Molecular hydrogels from bolaform amino acid derivatives: A structure-properties study based on the thermodynamics of gel solubilization

Article abstract of DOI:10.1002/chem.201103193

Insight is provided into the aggregation thermodynamics associated to hydrogel formation by molecular gelators derived from L-valine and L-isoleucine. Solubility data from NMR measurements are used to extract thermodynamic parameters for the aggregation i

Full text of DOI:10.1002/chem.201103193

FULL PAPER
DOI: 10.1002/chem.201103193
Molecular Hydrogels from Bolaform Amino Acid Derivatives: A Structure–
Properties Study Based on the Thermodynamics of Gel Solubilization
Vicent J. Nebot,[a] Josꢀ Armengol,[a] Johan Smets,[b] Susana Fernꢁndez Prieto,[b]
Beatriu Escuder,*[a] and Juan F. Miravet*[a]
Abstract: Insight is provided into the
aggregation thermodynamics associated
to hydrogel formation by molecular ge-
lators derived from l-valine and l-iso-
leucine. Solubility data from NMR
measurements are used to extract ther-
modynamic parameters for the aggre-
gation in water. It is concluded that at
room temperature and up to 558C,
these systems form self-assembled fi-
brillar networks in water with quite
low or zero enthalpic component,
whereas the entropy of the aggregation
is favorable. These results are ex-
plained by considering that the hydro-
phobic effect is dominant in the self-as-
sembly. However, studies by NMR and
IR spectroscopy reveal that intermolec-
ular hydrogen bonding is also a key
issue in the aggregation process of
these molecules in water. The low en-
thalpy values measured for the self-as-
sembly process are ascribed to the
result of a compensation of the favora-
ble intermolecular hydrogen-bond for-
mation and the unfavorable enthalpy
component of the hydrophobic effect.
Additionally, it is shown that by using
the hydrophobic character as a design
parameter, enthalpy-controlled hydro-
gel formation, as opposed to entropy-
controlled hydrogel formation, can be
achieved in water if the gelator is polar
enough. It is noteworthy that these two
types of hydrogels, enthalpy-versus en-
tropy-driven hydrogels, present quite
different response to temperature
changes in properties such as the mini-
mum gelator concentration (mgc) or
the rheological moduli. Finally, the
presence of a polymorphic transition in
a hydrogel upon heating above 708C is
reported and ascribed to the weaken-
ing of the hydrophobic effect upon
heating. The new soft polymorphic ma-
terials present dramatically different
solubility and rheological properties.
Altogether these results are aimed to
contribute to the rational design of mo-
lecular hydrogelators, which could be
used for the tailored preparation of
this type of soft materials. The reported
results could also provide ground for
the rationale of different self-assembly
processes in aqueous media.
Keywords: hydrophobic effect
·
molecular gels · polymorphism · sol-
ubility · thermodynamics
Introduction
and to promote novel applications that require the combina-
tion of robustness and adaptability.[4] There is a general
agreement that the use of water as solvent in supramolec-
ular chemistry research makes a significant difference when
compared to any other solvent. For example, hydrophobic
interactions are fundamental for the study of binding mech-
anisms in supramolecular chemistry.[5] As highlighted by En-
gerts and Blokzijl,[6e] water presents unique properties such
as a small molecular volume, the capacity of forming hydro-
gen-bonded networks, and a very low isothermic compressi-
bility. These remarkable properties are responsible both of
water being the solvent of life processes[7] and of some awk-
ward properties that emerge in the presence of this solvent
such as an unusual reactivity[8] or the so called hydrophobic
effect.[6] Leaving aside the controversy on the proper defini-
tion of the hydrophobic effect, it is well known that it plays
a key role in the organization of living matter and, in gener-
al, of soft materials such as, for example, micelles or vesicle-
s,[6a] as well as in protein science[6g] or for the properties of
foldamers.[9]
The detailed rationalization of the structural, kinetic, and
thermodynamic parameters relevant for self-assembly pro-
cesses that yield soft materials represents a very appealing
goal.[1] In this context, aqueous environments are especially
interesting due to the biological relevance of water. For ex-
ample, a recent review highlights the interesting possibilities
that emerge from the marriage of synthetic supramolecular
chemistry and biological systems[2] and in other recent
papers emphasis is put on the diversity of supramolecular
complex assemblies that can be formed in aqueous media.[3]
It has been pointed out that water-based non-covalent mate-
rials have the potential to replace conventional polymers
[a] V. J. Nebot, J. Armengol, Dr. B. Escuder, Dr. J. F. Miravet
Department de Quꢀmica Inorgꢁnica i Orgꢁnica
Universitat Jaume I, 12071 Castellꢂ (Spain)
Fax : (+34)964728214
E-mail: escuder@uji.es
miravet@uji.es
Molecular gels represent an intriguing case of self assem-
bly of low molecular weight species into nanoACTHNUGRTENUNG(micro)fibrillar
[b] Dr. J. Smets, S. F. Prieto
Procter & Gamble, 1853 Strombeek-Bever (Belgium)
networks that percolate the solvent and transform it into a
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201103193.
viscoelastic material, namely into a molecular gel.[10] Key
Chem. Eur. J. 2012, 18, 4063 – 4072
ꢃ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4063

features of this type of soft
matter when compared to
most cases of gels formed by
polymers are the remarkable
lyo- and thermoreversibility
together with the precise struc-
tural arrangement associated
to the self-assembly process
that generates molecular gels.
These materials present appli-
cations in a wide range of
areas such as regenerative
medicine, controlled drug re-
lease, optoelectronic materials,
or catalysts among others.[11]
As an example of the above-
mentioned special behavior of
Scheme 1. Chemical structure of the compounds discussed in this paper.
water, commonly molecular gels are divided into organogels
(gels in organic solvents) and hydrogels. The preparation of
molecular hydrogels has been reviewed[12] and, among the
structural diversity found in hydrogelators, common struc-
tural motifs include the presence of amino acid or peptide
moieties, hydrophobic aliphatic tails, and aromatic moieties.
For example, a very detailed and illustrative report on the
hydrogelation capabilities of some amino acid derivatives
was reported by Menger and Caran.[13] The ab initio design
of molecules capable to generate molecular gels or the pre-
diction of the gel properties based on the structure of the
gelator are major challenges in this field.[14] In this sense, the
relationship among solubility, minimum gelator concentra-
tion (mgc), and gel–sol transition temperature (Tgel) was
highlighted for a series of molecular gels formed in tolu-
ene.[15] Very recent work also points to the use of solubility
parameters as predictors for the formation of molecular
gels.[16] It has been stressed that in order to achieve rational
design of molecular gels a combination of thermodynamic
and kinetic studies is required.[17] Although thermodynamic
parameters associated to molecular gel formation in organic
solvents are described in many cases,[15,18] in the case of mo-
lecular hydrogels there is a lack of data on this respect. For
example, no thermodynamic data are reported in the above-
mentioned reviews on molecular hydrogels and, up to our
knowledge, only in some cases thermodynamic parameters
have been extracted for the aggregation of molecular gela-
tors in mixtures of water with organic solvents.[19] Overall,
the role played by hydrophobic effects in molecular gels and
its regulation by means of structural changes has not been
studied in detail. Recently, Adams et al. reported a clear
correlation between the hydrophobic character and the gela-
tion efficiency for a series of 9-fluorenylmethoxycarbonyl
(Fmoc)-protected dipeptides.[20] In particular, the role
played by intermolecular hydrogen bonding versus hydro-
phobic effects in the formation of hydrogels by dipeptides
has been pointed out recently, as a source of interest in this
field and the presence of both hydrogen-bonding interac-
tions and hydrophobic effects in the formation of molecular
hydrogels has been pointed out in some cases.[21]
Here, we report on a detailed study of the formation of
molecular hydrogels by a derivative of l-valine 1 and the
closely related compounds 2–4 (Scheme 1). Additionally, hy-
drogels formed by compound 7 are analyzed and compared
to those formed by compounds 2–4. The results shown here
highlight the role that entropy and enthalpy may play in any
self-assembly process in water and could provide a ground
for the rational design of hydrogels and, in general, of soft
matter building blocks. In the present work, special empha-
sis is placed on the role that hydrophobic interactions play
in the aggregation process. This study takes advantage on
the fact that compound 1 is known to be an “ambidextrous”
gelator capable of forming gels in organic solvents and
water.[22] Therefore, the thermodynamic driving forces for
the aggregation of this molecule in both solvents are com-
pared.
Results and Discussion
Aiming to evaluate thermodynamic parameters for the pro-
cess of hydrogel formation by compound 1, the solubility of
the networked gelator in acetonitrile and water were com-
pared by means of NMR measurements.[23] As shown in
Figure 1, an exponential relationship of the solubility with
the temperature was measured for the gel formed in aceto-
nitrile, a pattern which has been commonly observed for dif-
ferent gelators. However, the solubility of the gelator in
water shows a much lower dependence on the temperature
values in the range T=20–558C. What is particularly re-
markable is that the measured solubility decreases slightly
from 9.3 to 8.7 mm upon going from 25 to 558C. At this
point, the solubility of the gelator increases with the temper-
ature in a related way to that observed in acetonitrile. A
similar solubility profile has been previously reported for
the aggregation of a modified amyloid beta heptapeptide,
and was ascribed to the enthalpy and entropy dependence
on temperature for the peptide solubilization in water.[6i]
Thermodynamic parameters involved in the aggregation
can be obtained from the data shown in Figure 1. Firstly, a
4064
www.chemeurj.org
ꢃ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 4063 – 4072

Molecular Hydrogels from Bolaform Amino Acid Derivatives
FULL PAPER
lack of an enthalpic component at this temperature, being
the entropy associated to this process equal to the natural
logarithm of the solubility divided by the gas constant (see
the Supporting Information for equations). Noticeably, a
negative, unfavorable, solubilization entropy value (see
Table 1) is obtained for the gel network formed in water.
This negative entropy value obtained can be considered in a
first analysis a counterintuitive result considering that the
solubilization process transforms a condensed phase into a
solution, which would present more degrees of freedom as-
sociated, for example, to molecular translation and rotation.
However, this result is closely related to those obtained in
the study of the solubility of alkanes in water and can be ra-
tionalized considering the so-called hydrophobic effect. In
brief, the solubilization of hydrophobic components in water
can result, as noted for example by Engberts and Blokzijl,[6e]
in negative entropy values associated to “the small molecu-
lar volume of the water molecules” and to the “reduction of
rotational freedom of water molecules in the first hydration
layers”. The decrease of solubility with temperature de-
scribed is to some point related with the so called “inverse
temperature transitions” reported in aqueous media that
result, for example, in crystallization of polypeptides or in
enhanced protein stability on raising the temperature.[24]
Noticeably, the different thermodynamic parameters asso-
ciated to gels formed in water and acetonitrile by compound
1 result in some unusual properties of the hydrogels when
compared to common molecular organogels. For example,
an interesting consequence of the flatness of the solubility
versus temperature graph in the temperatures range of ap-
proximately 25–708C is that the minimum gelator concen-
tration value, 17 mm, is virtually independent of the temper-
ature in this interval. A curious consequence of this proper-
ty is evidenced in the graphs of temperature for the gel-to-
solution transition (Tgel) versus concentration. Typically this
type of graph shows a marked slope in the initial range of
concentration and then a flat profile at higher concentra-
tions, which can be related to the exponential relationship
of concentration and temperature (Figure 2, top). Notably,
the Tgel versus concentration graph obtained for the hydro-
gel derived from 1 is atypical because Tgel =858C represents
an absolute minimum in the Tgel values. No gel can be pre-
pared with thermal stability below this point (in the temper-
ature range studied, which starts at 258C) because the con-
centration at which a gel melts at 858C can also be consid-
ered the mgc value at 258C.
Figure 1. Temperature dependence of the solubility of compound 1 in the
gels formed in acetonitrile (top) and water (bottom). The data were ob-
tained by NMR analysis.[23]
van’t Hoff treatment of the data for the gel formed in aceto-
nitrile with the thermodynamic parameters for the solubili-
zation of the fibrillar network, results in positive values for
the enthalpy change (unfavorable) and entropy change (fa-
vorable) associated to this process (Table 1). Regarding the
Table 1. Solubilization enthalpy and entropy for the fibrillar gel network
of compound 1 at 558C.[a,b]
Solvent
DH [kJmolÀ1
]
TDS [kJmolÀ1
]
water
acetonitrile
0 (1)
35 (2)
À11 (1)
19 (1)
[a] [1]=19 mm. [b] Estimated error in parentheses.
hydrogel, a similar analysis of the solubility data is not feasi-
ble as a result of the non-exponential relationship between
solubility and temperature for this gelator in water. Howev-
er, the solubility versus temperature profile in water allows
for the easy extraction of some thermodynamic parameters
of this system. The relative invariance of the solubility in
water from 25 to 558C must be associated to an almost neg-
ligible enthalpic component in the solubilization process[6c]
(see the Supporting Information for equations). As a matter
of fact, the solubility of the gel network in water at 558C
represents a minimum and, therefore, at this temperature
the enthalpy change associated to the solubilization process
must be zero. Therefore, the entropy change of the solubili-
zation can be calculated at 558C, taking advantage of the
Of special interest is the amazing variation of the rheolog-
ical properties of the hydrogel with the temperature. As can
be seen in Figure 3, the storage modulus, G’, measured for
the hydrogel formed by 1 shows an intriguing big increase
(notice the logarithmic scale) in the temperature range 25–
558C. Namely, the gel network seems to be “stronger” as
the temperature is raised. At higher temperatures the G’
module decreases clearly, a behavior which is expected for
this type of supramolecular systems. A related case has been
reported for an aqueous triblock copolymeric hydrogel[25a]
and a similar trend has been reported, although the ob-
Chem. Eur. J. 2012, 18, 4063 – 4072
ꢃ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
4065

J. F. Miravet, B. Escuder et al.
Summing up, the aggregation of compound 1 into fibrillar
networks in acetonitrile is controlled by the enthalpy and in
water (at temperatures below 558C) by the entropy. To as-
certain how this observation correlates with the structure of
the assemblies formed in both solvents structural studies by
using NMR and IR spectroscopy were performed. The ag-
gregation of compound 1 in organic solvents most likely
proceeds through a structural arrangement related to that
found for analogues, which present terminal benzoyl of ben-
zyloxycarbonyl (Boc) units. The proposed structural ar-
rangement shown in previous work involves multiple hydro-
gen bonds among the molecules of gelator yielding a b-
sheet-like structure (see Figure 4).[26] A similar aggregation
Figure 4. Schematic aggregation model of for compound 1 and molecular
mechanics energy minimized model (non-polar hydrogen atoms are re-
moved for clarity; MACROMODEL 9.0, AMBER*, GB/SA simulation
of chloroform as solvent).
Figure 2. Temperature measured for the gel-to-solution transition for the
gels formed by compound 1 in acetonitrile (top) and water (bottom). The
data were obtained from oscillatory shear rheology studies (see the Sup-
porting Information).
pattern is reported for the crystalline structure of closely re-
lated Boc-protected analogues.[27] Although the possibility of
pyridine units participating in intermolecular hydrogen
bonding is feasible, in the case of compound 1 previous
studies regarding the interaction of the gel network with
metal cations and guest species with hydrogen-bonding
donor capabilities seem to discard this possibility in our
system.[22,28]
1
A H NMR study of compound 6 in CD3CN, a fully solu-
ble analogue of compound 1, confirms that hydrogen bond-
ing is a driving force for the aggregation of this type of mol-
ecules in acetonitrile, as shown in the shift of the NH signals
upon increasing concentration (Figure 5). The aggregation
model proposed in acetonitrile (Figure 4) would be in ac-
Figure 3. Variation of the rheological storage modulus (Gꢄ, logarithmic
scale) with the temperature for the gels formed by compound 1 in water
^
^
(
, 19 mm) and acetonitrile ( , 6 mm). Oscillation frequency=1 Hz and
oscillatory stress=130 Pa. For clarity the variation of loss modulus (G’’)
is omitted. See the Supporting Information for a full set of data.
served variation is much smaller, for a peptide amphiphile
capable of forming hydrogels in the presence of calcium
chloride.[25b] Noticeably, the rheological profile of Figure 3
shows similarities with the solubility profile shown in
Figure 1, being the maximum G’ value coincident with the
minimum observed solubility. On the other hand, the rheo-
logical analysis for the gel formed in acetonitrile shows the
expected trend, becoming the gel progressively weaker upon
increasing the temperature, and revealing a nice correspond-
ence with the solubility increase shown in Figure 1.
Figure 5. Fragment of the 1H NMR spectrum of compound 6 in CD3CN,
showing the resonance of the amide NHs. [6]=1mM (top), 76 mm
(bottom).
4066
www.chemeurj.org
ꢃ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 4063 – 4072

Molecular Hydrogels from Bolaform Amino Acid Derivatives
FULL PAPER
cordance with the important enthalpic component
(35 kJmolÀ1) of the solubilization process which would be
related mainly to the breakage of several intermolecular H-
bonds.
On the other hand, the involvement of intermolecular hy-
drogen bonding in the aggregation of compound 1 in water
was nicely confirmed by IR spectroscopy studies.[29] As
shown in Figure 6, upon cooling down a hot solution of com-
tradictory results. As noted above, the solubilization of the
hydrogel formed by compound 1 in water presents a low en-
thalpic component, especially in the temperature range 25–
508C, being zero at 558C. These data at first glance suggest
that the energy associated to the breakage of hydrogen
bonds of the gelator in the solubilization process is almost
negligible. If we assume this statement to be correct, the in-
troduction of hydrogen-bonding groups in the design of hy-
drogelators would seem pointless. Fortunately for this dis-
cussion, the relevance of hydrogen bonding in aqueous envi-
ronment has been studied in detail due to the key role that
this non-covalent interaction plays in protein folding and
stability. As pointed out by Murphy and Habermann,[32] al-
though some authors have proposed that the loss of amide–
amide hydrogen bonds is compensated energetically by the
formation of amide–water hydrogen bonds (resulting in a
zero enthalpy gain for the process or disassembly), now
there is a general agreement that the enthalpy of amide–
amide hydrogen-bonding formation in water is favorable.
For example, the results shown by this authors in the solubi-
lization of crystalline cyclic dipeptides clearly reveal positive
(unfavorable) enthalpies associated to the dissolution
(breakage of intermolecular hydrogen bonds) of the cyclic
peptides in water. The values reported range from 13 to
Figure 6. Variation of the FTIR spectrum of a hot solution of compound
1 (14.5 mm) in water upon cooling down to room temperature. Arrows in-
dicate how the signals change with time.
pound 1 in water, new bands in the C=O region can be ob-
served, which should be ascribed to the formation of inter-
molecular hydrogen bonds as the aggregation proceeds. Sim-
ilar conclusions were obtained in the study of compounds 2
and 3 when their IR spectra were recorded in mixtures of
water and DMSO (these compounds presented very fast ag-
gregation kinetics that precluded the variable temperature
analysis performed for compound 1). Upon increasing the
percentage of water in the water/DMSO mixtures the C=O
stretching shifted to lower wavenumbers, in agreement with
the participation of hydrogen bonding in the aggregation
process (see the Supporting Information).[30] Additionally,
the IR spectrum reveals changes in the CH stretching
region, As can be seen in Figure 7 there is a shift of these
bands to lower wavenumbers as the concentration increases,
which has been ascribed to the packing of aliphatic
chains.[31]
26 kJmolÀ1 and, for example, the cyclic dipeptide
5
(Scheme 1) is reported to present a solubilization enthalpy
of 26 kJmolÀ1. Obviously, these data are in contradiction
with the zero enthalpy change observed for gelator 1 in
water. However, the consideration of the hydrophobic ef-
fects can rationalize these results. The zero enthalpic com-
ponent of the solubilization process of compound 1 in water
at 558C should be seen as the result of the sum of the unfav-
orable enthalpy associated to intermolecular bond dissocia-
tion (mainly hydrogen bonds) and the enthalpy associated
to the hydrophobic effect at that temperature, which must
be favorable. Indeed, it is well reported that the enthalpic
component of the hydrophobic effect is quite variable with
the temperature, being favorable for the dissolution of
apolar substrates into water at low temperatures (for exam-
ple 258C) and becoming unfavorable at higher temperatur-
es.[6b,f] As for the cyclic dipeptide 5, the unfavorable dissolu-
tion enthalpy reported, can be rationalized considering that
the hydrophobic effects associated to the dissolution of this
rather polar molecule are negligible. For example, a quanti-
tative estimation of polarity reveals a tremendous difference
in the polar character of compounds 1 and 5 being the
values of the calculated partition coefficients between water
and octanol (clogP) 1.9 and À1.7, respectively (in this scale
the value of clogP is 2.0 for benzene and À0.7 for metha-
nol).
The fact that the formation of hydrogel 1 is entropy
driven and, at the same time, hydrogen bonding is a driving
force for the aggregation, seem to be to a certain point con-
Further insight in the structural properties of the gel
formed by compound 1 in water was obtained from NMR
studies. The comparison of the 1H NMR spectra of com-
pound 1 in solution (2 mm) and after formation of a gel
(19 mm) in H2O is puzzling. As can be seen in Figure 8, the
spectrum of the solution does not show any NH resonance
but the spectrum corresponding to the gel system reveals an
Figure 7. FTIR spectra of compound 1 in D2O; solution (top, 8 mm) and
gel (bottom, 14.5 mm).
Chem. Eur. J. 2012, 18, 4063 – 4072
ꢃ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
4067

J. F. Miravet, B. Escuder et al.
to this aromatic stacking. Additionally, it is worth to men-
tion the intriguing disappearance of the NH resonances in
the top spectrum of Figure 9. The fact that the NHs reso-
nances of compound 6 can be observed at a concentration
of 1 mm but not at higher concentrations (76 mm) indicates
an acceleration of the exchange process between the hydro-
gen atoms from the amide NHs and water associated to a
concentration increase. A rationale for this concentration
effect is that the pyridine groups catalyze the NH–H2O
proton exchange and that a bimolecular event (collision of
two molecules of 6) is required for the catalysis to be effec-
tive.
All the results shown above can be used advantageously
for the rational design of low molecular weight hydrogela-
tors. Compound 1 is a hydrogelator with moderate efficiency
in terms of minimum gelator concentration at 258C (mgc=
14 mm). Considering that the entropy change associated
with the hydrophobic effect is the driving force for the ag-
gregation at this temperature, it seems reasonable to intro-
duce groups with enhanced hydrophobicity in the molecule
to boost the gelation properties as a result of an improved
favorable entropic term. With this purpose compounds 2–4
were prepared and studied (see Scheme 1). In compound 2
the hydrophobic character is increased by employing a cen-
tral pentamethylenic spacer as compared to compound 1,
which presents a propylenic one. Compound 3 is an isomer
of compound 2, which derives from l-isoleucine instead of
valine and presents a propylenic aliphatic spacer. Finally,
compound 4 presents both a pentamethylenic spacer and de-
rives from the amino acid l-isoleucine. The study of these
compounds permits to evaluate how the introduction of hy-
drophobic moieties in different parts of the molecule
(amino acid side chain vs. aliphatic spacer) affects the hy-
drogelation properties. It was found that compounds 1–4
form hydrogels that were made of fibrillar networks (see the
Supporting Information for electron microscopy images).
The results obtained in the study of the hydrogelation capa-
bilities of the compounds mentioned above are summarized
in Table 2. Remarkably, the rational design of improved hy-
drogelators by means of increasing the hydrophobic charac-
ter of the molecule was very satisfactory and agrees with a
recent report from Adams et al. on the correlation of the
hydrophobic character and the gelation efficiency.[20] On
Figure 8. Fragment of the 1H NMR spectra of compound 1 in solution
(bottom, 2 mm) and gel (top, 19 mm) in H2O. Solvent signal was presatu-
rated.
NH signal corresponding to the central amide unit. This fact
should be ascribed to the exchange of the amide NHs with
the solvent. If this process is fast enough, the NH signals
vanish as a result of the solvent presaturation technique
used to record the spectrum. The results obtained show that
this exchange is slowed down in the gel system, most likely
due to the involvement of the amide in intermolecular hy-
drogen bonding. Therefore, in an indirect way, this result
confirms that hydrogen bonding plays a role in the self-asso-
ciation of this molecule in water. Particularly, the central
NH unit would be involved in the aggregation mechanism,
whereas the terminal one, which is not visible in any spectra,
would be weakly involved, if at all, in intermolecular hydro-
gen bonding.
An interesting difference in the intermolecular interac-
tions found for compound 1 in acetonitrile and water is the
probable presence of aromatic p–p stacking in water. The
1H NMR spectrum of the model compound 6, which is solu-
ble in water in the studied concentration range, reveals a
shielding of the aromatic pyridine units upon increasing the
concentration, which is indicative of p–p stacking interac-
tions (Figure 9). This behavior was not detected in acetoni-
trile, suggesting that the hydrophobic effect may contribute
Table 2. Comparison of the minimum gelator concentration, solubility,
and entropy of the solubilization values for the hydrogels formed by
compounds 1–4.[a]
Compound
mgc [mm]
Solubility (258C) [mm]
DSgelÀsol (558C)[b]
[JmolÀ1 KÀ1
]
1
2
3
4
14 (17)[c]
9
3
1
À35
À46
8
5
ꢀ2
À61
ꢀ0.1[d]
<À77
[a] Estimated errors: mgc=0.5 mm, mgc=minimum gelator concentra-
tion; solubility=5%; DSgelÀsol =1 JmolÀ1 KÀ1. [b] Calculated considering
DH=0 at this temperature. [c] 14 mm: gel formed after 72 h; 17 mm: gel
formed after 24 h. [d] Estimated from the detection limit by 1H NMR
spectroscopy under the conditions of the experiment.
Figure 9. Fragment of the 1H NMR spectra of compound 6 showing the
shielding experienced by the protons of the pyridine unit upon increasing
the concentration (see Figure 8 for labeling).
4068
www.chemeurj.org
ꢃ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 4063 – 4072

Molecular Hydrogels from Bolaform Amino Acid Derivatives
FULL PAPER
going from compound 1 to the most hydrophobic compound
in the series, that is compound 4, a very significant improve-
ment in the gelation capabilities was observed. The mini-
mum gelator concentration required to form a gel was re-
duced from 17 to less than 2 mm. The gelation efficiency in
water was found to be directly related to the solubility, as
shown in previous work in organic solvents.[15]
The results in Table 2 reveal that the solubility in water at
258C experienced a dramatic reduction upon going from ge-
lator 1 (9 mm) to gelator 4 (0.1 mm), which must be associat-
ed to the entropic term of the aggregation free energy that
is much higher in absolute value for the more hydrophobic
compound 4 (DS<À77 JKÀ1 molÀ1) than in the case of the
system formed by compound
1
(DS=À35 JKÀ1 molÀ1).
These results agree with the different water–octanol parti-
tion coefficients (clogP) estimated for compounds 1 and 4,
which are 1.9 and 2.9, respectively. The data from Table 2
corresponding to the isomeric compounds 2 and 3, show
that the introduction of hydrocarbonated fragments in the
amino acid side chain (isoleucine instead of valine residue,
compound 2) results in quite stronger hydrophobic effects
than those obtained by increasing the aliphatic central
spacer (pentamethylenic instead of propylenic central ali-
phatic spacer, compound 3). The relevant difference be-
tween the amino acids valine and isoleucine in terms of hy-
drophobic effects has been pointed out previously in the
context of protein science.[6g]
It is worthy to comment on the solubility versus tempera-
ture profile obtained for compounds 1–3 (compound 4 was
too insoluble for the determination of its solubility by NMR
spectroscopy). It can be seen in Figure 10 that compounds 1
and 2 show related solubility profiles, both presenting a min-
imum value at 558C. In the case of compound 2, the solubil-
ity increase above this temperature is significantly smaller in
absolute terms when compared to that experienced for com-
pound 1, but comparable in relative terms. Having in mind
the exponential relationship between the solubility and the
entropy changes, this result agrees with a similar absolute
change in the entropic term of the aggregation on going
from 55 to 708C in both systems.
Compound 3 deserves a special mention due to its weird
solubility versus temperature profile. As shown in Figure 10,
the solubility of the hydrogelator is almost constant from 25
to 608C (this implies a negligible association enthalpy in this
temperature range) but at higher temperatures a dramatic
decrease of the solubility is observed, making the compound
undetectable by NMR spectroscopy above 708C. This be-
havior is just the opposite of that shown for compound 1,
and instead of the expected solubility increase with the tem-
perature the formation of a very insoluble gel network is ob-
served. Considering the precedents in supramolecular gela-
tion, this result should most likely be ascribed to a structural
rearrangement of the aggregates (formation of a different
polymorph) as reported in some previous cases, including
closely related analogues of compound 3, which were stud-
ied in organic solvents.[33] In the present case, the significant
changes in the hydrophobic effect expected to take place
Figure 10. Solubility variation with the temperature for the hydrogels
*
^
^
formed by compounds 1 ( , 19 mm), 2 ( , 14 mm), and 3 ( , 6 mm).
above 558C could be the driving force for this structural re-
arrangement. A nice experimental proof of the polymorphic
change experienced by the hydrogel formed by compound 3
was obtained from CD spectroscopy (Figure 11). The CD
spectra obtained for hydrogels of compound 3 at 20 and
Figure 11. CD spectra for the hydrogel formed by compound 3 (4 mm) at
~
^
*
different temperatures ( =20, =60, =708C).
Chem. Eur. J. 2012, 18, 4063 – 4072
ꢃ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
4069

J. F. Miravet, B. Escuder et al.
morph II presented the forma-
tion of tapes of approximately
1 mm width (see the Supporting
Information).
An interesting consequence
of the polymorphic change ex-
perience by the hydrogel
formed by compound 3 is its
unusual variable temperature
rheological profile. As shown
in Figure 13, the gel formed by
cooling a hot solution to 208C
(polymorph I) is strengthened
upon heating from 25 to 708C
and at this point an abrupt ten-
dency change is recorded,
which yields a gel with lower
storage module and poor tem-
Figure 12. DSC thermograms of the xerogels prepared from hydrogels formed by compound 3. The gels were
obtained by cooling a hot solution either to 20 (polymorph I, solid line) or 708C (polymorph II, dashed line).
Left: Pictures of the polymorph I and II.
708C showed remarkable differences, for example, the
molar ellipticity at 260 nm being of opposite sign at these
temperature values. This result indicates a significant reor-
ganization of the aggregates that form the fibrillar network.
The positive Cotton effect observed for the samples corre-
sponding to the hydrogel at 20 and 608C agrees with previ-
ous work on aggregation of pyridine derivatives.[34] Further
evidence of this polymorphic transition was provided by dif-
ferential scanning calorimetry (DSC) analysis of xerogels
prepared from a hydrogel of compound 3. In one case, a hy-
drogel was obtained by cooling down the hot solution of 3
in a thermostated bath at 258C (polymorph I) and in the
other case the system was thermostated at 708C (polymor-
ph II). Notice in Figure 12 the different aspect that both gels
present, being one transparent and the other one opaque.
The DSC graphs obtained revealed that the xerogel ob-
tained at 208C presented an exothermic peak at 2208C,
which should be ascribed to a polymorphic transition
(Figure 12).
perature dependence ascribed to the progressive formation
of polymorph II. In fact, when the gel was formed by cool-
ing down a hot solution to 708C, this hydrogel (polymor-
ph II) presented a different rheological profile (Figure 13).
Up to this point, the improvement of gelation capabilities
by using a rational design based on the consideration of hy-
drophobic effects has been shown. Following the line of rea-
soning presented above, it seems feasible to move from en-
tropy- to enthalpy-directed hydrogelation. For this purpose
the hydrophobic effect should be minimized, namely, hydro-
philic gelators should be considered. With this idea in mind,
compound 7, which was reported in the literature as hydro-
gelator was chosen for the study.[35] This compound is quite
more hydrophilic than, for example, compound 1 (clogP=
0.8 and 1.9, respectively). Indeed, the solubility profile ob-
tained for compound 7 in water is significantly different
from that of compound 1 (Figure 14). The solubility of com-
pound 7 shows a very strong, exponential dependence with
the temperature, revealing an enthalpy-driven aggregation.
As a matter of fact, the solubility data can be fitted to a
van’t Hoff analysis, yielding an unfavorable solubilization
enthalpy of 49 kJmolÀ1 and a favorable solubilization entro-
py of 116 JKÀ1 molÀ1. The strong enthalpy component when
compared to that obtained for compound 1 in acetonitrile
Additionally, scanning electron microscopy (SEM) images
of the xerogels were remarkably different. In the case of
polymorph I quite thin fibers were observed, whereas poly-
Figure 13. Variation of the rheological storage modulus (Gꢄ) with the
temperature for the gels formed by compound 3 (6 mm) in water, by
^
cooling down a hot solution to 20 (polymorph I, ) and to 708C (poly-
Figure 14. Solubility variation with temperature for the hydrogels formed
^
^
^
morph II, ). Oscillation frequency=1 Hz and oscillatory stress=1 Pa.
by compounds 1 ( , 19 mm) and 7 ( , 106 mm).
4070
www.chemeurj.org
ꢃ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 4063 – 4072

Molecular Hydrogels from Bolaform Amino Acid Derivatives
FULL PAPER
(38 kJmolÀ1) is difficult to rationalize if an aggregation
based mainly on intermolecular hydrogen bonding is pro-
posed for 7. This compound presents only one amide group
available for this purpose as compared to the four amide
groups present in compound 1. Most likely, aromatic stack-
ing is a major driving force for the aggregation of compound
7, which presents a planar geometry and two conjugated ar-
omatic rings, as suggested in the original study of hydrogel
formation by this molecule.[35]
Obviously, switching from an entropy-driven hydrogel to
and enthalpy-driven one also has consequences on practical
properties of these soft materials such as the rheological
profiles. As shown in Figure 15, the enthalpy-driven hydro-
gel formed by compound 7, presents a dependence of the
storage modulus on the temperature, which could be consid-
ered as usual for molecular gels, and which correlates well
with the solubility pattern depicted above. On the other
hand, the poor temperature dependence of the solubility of
compound 5 results in good mechanical properties (high
storage modulus) in the range of temperatures shown in
Figure 15.
This fact is demonstrated by the study of hydrogelator 7,
whose aggregation is controlled by enthalpy. The low hydro-
phobic character of this compound explains this property. It
is outstanding that moving from an enthalpy-driven to an
entropy-driven aggregation process results in some noticea-
ble changes in the properties of the hydrogels, which can po-
tentially be used for different applications. Firstly, the entro-
py-driven hydrogels studied here, show poor temperature
dependence of their properties below 558C, resulting in low
solubility changes and therefore, almost invariable minimum
gelation concentration values. As a result of this, in differ-
ence with common molecular gels, it is not possible to mod-
ulate the thermal stability with concentration changes in the
range of about 25–808C in the entropic-driven hydrogels,
being the gels either stable up to 858C or not formed. In ad-
dition to this, rheological studies reveal a quite unusual pro-
file of the temperature-dependant viscoelastic properties, re-
sulting in a strengthening of these materials upon heating up
to approximately 558C. Another point of interest is the fact
that hydrogen bonding is shown to play a key role in the ag-
gregation of hydrogelator 1 in water, as demonstrated by IR
spectroscopy. This fact agrees with previous work on the
relevance that intermolecular hydrogen bonding may have
in aqueous environments. All these results seem to contrib-
ute to provide a ground for the rational design of molecular
gelators, a task that has been considered by different au-
thors as a major challenge. The results reported should con-
tribute to achieve a tailored design of molecular gels in
order to prepare materials with the required properties. For
example, enthalpic gels could be useful in controlled release
based on temperature changes, whereas entropic gels would
yield robust materials relatively insensitive to temperature
changes. Noticeably, the identification of entropic or en-
thalpic gels is experimentally simple. For example, solubility
versus temperature studies or, variation of Tgel with temper-
ature can be used for this purpose.
Figure 15. Variation of the rheological storage modulus (Gꢄ) with the
^
temperature for the hydrogels formed by compounds 2 ( , 9.8 mm) and 7
Aside of these considerations, the study of the hydrogel
formed by compound 3 underlines the importance that poly-
morphic transitions may have in soft materials. The reported
structural change above 708C results in a dramatic change
in the solubility and rheological properties.
^
(
, 75 mm). Oscillation frequency=1 Hz and oscillatory stress=1 Pa.
Conclusion
Remarkable differences in the self-assembly process that
takes place in acetonitrile and water have been found for
the ambidextrous gelator 1. Thermodynamically, the aggre-
gation process is found to be enthalpy driven in the organic
solvent but entropy driven in water. Therefore, the hydro-
phobic effect is shown to play a key role in the formation of
the hydrogel by compound 1. Improvement of the gelation
efficiency is easily achieved by increasing the hydrophobic
character of the gelators, by using, for example, isoleucine
derivatives instead of valine ones. Noticeably, regarding the
intermolecular driving forces for aggregation, it is found
that aromatic stacking plays a role in the aggregation of
model compound 6 in water, but this interaction is not de-
tected in acetonitrile. Among the results reported, it should
be highlighted that it is possible to have either enthalpy-
driven or entropy-driven molecular gel formation in water.
Experimental Section
For experimental details, see the Supporting Information.
Acknowledgements
We gratefully acknowledge financial support from the Spanish Ministry
of Science and Innovation (CTQ2009-13961) and Universitat Jaume I
(P1-1B2009-42 and P1-1B2007-11). We are thankful for technical support
from SCIC (Universitat Jaume I).
[1] a) D. F. Evans, H. Wennerstrçm, The Colloidal Domain: Where
Physics, Chemistry, Biology, and Technology Meet, WILEY-VCH,
Chem. Eur. J. 2012, 18, 4063 – 4072
ꢃ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
4071

J. F. Miravet, B. Escuder et al.
Weinheim, 1999; b) I. W. Hamley, Introduction to Soft Matter: Poly-
mers, Colloids, Amphiphiles and Liquid Crystals, Wiley, New York,
2000.
[15] A. R. Hirst, I. A. Coates, T. R. Boucheteau, J. F. Miravet, B. Escud-
er, V. Castelletto, I. W. Hamley, D. K. Smith, J. Am. Chem. Soc.
2008, 130, 9113.
[2] D. A. Uhlenheuer, K. Petkau, L. Brunsveld, Chem. Soc. Rev. 2010,
39, 2817.
[16] M. Raynal, L. Bouteiller, Chem. Commun. 2011, 47, 8271.
[17] J. H. van Esch, Langmuir 2009, 25, 8392.
[3] a) J. M. Zayed, N. Nouvel, U. Rauwald, O. A. Scherman, Chem. Soc.
Rev. 2010, 39, 2806; b) J. H. Ryu, D. J. Hong, M. Lee, Chem.
Commun. 2008, 1043.
[4] E. Krieg, B. Rybtchinski, Chem. Eur. J. 2011, 17, 9016.
[5] a) G. V. Oshovsky, D. N. Reinhoudt, W. Verboom, Angew. Chem.
2007, 119, 2418; Angew. Chem. Int. Ed. 2007, 46, 2366; b) H. J.
Schneider, Angew. Chem. 2009, 121, 3982; Angew. Chem. Int. Ed.
2009, 48, 3924.
[6] a) C. Tanford, Science 1978, 200, 1012; b) R. L. Baldwin, Proc. Natl.
Acad. Sci. USA 1986, 83, 8069; c) P. L. Privalov, S. J. Gill, Pure
Appl. Chem. 1989, 61, 1097; d) K. A. Dill, Science 1990, 250, 297;
e) W. Blokzijl, J. Engberts, Angew. Chem. 1993, 105, 1610; Angew.
Chem. Int. Ed. Engl. 1993, 32, 1545; f) J. A. Schellman, Biophys. J.
1997, 73, 2960; g) P. A. Karplus, Protein Sci. 1997, 6, 1302; h) N. T.
Southall, K. A. Dill, A. D. J. Haymet, J. Phys. Chem. B 2002, 106,
521; i) I. W. Hamley, D. Nutt, G. Brown, J. F. Miravet, B. Escuder, F.
Rodrꢀguez-Llansola, J. Phys. Chem. B. 2010, 114, 940.
[18] See for example: M. Suzuki, H. Saito, K. Hanabusa, Langmuir 2009,
25, 8579.
[19] a) J. Makarevic, M. Jokic, B. Peric, V. Tomisic, B. Kojic-Prodic, M.
Zinic, Chem. Eur. J. 2001, 7, 3328; b) J. Chen, J. W. Kampf, A. J.
McNeil, Langmuir 2010, 26, 13076.
[20] D. J. Adams, L. M. Mullen, M. Berta, L. Chen, W. J. Frith, Soft
Matter 2010, 6, 1971.
[21] a) E. K. Johnson, D. J. Adams, P. J. Cameron, J. Mater. Chem. 2011,
21, 2024; b) A. Ghosh, J. Dey, Langmuir 2009, 25, 8466.
[22] J. F. Miravet, B. Escuder, Chem. Commun. 2005, 5796.
[23] B. Escuder, M. Llusar, J. F. Miravet, J. Org. Chem. 2006, 71, 7747.
[24] a) D. W. Urry, Angew. Chem. 1993, 105, 859; Angew. Chem. Int. Ed.
Engl. 1993, 32, 819; b) D. W. Urry, Chem. Phys. Lett. 2004, 399, 177;
c) J. M. Vinther, S. r. M. Kristensen, J. J. Led, J. Am. Chem. Soc.
2011, 133, 271.
[25] a) M. J. Hwang, J. M. Suh, Y. H. Bae, S. W. Kim, B. Jeong, Biomac-
romolecules 2005, 6, 885; b) Y. S. Dagdas, A. Tombuloglu, A. B. Te-
kinay, A. Dana, M. O. Guler, Soft Matter 2011, 7, 3524.
[26] a) K. Hanabusa, R. Tanaka, M. Suzuki, M. Kimura, H. Shirai, Adv.
Mater. 1997, 9, 1095; b) B. Escuder, S. Martꢀ, J. F. Miravet, Langmuir
2005, 21, 6776.
[27] M. Doi, A. Asano, H. Yoshida, M. Inouguchi, K. Iwanaga, M.
Sasaki, Y. Katsuya, T. Taniguchi, D. Yamamoto, J. Pept. Res. 2005,
66, 181.
[28] ; B. Escuder, J. F. Miravet, J. A. Saez, Org. Biomol. Chem. 2008, 6,
4378.
[29] A. Saha, B. Roy, A. Garai, A. K. Nandi, Langmuir 2009, 25, 8457.
[30] M. Suzuki, M. Yumoto, M. Kimura, H. Shirai, K. Hanabusa, Chem.
Eur. J. 2003, 9, 348.
[31] J. van Esch, F. Schoonbeek, M. de Loos, H. Kooijman, A. L. Spek,
R. M. Kellogg, B. L. Feringa, Chem. Eur. J. 1999, 5, 937.
[32] S. M. Habermann, K. P. Murphy, Protein Sci. 1996, 5, 1229.
[33] a) J. F. Douglas, Langmuir 2009, 25, 8386; b) F. Rodrꢀguez-Llansola,
J. F. Miravet, B. Escuder, Chem. Commun. 2009, 209; c) D. S. Tseko-
va, B. Escuder, J. F. Miravet, Cryst. Growth Des. 2008, 8, 11; d) D. S.
Tsekova, J. A. Saez, B. Escuder, J. F. Miravet, Soft Matter 2009, 5,
3727; e) J. R. Moffat, D. K. Smith, Chem. Commun. 2008, 2248.
[34] P. Duan, X. Zhu, M. Liu, Chem. Commun. 2011, 47, 5569.
[35] D. K. Kumar, J. D. Amilan, P. Dastidar, A. Das, Langmuir 2004, 20,
10413.
[7] P. Ball, Chem. Rev. 2008, 108, 74.
[8] C. J. Li, L. Chen, Chem. Soc. Rev. 2006, 35, 68.
[9] I. Saraogi, A. D. Hamilton, Chem. Soc. Rev. 2009, 38, 1726.
[10] a) R. G. Weiss, P. Terech, Molecular Gels: Materials with Self-Assem-
bled Fibrillar Networks, Springer, Heidelberg, 2005; b) F. Fages,
Low Molecular Mass Gelators, Springer, Heidelberg, 2005; c) P.
Terech, R. G. Weiss, Chem. Rev. 1997, 97, 3133; d) D. J. Abdallah,
R. G. Weiss, Adv. Mater. 2000, 12, 1237; e) J. H. van Esch, B. L. Fer-
inga, Angew. Chem. 2000, 112, 2351; Angew. Chem. Int. Ed. 2000,
39, 2263; f) O. Gronwald, E. Snip, S. Shinkai, Curr. Opin. Colloid In-
terface Sci. 2002, 7, 148; g) A. R. Hirst, D. K. Smith, Chem. Eur. J.
2005, 11, 5496; h) M. O. M. Piepenbrock, G. O. Lloyd, N. Clarke,
J. W. Steed, Chem. Rev. 2010, 110, 1960; i) A. Dawn, T. Shiraki, S.
Haraguchi, S.-i. Tamaru, S. Shinkai, Chem. Asian J. 2011, 6, 266;
j) J. W. Steed, Chem. Commun. 2011, 47, 1379.
[11] a) A. R. Hirst, B. Escuder, J. F. Miravet, D. K. Smith, Angew. Chem.
2008, 120, 8122; Angew. Chem. Int. Ed. 2008, 47, 8002; b) S. Bane-
rjee, R. K. Das, U. Maitra, J. Mater. Chem. 2009, 19, 6649; c) B. Es-
cuder, F. Rodrꢀguez-Llansola, J. F. Miravet, New J. Chem. 2010, 34,
1044.
[12] a) L. A. Estroff and A. D. Hamilton, Chem. Rev. 2004, 104, 1201;
b) M. de Loos, B. L. Feringa, J. H. van Esch, Eur. J. Org. Chem.
2005, 3615.
[13] F. M. Menger, K. L. Caran, J. Am. Chem. Soc. 2000, 122, 11679.
[14] P. Dastidar, Chem. Soc. Rev. 2008, 37, 2699.
Received: October 10, 2011
Published online: February 22, 2012
4072
www.chemeurj.org
ꢃ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 4063 – 4072