- Email: [email protected]
or calcium in nickel–pectin interaction by potentiometric and voltammetric techniques
Bioelectrochemistry 77 (2009) 31–36
Contents lists available at ScienceDirect
Bioelectrochemistry j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / b i o e l e c h e m
Competition of copper and/or calcium in nickel–pectin interaction by potentiometric and voltammetric techniques C. Vilhena 1, M.L.S. Gonçalves, A.M. Mota ⁎ Centro de Química Estrutural (Torre Sul), Instituto Superior Técnico, Av. Rovisco Pais, 1049-001 Lisboa, Portugal
a r t i c l e
i n f o
Article history: Received 11 February 2009 Received in revised form 7 May 2009 Accepted 23 May 2009 Available online 30 May 2009 Keywords: Ni(II) Cu(II) Ca(II) Pectin Competition Voltammetry
a b s t r a c t The interaction between nickel and pectin extracted from citrus fruit was studied in 0.10 M KNO3, at pH 5.5 and 25 °C. Differential pulse and/or square wave polarography were used to determine free nickel. For a high coverage degree (θ) of the pectin by the metal ion a good fitting was observed between experimental results and the model that includes both complex species, ML and ML2 (M for the metal ion and L for the ligand). In the ML2 species, Ni(II) interacts with two carboxylate groups of different chains, resulting in an inter-chain association. For low θ values, the formation of ML2 is hindered due to the repulsion between the negative charges of carboxylic groups in two independent segments of pectin. The influence of calcium or copper ions on the free nickel concentration, in the presence of pectin, may lead to a decrease in free nickel concentration, contrary to what would be expected from direct competition between Ca(II) or Cu(II) and Ni(II) for the pectin binding sites. This is due to the partial neutralisation of the negative carboxylic charges by the positive charges of the divalent cations, which favours NiL2 formation through the association of independent chains. © 2009 Elsevier B.V. All rights reserved.
1. Introduction Partially esterified polygalacturonic acid (Fig. 1) is the main component of pectins in higher plants [1], and is an important constituent of the apoplastic transport system. Polygalacturonic acid (PGA) acts as a selective filter of nutritive elements and regulates the movement of ions through and out of the cells [2,3]. The presence of carboxyl groups and methyl ester has a determinant influence on the physical–chemical properties of PGA, mainly in terms of complexation and gel formation. To interpret Cu and/or Ni–PGA interaction, some authors assumed only the formation of the ML complex [4,5], and others the ML2 species [6]. In reference 7 the species ML is proposed for Ni and ML2 for Cu. Species formation depends on the metal ion, pectin origin and esterified fraction, as well as on the metal ion and pectin concentrations. In the interaction of Ca with pectins a great emphasis is given on the interchain association due to stronger cooperative binding of Ca [8]. Although the binding affinity of polygalacturonic acid (PGA) or derivatives with metal ions has been extensively investigated, namely with calcium [3,9–15], copper [4,7,16–18], and nickel [4–7], there is a total lack of information about the competition between different metal ions for the binding sites of PGA. Further investigation should
⁎ Corresponding author. Tel.: +351 21 8419177; fax: +351 21 8464457. E-mail addresses: [email protected] (C. Vilhena), [email protected] (M.L.S. Gonçalves), [email protected] (A.M. Mota). 1 Present address: Lusomedicamenta, Consiglieri Pedroso 69-B, Queluz de Baixo, 2730-055 Barcarena, Portugal. 1567-5394/$ – see front matter © 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.bioelechem.2009.05.010
be done in this context. On the other hand, although voltammetric methods have been widely used in metal ion speciation in aqueous solution, in the presence of pectins only very few applications can be found, and none about Ni(II)–pectin interaction, to the best of our knowledge. The serpentine-derived soils are known to be less favourable to plant growth and productivity due to the large content of Ni(II) present in these type of soils, although in a low dose Ni(II) is an essential tracenutrient for higher plants [19]. The uptake of Ni(II) by a very common tree in Mediterranean countries, Quercus ilex, on highly contaminated serpentine soils of NE Portugal (about 25 × 10− 6 mol/g of soil), has been studied by Nabais et al. [20,21]. These authors found Ni concentrations of (3 to 7) × 10− 6 M in the xylem of Q. ilex grown in serpentine soils, which is about nine times higher than the concentration in sandy loam soils. In that range of concentrations, Ni(II) may compete with Ca(II) for the same binding sites of PGA. Since Cu(II) is another heavy metal often present in this type of soils, it is also important to know how its presence affects Ni complexation with pectins. PGA is responsible for important interactions that occur between solutes and cell walls of xylem vessels, since pectins are one of the main constituents of that tissue. The interactions between cations and the negatively charged groups of the cell walls can delay the ion transport in the xylem sap [21]. On the other hand, PGA may be released from the walls of root vascular cells, being present at lower levels in the xylem sap [22]. In this way, cations may be partly transported from roots to shoots of higher plants complexed to PGA. The knowledge of metal ions interaction with PGA has important implications in the physiology of plants exposed to those metals.
32
C. Vilhena et al. / Bioelectrochemistry 77 (2009) 31–36
A Metrohm pH electrode (ref. 6.1033.500) with a Metrohm reference electrode (ref. 6.0733.100) and an Orion pH semi-micro electrode (91-03 SCE) were used in the pectin acid/base titration and in the pH control at 5.5, respectively.
Fig. 1. Polygalacturonic acid (PGA) structure.
Chelating agents as PGA are able to keep free trace metal ions within certain limits, thus reducing their toxicity. On the other hand, data on the interaction of metal ions with pectins allow a better understanding in the transfer of macro and micro metal ion nutrients through the soil–root interface and in the modifications induced by those ions on the fibrillar structure of PGA [23]. If the presence of Ca and/or other divalent ions decrease the distance between adjacent chain-segments, the probability to form more energetic cross-links increases, enhancing the strength of the network [24,25]. Also, due to its metal binding capacity, pectins have been considered a very interesting alternative to remove transition and heavy metals from wastewater and soils [26,27]. In the present work, the interaction of pectin with Ni(II), in the absence or presence of Ca(II) or Cu(II), was investigated at pH 5.5, similar to that of Q. ilex xylem sap. Voltammetric techniques were used to determine free nickel and free copper concentrations, for ligand concentrations in the absence of adsorption on the mercury electrode. Potentiometry with specific electrodes was also used in the determination of free Ca(II) or Cu(II). A large range of concentrations were tried to find the best models that fit the experimental results.
2.2.2. Voltammetric experiments Differential pulse polarography (DPP) and square wave voltammetry (SWV), used to determine free Ni(II) concentration, were performed with an Eco Chemie Autolab-PGST12 attached to a Metrohm 663 VA stand (with a hanging mercury electrode as working electrode (HMDE ref. 6.1246.020), Ag/AgCl reference electrode with double junction and a glassy carbon counter electrode) and to a personal computer using GPES 4.7 software (Eco Chemie). The experimental conditions in DPP were: initial potential of −0.8 V, final potential of −1.3 V, step potential of 2 mV, and amplitude of 50 mV. In SWV, the experimental conditions were: frequency of 10 Hz, initial potential of −0.8 V, final potential of − 1.3 V, step potential of 5 mV, and amplitude of 25 mV. 3. Results and discussion 3.1. Working concentration range of pectin
2. Experimental
Pectin contains a concentration of (1.90 ± 0.02) × 10− 3 mol of carboxylic groups per gram of pectin. From its dissociation constant (pKH equal to 3.76 [18]) it can be found that at pH 5.5 the pectin is about 98% de-protonated. Voltammetric measurements of Cu(II) were limited to a maximum ligand concentration of 0.056 g dm− 3 due to pectin adsorption on mercury in the potential range of the Cu(II) reduction peak (peak potential Ep ~− 0.1 V) [18]. For higher ligand concentrations, free copper was measured by potentiometry with a Cu(II) selective electrode. The highest concentration used in Ni(II)–pectin studies was 1 g of pectin dm− 3, since for higher concentrations, the differential pulse base line (obtained in the absence of nickel) began to present two broad peaks in the potential range of −1.0 V, close to the reduction peak potential of Ni(II).
2.1. Reagents
3.2. Voltammetric signal of Ni(II)
In all solutions, water was distilled and passed through a Millipore Milli-Q system. A potassium salt of esterified (25%) pectin extracted from citrus fruit was purchased from Sigma (P-9311) (www.SigmaAldrich.com). Stock solutions of Ca(II), Cu(II), and Ni(II) (as nitrates), with a concentration of 0.10 M, were prepared from analytical grade reagents. The concentration ranges used for pectin and for metal ions are presented in Table 1. Mercury used in the working voltammetric electrode was suprapure and purchased from Merck. Dissolved oxygen was removed from solutions in the electrochemical cell (potentiometric and voltammetric experiments) by bubbling nitrogen of 99.995% purity.
The redox couple Ni(II)/Ni0(Hg) had an irreversible behaviour in the absence or presence of pectin, with a peak width of 103 ± 5 mV. The peak potential was not affected by the presence of pectin, but the peak current decreased with the increase in ligand concentration, i.e., the complex showed an inert behaviour within the timescale of the technique. The lability criteria [28] also predict an inert behaviour for the Ni(II)–pectin system if the Eigen mechanism [29] (ML complex formation is kinetically limited by the dehydration of the metal ion M) is assumed. For inert complexes, the free metal concentration is directly determined from the peak current: M + L
½M = ip
M
= ip × ½Mt
2.2. Equipment and experimental conditions All of the experiments (potentiometric and voltammetric) were carried out in a thermostatic cell under a purified nitrogen atmosphere at 25.0 °C in 0.1 M KNO3. The experiments in the presence of metal ions were performed either by titration (with a waiting time of 5 min for each point) or using batch solutions (2 h or 1 day). 2.2.1. Potentiometric experiments For potentiometric measurements, a Sentek-Denver potentiometer was used with a Metrohm Ag/AgCl reference electrode. Free calcium and free copper concentrations were measured using, respectively, Ca(II) (Orion 93-20) and Cu(II) (Orion 94-29) electrodes.
M+L where iM are the peak currents in the absence and presence p and ip of pectin (M stands for metal ion and L for ligand), respectively, for the same total concentration of Ni(II) in solution, [M]t.
Table 1 Concentration ranges used in this work. Pectin (g dm− 3)
Ca(II) (M)
0.45–1.0 0.50–0.80 0.50–1.0
(5.0–10) × 10− 4
Cu(II) (M)
Ni(II) (M)
(1.0–5.0) × 10− 4
(0.02–20) × 10− 4 (0.1–0.8) × 10− 4 (2.0–10) × 10− 4
C. Vilhena et al. / Bioelectrochemistry 77 (2009) 31–36
33
3.3. Ni(II) interaction with pectin Pectin concentrations from 0.2 to 1 g dm− 3, i.e., (0.4 to 2)× 10− 3 mol COO− dm− 3, and Ni(II) concentrations from 2 × 10− 6 to 2 × 10− 3 M were used. The highest pectin concentration is limited by the adsorption on the Hg electrode. The lowest concentration, 0.2 g dm− 3, was chosen so that the free metal ion was always lower than 95% of the total metal ion concentration. Differential pulse polarography (DPP) and/or square wave voltammetry (SWV) were used to determine the free metal ion from the peak current. No differences where observed in free Ni(II) concentration from DPP or SWV measurements. In terms of Ni(II)–pectin kinetics, no systematic deviation was observed between experimental results from batch solutions or from ligand titrations with Ni(II), i.e., equilibrium was reached within 5 min (waiting time after Ni(II) addition and before the voltammetric analysis, during the titration of the pectin with the metal ion). Mass balance equations of the metal ion and ligand were used to determine the stability constants from the models including only one species, ML or ML2, or both species, ML + ML2. In the concentration range where θ ≥ 0.1, the models based on only one species, ML or ML2, are not suitable to explain the experimental results, either for Cu(II)– pectin or for Ni(II)–pectin systems. Assuming the ML + ML2 model, the best fitting between experimental and theoretical free metal concentrations for all ligand concentrations and θ ≥ 0.1 led to the following stoichiometric stability constants: log(βNiL/M− 1) = 2.6 ± 0.1 and log(βNiL2/M− 2) = 5.8 ± 0.1. In this case a good adjustment is obtained between experimental values and theoretical curves as presented in Fig. 2, where free Ni values (experimental and theoretical) were plotted against total Ni concentration. This model had been previously found for the Cu(II)–pectin system with θ ≥ 0.1: log(βCuL/M− 1) = 3.5 and log(βCuL2/M− 2) = 8.0 were determined in 0.10 M KNO3, at pH 5.5 and 25 °C [18]. However, the ML + ML2 model does not explain the results obtained for θ b 0.1, for none of these systems. In Fig. 3 experimentalfree Ni concentration ([Ni]exp) is plotted against theoretical values ([Ni]theor) for the ML + ML2 model with log(βNiL/M− 1) = 2.6 and log(βNiL2/M− 2) = 5.8, determined for the same total ligand and metal concentrations as [Ni]exp. If the model was appropriated, experimental points should fit the straight line with slope one. Since [Ni]exp is above the straight line with slope one, this indicates a lower complexation compared to the one expected from the ML + ML2 model. Theoretical values obtained from the ML model with log(βNiL/M− 1) = 2.6 were also included as comparison. A good fitting is observed between this model and the experimental points.
The decrease of Cu(II)–pectin interaction in the low θ range was previously attributed to a different conformation of the complex species [18]. For this system, the pectin adsorption on the Hg electrode in the range of the Cu(II) peak potential does not allow pectin concentrations higher than 0.056 g dm− 3. However, for Ni(II)–pectin interaction, a lower range of θ values was able to be explored, and so a better insight on the system was possible. For very low θ values (θ ≤ 0.01), the experimental points of the Ni(II)–pectin system fitted the ML model quite well, for log(βNiL/M− 1) = 2.56 ± 0.09 (Fig. 4). The validity of the ML model is supported by βNiL value, similar to the one obtained in the high θ range where ML + ML2 was assumed, and by the straight line obtained from experimental versus theoretical free nickel concentration (Fig. 4), with slope 1.00 ± 0.01 and intercept (2 ± 4) × 10− 7 (R2 = 0.996). Optimum values should have an unitary slope and a null intercept. Assuming the ML2 model, the straight line [Ni]exp versus [Ni]theor showed R2 = 0.991 with slope (0.97 ± 0.09) and intercept (11 ± 5) × 10− 7. In this case, log(βNiL2/M− 2) was equal to 5.4 ± 0.2, which is different from the value obtained in the high θ range. These results point to the formation of only one species, ML, in the very low θ range. Although in a classical approach the excess of ligand (low θ values) should favour the ML2 versus ML species, the electrostatic effect in pectin (repulsion between negatively charged
Fig. 2. Experimental-free Ni concentration (points) and theoretical values from the ML + ML2 model (dashed lines) versus total Ni concentration in the high θ range. [Ni]total, [Ni]: total and free Ni(II) concentration, respectively. Pectin concentration (g dm− 3): 0.5 (Δ); 1.0 (Ο). (—): straight line with slope 1 and null intercept.
Fig. 4. Fitting between experimental points and the theoretical ML model in the θ range b 0.01. (—): straight line with slope 1 and null intercept. [Ni]exp — experimental-free Ni(II) concentration for pectin concentration (g dm− 3): 0.224 (⋄), 0.448 (□), 0.493 (+), 0.500 (△), 0.806 (✱), 1.00 (○). [Ni]theor — free Ni concentration determined from the ML model.
Fig. 3. Fitting between experimental points and theoretical curves in the low θ range. (—): straight line with slope 1 and null intercept. [Ni]theor — theoretical free Ni(II) concentration from the ML + ML2 model. ⋄, △: points where [Ni] represents the experimentalfree Ni(II) concentration for pectin concentration (g dm− 3) of 0.806 (⋄) and 0.448 (△). (——): curves where [Ni] represents the theoretical free Ni(II) concentration for the ML model. (a, b, c, d): (2, 4, 6, 8) × 10− 6 M of total Ni(II) concentration.
34
C. Vilhena et al. / Bioelectrochemistry 77 (2009) 31–36
Fig. 5. Species distribution obtained from the ML + ML2 model. [Ni]total: total Ni concentration. Thin and thick lines: 0.5 and 1.0 g of pectin dm− 3, respectively. Solid line: percentage of free ligand (versus total ligand concentration). - - -, —–—, —– –— : percentage of M, ML and ML2 (versus total metal concentration), respectively.
addition of this metal ion may favour the formation of ML2 species through the partial neutralisation of pectin negative charges. For example, for 0.50 g pectin dm− 3 and 2 × 10− 4 M of total Ni(II) concentration, the θexp is equal to 0.045, below 0.1, and so [Ni]exp is 40% higher than [Ni]theor determined from the ML + ML2 model with the stability constants presented above (Fig. 6, first column, where the dashed line is the theoretical value from the ML + ML2 model). When Cu(II) is added, θexp (=([Ni]t − [Ni] + [Cu]t − [Cu])/total ligand concentration) increases to values higher than 0.18. Thus, the experimental-free Ni(II) is similar, within the experimental error, to the theoretical value determined from the ML + ML2 model (Fig. 6, first set of columns excluding the first one, in the absence of copper). The increase in the complexing power of the ligand in the presence of Cu(II), which allows the NiL2 formation, led to a decrease in free Ni(II) concentration, as was experimentally observed. For the other two sets of total Ni concentrations in Fig. 6, θ is always N0.1 in the absence or presence of Cu(II), and so [Ni]exp is similar to [Ni]theor for the model ML + ML2, within the experimental error. 3.5. Influence of Ca(II) in Ni(II)–pectin interaction
pectin chains) hinders the formation of ML2 species in the low θ range, where each L corresponds to a separate chain. Increasing the metal to ligand concentration ratio so that θ ≥ 0.1, the carboxylic negative charges of the pectin chain are partial neutralised by the positive charge of the metal ion. It allows another pectin chain to get closer to the metal ion, and a “sandwich” with Ni(II) ions in the middle of the two chains may then be formed. Siew and Williams [30] also described inter-chain association, where divalent ions like Ni(II) interact with carboxylate groups of more than one polymer chain, inducing cross-linking. Wittmer et al. [31] argued that divalent cations can interact with one carboxylate group along the polymer chain to form a monocomplex or with two carboxylate groups on different chains, which is in accordance with the results of this work. In Fig. 5, species distribution obtained from the ML + ML2 model with log(βNiL/M− 1) = 2.6 and log(βNiL2/M− 2) = 5.8 showed that for θ ≥ 0.1, and in the concentration range used, ML is never in a large excess compared to ML2. The ML2 species is very stable, presenting a value of log(KNiL2/M− 1) = 2.88 (KML2 = [ML2]/([ML] [L])), higher than log βNiL as already verified for the Cu(II)–pectin system [18]. Both systems, Cu(II)–pectin and Ni(II)–pectin, present similar behaviours, and so they should be interpreted in the same way, i.e., assuming ML formation in the low θ range and ML + ML2 in the high θ range. Besides metal/ligand concentrations, responsible for θ values, other parameters such as ionic strength, pectin origin and esterified fraction, should also influence the presence of ML and/or ML2 species. However this discussion is beyond the scope of this paper, where only one pectin and one ionic strength were used.
According to the results from potentiometric experiments with a calcium selective electrode, a value of log(βCaL/M− 1) = 2.4 ± 0.1 was obtained for calcium concentrations lower than 10− 4 M and pectin amounts in the range of 0.2–1.6 g dm− 3, so that θ b 0.03. To achieve θ values higher than 0.1, calcium concentration in the range of (0.5–1) × 10− 3 M was used (higher concentrations are not possible due to a gel formation). By imposing log(βCaL/M− 1) = 2.4, the value of log(βCaL2/M− 2) = 5.6 ± 0.1 was found. The interaction of Ca(II) in free Ni(II) concentration is presented in Fig. 7. The values were obtained from at least three different titrations (addition of copper for a constant nickel and pectin concentration or vice-versa) performed in different days for the same experimental conditions. Once more, the interaction of Ca(II) with the Ni(II)–pectin system did not increase free Ni(II) concentration, as would be expected from a classical approach. Experimental θ values, always below 0.1 in the absence of Ca(II) (θexp ≤ 0.02), became higher than 0.1 when Ca(II) was added (θexp =([Ni]t −[Ni]+[Ca]t −[Ca])/total ligand concentration). In Fig. 7, when θexp bb 0.1 (absence of calcium, first column in each set of columns), [Ni]exp was about 45% higher than [Ni]theor determined from the ML + ML2 model (dashed lines). In this θ range the formation of the ML2 species was hindered by the repulsion between
3.4. Influence of Cu(II) in Ni(II)–pectin interaction Fig. 6 presents the effect of Cu(II) interaction in free Ni(II) concentration for different sets of Ni(II)–pectin concentrations. The values were obtained from at least three different titrations (addition of copper for a constant nickel and pectin concentration or vice-versa) performed in different days for the same experimental conditions. The interaction of Cu(II) with the Ni(II)–pectin system may lead to a decrease in free Ni(II), as can be seen from the two first columns in Fig. 6. However, direct competition of Cu(II) for the binding sites of the pectin should increase the free Ni(II) concentration and not decrease it. The observed behaviour can be explained based on the species formation discussed above. In fact, if θ experimental values are lower than 0.1 for the Ni(II)–pectin system in the absence of Cu(II), the
Fig. 6. Interaction of Cu(II) in free Ni(II) concentration for different sets of Ni(II)–pectin concentrations. [Ni]total, [Ni]exp: total and experimental-free Ni(II) concentration, respectively. Pectin concentration (g dm− 3): 0.5 (A); 0.5 (B); 1.0 (C). Total copper concentration (M), from left to right: (A): (0,1, 2, 4) × 10− 4; (B): (0,1, 2, 5) × 10− 4; (C): (0,1, 2, 5) × 10− 4. Dashed lines: theoretical values assuming the ML + ML2 model for both cations.
C. Vilhena et al. / Bioelectrochemistry 77 (2009) 31–36
35
groups of more than one polymer chain. This association is only formed when the partial neutralisation of the negative carboxylic charges by the increase of the number of divalent ions or decrease of ligand concentration is sufficient in order to favour the association of separate chains through the metal ion. The affinity of the metal ions for the pectin follows the order: Cu(II) N Ni(II) ≥ Ca(II). Acknowledgements This work is under the Research Project POCTI/1999/MGS/35653 and FCT — SFRH/BD/6293/2001. References
Fig. 7. Interaction of Ca(II) in free Ni(II) concentration in the presence of 0.50 g of pectin dm− 3. Total calcium concentration (M) in the columns of each set, from left to right: 0; 5 × 10− 4; 7.5 × 10− 4; 1 × 10− 3.
the different pectin chains, and so nickel complexation is much lower than the one predicted from the ML + ML2 model. However, in the presence of Ca(II) (θexp ≥ 0.1), the shift between experimental and theoretical values, assuming the model ML + ML2 for both metal ions, is below 13% (dashed lines included in the error bar). The model NiL + NiL2 + CaL was also tried, but shifts between experimental and theoretical values of free Ni(II) increased up to 26%. Siew and Williams [30] also noticed that Ca(II) ions interact with anionic polymers containing carboxylate groups through chemical binding. The same authors concluded that Ca(II) may interact with carboxylate groups of separate pectin chains, resulting in crosslinking, or with carboxylate groups in a single chain, forming complexes at specific sites on the chains. With increasing Ca(II) concentrations, a gel formation occurs due to the association of uncharged or oppositely charged segments on separate polymer chains. 4. Conclusion Pectin in the presence of divalent cations such as Cu(II), Ni(II), and Ca(II), and in the very low θ range (very high excess of negative charges), forms only the ML complex, where one cation interacts with one carboxylate in a specific site of the pectin chain. In this θ range the species ML2, where M interacts with carboxylate groups of two different polymer chains, is hindered due to the repulsion between the negative charges of carboxylic groups in the two independent segments of pectin. For θ ≥ 0.1 the species ML2 is formed in solution, contrary to what would be expected from an increase in θ value, i.e., from a decrease of ligand and/or increase of metal ion concentrations. This should be due to the partial neutralisation of the carboxylic groups by the positive charges of divalent cations, which induces the association of separate chains through the metal ion, leading to the ML2 formation. The presence of Cu(II) or Ca(II) may increase Ni(II) complexation to the pectin, contrary to what would be expected from a direct competition between divalent cations for the pectin binding sites. This can be explained by the partial neutralisation of pectin's negative charges by Cu(II) or Ca(II), which favours NiL2 formation through the association of independent chains. The experimental evidence for (i) the ML2 formation when the pectin concentration decreases and (ii) the decrease of the experimental-free Ni(II) concentration in the presence of Ca(II) or Cu(II), is opposite to the usual trends presented by the majority of ligands. In this paper they have been interpreted assuming a pectin inter-chain association, where divalent ions like Ni(II) interact with carboxylate
[1] A. Darvill, M. McNeil, P. Albersheim, D.P. Delmer, The primary cell walls of flowering plants, in: N.E. Tolbert (Ed.), the Biochemistry of Plants: A Comprehensive Treatise, vol.1, Academic Press, New York, 1980, pp. 91–161. [2] J. Moorby, Transport Systems in Plants, Longman Inc, New York, 1981, pp. 28–32. [3] B. Gessa, S. Deiana, Ca-polygalacturonate as a model for a soil–root interface. Fibrillar structure and comparison with natural root mucilage, Plant Soil 140 (1992) 1–13. [4] V.M. Dronnet, C.M.G.C. Renard, M.A.V. Axelos, J.F. Thibault, Characterisation and selectivity of divalent metal ions binding by citrus and sugar beet pectins, Carbohydr. Polym. 30 (1996) 253–263. [5] K.A. Davarski, S.N. Manolov, I.N. Petrova, V.D.J. Mavrov, Complex equilibria in the system M2+-polygalacturonic acid (pectin) — H2O (M = Co, Ni, Cu, Zn, Cd and Pb), J. Coord. Chem. 33 (1994) 75–81. [6] S. Deiana, G. Micera, G. Muggiolu, C. Gessa, A. Pusino, Interaction of transitionmetal ions with polygalacturonic acid — a potentiometric study, Colloids Surf. 6 (1983) 17–25. [7] M.T. Kartel, L.A. Kupchik, B.K. Veisov, Evaluation of pectin binding of heavy metal ions in aqueous solutions, Chemosphere 38 (1999) 2591–2596. [8] A. Malovíková, M. Rinaudo, M. Milas, Comparative interactions of magnesium and calcium counterions with polygalacturonic acid, Biopolymers 34 (1994) 1059–1064. [9] G. Ravanat, M. Rinaudo, Investigation on oligogalacturonic and polygalacturonic acids by potentiometry and circular dichroism, Biopolymers 19 (1980) 2209–2222. [10] J.F. Thibault, M. Rinaudo, Chain association of pectin molecules during calciuminduction gelation, Biopolymers 25 (1986) 455–468. [11] R. Kohn, Binding of divalent-cations to oligomeric fragments of pectin, Carbohydr. Res. 160 (1987) 343–353. [12] C. Garnier, M.A.V. Axelos, J.F. Thibault, Selectivity and cooperativity in the binding of calcium ions by pectin, Carbohydr. Res. 256 (1994) 71–81. [13] B. Manunza, S. Deiana, M. Pintore, C. Gessa, Interaction of Ca2+ and Na+ ions with polygalacturonate chains: a molecular dynamics study, Glycoconj. J. 15 (1998) 297–300. [14] S.K. Wiedmer, A. Cassely, M. Hong, M.V. Novotny, M.L. Riekkola, Electrophoretic studies of polygalacturonate oligomers and their interactions with metal ions, Electrophoresis 21 (2000) 3212–3219. [15] M.C. Ralet, M.J. Crépeau, H.C. Buchholt, J.F. Thibault, Polyelectrolyte behaviour and calcium binding properties of sugar beet pectins differing in their degrees of methylation and acetylation, Biochem. Eng. J. 16 (2003) 191–2001. [16] S. Deiana, L. Erre, G. Micera, P. Piu, C. Gessa, Coordination of transition-metal ions by polygalacturonic acid — a spectroscopy study, Inorg. Chim. Acta. - Bioinorg. Chem. 46 (1980) 249–253. [17] G. Manzini, A. Cesaro, F. Delben, S. Paoletti, E. Reisenhofer, Copper(II) binding by natural ionic polysaccharides. 1. Potentiometric and spectroscopic, Bioelectrochem. Bioeng. 12 (1984) 443–454. [18] C. Vilhena, M.L.S. Gonçalves, A.M. Mota, Binding of copper(II) to pectins by electrochemical methods, Electroanalysis 16 (2004) 2065–2072. [19] J. Gerendás, J.C. Polacco, S.K. Freyermuth, B. Sattelmacher, Significance of nickel for plant growth and metabolism, J. Plant Nutr. Soil Sci. 162 (1999) 241–256. [20] C. Nabais, H. Freitas, J. Hagemeyer, S-W. Breckle, Radial distribution of Ni in stemwood of Quercus ilex L trees grown on serpentine and sandy loam (umbric leptosol) soils of NE-Portugal, Plant Soil 183 (1996) 181–185. [21] G.C.L. Araujo, S.G. Lemos, C. Nabais, Nickel sorption capacity of ground xylem of Quercus ilex trees and effects of selected ligands present in the xylem sap, J. Plant Physiol. 166 (2009) 270–277. [22] S. Satoh, Organic substances in xylem sap delivered to above-ground organs by the roots, J. Plant Res. 119 (2006) 179–187. [23] S. Deiana, A. Premoli, C. Senette, Oxidation of caffeic acid by Fe(III) trapped in a Capolygalacturonate network, Plant Physiol. Biochem. 46 (2008) 435–443. [24] A. Kjøniksen, M. Hiorth, B. Nyström, Temperature-induced association and gelation of aqueous solutions of pectin. A dynamic light scattering study, Eur. Polym. J. 40 (2004) 2427–2435. [25] J.B. Wehr, N.W. Menzies, F.P.C. Blamey, Inhibition of cell-wall autolysis and pectin degradation by cations, Plant Physiol. Biochem. 42 (2004) 485–492. [26] J.S. Melo, S.F. D'Souza, Removal of chromium by mucilaginous seeds of Ocimum basilicum, Bioresour. Technol. 92 (2004) 151–155. [27] A.L. Debbaudt, M.L. Ferreira, M.E. Gschaider, Theoretical and experimental study of M2+ adsorption on biopolymers. III. Comparative kinetic pattern of Pb, Hg and Cd, Carbohydr. Polym. 56 (2004) 321–332.
36
C. Vilhena et al. / Bioelectrochemistry 77 (2009) 31–36
[28] H.G. De Jong, H.P. van Leeuwen, Voltammetry of metal-complex systems with different diffusion-coefficients of the species involved. 2. Behaviour of the limiting current and its dependence on association dissociation kinetics and lability, J. Electroanal. Chem. 234 (1987) 17–29. [29] M. Eigen, K. Tamm, Fast elementary steps in chemical reaction mechanism, Pure Appl. Chem. 6 (1963) 97–115.
[30] C.K. Siew, P.A. Williams, New insights into the mechanism of gelation of alginate and pectin: charge annihilation and reversal mechanism, Biomacromolecules 6 (2005) 963–969. [31] J. Wittmer, A. Johner, J.F. Joanny, Precipitation of polyelectrolytes in the presence of multivalent salts, J. Phys. II 5 (1995) 635–654.
Contents lists available at ScienceDirect
Bioelectrochemistry j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / b i o e l e c h e m
Competition of copper and/or calcium in nickel–pectin interaction by potentiometric and voltammetric techniques C. Vilhena 1, M.L.S. Gonçalves, A.M. Mota ⁎ Centro de Química Estrutural (Torre Sul), Instituto Superior Técnico, Av. Rovisco Pais, 1049-001 Lisboa, Portugal
a r t i c l e
i n f o
Article history: Received 11 February 2009 Received in revised form 7 May 2009 Accepted 23 May 2009 Available online 30 May 2009 Keywords: Ni(II) Cu(II) Ca(II) Pectin Competition Voltammetry
a b s t r a c t The interaction between nickel and pectin extracted from citrus fruit was studied in 0.10 M KNO3, at pH 5.5 and 25 °C. Differential pulse and/or square wave polarography were used to determine free nickel. For a high coverage degree (θ) of the pectin by the metal ion a good fitting was observed between experimental results and the model that includes both complex species, ML and ML2 (M for the metal ion and L for the ligand). In the ML2 species, Ni(II) interacts with two carboxylate groups of different chains, resulting in an inter-chain association. For low θ values, the formation of ML2 is hindered due to the repulsion between the negative charges of carboxylic groups in two independent segments of pectin. The influence of calcium or copper ions on the free nickel concentration, in the presence of pectin, may lead to a decrease in free nickel concentration, contrary to what would be expected from direct competition between Ca(II) or Cu(II) and Ni(II) for the pectin binding sites. This is due to the partial neutralisation of the negative carboxylic charges by the positive charges of the divalent cations, which favours NiL2 formation through the association of independent chains. © 2009 Elsevier B.V. All rights reserved.
1. Introduction Partially esterified polygalacturonic acid (Fig. 1) is the main component of pectins in higher plants [1], and is an important constituent of the apoplastic transport system. Polygalacturonic acid (PGA) acts as a selective filter of nutritive elements and regulates the movement of ions through and out of the cells [2,3]. The presence of carboxyl groups and methyl ester has a determinant influence on the physical–chemical properties of PGA, mainly in terms of complexation and gel formation. To interpret Cu and/or Ni–PGA interaction, some authors assumed only the formation of the ML complex [4,5], and others the ML2 species [6]. In reference 7 the species ML is proposed for Ni and ML2 for Cu. Species formation depends on the metal ion, pectin origin and esterified fraction, as well as on the metal ion and pectin concentrations. In the interaction of Ca with pectins a great emphasis is given on the interchain association due to stronger cooperative binding of Ca [8]. Although the binding affinity of polygalacturonic acid (PGA) or derivatives with metal ions has been extensively investigated, namely with calcium [3,9–15], copper [4,7,16–18], and nickel [4–7], there is a total lack of information about the competition between different metal ions for the binding sites of PGA. Further investigation should
⁎ Corresponding author. Tel.: +351 21 8419177; fax: +351 21 8464457. E-mail addresses: [email protected] (C. Vilhena), [email protected] (M.L.S. Gonçalves), [email protected] (A.M. Mota). 1 Present address: Lusomedicamenta, Consiglieri Pedroso 69-B, Queluz de Baixo, 2730-055 Barcarena, Portugal. 1567-5394/$ – see front matter © 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.bioelechem.2009.05.010
be done in this context. On the other hand, although voltammetric methods have been widely used in metal ion speciation in aqueous solution, in the presence of pectins only very few applications can be found, and none about Ni(II)–pectin interaction, to the best of our knowledge. The serpentine-derived soils are known to be less favourable to plant growth and productivity due to the large content of Ni(II) present in these type of soils, although in a low dose Ni(II) is an essential tracenutrient for higher plants [19]. The uptake of Ni(II) by a very common tree in Mediterranean countries, Quercus ilex, on highly contaminated serpentine soils of NE Portugal (about 25 × 10− 6 mol/g of soil), has been studied by Nabais et al. [20,21]. These authors found Ni concentrations of (3 to 7) × 10− 6 M in the xylem of Q. ilex grown in serpentine soils, which is about nine times higher than the concentration in sandy loam soils. In that range of concentrations, Ni(II) may compete with Ca(II) for the same binding sites of PGA. Since Cu(II) is another heavy metal often present in this type of soils, it is also important to know how its presence affects Ni complexation with pectins. PGA is responsible for important interactions that occur between solutes and cell walls of xylem vessels, since pectins are one of the main constituents of that tissue. The interactions between cations and the negatively charged groups of the cell walls can delay the ion transport in the xylem sap [21]. On the other hand, PGA may be released from the walls of root vascular cells, being present at lower levels in the xylem sap [22]. In this way, cations may be partly transported from roots to shoots of higher plants complexed to PGA. The knowledge of metal ions interaction with PGA has important implications in the physiology of plants exposed to those metals.
32
C. Vilhena et al. / Bioelectrochemistry 77 (2009) 31–36
A Metrohm pH electrode (ref. 6.1033.500) with a Metrohm reference electrode (ref. 6.0733.100) and an Orion pH semi-micro electrode (91-03 SCE) were used in the pectin acid/base titration and in the pH control at 5.5, respectively.
Fig. 1. Polygalacturonic acid (PGA) structure.
Chelating agents as PGA are able to keep free trace metal ions within certain limits, thus reducing their toxicity. On the other hand, data on the interaction of metal ions with pectins allow a better understanding in the transfer of macro and micro metal ion nutrients through the soil–root interface and in the modifications induced by those ions on the fibrillar structure of PGA [23]. If the presence of Ca and/or other divalent ions decrease the distance between adjacent chain-segments, the probability to form more energetic cross-links increases, enhancing the strength of the network [24,25]. Also, due to its metal binding capacity, pectins have been considered a very interesting alternative to remove transition and heavy metals from wastewater and soils [26,27]. In the present work, the interaction of pectin with Ni(II), in the absence or presence of Ca(II) or Cu(II), was investigated at pH 5.5, similar to that of Q. ilex xylem sap. Voltammetric techniques were used to determine free nickel and free copper concentrations, for ligand concentrations in the absence of adsorption on the mercury electrode. Potentiometry with specific electrodes was also used in the determination of free Ca(II) or Cu(II). A large range of concentrations were tried to find the best models that fit the experimental results.
2.2.2. Voltammetric experiments Differential pulse polarography (DPP) and square wave voltammetry (SWV), used to determine free Ni(II) concentration, were performed with an Eco Chemie Autolab-PGST12 attached to a Metrohm 663 VA stand (with a hanging mercury electrode as working electrode (HMDE ref. 6.1246.020), Ag/AgCl reference electrode with double junction and a glassy carbon counter electrode) and to a personal computer using GPES 4.7 software (Eco Chemie). The experimental conditions in DPP were: initial potential of −0.8 V, final potential of −1.3 V, step potential of 2 mV, and amplitude of 50 mV. In SWV, the experimental conditions were: frequency of 10 Hz, initial potential of −0.8 V, final potential of − 1.3 V, step potential of 5 mV, and amplitude of 25 mV. 3. Results and discussion 3.1. Working concentration range of pectin
2. Experimental
Pectin contains a concentration of (1.90 ± 0.02) × 10− 3 mol of carboxylic groups per gram of pectin. From its dissociation constant (pKH equal to 3.76 [18]) it can be found that at pH 5.5 the pectin is about 98% de-protonated. Voltammetric measurements of Cu(II) were limited to a maximum ligand concentration of 0.056 g dm− 3 due to pectin adsorption on mercury in the potential range of the Cu(II) reduction peak (peak potential Ep ~− 0.1 V) [18]. For higher ligand concentrations, free copper was measured by potentiometry with a Cu(II) selective electrode. The highest concentration used in Ni(II)–pectin studies was 1 g of pectin dm− 3, since for higher concentrations, the differential pulse base line (obtained in the absence of nickel) began to present two broad peaks in the potential range of −1.0 V, close to the reduction peak potential of Ni(II).
2.1. Reagents
3.2. Voltammetric signal of Ni(II)
In all solutions, water was distilled and passed through a Millipore Milli-Q system. A potassium salt of esterified (25%) pectin extracted from citrus fruit was purchased from Sigma (P-9311) (www.SigmaAldrich.com). Stock solutions of Ca(II), Cu(II), and Ni(II) (as nitrates), with a concentration of 0.10 M, were prepared from analytical grade reagents. The concentration ranges used for pectin and for metal ions are presented in Table 1. Mercury used in the working voltammetric electrode was suprapure and purchased from Merck. Dissolved oxygen was removed from solutions in the electrochemical cell (potentiometric and voltammetric experiments) by bubbling nitrogen of 99.995% purity.
The redox couple Ni(II)/Ni0(Hg) had an irreversible behaviour in the absence or presence of pectin, with a peak width of 103 ± 5 mV. The peak potential was not affected by the presence of pectin, but the peak current decreased with the increase in ligand concentration, i.e., the complex showed an inert behaviour within the timescale of the technique. The lability criteria [28] also predict an inert behaviour for the Ni(II)–pectin system if the Eigen mechanism [29] (ML complex formation is kinetically limited by the dehydration of the metal ion M) is assumed. For inert complexes, the free metal concentration is directly determined from the peak current: M + L
½M = ip
M
= ip × ½Mt
2.2. Equipment and experimental conditions All of the experiments (potentiometric and voltammetric) were carried out in a thermostatic cell under a purified nitrogen atmosphere at 25.0 °C in 0.1 M KNO3. The experiments in the presence of metal ions were performed either by titration (with a waiting time of 5 min for each point) or using batch solutions (2 h or 1 day). 2.2.1. Potentiometric experiments For potentiometric measurements, a Sentek-Denver potentiometer was used with a Metrohm Ag/AgCl reference electrode. Free calcium and free copper concentrations were measured using, respectively, Ca(II) (Orion 93-20) and Cu(II) (Orion 94-29) electrodes.
M+L where iM are the peak currents in the absence and presence p and ip of pectin (M stands for metal ion and L for ligand), respectively, for the same total concentration of Ni(II) in solution, [M]t.
Table 1 Concentration ranges used in this work. Pectin (g dm− 3)
Ca(II) (M)
0.45–1.0 0.50–0.80 0.50–1.0
(5.0–10) × 10− 4
Cu(II) (M)
Ni(II) (M)
(1.0–5.0) × 10− 4
(0.02–20) × 10− 4 (0.1–0.8) × 10− 4 (2.0–10) × 10− 4
C. Vilhena et al. / Bioelectrochemistry 77 (2009) 31–36
33
3.3. Ni(II) interaction with pectin Pectin concentrations from 0.2 to 1 g dm− 3, i.e., (0.4 to 2)× 10− 3 mol COO− dm− 3, and Ni(II) concentrations from 2 × 10− 6 to 2 × 10− 3 M were used. The highest pectin concentration is limited by the adsorption on the Hg electrode. The lowest concentration, 0.2 g dm− 3, was chosen so that the free metal ion was always lower than 95% of the total metal ion concentration. Differential pulse polarography (DPP) and/or square wave voltammetry (SWV) were used to determine the free metal ion from the peak current. No differences where observed in free Ni(II) concentration from DPP or SWV measurements. In terms of Ni(II)–pectin kinetics, no systematic deviation was observed between experimental results from batch solutions or from ligand titrations with Ni(II), i.e., equilibrium was reached within 5 min (waiting time after Ni(II) addition and before the voltammetric analysis, during the titration of the pectin with the metal ion). Mass balance equations of the metal ion and ligand were used to determine the stability constants from the models including only one species, ML or ML2, or both species, ML + ML2. In the concentration range where θ ≥ 0.1, the models based on only one species, ML or ML2, are not suitable to explain the experimental results, either for Cu(II)– pectin or for Ni(II)–pectin systems. Assuming the ML + ML2 model, the best fitting between experimental and theoretical free metal concentrations for all ligand concentrations and θ ≥ 0.1 led to the following stoichiometric stability constants: log(βNiL/M− 1) = 2.6 ± 0.1 and log(βNiL2/M− 2) = 5.8 ± 0.1. In this case a good adjustment is obtained between experimental values and theoretical curves as presented in Fig. 2, where free Ni values (experimental and theoretical) were plotted against total Ni concentration. This model had been previously found for the Cu(II)–pectin system with θ ≥ 0.1: log(βCuL/M− 1) = 3.5 and log(βCuL2/M− 2) = 8.0 were determined in 0.10 M KNO3, at pH 5.5 and 25 °C [18]. However, the ML + ML2 model does not explain the results obtained for θ b 0.1, for none of these systems. In Fig. 3 experimentalfree Ni concentration ([Ni]exp) is plotted against theoretical values ([Ni]theor) for the ML + ML2 model with log(βNiL/M− 1) = 2.6 and log(βNiL2/M− 2) = 5.8, determined for the same total ligand and metal concentrations as [Ni]exp. If the model was appropriated, experimental points should fit the straight line with slope one. Since [Ni]exp is above the straight line with slope one, this indicates a lower complexation compared to the one expected from the ML + ML2 model. Theoretical values obtained from the ML model with log(βNiL/M− 1) = 2.6 were also included as comparison. A good fitting is observed between this model and the experimental points.
The decrease of Cu(II)–pectin interaction in the low θ range was previously attributed to a different conformation of the complex species [18]. For this system, the pectin adsorption on the Hg electrode in the range of the Cu(II) peak potential does not allow pectin concentrations higher than 0.056 g dm− 3. However, for Ni(II)–pectin interaction, a lower range of θ values was able to be explored, and so a better insight on the system was possible. For very low θ values (θ ≤ 0.01), the experimental points of the Ni(II)–pectin system fitted the ML model quite well, for log(βNiL/M− 1) = 2.56 ± 0.09 (Fig. 4). The validity of the ML model is supported by βNiL value, similar to the one obtained in the high θ range where ML + ML2 was assumed, and by the straight line obtained from experimental versus theoretical free nickel concentration (Fig. 4), with slope 1.00 ± 0.01 and intercept (2 ± 4) × 10− 7 (R2 = 0.996). Optimum values should have an unitary slope and a null intercept. Assuming the ML2 model, the straight line [Ni]exp versus [Ni]theor showed R2 = 0.991 with slope (0.97 ± 0.09) and intercept (11 ± 5) × 10− 7. In this case, log(βNiL2/M− 2) was equal to 5.4 ± 0.2, which is different from the value obtained in the high θ range. These results point to the formation of only one species, ML, in the very low θ range. Although in a classical approach the excess of ligand (low θ values) should favour the ML2 versus ML species, the electrostatic effect in pectin (repulsion between negatively charged
Fig. 2. Experimental-free Ni concentration (points) and theoretical values from the ML + ML2 model (dashed lines) versus total Ni concentration in the high θ range. [Ni]total, [Ni]: total and free Ni(II) concentration, respectively. Pectin concentration (g dm− 3): 0.5 (Δ); 1.0 (Ο). (—): straight line with slope 1 and null intercept.
Fig. 4. Fitting between experimental points and the theoretical ML model in the θ range b 0.01. (—): straight line with slope 1 and null intercept. [Ni]exp — experimental-free Ni(II) concentration for pectin concentration (g dm− 3): 0.224 (⋄), 0.448 (□), 0.493 (+), 0.500 (△), 0.806 (✱), 1.00 (○). [Ni]theor — free Ni concentration determined from the ML model.
Fig. 3. Fitting between experimental points and theoretical curves in the low θ range. (—): straight line with slope 1 and null intercept. [Ni]theor — theoretical free Ni(II) concentration from the ML + ML2 model. ⋄, △: points where [Ni] represents the experimentalfree Ni(II) concentration for pectin concentration (g dm− 3) of 0.806 (⋄) and 0.448 (△). (——): curves where [Ni] represents the theoretical free Ni(II) concentration for the ML model. (a, b, c, d): (2, 4, 6, 8) × 10− 6 M of total Ni(II) concentration.
34
C. Vilhena et al. / Bioelectrochemistry 77 (2009) 31–36
Fig. 5. Species distribution obtained from the ML + ML2 model. [Ni]total: total Ni concentration. Thin and thick lines: 0.5 and 1.0 g of pectin dm− 3, respectively. Solid line: percentage of free ligand (versus total ligand concentration). - - -, —–—, —– –— : percentage of M, ML and ML2 (versus total metal concentration), respectively.
addition of this metal ion may favour the formation of ML2 species through the partial neutralisation of pectin negative charges. For example, for 0.50 g pectin dm− 3 and 2 × 10− 4 M of total Ni(II) concentration, the θexp is equal to 0.045, below 0.1, and so [Ni]exp is 40% higher than [Ni]theor determined from the ML + ML2 model with the stability constants presented above (Fig. 6, first column, where the dashed line is the theoretical value from the ML + ML2 model). When Cu(II) is added, θexp (=([Ni]t − [Ni] + [Cu]t − [Cu])/total ligand concentration) increases to values higher than 0.18. Thus, the experimental-free Ni(II) is similar, within the experimental error, to the theoretical value determined from the ML + ML2 model (Fig. 6, first set of columns excluding the first one, in the absence of copper). The increase in the complexing power of the ligand in the presence of Cu(II), which allows the NiL2 formation, led to a decrease in free Ni(II) concentration, as was experimentally observed. For the other two sets of total Ni concentrations in Fig. 6, θ is always N0.1 in the absence or presence of Cu(II), and so [Ni]exp is similar to [Ni]theor for the model ML + ML2, within the experimental error. 3.5. Influence of Ca(II) in Ni(II)–pectin interaction
pectin chains) hinders the formation of ML2 species in the low θ range, where each L corresponds to a separate chain. Increasing the metal to ligand concentration ratio so that θ ≥ 0.1, the carboxylic negative charges of the pectin chain are partial neutralised by the positive charge of the metal ion. It allows another pectin chain to get closer to the metal ion, and a “sandwich” with Ni(II) ions in the middle of the two chains may then be formed. Siew and Williams [30] also described inter-chain association, where divalent ions like Ni(II) interact with carboxylate groups of more than one polymer chain, inducing cross-linking. Wittmer et al. [31] argued that divalent cations can interact with one carboxylate group along the polymer chain to form a monocomplex or with two carboxylate groups on different chains, which is in accordance with the results of this work. In Fig. 5, species distribution obtained from the ML + ML2 model with log(βNiL/M− 1) = 2.6 and log(βNiL2/M− 2) = 5.8 showed that for θ ≥ 0.1, and in the concentration range used, ML is never in a large excess compared to ML2. The ML2 species is very stable, presenting a value of log(KNiL2/M− 1) = 2.88 (KML2 = [ML2]/([ML] [L])), higher than log βNiL as already verified for the Cu(II)–pectin system [18]. Both systems, Cu(II)–pectin and Ni(II)–pectin, present similar behaviours, and so they should be interpreted in the same way, i.e., assuming ML formation in the low θ range and ML + ML2 in the high θ range. Besides metal/ligand concentrations, responsible for θ values, other parameters such as ionic strength, pectin origin and esterified fraction, should also influence the presence of ML and/or ML2 species. However this discussion is beyond the scope of this paper, where only one pectin and one ionic strength were used.
According to the results from potentiometric experiments with a calcium selective electrode, a value of log(βCaL/M− 1) = 2.4 ± 0.1 was obtained for calcium concentrations lower than 10− 4 M and pectin amounts in the range of 0.2–1.6 g dm− 3, so that θ b 0.03. To achieve θ values higher than 0.1, calcium concentration in the range of (0.5–1) × 10− 3 M was used (higher concentrations are not possible due to a gel formation). By imposing log(βCaL/M− 1) = 2.4, the value of log(βCaL2/M− 2) = 5.6 ± 0.1 was found. The interaction of Ca(II) in free Ni(II) concentration is presented in Fig. 7. The values were obtained from at least three different titrations (addition of copper for a constant nickel and pectin concentration or vice-versa) performed in different days for the same experimental conditions. Once more, the interaction of Ca(II) with the Ni(II)–pectin system did not increase free Ni(II) concentration, as would be expected from a classical approach. Experimental θ values, always below 0.1 in the absence of Ca(II) (θexp ≤ 0.02), became higher than 0.1 when Ca(II) was added (θexp =([Ni]t −[Ni]+[Ca]t −[Ca])/total ligand concentration). In Fig. 7, when θexp bb 0.1 (absence of calcium, first column in each set of columns), [Ni]exp was about 45% higher than [Ni]theor determined from the ML + ML2 model (dashed lines). In this θ range the formation of the ML2 species was hindered by the repulsion between
3.4. Influence of Cu(II) in Ni(II)–pectin interaction Fig. 6 presents the effect of Cu(II) interaction in free Ni(II) concentration for different sets of Ni(II)–pectin concentrations. The values were obtained from at least three different titrations (addition of copper for a constant nickel and pectin concentration or vice-versa) performed in different days for the same experimental conditions. The interaction of Cu(II) with the Ni(II)–pectin system may lead to a decrease in free Ni(II), as can be seen from the two first columns in Fig. 6. However, direct competition of Cu(II) for the binding sites of the pectin should increase the free Ni(II) concentration and not decrease it. The observed behaviour can be explained based on the species formation discussed above. In fact, if θ experimental values are lower than 0.1 for the Ni(II)–pectin system in the absence of Cu(II), the
Fig. 6. Interaction of Cu(II) in free Ni(II) concentration for different sets of Ni(II)–pectin concentrations. [Ni]total, [Ni]exp: total and experimental-free Ni(II) concentration, respectively. Pectin concentration (g dm− 3): 0.5 (A); 0.5 (B); 1.0 (C). Total copper concentration (M), from left to right: (A): (0,1, 2, 4) × 10− 4; (B): (0,1, 2, 5) × 10− 4; (C): (0,1, 2, 5) × 10− 4. Dashed lines: theoretical values assuming the ML + ML2 model for both cations.
C. Vilhena et al. / Bioelectrochemistry 77 (2009) 31–36
35
groups of more than one polymer chain. This association is only formed when the partial neutralisation of the negative carboxylic charges by the increase of the number of divalent ions or decrease of ligand concentration is sufficient in order to favour the association of separate chains through the metal ion. The affinity of the metal ions for the pectin follows the order: Cu(II) N Ni(II) ≥ Ca(II). Acknowledgements This work is under the Research Project POCTI/1999/MGS/35653 and FCT — SFRH/BD/6293/2001. References
Fig. 7. Interaction of Ca(II) in free Ni(II) concentration in the presence of 0.50 g of pectin dm− 3. Total calcium concentration (M) in the columns of each set, from left to right: 0; 5 × 10− 4; 7.5 × 10− 4; 1 × 10− 3.
the different pectin chains, and so nickel complexation is much lower than the one predicted from the ML + ML2 model. However, in the presence of Ca(II) (θexp ≥ 0.1), the shift between experimental and theoretical values, assuming the model ML + ML2 for both metal ions, is below 13% (dashed lines included in the error bar). The model NiL + NiL2 + CaL was also tried, but shifts between experimental and theoretical values of free Ni(II) increased up to 26%. Siew and Williams [30] also noticed that Ca(II) ions interact with anionic polymers containing carboxylate groups through chemical binding. The same authors concluded that Ca(II) may interact with carboxylate groups of separate pectin chains, resulting in crosslinking, or with carboxylate groups in a single chain, forming complexes at specific sites on the chains. With increasing Ca(II) concentrations, a gel formation occurs due to the association of uncharged or oppositely charged segments on separate polymer chains. 4. Conclusion Pectin in the presence of divalent cations such as Cu(II), Ni(II), and Ca(II), and in the very low θ range (very high excess of negative charges), forms only the ML complex, where one cation interacts with one carboxylate in a specific site of the pectin chain. In this θ range the species ML2, where M interacts with carboxylate groups of two different polymer chains, is hindered due to the repulsion between the negative charges of carboxylic groups in the two independent segments of pectin. For θ ≥ 0.1 the species ML2 is formed in solution, contrary to what would be expected from an increase in θ value, i.e., from a decrease of ligand and/or increase of metal ion concentrations. This should be due to the partial neutralisation of the carboxylic groups by the positive charges of divalent cations, which induces the association of separate chains through the metal ion, leading to the ML2 formation. The presence of Cu(II) or Ca(II) may increase Ni(II) complexation to the pectin, contrary to what would be expected from a direct competition between divalent cations for the pectin binding sites. This can be explained by the partial neutralisation of pectin's negative charges by Cu(II) or Ca(II), which favours NiL2 formation through the association of independent chains. The experimental evidence for (i) the ML2 formation when the pectin concentration decreases and (ii) the decrease of the experimental-free Ni(II) concentration in the presence of Ca(II) or Cu(II), is opposite to the usual trends presented by the majority of ligands. In this paper they have been interpreted assuming a pectin inter-chain association, where divalent ions like Ni(II) interact with carboxylate
[1] A. Darvill, M. McNeil, P. Albersheim, D.P. Delmer, The primary cell walls of flowering plants, in: N.E. Tolbert (Ed.), the Biochemistry of Plants: A Comprehensive Treatise, vol.1, Academic Press, New York, 1980, pp. 91–161. [2] J. Moorby, Transport Systems in Plants, Longman Inc, New York, 1981, pp. 28–32. [3] B. Gessa, S. Deiana, Ca-polygalacturonate as a model for a soil–root interface. Fibrillar structure and comparison with natural root mucilage, Plant Soil 140 (1992) 1–13. [4] V.M. Dronnet, C.M.G.C. Renard, M.A.V. Axelos, J.F. Thibault, Characterisation and selectivity of divalent metal ions binding by citrus and sugar beet pectins, Carbohydr. Polym. 30 (1996) 253–263. [5] K.A. Davarski, S.N. Manolov, I.N. Petrova, V.D.J. Mavrov, Complex equilibria in the system M2+-polygalacturonic acid (pectin) — H2O (M = Co, Ni, Cu, Zn, Cd and Pb), J. Coord. Chem. 33 (1994) 75–81. [6] S. Deiana, G. Micera, G. Muggiolu, C. Gessa, A. Pusino, Interaction of transitionmetal ions with polygalacturonic acid — a potentiometric study, Colloids Surf. 6 (1983) 17–25. [7] M.T. Kartel, L.A. Kupchik, B.K. Veisov, Evaluation of pectin binding of heavy metal ions in aqueous solutions, Chemosphere 38 (1999) 2591–2596. [8] A. Malovíková, M. Rinaudo, M. Milas, Comparative interactions of magnesium and calcium counterions with polygalacturonic acid, Biopolymers 34 (1994) 1059–1064. [9] G. Ravanat, M. Rinaudo, Investigation on oligogalacturonic and polygalacturonic acids by potentiometry and circular dichroism, Biopolymers 19 (1980) 2209–2222. [10] J.F. Thibault, M. Rinaudo, Chain association of pectin molecules during calciuminduction gelation, Biopolymers 25 (1986) 455–468. [11] R. Kohn, Binding of divalent-cations to oligomeric fragments of pectin, Carbohydr. Res. 160 (1987) 343–353. [12] C. Garnier, M.A.V. Axelos, J.F. Thibault, Selectivity and cooperativity in the binding of calcium ions by pectin, Carbohydr. Res. 256 (1994) 71–81. [13] B. Manunza, S. Deiana, M. Pintore, C. Gessa, Interaction of Ca2+ and Na+ ions with polygalacturonate chains: a molecular dynamics study, Glycoconj. J. 15 (1998) 297–300. [14] S.K. Wiedmer, A. Cassely, M. Hong, M.V. Novotny, M.L. Riekkola, Electrophoretic studies of polygalacturonate oligomers and their interactions with metal ions, Electrophoresis 21 (2000) 3212–3219. [15] M.C. Ralet, M.J. Crépeau, H.C. Buchholt, J.F. Thibault, Polyelectrolyte behaviour and calcium binding properties of sugar beet pectins differing in their degrees of methylation and acetylation, Biochem. Eng. J. 16 (2003) 191–2001. [16] S. Deiana, L. Erre, G. Micera, P. Piu, C. Gessa, Coordination of transition-metal ions by polygalacturonic acid — a spectroscopy study, Inorg. Chim. Acta. - Bioinorg. Chem. 46 (1980) 249–253. [17] G. Manzini, A. Cesaro, F. Delben, S. Paoletti, E. Reisenhofer, Copper(II) binding by natural ionic polysaccharides. 1. Potentiometric and spectroscopic, Bioelectrochem. Bioeng. 12 (1984) 443–454. [18] C. Vilhena, M.L.S. Gonçalves, A.M. Mota, Binding of copper(II) to pectins by electrochemical methods, Electroanalysis 16 (2004) 2065–2072. [19] J. Gerendás, J.C. Polacco, S.K. Freyermuth, B. Sattelmacher, Significance of nickel for plant growth and metabolism, J. Plant Nutr. Soil Sci. 162 (1999) 241–256. [20] C. Nabais, H. Freitas, J. Hagemeyer, S-W. Breckle, Radial distribution of Ni in stemwood of Quercus ilex L trees grown on serpentine and sandy loam (umbric leptosol) soils of NE-Portugal, Plant Soil 183 (1996) 181–185. [21] G.C.L. Araujo, S.G. Lemos, C. Nabais, Nickel sorption capacity of ground xylem of Quercus ilex trees and effects of selected ligands present in the xylem sap, J. Plant Physiol. 166 (2009) 270–277. [22] S. Satoh, Organic substances in xylem sap delivered to above-ground organs by the roots, J. Plant Res. 119 (2006) 179–187. [23] S. Deiana, A. Premoli, C. Senette, Oxidation of caffeic acid by Fe(III) trapped in a Capolygalacturonate network, Plant Physiol. Biochem. 46 (2008) 435–443. [24] A. Kjøniksen, M. Hiorth, B. Nyström, Temperature-induced association and gelation of aqueous solutions of pectin. A dynamic light scattering study, Eur. Polym. J. 40 (2004) 2427–2435. [25] J.B. Wehr, N.W. Menzies, F.P.C. Blamey, Inhibition of cell-wall autolysis and pectin degradation by cations, Plant Physiol. Biochem. 42 (2004) 485–492. [26] J.S. Melo, S.F. D'Souza, Removal of chromium by mucilaginous seeds of Ocimum basilicum, Bioresour. Technol. 92 (2004) 151–155. [27] A.L. Debbaudt, M.L. Ferreira, M.E. Gschaider, Theoretical and experimental study of M2+ adsorption on biopolymers. III. Comparative kinetic pattern of Pb, Hg and Cd, Carbohydr. Polym. 56 (2004) 321–332.
36
C. Vilhena et al. / Bioelectrochemistry 77 (2009) 31–36
[28] H.G. De Jong, H.P. van Leeuwen, Voltammetry of metal-complex systems with different diffusion-coefficients of the species involved. 2. Behaviour of the limiting current and its dependence on association dissociation kinetics and lability, J. Electroanal. Chem. 234 (1987) 17–29. [29] M. Eigen, K. Tamm, Fast elementary steps in chemical reaction mechanism, Pure Appl. Chem. 6 (1963) 97–115.
[30] C.K. Siew, P.A. Williams, New insights into the mechanism of gelation of alginate and pectin: charge annihilation and reversal mechanism, Biomacromolecules 6 (2005) 963–969. [31] J. Wittmer, A. Johner, J.F. Joanny, Precipitation of polyelectrolytes in the presence of multivalent salts, J. Phys. II 5 (1995) 635–654.