture was stirred vigorously for 2 hours at room temperature.

OCH₂, J ¼ 6.3 Hz), 6.55 (d, 2H, 4,6-benzene-H, J ¼ 8.1 Hz), 6.90 (t, Aerwards, the solvent was removed under vacuum using a

13

ꢃ

1H, 5-benzene-H, J ¼ 8.4 Hz); C NMR (75 MHz, CDCl₃, 20 C, membrane pump. The solid obtained was freeze-dried from

TMS, d, ppm): 14.1 (CH₃), 22.7 (CH₃CH₂), 26.1 (OCH₂CH₂CH₂), benzene in vacuo to yield a white powder of II. Yield: 0.21 g (82%

29.4 (CH₃CH₂CH₂CH₂), 29.5 (CH₃CH₂CH₂CH₂(CH₂)₅), 29.9 (1,3- of theory); m.p. 7.7 ^ꢃC (by DSC); TLC (EtOAc–MeOH ¼ 6 : 1): R_f

OCH₂CH₂), 30.3 (2-OCH₂CH₂), 31.9 (CH₃CH₂CH₂), 69.0 (1,3- ¼ 0.34; ¹H NMR (300 MHz, CDCl₃, 20 ^ꢃC, TMS, d, ppm) (Fig. S4

OCH₂), 73.3 (2-OCH₂O), 106.7 (4,6-benzene-C), 123.1 (5-benzene- of ESI†): 0.88 (overlapped t, 9H, CH₃), 1.26 (overlapped peaks,

C), 138.4 (2-benzene-C), 153.4 (1,3-benzene-C).

54H, CH₃(CH₂)₉), 1.67 (broad, 6H, O(CH₂)₂CH₂), 3.79 (broad,

4H, 2,4-OCH₂), 4.03 (broad, 2H, 3-OCH₂), 6.18 (broad, 1H,

2,3,4-Tris(dodecyloxy)benzenesulfonic acid (2). Under

a

nitrogen atmosphere, 1.27 g of 1 (2.0 mmol) was added into 10 5-benzene-H), 7.28 (broad, 1H, 6-benzene-H); ¹³C NMR

ml of concentrated sulfuric acid (95–97%) at 0 ^ꢃC. The reaction (75 MHz, CDCl₃, 20 C, TMS, d, ppm) (Fig. S5 of ESI†): 14.12,

ꢃ

suspension was stirred for 2 hours while the temperature of the 22.74, 25.74, 26.42, 29.51, 29.60, 29.70, 29.95, 30.63, 32.01,

ꢃ

water was gradually increased until 20 C. The suspension was 68.54, 73.47, 75.42, 106.61, 123.85, 128.90, 141.77, 150.20,

poured slowly into 50 ml of ice water. The resulting light-yellow 155.30; elemental analysis calculated (%): C 68.61, H 10.59;

suspension was kept at 4 ^ꢃC for 1 hour and the precipitate was found (%): C 67.32, H 11.16.

isolated by ltration and pre-dried in vacuo. The crude product

was recrystallized from 30 ml acetone at 4 C three times and nitrogen atmosphere, 61.6 mg (0.88 mmol) potassium meth-

Potassium 2,3,4-tris(dodecyloxy)benzenesulfonate (III). In a

ꢃ

dried in vacuo to yield a white powder. Yield: 1.24 g (87.2% of oxide was dissolved in 10 ml of freshly dried methanol.

1

theory); m.p. 65.0–65.5 ^ꢃC; H NMR (300 MHz, CDCl₃, 20 ^ꢃC, Subsequently, 625.4 mg (0.88 mmol) of compound 2 was added,

TMS, d, ppm): 0.88 (t, 9H, CH₃, J ¼ 6.7 Hz), 1.27 (overlapped and the reaction mixture was stirred vigorously for 2 hours at

peaks, 48H, CH₃(CH₂)₈), 1.45 (m, 6H, O(CH₂)₂CH₂), 1.83 (m, 6H, room temperature. Finally, the solvent was removed under

O(CH₂)CH₂), 4.00 (overlapped t, 4H, 3- and 4-OCH₂), 4.29 (t, 2H, vacuum using a membrane pump. The solid obtained was

2-OCH₂, J ¼ 6.98 Hz), 6.69 (d, 1H, 5-benzene-H, J ¼ 9.06 Hz), freeze-dried from benzene in vacuo to yield a white powder of

7.54 (d, 1H, 6-benzene-H, J ¼ 8.85 Hz); ¹³C NMR (75 MHz, III. Yield: 0.62 g (94% of theory); m.p. 17.8 ^ꢃC (by DSC); TLC

CDCl₃, 20 C, TMS, d, ppm): 14.13, 22.70, 25.77, 26.10, 29.11, (EtOAc–MeOH ¼ 6 : 1): R_f¼ 0.16; ¹H NMR (300 MHz, CDCl₃, 20

ꢃ

This journal is © The Royal Society of Chemistry 2014

Soft Matter, 2014, 10, 1746–1757 | 1749

View Article Online

Soft Matter

Paper

^ꢃC, TMS, d, ppm) (Fig. S6 of ESI†): 0.88 (overlapped t, 9H, CH₃),

1.26 (overlapped peaks, 54H, CH₃(CH₂)₉), 1.75 (m, 6H,

OCH₂CH₂), 3.84 (broad, 4H, 3- and 4-OCH₂), 4.07 (broad, 2H,

2-OCH₂), 6.29 (d, 1H, 5-benzene-H, J ¼ 8.28 Hz), 7.35 (d, 1H,

6-benzene-H, J ¼ 8.70 Hz); ¹³C NMR (75 MHz, CDCl₃, 20 ^ꢃC,

TMS, d, ppm) (Fig. S7 of ESI†): 14.12, 22.73, 25.83, 26.39, 29.49,

29.58, 29.92, 30.62, 32.03, 68.56, 73.55, 74.96, 106.79, 123.63,

130.71, 142.00, 150.34, 155.23; elemental analysis calculated

(%): C 67.33, H 10.36; found (%): C 66.36, H 10.49.

Caesium 2,3,4-tris(dodecyloxy)benzenesulfonate (IV). In a

100 ml two-necked ask with a magnetic stirrer in a nitrogen

atmosphere, 0.71 g (1.00 mmol) of 2 was dissolved in 50 ml

methanol at 60 ^ꢃC. Subsequently, 0.34 g (2.0 mmol) of caesium

hydroxide was added. The reaction mixture was reuxed for 30

minutes, and was ltered hot. The ltrate was cooled to room

temperature, the precipitate was isolated by ltration and

washed with water until the washing water became neutral.

Aer pre-drying in vacuo, the crude product was twice crystal-

lized from n-hexane to obtain a white powder of IV. Yield: 0.8 g

Fig. 1 DSC scans of lithium 2,3,4-tris(dodecyloxy)benzenesulfonate.

The inset shows high-temperature low-heat transformation on ﬁrst

heating.

a crystalline one. The proposed structure, then, could be

described either by two coexistent crystalline layer-type meso-

phases or by the formation of a complex smectic mesophase in

which ordered layers of in-plane benzenesulfonate dendrons

are separated by water. With increasing temperature water

evaporates leading to the drop in smectic interlayer distance

with a subsequent reorganization of the benzenesulfonic

groups which manifests itself in a drastic change of the wide-

angle scattering patterns as well (inset in Fig. 2).

It should be noted, however, that nascent crystalline phases

formed by the “as-received” samples undergo irreversible tran-

sitions and thus can be ascribed to the eﬀect of solvents or

hydrate complexes formed during the synthesis of the

ꢃ

(98% of theory); m.p. 50.3 C (by DSC); TLC (CHCl₃–MeOH ¼

6 : 1): R_f¼ 0.44; ¹H NMR (300 MHz, CDCl₃, 20 ^ꢃC, TMS, d, ppm)

(Fig. S8 of ESI†): 0.88 (t, 9H, CH₃, J ¼ 6.6 Hz), 1.26 (overlapped

peaks, 48H, CH₃(CH₂)₈), 1.47 (broad, 6H, O(CH₂)₂CH₂), 1.72 (m,

6H, OCH₂CH₂), 3.90 (overlapped peaks, 6H, OCH₂, J ¼ 6.6 Hz),

7.03 (s, 2H, 2,6-position); ¹³C NMR (75 MHz, CDCl₃, 20 ^ꢃC, TMS,

d, ppm) (Fig. S9 of ESI†): 14.12, 22.71, 26.27, 29.44, 29.62, 29.80,

31.96, 73.42, 106.7, 139.52, 140.03, 152.99; elemental analysis

calculated (%): C 59.84, H 9.21; found (%): C 59.71, H 9.23.

Results and discussion

Structure and phase behaviour of lithium 2,3,4-

tris(dodecyloxy)benzenesulfonate

The lithium cation is the smallest amongst the alkali metals,

and the shape of an isolated lithium 2,3,4-tris(dodecyloxy)ben-

zenesulfonate (I) molecule can be reliably described as a sharp

V-cone. As revealed by diﬀerential scanning calorimetry (Fig. 1),

three subsequent phase transitions of compound I take place

upon its heating at 93 ^ꢃC, 118 ^ꢃC and 161 ^ꢃC (DH₁¼ 48.1 kJ

mol^ꢂ1; DH₂¼ 4.7 kJ mol^ꢂ1and DH₃¼ 0.27 kJ mol^ꢂ1respec-

tively). The latter one corresponds to the material isotropiza-

tion, and the second signal is accompanied by the simultaneous

loss of birefringence, corresponding to the transition into an

optically isotropic phase.

X-Ray diﬀractograms of the “as-received” compound I are

characterized by two sets of reections, consisting of three

peaks with d-spacing ratio 1 : 2 : 3 (Fig. 2, curve 1, see ESI, Table

S1†). They correspond to interlayer distances of L₁¼ 3.64 nm

and L₂¼ 3.27 nm respectively. Separate analysis of the phase

behaviour of materials either in the vacuum of the X-ray camera

or in an especially designed air-conditioned holder allowed us

to explain unequivocally this pattern by the presence of two

diﬀerent layered structures, the latter one (L₂¼ 3.27 nm)

formed in the presence of water molecules by self-assembly of

benzenesulfonate derivatives' hydrates, and the former one

organized by benzenesulfonate dendrons only. The analysis of

Fig. 2 Small-angle scattering for lithium 2,3,4-tris(dodecyloxy) ben-

zenesulfonate at 20 ^ꢃC (1), 95 ^ꢃC (2), 130 ^ꢃC (3), 170 ^ꢃC (4), 220 ^ꢃC (5)

and 200 ^ꢃC cooling (6); one-digit indices show the reﬂections of layer

structures, three-digit ones correspond to a high-temperature cubic

ꢀ

mesophase of Pm3n symmetry. The inset shows WAXS patterns of the

wide-angle X-ray patterns revealed that the as-received phase is material at 20 ^ꢃC (black) and 90 ^ꢃC (red).

1750 | Soft Matter, 2014, 10, 1746–1757

This journal is © The Royal Society of Chemistry 2014

View Article Online

Paper

Soft Matter

compounds. Such “virgin phases” hence do not play a signi- from a tapered slice, found in the smectic layers, to a conic one

cant role in the studies of the materials phase behaviour. as present in the cubic mesophase. Reconstruction of electron

Heating the samples up to 95 ^ꢃC is accompanied by a density distribution revealed that the unit cell of the cubic

decrease of the interlayer distances with thermal expansion mesophase contained eight diﬀerent scatterers, two of them

coeﬃcients b₁¼ ꢂ2.2 ꢁ 10^ꢂ4K^ꢂ1and b₁¼ ꢂ2 ꢁ 10^ꢂ4K^ꢂ1(type A) occupying special positions of the bcc lattice, while six

down to L₁¼ 3.58 nm and L₂¼ 3.22 nm respectively (see ESI, others (type B) occupied special positions on cubic faces (Fig. 4).

Table S2†).

It should be noted that these two types of scatterers have

This eﬀect is rather common for liquid crystalline materials, substantially diﬀerent structures with diﬀerent distributions of

and can be explained by eﬀective shrinking of aliphatic tails the electron density in their cores. The higher the temperature

with increasing temperature due to the accumulation of gau- the less such diﬀerence is pronounced. Such an eﬀect is fol-

che-conformers.

lowed by a merger of micelles in closest special position e (note

According to molecular modelling, the length of the lithium an arrow in Fig. 4a) resulting in a phase transition to a primitive

2,3,4-tris(dodecyloxy) benzene-sulfonate ion-pair is 2.25 nm if cubic lattice with unit cell parameter a ¼ 2.585 nm (curves 4 and

alkyl tails are fully extended. This is substantially smaller than 5 in Fig. 2, corresponding data summarized in Tables S4 and S5

the interlayer distance, thus it was concluded that the observed of ESI†). It is important to note that in the transition point, the

ꢀ

smectic layers have a bilayer structure (cf. Fig. 3). The diﬀerence primitive cubic lattice parameter is 0.56 of that in the Pm3n

between the observed interlayer distance and the doubled phase, the value being in very good agreement with the value of

molecule length (Dl ¼ 0.9 nm) can be attributed to a partial 5^1/2/4 calculated from stereometric considerations. This tran-

inter-penetration of alkyl tails of opposing molecules. It should sition is accompanied by a sharp increase of the average

be noted that in the temperature interval of the rst transition domain size, manifesting itself by a grain structure of reec-

ꢃ

(ꢀ95 ^ꢃC) reections of the crystal-hydrate structure lose their tions at temperatures higher than 170 C.

intensity abruptly revealing the order–disorder transition in the

Further heating of compound I leads to a decrease of the

layer structure due to the loss of water molecules which has cubic lattice parameter down to a ¼ 2.56 nm at 220 ^ꢃC. The

been conrmed by the results of thermogravimetric analysis thermal expansion coeﬃcient of the micellar primitive cubic

(see ESI, Fig. S3 and S10†).

phase was calculated to be b ¼ ꢂ1.5 ꢁ 10^ꢂ4K^ꢂ1, which, by the

Moreover, a new set of four reections appears, yielding order of value, is very common for liquid crystalline phases.^31,32

2

d-spacing square ratio of d₂₀₀²: d₂₁₀²: d₂₁₁

:

d₃₁₀

¼

For example, the high-temperature cubic mesophase of Ia3d

ꢀ

a

ꢀ

4 : 5 : 6 : 10, which is highly specic for the cubic phase of Pm3n symmetry formed by (C₆F₁₃–C₂H₄O–C₆H₄–CH₂O)₃–C₆H₃–CO–

symmetry (a ¼ 5.95 nm), see ESI, Table S3.† On further heating (OCH₂CH₂)₄–OH, i.e. gallic acid cores with long partially uo-

ꢃ

to 130 C the transition into the cubic mesophase has become rinated alkyl tails, is characterized by an expansion coeﬃcient

fully developed, one can observe at least eleven reections of b ¼ ꢂ7 ꢁ 10^ꢂ4K^ꢂ1

.

^33–35Such thermal behaviour of a material

ꢀ

(Fig. 2, curve 3) indexed by the Pm3n symmetry unit cell. Thus,

the loss of optical birefringence can unambiguously be related

to the transition from a smectic bilayer structure, formed during

the synthesis of the compounds with participation of some

crystal-hydrates and subsequent drying, into a cubic meso-

phase. Such a transition is probably due to the change of the

volume fractions of the aliphatic chains within the mesogenic

groups, which, in turn, denes an equilibrium interphase

curvature.^29,30It should also be noted that an increasing mobility

of alkyl tails leads to the change of the general molecular shape

Fig. 4 Electron density maps of lithium 2,3,4-tris(dodecyloxy)ben-

zenesulfonate at 100 ^ꢃC (a), 130 ^ꢃC (b) and 170 ^ꢃC (c). Electron

ꢀ

density distributions in the Pm3n phase were calculated using –, +

(ꢂ), +, +, ꢂ (+) phase combination of reﬂections 200, 210 (120), 211,

422, 430 (340) leading to minimum values of the fourth moment of

electron density ﬂuctuations (3.89 and 2.44 at 100 ^ꢃC and 130 ^ꢃC

respectively). On (100) planes, cross-sections of nodal surfaces are

shown in (c), which were calculated using +, +, ꢂ, ꢂ, ꢂ, ꢂ phase

Fig. 3 Molecular modeling of a bilayer structure in lithium 2,3,4- combination of 100, 110, 111, 200, 210, 222, 211 of a primitive cubic

tris(dodecyloxy)benzenesulfonate.

lattice (h(Dr)⁴i ¼ 2.13).

This journal is © The Royal Society of Chemistry 2014

Soft Matter, 2014, 10, 1746–1757 | 1751

View Article Online

Soft Matter

Paper

can be explained by increasing concentrations of gauche- temperatures. During the rst heating run two peaks are

conformers in the alkyl tails of micellar periphery with observed at 54 ^ꢃC and 103.5 ^ꢃC (compare with 93 ^ꢃC and 118 ^ꢃC

increasing temperature. This in turn leads to a soening of the for compound I). Fusion heats are also lower for the sodium salt

supramolecular aggregates, followed by their increased mobility (DH₁¼ 19.8 kJ mol^ꢂ1; DH₂¼ 3.7 kJ mol^ꢂ1).

and by corresponding compactization of the cubic mesophase.

On cooling scans one can see two exothermal peaks at 80 ^ꢃC

ꢂ1

ꢃ

Cooling the samples of compound I down to room temperature and ꢂ8 C with heats of fusion of 6.6 and 2.1 kJ mol respec-

resulted in the disappearance of the grain structure of cubic tively. The substantial diﬀerence of the high-temperature peak

mesophase reections, and to the redistribution of intensities position on heating and cooling scans can be explained by

ꢀ

of the reections of Pm3n symmetry.

retarded kinetics of mesophase development at the heating/

For instance, the 310 reection completely disappeared at T cooling rates used in experiments. Similar to lithium 2,3,4-

ꢃ

¼ 30 C. It should be emphasized that during all X-ray experi- tris(dodecyloxy) benzene sulfonate, the low-temperature peak of

ments, the studied samples were kept in a vacuum. At room compound II can be ascribed to a crystal–mesophase transition.

temperature the cubic mesophase did not disappear over a

Polarizing optical microscopy of as-received samples

period of several days of investigation, while annealing in the revealed the existence of crystalline domains up to 50 ^ꢃC. In

presence of air causes its collapse within one day. Diﬀraction support to these observations, SAXS patterns of sodium 2,3,4-

patterns of such annealed samples contain only one broad tris(dodecyloxy)benzenesulfonate at room temperature (Fig. 6

reection, which corresponds to an interlayer distance of L ¼ and Table S6 of ESI†) contain 15 narrow reections, which may

3.45 nm, indicating a partial recovery of the smectic order.

be indexed by a cubic crystal lattice of I23 symmetry with lattice

ꢃ

Thus, lithium 2,3,4-tris(dodecyloxy)benzenesulfonate is parameter a ¼ 6.18 nm. Heating to 50 C is followed by its

characterized by a bilayer crystalline structure of smectic type, increase up to 6.24 nm (b ¼ 4.2 ꢁ 10^ꢂ4K^ꢂ1).

ꢀ

which transforms to the micellar cubic mesophase of Pm3n

Further heating leads to the development of a new set of

symmetry at 130 ^ꢃC. The transition is probably driven by a reections with d-spacing squares ratio d₁₀²: d₁₁²: d₂₀²: d₂₁

¼

2

change of the over-all shape of the mesogenic unit due to an 1 : 3 : 4 : 7 (Fig. 7 and Table S7 of ESI†), which is characteristic

eﬀective decrease of the length of its aliphatic tails with for a 2D hexagonal l_ꢃattice. Wide-angle reections disappear

increasing temperature. At 160 ^ꢃC, a more uniform distribution simultaneously at 50 C, unambiguously indicating the devel-

of alkyl-material between the micelles in diﬀerent positions opment of a disordered columnar mesophase. Its domains have

leads to the transformation to a primitive cubic mesophase.

an average size of 45.5 nm, being estimated by means of the

Hoseman equation from the integral half-widths of 10 and 20

reections, and its paracrystallinity parameter was equal to 3%.

The cylinder diameter near the transition temperature was

calculated to be 3.44 nm at 100 ^ꢃC, close to the observed

interlayer distance in the smectic phase, formed by compound

I. Thus, we can draw an important conclusion that in sodium

2,3,4-tris(dodecyloxy)benzenesulfonate the eﬀective volume

fraction of the mesogens' anionic groups is larger than that of

compound I. So, the microphase segregation of the aromatic

and aliphatic parts of benzenesulfonate molecules requires the

Structure and phase behaviour of sodium 2,3,4-

tris(dodecyloxy)benzenesulfonate

The sodium ion, compared to the lithium cation, has a larger

˚

diameter (1.02 A and 0.76 A respectively). In this section we shall

explore how such a change of monodendrons' counter-ion size

would aﬀect the materials' phase behavior. DSC curves of

sodium 2,3,4-tris(dodecyloxy)benzenesulfonate have much in

common with thermograms of compound I (Fig. 5), though in

the former case the transition peaks are found at lower

Fig.

6 Small-angle scattering diﬀractograms of sodium 2,3,4-

Fig. 5 DSC scans of sodium 2,3,4-tris(dodecyloxy)benzenesulfonate: tris(dodecyloxy)benzenesulfonate at room temperature. Indices of

(1) ﬁrst heating, (2) second heating, and (3) ﬁrst cooling.

1752 | Soft Matter, 2014, 10, 1746–1757

reﬂections of the cubic crystalline I23 phase are shown.

This journal is © The Royal Society of Chemistry 2014

View Article Online

Paper

Soft Matter

Fig. 8 (111) projection of electron density distribution in the cubic

Pm3n phase of 2,3,4-tris(dodecyloxy)benzenesulfonate at 170 ^ꢃC.

ꢀ

Fig. 7 Small-angle diﬀractograms of sodium 2,3,4-tris(dodecyloxy)

benzenesulfonate at 100 ^ꢃC (1), 140 ^ꢃC (2), 240 ^ꢃC (3) and at room

temperature after cooling from 250 ^ꢃC (4).

Annealing of sodium 2,3,4-tris(dodecyloxy)benzenesulfonate

samples at room temperature in an air atmosphere leads to the

substantial increase of column diameter up to D ¼ 3.77 nm,

possibly due to the relaxation of the aliphatic matrix.

Another explanation of this fact was mentioned when the

temperature behaviour of lithium benzenesulfonate has been

discussed, i.e. an impact of water molecules on the structure of

supramolecular aggregates formed by benzenesulfonate meso-

gens. A detailed study of water absorption eﬀects is a topic for a

separate publication, which will follow later.

curvature of the interface between the ionic and electrically

neutral micro-segregated phases to be of non-zero value at lower

temperatures. This explains why in compound I a smectic phase

is developed at low temperatures, while in compound II a

columnar phase is formed. It should also be noted that the

observed column diameter is smaller than double the length of

sodium 2,3,4-tris(dodecyloxy)benzenesulfonate due to the

reasons aforementioned.

Further heating of the samples of compound II to the

temperatures higher than 230 ^ꢃC leads to the transformation of

The disordered columnar mesophase exists up to 118 ^ꢃC. At

140 ^ꢃC its high reections disappear, and the rst, strongest one

shis to the wider angle region (d ¼ 2.91 nm), gaining simul-

taneously in half-width. With further heating, three reections

start to be resolved, the d-spacing squares ratio being equal to

ꢀ

the Pm3n phase to the plastic crystal mesophase with a primi-

tive cubic lattice (Fig. 7, curve 3), the eﬀect is similar to that in

compound I. However, the transition temperature is substan-

tially_ꢃhigher for_ꢃsodium 2,3,4-tris(dodecyloxy)benzenesulfonate

(235 C vs. 161 C for the lithium salt). It should also be noted

that a high-temperature primitive cubic mesophase is preserved

until sample destruction – cooling the samples aer 250 ^ꢃC does

not lead to the restoration of any ordered state due to the

compound disintegration (Fig. 7, curve 4).

2

d₂₀₀²: d₂₁₀²: d₂₁₁¼ 4 : 5 : 6, testifying to the development of

ꢃ

ꢀ

the cubic mesophase of Pm3n symmetry (a ¼ 6.51 nm at 140 C:

Fig. 7, curve 2, and Table S8 of ESI†). In compound I, the

transition to the cubic mesophase of the same symmetry was

observed (a ¼ 5.95 nm) at the same temperature. The larger

micellar size in compound II is due to the larger cations in the

centres of the molecular aggregates. The temperature behaviour

of the cubic phase of lithium- and of sodium benzenesulfonate

salts is quite similar – in both materials a drastic growth of

Structure and phase behaviour of potassium 2,3,4-

tris(dodecyloxy)benzenesulfonate

ꢃ

domain size has been observed at temperatures above 150 C.

The phase behavior of potassium 2,3,4-tris(dodecyloxy)benze-

It is important to note the nature of the transition from nesulfonate (III) diﬀers substantially from the temperature

columnar to cubic phase revealed by Fig. 8, which represents behaviour of the lithium and sodium benzenesulfonate salts.

ꢀ

(111) projection of the electron density distribution in the Pm3n Thermograms of the rst heating of compound III (Fig. 9) reveal

phase of compound II at 170 C. One can see the remnants of four endothermal peaks at 18 ^ꢃC (DH₁¼ 22 kJ mol^ꢂ1), 47 ^ꢃC

ꢃ

late cylinders revealing the presence of a rotation axis of the (DH₂¼ 5.8 kJ mol^ꢂ1), 123 ^ꢃC (DH₃¼ 4.6 kJ mol^ꢂ1), and 171 ^ꢃC

sixth order in the cubic lattice.

(DH₄¼ 3.2 kJ mol^ꢂ1). Fusion heats of the last three of them are

ꢃ

Cooling compound II to 120 C leads to the reversible tran- common for mesophase–mesophase transitions. However,

sition into the disordered columnar mesophase, which is stable polarizing optical microscopy reveals a fan-shape texture, which

down to room temperature. Its column diameter changed to D ¼ is characteristic for smectic and columna_ꢃr mesophases, to be

3.68 nm, corresponding to a temperature expansion coeﬃcient stable from room temperature up to 170 C. Above this point

being equal to ꢂ3 ꢁ 10^ꢂ4K^ꢂ1. Negative values of temperature samples of compound III are optically isotropic. This disap-

expansion coeﬃcient in columnar mesophases have also been pearance of the liquid crystalline texture is reversible.

observed in compounds on the basis of gallic acid.^36–38This fact

was explained by an extension of columns along their axis. tris(dodecyloxy)benzenesulfonate powder reveal more than

Small-angle X-ray patterns of nascent potassium 2,3,4-

This journal is © The Royal Society of Chemistry 2014

Soft Matter, 2014, 10, 1746–1757 | 1753

View Article Online

Soft Matter

Paper

Fig. 11 Small-angle scattering for potassium 2,3,4-tris(dodecyloxy)

benzenesulfonate at 100 ^ꢃC (1), 120 ^ꢃC (2), 170 ^ꢃC (3) and 200 ^ꢃC (4);

two-digit indices show the reﬂections of the disordered columnar

mesophase, three-digit ones correspond to the 3D hexagonal crys-

talline phase.

Fig. 9 DSC scans of potassium 2,3,4-tris(dodecyloxy)benzenesulfo-

nate: (1) ﬁrst heating, (2) second heating, and (3) ﬁrst cooling.

twenty narrow reections, indexed by a cubic crystal lattice of

I23 symmetry with a lattice parameter a ¼ 6.39 nm (Fig. 10 and

Table S9 of ESI†).

columnar phase. The transition includes the loss of crystalline

order inside the columns, followed by the disappearance of

intracolumnar register. The arrangement of the columns within

the 2-D lattice, however, is retained (see ESI, Table S11†).

Aer cooling down to room temperature the intra-columnar

order recovers, though with axial repeat c ¼ 4.84 nm, compared

with c ¼ 4.80 nm at 50 ^ꢃC on the transition to the columnar

phase during the rst heating.

ꢃ

Heating up to 50 C is followed by the transition to a crys-

talline columnar mesophase with a ¼ 3.60 nm and c ¼ 4.83 nm

(P6/mmm symmetry), as is revealed by SAXS and WAXS patterns

(Fig. 11 and Table S10 of ESI†). The average domain size, esti-

mated with the use of Hosemann's equation, is 52 nm, the

calculated paracrystallinity parameter equal to 2.9%. As it is

common for supramolecular aggregates formed by alkali salts

of asymmetric benzenesulfonate acid, the column diameter is

substantially smaller than twice the length of the molecule (2.34

nm per molecule as is revealed by molecular modeling). This

eﬀect can either be explained by a tilt of the dendrons relative to

the column axis (pine-tree structure) or by the interdigitation of

aliphatic tails of dendrons belonging to neighbouring columns.

Further heating up to 123 ^ꢃC leads to the loss of wide-angle

reections, as well as of non-equatorial small-angle ones

(Fig. 11, curve 2), indicating the development of the disordered

Heating compound III up to 171 ^ꢃC leads to the development

ꢀ

of a cubic phase of Pm3n symmetry with the lattice parameter a

¼ 6.48 nm (Fig. 11, curve 4, and Table S12 of ESI†) followed by a

primitive cubic mesophase. The micellar size follows the trend

of the other investigated alkali 2,3,4-tris(dodecyloxy)benzene-

sulfonates. With heating, the domain sizes become larger, as is

manifested by the grain structure of small-angle reections.

Cooling samples of compound III from the primitive cubic

mesophase leads to the formation of a disordered columnar

mesophase D_hd(d ¼ 3.37 nm), which is stable down to the room

temperature. On subsequent heating, a high-temperature tran-

sition to a cubic phase is observed, recognized from the increase

of the relative intensity of 210 reection due to the redistribution

of electron density in the material. Thus, in compound III,

compared with the other investigated materials, a low-tempera-

ture ordered columnar mesophase is observed. Moreover, the

stability of the disordered columnar mesophase increases up to

171 ^ꢃC (104 ^ꢃC for sodium salt). However, the temperature

behavior of potassium 2,3,4-tris(dodecyloxy)benzenesulfonate is

quite complex: the “as-received” material forms a nascent crystal

phase of I23 symmetry, which transforms at 50 ^ꢃC to a crystalline

columnar phase. At 123 ^ꢃC the order inside columns is lost, and

ꢃ

ꢀ

at 173 C a cubic mesophase of Pm3n symmetry is developed.

Structure and phase behaviour of caesium 2,3,4-

tris(dodecyloxy)benzenesulfonate

Fig. 10 Small-angle scattering of potassium 2,3,4-tris(dodecyloxy)

benzenesulfonate at room temperature. Indices of reﬂections of the

cubic crystalline I23 phase are shown.

From the data shown above it is clear that gradual changes of

the radius of the cation, which lies in the vicinity of the focal

1754 | Soft Matter, 2014, 10, 1746–1757

View Article Online

Paper

Soft Matter

point of the anionic part of the molecule, dene changes of the

eﬀective mesogen shape, which, in turn, aﬀects the phase

behavior of the material. Caesium is the largest of alkali cations,

whose salt has been studied in this paper. On the rst heating

scans of caesium 2,3,4-tris(dodecyloxy)benzenesulfonate (IV)

ꢃ

one can see the only transition at 54 C, having a substantial

heat eﬀect of 53 kJ mol^ꢂ1. It is not reproduced in subsequent

heating–cooling scans. The second heating thermograms

revealed two reversible transitions, one at 8 ^ꢃC (DH ¼ 9.3 kJ

mol^ꢂ1) and a second at 216 C (DH ¼ 3.4 kJ mol^ꢂ1). At about

ꢃ

210 ^ꢃC a broken fan texture also disappears, which is charac-

teristic for optical micrographs of compound IV in a wide

temperature range between the rst and second transitions.

In “as-received” samples of caesium 2,3,4-tris(dodecyloxy)-

benzenesulfonate a cubic phase of I23 symmetry with a lattice

parameter a ¼ 6.69 nm is observed. It should be remembered

that at room temperature a cubic mesophase of the same

symmetry has been observed in compounds II and III with the

lattice parameters a ¼ 6.18 nm and 6.39 nm respectively. Thus

the structure of crystal-hydrates formed in “as-received”

samples of alkali salts of 2,3,4-tris(dodecyloxy)benzenesulfonic

acid is similar.

To verify this hypothesis we have carried out comparative

analysis of the structure of the as-received sample of compound

IV and of the same sample aer a day in a vacuum. In the latter

case small-angle reections were signicantly wider, high

reections have not been observed at all, pointing to the dete-

rioration of crystal order due to the loss of hydrate crystals.

Heating up to 25 ^ꢃC leads to the irreversible transition to the

Fig. 12 Small-angle scattering for caesium 2,3,4-tris(dodecyloxy)-

benzenesulfonate at 50 ^ꢃC (1), 230 ^ꢃC (2), and 30 ^ꢃC after cooling from

230 ^ꢃC (3); reﬂections of ordered and disordered columnar meso-

phases (HO and HD respectively) are shown.

transition peak has a maximum at +8 ^ꢃC; however the transition

is rather broad and extends to 20–25 ^ꢃC. Thus, it can be

supposed that this DSC peak corresponds to an order–disorder

transition inside the columns.

The detailed analysis of the structure of the columns in the

ordered and disordered columnar mesophases, as well as the

mechanism of transition between them will be the subject of a

diﬀerent paper in preparation now.

To conclude, in caesium 2,3,4-tris(dodecyloxy)benzenesul-

fonate at room temperature a mix of coexisting ordered and

disordered columnar phases is observed. It should also be

noted that in compound IV, unlike lithium, sodium and

potassium salts of 2,3,4-tris(dodecyloxy)benzenesulfonic acid, a

high-temperature cubic mesophase does not exist.

This is attributed to the substantial divergence of the shape

of mesogen IV from the cone one. Therefore, development of

spherical micelles becomes impossible, while columnar meso-

phases are stable up to isotropisation.

disordered columnar phase, characterized by four observed

2

reections with d-spacing square ratio d₁₀²: d₁₁²: d₂₀²: d₂₁

¼

1 : 3 : 4 : 7 (Fig. 12).

Column diameters change linearly from 3.82 nm at 30 ^ꢃC to

3.43 nm at 230 ^ꢃC, the corresponding expansion coeﬃcient

being equal to b ¼ (ꢂ6.6 ꢅ 0.3) ꢁ 10^ꢂ4K^ꢂ1. Crystalline domains

grow simultaneously from 37.5 nm to 76.8 nm, while the par-

acrystalline parameter only weakly changed (3% at 30 ^ꢃC and

3.5% at 230 ^ꢃC).

The intensities of reections of the disordered columnar

phase did not change up to 230 ^ꢃC. At this temperature a

transition to the isotropic state has been observed. The rst

reection of the columnar mesophase widened up to form an

amorphous halo with a corresponding d-spacing of 2.87 nm.

Cooling the material down to room temperature leads to the

development of complex scattering patterns, consisting of two

sets of reections, one of them corresponding to a column

diameter of 4.2 nm, and the other one to a diameter of 3.82 nm.

It is important that the relative intensities of reections

belonging to the diﬀerent sets strongly depend on both

temperature and the thermal history of the sample. The higher

the temperature, the lower the intensity of the former set of

reections, and the stronger reections of the latter set. At

ꢀ35 ^ꢃC only one set of reections remains. Such behavior of

small-angle reections is also in trend with the intensities of

ꢃ

wide-angle reections, which disappear at ꢀ35 C. To explain

this behavior, one should investigate in detail DSC scans of the

Fig. 13 DSC scans of caesium 2,3,4-tris(dodecyloxy)benzenesulfonate:

second heating run of compound IV (Fig. 13). The rst (1) ﬁrst heating and (2) second heating.

Soft Matter, 2014, 10, 1746–1757 | 1755

View Article Online

Soft Matter

Paper

organic anion to the inorganic cation leads to systematic

changes in material temperature behaviour. The diameter of

the cation varied in the range from ꢀ0.102 nm to 0.167 nm,

while the length of the aliphatic tails was the same in all four

compounds studied. Mesophases of diﬀerent types and

dimensionalities were discovered: 1-D (smectic bilayers), 2-D

(ordered and disordered columnar phases), and 3-D (crystalline

ꢀ

hexagonal, high-temperature plastic crystals of Pm3n symmetry,

low-temperature cubic phase of I23 symmetry). Below we carry

out a comparative analysis of the structural and thermal char-

acteristics of such mesophases, as well as of transitions between

them. Phase sequences are represented in the phase diagram,

associating the variation of the cation size with the temperature

behaviour of the material as shown in Fig. 14. It should be

emphasized that the phase behaviour of compounds I–IV on the

rst heating (Fig. 14a) is complicated by the formation of crystal

hydrates and by their transformation due to the loss of water

molecules. Thus, to establish an eﬀect of the size of the focal

group on the temperature behaviour of the material, we focus

on the phase diagrams derived from the second heating scans

(Fig. 14b) in our following considerations.

Several main conclusions can be drawn.

(1) Cubic types prevail, when the size of the cation is small,

and the shape of the mesogenic group is close to a conic one.

With increasing cation size, the shape becomes more tapered,

and the columnar mesophase appears to be more stable.

(2) Increasing the cation size leads to a higher temperature of

transition to the cubic plastic crystal mesophase. This fact can

be explained by a variation of the free energy functional of self-

assembled aggregates formed by diﬀerent salts of benzene-

sulfonic acid. Obviously, an increasing dendron size leads to a

stabilization of the columnar mesophase. Geometric aspects of

this explanation can be found in ref. 18.

Fig. 14 Phase diagrams (cation radius R_ionvs. temperature) of alkali

salts of 2,3,4-tris(dodecyloxy)benzenesulfonic acid as derived from the

ﬁrst (a) and following (b) heating cycles.

(3) The column diameters of the disordered columnar mes-

ophases do not depend signicantly on the size of the cation in

alkali salts of 2,3,4-tris(dodecyloxy)benzenesulfonic acid.

(4) Ordered and disordered columnar phases can coexist in a

rather wide temperature range, their lattice parameters (i.e.

cylinder diameters) being substantially diﬀerent. Moreover,

intensity ratios of reections are diﬀerent as well, indicating

that order-disorder transition inside columns is followed by (or

caused by) the redistribution of electron densities.

It is also interesting to analyze the temperature dependen-

cies of the lattice parameters in the studied compounds

(Fig. 15). In any case at ambient temperature the cylinder

diameter is in the range of 3.4–4.4 nm.

Fig. 15 Temperature dependencies of lattice parameters for alkali

salts of 2,3,4-tris(dodecyloxy)benzenesulfonic acid: interlayer distance

ꢀ

for smectic structures L (-); parameter a of the Pm3n cubic meso-

phase (C); column diameter D in crystalline (+), ordered (;) and

disordered columnar hexagonal mesophases (:). Cyan markers stand

for lithium salt, red – for sodium, black – for potassium and blue – for

caesium.

The loss of the order inside the columns does not lead to a

considerable change of the thermal expansion coeﬃcient b.

However, the most pronounced feature of the D_hoto D_hdtran-

sition is a substantial (up to 10% with compound IV) jump

down of the column diameter. This eﬀect may be due to the

Discussion and conclusions

Summarizing the discussion of the interrelation between the shi of tapered dendrons to the centre of the cylinder, caused by

shape of the molecule, structure of supramolecular aggregates an increased mobility of the dendrons and a corresponding loss

and phase behaviour of the material, one can make several of the order in the interior ion rim. Consequently, the most

important generalizations. Our investigation of the structure dramatic change is observed close to the column axis, where the

and phase behaviour of alkali 2,3,4-tris(dodecyloxy)benzene- low electron density of the ordered columnar mesophase

sulfonate revealed that variation of the volume ratio of the switches to a high electron density in the disordered one.³⁵Such

1756 | Soft Matter, 2014, 10, 1746–1757

View Article Online

Paper

Soft Matter

a change manifests itself in diﬀerent ratios of the reection 16 L. Leibler, Macromolecules, 1980, 13, 1602.

intensities described above. The supramolecular cylinder can 17 M. W. Matsen and F. S. Bates, Macromolecules, 1996, 29,

be regarded as a switchable ion channel, to be closed at higher

temperatures and opened at lower ones.

1091.

18 G. Ungar, Y. Liu, X. B. Zeng, et al., Science, 2003, 299, 1208.

In conclusion, the synthesized 2,3,4-tris(alkoxy)benzenesul- 19 R. Hosemann and A. M. Hindeleh, J. Macromol. Sci., Part B:

fonates may be considered to belong to the class of “smart” Phys., 1995, 34(4), 327.

thermotropic materials, exhibiting a rich variety of phase 20 A. Guinier, Theorie et Technique de la Radiocristallographie,

behavior, including phases of diﬀerent symmetries and

Dunod, Paris, 1956.

dimensionalities.

21 V. S. K. Balagurusamy, G. Ungar, V. Percec and G. Johanson,

J. Am. Chem. Soc., 1997, 119, 1539.

22 P. Mariani, V. Luzzati and H. Delacroix, J. Mol. Biol., 1988,

204, 165.

Notes and references

23 H. Sun, J. Phys. Chem. B, 1998, 102, 7338.

24 H. Sun, P. Ren and J. R. Fried, Comput. Theor. Polym. Sci.,

1998, 8, 229.

25 D. Rigby, H. Sun and B. E. Eichinger, Polym. Int., 1998, 44,

311.

1 U. Beginn, L. Yan, S. N. Chvalun, M. A. Shcherbina,

¨

A. V. Bakirov and M. Moller, Liq. Cryst., 2008, 35, 1073.

2 U. Beginn, Prog. Polym. Sci., 2003, 28, 1049.

3 V. Percec, C.-H. Ahn, T. K. Bera, G. Ungar and

D. J. P. Yeardley, Chem. – Eur. J., 1999, 16, 1070.

4 U. Beginn and G. Lattermann, Mol. Cryst. Liq. Cryst., 1994,

241, 215.

´

26 A. K. Rappe, C. J. Casewit, K. S. Colwell, W. A. Goddard and

W. M. Skiﬀ, J. Am. Chem. Soc., 1992, 114, 10024.

´

27 C. J. Casewit, K. S. Colwell and A. K. Rappe, J. Am. Chem. Soc.,

5 G. Staufer, M. Schellhorn and G. Lattermann, Liq. Cryst.,

1995, 18, 519.

1992, 114, 10035.

´

¨

28 A. K. Rappe, K. S. Colwell and C. J. Casewit, Inorg. Chem.,

6 K. Borisch, C. Tschierske, P. Goring and S. Diele, Chem.

1993, 32, 3438.

Commun., 1998, 2711.

¨

29 M. W. Matsen and F. S. Bates, Macromolecules, 1996, 29,

7641.

7 K. Borisch, S. Diele, P. Goring, H. Kresse and C. Tschierske, J.

Mater. Chem., 1998, 8, 529.

30 L. Leibler, Macromolecules, 1980, 13, 1602.

31 E. Fontes, P. Heiney, M. Ohba, J. N. Haseltine and

A. B. Smith, Phys. Rev. A, 1988, 37, 1329.

32 T. Sauer, Macromolecules, 1993, 26, 2057.

33 S. N. Chvalun, M. A. Shcherbina, I. V. Bykova, J. Blackwell

and V. Percec, Polym. Sci., Ser. A, 2002, 45, 1281.

34 S. N. Chvalun, M. A. Shcherbina, A. N. Yakunin, J. Blackwell

and V. Percec, Polym. Sci., Ser. A, 2007, 49, 266.

35 M. A. Shcherbina, A. V. Bakirov, A. N. Yakunin, V. Percec,

8 L. D. Landau and E. M. Lifshitz, Statistical Physics, (Fizmatlit,

Moscow, 1995; Pergamon, Oxford, 1980), parts 1, 4.

9 P. A. Smith, C. D. Nordquist, T. N. Jackson, et al., Appl. Phys.

Lett., 2000, 77, 1399.

10 T. Thurn-Albrecht, J. DeRouchey, T. P. Russell and

H. M. Jaeger, Macromolecules, 2000, 33(9), 3250.

11 X. L. Xu, S. A. Majetich and S. A. Asher, J. Am. Chem. Soc.,

2002, 124(46), 13864.

12 A. V. Bakirov, A. N. Yakunin, M. A. Shcherbina,

¨

U. Beginn, M. Moller and S. N. Chvalun, Crystallogr. Rep.,

S. N. Chvalun, X. Zhu, U. Beginn and M. Moller,

2012, 57(2), 151.

Nanotechnologies in Russia, 2010, 5(9–10), 41.

¨

36 S. N. Chvalun, J. Blackwell and Y.-K. Kwon, Macromol. Symp.,

1997, 118, 663.

13 X. Zhu, B. Tartsch, U. Beginn and M. Moller, Chem. – Eur. J.,

2004, 10, 3871.

37 S. N. Chvalun, J. Blackwell and J. D. Cho, Polymer, 1998,

39(19), 4515.

38 S. N. Chvalun, M. A. Shcherbina, I. V. Bykova, J. Blackwell,

V. Percec, Y. K. Kwon and J.-D. Cho, Polym. Sci., Ser. A,

2001, 43, 33.

14 V. Percec, M. Glodde, T. K. Bera, Y. Miura, I. Shiyanovskaya,

K. D. Singer, V. S. K. Balagurusamy, P. A. Heiney, I. Schnell,

A. Rapp, et al., Nature, 2002, 419, 384.

15 J. N. Izraelachvili, Intermolecular and Surface Forces, Acad.

Press, New York, 1992.

Soft Matter, 2014, 10, 1746–1757 | 1757

Article Doi

The effect of the shape of the mesogenic group on the structure and phase behavior of 2,3,4-tris(dodecyloxy)benzenesulfonates with alkaline cations

DOI: 10.1039/c3sm51821c

Source and publish data:

Authors:

Article abstract of DOI:10.1039/c3sm51821c

Full text of DOI:10.1039/c3sm51821c

Products guided by the article

R&D Labs maybe for 878791-45-8

Relevant to this article

Hot Product