1664
H. Tamiaki et al. / Tetrahedron 66 (2010) 1661–1666
was soluble in hot water (60 mM) but gave white precipitates after
cooling to room temperature, as in 1(14,23). In contrast to 1(14,32)
and 2(14,23), less hydrophobic 2(14,32) gave a transparent hydro-
gel under the same conditions as described above [at 60 mM after
cooling to room temperature from its hot aqueous solution at 60 ꢀC]
(see Fig. 1D). Although 2(14,45) afforded a highly viscous sol, i.e.,
a slightly fluid gel (see Fig. 1E), 2(14,77) yielded a similar clear gel as
2(14,32); the former is referred to a viscous sol hereafter. As shown
in the fourth column of Table 2, hydrogelation of 2(14,n) is de-
pendent on the oligo- or polymerization number (n) as is that of
1(14,n). Compared to 1(14,n), even a shorter poly(ethylene glycol)
chain as in 2(14,32) could gelate water and the total balance be-
tween hydrophilic and hydrophobic parts in a gelator molecule
would be important for the formation of supramolecular hydrogels.
It is noteworthy that hydrogels induced by 2(14,n) were clear due to
fewer of their large aggregates being formed.
chains, p–p stacking of the benzene rings and hydrogen-bonding of
the amide groups to form large supramolecules, which self-as-
sembled further to give the fibers as mentioned above.
Dodecylated 2(12,n) [n¼23, 32, and 45] gave no hydrogels under
the same conditions, due to their lower hydrophobicity and higher
water-solubility (third column of Table 2). Hexadecylated 2(16,n)
self-assembled in water (60 mM) to give a white gel at n¼23 and
viscous sols at n¼32 and 45 (fifth column of Table 2). In 2(m,32),
dodecylated compound (m¼12) was too water-soluble to gelate
water, tetradecylated compound (m¼14) gave a homogeneous
hydrogel with transparency due to the good balance of its hydro-
philic with hydrophobic substituents and self-aggregates of hexa-
decylated compound (m¼16) held water less efficiently to afford
a viscous sol (fifth row of Table 2). Hydrophobicity by an increase in
the length of alkyl chains is necessary for the formation of hydro-
gels by self-assemblies of the present synthetic amphiphiles but too
long alkyl chains made the hydrogels fluid.
Figure 2. A SEM image of self-assembly of 2(14,32) after freeze-drying its hydrogel
(60 mM, room temperature).
Addition of one more amide bond to 2(14,32) gave a diamide
6(14,1,32), which yielded a similar hydrogel. Minimum concen-
trations for the hydrogel formation of 2(14,32) and 6(14,1,32) were
36 and 15 mM, respectively: 15 mM of 6(14,1,32) was ca. 3%(wt/v).
Introduction of an amide bond would strengthen the in-
termolecular interaction through hydrogen bonding to suppress
the minimum concentration. The result also indicates the impor-
tance of hydrogen bonding of the amide bonds for hydrogels.
Moreover, such intermolecular interaction was supported by the
following experiments. Hydrogel of 2(14,32) was transformed to
a sol state by elevating the temperature. The transition tempera-
ture (Tgel) increased with an increase of concentration: Tgel¼28, 31,
and 43 ꢀC for 36, 40, and 60 mM, respectively. Higher concentra-
tion of gelators enhanced the intermolecular interaction in
a hydrogel to increase Tgel. It is noteworthy that hydrogel by
2(14,32) at 60 mM was formed under both the acidic and alkaline
conditions (at pH¼1 and 13).
2.4. Behavior of doubly tetradecylated amphiphiles in water
Since 2(14,32) is a good hydrogelator as mentioned above, the
hydrogelation ability of other related compounds was examined.
Removal of the 4-methoxy group from 2(14,32) as in 3(14,32) in-
duced the formation of a viscous sol under the same conditions
described above. This change from a hydrogel to a viscous sol would
be explained by three factors: 1) an increase of the molecular hy-
drophobicity, 2) an electronic factor to the intermolecularly in-
teractive phenyl and amide groups, and 3) a steric factor to the
conformation of the two tetradecoxy chains on the phenyl group.
Another amphiphile 4(14,32) possessing 3,4,5-alkoxy groups,
which was regioisomeric to 2(14,32) gave an aqueous solution but
no more hydrogel at room temperature, indicating that the first two
factors 1) and 2) could be ruled out. A shift of tetradecoxy group
from the 5- to 4-positions on the phenyl group in 3(14,32)/
5(14,32) disturbed hydrogelation to give an aqueous solution,
which supported that the 3- and 5-tetradecoxy substituents were
requisite for the formation of hydrogels at room temperature.
Substitution of amide bond of 2(14,32) with ester bond as in
20(14,32) led to no formation of the hydrogel at room temperature;
the amide bond is thus necessary for the hydrogelation. Monomeric
2(14,32) in dichloromethane gave a vibrational peak at 1658 cmꢃ1
for its amide carbonyl stretching band, while the IR peak for the
C]O of its hydrogel was situated at 1647 cmꢃ1. The 11-cmꢃ1 low
wavenumber shift by hydrogelation indicated that self-assemblies
of 2(14,32) in the hydrogel were formed by bonding of the amide
carbonyl group (probably hydrogen-bonding between amide sub-
stituents). After freeze-drying the hydrogel, fibrous networks were
observed as shown in a SEM image (see Fig. 2): the fibrous width
was about 200 nm and the porous diameter was about 600 nm.
Such fibrous self-assemblies of 2(14,32) would be prepared in its
hydrogel and the networks would contain a large amount of water
to make the hydrogel. The amphiphilic compounds self-aggregated
in water through hydrophobic interaction of the long tetradecyl
2.5. Temperature dependency for hydrogelation
As mentioned above, 60 mM of 2(14,32) in water gave a clear
hydrogel in the range from room temperature to 43 ꢀC, above
which the normal gel was transformed to the viscous sol. Similar
transformation was observed by rheological technique. During
heating the hydrogel of 2(14,32) from 0 ꢀC, storage modulus G0 was
larger than loss modulus G00 below 53 ꢀC and G0 was smaller than G00
in the range between 53 and 69 ꢀC (see Fig. 3A). The elastic mod-
ulus data indicate that Tgel from gel to sol states was 53 ꢀC, which
was higher than the above value from fluidity checked by the eye:
the difference might be ascribable to the observation techniques.
Such temperature-dependent gels are often observed because
thermal energy could collapse the gel network. In contrast, the
viscous sol induced by 3(14,32) at room temperature (60 mM) was
changed to a hydrogel at 60 ꢀC (see also Table 3). Such a reverse
dependency is sometimes observed in a hydrogel by polymeric
gelators including PEG and poly(N-isopropylacrylamide) de-
rivatives.7 During the heating of aqueous 60 mM of 4(14,32),
transformation of solution to gel states was similarly observed at
52 ꢀC. Even with heating, no state transformation occurred in
5(14,32) to make it an aqueous solution due to its weaker in-
termolecular interaction (vide supra). It is noted that the resulting