Catechol immobilized on crosslinked polystyrene resins by grafting or copolymerization: Incidence on metal ions adsorption

Article abstract of DOI:10.1016/j.reactfunctpolym.2011.11.001

Porous poly(styrene-co-divinylbenzene) resins functionalized with catechol were prepared in two different ways. Catechol was either grafted on commercial Amberlite XAD-4 resin via a reduced imine and diazo bridges or incorporated by direct copolymerizatio

Full text of DOI:10.1016/j.reactfunctpolym.2011.11.001

Reactive & Functional Polymers 72 (2012) 98–106
Contents lists available at SciVerse ScienceDirect
Reactive & Functional Polymers
journal homepage: www.elsevier.com/locate/react
Catechol immobilized on crosslinked polystyrene resins by grafting or
copolymerization: Incidence on metal ions adsorption
c
c
a
Julie Bernard a, Catherine Branger a,b, , Isabelle Beurroies , Renaud Denoyel , André Margaillan
⇑
a Laboratoire MAPIEM, EA 4323, Université du Sud Toulon Var, ISITV, Avenue George Pompidou, BP 56, 83162 La Valette du Var, France
b IUFM Célestin Freinet, Université de Nice Sophia-Antipolis, 89 Avenue George V, 06046 Nice Cedex 1, France
c Universités d’Aix-Marseille I, II et III, CNRS UMR 6264, Laboratoire Chimie Provence, Site Madirel, Centre de St Jérôme, 13397 Marseille Cedex 20, France
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 7 July 2011
Received in revised form 18 October 2011
Accepted 2 November 2011
Available online 9 November 2011
Porous poly(styrene-co-divinylbenzene) resins functionalized with catechol were prepared in two differ-
ent ways. Catechol was either grafted on commercial AmberliteÒ XAD-4 resin via a reduced imine and
diazo bridges or incorporated by direct copolymerization of divinylbenzene with dimethoxystyrene fol-
lowed by deprotection of the methoxy groups. The efficiency of functionalization was evidenced by pyro-
lysis coupled with gas phase chromatography and infra-red spectroscopy. The amount of incorporated
catechol inside the different resins was determined by acido-basic back titration and varies between
0.27 and 1.38 mmol/g of resin. Grafting resulted in a decrease of the surface area due to the blocking
of some connections between the mesopores. For synthesized copolymers, high contents of divinylben-
zene monomer led to high surface areas. Comparing metal retention properties of both kind of materials
towards Pb(II), Cu(II), Ni(II) and Cd(II) ions proved the equivalence of these sorbents. At low metal con-
centration, interactions with the sorbents seem to be non specific whereas at higher concentration cat-
echol is responsible for the retention properties. Since the synthesis of copolymers is easier than the
grafting of sorbents, it appears to be a more attractive procedure to obtain chelating sorbents.
Ó 2011 Elsevier Ltd. All rights reserved.
Keywords:
Porous copolymers
Catechol
Grafting
Suspension copolymerization
Solid-phase extraction
1. Introduction
is to introduce complexing groups on sorbent surface either by the
impregnation technique [16,17] or by grafting [18–21]. Although
Highly crosslinked macroporous resins have been widely studied
for their properties linked to their porosity [1]. Introducing func-
tional groups on such materials makes them suitable for applica-
tions in various areas such as ion exchangers resins [2–4],
catalysts supports [5–7], solid-phase extraction sorbents [8,9],
materials for biomedical applications [10,11] and solid-supported
purification [12,13]. The performances of the resins are highly
dependent on their structure.
In the field of metal trace analysis in liquids, functional resins
are used to enhance the sensitivity of detection techniques by
introducing a preconcentration step prior to analysis [14]. The
so-called solid-phase extraction process is based on the interac-
tions between a porous solid sorbent and the target molecules
present in the liquid phase. The nature of the sorbent functional
groups will directly influence the properties of the material: the
use of chelating functions is particularly well suited to the extrac-
tion of metal ions [15]. The first route for preparing chelating resins
impregnation is very simple to achieve, it involves a weak binding
between the ligand and the sorbent and therefore poor stability of
the material. Grafting enables to improve the stability and reus-
ability of the sorbents thanks to the formation of covalent bonds
between ligand and sorbent. It can also be achieved on home-made
copolymers [22,23]. This approach is time consuming and difficult
to realize. The second route to obtain porous chelating sorbents is
to directly copolymerize a functional monomer with a crosslinker
[24–26]. Suspension polymerization is usually performed and a
porogen agent (a diluent, an oligomer or a polymer) is introduced
to control the porosity of the prepared beads. The number of syn-
thetic steps is thus reduced and the reproducibility improved.
Grafting is more widespread in the literature since it does not in-
volve the preparation of specific monomers but copolymerization,
thanks to its convenience, appears more attractive providing
experimental parameters (type and amount of the porogen agent,
crosslinker concentration. . .) are optimized.
Catechol (1,2-benzenediol) is a metal complexing agent since its
two hydroxyl functions enable chelation of metals such as cad-
mium and copper with relatively high complexation constants
(log K = 10.8 and 14.1, respectively) [27]. Catechol containing mol-
ecules and their metal complexes are also of wide interest in
⇑
Corresponding author at: Laboratoire MAPIEM, EA 4323, Université du Sud
Toulon Var, ISITV, Avenue George Pompidou, BP 56, 83162 La Valette du Var, France.
Tel.: +33 4 9414 2580; fax: +33 4 9414 2598.
E-mail address: branger@univ-tln.fr (C. Branger).
1381-5148/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.reactfunctpolym.2011.11.001

J. Bernard et al. / Reactive & Functional Polymers 72 (2012) 98–106
99
biology since they are responsible for metal ion solubilization,
transport, and storage [28,29]. Functional polymers have been pre-
pared with catechol with various aims such as miming mussel
adhesive proteins [30], preparing helical functional polymer [31]
or antifouling macromolecules [32].
4-[(4-ethylphenylimino)methyl]benzene-1,2-diol
(250 mg,
1,04 mmol) was dissolved in dried methanol. NaBH4 (194 mg,
5.10 mmol) was slowly added while maintaining the temperature
below 5 °C. The reaction mixture was stirred at that temperature
for 30 min. 50 mL of deionised iced water were added and the
product was extracted with chloroform. After drying on sulfate
magnesium and removal of the solvent, 189 mg of 4-[(4-ethylphe-
nylamino)methyl]benzene-1,2-diol as a yellow liquid were ob-
tained (yield 78%). 1H NMR (DMSO-d6; 400 MHz; d in ppm) 8.58
(s, 2H, AOH protons), 6,87 (d, 2H, aromatic protons ortho to ethyl
group), 6.74–6.60 (m, 3H, aromatic protons of the hydroxylated
benzene), 6.49 (d, 2H, aromatic protons ortho to ANHA), 4.40 (s,
2H, ACH2A protons of reduced imine), 2.41 (q, 2H, ACH2A protons
of ethyl group), 1.09 (t, 3H, ACH3).
This paper presents the introduction of catechol on crosslinked
polystyrene resins by either grafting or copolymerization. Grafting
is achieved on commercial AmberliteÒ XAD-4 resin via two differ-
ent bridges: reduced imine or diazo. In the case of the imine bridge,
reduction to an amine function is performed to increase the stabil-
ity of the functionalized sorbent toward aqueous phase. In the case
of copolymerization, commercial 3,4-dimethoxystyrene (DMS) is
used as a functional monomer and copolymerized in suspension
with divinylbenzene (DVB). The dimethoxy groups are further
cleaved to recover the hydroxyl functions and obtain a poly(vinyl-
catechol-co-divinylbenzene) resin. The extent of incorporation of
catechol in both types of materials is determined, as well as its
influence on porous structure. Their complexing properties for
Cd(II), Cu(II), Ni(II) and Pb(II) ions are compared in aqueous
solution.
2.2.2. Preparation of Resin I (Scheme 2)
4.43 g of AmberliteÒ XAD-4 bearing an amine function, pre-
pared according to Ref. [33], and 3,4-dihydroxybenzaldehyde
(8.38 g, 60.7 mmol) were introduced in 100 mL of dried diethyl
ether. The mixture was stirred at room temperature. After 6 days,
the beads were filtrated and washed successively with 100 mL of
dried diethyl ether and 100 mL of dried ethanol. The beads were
then dried under vacuum at room temperature for 3 days (5.25 g).
3.74 g of the previous resin were suspended in 100 mL of anhy-
drous methanol, cooled to 0–5 °C and mechanically stirred for
30 min (100 rpm). NaBH4 (3.66 g, 96.3 mmol) was slowly added
while maintaining the temperature below 5 °C. The mixture was
stirred at that temperature for 1 h. 200 mL of deionised iced water
were added and stirring was carried on at 5 °C for 20 min. The
beads were filtrated and washed with water until reaching neutral
pH. The resin beads were finally washed with a Soxhlet system
24 h with water and 24 h with methanol. They were dried under
2. Experimental
2.1. Materials
AmberliteÒ XAD-4 was purchased from Acros Organics. The
beads were washed before use to eliminate NaCl and Na2CO3 used
for storage. In that aim, AmberliteÒ XAD-4 was suspended under
orbital agitation successively in methanol (24 h), 10% HCl solution
(4 h), pH 9 NaOH solution (4 h), deionised water (for neutraliza-
tion) and methanol (4 h) and finally dried under vacuum at room
temperature.
3,4-dihydroxybenzaldehyde (3,4-DHB), 2-methoxyphenol (guaiacol),
sodium nitrite, boron tribromide, anhydrous SnCl2, 4-ethylaniline,
toluene, acetonitrile, polyvinylalcohol (PVA, 88% hydrolyzed, Mw
ꢀ8800 g molꢁ1), HCl 37%, H2SO4 65%, HClO4 70% and NaBH4 were
purchased from Acros Organics and were used without purification.
Divinylbenzene (DVB) (80% technical grade), styrene (>99%), 3,4-
dimethoxystyrene (DMS) (technical grade), azobis(isobutyronitrile)
(AIBN) and NaCl were supplied by Sigma Aldrich. Diethyl ether,
methanol and ethanol were obtained from Carlo Erba and dried on
molecular sieves before use.
vacuum for 3 days at room temperature and sieved at 300
(99% of the resin beads, 3.46 g of Resin I).
lm
2.2.3. Synthesis of the model molecule for Resin II: 4-[(4-
ethylphenyl)diazenyl]benzene-1,2-diol (Scheme 3)
4-Ethylaniline (1.21 g, 10.0 mmol) was dissolved in a mixture of
20 mL of deionised water and 5 mL of HCl. The mixture was cooled
at a temperature between 0 and 5 °C and placed under inert atmo-
sphere (N2 gas). A solution of sodium nitrite (0.83 g in 5 mL of
water) was added dropwise while controlling the temperature
and stirring was maintained for 40 min. A solution of guaiacol
(1.10 g, 8.87 mmol) was prepared in a mixture of 10 mL of water
and 10 mL of acetonitrile and put under nitrogen flux to prevent
the oxidation of the hydroxyl function of guaiacol. This solution
was slowly added on the diazonium salt while controlling the tem-
perature (<5 °C) and the pH (at 6–7 with a K2CO3 solution at
1 mol Lꢁ1). The reaction mixture was stirred for 30 min at a tem-
perature between 0 and 5 °C. At the end of the reaction, a few drops
of HCl were added to help precipitation of the product. It was ex-
tracted by diethyl ether. The obtained dark oil was purified by
chromatography on silica gel eluting with pentane to which tolu-
ene was added until reaching 100% v/v. 1.34 g of 4-[(4-ethylphe-
nyl)diazenyl]-2-methoxyphenol as a red liquid were obtained
(yield 63%). 1H NMR (DMSO-d6; 400 MHz; d in ppm) 9.93 (s, 1H,
AOH proton), 7.76 (d, 2H, aromatic protons meta to ethyl group
and ortho to AN@NA), 7.48–7.45 (m, 2H, aromatic protons ortho
to AN@NA on the guaiacol benzene ring), 7.38 (d, 2H, aromatic
protons ortho to ethyl group), 6.97 (d, 1H, aromatic proton ortho
to AOH), 3.86 (s, 3H, ACH3 protons of guaiacol), 2.68 (q, 2H,
ACH2A), 1.21 (t, 3H, ACH3 protons of ethyl group).
For the determination of complexing properties, Cd(II), Cu(II),
Ni(II) and Pb (II) solutions (1 g Lꢁ1) from Panreac and HNO3 supra-
pure 65% from Carlo Erba were used. All metal solutions were pre-
pared with ultra-high quality deionised water (Millipore,
resistivity >18 m
X cm).
2.2. Grafted resins preparation (Scheme 1)
2.2.1. Synthesis of the model molecule for Resin I:
4-[(4-ethylphenylamino)methyl]benzene-1,2-diol (Scheme 2)
4-Ethylaniline (440 mg, 3.63 mmol) and 3,4-dihydroxybenzal-
dehyde (500 mg, 3.63 mmol) were dissolved in 70 mL of dried
diethyl ether. The reaction mixture was stirred at room tempera-
ture for 24 h. After solvent removal, the crude product was recrys-
tallized in ethanol leading to 820 mg of an orange solid. (820 mg of
4-[(4-ethylphenylimino)methyl]benzene-1,2-diol, yield 94%). 1H
NMR (DMSO-d6; 400 MHz; d in ppm) 9.44 (s, 2H, AOH protons),
8.37 (s, 1H, imine proton), 7.44 (s, 1H, aromatic proton ortho to
AOH and imine), 7.20–7.11 (m, 5H, aromatic protons ortho to
@NA, to ethyl group and to imine), 6.86 (d, 1H, aromatic proton
ortho to AOH and meta to imine), 2.58 (q, 2H, ACH2A), 1.17 (t,
3H, ACH3).
4-[(4-ethylphenyl)diazenyl]-2-methoxyphenol
(280 mg,
1.09 mmol) was dissolved in 20 mL of anhydrous dichloromethane
under nitrogen atmosphere. The solution was cooled to 0–5 °C and
stirred for 30 min. 1.00 mL of BBr3 (10.4 mmol) were added drop-

100
J. Bernard et al. / Reactive & Functional Polymers 72 (2012) 98–106
CH CH2
CH CH2
OH
OH
x
y
OH
OH
N N
NH CH2
OH
OH
z
Resin I
Resin II
DVB-DOH
Scheme 1. Structures of grafted resins (resin I and II) and synthesised copolymers (DVB-DOH).
OH
OH
OH
3,4-DHB,
Ether, RT
NaBH ,
4
R
NH2
R
N CH
OH
R
NH CH2
Methanol,
0-5°C
R = C H
4-[(4-ethylphenylimino)methyl]benzene-1,2-diol
for R = C H
4-[(4-ethylphenylamino)methyl]benzene-1,2-diol
for R = C H
2
5
or
R = polymer backbone
2
5
2 5
or
or
Resin I for R = polymer backbone
Resin I before reduction
for R = polymer backbone
Scheme 2. Synthesis scheme of 4-[(4-ethylphenylamino]methyl)benzene-1,2-diol (R = C2H5) and Resin I (R = polymer backbone).
OCH3
OH
OH
OH
1/ HCl, NaNO ,
2
BBr ,
3
0-5°C
R
NH2
R
N
N
R
N
N
2/ Guaiacol,
CH Cl ,
2
2
H O/CH CN,
2
3
0-5°C
0-5°C
R = C H
2
4-[(4-ethylphenyl)diazenyl]-2-methoxyphenol
for R = C H
5
4-[(4-ethylphenyl)diazenyl]benzene-1,2-diol
for R = C H
or
2
5
2
5
R = polymer backbone
or
or
Resin II for R = polymer backbone
Resin II before deprotection
for R = polymer backbone
Scheme 3. Synthesis scheme of 4-[(4-ethylphenyl)diazenyl]benzene-1,2-diol (R = C2H5) and Resin II (R = polymer backbone).
wise while maintaining the temperature under 5 °C. The mixture
was then stirred for 1 h. It was then hydrolyzed with 200 mL of
deionised iced water and the product was extracted with dichloro-
methane. After drying on sulfate magnesium and removal of the
solvent, 240 mg of 4-[(4-ethylphenyl)diazenyl]benzene-1,2-diol
as a yellow liquid were obtained (yield 90%). 1H NMR (DMSO-d6;
400 MHz; d in ppm) 9.81–9.49 (broad peaks, 2H, AOH protons),
7.72 (d, 2H, aromatic protons meta to ethyl group and ortho to
AN@NA), 7.34 (m, 4H, aromatic protons ortho to AN@NA on the
guaiacol benzene ring and aromatic protons ortho to the ethyl
group and meta to AN@NA), 6.94 (d, 1H, aromatic proton ortho
to AOH and meta to AN@NA), 2.65 (q, 2H, ACH2A), 1.19 (t, 3H,
ACH3 protons of ethyl group).
ture between 0 and 5 °C. The beads were filtrated and rinsed with
acetone. The resin beads were finally washed with a Soxhlet
system 24 h with water and 24 h with ethanol and then dried un-
der vacuum at room temperature for 3 days (5.60 g).
3.20 g of the previous resin were suspended in 50 mL of anhy-
drous dichloromethane, placed under nitrogen atmosphere, cooled
to 0–5 °C and mechanically stirred for 30 min (100 rpm). 3.2 mL of
BBr3 (33 mmol) were slowly added while maintaining the temper-
ature under 5 °C. The mixture was then stirred for 7 h at that tem-
perature. It was then hydrolyzed with deionised iced water. The
beads were filtrated and washed with water until reaching neutral
pH. The resin beads were finally washed with a Soxhlet system
24 h with water and 24 h with ethanol. They were dried under vac-
uum for 3 days at room temperature and sieved at 300
the resin beads, 2.91 g of Resin II).
lm (99% of
2.2.4. Preparation of Resin II (Scheme 3)
5.66 g of AmberliteÒ XAD-4 bearing an amine function were
suspended in a mixture of 97 mL of deionised water and 23 mL
of HCl. The reaction mixture was cooled at a temperature between
0 and 5 °C and placed under inert atmosphere (N2 gas). A solution
of sodium nitrite (3.88 g in 23 mL of water) was added dropwise
while controlling the temperature and stirring was maintained
for 2 h. A solution of guaiacol (5.10 g, 41.1 mmol) was prepared
in a mixture of 15 mL of water and 15 mL of acetonitrile. This solu-
tion was slowly added on the diazotated resin while controlling the
temperature (<5 °C) and the pH (at 6–7 with a K2CO3 solution at
1 mol Lꢁ1). The reaction mixture was stirred for 4 h at a tempera-
2.3. Synthesis of the DVB-DOH resin (Scheme 1)
Polymerization was carried out in a 250 mL round-buttomed
flask under mechanical agitation (200 rpm). 0.4 g of PVA were dis-
solved in 80 mL of deionized water at 80 °C for 4 h then 0.4 g
(6.8 mmol) of NaCl were introduced and the reaction mixture
was placed under nitrogen atmosphere (purge of 15 min). Prior
to polymerization, the mixture of 10 mL of monomers (composi-
tion detailed in Table 2) was extracted with 10% NaOH solution
and rinsed with ultra pure water. After addition of 0.1 g (6.1 mmol)

J. Bernard et al. / Reactive & Functional Polymers 72 (2012) 98–106
101
of AIBN to the organic mixture (10 mL of toluene and 10 mL of
monomers), it was introduced in the polymerization flask under
nitrogen atmosphere. The polymerization was run at 80 °C for
6 h. The DVB-DMS resin beads were collected by filtration and
washed with hot water. They were then extracted in a Soxhlet
24 h with water and 24 h with ethanol and then dried under vac-
uum for 3 days at room temperature (conversion rates varying
from 90% to 98%).
3.5 g of DVB-DMS resin beads were suspended in 50 mL of
anhydrous dichloromethane under low mechanical agitation
(100 rpm). The resin beads were placed at 0–5 °C under nitrogen
atmosphere. 2.8 mL (29.1 mmol) of BBr3 were added. After 7 h,
the reaction mixture was poured slowly into ultra pure iced water.
The resin beads of DVB-DOH were collected by filtration and ex-
tracted in a Soxhlet 24 h with water and 24 h with ethanol and
then dried under vacuum for 3 days at room temperature.
(DVB-DOH of Scheme 1) were prepared in two steps. First, 3,4-
dimethoxystyrene was copolymerized in suspension with divinyl-
benzene using toluene as an inert diluent. The toluene is used to
generate porous materials with a high surface area since toluene
has a good thermodynamic compatibility with the polymer net-
work and will cause late phase separation of solvent porogen dur-
ing the copolymerization [35]. In a second step, catechol functions
were deprotected using boron tribromide. In a previous paper, the
synthesis and characterization of these DVB-DOH resins were de-
scribed [36] so it will not be detailed in this article.
As the grafting reactions occur in heterogeneous phase, it com-
plicates both the reaction kinetics and the characterization of the
grafted sorbents (insoluble in all organic solvents). Synthesis of a
model compound enables the optimization of the experimental
conditions. Moreover, since the molecule can be characterized by
classical techniques in organic chemistry (such as NMR), it can
be used as a reference for the characterization of the grafted resin.
4-Ethylaniline was selected to prepare the model molecules for its
primary amine function which is necessary for the synthesis of an
imine or a diazo link. Moreover, its ethylbenzene synthon brings
the aromatic character specific to AmberliteÒ XAD-4 and its ethyl
group can be easily located on 1H NMR spectra allowing easy signal
integration.
2.4. Instrumentation and resins characterization
1H NMR analysis was performed on a Bruker AVANCE 400 NMR
spectrometer using DMSO-d6 as solvent.
The chemical composition of resins was confirmed by FTIR
spectroscopy (Nicolet, Nexus). It was performed in transmission
mode on KBr pellets (32 scans, resolution 4 cmꢁ1).
The sample pyrolysis at 750 °C during 15 s under He atmo-
sphere and the coupling with a gas phase chromatograph/mass
spectrometer (HP 6890 (GPC)/HP5972A (MS), HEWLETT–PACK-
3.1.1. Resin I
3.1.1.1. 4-[(4-ethylphenylamino)methyl]benzene-1,2-diol. The model
molecule 4-[(4-ethylphenylamino)methyl]benzene-1,2-diol was
synthesized according to the synthesis Scheme 2. 1H NMR allowed
the validation of each synthesis step. FTIR analysis of 4-[(4-ethyl-
phenylimino)methyl]benzene-1,2-diol and 4-[(4-ethylphenylami-
no)methyl]benzene-1,2-diol (Fig. 1) was performed in order to
determine the characteristic bands that prove chemical modifica-
tion and to help interpretation of FTIR spectra for grafted sorbent.
Imine formation was assessed by the absence of the C@O vibration
band of 3,4-dihydroxybenzaldehyde (3,4-DHB) at 1648 cmꢁ1, by an
intense band at 1574 cmꢁ1 (vibration band of aromatic C@C en-
hanced by conjugation with C@N bound of imine) and by bands
at 1337 and 1282 cmꢁ1 (vibration bands of CAO and OAH of
catechol). The specific band of C@N bound of aromatic imine ap-
ARD, CP-SIL 8CB 30 m ꢂ 0.25
lm column) was performed on frac-
tion product (0.5 mg) placed into quartz tubes.
Nitrogen adsorption–desorption isotherms at 77 K were deter-
mined with a Micrometrics ASAP 2010 apparatus. Before the
adsorption experiment the samples were evacuated several hours
at a pressure lower than 10ꢁ3 Pa and a temperature of 50 °C.
Hydroxyl functions of catechol resins were determined by back
titration (pH meter Methrom 702). 50 mg of resin beads were
swelled with 200lL of ethanol and then put in a NaOH solution
(5.10ꢁ3 mol Lꢁ1, 50 mL) during 72 h under orbital agitation. Excess
NaOH was titrated by a hydrochloric acid solution (2.10ꢁ3 mol Lꢁ1).
Amine titration was performed in the same way using perchloric
acid (5.10ꢁ3 mol Lꢁ1, 50 mL) which excess was titrated by NaOH
solution (5.10ꢁ3 mol Lꢁ1). Each titration was repeated three times.
Water uptake ratios of resins were determined by thermogravi-
metrical analysis (TGA Q600, TA Instrument). The beads were put
in water during 72 h (ambient temperature and orbital agitation),
dried under atmospheric conditions during 24 h and then analyzed
by TGA (isotherm at 110 °C during 1 h).
(a)
The quantities of metal ions retained per gram of resins were
determined in batch experiments as follow: 50 mg of resins were
put in contact with 100 mL of a metal ion solution (5 mg Lꢁ1) dur-
ing 14 h. The remaining metal concentrations in the solution were
measured by ICP-AES (ICP Vista MPX Axial EL 02024632, SPS 5 FL
02014805, AA 800S/N 94121123) with following parameters:
1648
(b)
power 1.2 kW, pump speed 15 mL minꢁ1
13.5 L minꢁ1
, plasma flow rate
1609
(c)
1574
1337
.
1282
1277
3. Results and discussion
3.1. Synthesis and chemical characterization of the resins
AmberliteÒ XAD-4 resin was used for grafting of catechol. It is a
macroporous poly(styrene-co-divinylbenzene) commercial resin
used for solid-phase extraction applications [34]. The grafting of
catechol on AmberliteÒ XAD-4 resin was performed via an imine
reduced bridge (Resins I of Scheme 1) and a diazo bridge (Resin
II of Scheme 1). Poly(vinylcatechol-co-divinylbenzene) resin beads
2 000
1 500
1 000
500
Wavenumbers (cm-1)
Fig. 1. Transmittance FTIR spectra of 3,4-DHB (a), 4-[(4-ethylphenylimi-
no)methyl]benzene-1,2-diol (b) and 4-[(4-ethylphenylamino)methyl]benzene-1,2-
diol (c) (KBr pellets).

102
J. Bernard et al. / Reactive & Functional Polymers 72 (2012) 98–106
peared simultaneously with aromatic C@C band at 1609 cmꢁ1. The
reduction of the imine function was assessed by the important
intensity decrease of the band at 1574 cmꢁ1 (due to the conjuga-
tion loss between aromatic C@C and C@N bound of imine) and
by the appearance of the stretching vibration band of the CAN
(a)
(b)
(c)
bound of a secondary aromatic amine at 1277 cmꢁ1
.
3.1.1.2. Resin I. AmberliteÒ XAD-4 was activated by the introduc-
tion of an amine function by successive steps of nitration and
reduction as previously described [33]. The coupling with 3,4-
DHB and the following reduction were realized in similar condi-
tions as for the synthesis of the model compound (Scheme 2).
Commercial and modified resins were characterized by Py-GC–
MS. The degradation products are shown in Table 1. They are in
agreement with the poly(ethylvinylbenzene-co-divinylbenzene)
structure of AmberliteÒ XAD-4. The activation by the amine func-
tion was confirmed by the presence of aminated aromatic com-
pounds after pyrolysis. Catechol and derivatives were found as
degradation products of Resin I.
1619
(d)
(e)
(f)
1290
1587
1380
1621
1605
The resin was characterized by FTIR at each grafting step
(Fig. 2). The activation of AmberliteÒ XAD-4 by ANH2 groups was
assessed by appearance of a band at 1619 cmꢁ1 and disappearance
of ANO2 bands (1529 and 1346 cmꢁ1) introduced during the first
step of nitration. The imine formation by coupling of AmberliteÒ
XAD-4 functionalized by ANH2 with 3,4-DHB was evidenced by
the intense band at 1587 cmꢁ1 corresponding to the vibration band
of aromatic C@C enhanced by conjugation with the C@N imine
bound: this band was already observed for the formation of the
imine model compound. As for the molecule, the vibration bands
of CAO and OAH of catechol were present at 1380 and
1290 cmꢁ1. The reduction of the imine bridge was assessed by
the decrease of the band at 1587 cmꢁ1 (lost of conjugation be-
tween C@C and C@N bounds) and the intense band at 1621 cmꢁ1
(aromatic C@C stretching vibration band and NAH deformation
vibration band).
1457
1267
1457
1 500
2 000
1 000
500
Wavenumbers (cm-1)
Fig. 2. Transmittance FTIR spectra of AmberliteÒ XAD-4 (a), AmberliteÒ XAD-4
bearing an amine function (b), Resin I before reduction (c), Resin I (d), Resin II before
deprotection (e) and Resin II (f) (KBr pellets).
been and Tan [37]. These authors proved that guaiacol (2-methoxyphe-
nol) could react with 4-methylbenzene diazonium cation with a good
yield to give 2-methoxy-4-[(4-methylphenyl)azo]phenol. The methoxy
group of this compound was further deprotected by boron tribromide
to recover catechol functions. As described in Scheme 3, we applied the
same synthesis scheme to synthesize 4-[(4-ethylphenyl)diazenyl]ben-
zene-1,2-diol in order to characterize it by FTIR (Fig. 3).
The formation of the diazo bridge induces an intense C@C aro-
matic stretching band at 1594 cmꢁ1 because of the superimposi-
tion of the N@N stretching band. Characteristic bands of guaiacol
are also found in 2-ethoxy-4-[(4-methylphenyl)azo]phenol spec-
trum: the CAO stretching band of phenol at 1265 cmꢁ1 and the
CAH deformation band of OCH3 group at 1457 cmꢁ1. This last band
disappears after the deprotection by cleavage of the methoxy
group. This synthesis step is also assessed by the appearance of
the 1322 cmꢁ1 band of the combination of OAH deformation vibra-
tion and CAO stretching vibration.
3.1.2. Resin II
3.1.2.1. 4-[(4-ethylphenyl)diazenyl]benzene-1,2-diol. The model molecule
corresponding to Resin II cannot be synthesized by direct diazoic coupling
of 4-ethylbenzene diazonium cation with catechol as evidenced by Hagh-
Table 1
Identification of degradation products obtained by Py–GC–MS.
Resins
Degradation products
AmberliteÒ XAD-4
Toluene
Styrene
1-methylmethylbenzene
1,4-divinylbenzene
1-methyl-2-prop-2-
enylbenzene
AmberliteÒ XAD-4 bearing an amine
function
Aniline
2-ethylaniline
2-ethyl-6-methylaniline
4-methylaniline
3.1.2.2. Resin II. Resin II was prepared according to the same syn-
thesis scheme as its corresponding model compound (Scheme 3).
It was characterized by Py-GC–MS before and after deprotectrion
by boron tribromide (Table 1). The coupling with guaiacol induced
the presence of this compound as a degradation product of Resin II
before deprotection. 3-methylcatechol was found as a rearrange-
ment product of degraded Resin II, proving the efficiency of the
deprotection step.
FTIR spectra of Resin II before deprotection and Resin II are
shown in Fig. 2. As for the model molecule, an intense band at
1605 cmꢁ1 is characteristic of the superimposition of the N@N
stretching band with a C@C aromatic stretching band. The band
at 1457 cmꢁ1 already present in the AmberliteÒ XAD-4 aminated
spectrum is amplified by the CAH deformation band of the guaia-
Resin I
Catechol
3-methylcatechol
4-methylcatechol
Aniline
4-ethylaniline
2-ethylaniline
Resin II before deprotection
Resin II
Guaiacol
Aniline
4-ethylaniline
3-methylcatechol
Aniline
4-ethylaniline

J. Bernard et al. / Reactive & Functional Polymers 72 (2012) 98–106
103
On the contrary, a decrease of the surface area and the pore vol-
ume is observed upon grafting by comparing AmberliteÒ XAD-4
surface area and pore volume with those of Resins I and II. The sur-
face areas of the grafted resins are around 600 m2 gꢁ1 when the
XAD-4 is 865 m2 gꢁ1 and the pore volume decrease from
1.2 cm3 gꢁ1 for XAD-4 to 0.8 cm3 gꢁ1 for the Resin I and II. This can-
not be attributed to a post-crosslinking reaction due to the use of a
Lewis acid (tin chloride during the amination step of AmberliteÒ
XAD-4 or boron tribromide during the deprotection step of guaia-
col for Resin II). Such a species can indeed catalyze a Friedel-Craft’s
alkylation reaction between residual double bonds of divinylben-
zene and aromatic rings of the polymer backbone [40]. This reac-
tion surely occurs but is not responsible for the modification of
surface area since it has been proved that it should enhance it by
some shrinkage of the walls between pores [41]. More probably,
some connections between the pores might be obstructed during
grafting steps thus resulting in a decrease of the surface area and
the pore volume.
(a)
(b)
1622
1597
(c)
(d)
843
865
1457
1594
1265
1322
From the shape of the adsorption branch of the isotherm, it
seems that microporosity is present. The microporous volumes
were calculated using the t-plot from the de Boer method. In all
cases, this volume is about 0.08 cm3 gꢁ1 so this microporosity will
not influence the sorption properties.
Comparing the pore size distributions obtained by applying the
BJH model to the desorption branch (assumed to be closer to the
equilibrium state) on the different resins, a small shift of the pore
diameter is observed. The average pore diameter calculated from
r = 2Vp/SBET gives 8 nm for XAD-4 and 11 nm is observed for Resin
I although no change is noticed for Resin II (Fig. 5). Nevertheless
the pore sizes are in the same range. Due to the grafting the porous
structure might be more rigid resulting in the decrease of the sur-
face area and porous volume resulting in less accessibility.
2 000
1 500
1 000
)
500
Wavenumbers (cm-1
Fig. 3. Transmittance FTIR spectra of 4-ethylaniline (a), 2-methoxyphenol (b), 2-
methoxy-4-[(4-ethylphenyl)diazenyl]phenol and 4-[(4-ethylphenyl)diazenyl]ben-
zene-1,2-diol (d) (KBr pellets).
col OCH3 group. Moreover, the CAO stretching band of phenol is
observed at 1267 cmꢁ1. Deprotection was difficult to assess by
FTIR: it only induces a slight decrease of the band at 1457 cmꢁ1
(because of the superimposition with the C@C band).
3.2. Resins structure characterizations
3.2.1. Quantification of AOH and ANHA functions
3.3. Extraction experiments
The evaluation of the amount of amine and catechol functions
was carried out by back titration. The resins were first swollen
by a small amount of ethanol before putting them in contact with
the titrating aqueous solution [36]. This first step of swelling was
essential since DVB-DOH copolymers were not enough hydrophilic
for direct measurements. DVB-DOH(9/1) copolymer beads, in par-
ticular, float on water surface in agreement with the very small
water uptake ratio measured by TGA (Table 2).
The comparison of the amount of amine and catechol functions
for Resin I indicates a high yield of grafting of 3,4-DHB on the ami-
nated XAD-4 resin. However, catechol is more incorporated in
Resin II (diazo bridge). For the DVB-DOH copolymer series, DVB-
DOH(5/5) presents, as expected, the highest content of catechol
units.
Catechol is known for its chelating properties toward some me-
tal ions thanks to its O,O-donating system: it will act as a bidentate
ligand. Complex formation is in competition with protonation and
will therefore be more efficient in basic medium. The affinity of a
polystyrenic resin functionalized by catechol with Pb(II), Cu(II),
Ni(II) and Cd(II) ions has been previously studied and it was shown
that the optimum pH for retention of these ions was 8.0 [33]. In the
present work, the adsorption of those metals for the prepared res-
ins was studied at pH 8.0 under steady state conditions in two dif-
ferent domains of concentration: diluted with metallic solutions of
50 l
g Lꢁ1 or concentrated, at 5 mg Lꢁ1. However, for copper at this
latter concentration, pH was fixed at 6.5 to avoid precipitation of
hydroxide species. The retention properties of AmberliteÒ XAD-4
were also determined for comparison. The adsorption in the low
concentration range is used to calculate a distribution coefficient
characteristics of the affinity, whereas at high concentration, the
capacity is evaluated.
For diluted solutions, the amount of catechol in the beads is in
large excess compared to the metal quantities and the distribution
coefficient Kd was calculated by Eq. (1):
3.2.2. Resins pore structure
Nitrogen adsorption measurements were performed on the dif-
ferent resins. For grafted resins, the adsorption–desorption iso-
therms are of type IV [38] (Fig. 4) as for DVB-DOH copolymers,
indicating that those solids are mainly mesoporous according to
the IUPAC nomenclature. The surface areas were determined from
nitrogen adsorption isotherms by using the Brunauer, Emmet and
Teller (BET) treatment [39] and are given in Table 1.
Since adsorption and desorption curves are not parallel, it is
likely that the mesopores are interconnected or have irregular
shapes. The return of the desorption branch on the adsorption is
not perfect as often observed with organic adsorbents because of
the slow diffusion of nitrogen between polymer chains.
The previous results for copolymers DVB-DOH proved that the
use of boron tribromide during the deprotection step induces a
post-crosslinking reaction leading to an increase of the surface area
[36].
ꢀ
ꢁ
C0 ꢁ Cf
V
m
Kd
¼
ꢃ
ð1Þ
Cf
where C0 and Cf represent the initial and final concentrations of me-
tal ions (
[42].
l
g Lꢁ1), respectively and m is the mass of dried beads (g)
Fig. 6 represents log(Kd) for the different metal ions and resins.
For lead and copper, the best sorbent is AmberliteÒ XAD-4, whereas
it is copolymer DVB-DOH(9/1) for nickel and cadmium. AmberliteÒ
XAD-4 and DVB-DOH(9/1) are the resins presenting the largest sur-

104
J. Bernard et al. / Reactive & Functional Polymers 72 (2012) 98–106
Table 2
Composition of DVB-DOH copolymers, quantification of NH and catechol functions, water uptake ratio, surface area determined by BET (SBET), porous volume and average pore
diameter.
Resin
AmberliteÒ
XAD-4
Resin I
Resin II
–
DVB-
DOH
(9/1)
DVB-
DOH
(7/3)
DVB-DOH
(5/5)
Composition of the monomer mixture for the preparation
of DVB-DOH copolymers
DVB (mL)
DMS (mL)
–
–
9
1
–
0.27
1.5
658
0.62
3.8
7
3
–
0.40
2.8
464
0.38
3.3
5
5
–
0.61
4.9
0
Quantity of NH function (mmol gꢁ1
)
–
–
0.6
865
1.19
8
0.65
0.61
8.4
600
0.84
11
–
Quantity of catechol function (mmol gꢁ1
Water uptake ratio (%)
)
1.38
7.5
577
0.82
7
SBET (m2 gꢁ1
)
Porous volume (cm3 gꢁ1
)
Non-mesoporous material
Non-mesoporous material
Average pore diameter (nm)
For more concentrated solutions, the sorption capacities CS (in
910
l
mol gꢁ1) were calculated by Eq. (2):
810
Amberlite XAD-4
710
ðC0 ꢁ Cf Þ
V
m
Resin I
CS ¼
ꢃ
ð2Þ
610
M
Resin II
510
where C0, Cf and m were previously described, M is the molar mass
of the considered metal (g molꢁ1) and V is the volume of the metal-
lic solution (L).
410
310
210
110
10
Sorption capacities for the different metal ions are presented in
Fig. 7. In these conditions, the best results are obtained either by
Resin II for nickel, by Resin II and DVB-DOH(7/3) for copper or by
DVB-DOH(7/3) for cadmium. Lead seems to be indifferent to the
sorbent. In comparison with the results in diluted medium, this
proves the chelating role of catechol at higher concentration range
since the highest sorption capacities are not found for resins with
the largest surface areas but with the highest amounts of catechol
entities. DVB-DOH(5/5) is an exception most probably because it is
a non-mesoporous material. In the series of DVB-DOH copolymers,
it clearly appears that the best sorbent is DVB-DOH(7/3). It is cer-
tainly a good compromise between the quantity of ligand and the
porous structure. It can be noticed that nickel might also interact
with the nitrogen atoms of the diazo link of Resin II.
0,8
1
0
0,2
0,4
0,6
P/P°
Fig. 4. N2 adsorption- desorption isotherms at 77 K of AmberliteÒ XAD-4, Resin I
and Resin II.
1,2
1
0,3
0,25
0,2
0,15
0,1
0,05
0
Amberlite XAD-4
Resin I
Resin II
Amberlite XAD-4
Resin I
0,8
0,6
0,4
0,2
0
Resin II
4,5
4
3,5
3
0
5
10
15
20
D (nm)
2,5
Kd
Fig. 5. Pore size distributions and cumulative pore volume calculated by the BJH
method applied on the desorption branch for AmberliteÒ XAD-4, Resin I and
Resin II.
2
1,5
1
face areas and the lowest amount of catechol. This could indicate
that for low metal ion concentrations, the determining factor is
not the ligand proportion but structural factors. Metal retention
would result from non specific interactions or slow diffusion inside
the porous structure. The high level of adsorption by DVB-DOH(5/5)
which has zero surface area as measured in a dry state by nitrogen
adsorption, indicates that the porosity is opened by swelling in
aqueous solution.
0,5
0
Pb
Cu
Ni
Cd
Fig. 6. Evolution of the distribution coefficient (Kd).

J. Bernard et al. / Reactive & Functional Polymers 72 (2012) 98–106
105
120
100
80
60
40
20
0
The retention properties of the grafted resins were compared to
those of poly(vinylcatechol-co-divinylbenzene) copolymers. Those
materials were obtained in two steps by copolymerization of divi-
nylbenzene with dimethoxystyrene followed by deprotection of
the methoxy groups. The extraction experiments at low metal con-
centrations evidenced that retention is due to non specific interac-
tions or mechanical retention inside the porous polymer backbone.
It can even be concluded that, in such conditions, AmberliteÒ XAD-
4 resin is sufficiently efficient. Nevertheless, for more concentrated
solutions, catechol is mostly responsible for metal extraction. It
was evidenced that best results were obtained for resins contain-
ing the highest amount of catechol providing that the material is
mesoporous.
This work also led to the conclusion that functionalized resins
obtained by the copolymerization route were equivalent or more
efficient for metal retention than grafted resins. These copolymers
are easier to prepare, less time consuming and more eco-friendly
since their preparation implies less toxic reagents. Moreover, their
synthesis is more reproducible and leads to non colored materials.
This proves that direct copolymerization is a more attractive route
for functionalized materials on the condition that functionalized
monomers are easily accessible.
Pb
Cu
Ni
Cd
mol/g).
Fig. 7. Evolution of the sorption capacity (CS in
l
Acknowledgement
Authors thank the Analysis Laboratory of the Var Department
(Toulon branch) for ICP-AES measurements carried out during this
plan and PACA Region for financial support.
References
[1] S. Dubinsky, J.I. Park, I. Gourevich, C. Chan, M. Deetz, E. Kumacheva,
Macromolecules 42 (2009) 1990–1994.
[2] N. Fontanals, P. Cormack, D. Sherrington, J. Chrom. A 1215 (2008) 21–29.
[3] C. Toro, R. Rodrigo, J. Cuellar, React. Func. Polym. 68 (2008) 1325–1336.
[4] E.S. Dragan, E. Avram, D. Axente, C. Marcu, J. Polym. Sci. A Polym. Chem. 42
(2004) 2451–2461.
(a)
(b)
(c)
(d)
Fig. 8. Colour variation of DVB-DOH(9/1) resin upon complexation with varying
lead concentration solutions: (a) 50 (b) 100 (c) 500 (d)
g Lꢁ1 g Lꢁ1 g Lꢁ1
1000
g Lꢁ1
l
,
l
,
l
,
l
.
[5] R. Albright, React. Polym. Ion Exchangers Sorbents 4 (1986) 155–174.
[6] B. Corain, M. Zecca, K. Jerabek, J. Mol. Catal. A Chem. 177 (2001) 3–20.
[7] P. Barbaro, F. Liguori, Chem. Rev. 109 (2009) 515–529.
[8] N. Fontanals, R.M. Marcé, M. Galià, F. Borrull, J. Polym. Sci. A Polym. Chem. 41
(2003) 1927–1933.
[9] J. Fritz, P. Dumont, L. Schmidt, J. Chrom. A 691 (1995) 133–140.
[10] H.P. Hentze, M. Antonietti, Rev. Mol. Biotech. 90 (2002) 27–53.
[11] S. Gupta, P. Benien, P. Sahoo, Curr. Drug Delivery 7 (2010) 252–262.
[12] R.J. Booth, J.C. Hodges, Acc. Chem. Res. 32 (1999) 18–26.
[13] A. Guyomard, D. Fournier, S. Pascual, L. Fontaine, J.-F. Bardeau, Eur. Polym. J. 40
(2004) 2343–2348.
[14] V. Camel, Spectrochim. Acta Part B 58 (2003) 1177–1233.
[15] A.N. Pustam, S.D. Alexandratos, React. Func. Polym. 70 (2010) 545–554.
[16] A. Ramesh, K. Ramamohan, K. Seshaiah, Talanta 57 (2002) 243–252.
[17] M.S. Hosseini, A. Hosseini-Bandegharaei, H. Raissi, F. Belador, J. Hazard. Mat.
169 (2009) 52–57.
[18] M.E. Leon-Gonzalez, L.V. Perez-Arribas, J. Chrom. A 902 (2000) 3–16.
[19] P. Metilda, K. Sanghamitra, J.M. Gladis, T.P. Rao, Talanta (2004).
[20] G. Depecker, C. Branger, A. Margaillan, T. Pigot, S. Blanc, F. Robert-Peillard, B.
Coulomb, J.-L. Boudenne, React. Func. Polym. 69 (2009) 877–883.
[21] J.-L. Boudenne, S. Boussetta, C. Brach-Papa, C. Branger, A. Margaillan, F.
Théraulaz, Polym. Int. 51 (2002) 1050–1057.
[22] M.V. Dinu, E.S. Dragan, React. Func. Polym. 68 (2008) 1346–1354.
[23] D. Türkmen, E. YIlmaz, N. Öztürk, S. Akgöl, A. Denizli, Mater. Sci. Eng. C 29
(2009) 2072–2078.
[24] R. Beauvais, S. Alexandratos, React. Func. Polym. 36 (1998) 113–123.
[25] S. Köytepe, S. Erdogan, T. Seçkin, J. Hazard. Mat. 162 (2009) 695–702.
[26] M. Maciejewska, J. Gawdzik, J. Liq. Chromatogr. Rel. Technol. 31 (2008) 950.
[27] L.G. Sillén, A.E. Martell, Stability Constants of Metal-Ion Complexes, The
Chemical Society, Burlington House, London, 1964. pp. 472.
[28] V. Nurchi, T. Pivetta, J. Lachowicz, G. Crisponi, J. Inorg. Biochem. 103 (2009)
227–236.
An important conclusion from this study is that the retention
properties of copolymer DVB-DOH(7/3) are equivalent or superior
to those of the grafted resins. However, this copolymer was ob-
tained in only two synthesis steps compared to four for the grafting
(two for the activation by an amine group and two for the incorpo-
ration of catechol). In particular, the use of the sulfonitric toxic
mixture during the nitration step of AmberliteÒ XAD-4 step could
be avoided. Moreover, it prevented the resins from getting the dark
coloration linked to that nitration step. DVB-DOH copolymers have
a pale yellow color which changes upon complexation (Fig. 8).
4. Conclusions
In the present paper, we introduced catechol on a poly(styrene-
co-divinylbenzene) resin (AmberliteÒ XAD-4) by grafting. To
optimize the synthesis yields and help characterization of the func-
tionalized resins, the synthesis scheme was first applied on a mod-
el molecule. Grafting by two different bridges (a reduced imine and
a diazo) was evidenced by Py-GC–MS and FTIR analysis. Acido-
basic back titration of catechol units was performed to prove the
incorporation of this group. The results obtained by adsorption/
desorption of nitrogen to study the porous structure of the resins
revealed a decrease of the surface area and the pore volume indi-
cating that some connections between mesopores must be blocked
during the grafting procedures. However the swelling of the mate-
rial in presence of solvent is not taken into account and could affect
both adsorption and diffusion properties.
[29] M.J. Sever, J.J. Wilker, Dalton Trans. (2004) 1061.
[30] G. Westwood, T.N. Horton, J.J. Wilker, Macromolecules 40 (2007) 3960–3964.
[31] L. Li, Y. Li, X. Luo, J. Deng, W. Yang, React. Funct. Polym. 70 (2010) 938–943.
[32] A.R. Statz, R.J. Meagher, A.E. Barron, P.B. Messersmith, J. Am. Chem. Soc. 127
(2005) 7972–7973.
[33] J. Bernard, C. Branger, T.L.A. Nguyen, R. Denoyel, A. Margaillan, React. Func.
Polym. 68 (2008) 1362–1370.

106
J. Bernard et al. / Reactive & Functional Polymers 72 (2012) 98–106
[34] W.-lu Song, Z.-liang Zhi, L.-sheng Wang, Talanta 44 (1997) 1423–1433.
[35] F.S. Macintyre, D.C. Sherrington, Macromolecules 37 (2004) 7628–7636.
[36] J. Bernard, C. Branger, I. Beurroies, R. Denoyel, S. Blanc, A. Margaillan, Polymer
51 (2010) 2472–2478.
[39] S. Brunauer, P.H. Emmett, E. Teller, J. Am. Chem. Soc. 60 (1938) 309–319.
[40] V.A. Davankov, M.P. Tsyurupa, React. Polym. 13 (1990) 27–42.
[41] K. Aleksieva, J. Xu, L. min Wang, A. Sassi, Z. Pientka, Z. Zhang, K. Jerabek,
Polymer 47 (2006) 6544–6550.
[37] K. Haghbeen, E.W. Tan, J. Org. Chem. 63 (1998) 4503–4505.
[38] K.S.W. Sing, D.H. Everett, R.A.W. Haul, L. Moscou, J. Rouquerol, T.
Siemieniewska, Pure Appl. Chem. 57 (1985) 603–619.
[42] M. Andaç, R. Say, A. Denizli, J. Chrom. B 811 (2004) 119–126.