S. Bhattacharya et al.
ture (v/v) followed by heating at 70–758C gave a transparent
solution, which upon subsequent cooling and sonication led
to the formation of a white colored, thermoreversible,
opaque gel (minimum gelator concentration [mgc]=
16.4 mm=[BP]) (Figure 1 h). It is important to note that the
mixture of 2,2’-bipyridine or 3,3’-bipyridine and DA did not
lead to gelation under the same conditions. This observation
prompted us to synthesize OPV derivatives functionalized
with various isomeric pyridyl termini (NP, NM, NO) to
obtain salt-type OPV-based gels. Among them, notably only
mixture of NP and DA (1:2) was able to form an orange-col-
ored, fluorescent, thermoreversible, opaque gel in 2:1 water/
ethanol (v/v) ([mgc]=3.31 mm=[NP]) (Figure 1g), whereas
other members did not lead to gelation with DA in that sol-
vent system. Interestingly, a 1:2 mixture of NP and DA
formed a gel at much lower mgc compared with that of the
BP–DA gel. Accordingly, 1 unit of NP–DA (1:2) system can
hold almost 8000 water molecules and 1300 ethanol mole-
cules inside the three-dimensional gel network. The chain
length of the OPV backbone is also important in determin-
ing the effect on gelation. Accordingly a control compound,
8-NP was synthesized (Figure 1a). Interestingly, 8-NP
formed a precipitate with DA in a 2:1 water/ethanol mix-
ture. This result clearly elucidates that a hydrophilic/hydro-
phobic balance is crucial for the gelation of this system.
Salt formation of each of NP, NM, and NO was also
checked in presence of different fatty acids of variable chain
lengths as well as aromatic acids, such as benzoic acid.
Among them, only NP resulted in the salt formation with
these acids, whereas NM and NO failed. Interestingly, the
salt of NP with fatty acids of variable chain lengths (no. of
carbon atoms (n)=8, 10, 12, 14, 16) led to the formation of
gelatinous precipitates in the 2:1 water/ethanol mixture.
This signifies that the hydrogen bonding through amide
units of DA is very important and plays a crucial role in the
aggregation as well as hydrogelation. Benzoic acid was
chosen as the simplest among the aromatic acids, which also
resulted in the formation of precipitate with NP in that sol-
vent system, because of the lack of both flexible aliphatic
hydrocarbon chain and hydrogen-bonding motif. Thus apart
from the acidity, the lipophilic/hydrophilic balance as well as
the flexibility and hydrogen bonding all contribute to the ge-
lation.
the pKa value of each of these series of compounds was ob-
tained from the plots of the absorbance versus pH. The ex-
perimentally determined pKa values of NP in 0.1m sodium
formate/formic acid and phosphate/citrate buffer solutions
are 3.77 and 3.72, respectively (Figures S4 and S5, the Sup-
porting Information). So, the average pKa value of NP is
3.75Æ0.04. Similarly, the experimentally determined pKa
value of NM is 3Æ0.11 (Figures S6 and S7, the Supporting
Information), whereas NO remains unprotonated even at
pH 2.8. This is clearly evident from the broad absorption
spectra of NO in sodium formate/formic acid buffer solution
(Figure S8, the Supporting Information). According to the
experimentally determined pKa values, the order of basicity
of these OPVs is NP>NM>NO.
Proton transfer during an acid/base reaction depends on
the DpKa value (pKa of the conjugate acid of pyridine
baseÀpKa of the carboxylic acid).[24] When DpKa <0, it re-
À
À
sults in a neutral O H···N H bond. On the other hand, 0<
+
À
À
À
DpKa <3 gave an intermediate O H···N/N H···O hydro-
gen-bond character. However, DpKa >3 results in complete
proton transfer to induce an ionic interaction N+ H···O .
À
À
Therefore, NP with relatively higher basicity should be able
to form a salt with DA with an intermediate O H···N/N
+
À
À
H···OÀ hydrogen-bond character, which accounts for the
reason of selective gel formation by NP. Whereas, NM and
NO, being relatively less basic compared with NP, most
likely fail to form a salt with DA. Furthermore, in case of
NO, the pyridyl ’’N’’ atom is probably buried under the p-
cloud of the OPV backbone. This results in a crowding
effect, which may be an alternative reason of the inability of
NO towards salt formation. It is important to note that pKa
was measured in pure buffered aqueous media. However,
the gelation was performed in 2:1 water/ethanol mixture,
and there will be changes in the pKa values in ethanolic
aqueous media. Moreover, it is known that pKa values of
different pyridine derivatives show changes at interfaces.[25]
Thus, the pKa of various isomeric OPV derivatives may also
change upon self-assembly. However, the extent of change
in pKa would be probably similar for the three isomers. So,
the trend in these values may explain the selective salt for-
mation and gel formation phenomena as observed here.
Computational studies: The main factor that governs the se-
lective gel formation is the protonation of NP by DA, and
thereby self-aggregation of NP–DA (1:2) using several weak
non-covalent interactions leads to robust gel formation. To
obtain a reasonable explanation of these observations, each
of these molecules in the form of the HCl salt were opti-
mized by using the B3LYP G/6-31G* level of theory
(Figure 2). HCl was chosen as a proton source to mimic the
real situation. For these purposes, the two most relevant
conformations were considered; one in which the attached
N atom of the end pyridyl ring was placed facing towards
the vinylic bond and one in which the attached N atom was
placed away from the vinylic bond (in case of NM-HCl and
NO-HCl; Figure 2b and c). In striking contrast, only one
conformation was possible for NP-HCl, in which the at-
The role of basicity in selective gelation: The gelation prop-
erties of the OPV derivatives in presence of DA was found
to depend on the basicity of the ’’N’’ atom of the pyridyl
ring. We thought that it is important to determine the pKa
values of these compounds to find out the reason of selec-
tive gelation by NP. Accordingly, we obtained the pKa
values of these pyridine derivatives experimentally.
Since the protonated and non-protonated forms of these
OPV derivatives absorb differently, UV/Vis spectroscopy
was used as an efficient tool to determine their pKa
values.[23] We recorded the UV/Vis absorption spectra of
these compounds in 0.1m sodium formate/formic acid and
phosphate/citrate buffer solutions of various pH values and
16674
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 16672 – 16681