Release of bioactive volatiles from supramolecular hydrogels: Influence of reversible acylhydrazone formation on gel stability and volatile compound evaporation

Article abstract of DOI:10.1039/c0ob01139h

In the presence of alkali metal cations, guanosine-5′-hydrazide (1) forms stable supramolecular hydrogels by selective self-assembly into a G-quartet structure. Besides being physically trapped inside the gel structure, biologically active aldehydes or ketones can also reversibly react with the free hydrazide functions at the periphery of the G-quartet to form acylhydrazones. This particularity makes the hydrogels interesting as delivery systems for the slow release of bioactive carbonyl derivatives. Hydrogels formed from 1 were found to be significantly more stable than those obtained from guanosine. Both physical inclusion of bioactive volatiles and reversible hydrazone formation could be demonstrated by indirect methods. Gel stabilities were measured by oscillating disk rheology measurements, which showed that thermodynamic equilibration of the gel is slow and requires several cooling and heating cycles. Furthermore, combining the rheology data with dynamic headspace analysis of fragrance evaporation suggested that reversible hydrazone formation of some carbonyl compounds influences the release of volatiles, whereas the absolute stability of the gel seemed to have no influence on the evaporation rates.

Full text of DOI:10.1039/c0ob01139h

View Article Online / Journal Homepage / Table of Contents for this issue
Organic &
Biomolecular
Chemistry
Dynamic Article Links
Cite this: Org. Biomol. Chem., 2011, 9, 2906
www.rsc.org/obc
PAPER
Release of bioactive volatiles from supramolecular hydrogels: influence of
reversible acylhydrazone formation on gel stability and volatile compound
evaporation†
a
a
b
c
Barbara Buchs (n e´ e Levrand), Wolfgang Fieber, Florence Vigouroux-Elie, Nampally Sreenivasachary,
c
a
Jean-Marie Lehn* and Andreas Herrmann*
Received 8th December 2010, Accepted 1st February 2011
DOI: 10.1039/c0ob01139h
In the presence of alkali metal cations, guanosine-5¢-hydrazide (1) forms stable supramolecular
hydrogels by selective self-assembly into a G-quartet structure. Besides being physically trapped inside
the gel structure, biologically active aldehydes or ketones can also reversibly react with the free
hydrazide functions at the periphery of the G-quartet to form acylhydrazones. This particularity makes
the hydrogels interesting as delivery systems for the slow release of bioactive carbonyl derivatives.
Hydrogels formed from 1 were found to be significantly more stable than those obtained from
guanosine. Both physical inclusion of bioactive volatiles and reversible hydrazone formation could be
demonstrated by indirect methods. Gel stabilities were measured by oscillating disk rheology
measurements, which showed that thermodynamic equilibration of the gel is slow and requires several
cooling and heating cycles. Furthermore, combining the rheology data with dynamic headspace
analysis of fragrance evaporation suggested that reversible hydrazone formation of some carbonyl
compounds influences the release of volatiles, whereas the absolute stability of the gel seemed to have
no influence on the evaporation rates.
Introduction
mixture reversibly react to form non-volatile condensation prod-
ucts in equilibrium with the corresponding unreacted species.
5–8
The generation of dynamic combinatorial libraries based on
reversible covalent bond formation has recently attracted consid-
erable interest for the development of enzyme inhibitors in drug
discovery, the selective preparation of macrocycles by templated
synthesis, the development of catalytic ligands, the design of
stimuli-responsive polymer materials and other systems which
The position of the equilibrium depends on external conditions
such as temperature, pH, or the concentration of the different reac-
tants. Once deposited onto a surface, the non-bound volatiles start
to evaporate and shift the equilibrium towards the hydrolysis of the
condensation products. Particularly interesting in this context is
the formation of acylhydrazones, which comprise both a reversibly
1–4
react by adaptation to external influences. In particular, the
condensation of aldehydes and ketones with amine derivatives
to form imines was identified as a versatile reaction for dynamic
4
formed imine unit and a peptide bond. Dynamic mixtures based
on reversible acylhydrazone formation were found to be highly
efficient to slow down the evaporation of volatile fragrance
3,4
combinatorial chemistry. Similarly, dynamic mixtures composed
of carbonyl compounds and amines or specific amine derivatives
have been identified as efficient delivery systems for the controlled
release of volatile aldehydes and ketones in different practical
5,6
aldehydes or ketones in functional perfumery applications. In
the presence of a hydrazide, headspace concentrations of single
raw materials released from an equilibrated dynamic mixture from
a cotton surface increased by up to two orders of magnitude when
5–8
applications.
6
compared to a standard reference sample without hydrazide.
The concept of using dynamic mixtures for the release of
volatiles is based on the fact that the various constituents of the
As a further step to develop novel delivery systems for
bioactive volatiles, we became interested in investigating the
release of bioactive volatiles from dynamic mixtures within
a
Firmenich SA, Division Recherche et D e´ veloppement, 1 route des Jeunes,
9
ordered supramolecular assemblies. Besides reversibly reacting
B.P. 239, CH-1211 Gen e` ve 8, Switzerland. E-mail: andreas.herrmann@
firmenich.com
with carbonyl compounds to form acylhydrazones, bifunctional
10–13
b
Firmenich SA, Division Parfumerie, 1 route des Jeunes, B.P. 239, CH-1211
guanosine-5¢-hydrazide (1)
also forms stable supramolecular
Gen e` ve 8, Switzerland
hydrogels in the presence of alkali metal cations by selective
c
ISIS, Universit e´ de Strasbourg, 8 all e´ e Gaspard Monge, B.P. 70028, F-67083
self-assembly of a guanosine quartet (G-quartet) structure via
Strasbourg cedex, France. E-mail: lehn@isis.u-strasbg.fr
14,15
intermolecular Hoogsten-type hydrogen bonding
(Scheme 1).
†
Electronic supplementary information (ESI) available: synthesis and
The stability of the G-quartet structure is mainly influenced
spectroscopic data of compounds 1, 4, 5 and 23 and numerical data for
Fig. 8. See DOI: 10.1039/c0ob01139h
+
2+
by the choice of cation; K or Sr seem to form particularly
2
906 | Org. Biomol. Chem., 2011, 9, 2906–2919
This journal is © The Royal Society of Chemistry 2011

View Article Online
Scheme 1 Self-assembly of bifunctional guanosine-5¢-hydrazide (1) to G-quartets in the presence of cations and reversible acylhydrazone formation
between the G-quartet of the hydrogel and an aldehyde or ketone.
+
+
+
stable hydrogels, but (CH
3
)
4
N , NH
4
or Na are also suitable
carried out with relatively polar and almost non-volatile carbonyl
14
10,12
for gel formation. Besides the nature of the cation, the structure
of carbonyl compounds used for reversible hydrazone formation
compounds,
the inclusion of apolar and highly volatile flavours
and fragrances is expected to reduce the water-solubility of the
11
18
also strongly influences the stability of the hydrogel. It was shown
that the addition of stoichiometric amounts of certain aldehydes
in a sodium acetate buffer resulted in highly viscous gels, whereas
others only gave solutions. Furthermore, if a mixture of aldehydes
or ketones was added to a hydrogel of 1 the dynamic mixture
selects the compound forming the most stable gel structure in a
thermoreversible process.
The combination of gelation-driven self assembly with re-
versible covalent bond formation in an aqueous environment
makes 1 an interesting compound to control the release of
bioactive substances, in particular those with aldehyde or ketone
entire system and thus influence the stability of the hydrogels.
Results and discussion
Synthesis and structural characterisation
10
Guanosine-5¢-hydrazide (1) can be prepared from guanosine
(2) in a four-step sequence, or from commercially available
16
1–3
10
ketal 3 in three steps (Scheme 2). Starting from less expen-
sive 2, we protected the free 3¢,4¢-hydroxyl groups with 2,2-
dimethoxypropane in the presence of p-toluenesulfonic acid in
12,17
19
functions.
In such systems the evaporation of volatile aldehydes
acetone to give the corresponding isopropylidene derivative 3.
and ketones is not only expected to be slowed down by reversible
covalent bond formation with the hydrazide function at the
periphery of the G-quartet, but also by physical inclusion of the
compounds inside the gel structure by non-covalent interactions.
In this work we complete our previous investigations on
Oxidation of the free hydroxymethyl group to the corresponding
carboxylic acid 4 was carried out using the 2,2,6,6-tetramethyl-1-
piperidinyloxyl (TEMPO) radical and [bis(acetoxy)iodo]benzene
(BAIB) in aqueous acetonitrile. In this step it was important
to add the solvent to a mixture of all the other compounds
20
10,12
the stability of the hydrogels
in the presence of various
to avoid an excessive formation of CO
then transformed to its corresponding methyl ester with SOCl
in methanol by simultaneous removal of the hydroxyl protecting
2
. Carboxylic acid 4 was
fragrance aldehydes and ketones by rheology measurements, and
compare the release of volatile carbonyl compounds from the
gels by dynamic headspace analysis under realistic conditions
encountered in practical applications, such as hydrogels for air
freshener formulations. In contrast to previous investigations
2
21
group. Ester 5 was obtained in its protonated form as shown by
NMR analysis. Addition of NaOD to a solution of protonated 5
in DMSO-d
6
shifted the signal of neighbouring C(5) from 110.2
This journal is © The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 2906–2919 | 2907

View Article Online
Scheme 2 Preparation of guanosine-5¢-hydrazide (1) and its corresponding acylhydrazone 6.
to 116.4 ppm, with the latter value being the one found for the
of two isomers (ca. 64 : 36) with respect to the amide bond
conformation, with the syn isomer (Scheme 3) being the major
isomer. The structure of the compounds was confirmed by 1D
other guanosine derivatives. Finally, treatment of 5 with hydrazine
10,22
hydrate in ethanol or methanol
afforded hydrazide 1 in good
1
13
yield, allowing its preparation on a 10 g scale.
As a reference compound we also prepared acylhydrazone 6
and 2D homonuclear (COSY, NOESY) and H-, C-heteronuclear
(HSQC, HMBC) NMR experiments in DMSO which showed that
the two species were conformational rather than constitutional
(
Scheme 2) which was obtained by heating a suspension of 1 and
6,23
benzaldehyde in ethanol followed by filtration of the precipitated
product as previously described for the preparation of other
isomers. In the NOESY experiment crosspeaks between equiv-
alent protons of both isomers were observed that had opposite
signs with respect to regular NOE peaks. These are due to a
6
hydrazones. Benzaldehyde derivative 6 was obtained as a mixture
1
Scheme 3 Measured NOEs (plain arrows), HMBC (dotted arrow) and chemical exchange (bold arrow) between the syn and anti isomers of 6 by H-
13
and C-NMR spectroscopy in DMSO.
2
908 | Org. Biomol. Chem., 2011, 9, 2906–2919
This journal is © The Royal Society of Chemistry 2011

View Article Online
conformational exchange between the two molecules that takes
place in the mixing period of the NOESY pulse sequence (800 ms).
The exchange process was found to be slow on the NMR timescale,
as both molecules give rise to separate sets of signals in the
NMR spectrum. In addition, NOE crosspeaks were observed
between the amide proton of one isomer and protons of the other
isomer and vice versa, respectively, demonstrating the presence of
a conformational exchange. The observed NOEs and chemical
exchanges are illustrated in Scheme 3.
Hydrogel formation and visual evaluation of the gel stability
To get a first insight into the influence of the gel stability in the
presence of different volatile carbonyl compounds, we prepared
a series of hydrogels with guanosine-5¢-hydrazide (1) (where both
physical inclusion and covalent bond formation are possible) and
a series of typical fragrance raw materials in a molar ratio of
2
: 1 (15 mM for the hydrogelator and 7.5 mM for the fragrance
material) and compared them to hydrogels formed from guanosine
2) (where physical inclusion but no covalent bond formation is
(
+
possible). As more stable G-quartets are usually obtained with K
rather than with Na as the cation, we decided to use a potassium
+
acetate buffer (0.5 M, pH 6) for hydrogel preparation.
Hydrogels were formed by dissolving (using sonication!)
0
.12 mmol of the hydrogelator (1 or 2) in 7 mL of a potassium
acetate buffer at pH 6. The mixture was heated on a water bath to
◦
ca. 70 C until a homogenous solution was obtained. Then 1 mL
of a 0.06 M solution of the volatile carbonyl compound(s) (e.g.
7–22, Fig. 1) in the acetate buffer was added, the sample was left
to cool to room temperature (r.t.), and allowed to stand overnight
to give a gel containing 15 mM of the hydrogelator and 7.5 mM of
the carbonyl derivative. The volatiles were chosen to cover a broad
range of vapour pressures (volatilities), ranging from 0.06 Pa for
the least volatile of the series (7) to 309 Pa for the most volatile
one (15). Similarly, compounds with different water-solubility
24
(
expressed as the octanol/water partition coefficient logPo/w),‡
varying from 0.41 for the most soluble compound (15) to 4.36 for
the least soluble one (10). To see whether a stable gel was obtained
or whether the compound precipitated, the samples were inverted
and/or analysed visually after cooling to room temperature and,
in view of the targeted practical applications, again after storing
for 5 days, while allowing evaporation of the volatiles. The data
obtained are summarised in Table 1, also showing typical examples
for a precipitate, a strong, or a weak gel. Weak gels can be re-
dissolved by agitation of the sample.
The experiments showed that in all cases investigated so
far the hydrogels formed from guanosine-5¢-hydrazide (1) were
significantly more stable than those obtained from guanosine (2),
where precipitation occurred after storing the samples for 5 days.
Furthermore, according to the choice of the carbonyl compound
added to the hydrogelator, gels of different stability were obtained.
It is interesting to note that the evaporation of the fragrances does
not seem to generally influence the stability of hydrogels formed
Fig. 1 Structures and trivial names of typical volatile carbonyl com-
pounds used as fragrances with their calculated vapour pressures (in Pa)
and logPo/w values (in brackets).‡
was observed) whereas the weak hydrogels of 2 all precipitated
after evaporation of the fragrance for 5 days (Table 1).
ꢀ
R
ꢀR
from 1 (only in the presence of Lilial (10) and Trifernal (11) the
gel stability decreased; in the case of benzylacetone (18) an increase
In some cases, gels were prepared in duplicate or triplicate
and, based on the visual analysis of the gels after cooling to
room temperature and allowing them to stand for 5 days, a
good reproducibility was achieved, with the exception of the gels
prepared in the presence of cinnamaldehyde (8) and (R)-citronellal
‡
Vapour pressures and octanol/water partition coefficients (logPo/w) were
calculated with the PBT Calculator (v. 1.0.0, EPA’s Office of Pollution
Prevention and Toxics and Syracuse Research Corporation) based on the
EPIwin program (US Environmental Protection Agency).
This journal is © The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 2906–2919 | 2909

View Article Online
Table 1 Visual evaluation of hydrogels formed in the presence of fragrance aldehydes or ketones (7–22, Fig. 1) using a molar ratio between the
hydrogelator and the fragrance of 2 : 1. The photographs show the formation of a precipitate of 2 in the presence of citronellal (14, left), of a strong gel of
ꢀ
R
1
in the presence of (-)-menthone (19, right) and of a weak gel of 1 in the presence of Trifernal (11, middle) (after 5 days)
Guanosine-5¢-hydrazide (1)
Guanosine (2)
Aldehyde or ketone
after cooling
after 5 d
after cooling
after 5 d
—
weak gel
weak gel
strong gel
strong gel
strong gel
weak gel
weak gel
weak gel
weak gel
strong gel
strong gel
strong gel
strong gel
strong gel
strong gel
strong gel
strong gel
strong gel
precipitate
weak gel
weak gel
weak gel
weak gel
weak gel
precipitate
weak gel
weak gel
weak gel
precipitate
weak gel
precipitate
precipitate
precipitate
precipitate
precipitate
precipitate
precipitate
precipitate
precipitate
precipitate
precipitate
precipitate
precipitate
precipitate
precipitate
precipitate
precipitate
precipitate
precipitate
precipitate
precipitate
precipitate
Vanillin (7)
Cinnamaldehyde (8)
strong gel
strong gel
strong gel
strong gel
strong gel
weak gel
1
0-Undecanal (9)
ꢀ
R
Lilial (10)
ꢀ
R
Trifernal (11)
,5,5-Trimethylhexanal (12)
3
ꢀ
R
Triplal (13)
weak gel
(
R)-Citronellal (14)
strong gel
strong gel
strong gel
strong gel
weak gel
strong gel
strong gel
strong gel
strong gel
Furfural (15)
Benzaldehyde (16)
4
-Methylacetophenone (17)
Benzylacetone (18)
(
-)-Menthone (19)
a-Damascone (20)
Delphone (21)
ꢀ
R
Hedione (22)
(
14), where either strong or weak gels were obtained for different
Table 2 Amounts of guanosine-5¢-hydrazide (1) and benzaldehyde (16)
1
determined in the water phase (in mol%) by H-NMR spectroscopy with
samples. At the given concentrations, neither the vapour pressure
of the fragrance, nor its hydrophilicity seems to influence the
stability of the hydrogels obtained.
respect to the total amount of compounds in the NMR tube using dioxane
as the internal standard
In the case of 1 the carbonyl compounds can react with the
hydrogelator, being physically trapped inside the gel structure
and/or remaining as free molecules in the solvent. However, in
the case of 2, carbonyl compounds can only be physically trapped
inside the gel structure or remain in the solvent. As a consequence
of the reduced mobility of the gel with respect to the NMR
timescale, the amount of hydrazide 1 or aldehydes and ketones
which are trapped in the hydrogel structure give rise to broad
Amount of Amount of
Total amount Total amount Total amount free 1
free 16
[mol-%]
of 1
of dioxane
of 16
[mol-%]
1
.0 eq.
0.5 eq.
0.5 eq.
0.5 eq.
0.5 eq.
1.0 eq.
1.5 eq.
4.7
4.9
4.3
13.7
8.9
5.6
1.0 eq.
1
.0 eq.
1
10,12
signals in the H-NMR spectrum which cannot be quantified.
This amount can thus only be estimated indirectly by integrating
the signals of the compounds remaining in the water phase
dioxane. The tubes were heated and cooled to room temperature
overnight to form the gels. H-NMR spectra were recorded on the
1
fully solidified samples and the percentage of free hydrogelator (in
mol%) was determined by integrating the proton signal of H(8) on
the guanine group with respect to the internal dioxane reference,
and that of free 16 by integrating the H(3) proton signals of the
aromatic ring. Table 2 lists the amounts of free 1 and free 16
determined in the water phase of the different samples.
(
= “free” compounds), giving rise to sharp peaks which can be
integrated with respect to an internal standard that does not
interact with the hydrogel. In our experiments we chose dioxane
as the internal standard, which was added to the buffer solution to
correspond to 0.5 equivalents with respect to the total amount of
1
used to form the gel. Hydrogels were prepared in an NMR tube
After 5 days at room temperature, the NMR samples were re-
heated until the gel dissolved, and then cooled again to room
temperature. Re-measuring the H-NMR spectra indicated 13.9%,
by adding 100 mL of a 0.5 M deuterated potassium acetate buffer
containing dioxane and different amounts of benzaldehyde (16) to
1
7
00 mL of the hydrogelator in a deuterated potassium acetate buffer
8.3% and 5.6% of free 16, and 4.2%, 3.2% and 3.2% of free
1, respectively, thus showing good reproducibility of the gel
formation.
without dioxane, to give 15 mM solutions of 1 with 0.5, 1.0 or 1.5
molar equivalents of 16, respectively, and 0.5 molar equivalents of
2
910 | Org. Biomol. Chem., 2011, 9, 2906–2919
This journal is © The Royal Society of Chemistry 2011

View Article Online
1
0
In agreement with previous measurements, the data show that
more than 95% of the hydrogelator were incorporated in the gel
structure. The fact that only 5–14% of free 16 remained in the
trapped liquid phase, although it was used in excess with respect
to the hydrogelator in one case, indicates that most of the active
aldehyde or ketone is in fact incorporated into the supramolecular
hydrogel structure. It can either be physically trapped inside the
gel and/or be covalently linked to 1.
To investigate this option and, in particular, to verify the stability
of the hydrazide in basic media, we prepared an aqueous solution
of guanosine-5¢-hydrazide (1) in D
2
O. Addition of two drops
of NaOD to the solution of 1 resulted in rapid degradation
21
and formation of carboxylic acid 23. The structure of 23 was
confirmed by reaction of previously prepared guanosine derivative
25
4 with formic acid (Scheme 2).
Under the same conditions (D
2
O and a few drops of NaOD),
It is interesting to note that an increasing amount of 16 added
to the samples results in a lower percentage of the respective
compounds outside the gel, whereas a more or less constant
amount of hydrazide 1 was detected in the three samples. The
observed decrease of the amount of 16 outside the gel structure
might be the result of its limited solubility in the aqueous buffer.
Nevertheless, due to the fact that an excess of 16 with respect to
hydrogelator 1 was used, the data in Table 2 show that at least a
part of the aldehyde is physically trapped within the gel structure.
However, the measurements did not prove the formation of the
corresponding hydrazone.
As already mentioned above, molecules that are engaged in the
gel formation are not visible in the NMR spectrum, due to reduced
mobility. A possible reason for this might be incomplete averaging
of dipolar couplings or magnetic susceptibility gradients. In
order to elucidate this further we carried out High Resolution
Magic Angle Spinning (HRMAS) NMR spectroscopy, where such
unfavourable interactions can by eliminated by rotation of the
sample at the magic angle. However, with regard to the relative
peak intensities, the spectra are identical to the ones obtained with
standard liquid state NMR. We therefore assume that the extensive
line broadening is caused by slow to intermediate motion on the
NMR time scale of the complex and a subsequent decrease of the
hydrazone 6 was found to be at least temporarily stable. In the
presence of NaOD the conformational exchange between the syn
and anti isomer was accelerated giving rise to a single set of
peaks for the two structures. We thus prepared a gel by dissolving
guanosine-5¢-hydrazide (1) by sonication in a 0.5 M potassium
◦
acetate buffer (pH 6). The sample was heated to 80 C prior
to the addition of a solution containing equimolar amounts of
carbonyl compounds 14–17, dissolved in the buffer, yielding a
15 mM solution of 1 and a 3.3 mM solution for each aldehyde or
ketone. The glass container was closed, left cooling to r.t. overnight
to form the gel which then submitted to four heating and cooling
cycles and left standing at r.t. for 60 h. The gel was dissoved
by addition of a few drops of NaOD and the resulting solution
analysed by high-resolution electrospray mass spectrometry (in the
positive and negative ion mode). No mass peaks corresponding
to either one of the hydrazones possibly formed and no peaks
corresponding to 1 or its degradation product 23 could be detected.
Hydrazone formation in the gels could thus not be proved with
this method. Attempts to follow the hydrazone formation by IR
spectroscopy were also not successful.
Mixtures containing various fragrance aldehydes or ketones
were prepared by adding a total of 1.5 equivalents of three
compounds (0.06 M for each component) to the hydrogelator.
A solution of the carbonyl compound mixture (8 mL) was added
transverse relaxation time T .
2
1
◦
HRMAS H-NMR spectra of the guanosine-5¢-hydrazide gels
show, besides the large peaks corresponding to the residual solvent
signal, dioxane, and the acetate buffer, the expected guanosine
peaks H(8) (8.03 ppm), H(R1) (6.01 ppm), and two other protons
of the sugar ring (4.68 ppm, 4.61 ppm). Our measurements showed
that, according to the visible peaks in the NMR spectrum, the
amount of free guanosine-5¢-hydrazide (1) in a 30 mM sample is
to hydrogelator (1). The sample was heated to 70 C on a water
bath and cooled to room temperature. Individual experiments with
furfural (15), benzaldehyde (16) and 4-methylacetophenone (17)
all gave stable gels. However, addition of a mixture of the three
◦
compounds to 1 gave a precipitate, even at 70 C (presumably due
to reasons of solubility). Filtration of the precipitate, drying under
vacuum, and NMR analysis in DMSO showed that a complex
1
3
1
0.3%. This value increased when the concentration of 1 in the
mixture of compounds was obtained. Comparison of the C-
NMR spectrum of the precipitate with those of 1 and 6 (see
Fig. 2) indicated that both compounds were part of the product
mixture. In particular, the formation of acylhydrazone 6 (syn and
anti isomers) shows that the expected reversible reaction with the
hydrogelator takes place spontaneously after only a short period
of heating.
gel decreased. In a 15 mM sample the amount of free 1 in the gel
structure was found to be 24.9%. Increasing the temperature to
◦
◦
6
0 C gave 88.1% of the free hydrogelator, and at 70 C 94.6% of
free 1 were measured, corresponding to the physical melting and
disruption of the gel structure.
Incorporation experiments were carried out with 1 (15 mM) and
bioactive volatiles (0.5 eq.) undecenal (9) and (R)-citronellal (14),
respectively, in 0.5 M potassium acetate buffer. The NMR spec-
trum of these gels displayed weak signal intensities corresponding
to the free fraction of acylhydrazones.
Decreasing the concentration of 1 to 7.5 mM (and those of
the fragrances to 3.75 mM) generally resulted in considerably less
stable hydrogels. Thus the concentration of the hydrogelators was
maintained at 15 mM in the following experiments.
Precipitation was also observed from mixtures of 1 with (R)-
citronellal (14), benzaldehyde (16), and 4-methylacetophenone
ꢀ
R
(17), or Triplal (13), (R)-citronellal (14) and benzaldehyde (16).
Analysis of the precipitates of these mixtures was found to
be complicated and was thus not further pursued. Since the
preparation of these mixtures was also complicated by solubility
problems of the fragrance molecules in the buffer solution, we
decided to reduce their amount by a factor of two with respect to
the hydrogelator.
As a further option to get insight into the composition of
the gel structure, we envisioned dissolving the hydrogel after
its equilibration by the addition of base. This was expected to
New gels were thus prepared and analysed visually after cooling
to room temperature, and after storing at room temperature
for 5 days as described above. The data obtained are sum-
marised in Table 3. Once again, the hydrogels formed from
“
freeze” equilibration by slowing down the possible hydrolysis of
hydrazones and thus allow NMR analysis of the gel composition.
This journal is © The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 2906–2919 | 2911

View Article Online
Fig. 3 Structure, vapour pressure in Pa and logPo/w value (in brackets)‡
of 2-formylbenzenesulfonic acid sodium salt (24).
we used it as a reference for the preparation of mixtures with a
variety of more volatile and less water soluble fragrance molecules.
Hydrogels of 1 and 2 were thus prepared in the presence of an
equimolar amount of a fragrance aldehyde (15 mM for both the
hydrogelator and the fragrance) and varying amounts of 24 (0,
1
3
Fig. 2 Expansion of the C-NMR spectrum of the precipitate isolated
from a mixture of furfural (15), benzaldehyde (16), 4-methylacetophenone
(
(
0
.125, 0.250 and 0.500 eq., respectively). The gels were prepared
17) and hydrogelator 1. Assigned peaks correspond to 1 (ꢀ) and 6
as described above by dissolving 1 or 2 in a 0.5 M potassium acetate
syn-isomer (ꢁ) and anti-isomer (᭺)).
◦
buffer (pH 6) and heating to 70 C. A solution of the fragrance
aldehyde and a given amount of 24, both dissolved in the buffer,
was added and the mixture allowed cooling to room temperature.
The samples were inverted and/or analysed visually after cooling
to room temperature and after storing for 5 days while allowing
the fragrance to evaporate. The observations are summarised in
Table 4.
guanosine-5¢-hydrazide (1) were significantly more stable than
those obtained from guanosine (2), where precipitation occurred
after storing the samples for 5 days.
To see whether the precipitation observed at higher fragrance
concentration is due to the formation of the relatively hydrophobic
12
hydrazone 6, we prepared a mixed G-quartet gel containing 1 and
in a molar ratio of ca. 3 : 1. The hydrogel formed in the acetate
From this data it is very difficult to estimate the influence of 24
ꢀ
R
6
on the stability of the different gels. In the case of Trifernal (11)
as the fragrance aldehyde, an increasing amount of 24 seemed to
give rise to more stable hydrogels. On the other hand, in the case
buffer at pH 6 after cooling to room temperature was found to
be more or less stable, and stabilised even further after storing
for 5 days. This phenomenon may be due to the hydrolysis of
the hydrazone followed by the slow evaporation of benzaldehyde,
which then increased the water-solubility of the hydrogelator.
Preparation of a gel composed of 2 and benzaldehyde (at 0.2 molar
equivalents) gave a precipitate after cooling to room temperature.
No gel was formed from a mixture consisting of 1 and 6 (ca. 3 : 1)
in the presence of citronellal and acetophenone (each at 0.2 molar
equivalents with respect to 1).
ꢀ
R
of Lilial (10) it seems that the slow evaporation of the aldehyde
also influenced the stability of the gel, as higher gel stability was
observed after 5 days. Under the same conditions guanosine (2)
did not form stable gels at all.
Oscillating disk rheology measurements
Our previous experiments have shown that the structure of
the carbonyl compound added to hydrogelator 1 influences the
stability of the gel. Furthermore, we could demonstrate that
carbonyl compounds are incorporated into the gel structure by
Previous studies have shown that 2-formylbenzenesulfonic acid
sodium salt (24, Fig. 3) stabilises the formation of the hydrogels of
10
1
. As aldehyde 24 is highly water-soluble and not very volatile,
Table 3 Formation of hydrogels in the presence of three fragrance aldehydes or ketones using a molar ratio between the hydrogelator and each fragrance
of 4 : 1
guanosine-5¢-hydrazide (1)
guanosine (2)
Aldehydes or ketones
after 1 d
after 5 d
stable gel
after 1 d
after 5 d
Furfural (15)
Benzaldehyde (16)
weak gel
stable gel
stable gel
stable gel
weak gel
precipitate
weak gel
weak gel
precipitate
4
-Methylacetophenone (17)
(
R)-Citronellal (14)
Benzaldehyde (16)
-Methylacetophenone (17)
stable gel
stable gel
stable gel
precipitate
precipitate
precipitate
4
ꢀ
R
Triplal (13)
R)-Citronellal (14)
Benzaldehyde (16)
(
ꢀ
R
Triplal (13)
R)-Citronellal (14)
-Methylacetophenone (17)
(
4
2
912 | Org. Biomol. Chem., 2011, 9, 2906–2919
This journal is © The Royal Society of Chemistry 2011

View Article Online
Table 4 Visual evaluation of hydrogels formed from 1 or 2 in the presence of fragrance aldehydes 10, 11, 12 or 13 and a different molar ratio of
2
-formylbenzenesulfonic acid sodium salt (24)
eq. of
guanosine-5¢-hydrazide (1)
guanosine (2)
Aldehyde
24
after cooling
after 5 d
after cooling
after 5 d
ꢀ
R
Lilial (10)
—
stable gel
weak gel
weak gel
stable gel
weak gel
weak gel
no gel
stable gel
stable gel
stable gel
stable gel
stable gel
stable gel
stable gel
stable gel
stable gel
stable gel
stable gel
stable gel
stable gel
weak gel
weak gel
weak gel
stable gel
stable gel
stable gel
stable gel
stable gel
stable gel
stable gel
stable gel
stable gel
weak gel
precipitate
precipitate
weak gel
weak gel/precipitate
weak gel/precipitate
precipitate
precipitate
precipitate
weak gel/precipitate
weak gel/precipitate
weak gel/precipitate
precipitate
weak gel
weak gel/precipitate
weak gel/precipitate
precipitate
precipitate
precipitate
precipitate
precipitate
precipitate
precipitate
precipitate
precipitate
precipitate
precipitate
precipitate
precipitate
weak gel/precipitate
precipitate
precipitate
0
0
0
.125
.250
.500
ꢀ
R
Trifernal (11)
—
0
0
0
.125
.250
.500
3
,5,5-Trimethylhexanal (12)
—
0
0
0
.125
.250
.500
ꢀ
R
Triplal (13)
—
0
0
0
.125
.250
.500
physical inclusion as well as by reversible reaction with the
hydrazide function of 1, as suggested by indirect evidence (see
above).
In order to get more reliable, quantitative data for the stability
of the gels, and to see whether particular fragrance aldehydes or
ketones have a stabilising or de-stabilising effect, we decided to
determine the viscosity of the gels by rheology measurements, in
10
analogy to similar experiments carried out previously. Despite
the fact that rotating disk measurements (performed with a
◦
stress controlled rotating cone plate (35 mm, 4 ) at shear rates
-
1 10
varying between 0.38 and 40 s ) were quite reproducible, the
interpretation of the results was found to be difficult, due to the fact
that the gels were easily destroyed when starting the measurements
with the rotating disk. Rheology measurements using a rotating
disk are thus not suitable for characterisation of the stability of the
different hydrogels. We therefore decided to analyse the stability
of the hydrogels using an oscillating disk.
Fig. 4 Determination of the domain of linearity for a hydrogel prepared
of pure 1 (15 mM in a potassium acetate buffer) by rheology measurements
in the dynamic oscillating mode.
Fig. 4 shows that the linearity domain was not followed when
strains aboveca. 0.5%were applied. In the followingmeasurements
the strain applied to the gels was fixed at 0.4%.
gel
In a typical measurement, the solution of the hydrogelator
◦
(
0.12 mmol) was heated to 80 C, prior to the addition of the
carbonyl compounds (0.06 mmol) dissolved in the buffer (or an
equivalent of pure buffer solution), to give a final solution of
To determine the gelation temperature T , measurements were
◦
◦
carried out while cooling the sample from 80 C to 20 C (at
◦
-1
1
5 mM of 1 and 7.5 mM (or 0 mM) of carbonyl compound(s).
1 C min ). From the interception of the curves of G¢ and G¢¢, the
gelation temperature of 1 was found to be about 68 C (15 mM
◦
◦
Measurements were carried out with a cone plate (40 mm, 4 ). The
sample was placed onto the rheometer ground plate at 80 C. To
◦
in a potassium acetate buffer), thus confirming the previously
10
avoid evaporation of the fragrance during the measurement, a film
determined value. The presence of benzaldehyde (16, 7.5 mM, 0.5
ꢀ
R
◦
of Neobee was placed at the border of the rheometer plate. The
eq.) considerably decreased Tgel to about 50 C as shown in Fig. 5.
◦
◦
gel was then rapidly cooled to 20 C and left equilibrating at 20 C
for 5 min. A sweep of strain from 0.1 to 10% was carried out with a
logarithmic variation of deformation and the storage modulus G¢
Furthermore, the presence of 16 results in less stable hydrogels
as compared to the reference without fragrance (lower G¢ and G¢¢
values).
(
measure of a sample’s ability to store energy, also called the elastic
With the possibility to control the temperature during the
rheology measurement, we also investigated the reproducibility
of the gelation by repeating the cooling and heating cycles. With
each cycle, G¢ and G¢¢ increased, until stable values were obtained
(usually after 3 to 5 cycles). Fig. 6 shows the evolution of G¢
and G¢¢ obtained for hydrogels formed from 1, and Fig. 7 those
measured for hydrogels of 2, after three to five consecutive cooling
and heating cycles in the presence and absence of different carbonyl
modulus) and the loss modulus G¢¢ (measure of the sample’s ability
to dissipate energy) were determined. G¢ and G¢¢ were measured
for 30 min at 1 Hz. To operate at a deformation which does not
disrupt the gel structure (and thus not influence the measurement
of G¢ and G¢¢), we first determined the domain of linearity of a
hydrogel prepared from 1 (in a 0.5 M potassium acetate buffer) in
the dynamic oscillating mode (Fig. 4).
This journal is © The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 2906–2919 | 2913

View Article Online
In the case of carbonyl compound mixtures, the most stable
gel was observed after addition of benzaldehyde (16) and 4-
methylacetophenone (17) (G¢ = 1726 Pa), a mixture of an aldehyde
ꢀ
R
and a ketone, the least stable for a mixture of Triplal (13)
and furfural (15) (G¢ = 652 Pa), a mixture of two aldehydes.
Comparison of the storage modules G¢ of single compounds and
compound mixtures shows that there must be, at least in some
cases, a cooperative effect of gel stabilisation or destabilisation.
(
R)-Citronellal (14) and furfural (15) gave rise to relatively weak
hydrogels with 1 (G¢ = 869 and 761 Pa, respectively), whereas the
presence of 13 resulted in relatively stable gels. Equimolar amounts
of 13 and 14 and 1 afforded a considerably more stable hydrogel
(
6
G¢ = 1210 Pa) than the corresponding mixture of 13 and 15 (G¢ =
Fig. 5 Temperature dependent variation of G¢ and G¢¢ for a hydrogel of 1
in the presence (ꢀ, ꢁ) and absence (᭹, ᭺) of benzaldehyde (16, 0.5 eq.).
52 Pa). The data given in Table 5 suggest that furfural (15) has
ꢀR
a tendency to destabilise hydrogels of 1, whereas Triplal (13) or
4
-methylacetophenone (17) seem to have a stabilising effect. This
compounds. An increase of the stability of the gel was observed
for the hydrogel of 1 alone, as well as for 1 in the presence of
volatile or non-volatile carbonyl compounds. This illustrates that
the stabilisation of the gel is a general observation,§ and that, in
these cases, the thermodynamic equilibration process within the
gel structure is relatively slow.
is interesting, as 15 is the most water-soluble compound of the
fragrances tested in the present work. Nevertheless, based on the
given data, the effect is not readily predictable, as a mixture of 13
and 16 affords a relatively weak gel (G¢ = 681 Pa) as compared
to the respective individual carbonyl compounds (G¢ = 1365 and
2
051 Pa, respectively).
The values for G¢ and G¢¢ listed in Table 5 were recorded after
several cooling and heating cycles. The data show that the stability
of the hydrogels varies over a broad range of values. In the case
of hydrazide 1, the most stable gel was obtained in the absence of
any carbonyl compound (G¢ = 3126 Pa), the least stable gel in the
presence of furfural (15, G¢ = 761 Pa) if the addition of a single
carbonyl compound is considered. It is interesting to note that,
in the present case, ketones on average give rise to more stable
hydrogels than aldehydes.
It was already shown by visual analysis of the gels that guanosine
2) forms less stable hydrogels than the corresponding hydrazide
and that the stability of the gel can be influenced by addition of
different carbonyl compounds or compound mixtures. In contrast
to hydrazide 1, guanosine (2) itself did not form the most stable
hydrogel of the series.
(
The fact that the stability of the hydrogels prepared from 1 and
can be influenced by addition of one or several other compounds
2
indicates that the gel stability is at least in part influenced by the
more or less favourable inclusion/reaction of these compounds
into/with the gel structure. However, based on the present data,
the effect of reversible covalent bond formation on the gel stability
cannot readily be estimated.
§
Please note that the data listed in Table 1, 3 and 4 were recorded without
repetitive heating and cooling cycles, which might explain the differences
of gel stabilities observed in some cases.
Table 5 Data of G¢ and G¢¢ measured for hydrogels prepared from 1 and 2 in the presence or absence of different fragrance aldehydes or ketones after
3
–5 cooling and heating cycles by rheology in the dynamic oscillating mode (values taken after ca. 1200 s, n.d. = not determined)
Guanosine-5¢-hydrazide (1) 15 mM
G¢ [Pa]
Guanosine (2) 15 mM
Aldehyde or ketone 7.5 mM
—
G¢¢ [Pa]
G¢ [Pa]
G¢¢ [Pa]
3126
945
1342
2051
869
145.1
50.1
53.6
97.8
48.1
32.9
67.2
81.2
60.0
64.3
65.3
29.0
29.6
74.7
40.3
45.8
55.8
73.8
61.0
36
97
136
22
234
173
68
70
45
9
310
34
1.1
2.6
3.7
1.0
5.7
3.7
2.1
2.2
1.6
0.6
5.9
1.1
1.5
2.5
11.0
2.7
2.5
n.d.
n.d.
ꢀ
R
Trifernal (11)
3
,5,5-Trimethylhexanal (12)
ꢀ
R
Triplal (13)
(
R)-Citronellal (14)
Furfural (15)
761
Benzaldehyde (16)
1365
1713
1433
1541
1210
652
681
1363
947
4
-Methylacetophenone (17)
Benzylacetone (18)
(
-)-Menthone (19)
1
1
1
1
1
1
1
1
2
3 + 14
3 + 15
3 + 16
4 + 15
4 + 17
5 + 16
5 + 17
6 + 17
4
20
116
327
106
136
n.d.
n.d.
690
1053
1726
1715
2
914 | Org. Biomol. Chem., 2011, 9, 2906–2919
This journal is © The Royal Society of Chemistry 2011

View Article Online
Fig. 6 Increase of G¢ (full symbols) and G¢¢ (empty symbols) for hydrogels of 1 (a) and 1 in the presence of 0.5 eq. of 24 (b), menthone (19, c),
ꢀ
R
(
R)-citronellal (14, d), Triplal (13, e) or an equimolar mixture of 13 and 14 (0.25 eq. each, f) after several consecutive cooling and heating cycles.
Dynamic headspace analysis
given carbonyl compound from the gel in the presence of one of
the other three volatiles are illustrated in Fig. 8 (numerical data to
Fig. 8, including standard deviations are given in the ESI†).
The hydrogels were prepared as described above. The hydroge-
As a further step, and in view of practical applications of the
hydrogels as air freshener formulations, we investigated the release
of an equimolar mixture of two volatile aldehydes and ketones
◦
lator (1) was mixed with the carbonyl compounds at 80 C, then
26
from the hydrogels by dynamic headspace analysis, which allows
monitoring the evaporation of the volatile carbonyl compounds
without destroying the hydrogel. Headspace analysis is a fast and
convenient tool for the analysis of mixtures of volatiles as long
as the individual compounds can be separated by GC. In view
of possible practical applications as dynamic fragrance delivery
systems, we were essentially interested to follow the evaporation of
several fragrances at the same time and in particular to see whether
or not the presence of other carbonyl compounds influences
the evaporation of a given aldehyde or ketone from a hydrogel
formed from 1. Four carbonyl compounds ((R)-citronellal (14),
furfural (15), benzaldehyde (16), and 4-methylacetophenone (17))
spanning a wide range of different vapour pressures (from 309 to
the glass vials were closed and left to cool to room temperature
overnight. The samples were then submitted to four heating
and cooling cycles and left standing at room temperature for
60 h before being opened, to allow evaporation of the fragrance.
Headspace samples were taken after 1, 2, 4, 7, 9 and 15 days and
the experiment was repeated three times. The data points of the
measured headspace concentrations (in Fig. 8) were connected
with a line to illustrate the continuity of fragrance evaporation
over the entire time frame.
In view of the targeted practical application in perfumery, the
headspace measurements were carried out under ambient con-
ditions without controlling the temperature (room temperature)
or ambient humidity during fragrance evaporation. The lack of
control of these external parameters might explain the relatively
large standard deviations observed in some cases. Especially in the
1
1 Pa, see Fig. 1) were thus chosen to be investigated pairwise.
The headspace concentrations measured for the evaporation of a
This journal is © The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 2906–2919 | 2915

View Article Online
ꢀ
R
Fig. 7 Increase of G¢ (full symbols) and G¢¢ (empty symbols) for hydrogels of 2 (a) and 2 in the presence of 0.5 eq. of (R)-citronellal (14, b), Triplal (13,
c), or an equimolar mixture of 13 and 14 (0.25 eq. each, d) after several consecutive cooling and heating cycles.
Fig. 8 Headspace concentrations of a given carbonyl compound in the presence of (R)-citronellal (14, ꢀ), furfural (15, ᭜), benzaldehyde (16, ᭹) or
4
-methylacetophenone (17, ᭡) evaporated from a hydrogel of 1 (average values of three measurements, numerical data are given in the ESI†).
case of the highly volatile furfural large standard deviations were
measured, in particular at the beginning of the measurements.
Considering the three individual measurements for each mix-
ture, it was found that the headspace concentrations for ben-
zaldehyde and citronellal are almost identical within the same
measurement, independent of the nature of the second carbonyl
compound in the mixture. However, in the case of furfural and
4-methylacetophenone, significantly different headspace concen-
trations were recorded in the presence of another volatile carbonyl
compound in at least one of the three measurements.
2
916 | Org. Biomol. Chem., 2011, 9, 2906–2919
This journal is © The Royal Society of Chemistry 2011

View Article Online
The data illustrated in Fig. 8 allow several conclusions. First
of all, the evaporation of a given volatile carbonyl compound
from the hydrogel is not always influenced by the presence
of another compound. Whereas in the case of furfural and
The fact that a detailed analysis of the gel composition was
not possible, limits the current understanding of the different
phenomena involved in the inclusion and release process. The
effects resulting from a combination of physical inclusion and
reversible covalent bonding of bioactive compounds is thus based
on indirect evidence only. This is particularly true in the case of
more complex mixtures, where a full understanding of the different
phenomena is so far not possible.
4
-methylacetophenone, different headspace concentrations were
measured in the presence of different other carbonyl derivatives,
the evaporation of benzaldehyde and citronellal were found
to be almost unaffected by the presence of other compounds.
Furthermore, when considering the rheology measurements sum-
marised in Table 5, it seems that the stability of the gel alone
Experimental
(
as expressed by G¢) does not significantly influence the rates
of fragrance evaporation. If the stability of the gel influenced
fragrance evaporation, an increased slow release effect would
be expected with increasing gel stability. Benzaldehyde (16) was
released from 1 at similar rates in the presence of either 4-
methylacetophenone (17) (stable gel, G¢ = 1762 Pa) or furfural
General
General aspects and the instrumentation used are described in
the ESI.† Compounds 1, 4, 5 and 23 were prepared as previously
described in the literature (see ESI†).
10,20,21,25
(
15) (less stable gel, G¢ = 690 Pa). This suggests that the lower
headspace concentrations measured, e.g. for the release of 17 in the
presence of 16 as compared to those determined in the presence of
Synthesis of (2S,3S,4R,5R)-5-(2-amino-6-oxo-1,6-dihydro-9H-
purin-9-yl)-3,4-dihydroxytetrahydro-2-furancarboxylic acid
benzylidene hydrazide (6)
1
4, might be due to different compositions of the equilibria set-up
between the different carbonyl compounds and the hydrogelator
1). This may be due to a reversible reaction of the ketone with
(
A suspension of 1 (300 mg, 0.96 mmol) and benzaldehyde (153 mg,
1.45 mmol) in ethanol (10 mL) was refluxed for 6 h. After cooling
to room temperature (r.t.) the mixture was filtered to give 250 mg
(74%) of a grey solid as a mixture of two isomers with respect
to the amide bond conformation (syn/anti ca. 64 : 36). UV/Vis
the hydrogelator upon (partial) formation of the corresponding
hydrazone, thus supporting previous results suggesting reversible
hydrazone formation within the hydrogel structure.
(
1
ethanol): l (e) 302 (sh, 9800), 275 (23000), 258 (22300), 224 (sh,
2600), 218 (sh, 16800), 203 nm (28400). IR (neat): n˜ max 3310 m
Conclusions
Reversible covalent bond formation combined with the generation
of supramolecular assemblies, in which the active molecules also
get physically trapped inside the supramolecular structure, are
interesting as delivery systems for bioactive compounds. Following
the successful release of fragrance aldehydes and ketones by re-
versible acylhydrazone formation in aqueous media, we have now
investigated the potential of bifunctional guanosine-5¢-hydrazide
(br.), 3189 m (br.), 3102 m, 2927 m (br.), 2760 w, 1731 w, 1672 s,
1650 m, 1635 s, 1507 s, 1571 s, 1545 s, 1489 m (br.), 1448 w, 1406 m,
1386 w, 1365 m (br.), 1311 m, 1287 w, 1266 w, 1229 w, 1204 w, 1171
m, 1112 m, 1089 m, 1081 m, 1058 s, 1023 m, 979 w, 956 m, 910 w,
891 w, 876 w, 866 w, 832 m, 792 m, 776 m, 753 m, 733 w, 689 s, 682 s,
-
1
1
635 s, 622 w, 607 s cm . H-NMR (400 MHz, DMSO-d , syn): d
6
11.69 (s, 1 H); 10.88 (s, 1 H); 8.35 (s, 1 H); 8.28 (s, 1 H); 7.77–7.66
(m, 2 H); 7.54–7.39 (m, 3 H); 6.65 (br. s, 2 H); 5.88 (d, J = 6.1, 1 H);
(
1) as a delivery system for volatile carbonyl compounds from
hydrogels.
5.67 (br. s, 2 H); 4.61 (dd, J = 6.1, 1.0, 1 H); 4.47 (d, J = 3.1, 1 H);
1
Previous and present work has shown that bioactive volatile
carbonyl derivatives are both physically trapped in the hydrogel
structure and, at least partially, covalently linked to the hydrogela-
tor by reversible hydrazone formation. Hydrogels of 1 were found
to be generally much more stable than hydrogels of guanosine (2),
where bioactive compounds can only be physically enclosed in the
gel structure. Stable hydrogels of 1 are formed with fragrances
having different vapour pressures and water solubilities. Both
parameters do not seem to directly influence the gel stability, but of
course limited water solubility also limits the amount of fragrance
that can be added to the gel.
Oscillating disk rheology measurements after several cooling
and heating cycles showed that the gel stability increases after
several cycles, indicating that the thermodynamic equilibration
of the gel is a rather slow process. Dynamic headspace analysis
of mixtures showed that the absolute stability of the hydrogel
is not decisive for the evaporation of the fragrances, which
means that the physical inclusion is not sufficient to achieve an
increased long-lastingness of fragrance evaporation. However, the
measurements suggest that equilibria resulting from reversible
hydrazone formation are responsible for the different evaporation
profiles of some of the compounds.
4.31 ppm (dd, J = 3.1, 1.0, 1 H). H-NMR (400 MHz, DMSO-d
6
,
anti): d 11.72 (s, 1 H); 10.81 (s, 1 H); 8.39 (s, 1 H); 8.05 (s, 1 H);
7.77–7.66 (m, 2 H); 7.54–7.39 (m, 3 H); 6.65 (br. s, 2 H); 5.96 (d,
J = 7.2, 1 H); 5.67 (br. s, 2 H); 5.34 (m, 1 H); 4.43 (dd, J = 4.1, 3.1,
1
3
1 H); 4.26 ppm (m, 1 H). C-NMR (100.6 MHz, DMSO-d , syn):
6
d 166.45 (s); 156.69 (s); 154.42 (s); 151.50 (s); 149.50 (d); 136.24
(d); 134.38 (s); 130.70 (d); 129.20 (d); 127.60 (d); 116.15 (s); 87.34
1
3
(d); 83.24 (d); 73.73 (d); 73.41 ppm (d). C-NMR (100.6 MHz,
DMSO-d , anti): d 171.51 (s); 156.72 (s); 154.42 (s); 152.03 (s);
6
144.82 (d); 135.74 (d); 134.26 (s); 130.45 (d); 129.17 (d); 127.41
(d); 115.58 (s); 86.22 (d); 81.40 (d); 75.21 (d); 73.65 ppm (d). MS
+
+
(ESI): m/z 401 [M+2] , 400 [M+H] . HRMS (ESI): m/z calc. for
+
+
+
C
17
H
18
N
7
O : 400.1364 ([M+H] ), found: 400.1258 ([M+H] ).
5
Preparation of buffer solutions
A 0.5 M potassium acetate buffer (pH 6) was prepared from
potassium acetate (46.34 g), glacial acetic acid (1.65 g) and
demineralised water (969.55 g, filled up to 1000 mL). For a buffer of
a similar composition, a pH value of 5.95 (± 0.045) was measured
◦
at 25.0 C (± 0.21) (Mettler-Toledo MP220 with an InLab 410-
Ag/AgCl glass electrode) after two-point calibration.
This journal is © The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 2906–2919 | 2917

View Article Online
Similarly, 0.5 M buffer solutions in D
dissolving potassium acetate (1.1544 g, 11.77 mmol) and glacial
acetic acid (0.0397 g, 0.66 mmol) in D O (26.79 g, 25 mL), or
by dissolving potassium acetate (2.3056 g), glacial acetic acid
0.0794 g), and dioxane (0.266 g, serving as internal standard for
NMR measurements) in D O (53.49 g, 50 mL).
2
O (pD 6) were obtained by
(33.8 mg, 0.12 mmol) instead of 1 was carried out under the same
conditions.
2
At higher fragrance concentrations the mixtures precipitated. A
total of 8 mL of a solution containing 0.075 mol of furfural (15),
benzaldehyde (16), and 4-methylacetophenone (17) in 10 mL of
the above described 0.5 M potassium acetate buffer were added
(
2
◦
to 0.0374 g of 1 at 70 C. Filtration of the precipitate, and
drying under vacuum afforded 0.04 g of a grey solid containing
compounds 1 and 6 (syn and anti isomers).
Mixtures containing a fragrance aldehyde and a variable
General procedure for the preparation of hydrogels for visual
evaluation
amount of 24 were prepared by dissolving guanosine-5¢-hydrazide
In a glass vial, guanosine-5¢-hydrazide (1, 37.3 mg, 0.12 mmol) was
dissolved using sonication in 7 mL of the above described 0.5 M
potassium acetate buffer solution at pH 6. The vial was heated
on a water bath (at ca. 70 C) until the hydrazide completely
dissolved, then 1 mL of the corresponding aldehyde or ketone
(
1, 37.3 mg, 0.12 mmol) using sonication in 7 mL of the above
described 0.5 M potassium acetate buffer solution. The vial was
◦
heated on a water bath (at ca. 70 C) until the hydrazide completely
◦
dissolved, then 1 mL of a fragrance aldehyde (0.12 M) and 24
(
0 mg, 29.6 mg = 0.015 M, 59.2 mg = 0.030 M or 118.8 mg = 0.06
(
0.06 M) dissolved in the potassium acetate buffer was added,
M) dissolved in the potassium acetate buffer was added, yielding
a concentration of 1 and the fragrance aldehyde at 15 mM each
and variable concentrations of 24. The samples were left cooling
to r.t. and analysed visually.
yielding a 15 mM solution of 1 and a 7.5 mM solution of aldehyde
or ketone. The sample was left cooling to r.t. In a comparison
experiment guanosine (2, 33.8 mg, 0.12 mmol) was used instead
of 1. To see whether a gel was obtained or whether the compound
precipitated, the samples were inverted and/or analysed visually
after cooling to r.t. and after storing at r.t. for 5 days.
Oscillating disk rheology measurements
As described above, guanosine-5¢-hydrazide (1, 37.3 mg,
0
.12 mmol) was dissolved using sonication in 7 mL of a 0.5 M
General procedure for the determination of the amounts of free
fragrances and of fragrances included into the hydrogel
potassium acetate buffer solution at pH 6. The sample was heated
on a water bath (at ca. 80 C) until the hydrazide completely
◦
To determine the amount of free hydrogelator and free benzalde-
hyde within the gel structures, guanosine-5¢-hydrazide (1, 37.1 mg,
dissolved, then 1 mL of a solution of fragrance aldehyde or ketone
(0.06 M) in the potassium acetate buffer (or 1 mL of the buffer
solution as a reference) was added, yielding a concentration of 1
at 15 mM and the (total) fragrance aldehyde or ketone at 7.5 mM
(or 0 mM, reference) each. Alternatively, if two different fragrance
compounds were added, 0.5 mL of each of the 0.06 M solutions
were used. The sample was placed on the pre-heated rheometer (at
80 C) and cooled to 20 C for the measurements. In a comparison
experiment guanosine (2, 33.9 mg, 0.12 mmol) was used instead
of 1.
0
.12 mmol) was dissolved in 7 mL of a 0.5 M potassium acetate
buffer in D O (pD 6, see above). Then different amounts of
benzaldehyde (16, 31.6 mg, 0.30 mmol; 62.9 mg, 0.59 mmol or
4.7 mg, 0.89 mmol) were dissolved in 5 mL of a 0.5 M potassium
acetate buffer in D O containing dioxane (60.4 mM) as internal
standard. For the measurement, 700 mL of the buffer containing
2
9
2
◦
◦
1
1
and 100 mL of either one of the buffer solutions containing
6, respectively, were added to a NMR tube together with some
sodium 3-trimethylsilyl-tetradeuteriopropionate (as internal lock),
yielding a 15 mM solution of 1 with 0.5, 1.0 or 1.5 molar
equivalents of 16, and 0.5 molar equivalents of dioxane. The tubes
Rheology measurements were carried out in the dynamic oscil-
lation mode on a TA instruments AR 1500 rheometer equipped
◦
with a Peltier element and a cone plate (40 mm, 4 ). The sample
◦
◦
◦
were heated on a water bath (ca. 70 C) until a clear solution was
was placed on the rheometer at 80 C, rapidly cooled to 20 C and
◦
obtained and then left cooling to room temperature overnight.
left equilibrating at 20 C for 5 min. To avoid evaporation of the
volatiles during the measurements, a film of Neobee was placed
1
ꢀR
The H-NMR spectra (500 MHz) were recorded on the fully
solidified samples. The percentage of free hydrogelator (in mol-
between the border of the rheometer ground plate and the cone
plate. G¢ and G¢¢ were measured over 30 min at 1 Hz and a strain of
0.4%. The heating and cooling cycles were repeated (3 to 5 times)
until stable values were obtained for G¢ and G¢¢. Values for G¢ and
G¢¢ were taken after 1200 s.
%
) was determined by integrating the proton signal of H(8) on
the guanine group (8.0 ppm, 1 H) with respect to the internal
dioxane reference (3.8 ppm, 8 H), and that of free benzaldehyde by
integrating the H(3) proton signals of the aromatic ring (7.9 ppm,
2
H).
To investigate the formation of gels with mixtures of fragrance
Dynamic headspace sampling
aldehydes and/or ketones 0.075 mol of three different aldehydes
or ketones, respectively, were dissolved in 20 mL of the above
described 0.5 M potassium acetate buffer (pH 6). In a glass vial, 1
37.1 mg, 0.12 mmol) was dissolved using sonication in 8 mL of the
.5 M potassium acetate buffer solution containing the aldehydes,
yielding a 15 mM solution of 1 and a 3.75 mM solution for each
aldehyde or ketone. To see whether a gel was obtained or whether
the compound precipitated, the sample was heated and cooled
to r.t. as described above and a comparison experiment using 2
As described above, guanosine-5¢-hydrazide (1, 37.3 mg,
0.12 mmol) was dissolved by sonication in 7 mL of a 0.5 M
potassium acetate buffer (pH 6). The sample was heated to 80 C
prior to the addition of 0.5 mL of two of the fragrance solutions
(0.60 mmol) containing (R)-citronellal (14, 92.1 mg), furfural (15,
57.8 mg), benzaldehyde (16, 63.6 mg) or 4-methylacetophenone
(17, 80.5 mg) dissolved in 10 mL of buffer, yielding a 15 mM
solution of 1 and a 3.3 mM solution for each aldehyde or ketone
◦
(
0
2
918 | Org. Biomol. Chem., 2011, 9, 2906–2919
This journal is © The Royal Society of Chemistry 2011

View Article Online
(
corresponding to a total fragrance content of 7.5 mM). The glass
5 (a) A. Herrmann, Angew. Chem., Int. Ed., 2007, 46, 5836–5863; (b) A.
Herrmann, in “The Chemistry and Biology of Volatiles”, A. Herrmann,
container was closed (to avoid evaporation of the fragrances) and
left cooling to r.t. overnight to form the gels. The gels were then
submitted to four heating and cooling cycles and left standing at r.t.
for 60 h. The glass containers were then opened, and the headspace
measured after 1, 2, 4, 7, 9 and 15 d, by placing the samples
inside a headspace sampling cell (ca. 650 mL), respectively, and
exposed to a constant air flow of 200 mL min . The air was
filtered through active charcoal and aspirated through a saturated
solution of NaCl. After equilibration for 15 min, the volatiles were
adsorbed during 15 min (after 1 and 2 d), 30 min (after 5, 7 and 9
d) or 40 min (after 15 d) on a clean Tenax cartridge, respectively.
The cartridges were desorbed and analysed as described in the
ESI† (General). Headspace concentrations were obtained by
external standard calibrations of the corresponding fragrance
aldehydes and ketones using ethanol solutions of five different
concentrations. 0.1 mL of each calibration solution was injected
onto Tenax cartridges, which were immediately desorbed under
the same conditions as those resulting from the headspace
sampling. All samples were prepared and analysed in triplicate.
(
Ed.), John Wiley & Sons, Chichester, 2010, pp. 333-362.
6
(a) B. Levrand, Y. Ruff, J.-M. Lehn and A. Herrmann, Chem. Commun.,
2006, 2965–2967; (b) B. Levrand, W. Fieber, J.-M. Lehn and A.
Herrmann, Helv. Chim. Acta, 2007, 90, 2281–2314.
7
(a) G. Godin, B. Levrand, A. Trachsel, J.-M. Lehn and A. Herrmann,
Chem. Commun., 2010, 46, 3125–3127; (b) B. Buchs, G. Godin,
A. Trachsel, J.-Y. de Saint Laumer, J.-M. Lehn and A. Herrmann,
Eur. J. Org. Chem., 2011, 681–695.
-
1
8
9
(a) N. Giuseppone and J.-M. Lehn, Angew. Chem., Int. Ed., 2006, 45,
4
619–4624; (b) A. Herrmann, N. Giuseppone and J.-M. Lehn, Chem.–
Eur. J., 2009, 15, 117–124.
A. G. Orrillo and R. L. E. Furlan, J. Org. Chem., 2010, 75, 211–214.
ꢀ
R
10 N. Sreenivasachary and J.-M. Lehn, Proc. Natl. Acad. Sci. U. S. A.,
2
005, 102, 5938–5943.
1
1 E. Buhler, N. Sreenivasachary, S.-J. Candau and J.-M. Lehn, J. Am.
Chem. Soc., 2007, 129, 10058–10059.
6
12 N. Sreenivasachary and J.-M. Lehn, Chem.–Asian J., 2008, 3, 134–139.
13 (a) V. Setni
cˇ ka, M. Urbanov a´ , K. Volka, S. Nampally and J.-M. Lehn,
Chem.–Eur. J., 2006, 12, 8735–8743; (b) V. Setni cˇ ka, J. Nov y´ , S. B o¨ hm,
N. Sreenivasachary, M. Urbanov a´ and K. Volka, Langmuir, 2008, 24,
ꢀ
R
7
520–7527.
14 (a) W. Guschlbauer, J.-F. Chantot and D. Thiele, J. Biomol. Struct. Dyn.,
990, 8, 491–511; (b) S. Pieraccini, T. Giorgi, G. Gottarelli, S. Masiero
1
and G. P. Spada, Mol. Cryst. Liq. Cryst., 2003, 398, 57–73; (c) G. P.
Spada and G. Gottarelli, Synlett, 2004, 596–602; (d) J. T. Davis, Angew.
Chem., Int. Ed., 2004, 43, 668–698; (e) J. T. Davis and G. P. Spada,
Chem. Soc. Rev., 2007, 36, 296–313.
Acknowledgements
1
1
5 See also: M. S. Kaucher, W. A. Harrell Jr. and J. T. Davis, in
“Quadruplex Nucleic Acids”, S. Neidle and S. Balasubramanian, (Eds.),
The Royal Society of Chemistry, Cambridge, 2006, pp. 253-296.
6 For a selection of recent reviews on supramolecular hydrogels, see for
example: (a) M. de Loos, B. L. Feringa and J. H. van Esch, Eur. J. Org.
Chem., 2005, 3615–3631; (b) N. M. Sangeetha and U. Maitra, Chem.
Soc. Rev., 2005, 34, 821–836; (c) H. S. Choi and N. Yui, Prog. Polym.
Sci., 2006, 31, 121–144; (d) S. Chaterji, I. K. Kwon and K. Park, Prog.
Polym. Sci., 2007, 32, 1083–1122; (e) Z. Yang and B. Xu, J. Mater.
Chem., 2007, 17, 2385–2393; (f) A. R. Hirst, B. Escuder, J. F. Miravet
and D. K. Smith, Angew. Chem., Int. Ed., 2008, 47, 8002–8018; (g) P.
Dastidar, Chem. Soc. Rev., 2008, 37, 2699–2715; (h) J. Li and X. J. Loh,
Adv. Drug Delivery Rev., 2008, 60, 1000–1017; (i) H.-J. Yoon and W.-D.
Jang, J. Mater. Chem., 2010, 20, 211–222.
We are grateful to Maude Gaillard for her assistance in the prepa-
ration of the compounds and to Daniel Grenno for measuring the
ESI mass spectra. We thank Dr Roger Snowden for constructive
comments on the manuscript and Drs Val e´ ry Normand and Otto
Gr a¨ ther for discussions concerning the rheology measurements.
Notes and references
1
(a) “Dynamic Combinatorial Chemistry”, J. N. H. Reek, and S. Otto,
Eds.), Wiley-VCH, Weinheim, 2010; (b) “Dynamic Combinatorial
(
Chemistry in Drug Discovery, Bioorganic Chemistry, and Materials
Science”, B. L. Miller, (Ed.), John Wiley & Sons, Hoboken, 2010.
(a) A. V. Eliseev and J.-M. Lehn, Curr. Top. Microbiol. Immunol., 1999,
17 Parts of this publication are the subject of a patent application: J.-
M. Lehn, N. Sreenivasachary and A. Herrmann, (to Firmenich SA,
Universit e´ Louis Pasteur and CNRS), WO 2006/100647, 2006; Chem.
Abstr., 2006, 145, 383495.
2
2
43, 159–172; (b) B. Linton and A. D. Hamilton, Curr. Opin. Chem.
Biol., 1999, 3, 307–312; (c) G. R. L. Cousins, S.-A. Poulsen and J. K. M.
Sanders, Curr. Opin. Chem. Biol., 2000, 4, 270–279; (d) J.-M. Lehn and
A. V. Eliseev, Science, 2001, 291, 2331–2332; (e) O. Ramstr o¨ m and J.-M.
Lehn, Nat. Rev. Drug Discovery, 2002, 1, 26–36; (f) J.-M. Lehn, Chem.
Soc. Rev., 2007, 36, 151–160; (g) S. Otto and K. Severin, Top. Curr.
Chem., 2007, 277, 267–288; (h) S. Ladame, Org. Biomol. Chem., 2008,
18 A. R. Hirst, I. A. Coates, T. R. Boucheteau, J. F. Miravet, B. Escuder,
V. Castelletto, I. W. Hamley and D. K. Smith, J. Am. Chem. Soc., 2008,
130, 9113–9121.
19 E. Defrancq, A. Leterme, N. Pelloux, M.-F. Lhomme and J. Lhomme,
Tetrahedron, 1991, 47, 5725–5736.
20 J. B. Epp and T. S. Widlanski, J. Org. Chem., 1999, 64, 293–295.
21 K. E. Norris, O. Manscher, K. Brunfeldt and J. Brammer Petersen,
Nucleic Acids Res., 1975, 2, 1093–1100.
22 K. P. Vercruysse, D. M. Marecak, J. F. Marecek and G. D. Prestwich,
Bioconjugate Chem., 1997, 8, 686–694.
6
, 219–226; (i) R. F. Ludlow and S. Otto, Chem. Soc. Rev., 2008, 37,
01–108; (j) K. Severin, Dalton Trans., 2009, 5254–5264; (k) U. L u¨ ning,
1
Polish J. Chem., 2008, 82, 1161–1174; (l) M. Mastalerz, Angew. Chem.,
Int. Ed., 2010, 49, 5042–5053.
3
(a) J.-M. Lehn, Chem.–Eur. J., 1999, 5, 2455–2463; (b) S. J. Rowan, S. J.
Cantrill, G. R. L. Cousins, J. K. M. Sanders and J. F. Stoddart, Angew.
Chem., Int. Ed., 2002, 41, 898–952; (c) P. T. Corbett, J. Leclaire, L. Vial,
K. R. West, J.-L. Wietor, J. K. M. Sanders and S. Otto, Chem. Rev.,
23 See also: (a) G. Palla, C. Pelizzi, G. Predieri and C. Vignali, Gazz.
Chim. Ital., 1982, 112, 339–341; (b) G. Palla, G. Predieri, P. Domiano,
C. Vignali and W. Turner, Tetrahedron, 1986, 42, 3649–3654; (c) E.
Kleinpeter, I. Starke, D. Str o¨ hl and H.-J. Holdt, J. Mol. Struct., 1997,
404, 273–290.
24 (a) A. Leo, C. Hansch and D. Elkins, Chem. Rev., 1971, 71, 525–616;
(b) A. J. Leo, Chem. Rev., 1993, 93, 1281–1306.
2
006, 106, 3652–3711; (d) C. D. Meyer, C. S. Joiner and J. F. Stoddart,
Chem. Soc. Rev., 2007, 36, 1705–1723; (e) A. Herrmann, Org. Biomol.
Chem., 2009, 7, 3195–3204; (f) M. Barboiu, Chem. Commun., 2010, 46,
7
466–7476; (g) G. Gasparini, M. Dal Molin and L. J. Prins, Eur. J. Org.
25 G. Cristalli, A. Eleuteri, S. Vittori, R. Volpini, M. J. Lohse and K.-N.
Klotz, J. Med. Chem., 1992, 35, 2363–2368.
Chem., 2010, 2429–2440; (h) V. T. Bhat, A. M. Caniard, T. Luksch, R.
Brenk, D. J. Campopiano and M. F. Greaney, Nat. Chem., 2010, 2,
26 See for example: (a) “Headspace Analysis of Foods and Flavors: Theory
and Practice”, R. L. Rouseff and K. R. Cadwallader (Eds.), Kluwer
Academic/Plenum Publishers, New York, 2001; (b) P. Rubiolo, B.
Sgorbini, E. Liberto, C. Cordero and C. Bicchi, in “The Chemistry
and Biology of Volatiles”, A. Herrmann, (Ed.), John Wiley & Sons,
Chichester, 2010, pp. 49–93.
4
90–497.
(a) W. G. Skene and J.-M. P. Lehn, Proc. Natl. Acad. Sci. U. S. A., 2004,
01, 8270–8275; (b) J.-M. Lehn, Prog. Polym. Sci., 2005, 30, 814–831;
c) T. Maeda, H. Otsuka and A. Takahara, Prog. Polym. Sci., 2009, 34,
81–604.
4
1
(
5
This journal is © The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 2906–2919 | 2919