Resolution of the acid-base fraction curves of the calixarene derivatives with chemometric methods

Article abstract of DOI:10.1246/cl.2004.182

The acid-base properties of several new calix[4]arene derivatives were studied with chemometric methods by measurement of their UV absorbances under different pH. Resolving the acid-base fraction curves with chemometric methods, the pK_a of thes

Full text of DOI:10.1246/cl.2004.182

182
Chemistry Letters Vol.33, No.2 (2004)
Resolution of the Acid-base Fraction Curves of the Calixarene Derivatives
with Chemometric Methods
Li Wang, Zhong-Liang Zhu, and Xian-Fa Shi
Department of chemistry, Tongji University, Shanghai, 200092, P. R. China
(Received September 12, 2003; CL-030860)
The acid-base properties of several new calix[4]arene deriv-
acid in the presence of glacial acetic acid gave the derivative 2 in
good yield. The reduction of derivative 2 with SnCl2 in ethanol
yielded the amino product 4. Treatment of derivative 2 with
strong base NaOH afforded derivative 3. All the derivatives gave
satisfactory analytical results. The absorption spectra were ob-
tained on a Lambada Bio40 UV–vis spectrophotometer (Perkin
Elmer Corp.). The measurement was carried out in a solution
of mixed solvent, water and tetrahydrofuran (V:V = 1:1) for 2
and 4, water and ethanol (V:V = 3:7) for 3. The pH values were
adjusted using HCl, the buffer or NaOH.
atives were studied with chemometric methods by measurement
of their UV absorbances under different pH. Resolving the acid-
base fraction curves with chemometric methods, the pKa of these
derivatives were determined and the fraction curves and pure ab-
sorbing spectra of each absorbing component in the calixarene
system were obtained. The proton dissociation behavior of these
derivatives was also discussed in this paper.
Calixarenes are cyclic oligomers composed of phenol units
linked by methylene. The hydroxy of phenol units frequently
play one of the decisive roles in the supramolecular properties
of calixarenes such as the stabilization of a cone conformation
and the substrate selectivity.1,2 In addition, the phenol proton
dissociation was also often applied in the selective functionaliza-
tion of clixarenes and syntheses of calixarene derivatives.3–5 So
to know the proton dissociation behavior of calixarene is indis-
pensable to understand the chemistry of calixarene. However,
the related researches have been very limited. Among literatures
R
1
2
R = H, R' = COPh;
R = NO ,
R' = COPh;
2
CH2
CH2
R = NO ,
3
4
R' = H;
2
3
R = NH ,
OH
R' = COPh
OR'
2
Figure 1. Structure of the calixarene derivatives.
All the three derivatives displayed a pH-dependent spectral
change. Owing to this chromophoric nature, we could obtain
two-way data matrix and used the chemometric methods to study
the proton dissociation of calixarenes. For each derivative, we
first used principal component analysis (PCA)11 to determine
the number of the principal components of the measured sys-
tems. Operating PCA on the data matrix respectively, we can af-
firm there are both 3 absorptive components for derivative 3 and
4, and 2 absorptive components for derivative 2 in the measured
pH ranges. The existing pH range of each component was com-
puted with evolving factor analysis (EFA).12 For derivative 2,
the existing ranges of the two components are 3–7 and 4–7, re-
spectively. For derivative 3, the existing ranges of the three com-
ponents are 0–4, 2–13.5, and 11–14, respectively. For derivative
4, they are 0–2, 0–7, and 4–8, respectively.
The two-way data matrix of derivative 3 was further ana-
lyzed using iterative target transformation factor analysis (ITT-
FA).13 ITTFA is a modelless method derived from the self-mod-
eling curve resolution (SMCR). It takes advantages of the
features of the two-way data itself and resolves the data matrix
without any model. So it can overcome the drawbacks of the
model-based methods and attain great accuracy. As for the deriv-
atives 2 and 4, because their first components existing in the acid
range have only small absorption, we couldn’t get satisfactory
results with ITTFA. With target testing factor analysis
(TTFA),14 we still could gain perfect results for the two deriva-
tives. Unlike ITTFA, TTFA is applicable to resolve the model-
known data. Here a set of pKa values is first supposed and then
is optimized continually. When the difference between the cal-
culated reproduction data and the original data reaches the mini-
mum, the results are obtained. For each derivative, the fraction
so far appeared, we only find that Bohmer’s group and Shinkai’s
¨
group have estimated the pKa of several calixarene derivatives
such as p-sulfonatocalix[4]arne and p-nitrocalix[4]arene with
spectroscopic or potentiometric titration.6–9 Furthermore, these
studies just simply estimated the pKa values of the calixarene de-
rivatives. As far as we know, no further work appears dealing
with the dissociation behavior of the phenol proton in the calix-
arene derivatives until now. The scarcity of the related research-
es may be ascribed not only to the difficulty in solubility of most
calixarenes in water but also to the complication of the proton
dissociation of calixarenes.
Here we have synthesized several new calixarene deriva-
tives partly soluble in water and successfully introduced the che-
mometric methods to study their proton dissociations. Calix-
arenes frequently have several protons in one molecule
resulting in the complication of the proton dissociation. If the
pKa values are at close range or the spectra of each absorbing
species overlap seriously, the conventional computing methods
can’t give a satisfied result, even can’t give any resolution.
The chemometric methods display their advantages in this case,
especially when the number of ionized proton is unknown under
certain pH. Applying the chemometric methods allowed us to ac-
curately resolve the acid-base fraction curves of the calixarenes.
Besides determining their pKa values, with these methods we
could achieve the detailed information about the proton dissoci-
ation behavior, such as the pH range of each component and their
pure absorption spectra.
The measured calixarene derivatives 2–4 were first synthe-
sized (Figure 1). The starting derivative 1 was prepared accord-
ing to the reported method.10 Nitration of derivative 1 with nitric
Copyright Ó 2004 The Chemical Society of Japan

Chemistry Letters Vol.33, No.2 (2004)
183
curves obtained with ITTFA or TTFA are illustrated in Figure 2
(shown as mark symbol). According to the fraction curves of
each absorbing component, we can obtain the pKa values. The
pKa value of derivative 2 was calculated to be 5.24. The first
two pKa values of derivative 3 were 3.63 and 12.53, and the
two pKa values of derivative 4 were 0.85 and 5.86. Inversely, ac-
cording to the chemical equilibrium, using the pKa values we can
calculate the fraction curves. They were expressed as solid lines
in Figure 2. In other words, we have used the chemometric meth-
ods ITTFA or TTFA in analyzing the measured data to obtain the
fraction curves (experimental data, mark symbol in Figure 2) as
well as the pKa values, then used these pKa values to calculate
the fraction curves (theoretic data, solid line in Figure 2) in-
versely. The good accordance shown in all the figures implies
the reliability of the results obtained with these chemometic
methods. Treating the fraction carves obtained with the ITTFA
or TTFA with least square regression, we could extract the pure
absorption spectra of each absorbing component of these deriv-
atives.
of calixarene derivatives and find the rules of the proton dissoci-
ation of calixarenes and their derivatives.
The calculated results show us some interesting information
of their proton dissociations. Derivative 4 has two pKa values,
although it has only one phenol proton. But its amino group
on the upper rim can form ammonium ion in the acidic solution.
So we attributed its pKa1 to the proton of this ammonium ion.
Then the pKa value of the phenol proton of derivative 4 should
be pKa2 (5.86). The pKa of derivative 2 (5.24) is little smaller
than the pKa2 (attributed to the phenol proton) of derivative 4.
This can be rationalized with their different groups on the upper
rim. The group of derivative 2 is nitro that has the ability of at-
tracting electron, which can help the dissociation of the proton.
On the contrary, derivative 4 has an electron-repulsive amino ni-
trogen on the upper rim. Derivative 3 has two protons ionized in
the measured pH range (0–14). The first proton should be the
proton of the phenol unit with the electron-attracting nitro group.
Its pKa1 (3.63) is far smaller than the pKa of derivative 2, which
is ascribed to the hydrogen bond formed by the oxide anion with
the neighbor phenol hydroxy. Owing to the same reason, the oth-
er ionized proton should be the proton of the opposite phenol
unit. The great gap between pKa1 and pKa2 (12.53) also is easily
interpreted in terms of the hydrogen-bond interactions. When
one proton ionized, the oxide anion is strongly stabilized by
two hydrogen bonds. While two protons ionized, each oxide
anion is stabilized by only one hydrogen bond. From above dis-
cussion, we could conclude that the group opposite phenol hy-
droxyl and the hydrogen bond formed by the oxide anion with
the neighbor phenol hydroxy play important roles in the proton
dissociation of calixarenes.
1.0
0.8
0.6
A
0.4
0.2
0.0
3
4
5
6
7
pH
The author Li Wang is grateful for the financial support of
Science Development Foundation of Tongji University.
1.0
0.8
0.6
0.4
0.2
0.0
References
1
2
3
C. D. Gutsche, Acc. Chem. Res., 16, 161 (1983).
S. Shinkai, Tetrahedron, 49, 8933 (1993).
L. C. Groenen, B. H. M. Ruel, A. Casnati, W. Verboom, A.
Pochini, R. Ungaro, and D. N. Reinhoudt, Tetrahedron, 47,
8379 (1991).
B
8
4
E. M. Collins, M. A. Mckervey, E. Madigan, M. B. Moran, M.
Owens, G. Ferguson, and S. J. Harris, J. Chem. Soc., Perkin
Trans. 1, 1991, 3137.
6
0
2
4
10
12
14
pH
1.0
0.8
0.6
0.4
0.2
0.0
5
6
7
8
9
J. D. van Loon, W. Verboom, and D. N. Reinhoudt, Org. Prep.
Proced. Int., 24, 437 (1992).
V. Bohmer, E. Schade, and W. Vogt, Makromol. Chem., Rapid
¨
Commun., 5, 221 (1984).
S. Shinkai K. Araki, H. Koreoshi, T. Tsubaki, and O. Manabe,
Chem. Lett., 1986, 1351.
S. Shinkai, K. Araki, J. Shibata, D. Tsugawa, and O. Manabe,
Chem. Lett., 1989, 931.
K. Araki, K. Iwamoto, S. Shinkai, and T. Matsuda, Bull. Chem.
Soc. Jpn., 63, 3480 (1990).
C
4
8
0
2
6
10
10 C. D. Gutsche and L. G Lin, Tetrahedron, 42, 1633 (1986).
11 E. R. Malinowski and D. G. Howery, ‘‘Factor Analysis in Chem-
istry,’’ Wiley- Interscience Publication, New York (1980),
pp 73–79.
12 Y. Z. Liang, ‘‘White, Gray, and Black Multicomponent Systems
and Their Chemometric Algorithms,’’ Hunan Publishing House
of Science and Technology, Changsha (1996), pp 168–174.
13 Z. L. Zhu, W. Z. Cheng, and Y. Zhao, Chemom. Intell. Lab.
Syst., 64, 157 (2002).
pH
Figure 2. Distribution of the components (mark symbol: ITT-
FA or TTFA, solid line: calculated) A: derivatives 2; B: deriva-
tive 3; C: derivative 4.
Thus, with above chemometric methods, we have analyzed
the proton dissociation behavior of the calixarene derivative 2–4.
Chemometric methods provide us useful techniques to deal with
the complicated calixarene systems. With these methods, we
could systematically study the proton dissociation of a series
14 Z. L. Zhu, J. Xia, S. G. Mi, and T. H. Li, Chem. J. Chin. Univ.,
23, 546 (2002).
Published on the web (Advance View) January 19, 2004; DOI 10.1246/cl.2004.182