Spectroscopic properties and gelling ability of a set of rod-like 2,3-disubstituted anthracenes

Article abstract of DOI:10.1039/b306961c

The UV absorption spectra of 2,3-didecyloxyanthracene (DDOA) recorded in methanol, ethanol, 1-propanol, acetonitrile and methylcyclohexane reveal interesting features: they show a striking contrast between the isotropic solution and the gel state where, at ambient temperature, a fine structure appears with an apparent bathochromic shift (Δυ ≈ -700 cm ^-1), as observed in the solid state. Such an effect was mimicked by 2,3-dioxydi- (and -tri-) methyleneanthracenes in which the conformational mobility of the two juxtanuclear oxygen atoms is reduced in a manner similar to that assumed in the gel state. The fluorescence emission of DDOA (10 ^-5 M) at very low temperature exhibits a loss of fine structure and a bathochromic shift, for the gel state, in agreement with the presence of aggregates; the excitation spectra were found to be superimposable upon the absorption spectra of the isotropic and gel phases, respectively. Solvent screening for DDOA gelling ability has shown that the most efficient solvents are CH₃OH and CH₃CN. From the phase transition diagram (temperatures of gel setting, T_gel, and gel melting, T_m, versus concentration), thermodynamic parameters were derived: ΔH _gel0/kJ mol^-1 = -70 (CH₃OH), -66 (CH ₃CN); ΔS_gel⁰/(J K^-1 mol ^-1 = -147 (CH₃OH), -140 (CH₃CN) and ΔG_gel⁰/kJ mol^-1 (at 300 K) ≈ -26 (CH₃OH), -24 (CH₃CN). These parameters attest to the good stability of these gel systems. Finally, the influence of the chain length (n = C₇H₁₅ to C₁₂H₂₅) on the efficiency of gel formation (or melting) was investigated in methanol, ethanol, acetonitrile and heptane. It emerges that, to form gels, n-decyl and n-undecyl were found to be the most suitable chains and methanol and ethanol the most efficient solvents. It should be noted that the ability to form gels in methanol at a concentration of 0.6 mM at ambient temperature qualifies DDOA as a supergelator.

Full text of DOI:10.1039/b306961c

View Article Online / Journal Homepage / Table of Contents for this issue
P A P E R
Spectroscopic properties and gelling ability of a set of rod-like
2,3-disubstituted anthracenesy
´ ´
Jean-Pierre Desvergne,* Thierry Brotin, Danny Meerschaut, Gilles Clavier, Frederic Placin,
Jean-Luc Pozzo and Henri Bouas-Laurent*
´
Laboratoire de Chimie Organique et Organometallique (CNRS UMR 5802),
Universite Bordeaux 1, F-33 405, Talence cedex, France.
´
E-mail: jp.desvergne@lcoo.u-bordeaux1.fr
Received (in Montpellier, France) 17th June 2003, Accepted 6th October 2003
First published as an Advance Article on the web 6th January 2004
The UV absorption spectra of 2,3-didecyloxyanthracene (DDOA) recorded in methanol, ethanol, 1-propanol,
acetonitrile and methylcyclohexane reveal interesting features: they show a striking contrast between the
isotropic solution and the gel state where, at ambient temperature, a fine structure appears with an apparent
bathochromic shift (Dn ꢀ ꢁ700 cmꢁ1), as observed in the solid state. Such an effect was mimicked by
2,3-dioxydi- (and -tri-) methyleneanthracenes in which the conformational mobility of the two juxtanuclear
oxygen atoms is reduced in a manner similar to that assumed in the gel state. The fluorescence emission of
DDOA (10ꢁ5 M) at very low temperature exhibits a loss of fine structure and a bathochromic shift, for the gel
state, in agreement with the presence of aggregates; the excitation spectra were found to be superimposable
upon the absorption spectra of the isotropic and gel phases, respectively. Solvent screening for DDOA gelling
ability has shown that the most efficient solvents are CH3OH and CH3CN. From the phase transition diagram
(temperatures of gel setting, Tgel , and gel melting, Tm , versus concentration), thermodynamic parameters were
derived: DH0gel/kJ molꢁ1 ¼ ꢁ70 (CH3OH), ꢁ66 (CH3CN); DSg0el/(J Kꢁ1 molꢁ1 ¼ ꢁ147 (CH3OH), ꢁ140
(CH3CN) and DG0gel/kJ molꢁ1 (at 300 K) ꢀ ꢁ26 (CH3OH), ꢁ24 (CH3CN). These parameters attest to the
good stability of these gel systems. Finally, the influence of the chain length (n ¼ C7H15 to C12H25) on the
efficiency of gel formation (or melting) was investigated in methanol, ethanol, acetonitrile and heptane. It
emerges that, to form gels, n-decyl and n-undecyl were found to be the most suitable chains and methanol and
ethanol the most efficient solvents. It should be noted that the ability to form gels in methanol at a
concentration of 0.6 mM at ambient temperature qualifies DDOA as a supergelator.
Introduction
Gels are heterogeneous soft materials constituted of three
dimensional (3D) networks imprisoning a liquid.1–4 They are
used in everyday life and continuous research efforts are being
directed to extend their applications.3–8
For the last two decades, there has been a growing interest in
gelation induced by the addition of low molar mass (M ꢂ 1000)
substrates (the so-called LMOGs9,27) to organic liquids or,
Scheme 1 Cholesteryl-4-(2-anthryloxy)butanoate (CAB), an example
more rarely, water.9–29 These compounds show a wide struc-
tural diversity. For convenience they can be classified into
two categories: (1) gelators involving hydrogen bonds, ion-pairs
or metal coordination and (2) systems involving no H bonds and
based on van der Waals and weak electrostatic interactions. In
the latter category, a comprehensive study has been devoted to
the family of gelators containing steroidal and condensed aro-
matic rings, denoted ALS (aromatic linked steroids;19 an exam-
ple is shown in Scheme 1), whose discovery in 1987 gave a
strong impetus to the development of the field.19–49
of the ALS family (see text); the presence of an anthracene containing
tether was found to induce gelling properties and luminescence for a
variety of organic liquids.19
in the 2 and 3 positions was found to be compelling for gela-
tion to occur (Scheme 2). Furthermore, a rheological study29
has shown that DDOA gels formed in 1-octanol could be
described as soft solids exhibiting high yield stress values.
The elasticity of the DDOA gel (G0 ꢀ 104 Pa at 1 wt %) was
analysed as resulting from a 3D network of fibers or bundles of
fibers with extended interaction zones, as borne out by small
angle neutron scattering (SANS) investigations.29 Various
aspects of DDOA gel properties have been recently examined:
aerogel formation in supercrital CO2 ,30 electrochemical prop-
erties in propylene carbonate,31 porosity templating of sol-gel
A few years later28 some of us reported the gelling properties
of 2,3-di-n-decyloxyanthracene (DDOA) for several organic
solvents, especially methanol. Similarly to CAB (Scheme 1),
it so happens that DDOA contains an anthracene nucleus that
confers fluorescing properties onto the gels, but for DDOA, in
contrast to the ALS family, the presence of alkoxy substituents
derived silica32a and alumina,32b gelation of liquid crystals.33
However, since the preliminary results,28 no detailed spec-
troscopic study has been published. Such a study is reported
here, together with a systematic exploration of the role played
by the chain length (R ¼ CnH2n þ 1) and an investigation of the
y Electronic supplementary information (ESI) available: UV absorp-
tion spectra of DDOA in 1-propanol and acetonitrile as a function
of temperature; phase transition diagrams and related plots for DDOA
in various solvents. See http://www.rsc.org/suppdata/nj/b3/
b306961c/
T h i s j o u r n a l i s Q T h e R o y a l S o c i e t y o f C h e m i s t r y a n d t h e
C e n t r e N a t i o n a l d e l a R e c h e r c h e S c i e n t i f i q u e 2 0 0 4
234
N e w . J . C h e m . , 2 0 0 4 , 2 8 , 2 3 4 – 2 4 3

View Article Online
Scheme 2 2,3-Di-n-alkoxyanthracenes (DAOA) 1 and related hetero-
cyclic compounds 2. The compounds will be denoted 1C7 , 1C8 , etc.
and 21 , 22 , 23 , respectively. The di-n-decyloxy derivative (DDOA)
was first shown to form a gel with alcohols and alkanes.28
thermodynamic parameters of the phase transitions (sol-to-gel
and gel-to-sol).
Fig. 2 Fluorescence and excitation spectra (arbitrary scale) of
DDOA in degassed CH3OH (10ꢁ4 M). Fluorescence lexc : 370 nm.
Excitation: lem : 435 nm. 25 ꢃC isotropic (Á Á Á) and ꢁ70 ꢃC gel (—)
phases, respectively. Insert: Fluorescence spectrum of 2,3-dimeth-
oxyanthracene in degassed ethanol (ꢀ3 Â 10ꢁ5 M) at ꢁ65 ꢃC (lexc.
:
Results and discussion
365 nm), for comparison.
1. UV absorption and fluorescence spectroscopic properties
of DDOA
of 2,3-dimethoxyanthracene taken as reference and its excita-
tion spectrum is in line with that of Fig. 1. When the tempera-
ture is lowered, the gel forms and the fluorescence spectra are
red-shifted, as usually seen for aggregates52,53 and consistently
the excitation spectrum becomes that of the gel as in Fig. 1.
UV spectra of DDOA recorded in ethanol and 1-propanol
vs. temperature [ꢁ15 ꢃC (gel phase) to 20–50 ꢃC (isotropic
solution)] were found to show the same features as in metha-
nol. An example of the spectra (300–400 nm) in 1-propanol
(ꢁ15ꢃ to 26 ꢃC) exhibiting clear isosbestic points is given as
supplementary material (Fig. S1 in the ESI).
In methylcyclohexane (MCH), the fluorescence was studied
from þ20 ꢃC to ꢁ104 ꢃC. The fluorescence and excitation spec-
tra of DDOA in degassed MCH at very low concentration
(10ꢁ5 M) are represented in Fig. 3. The excitation spectrum
of the gel shows a remarkably fine structure. The maximum
wavelengths do not differ from those in CH3OH. The emission
spectrum of the gel is composed of ‘‘free’’ molecules dissolved
in the imprisoned solvent and molecular aggregates immobi-
lized within the fibrils, filaments and fibers, which experience
different environments and therefore exhibit a variety of spec-
tra as shown in the time-resolved spectra of Fig. 3 (insert). This
is borne out by the results of a transient kinetic analysis,
reported in Table 1. The wavelength dependent fluorescence
decay can be fitted with a sum of two exponentials at all the
wavelengths investigated. These decay times are all longer than
that of the isolated molecule (ffi 4 ns) and are in agreement
with the occurrence of different excited species displaying
slightly different emission spectra, as recently discussed by
De Schryver and coworkers52,53 for anthracene chromophores
The gelling property of DDOA was discovered through an
unsuccessful crystallization in CH3OH.28 Subsequent rapid
screening of gel-forming solvents showed that alcohol and
alkanes are the best-suited. For this reason, the electronic
absorption and emission spectra were studied essentially in
methanol and methylcyclohexane (MCH); the latter was cho-
sen for its ability to stay liquid at very low temperatures
(m.p. ꢁ126 ꢃC).
The absorption spectra of DDOA in CH3OH are repre-
sented in Fig. 1. At 26 ꢃC, the fluid isotropic solution exhibits
the typical spectrum of 2,3-dialkoxy substituted anthracenes50
but, at ꢁ15 ꢃC, the medium is a translucent gel at the concen-
tration used (10ꢁ3M) and the spectrum appears to be strongly
different. The first absorption band (300–400 nm), essentially
constituted of the 1La electronic transition, has a nice fine
structure with well-defined vibronic maxima at 338, 347, 365
and 385 nm. The latter band, the most intense, shows an
apparent bathochromic shift (Dn ffi ꢁ700 cmꢁ1); it probably
corresponds to the 0-0 transition. This feature, combined with
the enhanced fine structure as is the case for low temperature
spectra, points to a great loss of freedom of orientation for
the molecules in the gel. This statement is reinforced by the
examination of the second band (200–300 nm), which, peaking
as expected at 260 nm for the solution, shows a strong decrease
of intensity in the gel; this change points to some parallelism
between the long axes of the molecules due to p-p stacking.51
Typical spectra of the right angle fluorescence of a diluted
solution of DDOA in methanol as a function of temperature
are represented in Fig. 2. At 20 ꢃC the spectrum resembles that
Fig. 3 Excitation (left) and fluorescence (right) spectra of DDOA
(10ꢁ5 M) in degassed methylcyclohexane: solution (Á Á Á) and gel (—).
lexc 370 nm; lobs 410 nm (þ20 ꢃC) and 480 nm (ꢁ100 ꢃC). Insert:
time-resolved fluorescence spectra in MCH: þ20 ꢃC (Á Á Á) 0 ns delay;
ꢁ104 ꢃC (---) 0 ns delay, ꢁ104 ꢃC (—) 40 ns delay. The population
of emitting aggregates changes with time.
Fig. 1 Isotropic (Á Á Á) and gel (—) phase absorption spectra of DDOA
(ffi10ꢁ3 M) in CH3OH, at 26 ꢃC and ꢁ15 ꢃC, respectively.
T h i s j o u r n a l i s Q T h e R o y a l S o c i e t y o f C h e m i s t r y a n d t h e
C e n t r e N a t i o n a l d e l a R e c h e r c h e S c i e n t i f i q u e 2 0 0 4
N e w . J . C h e m . , 2 0 0 4 , 2 8 , 2 3 4 – 2 4 3
235

View Article Online
Table 1 Fluorescence decay of DDOA (ffi8 Â 10ꢁ5 M) in degassed
methanol at ꢁ45 ꢃC. t1 , t2 in ns; A1 and A2 are the preexponential
factors of the fluorescence intensity decay: I(t) ¼ A1exp(ꢁt/t1) þ
A2exp(ꢁt/t2)
lobs/nm
420
440
460
480
t1 (A1)
t2 (A2)
w2
5 (0.4)
14 (0.14)
1.44
8 (0.4)
20 (0.07)
1.09
9.5 (0.4)
25 (0.05)
1.05
11.5 (0.4)
37 (0.04)
1.32
in Langmuir–Blodgett films. The non-structured fluorescence
spectra showing maximum intensity at 460 nm suggest the for-
mation of excited aggregates from overlapping monomers as
displayed by some non-sandwich overlapping excimers.54
The mutual arrangements of the anthracene nuclei in the gel
are reminiscent of those in solid DDOA, as reflected in the
compared UV spectra (Fig. 4). One might speculate about
the correspondence between the crystal structure and the struc-
ture of aggregates in the gel as shown for instance by Terech,9b
Weiss et al.27a,b,41b and Dastidar et al.27c but no distinct X-ray
diffractograms could be obtained on pure DDOA gels and
we never could grow any single crystals of DDOA. In the series
the only single crystals suitable for X-ray determination were
obtained for di-n-hexyloxyanthracene (DHOA),55 but no X-ray
studies were performed on the gels of DHOA. The question
of the structure at the molecular level has been addressed in a
recent paper showing some unique features of the gel phase.56
The UV absorption spectrum of the gel phase does not
depend on the gelling solvent and it was used to follow-up
the gel formation of DDOA in a matrix of silica gel.32a Such
a gel could be formed from Si(OCH3)4 (TMOS), methanol,
water and a base (to obtain pH 12.7) by adding a methanolic
solution of DDOA (10ꢁ3M) and heating at 50 ꢃC for several
minutes, then allowing to cool down to ambient temperature.
Under these conditions, the inorganic gel was formed first, fol-
lowed by gradual formation of the organic gel. The inclusion
of the DDOA gel could be traced by recording the UV spectra
as a function of temperature (Fig. 5).
Fig. 5 UV absorption spectra of DDOA in a silica gel matrix in a 1
mm quartz cell as a function of temperature: (—) 25 ꢃC (organogel),
(---) 40 ꢃC (mixture of organogel and isotropic phase), (Á Á Á) 45 ꢃC
(the same mixture with a higher proportion of isotropic phase).
freely rotating oxygen atoms; this is strongly inhibited in the
methylenedioxy compound 21 , which results in an hypsochro-
mic shift (370 nm). In contrast, the propylenedioxy substituent
in 23 induces a bathochromic shift (382 nm); another red shift
(390 nm) is exhibited by the six-membered ring compound 22
in which the oxygen atoms appear to be maintained in the best
geometry for conjugation. From these data, it emerges that, in
the gel, the oxygen atoms should have a geometry intermediate
between those of 2,3-propylenedioxyanthracene (23) and
2,3-ethylenedioxyanthracene (22).
2. Photochemical reactivity of DDOA
Anthracenes are known to have photochromic properties57 but
our attempts to confer this property to the DDOA gels met
with failure. No clear change was noted after 2 h irradiation
of a methanolic gel in a Pyrex tube with a medium pressure
mercury lamp. The photoreactivity of DDOA was thus inves-
tigated in isotropic solution. Irradiation of a degassed benzene
solution (10ꢁ3 M) with a medium pressure mercury lamp,
using a Pyrex filter, was followed by UV spectroscopy. The
The UV absorption spectra of a solution of DDOA in
CH3CN (6 Â 10ꢁ4M) was recorded as a function of tempera-
ture (2.4 ꢃC to 50 ꢃC) and showed the same characteristic
changes (including isosbestic points) as observed in methanol
and 1-propanol (Fig. S2 in the ESI).
The difference in the vibronic structure between the isotropic
solutions (in methanol, MCH or acetonitrile) and the gel (and
the solid state) might be in line with the ability of oxygen lone
pairs to align parallel with the p electrons of the anthracene
nucleus. A qualitative piece of evidence to support this hypoth-
esis comes from a comparison of the UV absorption spectra of
the heterocyclic derivatives 21,2,3 with those of the acyclic com-
pounds 1CH3 and 1C10 (see Fig. 6). The acyclic compounds
exhibit lmax 374–376 nm, irrespective of the solvent and
the chain length. It thus reflects an average orientation of the
Fig. 6 UV absorption spectra, at ambient temperature, in MCH
(c ffi 10ꢁ46 M) of 2,3-methylenedioxyanthracene (21), 2,3-ethylenedi-
oxyanthracene (22) and 2,3-trimethylenedioxyanthracene (23). The
onset wavelength increases from 370 to 382 to 390 nm for 21 , 23 and
22 , respectively.. Compare these UV spectra with those of Figs. 1–4.
Fig. 4 UV absorption spectra of DDOA in a KBr pellet (Á Á Á) and of
its gel (—) in methanol at ambient temperature. The diffused light is
more important in KBr.
T h i s j o u r n a l i s Q T h e R o y a l S o c i e t y o f C h e m i s t r y a n d t h e
C e n t r e N a t i o n a l d e l a R e c h e r c h e S c i e n t i f i q u e 2 0 0 4
236
N e w . J . C h e m . , 2 0 0 4 , 2 8 , 2 3 4 – 2 4 3

View Article Online
gels using DDOA gel as a template32 and as experienced with
300–400 nm absorption slowly disappeared while another
absorption band, at l < 300 nm tailing down to 450 nm, grew
in after 14 h. After allowing the solution to stand for 12 h, the
UV spectrum of DDOA was recovered. Attempts to isolate the
photoproduct(s) were unsuccessful. No definite products could
be obtained either from irradiation in the presence of oxygen.
As a result of their photoreactivity, it is advisable to keep the
gels in the dark and under inert gas. Under these conditions,
methanolic gels (ꢀ5 Â 10ꢁ3 M) were found to be stable for
more than one year.
other gelators58,59 (see Section 1). DDOA was found to be
very soluble in CH2Cl2 and CHCl3 at ambient temperature;
addition of methanol or ethanol to such solutions allows
formation of gels; however, this method was not explored
further.
4. Phase transition diagrams and thermodynamic parameters
In order to gain more insight into the efficiency of gel forma-
tion and the melting of DDOA, phase transition temperatures
were measured as a function of concentration in methanol,
ethanol, heptane, acetonitrile and propylene carbonate. The
temperatures [Tm (gel melting) and Tgel (gel formation)] were
measured with an accuracy of Æ0.5 ꢃC, using a thermostatted
bath (heating and cooling rates of 1.5 ꢃC minꢁ1) and an auto-
mated apparatus allowing the registration of transmittance of
the diffused light. (see Experimental). As the dimensions of the
fibers are of the same order of magnitude as the visible light
wavelengths, the diffused light intensity increases with their
concentration. These measurements seem to be less arbitrary
than those based on mechanical tests of gelification.
3. Solvent screening for gel-forming ability
DDOA was found to form gels in methanol at ambient tem-
perature at a concentration as low as 6 Â 10ꢁ4 M (0.04 wt
%). Most gels are known to be formed at a concentration of
ca. 1 wt %; the efficiency of DDOA in methanol is therefore
outstanding (comparable values for supergelators were
reported by others35e,46a) and systematic tests of gel-forming
solvent ability were undertaken using the inverted tube method
described in the Experimental. Some salient results are col-
lected in Table 2; the molar concentration was chosen with a
view to obtaining thermodynamic data (see Section 4) and as
a common basis for comparison between the different 2,3-
dialkoxyanthracenes. It is remarkable that, in many cases, gels
can be formed at concentrations as low as 10ꢁ3 M. One
observes that alcohols and alkanes are less easily gelified by
DDOA as the molecular weight increases. Among alcohols,
gel-forming ability decreases with branching (e.g. butane-2-ol
vs. 1-butanol); however, 2,3-butanediol behaves like ethanol.
Acetonitrile shows good performance but the solubility is limi-
ted to 10ꢁ2 M. For potential applications to electrochemistry,
propylene carbonate was found to deserve special study31 and
the methacrylates were tested for possible polymerization
using DDOA gel as a template.11,12 It is also worth noting that
supercritical CO2 (not reported in Table 2) was explored as a
solvent to produce aerogels,30 as mentioned in the
Introduction.
Phase transition diagrams. A typical phase transition dia-
gram, that of DDOA in methanol, is presented in Fig. 7,
accompanied by the thermodynamic correlation expressed in
eqns (1) and (2) below (see Fig. 8). In this solvent, the maxi-
mum temperature is limited by the boiling point of methanol
(64.5 ꢃC); it is thus useless to increase the DDOA concentra-
tion to increase Tm . In principle, the maximum gelification
temperature should not exceed the melting point of the gelator
(m.p. DDOA 87 ꢃC). One notes a clear hysteresis between the
melting and gelification processes; this phenomenon depends
on the temperature variation rate and points to the difficulty
in characterizing a reversible process dealing with partial
heterogeneity. The temperatures of gelification (Tgel) were
observed to vary very little between different experiments
whereas the melting temperatures (Tm) depend on the thermal
history of the sample.
DDOA is not soluble in water but addition of some water
to alcohols was found to be possible, as shown in silica sol
Gel formation can be assumed to resemble a crystalliza-
tion; thus, it will be described by the following thermodynamic
equations (see Appendix):
Table 2 Organic solvents tested for gelation of DDOA at low concen-
trationa by the inverted tube method:. (þ) denotes gel formation and
(ꢁ) no gel formation observed
Gel melting : lnf ¼ ꢁDH0=R Â 1=Tm þ DS0=R
ð1Þ
Gel formation : lnf ¼ þDH0=R Â 1=Tgel ꢁ DS0=R ð2Þ
Ambient
temperature
Solvent
Concentration/M
ꢁ18 ꢃC
DH 0 and DS 0 are positive for the melting and negative for the
gelification processes, respectively; f denotes the molar frac-
tion of the solute in the appropriate solvent (see Appendix).
The diagram for gelification of DDOA in n-heptane (Fig. 9)
clearly shows that the concentrations used are much higher
than in methanol (as qualitatively suggested in Section 3). It
is accompanied with the linear plot according to eqn. (2) (see
Fig. 10). Other diagrams (Fig. S3a–d) and linear graphs
Methanol
Ethanol
6 Â 10ꢁ4
1 Â 10ꢁ3
1.7 Â 10ꢁ3
1.2 Â 10ꢁ3
2 Â 10ꢁ3
1.3 Â 10ꢁ2
1.1 Â 10ꢁ3
10ꢁ2
þ
þ (8–10 ꢃC)
1-Propanol
2-Propanol
1-Butanol
þ (0 ꢃC)
–
þ
þ
þ (0 ꢃC)
2-Butanol
2,3-Butanediol
1-Octanol
–
þ
þ
þ
–
Propylene carbonate
n-Pentane
n-Hexane
10ꢁ2
2.8 Â 10ꢁ3
5 Â 10ꢁ3
10ꢁ2
þ
þ
þ
þ
þ
–
n-Heptane
Isooctane
Acetonitrileb
Acetone
DMF
10ꢁ2
7.5 Â 10ꢁ4
3 Â 10ꢁ3
1.1 Â 10ꢁ3
1.4 Â 10ꢁ3
10ꢁ2
þ
þ
–
DMSO
Methyl methacrylate
Lauryl methacrylate
þ
þ
þ
10ꢁ2
a
Concentration in mol (M) was adopted as it is used in thermo-
dynamics (see phase transition diagrams below). [DDOA] in CH3OH
(molar mass: 490): 5 Â 10ꢁ4 M ¼ 0.03 wt %; 10ꢁ3 M ¼ 0.06 wt %;
Fig. 7 Phase transition diagram of DDOA in methanol. Tm and Tgel
denote the gel-to-sol and the sol-to-gel transition temperatures, respec-
tively. A hysteresis with DT ꢀ 10 ꢃC was observed. T accuracy Æ0.5 ꢃC.
10ꢁ2 M ¼ 0.62 wt % Solubility limited to 10ꢁ2
M
b
T h i s j o u r n a l i s Q T h e R o y a l S o c i e t y o f C h e m i s t r y a n d t h e
C e n t r e N a t i o n a l d e l a R e c h e r c h e S c i e n t i f i q u e 2 0 0 4
N e w . J . C h e m . , 2 0 0 4 , 2 8 , 2 3 4 – 2 4 3
237

View Article Online
Fig. 10 lnf vs. 1/T plot for DDOA gelification in n-heptane
[eqn. (2)].
Fig. 8 lnf vs. 1/T plots for DDOA in methanol, showing a good
linearity (r2 ¼ 0.987). 1 represents the gel-to-sol transition and 2 the
sol-to-gel process. See eqns (1) and (2).
been selected for a variety of gelators and are listed below;
most of them (except for the cholesterol-based systems) are
of the H-bonded type. Shinkai and coworkers60 determined
the melting enthalpy for a number of derivatives (differing in
the alkoxy group) of para-alkoxy-substituted azobenzene-
appended cholesterol-based gelators (aC-3): 41 < DH0m/kJ
molꢁ1 < 160 in ethanol and 41 < DH0m/kJ molꢁ1 < 110 in
1-butanol. For N-benzoyloxycarbonyl-L-alanine-4-hexadeca-
noyl-2-nitrophenyl esters (BLAHN), Hanabusa and cowor-
kers61 reported DHg0el ¼ ꢁ23 and ꢁ25 kJ molꢁ1, DSg0el ¼ ꢁ7
and ꢁ10 J Kꢁ1 molꢁ1and DGg0el (at 300 K) ¼ ꢁ21 and ꢁ22
kJ molꢁ1, in methanol and cyclohexane, respectively. In addi-
tion, a more substituted derivative of the same family exhibited
the following parameters in cyclohexane: DH0gel ¼ ꢁ42 kJ
molꢁ1, DS0gel ¼ ꢁ56 J Kꢁ1 molꢁ1 and DG0gel (at 300 K) ¼
ꢁ26 kJ molꢁ1. Garner, Terech and coworkers15 have derived
the parameters for the trans isomer of 4-tert-butyl-1-phenyl-
cyclohexanol (BACOl) in several solvents: DH0m ¼ 35.6 and
28.5 kJ molꢁ1 and DS0m ¼ 75.5 and 66 J Kꢁ1 molꢁ1; hence the
free enthalpies at 300 K are DG0m ¼ 13.5 and 9.2 kJ molꢁ1, in
n-heptane and dichloromethane, respectively. Typical values
of gel formation enthalpies for hexose derivatives were derived
by Shinkai and coworkers:62 for a-1-O-methyl-4,6-O-benzyli-
dene-D-glucose DHg0el/kJ molꢁ1 ¼ ꢁ43 (toluene), ꢁ40 (tetra-
chloromethane) whereas for the D-galactose isomer DHg0el/kJ
molꢁ1 ¼ ꢁ40 (toluene), ꢁ45 (tetrachloromethane). Interest-
ingly, for some dithienylcyclopentenediamides in toluene
DH0m values of up to 98 kJ molꢁ1 were found, presumably
due to a strong H-bond network.63
(Fig. S4a–c) obtained in n-heptane, acetonitrile and propylene
carbonate are given as supplementary material.
From these data, the thermodynamic parameters DH0 and
DS0 were calculated and are listed in Table 3. Are also included
the values obtained from a study of 2,3-dialkoxyanthracenes
incorporating chains of different lengths (vide infra). From the
upper part of Table 3, it emerges that n-heptane seems to have
a lower ability to form gels with DDOA (the free enthalpy
at 27 ꢃC clearly has a lower value) than the other solvents,
presumably because of a better solvation power. On the other
hand, methanol appears to be the most appropriate solvent.
Influence of the chain length on the gelification efficient. After
discovering the gelling ability of DDOA and having shown
that it is connected to the presence of two alkoxy chains in
positions 2 and 3 of anthracene, an obvious question was that
of the optimum chain length. The whole series of derivatives
1(C6–C12) was prepared and preliminary tests show that 1C6
could not give gels at room temperature but provided single
crystals suitable for X-ray structure determination.55 The other
compounds formed gels in alcohols, though it is noted that
1C7 was clearly borderline and 1C12 looked less efficient in
ethanol. Phase transition diagrams were therefore established
for 1C8–1C11 in ethanol (see Figs. 11 and 12) and the thermo-
dynamic parameters were determined (Table 3). Despite some
discrepancies in the enthalpy and entropy values, the free
enthalpies show a trend of increasing stability (at 27 ꢃC) from
1C8 to 1C11 but the differences are rather small.
The enthalpy values reflect the network strength and the
importance of the entropy in the phase transition is clearly
Another appraisal of the chain length influence can be
obtained from the melting temperatures (although these are
less reliable than the gelification temperatures, they can be
easily determined by the inverted tube method) for a fixed con-
centration (10ꢁ2 M) in different solvents, as shown in Fig. 13.
In methanol, ethanol and acetonitrile, 1C10 and 1C11 behave as
the most efficient compounds if one uses the melting tempera-
ture as the main criterion.
Table 3 Thermodynamic parameters (DH0, DS0) obtained from the
phase transition diagrams (Tgel or Tm vs. concentration) and free
enthalpy (DG0) at 300 K in CH3OH, n-heptane, CH3CN and propylene
carbonate for DDOA. In EtOH, the parameters were determined for
compounds with different chain lengths. Accuracy on DH0 and DS0
is Æ10%
DH0/kJ molꢁ1
DS0/JKꢁ1 molꢁ1 DG0 (27 ꢃC)/kJ molꢁ1
The data of Table 3 are to be compared with the published
thermodynamic values of other LMOGs. Some of these have
Solvent and
compound
Gelif. Melting Gelif. Melting Gelif.
Melting
Methanol
1C10
ꢁ70
ꢁ60
ꢁ66
þ86
þ92
þ72
ꢁ147 þ188
ꢁ155 þ246
ꢁ140 þ152
ꢁ25.9
ꢁ13.5
ꢁ24.0
þ29.5
þ18.2
þ26.4
b-Heptane
1C10
Acetonitrile
1C10
Propylene
carbonate
1C10
ꢁ58
–
ꢁ115
–
ꢁ24.5
–
Ethanol
1C8
ꢁ52
ꢁ53
ꢁ64
ꢁ48
þ74
þ85
þ93
þ81
ꢁ103 þ165
ꢁ104 þ190
ꢁ139 þ218
ꢁ82 þ171
ꢁ21.1
ꢁ21.8
ꢁ22.3
ꢁ23.4
þ24.5
þ28.0
þ27.6
þ29.7
1C9
1C10
Fig. 9 Phase transition diagram for the gelification of DDOA in
n-heptane. Temperature accuracy: Æ0.5 ꢃC.
1C11
T h i s j o u r n a l i s Q T h e R o y a l S o c i e t y o f C h e m i s t r y a n d t h e
C e n t r e N a t i o n a l d e l a R e c h e r c h e S c i e n t i f i q u e 2 0 0 4
238
N e w . J . C h e m . , 2 0 0 4 , 2 8 , 2 3 4 – 2 4 3

View Article Online
Fig. 11 Phase transition diagram for the gelification of 1(SC8 , ˘C9 ,
LC10 , /C11) in EtOH. Temperature accuracy: Æ0.5 ꢃC.
Fig. 13 Gel melting temperatures (Tm) as a function of the chain
length in 1(C7–C12)
Lacetonitrile and /n-heptane.
at c ¼ 10ꢁ2
M
in Smethanol, ˘ethanol,
apparent. The data listed in Table 3 compare well with those
reported above and attest to a good stability of DDOA gels
in methanol, ethanol and acetonitrile.
were also observed in other members of the family with
R ¼ C7H15 to C12H25 , in ethanol, acetonitrile and heptane;
1C10 and 1C11 were found to be the most efficient gelators
and n-heptane the least appropriate solvent. A test of solvent
efficiency was obtained by plotting the phase transition tem-
perature diagram versus concentration; from these diagrams,
the thermodynamic parameters DH0m , DSg0el and DSm0 (gel
denotes gel formation and m is for melting) that have been
determined are in keeping with a good stability of these sys-
tems in alcohols and acetonitrile and a fair stability in n-
heptane. The present results confirm and extend previous work
on DDOA. Other recent studies dealing with infrared spectra
and polarized fluorescence spectra of aerogels help in getting
a deeper insight into the gel structure at the molecular level.56
Differential scanning calorimetry. This technique is generally
used to determine phase transition temperatures and enthal-
pies. Measurements effectuated on pure DDOA powder for
the melting and resolidification led to a m.p. of 87.7 ꢃC and
DH0m ¼ 42 kJ molꢁ1 while the solidification point is 87.1 ꢃC
with DH0solid ¼ 37 kJ molꢁ1. Because methanol has a relatively
low melting point, propylene carbonate (boiling point at 1 bar
of 240 ꢃC) was chosen to measure gel transition temperatures.
The thermograms are shown in Fig. 14. The transition enthal-
pies were determined to be DH0m ¼ 65 kJ molꢁ1 and
DH0gel ¼ ꢁ37 kJ molꢁ1. Although one finds a hysteresis gap
of about 10 ꢃC, as in the preceding measurements based on dif-
fused light, the gelification temperature (ꢀ 55 ꢃC) is different
from that found with the automated apparatus (ꢀ 60 ꢃC,
Fig. S3d in ESI). The discrepancy observed for the DHg0el
values in propylene carbonate (ꢁ58 kJ molꢁ1 reported in Table
3) and the above value (ꢁ37 kJ molꢁ1 obtained by DSC) is not
clearly understood. Moreover, we cannot offer a satisfactory
explanation for the fact that DH0m was found to be 42 kJ molꢁ1
(DDOA powder) and 65 kJ molꢁ1 (gel in propylene carbo-
nate). The DSC method should be applied to all samples under
various heating and cooling rates. These thermodynamic data
should be considered as preliminary.
Experimental
General methods and materials
Melting points were determined in capillary tubes on a Buchi
510 apparatus and are uncorrected. FTIR spectra were
recorded on a Perkin Elmer Paragon 1000 spectrophotometer.
1H NMR spectra at 250 MHz and 13C NMR spectra at 62.9
MHz were recorded on a Bruker 250 AC instrument for solu-
tions in CDCl3 unless stated otherwise. Chemical shifts are
given in ppm and J values are given in Hz. Chromatographic
separations were performed on SDS silica chromagel (70–210
mm). Mass spectra were obtained on an AutoSpeq EQ spectro-
meter. Elemental analyses were performed by the Microanaly-
tical Service, Institut du Pin, University Bordeaux I. Electronic
absorption spectra and transmission spectra were measured on
Summary and conclusions
In exploring further the spectroscopic properties of DDOA
(1C10), this study has revealed that the UV spectra of the gel
state exhibits a clear fine structure and an apparent bathochro-
mic shift (ꢀꢁ700 cmꢁ1) irrespective of the solvent used
(methanol, ethanol, 1-propanol, acetonitrile, methylcyclo-
hexane or n-heptane). It is suggested that these features are
induced by a reduced mobility of the conformation of the jux-
tanuclear oxygen atoms, as in 2,3-dioxydi- (and -tri-) methyl-
eneanthracenes. The remarkable gel-forꢁm4ing properties of
DDOA (in methanol, at 20 ꢃC, at 6 Â 10 M ¼ 0.04 wt %)
Fig. 14 Differential scanning calorimetry of a 10ꢁ2 M solution of
DDOA in propylene carbonate: heating (!) and cooling ( ) rates
Fig. 12 Phase transition diagram for the melting of 1(SC8 , ˘C9 ,
LC10 and /C11) in EtOH. Temperature accuracy: Æ0.5 ꢃC.
of 5 ꢃC minꢁ1
.
T h i s j o u r n a l i s Q T h e R o y a l S o c i e t y o f C h e m i s t r y a n d t h e
C e n t r e N a t i o n a l d e l a R e c h e r c h e S c i e n t i f i q u e 2 0 0 4
N e w . J . C h e m . , 2 0 0 4 , 2 8 , 2 3 4 – 2 4 3
239

View Article Online
an Hitachi U-3300 spectrophotometer equipped with a thermo-
controlled cell monitored by a Lauda thermostat. The mea-
surements were realized using cells having 10 mm optical
pathlength and a platinum temperature sensor. For differential
scanning calorimetry measurements, a Perkin Elmer DSC 7
equipped with a stainless steel capsule having a rubber O-ring
was used.
The temperature-dependent fluorescence emission and exci-
tation spectra were obtained on a Hitachi F-4500 fluorescence
spectrophotometer corrected for excitation and emission using
quartz cells with a 1 cm optical pathlength, which were placed
in a copper holder inside a quartz Dewar flushed with cold
nitrogen gas. The temperature was adjusted by the nitrogen
flow. The fluorescence decays were measured by single photon
counting(AppliedPhotophysics) asdescribedelsewhere.64 Spec-
troscopic grade solvents were used for spectrophotometric
measurements. No fluorescent contaminants were detected
upon excitation in the wavelength region of experimental inter-
est. The samples were degassed by freeze-pump-thaw cycles on
a high-vacuum line and sealed under vacuum.
Fig. 15 Typical light transmission diagram as a function of tempera-
ture (variation 1 ꢃC minꢁ1) as recorded by the self-monitoring set-up.31
following, the preparations of 1(C7 , C8 , C9 and C11) and 23
are detailed.
Solvents used for synthesis and gelation tests (alcohols,
alkanes, acetonitrile, acetone, DMF, DMSO, propylene carbo-
nate, methyl and laurylmethacrylate, etc.) were purchased
Compounds 3Cn : general procedure for the alkylation of
2,3-dihydroxyanthraquinone. In a three-neck round-bottom
flask equipped with a reflux condenser and a dropping funnel,
the required amount of 4 and potassium carbonate (2.5 molar
equiv. per hydroxy group) were added to freshly distilled DMF
(10 cm3 per mmol of 4); the medium was warmed until com-
plete dissolution then the appropriate alkylbromide (2.25
molar equiv. per hydroxy group) was added dropwise over
20 min. The reaction mixture was refluxed for 12 h before
being allowed to cool to room temperature. The solvent was
evaporated under reduced pressure and the residue was treated
with water and thoroughly extracted with dichloromethane.
The organic layers were combined, dried (MgSO4) and filtered;
the solvent was evaporated and the raw solid was chromato-
graphed on a column of silica gel after dissolution in one of
the following eluting mixtures of CH2Cl2–petroleum ether
(v/v): (A) 20:80; (B) 40:60. Average yields were 80–85%.
2,3-Diheptyloxy-9,10-anthraquinone (3C7). Starting from 1.5 g
of 4, compound 3C7 (2.4 g, 88%), m.p. 92 ꢃC (methanol), was
isolated with elution system B. Anal. calcd for C28 H36 O4 : C,
77.03; H, 8.31%; found: C, 76.92; H, 8.37%. nmax (KBr): 3076,
2956, 2924, 2854, 1669, 1577, 1514, 1466, 1377, 1332, 1312,
ˆ
from SDS (Peypin, Bouches du Rhone, France) and are of
synthesis grade. Solvents of spectroscopic grade (methanol,
ethanol, 1-propanol, acetonitrile, heptane, methylcyclohexane,
etc.) from Aldrich were used for UV and fluorescence studies.
Gelation tests
In a typical gelation experiment, 1 cm3 of solvent of special
grade was added to a weighed amount of gelator in a sep-
tum-capped test tube (ca. 4 cm length and 0.7 cm diameter)
in order to obtain the desired concentration. The mixture
was warmed to the boiling point until the solid was dissolved
and then was allowed to cool down to room temperature. In
those cases in which a gel was formed, which generally
occurred within several minutes and was confirmed by obser-
ving that the sample did not flow when the test tube was
inverted, the tube was weighed to quantify the amount of gela-
tinized solvent. When the solution was turbid but the solvent
was not entirely gelatinized, the mixture was warmed again
and then cooled in a freezer (ꢁ18 ꢃC) for 24 h.
1219, 1111, 1087, 713 and 621 cmꢁ1 1H NMR (CDCl3): d
.
8.16 (2H, m, H5 and H8); 7.65 (2H, m, H6 and H7); 7.55
(2H, s, H1 and H4); 4.11 (4H, t, J ¼ 6.7 Hz, OCH2); 1.85
(4H, m, OCH2CH2); 1.47 (4H, m, OCH2CH2CH2); 1.21
(12H, m); 0.87 (6H, t, J ¼ 6.7 Hz, CH3). 13C NMR: d 182.4
(C9 and C10); 153.7 (C2 and C3); 133.6 (C6 and C7); 133.5
(C8a and C10a); 128.0 (C4a and C9a); 126.8 (C5 and C8);
109.2 (C1 and C4); 69.3 (OCH2); 31.8 (CH2CH2CH3); 29.0
(t); 28.9 (t); 25.9 (t); 22.6 (CH2CH3); 14.1 (CH3).
Phase transition temperature measurements
The measure of temperatures was based on the variation of dif-
fused visible light intensity as a function of the temperature,
using an automated apparatus, as described elsewhere.31 The
monitoring light was that of a laser diode beam, at 670 nm,
a wavelength at which the samples do not absorb, and the
transmitted light was measured by a photodiode. The heat-
ing/cooling cycles were performed at a rate of 1 ꢃC minꢁ1
and repeat measurements (at least 3 per compound) indicate
that the phase transition temperatures [Tgel (sol ! gel) and Tm
(gel ! sol)] could be determined with an accuracy of Æ0.5 ꢃC.
A hysteresis was systematically observed (see Fig. 15 for a typi-
cal diagram), related to the existence of an enthalpy for the
sol-gel transition and its magnitude depends on the heating/
cooling rates.
2,3-Di-n-octyloxy-9,10-anthraquinone (3C8). Starting from 1.5
g of 4, compound 3C8 (2.45 g, 85%), m.p. 97 ꢃC (methanol),
Preparations
Compounds 1(C1 and C7–C12) and 21–3 , studied in this article,
have been synthesized from a common substrate: 2, 3-dihy-
droxyanthraquinone (4), as outlined in Scheme 3.
The preparations of 4 (m.p. 393–394 ꢃC, sublim.), 1C1 (m.p.
133 ꢃC), 1C10 (m.p. 84 ꢃC) and 1C12 (m.p. 74 ꢃC) were
described in a preceding publication.50 Compounds 21 (m.p.
250–252 ꢃC) and 22 (m.p. 204–205 ꢃC) are known.65 In the
Scheme 3 Synthetic pathways for the preparation of 1(C1 , C7–C12)
and 21–3 : (a) RBr, DMF, K2CO3 ; (b: i) NaBH4 , EtOH; (b: ii) HCl;
(c) Br(CH2)nBr, DMF, K2CO3 ; (d) Zn, NH4OH.
T h i s j o u r n a l i s Q T h e R o y a l S o c i e t y o f C h e m i s t r y a n d t h e
C e n t r e N a t i o n a l d e l a R e c h e r c h e S c i e n t i f i q u e 2 0 0 4
240
N e w . J . C h e m . , 2 0 0 4 , 2 8 , 2 3 4 – 2 4 3

View Article Online
was isolated with elution system B. nmax (KBr): 3076, 2921,
2851, 1671, 1576, 1515, 1465, 1377, 1332, 1315, 1220, 1111,
OCH2CH2CH2); 1.43 (m, CH2); 1.02 (t, J ¼ 6.4 Hz, CH3).
13C NMR: d 150.2 (C2 and C3); 130.9 (C8a and C10a);
128.9 (C4a and C9a); 127.7 (C5 and C8); 124.4 (C6 and C7);
123.8 (C9 and C10); 106.0 (C1 and C4); 68.8 (OCH2); 32.0
(t); 29.3 (t); 29.2 (t); 26.2 (CH2CH3); 14.2 (CH3).
1087 and 712 cmꢁ1 1H NMR (CDCl3): d 8.15 (2H, m, H5
.
and H8); 7.65 (2H, m, H6 and H7); 7.55 (2H, s, H1 and
H4); 4.11 (4H, t, J ¼ 6.7 Hz, OCH2); 1.85 (4H, m, OCH2CH2);
1.47 (4H, m, OCH2CH2CH2); 1.30–1.20 (16H, m, CH2); 0.86
(6H, t, J ¼ 6.7 Hz, CH3). 13C NMR: d 182.4 (C9 and C10);
153.7 (C2 and C3); 133.6 (C6 and C7); 133.5 (C8a and
C10a); 128.0 (C4a and C9a); 26.8 (C5 and C8); 109.2 (C1
and C4); 69.3 (OCH2); 31.8 (t); 29.3 (t); 29.0 (t); 25.9
(OCH2CH2CH2); 22.7 (OCH2CH3); 14.1 (CH3). FABþ-MS:
m/z (%) 464 (100), 353 (44), 240 (85).
2,3-Di-n-octyloxyanthracene (1C8). Starting from 0.81 g of
3C8 , 0.37 g (49%) of 1C8 , m.p. 105 ꢃC, was obtained. nmax
(KBr): 2952, 2930, 2855, 1627, 1568, 1493, 1465, 1387, 1291,
1261, 1230, 1193, 1166, 1006, 889, 832, 743, 595 and
474 cmꢁ1 1H NMR: d 8.22 (s, H9 and H10); 7.95 (m, H5
.
and H8); 7.43 (m, H6 and H7); 7.20 (s, H1 and H4);
4.17 (t, J ¼ 6.4Hz, OCH2); 1.97 (m, OCH2CH2); 1.58 (m,
OCH2CH2CH2); 1.39 (m, CH2); 0.96 (t, J ¼ 6.4 Hz, CH3).
13C NMR: d 150.1 (C2 and C3); 130.8 (C8a and C10a);
128.8 (C4a and C9a); 127.7 (C5 and C8); 124.4 (C6 and C7);
123.8 (C9 and C10); 106.0 (C1 and C4); 68.8 (OCH2); 31.9
(t); 29.5 (t); 29.4 (t); 29.2 (t); 26.2 (CH2CH3); 22.8 (t); 14.2
(CH3). HRMS: calcd for C30H42O2 : 434.3184; found:
434.3148.
2,3-Di-n-nonyloxy-9,10-anthraquinone (3C9). Starting from 1.0
g of 4, compound 3C9 (1.75 g, 85%), m.p. 103—104 ꢃC (metha-
nol), was isolated with elution system B. nmax (KBr): 3078,
2957, 2920, 2850, 1670, 1575, 1514, 1466, 1332, 1220, 1111,
1
712 et 621 cmꢁ1. H NMR (CDCl3): d 8.17 (2H, m, H5 and
H8); 7.66 (2H, m, H6 and H7); 7.59 (2H, s, H1 and H4);
4.13 (4H, t, J ¼ 6.7 Hz, OCH2); 1.82 (4H, m, OCH2CH2);
1.45 (4H, m, OCH2CH2CH2); 1.30–1.20 (20H, m, CH2); 0.83
(6H, t, J ¼ 6.7 Hz, CH3). 13C NMR: d 183.1 (C9 and C10);
153.5 (C2 and C3); 133.6 (C6 and C7); 129.5 (C8a and
C10a); 127.7 (C4a and C9a); 126.9 (C5 and C8); 109.3 (C1
and C4); 69.4 (OCH2); 31.9 (t); 29.6 (t); 29.4 (t); 29.3 (t);
28.9 (t); 25.9 (OCH2CH2CH2); 22.7 (OCH2CH2); 14.1 (CH3).
FABþ-MS: m/z (%) 492 (100), 367 (24), 240 (53).
2, 3-Di-n-nonyloxyanthracene (1C9). Starting from 3C9 (1.23 g),
0.64 g (55%) of 1C9 , m.p. 100 ꢃC, was isolated. Anal. calcd
for C32H46O2 : C, 83.12; H, 9.96%; found: C, 83.29; H, 9.94%.
nmax (KBr): 2953, 2919, 2851, 1629, 1568, 1530, 1491, 1481,
1466, 1384, 1287, 1223, 1165, 1005, 889, and 739 cmꢁ1 1H
.
NMR: d 8.24 (s, H9 and H10); 7.98 (m, H5 and H8); 7.47
(m, H6 and H7); 7.21 (s, H1 and H4); 4.17 (t, J ¼ 6.6Hz,
OCH2); 1.99 (m, OCH2CH2); 1.58 (m, OCH2CH2CH2); 1.25
(m, 20H, CH2); 0.98 (t, J ¼ 6.6 Hz, CH3). 13C NMR: d
150.0 (C2 and C3); 130.7 (C8a and C10a); 128.7 (C4a and
C9a); 127.6 (C5 and C8); 124.4 (C6 and C7); 123.7 (C9 and
C10); 105.9 (C1 and C4); 68.7 (OCH2); 32.0 (CH2CH2CH3);
29.7 (t); 29.5 (t); 29.4 (t); 29.1 (t, OCH2CH2); 26.2
(t, OCH2CH2CH2); 22.8 (CH2CH3); 14.2 (CH3).
2,3-Di-n-undecyloxy-9,10-anthraquinone (3C11). Starting from
1.0 g of 4, compound 3C11 (1.8 g, 79%), m.p. 97 ꢃC (hep-
tane–benzene), was obtained with elution system A. nmax
(KBr): 2955, 2922, 2849, 1669, 1576, 1515, 1466, 1378, 1329,
1221, 1113, 713 and 621 cmꢁ1 1H NMR (CDCl3): d 8.26
.
(2H, m, H5 and H8); 7.74 (2H, m, H6 and H7); 7.67 (2H, s,
H1 and H4); 4.18 (4H, t, J ¼ 6.9Hz, OCH2); 1.89 (4H, m,
OCH2CH2); 1.48 (4H, m, OCH2CH2CH2); 1.25 (28H, m,
CH2); 0.87 (3H, t, J ¼ 6.6 Hz, CH3). 13C NMR: d 182.7 (C9
and C10); 153.5 (C2 and C3); 133.7 (C6 and C7); 133.5 (C8a
and C10a); 128.0 (C4a and C9a); 126.9 (C5 and C8); 109.3
(C1 and C4); 69.4 (OCH2); 32.0 (t); 29.8 (t); 29.7 (t); 29.6
(t); 29.4 (t); 28.9 (t); 26.0 (t); 22.7 (CH2CH3); 14.2 (CH3).
FABþ-MS: m/z (%) 548 (100); 389 (37); 240 (59).
2, 3-Di-n-undecyloxyanthracene (1C11). Starting from 3C11
(0.73 g), 0.43 g (62%) of 1C11 , m.p. 102–103 ꢃC (pentane),
was obtained. nmax (KBr): 2917, 2849, 1630, 1586, 1491,
1481, 1466, 1384, 1287, 1222, 1164, 890 and 739 cmꢁ1 1H
.
NMR: d 8.25 (s, H9 and H10); 7.98 (m, H5 and H8); 7.45
(m, H6 and H7); 7.23 (s, H1 and H4); 4.20 (t, J ¼ 6.8Hz,
OCH2); 2.00 (m, OCH2CH2); 1.55 (m, OCH2CH2CH2); 1.37
(m, CH2); 0.98 (t, J ¼ 6.8 Hz, CH3). 13C NMR: d 150.1 (C2
and C3); 130.8 (C8a and C10a); 128.8 (C4a and C9a); 127.6
(C5 and C8); 124.4 (C6 and C7); 123.8 (C9 and C10); 106.0
(C1 and C4); 68.8 (OCH2); 32.0 (CH2CH2CH3); 29.7 (t); 29.5
(t); 29.4 (t); 29.1 (t); 26.2 (t); 22.8 (CH2CH3); 14.2 (CH3).
FABþ-MS: m/z (%) 518 (100); 329 (33); 307 (45); 289 (52);
210 (68). HRMS: calcd for C36H54O2 : 518.4123; found
518.4116.
Compounds 1Cn : general procedue for the reduction of
anthraquinones 3Cn. In a three-neck round-bottom flask
equipped with a reflux condenser, a drying column and a mag-
netic stirrer, the required amount of 2,3-dialkoxy-9,10-anthra-
quinone and 2-propanol (20 cm3 per mmol) were placed and a
large excess of sodium borohydride (ca. 20 molar equiv.) was
added in small portions at such a rate as to avoid a rapid tem-
perature rise. The medium was refluxed under nitrogen for 3 h
and allowed to cool to room temperature. The mixture was
then hydrolyzed with 35% hydrochloric acid in crushed ice
and extracted repeatedly with dichloromethane. The organic
layers were combined, washed with a sodium hydroxide solu-
tion (pH ꢀ 9) and water until neutral, dried (MgSO4) and fil-
tered. The solvent was evaporated and the solid residue
(which is the anthrone intermediate), which was not purified
further, was treated in the same way, using the same set-up
as above, except that the reflux was maintained for ca. 12 h.
The final residue was chromatographed on a column of silica
gel after dissolution in a solvent mixture of CH2Cl2–petroleum
ether (v/v 30:70). Average yields were 50–60%.
2,3-Trimethylenedioxyanthracene (23). 2,3-Dihydroxy-9,10-
anthraquinone (4; 3 g, 12.5 mmol),50 potassium carbonate (5
g, 36 mmol) and 100 cm3 DMF were placed in a three-necked
round-bottom flask equipped with a dropping funnel and a
reflux condenser, under a nitrogen atmosphere. A solution of
1,3-dibromopropane (3 g, 15 mmol) in 50 cm3 DMF was
added slowly through the dropping funnel to the mixture
heated at 100 ꢃC. After the addition, the medium temperature
was raised to 160 ꢃC and maintained for 17 h. Another aliquot
of the anthraquinone solution (0.5 g, 3.6 mmol, in 25 cm3
DMF) was then added and allowed to react for a further 2
h. After cooling, the reaction mixture was diluted with 300
cm3 water, which generated a yellowish precipitate that was fil-
tered off and dissolved in CH2Cl2 , washed with water and
dried on Na2SO4 . After filtration and evaporation of the
solvent, the pale brown raw material (53 , 2.4 g) was collected
to be used in the next step without further purification.
2,3-Di-n-heptyloxyanthracene (1C7). Starting from 3C7 (0.75
g), 0.39 g (56%) of 1C7 , m.p. 110 ꢃC, was isolated. Anal. calcd
for C28H38O2 : C, 82.76; H, 9.36%; found: C, 82.66; H, 9.47%.
nmax (KBr): 2951, 2929, 2855, 1627, 1568, 1493, 1465, 1387,
1291, 1261, 1228, 1193, 1166, 1007, 889, 832, 743, 595 and
474 cmꢁ1
.
and H8); 7.45 (m, H6 and H7); 7.22 (s, H1 and H4);
1H NMR: d 8.24 (s, H9 and H10); 7.96 (m, H5
In a round-bottom flask equipped with a reflux condenser, a
ground mixture of the crude anthraquinone derivative (53) and
zinc powder (5 g, 7.7 mmol) was added to a 20% ammonia
4.17 (t, J ¼ 6.4Hz, OCH2); 2.00 (m, OCH2CH2); 1.58 (m,
T h i s j o u r n a l i s Q T h e R o y a l S o c i e t y o f C h e m i s t r y a n d t h e
C e n t r e N a t i o n a l d e l a R e c h e r c h e S c i e n t i f i q u e 2 0 0 4
N e w . J . C h e m . , 2 0 0 4 , 2 8 , 2 3 4 – 2 4 3
241

View Article Online
K. te Nijenhuis, Thermoreversible Networks, Springer-Verlag,
Berlin, 1997.
(a) T. Tanaka, C. Wang, V. Pande, A. Y. Grosberg, A. English, S.
Masamune, H. Gold, R. Levy and K. King, Faraday Discuss.,
1996, 102, 201; (b) ‘‘Physikalische Chemie der Gele und
Netzwerke’’, guest eds. H. Hoffman, H. Rehage and W.
Oppermann, Ber. Bunsenges Phys. Chem., 1998, 102, 1523–1717.
(a) M. Jacoby, Chem. Eng. News, 1997, 75(43), 10; (b) R. Dagani,
Chem. Eng. News, 1997, 75(23), 26.
aqueous solution (100 cm3). The mixture, which rapidly turned
dark red, was stirred at 90 ꢃC until the coloration vanished (i.e.
for about 10 h). The mixture was filtered and the solid was
washed with boiling ethanol, dissolved in dichloromethane
and chromatographed on a column of silica gel (eluant:
CH2Cl2–petroleum ether, v/v 50:50). Compound 23 was iso-
lated as a pale yellow powder (1.1 g, overall yield 35%), m.p.
218 ꢃC. Anal. calcd for C17H14O2 : C, 81.60; H, 5.60; O,
12.80%; found: C, 82.22; H, 5.60; O, 12.73%. nmax (KBr):
3040, 2950, 2860, 1475, 1455, 1380, 1300, 1275, 1210, 1135,
6
7
8
9
(a) P. Terech and R. G. Weiss, Chem. Rev., 1997, 97, 3133; (b)
P. Terech, Ber. Bunsenges. Phys. Chem., 1998, 102, 1630.
10 W. Gu, L. Lu, G. B. Chapman and R. G. Weiss, Chem. Commun.,
1997, 543.
11 J. van Esch, S. de Feyter, R. M. Kellog, F. de Schryver and B. L.
Feringa, Chem.-Eur. J., 1997, 3, 1238.
1
1040, 935, 885 and 730 cmꢁ1. H NMR (CDCl3): d 2.28 (m,
2H); 4.33 (t, 4H); 7.0–8.26 (m, 8H). MS: m/z (%) 250 (Mþ,
100), 222 (25), 181 (28), 165 (23), 164 (25), 152 (17), 28 (31).
12 R. J. H. Hafkamp, B. P. A. Kokke, I. M. Danke, H. P. M. Geurts,
A. E. Rowan, M. C. Feiters and R. J. M. Nolte, Chem. Commun.,
1997, 545.
13 J. E. Sohna-Sohna and F. Fages, Chem. Commun., 1997, 327.
14 M. de Loos, J. van Esch, I. Stokroos, R. M. Kellogg and B. L.
Feringa, J. Am. Chem. Soc., 1997, 119, 12 675.
15 C. M. Garner, P. Terech, J. J. Allegraud, B. Mistrot, P. Nguyen,
A. de Geyer and D. Rivera, J. Chem. Soc., Faraday Trans., 1998,
94, 2173.
16 (a) S. Yamasaki and H. Tsutsumi, Bull. Chem. Soc. Jpn., 1996, 69,
561; (b) M. Watase and H. Itagaki, Bull. Chem. Soc. Jpn., 1998,
71, 1457; (c) M. Watase, Y. Nakatani and H. Itagaki, J. Phys.
Chem. B, 1999, 103, 2366.
Appendix: determination of equilibrium constants
between solution and gel phase from the phase
transition diagrams
At a given temperature, the equilibrium constant of the
reversible reaction: gel 0 liquid (solution) is expressed by
K ¼ activity (liquid)/activity (gel) ꢀ a/1.
The melting of a gel into the liquid phase is assumed to
resemble the melting of a solid66 whose activity is 1. As the
concentration of DDOA is low, the activity of the liquid phase
will be represented by that of its molar fraction denoted f;
therefore:
17 K. Inoue, Y. Ono, Y. Kanekiyo, T. Ishi-i, K. Yoshihara and S.
Shinkai, Tetrahedron Lett., 1998, 39, 2981.
18 (a) R. Oda, I. Huc and S. J. Candau, Angew. Chem., Int. Ed.,
1998, 37, 2689; (b) R. Oda, I. Huc, M. Schmutz, S. J. Candau
and F. C. MacIntosh, Nature (London), 1999, 399, 566.
19 Y.-C. Lin and R. G. Weiss, Macromolecules, 1987, 20, 414.
20 Y.-C. Lin, B. Kachar and R. G. Weiss, J. Am. Chem. Soc., 1989,
111, 5542.
21 Y.-C. Lin and R. G. Weiss, Liq. Cryst., 1989, 4, 367.
22 I. Furman and R. G. Weiss, Langmuir, 1993, 9, 2084.
23 R. Mukkamala and R. G. Weiss, J. Chem. Soc. Chem. Commun.,
1995, 375.
24 P. Terech, I. Furman and R. G. Weiss, J. Phys. Chem., 1995, 99,
9558.
25 P. Terech, E. Ostuni and R. G. Weiss, J. Phys. Chem., 1996, 100,
3759.
26 E. Ostuni, P. Kamaras and R. G. Weiss, Angew. Chem., Int. Ed.
Engl., 1996, 35, 1324.
27 (a) D. J. Abdallah and R. G. Weiss, Langmuir, 2000, 16, 352; (b)
D. J. Abdallah, S. A. Sirchio and R. G. Weiss, Langmuir, 2000,
16, 7558; (c) A. Ballabh, D. R. Trivedi and P. Dastidar, Chem.
Mater., 2003, 15, 2136.
lnK ¼ lnf
lnf ¼ ꢁDG0=RT
ðA1Þ
ðA2Þ
ðA3Þ
where
and
lnf ¼ ꢁDH0=R Â 1=Tm þ DS0=R
with Tm being the melting temperature.
Conversely, the reversible formation of a gel from the liquid
phase (similar to a crystallization) will be represented by
K ꢀ 1/a, which leads to:
lnf ¼ þDH0=R Â 1=Tgel ꢁ DS0=R
ðA4Þ
where Tgel is the gelification temperature. DH0 and DS0 (stan-
dard conditions) will be positive for the melting and negative
28 T. Brotin, R. Utermo¨hlen, F. Fages, H. Bouas-Laurent and J.-P.
Desvergne, J. Chem. Soc., Chem. Commun., 1991, 416.
29 (a) P. Terech, J.-P. Desvergne and H. Bouas-Laurent, J. Colloid
Interface Sci., 1995, 174, 258; (b) P. Terech, I. Furman, R. G.
Weiss, H. Bouas-Laurent, J.-P. Desvergne and R. Ramasseul,
Faraday Discuss., 1996, 101, 345.
for the gelification. The molar fraction f ¼ xsolute/(xsolute
þ
x
solvent). As xsolute << xsolvent, f ꢀ xsolute/xsolvent and therefore,
f ¼ (m/M)/(m0/M0) where m and m0 are the masses of solute
and solvent, respectively while M and M0 are the molar masses.
30 F. Placin, J.-P. Desvergne and F. Cansell, J. Mater. Chem., 2000,
10, 2147.
Acknowledgements
`
31 F. Placin, J.-P. Desvergne and J.-C. Lassegues, Chem. Mater.,
2001, 13, 117.
D. Meerschaut is grateful to the CNRS for a postdoctoral
grant and to Dr. R. Lesclaux and to University Bordeaux 1
for financial assistance. We thank Dr. M. Lamotte for
assistance in gated spectroscopy as well as F. Fages and R.
Utermo¨hlen for valuable discussions. Financial support from
32 (a) G. M. Clavier, J.-L. Pozzo, H. Bouas-Laurent, C. Liere, C.
Roux and C. Sanchez, J. Mater. Chem., 2000, 10, 1725; (b) M.
Llusar, L. Pidol, C. Roux, J.-L. Pozzo and C. Sanchez, Chem.
Mater., 2002, 14, 5124.
33 T. Kato, T. Kutsuna, K. Yabuuchi and N. Mizoshita, Langmuir,
2002, 18, 7086.
34 (a) K. Hanabusa, R. Tanaka, M. Suzuki, M. Kimura and S.
Shirai, Adv. Mater., 1997, 8, 740; (b) Y. Hishikawa, K. Sada,
R. Watanabe, M. Miyata and K. Hanabusa, Chem. Lett.,
1998, 795.
´
the Region Aquitaine is acknowledged with gratitude.
References
35 (a) S. Shinkai and K. Murata, J. Mater. Chem., 1998, 8, 485; (b)
K. Yoza, Y. Ono, K. Yoshihara, T. Akao, H. Shinmori, M.
Takeuchi, S. Shinkai and D. N. Reinhoudt, Chem. Commun.,
1998, 907; (c) O. Gronwald and S. Shinkai, Chem.-Eur. J., 2001,
7, 4328; (d ) H. Kobayashi, A. Friggeri, K. Koumoto, M. Amaike,
S. Shinkai and D. N. Reinhoudt, Org. Lett., 2002, 4, 1423 and
references therein; (e) O. Gronwald, E. Snip and S. Shinkai, Curr.
Opin. Colloid Interface Sci., 2002, 7, 148. The lowest concentra-
tions reported for methyl-4,6-O-para-nitrobenzylidene-a-D-galac-
topyranoside and methyl-4,6-O-a-D-mannopyranoside in alkanes
were found to be 0.9–2.5 mM.
1
Reversible Polymeric Gels and Related Systems, ed. P. S. Russo,
ACS Symposium Series 350, American Chemical Society,
Washington, D.C., 1987.
Physical Networks, Polymers and Gels, eds. W. Buchard and S. B.
Ross-Murphy, Elsevier Applied Science, London, 1990.
Food Polymers, Gels and Colloids, ed. E. Dickinson The Royal
Society of Chemistry, Cambridge, UK, 1991.
J. M. Guenet, Thermoreversible Gelation of Polymers and
Biopolymers, Academic Press, London, 1992.
J. H. Fuhrhop and W. Helfrich, Chem. Rev., 1993, 93, 1565.
2
3
4
5
T h i s j o u r n a l i s Q T h e R o y a l S o c i e t y o f C h e m i s t r y a n d t h e
C e n t r e N a t i o n a l d e l a R e c h e r c h e S c i e n t i f i q u e 2 0 0 4
242
N e w . J . C h e m . , 2 0 0 4 , 2 8 , 2 3 4 – 2 4 3

View Article Online
was first prepared by: V. K. Lagodzinski, Ann. Chem., 1905,
342, 90.
36 (a) U. Maitra, P. V. Kumar, N. Chandra, L. J. D’Souza, M. D.
Prasanna and A. R. Raju, Chem. Commun., 1999, 595; (b) U.
Maitra, P. V. Kumar, N. M. Sangeetha, P. Babu and A. R. Raju,
Tetrahedron: Asymmetry, 2001, 12, 477.
51 D. Marquis, J.-P. Desvergne and H. Bouas-Laurent, J. Org.
Chem., 1995, 60, 7984.
37 (a) J. van Esch, F. Schoonbeek, M. De Loos, H. Kooijman, A. L.
Spek, R. M. Kellogg and B. L. Feringa, Chem.-Eur. J., 1999, 5,
937; (b) J. van Esch, F. Schoonbeek, M. De Loos, H. Kooijman,
A. L. Spek, R. M. Kellogg and B. L. Feringa, Chem.-Eur. J., 1999,
5, 6445; (c) L. N. Lucas, J. van Esch, R. M. Kellogg and B. L.
Feringa, Chem. Commun., 2001, 759.
52 G. Biesmans, G. Verbeek, B. Verschuere, M. van der Auweraer
and F. C. de Schryver, Thin Solid Films, 1989, 169, 127.
53 M. van der Auweraer, G. Biesmans and F. C. de Schryver, Chem.
Phys., 1988, 119, 355.
54 J. Ferguson, A. Castellan, J.-P. Desvergne and H. Bouas-Laurent,
Chem. Phys. Lett., 1981, 78, 446.
38 G. Clavier, M. Mistry, F. Fages and J. L. Pozzo, Tetrahedron
Lett., 1999, 40, 9021.
39 H. Ihara, M. Yoshitake, M. Takafuji, T. Yamada, T. Sagawa, C.
Hirayama and H. Hachisako, Liq. Cryst., 1999, 26, 1021.
40 L. A. Cuccia, J. M. Lehn, J. C. Homo and M. Schmutz, Angew.
Chem., Int. Ed., 2000, 39, 233.
41 (a) L. D. Lu, T. M. Cocker, R. E. Bachman and R. G. Weiss,
Langmuir, 2000, 16, 20; (b) M. George and R. G. Weiss, J. Am.
Chem. Soc., 2001, 123, 10 393.
42 (a) R. Wang, C. Geiger, L. H. Chen, B. Swanson and D. G.
Whitten, J. Am. Chem. Soc., 2000, 122, 2399; (b) D. C. Duncan
and D. G. Whitten, Langmuir, 2000, 16, 2399.
55 J.-L. Pozzo, J.-P. Desvergne, G. M. Clavier, H. Bouas-Laurent,
P. G. Jones and J. Perlstein, J. Chem. Soc., Perkin Trans. 2,
2001, 824.
56 F. Placin, J.-P. Desvergne, C. Belin, T. Buffeteau, B. Desbats,
`
L. Ducasse and J. C. Lassegues, Langmuir, 2003, 19, 4563.
57 (a) H. Bouas-Laurent, A. Castellan, J.-P. Desvergne and R.
Lapouyade, Chem. Soc. Rev., 2000, 29, 43; (b) H. Bouas-Laurent,
A. Castellan, J.-P. Desvergne and R. Lapouyade, Chem. Soc.
Rev., 2001, 30, 248; (c) H. Bouas-Laurent and J.-P. Desvergne,
in Photochromism, Molecules and Systems (revised edition), eds.
H. Durr and H. Bouas-Laurent, Elsevier, Amsterdam, 2003,
¨
ch. 14, pp 561–630.
43 R. P. Lyon and W. M. Atkins, J. Am. Chem. Soc., 2001, 123,
4408.
58 Y. Ono, K. Nakashima, M. Sano, Y. Kanekiyo, K. Inoue, J. Hojo
and S. Shinkai, Chem. Commun., 1998, 1477.
44 A. Ajayagosh and S. J. George, J. Am. Chem. Soc., 2001, 123,
5148.
59 J. H. Jung, M. Amaike and S. Shinkai, Chem. Commun., 2000,
2343.
45 (a) J. Makarevic, M. Jokic, B. Peric, V. Tomsic, B. Kojic-Prodic
and M. Zinic, Chem.-Eur. J., 2001, 7, 3328; (b) L. Frkanec, M.
Jokic, J. Makarevic, K. Wolsperger andZinic, J. Am. Chem.
Soc., 2002, 124, 9716.
46 (a) F. M. Menger and K. L. Caran, J. Am. Chem. Soc., 2000, 122,
11 679; amino acid derivatives were found to rigidify mixtures of
H2O–DMSO at concentrations <0.3 mM; (b) M. Ko¨lbel and
F. M. Menger, Chem. Commun., 2001, 275; (c) M. Ko¨lbel and
F. M. Menger, Langmuir, 2001, 17, 4490.
47 Y. Y. Waguespack, S. Barnerjee, P. Ramannair, G. C. Irvin,
V. T. John and G. L. McPherson, Langmuir, 2000, 16, 3036.
48 A. J. Carr, R. Melendez, S. J. Geib and A Hamilton, Tetrahedron
Lett., 1998, 39, 7447.
49 C. Shi, Z. Huang, S. Kilic, J. Xu, R. M. Enick, E. J. Beckman,
A. J. Carr, R. E. Melendez and A. D. Hamilton, Science, 1999,
286, 1540.
60 K. Murata, M. Aoki, T. Suzuki, T. Harada, H. Kawabata, T.
Komori, F. Ohseto, K. Ueda and S. Shinkai, J. Am. Chem.
Soc., 1994, 116, 6664.
61 K. Hanabusa, K. Okui, K. Karaki, M. Kimura and H. Shirai,
J. Colloid Interface Sci., 1997, 195, 86.
62 K. Yoza, N. Amanokura, Y. Ono, T. Akao, H. Shinmori, M.
Takeuchi, S. Shinkai and D. N. Rheinhoudt, Chem.-Eur. J.,
1999, 5, 2722.
63 L. N. Lucas, PhD Thesis, Groningen, The Netherlands, 2001.
64 F. Lahmani, A. Zehnacker, J.-P. Desvergne, H. Bouas-Laurent,
M. Colome`s and A. Krueger, J. Photochem. Photobiol. A, 1998,
113, 203.
65 T. Brotin, J. Waluk, J.-P. Desvergne, H. Bouas-Laurent and J.
Michl, Photochem. Photobiol., 1992, 55, 335.
66 (a) P. W. Atkins, Physical Chemistry, Oxford University Press, 5th
edn. 1994, pp. 226–227; (b) J. E. Eldridge and J. D. Ferry, J. Phys.
Chem., 1954, 58, 992.
`
50 J.-L. Pozzo, G. M. Clavier, M. Colomes and H. Bouas-Laurent,
Tetrahedron, 1997, 53, 6377; (b) 2,3-Dimethoxyanthracene 1C1
T h i s j o u r n a l i s Q T h e R o y a l S o c i e t y o f C h e m i s t r y a n d t h e
C e n t r e N a t i o n a l d e l a R e c h e r c h e S c i e n t i f i q u e 2 0 0 4
N e w . J . C h e m . , 2 0 0 4 , 2 8 , 2 3 4 – 2 4 3
243