Insights into the Formation and Structures of Molecular Gels by Diimidazolium Salt Gelators in Ionic Liquids or “Normal” Solvents

Article abstract of DOI:10.1002/chem.201600670

Insights are provided into the properties of molecular gels formed by diimidazolium salts both in “normal” solvents and ionic liquids. These materials can be interesting for applications in green and sustainable chemistry in which ionic liquids play a sig

Full text of DOI:10.1002/chem.201600670

DOI: 10.1002/chem.201600670
Full Paper
&
Ionic Liquids
Insights into the Formation and Structures of Molecular Gels by
Diimidazolium Salt Gelators in Ionic Liquids or “Normal” Solvents
[a]
[a]
[a]
[b]
[b, c]
Carla Rizzo, Francesca D’Anna,* Renato Noto, Mohan Zhang, and Richard G. Weiss
Abstract: Insights are provided into the properties of molec-
ular gels formed by diimidazolium salts both in “normal” sol-
vents and ionic liquids. These materials can be interesting
for applications in green and sustainable chemistry in which
ionic liquids play a significant role, like catalysis and energy.
In particular, two positional isomers of a diimidazolium
cation have been examined with a wide range of anions for
their ability to form gel phases. In particular, di-, tri-, and tet-
ravalent anions bearing aliphatic or aromatic spacers were
paired with the divalent cations. The properties of the
organo- and ionogels formed have been analyzed by means
of several different techniques, including calorimetry, rheolo-
gy, resonance light scattering, UV/Vis absorption, polarizing
optical microscopy, and powder X-ray diffraction measure-
ments. The investigations performed enabled us to obtain
a wide range of conductive materials characterized by
a high thermal stability and a low corrosiveness of the gela-
tor (organogels) or of both gelator and solvent (ionogels).
The information gained should be useful in the broader
quest to identify and promote their applications.
Introduction
and this should represent a significant advantage from the cat-
alysis point of view. However, despite the above-mentioned
advantages, possible solvent volatility and matrix damages
during the process still remain serious issues, whenever
organo- or hydrogels are concerned.
The search for new materials that could be applied in green
and sustainable chemistry area is one of the main goals of cur-
rent chemistry research. In particular, important fields like cat-
alysis and energy are continuously demanding new systems,
more stable and easier to handle than the liquid ones so far
developed. Gel phases, and supramolecular gels in particularly
seem good candidates to this aim. They are formed by low
molecular weight gelators (LMWGs) that possess the ability to
immobilize solvents by forming a 3-dimensional (3D) net-
On the other hand, as far as applications in energy are con-
sidered, satisfactory conductivity is a requested feature. Again
the nature of the electrolyte and the solvent volatility are the
main current limitations. Organic salts and ionic liquids (ILs)
may represent a good combination in attempt to solve all the
above-mentioned problems. In this regard, organic salts have
recently gained the attention of researchers as LMWGs to form
supramolecular gels. Indeed, many salts can be easily synthe-
sized, and small modifications of their structures can result in
[
1]
work. However, the processes by which self-assembly occurs
[
2]
are very complex. In fact, the majority of supramolecular gel
types has been discovered serendipitously, even though some
research groups have employed comprehensive approaches
that include experimental and computational methods in an
attempt to devise a priori recipes for the design of gelators
[
4]
formation of materials with very different properties. In addi-
tion, organic salts are less corrosive than inorganic molten
salts, which represents a significant advantage whenever appli-
cations of the corresponding gel materials, such as dye-sensi-
tized solar cells (DSSCs), electrocatalysis/supercapacitors and
fuel cells are taken into account.
[3]
and self-assembled fibrillar networks (SAFINs).
Supramolecular gels combine the ordered structure and
confined environment of solids with the properties of liquids
On the other hand, since their first appearance in literature,
ILs have been widely used both in catalysis and energy pro-
[
a] Dr. C. Rizzo, Prof. F. D’Anna, Prof. R. Noto
Università degli Studi di Palermo, Dipartimento STEBICEF
Sezione di Chimica, Viale delle Scienze, Parco O’Orleans II
[
5]
duction and storage. Self-assembly with ILs leads to the for-
mation of ionogels that are soft materials having both high
9
0128 Palermo (Italy)
[
6]
thermal stability and conductivity. These properties, in addi-
tion to low vapor pressures and low flammability, are found in
E-mail: francesca.danna@unipa.it
[
b] M. Zhang, Prof. R. G. Weiss
[
7]
Department of Chemistry, Georgetown University
Washington, DC, 20057-1227 (USA)
the ILs alone. In particular, most of typical IL properties, such
as conductivity, are not diminished in the ionogels; in some
cases, they are even enhanced. The beneficial ability to immo-
bilize ILs in a SAFIN opens several possibilities for new applica-
[
c] Prof. R. G. Weiss
Institute for Soft Matter Synthesis and Metrology, Georgetown University
Washington, DC, 20057-1227 (USA)
[
8]
tions. For instance, the use of ionogels in DSSCs avoids leak-
Supporting information and the ORCID identification number(s) for the au-
thor(s) of this article can be found under http://dx.doi.org/10.1002/
chem.201600670.
[
9]
age problems, which occur when ILs are used alone.
Chem. Eur. J. 2016, 22, 11269 – 11282
11269
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Kimizuka first defined “ionogels” referring to polymers able
dazolium (p-C im) and 3,3’-di-n-dodecyl-1,1’-(1,3-phenylenedi-
methylene)diimidazolium (m-C im) in which the counterions
12
12
[10]
to aggregate IL solutions.
Nowadays, this term comprises
a wide variety of materials, including polymeric gels, hydride
materials containing inorganic species, and recently supra-
are aliphatic or aromatic carboxylate anions that differ in the
extension of their p-surfaces and the number of negative
charges (2,6-naphtalenedicarboxylate (2,6-NDC), 1,4-benzenedi-
carboxylate (1,4-BDC), trimesate (Trim), citrate (Cit), ethylendia-
minetetracetate (EDTA)) (Scheme 1). Previously, we demonstrat-
ed by DFT calculations that the substitution pattern of the di-
cations is a key factor in determining the efficiency of the gela-
tion process due to the different packing modes that the meta
and para isomers require, especially when paired with their
[
11]
molecular gels too.
Literature on gels formed by ILs has been recently reviewed
[
12]
by Marr et al. However, few examples have been reported so
far regarding the gelation of organic salts or ILs in IL solution.
In the latter case, of course the domains formed are character-
ized by the presence of one IL acting as the liquid phase and
the other one as the solid continuous phase. The materials ob-
tained should be able to fulfill the requests of green and sus-
tainable chemistry.
[
18]
anions. In this work, we analyze how the combination of
these two dications with di-, tri-, or tetra anions influences the
gelating ability of the organic salts with several ILs and with
conventional solvents. Indeed, the different stoichiometries be-
tween the dications and their anions (1:1, 3:2, or 2:1) can lead
to different packing arrangements of the salts. Different stoi-
chiometric dication to anion ratios in the salt composition
were expected to have important consequences on the favora-
bility of interactions leading to gel formation.
In particular, Hao et al. reported the formation of an aniso-
tropic, thermally-reversible ionogel with sodium laurate as the
[
13]
LMWG.
Zhou et al. developed an anticorrosive/antioxidant
supramolecular gel, in which the LMWG is an imidazolium-type
[14]
“
ionic liquid” bearing a benzotriazole group.
Furthermore,
Yan et al. have made new conductive materials for DSSCs that
involve ionogel formation by host–guest interactions between
[15]
organic salts and b-cyclodextrins. In the framework of our in-
Among the ILs, imidazolium ILs having anions that differ in
terest in studying properties and applications of organic
size, symmetry, and coordinating ability ([bmim][BF ], [bmim]
4
[
16]
salts, we recently obtained supramolecular gels in IL solution
from dicationic organic salts that exhibit a high structural
order degree. The materials obtained showed similar or higher
[PF ], [bmim][SbF ], [bmim][NTf ]), and a pyrrolidinum IL
6
6
2
([bmpyrr][NTf ]) have been examined as gelation solvents
2
(Scheme 1). Specifically, the thermal properties of the neat or-
ganic salts have been investigated using thermogravimetric
(TGA) and differential scanning calorimetry (DSC) analyses. The
thermal stabilities of the gels have been investigated using the
[17]
conductivities as compared to neat ILs. Obviously, the large
number of cation/anion combination claimed for ionic liquids
can be also reasserted for organic salts. Consequently, gelation
of organic salts in ILs can give a huge number of soft materials
having significantly different properties. This brings about the
need to have a deep knowledge of the properties of ionogel
phases to better identify their practical potentialities.
[
19]
falling-ball method and DSC measurements, whereas their
mechanical and thixotropic properties were examined by
rheology. Furthermore, the ability of the gels to self-repair was
analyzed by subjecting them to ultrasound and magnetic stir-
ring. Finally, the gelation processes have been investigated by
resonance light scattering (RLS) and UV/Vis and rheology
measurements, and the morphologies of the SAFINs have
In the present work, we study the properties and the gelat-
ing abilities of two isomeric diimidazolium organic cations,
namely 3,3’-di-n-dodecyl-1,1’-(1,4-phenylenedimethylene)diimi-
Scheme 1. Structures of diimidazolium salts used as gelators and ionic liquids used as the solvent components.
Chem. Eur. J. 2016, 22, 11269 – 11282 www.chemeurj.org ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
11270

Full Paper
been studied by polarizing optical microscopy (POM) and
powder X-ray diffraction (PXRD).
charge, T or Tm decreased in the order, [2,6-NDC]>[1,4-BDC]
g
and [Trim]>[Cit]. This trend suggests that aromaticity and the
larger p-surface extension of the anion, together with its sym-
metry and rigidity, impart higher structural order.
Results and Discussion
Thermal analysis of organic salts
Gelation tests
Before testing the gelating abilities of the salts, their thermal
stabilities were evaluated by TGA and DSC measurements (the
Supporting Information, Figures S1, S2 and Table S1). In partic-
The gelating ability of the salts was tested in water, conven-
tional organic solvents and ionic liquids. The resistance to flow
of the gels, such as those shown in Figure 1, was tested by the
ular, the decomposition temperatures (T ), melting tempera-
[19]
d
tube inversion test. On this basis, the white opaque gels
tures (T ), variations of enthalpy corresponding to the melting
m
were stable for almost three months at room temperature in
sealed vessels.
processes (DH ), and of glass transition temperatures (T ) were
m
g
determined at the maximum slopes of TGA derivative curves,
as the maximum of heat flows in the DSC thermograms, as the
integrated areas of the melting calorimetric peaks, and as the
onset points of the sigmoidal curves of heat flow in the DSC
thermograms, respectively.
In all cases, a single-step decomposition process was ob-
served and T ranges were from 2068C for [p-C im] [Cit] to
d
12
3
2
2
458C for [m-C im][2,6-NDC]. Although the thermal stability of
12
these organic salts is high, it is lower than those reported by
Armstrong et al. for some diimidazolium salts bearing aliphatic
[
20]
spacers.
Perusal of the data in Table S1 indicates that the thermal
properties are affected significantly by the nature of both the
isomeric cation and the anions. In general, para isomers show
Figure 1. Gel phases at 5 wt% formed by (from left to right) [p-C12im][1,4-
BDC]/glycerol, [p-C im] [EDTA]/glycerol, [p-C im][1,4-BDC]/[bmim][BF ], [p-
lower T values than those of the corresponding meta isomers.
d
12
2
12
4
C
3 2 4 3 2 4
12im] [Trim] /[bmim][BF ] and [p-C12im] [Cit] /[bmim]BF ].
Td decreases on going from bivalent to trivalent anions and, in
the latter case, it decreases on going from aromatic to aliphatic
ones ([Trim]>[Cit]). However, for aliphatic anions, [EDTA] stabil-
izes better the salts. Furthermore, in agreement with thermal
Organic solvents
[21]
stability of some dioctylimidazolium salts, the increase in p-
Tables 1 and S2 and S3 (the Supporting Information) summa-
rize the results from gelation tests performed on a range of
conventional organic solvents with different polarities and hy-
drogen-bond donor abilities. Table 1 also includes the tempera-
surface area of aromatic dianions increased T , as evidenced
d
from the comparison between [m-C im][1,4-BDC] and [m-
1
2
C im][2,6-NDC].
12
Also, DSC thermograms revealed polymorphism in the com-
[
22]
pounds. Indeed, more than one endotherm can be detected
during the first heating. The second and third heating cycles
are reproducible and indicate the stability of the salts. This be-
havior appears to depend on the solvent of crystallization and
on the time needed for a salt to return to its initial morph (if it
is more stable thermodynamically than the one from solidifica-
tion of the melt; see below). In some cases, the first heating
thermograms revealed a small and second endotherm. Howev-
er, no evidence for mesophases was found using polarized op-
tical microscopy. Consequently, we ascribe these transitions to
melting processes. With the exception of the DSC thermo-
grams of [p-C im][2,6-NDC], [p-C im] [EDTA] and [m-
[
a]
Table 1. Values of CGC, Tgel and DH
m
for organogels.
Solvents
Glycerol
Octanol
CGC
[b]
[c]
[d]
[g]
[b]
[c]
CGC
3.2
T
gel
T
gel
DH
m
T
gel
[
[
[
p-C12im][1,4-BDC]
p-C12im] [2,6-NDC]
37.2
31.9
SM
36.2
42.6
33.8
[
[
e]
f]
1.8
–
2.3
2.0
3.7
2.0
57.0
p-C12 im]
2
[EDTA]
SM
5.0
33.0
ns
[e]
41.4
3.6
[m-C12im] [1,4-BDC]
23.7
24.0
27.6
ns
12
12
2
[
m-C12im]
2
[EDTA]
27.3
C im] [EDTA], the second and third heating thermograms
12
2
[e]
29.3
0.5
showed only glass transitions.
[
a] Appearances: SM=soft material; ns=no detectable DSC endotherm
In agreement with what is known about salts with similar
on heating. [b] %, w/w gelator/solvent. [c] Tgel values (8C) were reproduci-
ble within 18C; Tgel values determined at the CGC by the falling-ball
method. [d] Tgel values (8C) were reproducible within 18C; Tgel determined
at 5 wt% of gelator by the falling ball method. [e] Tgel values (8C) deter-
[
18,23]
structures,
the higher T and T values of [p-C im] isomers
m g 12
than those of the corresponding [m-C im] isomers are consis-
1
2
tent with the latter having lower structural order. However,
there is no clear relationship between salt stoichiometry and
the temperatures of the transitions. For anions with the same
mined by DSC. [f] Tgel value (8C) for a glass transition determined by DSC.
À1
[
g] DH
m
(Jg ) determined by DSC.
Chem. Eur. J. 2016, 22, 11269 – 11282
www.chemeurj.org
11271
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
tures corresponding to the gel-sol transitions (T ) at both
quent cooling of the mixtures did not induce soft materials for-
gel
[
24]
5
wt% and the critical gelation concentrations (CGC). The salts
mation. According to a previous report,
we centrifuged
are soluble in polar, protic solvents and insoluble in polar
aprotic ones. Some of the salts were able to gelate glycerol
and 1-octanol, and some of the gels were classified as soft ma-
terials (SM) because they were stable to the tube inversion
test, but were unable to support a lead ball (see below).
(6000 rpm, 1”10 min; 6000 rpm, 4”40 min) the [p-
C im] [Cit] /[bmim][NTf ] gels to check for phase separation to
12
3
2
2
solid and liquid components. However, the components of the
gelatinous suspensions could not be separated. From these re-
sults and the non-gelating ability of [p-C im][NTf ] in [bmim]
12
2 2
Irrespective of the isomeric cation, the CGCs reported in
Table 1 decrease with increasing p-surface area of the anion
[NTf ] (gelation tests were performed both at 2.6 and 5 wt%),
2
it is possible to identify [p-C im] [Cit] as the gelator. Addition-
12
3
2
(
[1,4-BDC]>[2,6-NDC]) and changing from divalent aromatic
al support for this assertion was found in the very different
anions to tetravalent aliphatic anions ([2,6-NDC]>[EDTA]).
transitions observed from DSC thermograms for [p-C im]
12
As noted above, we also determined the thermal stabilities
[NTf ] , [p-C im] [Cit] and [bmim] [Cit] (the Supporting Infor-
2 2 12 3 2 3
(
Tgel) of gels at 5 wt% gelator concentrations using data from
mation, Figure S2) and the one for the [p-C im] [Cit] /[bmim]
12 3 2
both the falling-ball method and DSC measurements. Thermo-
grams corresponding to gel–sol transitions are reported in Fig-
ure S3 (the Supporting Information). They exhibit large
changes in the heat capacities associated with second-order
phase transitions. All measurements were made in triplicate to
verify thermoreversibility and reproducibility. However, the
second run was measured 24 h after the first because gel refor-
mation was slow. As expected, Tm values for neat gelators (the
Supporting Information, Table S1) and Tgel values for the corre-
sponding organogels (Table 1) are very different; the media
into which the melting solids dissolve are not comparable.
DSC data for neat gelators are consistent with the observa-
tion that the thermal stability of the gel phases is dependent
on the isomer of the cation employed to make a salt. For both
the [EDTA]- and [1,4-BDC]-derivatives, thermal stability was
higher for the para than the meta cations; differences in the
isomeric nature of the cation affect packing in the neat solids
as well as in the gel SAFINs. However, the thermal stability of
the gels appears to be enhanced by the increase in p-surface
area of the anions associated with the para isomers: the Tgel
values for the [1,4-BDC]- and [2,6-NDC]-salts are 42.6 and
[NTf ] ionogel (Figure S4).
2
+
CGC values for gels in [bmim ]-based ILs were lower for
[1,4-BDC] than for [2,6-NDC] and for [Cit] than for [Trim], irre-
spective of the isomeric cation (3.4% and 4.3%, 3.5% and 5%,
respectively). Although the number of examples is small, the
data indicate that the gelation process is favored by a decrease
in the aromatic character of the anion. Previously, similar be-
havior was observed for ionogel formation from some func-
[
17]
tionalized diimidazolium salts. In that case (and in this one),
the aromatic cation of the gelator preferably gelated ILs
having an aliphatic cation. This suggests the benefit of struc-
tural complementarity between the gelator and IL solvent
structures.
Data collected for [Trim], [Cit], and [EDTA]-based organic
salts allow the effect of the IL anion structure on gel phase for-
mation to be explored. Irrespective of the gelator considered,
the largest CGC values were always observed in [bmim][BF4],
the IL having the largest hydrogen-bond accepting ability,
b (b=0.376, 0.207, 0.243, 0.146 for [bmim][BF ], [bmim][PF ],
4
6
[
25]
[bmim][NTf ] and [bmim][SbF ] respectively).
2
6
Gelation by LMWGs is the result of solvent entrapment
within a 3D SAFIN. In our case, such a process should involve
interactions between ions of the gelator and the solvent. Con-
sequently, interactions between ions of the solvent and gelator
should be favored when the b values are lower.
5
7.08C, respectively.
Ionic liquids
Gelation tests were performed also using ILs as the solvents
Also, as mentioned, data for [p-C im] [Cit] and [p-
12
3
2
(
[
[
the Supporting Information, Table S4). With the exception of
C im] [EDTA] in [bmim][NTf ] and [bmpyrr][NTf ] suggest that
12 2 2 2
m-C im] [Cit] , all of the salt gelators were able to gelate
decreasing the aromatic character of an IL cation leads to
higher CGC values. In that regard, Tgel values at the CGC and
5 wt% gelator concentrations were collected (Table 2 and the
Supporting Information, Figure S4).
12
3
2
bmim][BF ] and the CGC values were higher than with con-
4
ventional organic solvents. Whereas para isomers gelated most
of ILs tested, meta isomers precipitated or formed gelatinous
precipitates.
Some of the ionogels were “soft materials”. Their Tgel values
ranged from 22.2 to 53.08C and, similarly to organogels, values
determined using DSC measurements were higher than those
from the falling-ball method. Furthermore, according to data
collected for organogels and the corresponding neat salts, Tgel
values were higher for ionogels formed by para isomers than
Gelation of ILs by salts, especially those with different
anions, raises an important question: What is the nature of the
gelator? Specifically, do the anions remain with their original
cations? To answer this question, systems in which two cations
and two anions (one set from the gelator and the other from
the IL) were mixed and their gelation characteristics were com-
pared with those from the ionogel [p-C im] [Cit] /[bmim][NTf ]
meta ones. In [bmim][BF ] as the solvent, only minor differen-
4
ces in thermal stability were detected as a function of the
anion nature with the meta and para isomers of the cation,
and Tgel increased slightly on going from bivalent to trivalent
anions.
1
2
3
2
2
(
(
[
CGC=2.6%) as well as salts derived from exchange processes,
[p-C im][NTf ] and [bmim] [Cit]). The mixtures of [p-C im]
12
2 2
3
12
NTf ] and [bmim] [Cit] were prepared at the same cation and
2
2
3
anion mole ratios as those in the ionogel (n[bmim+]/n[p-C12im2+]
=
Data for [Trim]- and [Cit]-based gelators show that Tgel de-
creases as: [bmim][BF ]>[bmim][PF ]>[bmim][SbF ]>[bmim]
5
06 and n[NTf2]/n[Cit] =72). In both cases, heating and subse-
4
6
6
Chem. Eur. J. 2016, 22, 11269 – 11282
www.chemeurj.org
11272
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[
a]
pally by the nature of gelator anion rather than the
isomer of the cation. Based on comparisons of the
images in Figures 2d and S5b,c, Figures 2 f and S5o,
Figures 5d,e, Figures 2h,i, and Figures S5n and S5 f,
the morphologies were the same when only the cat-
m
Table 2. Tgel and DH values for ionogels at 5 wt% gelator concentrations.
[
p-C12im]
[p-C12im]
[p-C12im]
[Trim]
3
[p-C12im]
[Cit]
3
[p-C12im]
2
[1,4-BDC]
[2,6-NDC]
2
2
[b]
[EDTA]
[b]
[e]
[b]
[e]
[b]
[e]
[e]
[b]
[e]
Ionic liquids
T
gel
DH
m
T
gel
DH
m
T
gel
DH
m
T
gel
DH
m
T
gel
DH
m
4
6
4
4
1.4
SM
67.1 2.1
42.4
49.5 5.2
49.5
85.4 1.7
39.9
50.3 2.2
42.8
56.8
65.0 0.7
42.0
45.3 2.9
89.0
ns
49.0
ns
47.0
44.6 1.7
2+
2+
[
[
[
[
[
bmim][PF
bmim][NTf
bmim][BF
bmim][SbF
bmpyrr][NTf
6
]
ionic part of the gelator, [m-C im ] or [p-C im ],
12 12
[
c]
[c]
[c]
4.3 1.9
1.7
differed.
2
]
[
c]
[c]
[c]
4.2 2.7
However, the morphology of the gels was very de-
pendent on the specific IL employed. Using [p-
C im] [Cit] as the gelator, the images indicate that
5
8
3.4
1.7 1.9
SM
84.7 1.9
47.9
85.1 1.3
4
4
4
]
[
c]
[c]
[c]
[c]
–
–
12
3
2
3.7
9.8 1.2
6
]
[
c]
[c]
the gel networks were affected especially by the
symmetry of the IL anion. With the tetrahedral
3
3
6.3
9.2
2
]
[d]
[c]
[c]
–
49.1 2.6
anion, [BF ], a reticular network was observed (Fig-
4
ure 2g); with the octahedral anions, [PF ] and [SbF ],
[
m-C12im]
[m-C12im]
[m-C12im]
[Trim]
3
[m-C12im]
[Cit]
3
[m-C12im]
2
6
6
[1,4-BDC]
[2,6-NDC]
2
2
[b]
[EDTA]
thick textures characterized by spherulitic domains
were found (Figure 2d and S5j). In addition, as indi-
cated by the values of Tgel reported in Table 2, in-
creasing anion symmetry increased the ionogel sta-
bility.
[b]
[e]
[b]
[e]
[b]
[e]
[e]
m
[b]
[e]
Ionic liquids
T
gel
DH
m
T
gel
DH
m
T
gel
DH
m
T
gel
DH
T
gel
DH
m
3
6
9.9
2.0 2.1
35.4
62.4 1.9
34.7
ns
[
[
bmim][PF
6
]
[
c]
[c]
bmim][NTf
2
]
Finally, with the same IL as solvent, ([bmim][BF ]),
SM
47.2 1.5
28.4
47.4 2.0
SM
50.6 2.5
53.0
53.9 1.0
4
[c]
[c]
[c]
[c]
[
bmim][BF
4
]
reticular networks were detected for gels with [p-
C im][1,4-BDC] and [p-C im] [Cit] (Figures S5n and
1
2
12
3
2
2
6
8.2
5.9
SM
60.5
[
[
bmpyrr][NTf
2
]
2g), whereas spherulitic domains could be observed
in gels with ([2,6-NDC]-, [Trim]- and [EDTA]-based
salts (Figures S5g, S5o and 2h, respectively).
[
d]
[d]
–
–
a] SM=soft material. ns=no discernible DSC endotherm. [b] Tgel values (8C) were re-
producible within 18C; Tgel at 5 wt% of gelator by falling-ball method. [c] Tgel values
(
[
8C) were reproducible within 18C; Tgel at 5 wt% gelator using DSC thermograms.
À
d] For glass transition from DSC thermograms. [e] DH
m
(Jg ) from DSC thermograms.
Mechanical properties
Mechanical properties of both organo- and ionogels
were investigated at room temperature using rheol-
[
NTf ]>[bmpyrr][NTf ]. This order is not the one expected on
ogy and a frequency of 1 Hz for strain sweeps and strains from
0.02% to 0.5% for frequency sweeps; the values were chosen
to be within the linear viscoelastic region (LVR) of the gels. In
all strain sweeps, G’ was higher than G’’ at low values of strain,
indicating a solid-like behavior, and G’<G’’ at higher strains, in-
dicating liquid-like behavior. Furthermore, at the strain values
employed for frequency sweeps, (g), G’ was always higher than
G’’. These observations support the gel nature of the sam-
2
2
the basis of the anion coordination ability of the ILs ([BF ]>
4
[25]
[
SbF ]>[NTf ]>[PF ]). Rather, the IL anion symmetry appears
6 2 6
to be the more important factor.
In general, data collected using [p-C im] [Cit] and [p-
1
2
3
C im] [EDTA] as gelator also show slightly higher thermal sta-
12
2
bilities for ionogels formed in [bmim][NTf ] than in [bmpyrr]
2
[
NTf2].
[
26]
ples. A measure of the gels strength was extracted from the
crossover points of yield strain (g; i.e., where G’’ becomes
equal to G’) and the loss tangents (tan d=G’’/G’). Indeed, they
represent the level of stress needed to detect the flow of ma-
terial and strength of colloidal forces operating in the gel net-
works. In general, the increase in gel strength induces a de-
crease in tan d and a corresponding increase in g.
POM measurements
The gel phases were characterized further by several tech-
niques. In some cases, attempts to record POM images of gels
with 5 wt% of gelator between glass slides (Figures 2 and S5)
were unsuccessful because the pressure applied to the glass
while preparing the samples disrupted the SAFIN.
Comparisons of the images indicate that the nature of both
the solvent and the cation affected the morphologies. A clear
example is shown by the glycerol and octanol gels of [p-
C im] [EDTA] (Figures 2a,b). A very thick network is observed
Organogels
The rheology data are reported in Figure S6 and Table 3. No
rheological investigations were conducted for gels of [p-
C im] [EDTA] or [m-C im][1,4-BDC] in glycerol because of their
12
2
in glycerol whereas spherulitic domains are found in octanol.
12
2
12
2
+
2+
On the other hand, the gels of [p-C im ] and [m-C im
low thermal stability and one of them being a soft material
(Table 1).
1
2
12
]
-based salts with EDTA as the anion and glycerol as the sol-
vent were tangled fibrous systems (Figures 2a,c).
Tan d values ranged from 0.18 for [p-C im] [EDTA]/octanol
12
2
Three distinct morphologies were identified for the ionogels:
to 0.77 for [m-C im] [EDTA]/glycerol gels. Values less than
12 2
1
) thick textures, 2) spherulitic networks, and 3) reticular tex-
1 suggest SAFINs with strong association among particles. Ac-
tures. The actual morphology seems to be determined princi-
cording to g values and as accounted for by the increase in
Chem. Eur. J. 2016, 22, 11269 – 11282
www.chemeurj.org
11273
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Figure 2. POM images of organogels and ionogels at 5 wt%; a) [p-C12im]
perature after first heating and cooling of the sample; c) [m-C12im]
[Cit] in [bmim][PF ] at room temperature after first heating and cooling of the sample; e) [m-C12im]
the sample after first heating; f) [m-C12im] [Trim] in [bmim][BF ] at room temperature after first heating and cooling of the sample; g) [p-C12im]
] at room temperature; h) gel of [p-C12im] [EDTA] in [bmim][BF ] at 408C after first heating and cooling of the sample; i) [m-C12im] [EDTA] in [bmim]
] at room temperature.
2
[EDTA] in glycerol at room temperature; b) [p-C12im]
2
[EDTA] in octanol at room tem-
2
[EDTA] in glycerol at room temperature after first heating and cooling of the sample; d) [p-
[Trim] in [bmim][PF ] at 408C of cooling ramp of
[Cit] in
C
12im]
3
2
6
3
2
6
3
2
4
3
2
[
[
bmim][BF
BF
4
2
4
2
4
Table 3. G’ and G’’ at g=0.05%, tand=G’’/G’ and values of g at G’’=G’ for gels investigated at 5 wt% gelator concentrations and 258C. Error limits are
based on average of three different measurements with different aliquots.
Gel
G’ [Pa]
G’’ [Pa]
tan d
g at G’’=G’
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
p-C12im][1,4-BDC]/glycerol
p-C12im][2,6-NDC]/glycerol
16900Æ400
3340Æ330
5350Æ650
9940Æ640
4340Æ220
2200Æ800
840Æ50
6630Æ130
1830Æ70
4150Æ460
1850Æ370
590Æ80
0.39Æ0.01
0.55Æ0.02
0.77Æ0.01
0.18Æ0.06
0.14Æ0.01
0.13Æ0.02
0.14Æ0.01
0.18Æ0.06
0.16Æ0.01
0.22Æ0.02
0.29Æ0.03
0.180Æ0.001
0.23Æ0.01
0.13Æ0.02
0.17Æ0.04
0.11Æ0.01
0.11Æ0.02
0.19Æ0.02
0.18Æ0.01
4.8Æ0.2%
0.2Æ0.1%
4.3Æ0.2%
2.2Æ0.1%
3.2Æ0.6%
36Æ5%
m-C12im]
p-C12im]
p-C12im][1,4-BDC]/[bmim][BF
p-C12im][2,6-NDC]/[bmim][BF
2
[EDTA]/glycerol
2
[EDTA]/octanol
4
]
4
]
300Æ100
120Æ10
p-C12im]
p-C12im]
p-C12im]
3
3
2
[Trim]
[Cit] /[bmim][BF
[EDTA]/[bmim][BF
2
/[bmim][BF
4
]
21Æ2%
2
4
]
303Æ14
55Æ15
7.4Æ1.2%
32Æ9%
4
]
1500Æ700
20200Æ4700
19800Æ7500
28000Æ4600
3600Æ600
6630Æ870
8400Æ900
3340Æ560
3340Æ340
1390Æ20
260Æ70
m-C12im][1,4-BDC]/[bmim][BF
m-C12im][2,6-NDC]/[bmim][BF
4
]
]
4400Æ700
5800Æ800
4980Æ550
840Æ90
6.9Æ1.4%
15Æ3%
4
m-C12im]
m-C12im]
3
[Trim]
[EDTA]/[bmim][BF
/[bmim][PF
/[bmim][PF
[EDTA]/[bmim][PF
2
/[bmim][BF
4
]
]
0.5Æ0.3%
1.0Æ0.4%
6Æ1%
2
4
p-C12im]
p-C12im]
p-C12im]
3
3
2
[Trim]
[Cit]
2
6
]
850Æ120
1410Æ90
350Æ30
2
6
]
15Æ3%
6
]
3.4Æ0.1%
1.7Æ0.3%
0.4Æ0.1%
4.6Æ1.2%
m-C12im]
m-C12im]
m-C12im]
3
3
2
[Trim]
[Cit] /[bmim][PF
[EDTA]/[bmim][PF
2
/[bmim][PF
6
]
370Æ90
2
6
]
270Æ20
6
]
1400Æ300
250Æ60
Chem. Eur. J. 2016, 22, 11269 – 11282
www.chemeurj.org
11274
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
tan d values on going from [p-C im][1,4-BDC] to [p-C im][2,6-
For an analogous reason, the rheological properties of gels
1
2
12
NDC], gel strength decreased with the increase in p-surface
area of the gelator anion. Another indicator of gel strength is
Tgel. Interestingly, the decreases of Tgel values of organogels in
glycerol correlate with the rise in tan d values.
consisting of [p-C im] [Cit] and different ILs were investigated
12
3
2
(the Supporting Information, Figure S7).
The tan d values, from 0.11 for [p-C im] [EDTA]/[bmim][PF ]
12
2
6
to 0.29 for [m-C im][2,6-NDC] in [bmim][BF ], are lower than
12
4
the ones corresponding to organogels. They indicate that
stronger gels were formed in the IL solvents than in organic
solvents. A similar trend has been recently detected by Schal-
Ionogels
[
27]
ley et al. for a crown-ether functionalized bisurea gelator.
Viscoelastic properties of the IL solvents could, in principle,
contribute to gelation in the ionogels. However, the mechani-
Data reported in Table 3 show that tan d and g are affected
by the structures of both the solvent and gelator. Although
the differences in tan d values are small, the ionogels from the
para cation of the gelator are stronger than those from the
meta. This conclusion is consistent with the changes in the g
values. Interactions are stronger among gelator molecules with
the para isomer than with the meta.
cal properties of the most viscous IL employed, [bmim][PF6],
were those of a classical Newtonian liquid. Consequently, the
mechanical properties of the ionogels are attributed to the in-
teractions of both components. To compare the effects of
structural changes within the solvent portion, the rheological
properties of gels with [bmim][BF ] and [bmim][PF ] were in-
4
6
More complex behavior was detected when the gelator
anion was varied. Indeed, g values determined in [bmim][BF4]
indicate that increased aromatic character of the anion leads
to stronger gels: ionogel strength increases between [1,4-BDC]
and [2,6-NDC] and between [Cit] and [Trim]. However, the
vestigated (Figure 3 and Table 3).
effect of the anion in [bmim][PF ] solvent depended on the
6
isomer of the cation of the gelator. In general, and as found
for the organogels, the results with ionogels indicate that
higher rigidity and symmetry in both the cationic and anionic
parts of the gelator aid gel strength and thermal stability.
Table S5 summarizes data about g and tan d for ionogels
with [p-C im] [Cit] and different IL solvents.
12
3
2
The g values decrease in the order: [bmim][PF ]>[bmim]
6
[
BF ]>[bmim][SbF ]>[bmpyrr][NTf ]>[bmim][NTf ]. On the
4
6
2
2
other hand, analysis of tan d values indicates that the weakest
gel was formed in [bmim][NTf ] solution. This trend may
2
depend on the flexibility of the ionogel SAFINs. As discussed
above, increased symmetry of the IL anion seems to favor
spherulitic domains that should impart higher strength than
the other SAFIN types to the ionogels. However, G’ measured
in [bmim][BF ] was lower than in [bmim][NTf ]. In an attempt
4
2
to explain the trend, we also considered the anion coordina-
tion ability (b), polarity and viscosity (the Supporting Informa-
tion, Table S6). The two latter parameters have been used to
[
28]
interpret results of gelation tests with organic solvents. Of
these, viscosity appears to be the most important factor; its
decrease ([bmim][PF ]>[bmim][BF ]>[bmim][SbF ]>[bmim]
6
4
6
[
NTf ]) is in the same order as the values of g. Furthermore,
2
among the ILs investigated, [bmim][NTf ] forms ionogels that
2
flow most easily. This may possibly be related to the low cross-
[
29]
linking ability of the NTf anion.
2
Finally, comparison between tan d and g values of ionogels
with [bmim][NTf ] and [bmpyrr][NTf ] as the solvent suggest
2
2
that the aliphatic cation aids more gel strength.
Thixotropic and sonotropic behavior
Many reports have focused on creating thixotropic organogels
[
30]
because they may be useful in several applications. Exam-
ples of thixotropic organometallic LMWGs as well as toluene
gels with N-(3-hydroxypropyl)-dodecanamide or silicone oil
Figure 3. Strain (a) and frequency (b) sweep plots for ionogels obtained in
bmim][PF ] at 5 wt % gelator and 258C.
[
6
Chem. Eur. J. 2016, 22, 11269 – 11282
www.chemeurj.org
11275
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
gels of (R)-12-hydroxyoctadodecanamide have been report-
[31]
ed. Thus, the ability of our ionogels to regain their viscoelas-
tic properties after applying destructive strain from magnetic
stirring or ultrasound irradiation were explored.
Surprisingly few ionogels have been shown to be thixotrop-
ic. In these cases, their self-repairing ability was advantageous
[32]
for enhancing quasi-solid-state DSSC performance and for
[14]
improving solid-like lubricants. Furthermore, to the best of
our knowledge, the effect of ultrasound irradiation on the sta-
bility of ionogels has not been exploited extensively. We have
recently reported the thixotropic and sonotropic behavior of
[
17]
ionogels formed by some functionalized diimidazolium salts.
Data for the ionogels investigated here are reported in
Table S7 (the Supporting Information).
Firstly, gel phases were subjected to magnetic stirring (at
1
2
000 rpm for 5 minutes). With few exceptions (6 cases out of
2), gels reformed qualitatively after the samples were stored
at 48C overnight. The non-thixotropic gels showed the highest
thermal stability. We have recently detected a similar behavior
studying the thixotropic properties of two-component gels
formed by some diimidazolium salts in the presence of aro-
[
33]
matic guests. Based upon the results found here, thixotropic
recovery is enhanced when the SAFIN consists of a more flexi-
ble fibrous or spherulitic network in which network interac-
tions may be weaker.
The thixotropic behavior of the ionogels was also analyzed
more quantitatively by rheology. Gels were subjected initially
to conditions within the LVR (g=0.05–0.1% and f=1 Hz).
Thereafter, destructive strain was applied (g=5–50% and f=
1
Hz). The evolutions of G’ and G’’ after cessation of the de-
structive strain applying the strain are displayed in Figures 4
and S8 (the Supporting Information). In general, at least two
destruction–recovery cycles were performed on each sample.
Although the level of recovery was not complete in any of the
samples, G’ values higher than G’’ values, demonstrating the
presence of a gel, were found after applying strain levels in
the viscoelastic region.
Figure 4. G’ and G’’ at 258C as a function of time and application of low (G’
G“ regimes) and destructive stain (G’’>G’ regimes) to (a) [p-C im] [Cit] /
The percentage of recoveries indicated in Table S7 (the Sup-
porting Information) range from 0% to 15% in [m-
C im] [EDTA]/[bmim][BF ] to 73% in [p-C im] [Cit] /[bmim]
1
2
3
2
2 3 2 4
[bmim][NTf ] and (b) [p-C12im] [Cit] /[bmim][BF ] samples at 5 wt% of gelator
concentrations.
12
2
4
12
3
3
[
NTf2].
In some cases, G’ continued to rise over long periods (la-
As stated above, the sonotropic behavior of the ionogels
was also explored. Results collected are displayed in and they
allow distinguishing three different situations. In 5 out 22 gels
irradiated, ultrasound led to loss of viscoelasticity (the Support-
ing Information, Table S7); in 12 gels, irradiation did not affect
qualitatively the consistency of the samples; all of the 5 soft
materials were sonotropic. The sonotropic behavior was ob-
served predominantly for meta isomers of the gelator cation
bearing an aromatic anion. These materials also exhibit
medium or low thermal stabilities and mostly spherulitic
SAFINs (the Supporting Information, Figures S5n, 2 f and 2g).
beled as “G’ increasing” in Table S7) or oscillated with time (la-
beled as “oscillations” in Table S7). These behaviors did not
allow the percent of recovery to be determined. The oscillatory
behavior is especially bizarre. However, it was found on several
different samples prepared independently, with different
amounts of gels between the rheometer plates, and when
measurements were made on a different rheometer and with
plate–plate or cone–plate tool geometries.
[
34]
Nandi et al. have reported similar oscillatory behavior for
the moduli of gels. They ascribed the oscillations to a shear-as-
sisted re-aggregation process. That explanation seems logical
here as well. It could give rise to a different organization of the
gel network. Similar considerations have been made in consid-
eration of kinetic rheological and opacity measurements (see
below).
Chem. Eur. J. 2016, 22, 11269 – 11282
www.chemeurj.org
11276
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Kinetic study of gel phase formation: rheology measure-
ments
The kinetics of both organo- and ionogels formation were
studied as a function of the time by rheology. Gelation by
LMWGs can be described as a sequence of a stochastic nuclea-
tion events followed by growth of branched or unbranched
fibers and their entanglement or physical interaction to form
a 3D network (the SAFIN). The above processes have been
[35]
treated using the Avrami equation. This equation was initial-
ly developed to describe the crystallization of polymer and
other melts. However, it is frequently used in its log–log form
[
Eq. (1)] to study gelation processes at a constant temperature:
ln ½Àlnð1ÀFފ ¼ ½n ðln k þ ln tފ ð1Þ
In which k is a rate constant, n is the so-called Avrami expo-
nent that is associated with the type of growth leading to the
SAFIN, t is time and F, the volume fraction of gel, can be cal-
culated by the storage modulus at times 0 (i.e., when the ag-
gregation is first noted by an increase in G’), t and 1 (i.e.,
when G’ reaches a maximum, plateau value) [Eq. (2)].
F ¼ ðG0 ÀG0 Þ=ðG0 ÀG0ð0ÞÞ
ðtÞ
ð0Þ
ð1Þ
ð2Þ
The rheological kinetic measurements were carried out
within the LVR, using a low strain (ꢀ1%) and the data are
shown in Figures 5, S9 and S10 (the Supporting Information).
As mentioned above, it was possible to perform rheological
measurements only for organogels of [p-C im] [EDTA]/octanol
1
2
2
because the gel phases in glycerol using [p-C im][1,4-BDC]
1
2
and [p-C im][2,6-NDC] formed too slowly for G to be evalu-
12
(1)
ated. For the [p-C im] [EDTA]/octanol gel, values of k (=0.70”
12
2
À3 À2
1
0
s ) and n=1.89 were calculated. In particular, n=2 pre-
dicts the two-dimensional or rod-like growth of the SAFIN (as
is found in the POM images; Figure 2b).
[34]
For reason described previously,
G(1)
values could not be
Figure 5. Plots of G’ and G’’ as a function of time collected at 258C for: a) [p-
obtained in some experiments with ionogels, and therefore,
Avrami analyses were not possible. With the exception of gela-
tion by [p-C im][2,6-NDC], [p-C im] [EDTA] and [m-
C
5
12im]
3
[Trim]
2 4 3 2 6
/[bmim][BF ] and b) [p-C12im] [Cit] /[bmim][PF ] gel phases at
wt% gelator concentrations.
12
12
2
C im] [EDTA] in [bmim][BF ], plots of the ionogel data for
12
2
4
which Avrami plots are possible (Table 4 and the Supporting
Information, Figure S10) show an induction period, indicative
[36]
of slow, elementary nucleation steps. Induction periods were
also detected by RLS (see below).
Table 4. Rate constants (k) and kinetic order (n) of gel formation, derived
from rheology data treated by the Avrami Equation (1).
In most cases, the Avrami exponent predicts two-dimension-
al growth of the network, consistent with the spherulitic or re-
ticular objects observed in the POM images (Figures 2g, S5n–o
and S5m). However, n=3 is calculated for the [p-C im] [Cit] /
Àn
Entry
Gel
k [s
]
n
À3
À3
À3
À3
À3
1
[p-C12im][1,4-BDC]/
0.75”10
1.13”10
0.89”10
1.98”10
1.95”10
1.85
[bmpyrr][NTf
[p-C12im][1,4-BDC]/
bmim][BF
[p-C12im] [Trim]
bmim][BF
2
]
1
2
3
2
2
3
4
5
2.02
1.59
2.09
3.00
[
bmim][PF ] ionogel. The POM image in Figure 2d is consistent
6
[
4
]
with the predicted growth of 3D objects in the SAFIN.
The values of k indicate that the rate of gel phase formation
is not affected by the nature of the IL anions (see entries 4 and
3
2
/
[
4
]
[p-C12im]
3
[Cit]
2
/
4
[bmim][BF ]
5
in Table 4). On that basis, it appears from the experiments
3
[p-C12im] [Cit]
2
/
with [p-C im][1,4-BDC] that gel formation is slower when the
12
6
[bmim][PF ]
ILs have an aliphatic cation (see entries 1 and 2 in Table 4). Ob-
Chem. Eur. J. 2016, 22, 11269 – 11282
www.chemeurj.org
11277
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
viously, additional examples will be needed in the future to
test the generality of these hypotheses.
C im][1,4-BDC]/[bmim][BF ] and [p-C im] [Cit] /[bmim][SbF ]
12
4
12
3
2
6
show an induction period before aggregates become detecta-
ble.
A careful analysis of the plots shows that gelation occurs
through two different mechanisms that depend on the nature
of the gelator and the IL solvent. Data corresponding to time
and intensity of intermediate formation (t , I ) and gelation (t ,
Gel phase formation: RLS measurements
Ionogel phase formation was also investigated by RLS meas-
urements. RLS, a technique that detects aggregates in solution,
takes advantage of enhanced light scattering intensity ob-
served at wavelengths of light absorbed by aggregates when
m
m
e
Ie) are collected in Table S8. The data in Figure 6b indicate that
gelation proceeds by a single step mechanism: IRLS gradually
increases with time until reaching a plateau value. The data in
Figure 6a are consistent with a two-step mechanism. There, IRLS
increases (I ) and then decreases to a constant value (I ), corre-
[
37]
strong coupling exists among their chromophores.
RLS
measurements also provide an estimate of the size of the ag-
m
e
gregates; other factors being constant, RLS intensity (IRLS) is di-
sponding to the completion of gelation. This latter behavior
[
37b,38]
rectly proportional to the square of aggregate volume.
This technique has been used to study aggregation processes
[
17,18,41]
has been previously reported using different techniques.
On this basis, and bearing in mind that the self-assembly pro-
cess can involve fiber formation (i.e., 1D objects) before the 3D
networks of SAFINs are completed, we attribute the increase
and then decrease of the RLS intensity to initial formation of
larger aggregates that subsequently rearrange into smaller and
more stable ones. This is the opposite of Ostwald ripening, the
normal manner in which SAFIN structures age!
[39]
involving very different systems. Recently, we have used it to
[17,18,40]
follow gel phase formation.
Measurements of IRLS as a function of time were carried out
here at 258C (Figures 6 and S11). As found by rheological
measurements also, plots of IRLS as a function of time for [m-
Data collected in [bmim][BF ] solvent shows that gelation
4
times increase as the anions of the gelators are changed: [1,4-
BDC]<[EDTA]<[Trim]<[Cit]. For trivalent anions, SAFIN net-
works evolved faster with the aromatic anion ([Trim]), but
more extended aggregates were formed with the aliphatic
anion ([Cit]) (the Supporting Information, Figures S5o and 2g).
With [1,4-BDC] as the common gelator anion and [bmim]
[
BF ] as the common solvent, a one-step mechanism for SAFIN
4
2
+
formation was observed with [p-C im ] as the gelator cation
12
2
+
and a two-step mechanism occurred with [m-C im ] (en-
12
tries 1 and 9, Table S8). Furthermore, although the sizes of the
aggregates are very similar (IRLS ꢁ330 a.u.), the cation structure
affects gelation time significantly: t =600 and 4700 s for [p-
e
2
+
2+
C im ] and [m-C im ], respectively. Although other interest-
12
12
ing comparisons are available from the data in Table S8, but
they depend upon small numbers of examples, their generality
must be tested in future experiments.
Finally, the thermal reversibility of the gelation process of [p-
C im] [Cit] /[bmim][BF ] (i.e., are aggregates after consecutive
12
3
2
4
heating–cooling cycles of comparable size?) was investigated
using RLS measurements. Results from 3 cycles are displayed
in Figure S11 (the Supporting Information). Gel formation oc-
curred by a two-step mechanism and the times corresponding
to gel phase formation as well as to the maximum IRLS value
gradually decreased between cycles. However, because the IRLS
and Tgel values after each cycle (49.5, 49.7 and 47.98C for the
first, second and third cycle, respectively) were not affected by
reheating and recooling, we conclude that the gels and their
SAFINs remained virtually the same.
Opacity measurements
Gel phase formation was also investigated by recording the
opacity of gelating samples at 568 nm (Abs568 nm) as a function
of time (the Supporting Information, Figure S12).
Figure 6. Plots of IRLS as a function of time collected at 258C and correspond-
ing to ionogels at 5 wt% gelator concentrations.
Chem. Eur. J. 2016, 22, 11269 – 11282
www.chemeurj.org
11278
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Gel opacity can be related qualitatively to the number and
the size of polydisperse nanostructures in a system, including
[41b]
the crystallinity of a gel.
On that basis, there appears to be
no significant changes in the crystallinity of gels with different
gelators when [bmim][BF ] was the solvent. Similarly, only
4
small changes in the absorbance values were detected for gels
with [p-C im] [Cit] as the gelator and different ILs as the sol-
12
3
2
vent.
However, significant changes in crystallinity were found
when ionogels and organogels were compared. Abs568 nm
values increased on going from [p-C im] [EDTA]/octanol to [p-
1
2
2
C im] [EDTA]/[bmim][BF ] and [p-C im] [EDTA]/[bmim][PF ]
12
2
4
12
2
6
systems while crystallinity decreased on going from [p-C im]
12
[
1,4-BDC]/[bmim][BF ] to [m-C im][1,4-BDC]/[bmim][BF ], indi-
4 12 4
cating the importance of the isomer of the gelator cation.
Powder X-ray diffraction
Powder X-ray (PXRD) diffractograms of selected neat gelators
and their 5 wt% gels are compared in Figures 7 and S13 (the
Supporting Information). Spacings (d, nm) calculated from the
Bragg relationship are collected in Tables S9–S13, as well. Iono-
gels formed in [bmim][PF ] by the gelator consisting of the
6
para isomer of the cation and three different anions were in-
vestigated to gain insight into the influence of the anion on
molecular packing within the SAFINs (Figure S13). Also, the
PXRD patterns for both [p-C im] [EDTA] and [m-C im] [EDTA]
1
2
2
12
2
as neat solids and in [bmim][PF ] gels were compared to deter-
6
mine whether the neat solid and SAFIN morphs differ and the
role of the cation of the gelator on molecular packing within
the SAFINs (Figures 7 and S13). Finally, packing within SAFINs
of gels with [p-C im] [EDTA] as gelator and an IL ([bmim][PF ])
12
2
6
or a “normal” solvent (glycerol) were compared (Figure 7).
The 2q values of the diffraction peaks demonstrate that mo-
lecular packing of [p-C im] [EDTA] differs in its neat bulk solid,
1
2
2
[
42]
in its [bmim][PF ] ionogel and in its glycerol organogel. Also,
Figure 7. Diffractograms of neat gelators and of their 5 wt % gels in (a) glyc-
6
erol and (b) [bmim][PF
erol or [bmim][PF ].
6
] after empirical background subtraction of the glyc-
the diffractograms of neat [p-C im] [Trim] , [p-C im] [Cit] , and
1
2
3
2
12
3
2
6
[
[
m-C im] [EDTA] differ from those of their gels with [bmim]
12 2
PF ] as the liquid. Thus, all the gelators investigated are poly-
6
morphic; the processes by which they aggregate and nucleate
to form the 3D networks of their SAFINs depend acutely on
the nature of the solvent–gelator interactions.
their sol phases, and points to an obvious direction for future
investigations.
Although it is tempting to do so, we prefer not to use the
PXRD data to speculate about details of the specific packing
arrangements of the gelator molecules within their unit cells.
However, all neat gelators show diffraction peaks that are con-
sistent with (but not unique to) the presence of hydrogen
bonding: 2q=20.008 (4.43 Š), 20.688 (4.29 Š), 20.148 (4.40 Š)
and 20.928 (4.24 Š) for [p-C im] [EDTA], [p-C im] [Trim] , [p-
Although it was not possible to obtain diffraction quality
[42b]
single crystals of the gelators,
we were able to assign possi-
ble cell constants and lattice types in some cases by indexing
the PXRD peaks for the gelators and their gels (the Supporting
[42a]
Information, Table S14).
At least 40 diffraction peaks were
used to index each gelator, whereas 15–33 diffraction peaks
were used for the gels. A few peaks with very low intensity
could not be assigned; they may be from minor amounts of
a different morph or trace impurities. Regardless, most of gela-
tors and their gels pack in monoclinic lattices. As mentioned,
the assigned unit cells for the gelators and those of the gels
are different; the packing arrangements of the gelator mole-
cules change significantly from the neat solid phase to their
gel phases. Clearly, this observation constitutes a challenge to
an even deeper understanding of how the SAFINs form from
12
2
12
3
2
[
43]
C im] [Cit] and [m-C im] [EDTA], respectively. Also, a diffrac-
12
3
2
12
2
tion peak from [p-C im] [Cit] at 2q=25.088 (3.63 Š) may be
12
3
2
[
43–44]
due to p–p stacking.
The diffractograms of the organo- and ionogels contain
peaks that may be due to self-assembly through hydrogen
bonding interactions as well. Previously reported DFT calcula-
[
18]
tions on some diimidazolium salts indicate that the anions
of the gelators may form bridges among diimidazolium cat-
ions.
Chem. Eur. J. 2016, 22, 11269 – 11282
www.chemeurj.org
11279
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
As mentioned previously, the degree of additional stabiliza-
tion of the gels by p–p interactions seems to depend on both
the nature of the anion and the substitution pattern on the
cation of the gelators. Indeed, irrespective of solvent nature,
peaks at 2q=25.708 (3.46 Š) and 25.82 (3.54 Š) were ob-
served for [p-C im] [EDTA] gels in both glycerol and [bmim]
Experimental Section
Material and methods: Imidazole (ꢂ99%, Sigma–Aldrich), potassi-
um hydroxide (99%, Fluka), a,a’-p-dibromoxylene (97%, Sigma–Al-
drich), a,a’-m-dibromoxylene (97%, Sigma–Aldrich), 1-bromodode-
cane (97%, Sigma–Aldrich), Amberlite IRA-400 resin (chloride form)
12
2
(
99.8%, Sigma–Aldrich), sodium hydroxide (99%, Fluka), 1,4-benze-
[
PF6].
nedicarboxylic acid (98%, Sigma–Aldrich), 2,6-naphtalenedicarbox-
ylic acid (97%, Sigma–Aldrich), trimesic acid (95%, Sigma–Aldrich),
citric acid (99.5%, Fluka), ethylenediaminetetraacetic acid (98%,
Sigma–Aldrich), acetonitrile (99.8%, Sigma–Aldrich), dichlorome-
thane (99%, VWR), diethyl ether (99%, VWR) and all organic sol-
vents used for gelation tests were analytical reagents used as re-
ceived.
However, the gel of [m-C im] [EDTA] in [bmim][PF ] lacks
1
2
2
6
diffractions expected of p–p interactions and the gel of [p-
C im] [Cit] in [bmim][PF ] exhibits diffraction peaks consistent
12
3
2
6
with hydrogen bond formation only. Considering the non-aro-
matic nature of the [EDTA] anion, the occurrence of p–p inter-
actions in [p-C im] [EDTA] must be ascribed to interactions
12
2
ILs as [bmim][NTf ] (99%), [bmim][BF ] (99%), [bmim][PF ] (99%)
2
4
6
among the cations. In general, the data indicate more facile
bridging by [EDTA] than for the [Cit] anion. Also, DFT calcula-
were analytical reagents purchased from Iolitec and used as re-
ceived, while [bmim][SbF ] and [bmpyrr][NTf ] were prepared and
[
23]
6
2
tions
suggest a higher propensity for p-stacking by [p-
[45]
purified according to the reported procedures.
2
+
C im ]. Finally, evidence for p-p interactions was also found
12
1
13
H and C NMR measurements were recorded on Bruker instru-
in the diffraction peak of the [p-C im] [Trim] /[bmim][PF ] ion-
1
2
3
2
6
ments of 300 and 400 MHz.
ogel at 2q=25.838 (3.53 Š), where both cation/cation and
cation/anion interactions should be considered.
Melting points of products were measured by DSC. Precursor melt-
ing points were obtained on a Kofler melting point apparatus.
TGA measurements: TGA were performed on a TA 2910 differen-
tial scanning calorimeter interfaced to a TA Thermal Analyst 3100
controller while a slow stream of nitrogen flowed through the in-
strument cell. TGA measurements were performed equilibrating
the sample at 258C, isothermal 0.1 min, afterwards using a temper-
Conclusions
This work provides insights into the properties of gels formed
by some diimidazolium salts. Interestingly, these gelators are
able to gel solvents such as conventional solvents and ionic
liquids which are frequently claimed for their very different
properties. However, for both classes, solvents entrapped in
the SAFIN are highly polar and viscous. According to rheologi-
cal and opacity measurements, stronger and more crystalline
gels are formed in the ILs than in organic solvents.
À1
ature ramp with a rate of 108Cmin from 258C to 3008C.
DSC measurements: DSC was carried out on a DSC Q200 calorime-
ter (TA Instruments, New Castle, DE) interfaced to a TA Thermal An-
alyst 3100 controller connected to a RCS90 cooling system. Heat-
À1
ing and cooling cycles were done in a 50 mLmin stream of nitro-
gen. Samples were weighed (ꢁ5 mg for salts, ꢁ15 mg for gels) in
T-zero aluminum pans. Transition temperatures from DSC are re-
ported at the point of maximum heat flow.
The properties of the ionogels seem to depend mainly on
the nature of the IL solvent and of the gelator anion. These
factors also affect the morphology and size of the aggregates
characterizing the SAFINs as accounted for by POM images,
RLS and opacity measurements.
After equilibration of the sample at 258C, DSC measurements of
organic salts were performed heating the sample with a rate of
À1
2
08Cmin from 258C to 1808C and cooling it with the same rate
ramp to 08C. DSC measurements of gel phases were performed
equilibrating the sample at 258C and using a temperature ramp
À1
with a rate of 108Cmin from 258C to 1008C and a cooling ramp
Several gel phases show some thixotropic and sonotropic
behavior, which may be both useful in view of future applica-
tions. The temporal changes in the SAFINs, followed by UV/Vis
and RLS measurements, reveal significant differences in the ge-
lation mechanisms when the structure of the gelator or IL
anion are modified.
to 08C.
Preparation of gels and Tgel determination: Gels were prepared
by weighing into a screw-capped sample vial (diameter 1 cm) the
amount of salt and solvent (ꢁ250 mg). The sample vial was
heated in an oil bath at 90–958C until a clear solution was ob-
tained. The vial was then cooled and stored at 48C overnight. The
tube inversion test method was used to examine gel formation. To
determine the Tgel values, a lead ball (weighing 46.2 mg and 2 mm
of diameter) was placed on the top of the gel and the vial was put
Finally, XRD diffractograms, collected for selected neat gela-
tor and gel samples in glycerol and in [bmim][PF ], have al-
6
lowed hydrogen bond and p–p interactions to be identified as
some of the forces stabilizing the gels and controlling the mo-
lecular packing arrangements within the SAFINs. In particular,
the ability of the anion, whether aliphatic or aromatic, to con-
nect cation planes appears to be a major driving force for gel
stabilization.
À1
into a water bath. The bath temperature was gradually (28Cmin )
increased until the gel melted and the lead ball reached the
bottom of the vial (Tgel). The Tgel values were reproducible within
1
8C.
POM measurements: Samples were casted between two glasses
to record POM images. The samples were heated to their sol
phases and cooled. The instrument used is a Leitz 585 SM-LUX-POL
microscope equipped with crossed polarizers, a Leitz 350 heating
stage, a Photometrics CCD camera interfaced to a computer, and
an Omega HH503 microprocessor thermometer connected to a K
(Chrome-Alome) thermocouple (Omega Engineering, Inc.).
Data collected demonstrate how small modifications of gela-
tor structures and gelation solvents can modulate the proper-
ties of the soft materials. These, as a consequence of their
strength and crystallinity, hold great promise in the construc-
tion of new DSSCs and electrochemical devices.
Chem. Eur. J. 2016, 22, 11269 – 11282
www.chemeurj.org
11280
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Rheological measurements: Rheological measurements were re-
corded on an Anton Paar MCR 302 strain-controlled rheometer
using a Peltier temperature controller and a cone plate (CP 25-2)
tool, the sample was placed between the shearing plates of the
empirically subtracting the solvent components from the gel XRD
patterns.
[42a]
rheometer. Rheological properties, such as strain sweep and fre- Acknowledgements
quency sweep, were recorded before and after heating the
sample. Kinetics were performed in the linear-viscoelastic region,
We thank MIUR (FIRB 2010RBFR10BF5V) and the US National
heating the sample to 908C for 30 min to ensure the formation of
Science Foundation (Grants CHE-1147353 and 1502856) for fi-
nancial support. We thank Mr. Pasha Tabatabai of Georgetown
University for help with the rheological measurements.
À1
solution, then it was cooled at ꢁ208Cmin to 228C. When the
moduli presented a plateau, indicating the reformation of the gel,
strain sweep and frequency sweep were performed again to com-
pare the results with the previous ones.
Keywords: gels · green chemistry · ionic liquids · organic
salts · supramolecular gels
Self-healing properties of gel phases were also tested by rheome-
ter; they were carried out at room temperature and at a frequency
of 1 Hz, varying the strain from low to high percentage values for
fixed time intervals of 5 minutes (values of yield strain to apply
were chosen from linear viscoelastic region and destructive strain
of the gel). Rotational strain was kept at 0% for 0.05 s before
changing from destructive strain to linear viscoelastic region condi-
tions. When the moduli reached plateau values after the cessation
of disruptive strain, it was possible to calculate the percentage of
recovery of the initial G’ value. In those cases where the G’ values
oscillated or increased continuously, G’ and G’’ were measured at
a fix strain and frequency for longer periods.
[1] a) P. Terech, R. G. Weiss, Chem. Rev. 1997, 97, 3133–3160; b) R. G. Weiss,
P. Terech, Molecular Gels, Materials with Self-Assembled Fibrillar Networks,
Springer, Dordrecht, 2006.
[
[
2] R. G. Weiss, J. Am. Chem. Soc. 2014, 136, 7519–7530.
3] a) U. K. Das, V. G. Puranik, P. Dastidar, Cryst. Growth Des. 2012, 12, 5864–
5868; b) V. A. Mallia, R. G. Weiss, J. Phys. Org. Chem. 2014, 27, 310–315.
[
4] a) G. O. Lloyd, J. W. Steed, Soft Matter 2011, 7, 75–84; b) M. Rodrigues,
A. C. Calpena, D. B. Amabilino, M. L. Garduno-Ramirez, L. Perez-Garcia, J.
Mater. Chem. B 2014, 2, 5419–5429; c) P. Sahoo, N. N. Adarsh, G. E.
Chacko, S. R. Raghavan, V. G. Puranik, P. Dastidar, Langmuir 2009, 25,
8
742–8750; d) Y. Sun, X. Huang, Supramol. Chem. 2015, 27, 21–27.
5] G. G. Eshetu, M. Armand, B. Scrosati, S. Passerini, Angew. Chem. Int. Ed.
014, 53, 13342–13359; Angew. Chem. 2014, 126, 13558–13576.
6] a) K. Hanabusa, H. Fukui, M. Suzuki, H. Shirai, Langmuir 2005, 21,
RLS measurements: RLS measurements were carried out with
a spectrofluorophotometer using a synchronous scanning mode in
which the emission and excitation monochromators were preset to
identical wavelengths. The RLS spectrum was recorded from 300
to 600 nm with both the excitation and emission slit widths set at
[
[
2
10383–10390; b) J. Le Bideau, L. Viau, A. Vioux, Chem. Soc. Rev. 2011,
4
0, 907–925.
1
.5 nm. We chose as working wavelength the one corresponding
[7] N. V. Plechkova, K. R. Seddon in Methods and Reagents for Green Chemis-
try: An Introduction, (Eds.: P. Tundo, A. Perosa, F. Zecchini), chap 5„
Wiley, New York, 2007.
to the intensity maximum of the emission spectrum. Samples for
a typical kinetic measurement were prepared by injecting into
a quartz cuvette (light path 0.2 cm) the limpid hot solution of salt.
The measurements were carried out at 258C. Spectra were record-
ed until gel formation. The gel phase obtained at the end of the
measurement was stable after the tube inversion test.
[
8] D. Kim, G. Lee, D. Kim, J. S. Ha, ACS Appl. Mater. Interfaces 2015, 7,
608–4615.
4
[9] N. Mohmeyer, D. Kuang, P. Wang, H.-W. Schmidt, S. M. Zakeeruddin, M.
Gratzel, J. Mater. Chem. 2006, 16, 2978–2983.
10] N. Kimizuka, T. Nakashima, Langmuir 2001, 17, 6759–6761.
[
Opacity measurements: Opacity measurements were recorded
with a spectrophotometer. The opacity of the gel phases was de-
termined with UV/Vis measurements as a function of time, at
a wavelength of 568 nm and a temperature of 258C. As described
for RLS measurements, samples for a typical measurement were
prepared by injecting into a quartz cuvette (light path 0.2 cm)
a hot solution of salt. Spectra were recorded until gel formation.
The gel phase obtained at the end of the measurement was stable
according to the tube inversion test.
[11] a) Z. Fan, Y. Wang, Z. Xue, L. Zhang, Y. Chen, S. Zhang, J. Sol-Gel Sci.
Technol. 2014, 72, 328–333; b) Y. Ishioka, N. Minakuchi, M. Mizuhata, T.
Maruyama, Soft Matter 2014, 10, 965–971; c) K. Rawat, J. Pathak, H. B.
Bohidar, Soft Matter 2014, 10, 862–872; d) X. Zhu, H. Zhang, J. Wu,
Sens. Actuators B 2014, 202, 105–113; e) M. Czugala, C. O’Connell, C.
Blin, P. Fischer, K. J. Fraser, F. Benito-Lopez, D. Diamond, Sens. Actuators
B 2014, 194, 105–113; f) T. M. Benedetti, T. Carvalho, D. C. Iwakura, F.
Braga, B. R. Vieira, P. Vidinha, J. Gruber, R. M. Torresi, Sol. Energy Mater.
Sol. Cells 2015, 132, 101–106.
[12] P. C. Marr, A. C. Marr, Green Chem. 2016, 18, 105–128.
[
[
[
13] W. Jiang, J. Hao, Z. Wu, Langmuir 2008, 24, 3150–3156.
14] M. Cai, Y. Liang, F. Zhou, W. Liu, J. Mater. Chem. 2011, 21, 13399–13405.
15] W. Zhang, C. Yuan, J. Guo, L. Qiu, F. Yan, ACS Appl. Mater. Interfaces
Thixotropic and sonotropic behavior: The gel phases obtained
were subjected to two different external stimuli. The mechanical
stimulus involved stirring the gel phase at 1000 rpm for 5 min,
using a stirring bar (length 8 mm, height 3 mm). The sonotropic
behavior of the gel phases was tested by irradiating in an ultra-
sound water bath for 5 min with a power of 200 W and a frequency
of 45 kHz. Thereafter, the materials were stored at 48C overnight.
When the samples were stable to the tube-inversion test, the gels
were qualitatively defined as thixotropic or sonotropic.
2014, 6, 8723–8728.
[
16] F. D’Anna, R. Noto, Eur. J. Org. Chem. 2014, 4201–4223.
[17] F. D’Anna, C. Rizzo, P. Vitale, G. Lazzara, R. Noto, Soft Matter 2014, 10,
9281–9292.
[18] F. D’Anna, P. Vitale, F. Ferrante, S. Marullo, R. Noto, ChemPlusChem 2013,
7
8, 331–342.
[
19] S. Raghavan, B. Cipriano in Gel Formation: Phase Diagrams Using Tab-
letop Rheology and Calorimetry (Eds.: R. Weiss, P. Terech), Springer, Dor-
drecht, 2006, pp. 241–252.
Powder X-ray diffraction (XRD): Powder X-ray diffraction (XRD)
patterns of samples were obtained on a Rigaku Ultima IV X-Ray dif-
fractometer with CuKa X-rays (l=1.54 Š) generated by a Rigaku
generator operating at 40 kV and 30 mA with the collimator at
[20] J. L. Anderson, R. Ding, A. Ellern, D. W. Armstrong, J. Am. Chem. Soc.
2005, 127, 593–604.
[
21] F. D’Anna, H. Q. N. Gunaratne, G. Lazzara, R. Noto, C. Rizzo, K. R. Seddon,
Org. Biomol. Chem. 2013, 11, 5836–5846.
22] M. H. Nielsen, S. Aloni, J. J. De Yoreo, Science 2014, 345, 1158–1162.
23] F. D’Anna, F. Ferrante, R. Noto, Chem. Eur. J. 2009, 15, 13059–13068.
0
.5 mm and a D/Tex detector.
[
[
Salts, solvents and gels were placed on a silicon substrate as thin
films, and measurements were made overnight from 3 to 50 (8) in
[24] J. Kagimoto, N. Nakamura, T. Kato, H. Ohno, Chem. Commun. 2009,
2
q. The diffraction peaks for the gel networks were obtained by
www.chemeurj.org 11281
2405–2407.
Chem. Eur. J. 2016, 22, 11269 – 11282
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[
[
[
[
25] L. Crowhurst, P. R. Mawdsley, J. M. Perez-Arlandis, P. A. Salter, T. Welton,
Phys. Chem. Chem. Phys. 2003, 5, 2790–2794.
26] a) C. Daniel, C. Dammer, J.-M. Guenet, Polymer 1994, 35, 4243–4246;
b) A. Garai, A. K. Nandi, J. Polym. Sci. Part B 2008, 46, 28–40.
27] Z. Qi, N. L. Traulsen, P. M. de Molina, C. Schlaich, M. Gradzielski, C. A.
Schalley, Org. Biomol. Chem. 2014, 12, 503–510.
28] a) H. Bunzen, Nonappa, E. Kalenius, S. Hietala, E. Kolehmainen, Chem.
Eur. J. 2013, 19, 12978–12981; b) H. Xu, J. Song, T. Tian, R. Feng, Soft
Matter 2012, 8, 3478–3486.
[37] a) R. F. Pasternack, C. Bustamante, P. J. Collings, A. Giannetto, E. J. Gibbs,
J. Am. Chem. Soc. 1993, 115, 5393–5399; b) R. F. Pasternack, P. J. Col-
lings, Science 1995, 269, 935–939.
[38] J. Anglister, I. Z. Steinberg, J. Chem. Phys. 1983, 78, 5358–5368.
[39] a) Y. Feng, Z.-X. Liu, H. Chen, Z.-C. Yan, Y.-M. He, C.-Y. Liu, Q.-H. Fan,
Chem. Eur. J. 2014, 20, 7069–7082; b) L. Yu, Y. Zhang, R. Chen, D.
Zhang, X. Wei, F. Chen, J. Wang, M. Xu, Talanta 2015, 131, 475–479.
[40] C. Rizzo, F. D’Anna, S. Marullo, P. Vitale, R. Noto, Eur. J. Org. Chem. 2014,
1013–1024.
[
[
29] A. J. McLean, M. J. Muldoon, C. M. Gordon, I. R. Dunkin, Chem. Commun.
[41] a) M. Tena-Solsona, B. Escuder, J. F. Miravet, V. Casttelleto, I. W. Hamley,
A. Dehsorkhi, Chem. Mater. 2015, 27, 3358–3365; b) P. Terech, D. Pasqui-
er, V. Bordas, C. Rossat, Langmuir 2000, 16, 4485–4494.
[42] a) E. Ostuni, P. Kamaras, R. G. Weiss, Angew. Chem. Int. Ed. Engl. 1996, 35,
1324–1326; Angew. Chem. 1996, 108, 1423–1425; b) M. Zhang, S. Sel-
vakumar, X. Zhang, M. P. Sibi, R. G. Weiss, Chem. Eur. J. 2015, 21, 8530–
8543.
[43] S. Roy, A. Baral, A. Banerjee, Chem. Eur. J. 2013, 19, 14950–14957.
[44] a) A. Ajayaghosh, V. K. Praveen, Acc. Chem. Res. 2007, 40, 644–656;
b) S. J. George, A. Ajayaghosh, Chem. Eur. J. 2005, 11, 3217–3227.
[45] F. D’Anna, S. La Marca, P. Lo Meo, R. Noto, Chem. Eur. J. 2009, 15, 7896–
7902.
2
002, 1880–1881.
30] a) X. Huang, S. R. Raghavan, P. Terech, R. G. Weiss, J. Am. Chem. Soc.
006, 128, 15341–15352; b) Y. Ohsedo, M. Miyamoto, A. Tanaka, H. Wa-
2
tanabe, New J. Chem. 2013, 37, 2874–2880.
[
31] a) M. Lescanne, P. Grondin, A. O’AlØo, F. Fages, J. L. Pozzo, O. M. Monval,
P. Reinheimer, A. Colin, Langmuir 2004, 20, 3032–3041; b) V. A. Mallia,
M. George, D. L. Blair, R. G. Weiss, Langmuir 2009, 25, 8615–8625.
32] Y. Fang, J. Zhang, X. Zhou, Y. Lin, S. Fang, Electrochim. Acta 2012, 68,
[
[
2
35–239.
33] P. Vitale, F. D’Anna, S. Marullo, R. Noto, Soft Matter 2015, 11, 6652–
662.
6
[
[
[
34] B. Roy, P. Bairi, A. Saha, A. K. Nandi, Soft Matter 2011, 7, 8067–8076.
35] X. Y. Liu, P. D. Sawant, Appl. Phys. Lett. 2001, 79, 3518–3520.
36] X. Huang, P. Terech, S. R. Raghavan, R. G. Weiss, J. Am. Chem. Soc. 2005,
Received: February 12, 2016
1
27, 4336–4344.
Published online on June 30, 2016
Chem. Eur. J. 2016, 22, 11269 – 11282
www.chemeurj.org
11282
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim