Biomacromolecules 2005, 6, 189-195
189
Complexation Models of N-(2-Carboxyethyl)chitosans with Copper(II) Ions Yury A. Skorik,*,† Carlos A. R. Gomes,† Nina V. Podberezskaya,‡ Galina V. Romanenko,‡ Luiz F. Pinto,† and Yury G. Yatluk§ LAQUIPAI, Departamento de Quı´mica, Faculdade de Cieˆ ncas, Universidade do Porto, Rua do Campo Alegre 687, 4169-007 Porto, Portugal, Institute of Inorganic Chemistry, Siberian Division of the Russian Academy of Sciences, Acad. Lavrentyeva Avenue 3, 630090 Novosibirsk, Russian Federation, and Institute of Organic Synthesis, Urals Division of the Russian Academy of Sciences, S. Kovalevskoi Street 20, 620219 Ekaterinburg, Russian Federation Received July 13, 2004; Revised Manuscript Received September 20, 2004
The copper(II) complex formation equilibria of N-(2-carboxyethyl)chitosans with three different degrees of substitution (DS ) 0.42, 0.92, and 1.61) were studied in aqueous solution by pH-potentiometric and UVspectrophotometric techniques. It was demonstrated that the complexation model of CE-chitosans depends on DS: the [Cu(Glc-NR2)2] complexes are predominant for two lower substituted samples (“bridge model”, log β12 ) 10.06 and 11.6, respectively), whereas the increase of DS leads to formation mainly of the [Cu(Glc-NR2)] complexes (“pendant model”, log β11 ) 6.41). As a model for copper complexation with a disubstituted residue of CE-chitosan, the complex of N-methyliminodipropionate [CuMidp(H2O)]‚(H2O) was synthesized and structurally characterized by XRD. The unit cell consists of two crystallographically nonequivalent Cu atoms having slightly distorted square pyramidal coordination; Midp constitutes the basal plane of the pyramid and acts as a tetradentate NO3 chelate-bridging ligand by the formation of two sixmembered chelate rings (average Cu-O 1.99 Å, Cu-N 2.04 Å) and a bridge via carbonyl O atom (average Cu-O 1.99 Å), an apical position is occupied by a water molecule (average Cu-Ow 2.30 Å). 1. Introduction Chitosan, the deacetylated chitin derivative, has been the object of continuous study during the last few decades, and recent review articles outline much of the broad ranging research on this polymer to date.1-6 In particular, its metal complex formation and sorption properties7-9 have attracted much attention as a fundamental knowledge for selective removal and recovery of metal ions and/or radionuclides from aqueous solutions in a number of industrial applications, including mining, metallurgy, site remediation, water decontamination, and laboratory analysis. The complex-forming properties of chitosan have been compared for a large number of metal ions.10,11 However, the precise binding mechanism and, in particular, the local molecular geometry of complexation with the amino group is still being debated. Two conflicting models are currently being considered. In the “pendant model” it is proposed that a single metal ion is attached as a pendant to an amino group of chitosan.12,13 In the “bridge model”, the metal ions are believed to be coordinated to several amino groups originating either from the same or from different polymer chains.14 Though the complex-forming ability of chitosan is com* To whom correspondence should be addressed. E-mail:
[email protected];
[email protected]. Current address: Chemistry Department at Carnegie Mellon University, 4400 Fifth Avenue, Pittsburgh, PA 15213. † Universidade do Porto. ‡ Institute of Inorganic Chemistry. § Institute of Organic Synthesis.
Scheme 1. Structure of N-(2-Carboxyethyl)chitosan
pared favorably with that of other natural polysaccharides, the complexation selectivity has not been attained to a satisfactory level. Chemical modifications offer a wide spectrum of tools to enhance chelating properties, selectivity, and metal ion capacity of chitosan.2,7 It is known that the β-alanine derivatives such as iminodipropionic acid (Idp)15 or Idp-like ligands16-18 are more selective to copper(II) ions than well-known iminodiacetate (Ida) analogues or corresponding amines. To improve selectivity and metal ion capacity of chitosan, we have recently prepared a novel polysaccharide chelator, namely N-(2-carboxyethyl)chitosan (CE-chitosan, Scheme 1), by site-selective modification of chitosan with 3-halopropionic acids under mild alkaline conditions.19 It has also been found that novel CE-chitosan has many other useful features, such as water solubility in broad pH range,19 excellent antioxidant and significant antimutagenic activity,20 and better biodegradability than chitosan itself.21 CE-chitosan appears suitable for mediating transport of hydrophilic drugs such as vitamin B6 through the skin22 or as a nitric oxide carrier for use in a variety of medical applications, in which an effective dosage of nitric oxide is indicated as a preferred method of treatment.23
10.1021/bm049597r CCC: $30.25 © 2005 American Chemical Society Published on Web 11/19/2004
190
Biomacromolecules, Vol. 6, No. 1, 2005
Thus, we predict that the novel CE-chitosan due to its β-alanine chelate function will form stronger metal complexes and will have better selectivity for copper ions than chitosan itself. The novel polymer could have wide-ranging application in the separation and preconcentration of copper and other heavy metal ions, whereas metal complexes of CEchitosan have a potential to be used as a stationary phase in immobilized metal affinity chromatography for bio-separation and purification of peptides, proteins etc. In this paper, we present an investigation of the complex formation between copper(II) ions and CE-chitosans with different degrees of substitution (DS) in aqueous solutions by pH-potentiometry and UV-spectroscopy. Furthermore, as a model for copper(II) complexation with Idp-glucan residue of CE-chitosan, the copper(II) complex with N-methyliminodipropionic acid (H2Midp), viz. catena-aqua(N-methyliminodipropionato)copper(II) monohydrate [CuMidp(H2O)]‚ (H2O), was synthesized and structurally characterized by X-ray diffraction. 2. Experimental Section 2.1. Materials. The starting material was low molecular weight chitosan from coarse ground crab (Mw ) 150 000, Aldrich 44,886-9), the average degree of acetylation (DA) was 0.07 as determined by 1H NMR. 3-Chloropropionic acid (Fluka 26171), acrylic acid (Aldrich 14,723-0), 8 mol L-1 solution of methylamine in ethanol (Fluka 65590), 0.1000 mol L-1 copper(II) nitrate standard solution (Orion 942906), copper carbonate basic (Aldrich 20,789-6), potassium nitrate (puriss. p.a. Fluka 60419), potassium hydroxide (puriss. p.a. Fluka 60370), deuterium oxide (Aldrich 15,188-2), and concentrated deuterium chloride solution in D2O (Aldrich 54,304-7) were used as purchased. 2.2. General Methods. The 1H NMR spectra of the D2O/ DCl solutions of the CE-chitosan samples at a concentration of 10 mg mL-1 were performed using BRUKER DRX 400 spectrometer at T ) 70 °C with 3-(trimethylsilyl)-1-propanesulfonic acid as an internal standard. The electronic absorption spectra were run on a Techcomp 8500 spectrophotometer at room temperature (25 ( 2 °C) using 1.00 cm quartz cells. The elemental analyses were performed on a Carlo Erba EA 1108 analyzer. 2.3. Preparation of CE-Chitosans. Chitosan was treated with 3-chloropropionic acid in water in the presence of NaHCO3, and isolation and purification of the product were performed as described by Skorik et al.19 Using different ratios of the alkylating agent, 3-chloropropionic acid, to the glucosamine monomeric unit of chitosan (Glc-NH2) and different reaction times, three N-carboxyethylated derivatives with different DS were prepared. The degree of acetylation (DA) and the degree of substitution (DS, degree of carboxyethylation of aminogroup) of the prepared derivatives were determined by 1H NMR spectra. The integral of H-1 protons was considered as an internal standard. The experimental conditions and some characteristics of the prepared CE-chitosan samples are summarized in Table 1. For the sake of simplicity they were termed as samples CEQ1, CEQ2, and CEQ3.
Skorik et al. Table 1. Reaction Conditionsa and Some Parameters of CE-Chitosan Samples sample CEQ1 CEQ2 CEQ3
reagents ratiob 2 5 5×2
DFc
reaction time (h)
ne
x
y
z
DSd
72 72 72 × 2
0.07 0.07 0.07
0.55 0.16 0.0
0.39 0.69 0.35
0.0 0.08 0.57
0.42 0.92 1.61
a T ) 60 °C, pH ) 8-9 (NaHCO ). b 3-Chloropropionic acid/Glc-NH 3 2 residue molar ratio. c DF, degree of functionalization (ratio of functional 1 d groups) was determined from H NMR spectra. DS was calculated per Glc-NH2 residue of chitosan DS ) (y + 2z)/(x + y + z). e Represents degree of acetylation (DA).
2.4. Preparation of N-Methyliminodipropionic Acid H2Midp. It was prepared by aza-Michael 1,4-conjugate addition of acrylic acid to methylamine as follows: 30 mL (0.44 mol) of acrylic acid was added to 20 mL of an 8 mol L-1 solution of methylamine in ethanol (0.16 mol). The mixture was heated at boiling temperature for 5 h. The H2Midp precipitated from the reaction mixture as a white powder; yield 20.6 g (73%). mp ) 154-155 °C (156 °C).16 Found: C 47.94, H 7.35, N 7.99%. Calculated for C7H13NO4: C 47.99, H 7.48, N 8.00%. 1 H NMR (δ, ppm, DMSO-d6): 2.26 (s; 3 H; -CH3), 2.38 (t; 4 H; -CH2-COOH; J ) 7.1 Hz), 2.69 (t; 4 H; dNCH2-; J ) 7.1 Hz). 2.5. Preparation of Aqua(N-methyliminodipropionato)copper(II) Monohydrate Complex. The complex was prepared by mixing a hot saturated aqueous solution of H2Midp with (CuOH)2CO3 until carbon dioxide was removed. After filtration, the resulting solution was maintained at room temperature until evaporation resulted in the formation of deep-blue crystals suitable for X-ray diffraction analysis. Found: C 30.70; H 5.25; N 4.93%. Calculated for C7H15NO6Cu: C 30.83; H 5.54; N 5.14%. 2.6. Potentiometric Titration. A solution of each sample of CE-chitosan (∼3 mmol L-1 of amino groups) was prepared by dissolving a prescribed amount of the polymer in the 0.1000 mol L-1 HNO3 necessary to achieve the complete protonation of the CE-chitosan. After adding a prescribed amount of KNO3 in order to adjust ionic strength to 0.1 mol L-1, the solution was diluted up to a definite volume. Prior to titration, a standard 0.1000 mol L-1 Cu(NO3)2 solution was added to an aliquot of CE-chitosan solution at copper:Glc-NR2 ratio ranging from 0.25 to 2. A carbonate-free 0.1 mol L-1 KOH standard solution used as a titrant was prepared by ion-exchange technique.24 High precision potentiometric titrations were performed with a PC-guided system assembled with a Crison MicropH 2002 pH meter (equipped with an Orion 90-02-00 double junction AgCl/Ag reference electrode and a Russel SWL/ S7 glass electrode) and a Crison MicroBU 2030 micro buret. The titrations were performed under an atmosphere of CO2 and O2-free nitrogen at 25.0 ( 0.1 °C in a Metrohm EA880T double wall glass cell with a Metrohm E649 stirrer. The program was written in QuickBasic (Microsoft Corp., v. 7.0) to control the titration system via an RS232C interface. The glass electrode was calibrated in terms of hydrogen ion concentration with buffer solutions with ionic strength adjusted to 0.1 mol L-1 with KNO3 according to a procedure
Complexation Models of N-(2-Carboxyethyl)chitosans
Biomacromolecules, Vol. 6, No. 1, 2005 191
Figure 1. Potentiometric titration curves of 3 × 10-3 monomol L-1 CE-chitosan solutions for CEQ1 (a), CEQ2 (b), and CEQ3 (c) samples alone and in the presence of Cu(NO3)2; the ratios [Cu]/[Glc-NR2] are shown in the figures. R is the degree of neutralization; titrant KOH 0.1 mol L-1; T ) 25.0 °C; I ) 0.1 mol L-1 KNO3.
recommended by Vasconcelos & Machado.25 The stabilization criteria used for recording the potential is a critical parameter and it was ∆E/∆te 0.1 mV/min, which corresponds to a stabilization time for each point of around 1030 min. All pH titration experiments were run at least in duplicate to ensure reliability of the data. The acid-base and metal-ligand complexation parameters were calculated and refined with the SUPERQUAD program.26 2.7. X-ray Diffraction. The experimental data for a deepblue nonregular hexagonal plate with dimension of 0.66 × 0.34 × 0.20 mm was measured using Smart Apex CCD diffractometer: λ Mo KR ) 0.71073 Å, grafite monochromator, T ) 240 K. Total 8302 and 2755 unique reflections were collected. The ranges of h, k, l were -21 to +22, -7 to +7, and -16 to +14, respectively. The structure was solved by direct methods and refined by the full-matrix least-squares method in an anisotropic approximation for all non-hydrogen atoms and isotropically for H atoms with positions located on a difference electron density map. The final refinement cycles yielded R1 ) 0.0336 and wR2 ) 0.0922 for 2616 reflections with I > 2σ(I), R1 ) 0.0352 and wR2 ) 0.0931 for all 2755 reflections, and goodness of fit ) 1.079. All calculations were performed with the SHELX-97 program package (Sheldrick G. M. 1997. University of Go¨ttingen, Germany). Crystal data: C7H15NO6Cu (FW 272.74), orthorhombic, space group Pca21, a ) 19.837(9), b ) 7.084(3), c ) 14.896(7) Å, V ) 2093(2) Å3, Z ) 8, Dcalc ) 1.731 g cm-3, µ ) 2.096 mm-1. 3. Results and Discussion 3.1. pH Potentiometric Titration. The results of potentiometric titration of the CE-chitosans in the absence and in the presence of copper(II) ions at different Cu:Glc-NR2 ratios are shown in Figure 1a-c for three CE-chitosan samples CEQ1, CEQ2, and CEQ3 with different degrees of Ncarboxyethylation (DS, Table 1). The experimental data were fitted by the SUPERQUAD program in order to determine the type and the value of the
Table 2. Composition, Notation, and Formation Constants βpqr for the Copper Complexes ligand L CEQ1 CEQ2 CEQ3 Chitosan30 Metglu31 (β-Ala-)32
(Idp2-)15 (Midp2-)16 (Heidp2-)33
a
species (pqr) 011 01na 120 011 01na 120 011 01na 110 011 110 120 011 110 120 011 012 110 120 011 012 110 011 012 110 011 012 110
formula HL H nL CuL2 HL H nL CuL2 HL H nL CuL HL CuL CuL2 HL CuL CuL2 HL H 2L CuL CuL2 HL H 2L CuL HL H 2L CuL HL H 2L CuL
logβpqr 6.72 ( 0.02 9.90 ( 0.05 10.06 ( 0.02 7.02 ( 0.02 10.10 ( 0.07 11.6 ( 0.2 7.65 ( 0.03 10.90 ( 0.08 6.41 ( 0.11 6.0 5.47 8.09 7.689 4.126 7.52 10.14 13.71 6.99 12.45 9.61 13.72 9.4 9.48 13.83 8.05 8.91 12.84 8.4
temperature (°C); I (mol L-1) 25; 0.1(KNO3) 25; 0.1(KNO3) 25; 0.1(KNO3) Not specified 25; 0.15 (NaCl) 25; 0.1-0.2
30; 0.1 (KCl) 30; 0.1 (KCl) 30; 0.1 (KCl)
n is equal to DS + 1.
stability constants of complex species. The stability constants are denoted by βpqr, which correspond to the general notation pCu2+ + qL + rH+ ) [CupLqHr] βpqr ) [CupLqHr]/[Cu2+]p[L]q[H+]r where L represents the deprotonated forms of glucosamine units of CE-chitosan. A value of the water autoprotolysis constant pKw corresponding to logβ00-1 ) 13.86 was assumed for the experimental conditions (T ) 298 K, I ) 0.1 mol L-1).27 The formation constants of copper(II) hydroxocomplexes used in calculations were taken from the bibliography.28,29 The pH range embraced for SUPERQUAD calculations was limited to the first buffer region (pH up to 7). At pH higher than 7, insoluble aggregates or firm gels were formed that led to difficulties in titrant diffusion and poor reproducibility of glass electrode potential.
192
Biomacromolecules, Vol. 6, No. 1, 2005
Skorik et al.
Figure 2. Species distribution diagrams for Cu2+-CE-chitosan systems: CEQ1 (a), CEQ2 (b), and CEQ3 (c) as a function of pH. C(Cu)d C(Glc-NR2) ) 3 × 10-3 mol L-1, T ) 25.0 °C, I ) 0.1 mol L-1 KNO3.
Table 2 lists the most probable complex species together with the corresponding complex formation constants for CEchitosans and some model compounds. The calculated speciation of solutions containing both copper(II) ions and CE-chitosans is illustrated in Figure 2a-c. The protonation constants of CE-chitosans (β011 and β01n) listed in Table 2 are in good agreement with those reported in our previous paper.19 It is easy to observe that the basicity of the amino group of CE-chitosan expressed by β011 increases with the DS.19 The complexation behavior with copper is different for various CE-chitosan samples. In the case of the Cu-CEQ3 system, the model that gave a satisfactory fit to the experimental data was the one, which took into account the CuL species. Otherwise, for samples CEQ1 and CEQ2, it was observed that the best fit was found for the model considering CuL2 complex species. Hydroxide mixed-ligand complexes (as being proposed for copper complexes of chitosan30) have been rejected by SUPERQUAD species selection for all CE-chitosan samples over the pH range tested, generating a model that did not include them. The formation of other complexes could not be found in the systems either because they do not form or perhaps because their concentrations (and consequently their stability constants) are too low for potentiometric determination. Finally, it can be stated that the reported formation constants are reproducible and the statistical parameters are acceptable. A comparison of the distribution curves (Figure 2a-c) reveals that the complex stability is increasing in the following order CEQ1 < CEQ2 < CEQ3 as shown by the complex formation at a somewhat lower pH. Undoubtedly, it is connected with increasing of number of chelate rings formed by a monomeric unit with metal ion. In a recent paper, Rhazi et al.30 have suggested a formation of two copper-chitosan complexes, namely {CuL2+, 2OH-, H2O} and {CuL22+, 2OH-}, and have determined the stability constants of CuL and CuL2 species by pH-potentiometric titration (β110 and β120 respectively, Table 2). The stability constants shown in Table 2 for the copper complexes of CEchitosan samples are higher than those reported for chitosan itself that can be attributed both by increasing the ligand basicity (β011, Table 2) and by the β-alaninate chelate function. In comparison with methyl D-glucosaminide (meth-
Figure 3. Electronic absorption spectra of Cu(NO3)2-CEQ1 system vs pH: C(Cu) ) 5 × 10-5 mol L-1; C(Glc-NR2) ) 1 × 10-4 monomol L-1, T ) 25 °C, I ) 0.1 mol L-1 KCl.
yl 2-amino-2-deoxy-D-glucopyranoside, Metglu), CE-chitosans also form more stable complexes despite the lower basicity of their amino group. Here, the ligand basicity effect, which results in a decrease in the complex stability, is surpassed by the formation of six-membered β-alaninate chelate rings. A comparison with other model ligands such as β-alaninate or Idp-like ligands reveals that the stability constants of the corresponding complexes of CE-chitosans are lower due to the much lower basicity of the amino group (logβ011 values in 2-3 units less), since a decrease in the ligand basicity results in correspondingly lower affinity for metal ions. 3.2. UV Absorption Spectra. UV spectra recorded in aqueous solutions containing CEQ1 and Cu(NO3)2 at various pH are presented in Figure 3 together with a chart illustrating the dependence of absorbance A vs pH at λ ) 269 nm. In the acidic region (pH < 4), the spectra are identical with those of CE-chitosan that revealed no characteristic absorption in the UV region between 200 and 400 nm. Upon increasing pH value, an additional absorption band at 269 nm appeared testifying for the complex formation. The curve A vs pH (Figure 3) has a sigmoid shape that is typical for the formation of only one complex compound. The stoichiometry of the reactions between Cu2+ and CEQ1 was determined by means of the mole ratio method
Complexation Models of N-(2-Carboxyethyl)chitosans
Biomacromolecules, Vol. 6, No. 1, 2005 193
Scheme 2. Metal Coordination Modes of Different Residues of N-(2-Carboxyethyl)chitosana
a
L represents a monodentate ligand such as H2O, Cl-, etc.
Figure 5. Structure and numbering scheme of atoms in [CuMidp(H2O)] molecule. Figure 4. Spectrophotometric titration curve of 1 × 10-4 monomol L-1 CEQ1 solution by Cu(NO3)2: pH ) 6.00 ( 0.05, T ) 25 °C, I ) 0.1 mol L-1 KCl.
by measuring the intensity of the new absorbance at 269 nm as a function of metal concentration (R ranging from 0 to 2). The UV spectra shown in Figure 4 indicate that the absorbance reaches a maximum for R equal to 0.5 that corresponds to the formation of the CuL2 complex. On exceeding this ratio, slight linear increase of adsorption is observed, which is caused by increasing of the overlapped nitrate absorption band at ca. 220 nm. The UV spectra of the Cu2+-CEQ2 system are essentially the same as those of the Cu2+-CEQ1 system, the absorption maximum appears at 267 nm. The study of copper(II) complexes of CEQ3 sample was limited due to the precipitation of the complex formed, which made the spectrophotometric measurements unreliable. The spectra of the copper complexes of CE-chitosans are somewhat different from that observed for copper-chitosan system,30 where one broad absorption band at 257 nm observed at pH 5-5.8 was split into two bands at 246 and 270 nm at higher pH values. Authors have explained this phenomenon by the formation of two complex species with Cu:chitosan ratios of 1:1 and 1:2. The differences in the spectra observed upon coordination of copper(II) ions by chitosan30 and CE-chitosans seem to provide an additional proof of the different types of binding in these two cases. Thus, our results allow us to propose a “bridge model”, which corresponds to CuL2 complexes, as a predominant coordination mode of CE-chitosans with low degree of carboxyethylation (DS
