ELSEVIER

Reactive & Functional Polymers 3 1 ( 1996) l-9

REACTIVE ii FUNCTIONAL POLYMERS

Preconcentration of mercury(I1) on dithizone anchored poly(viny1 pyridine) support

Rupal Shah, Surekha Devi *
Department of Chemistry, MS. Universiry of Baroda, Baroda, 390 002, India

Accepted 31 January 1996

Abstract Dithizone, a selective and sensitive reagent for mercury, was anchored on cross-linked poly(vinyl pyridine) and was used for the preconcentration of nanogram levels of mercury. The polymeric support was characterised for the extent of coupled dithizone and thermal properties. The support was used in batch and column experiments for the preconcentration of mercury. Various conditions such as pH, temperature, flow rate and column parameters were optimised for the maximum extent of mercury loading. Figures exceeding 140 mg/g support were achieved at optimised conditions. Mixture of 0.2 M nitric acid and 2.0% ammonium nitrate in 1 : 1 proportion elutes mercury quantitatively. Separation of mercury from binary and ternary mixtures of mercury, palladium, zinc and cadmium was achieved without any cross-contamination.
Keywords:

Poly(vinyl

pyridine);

Dithizone;

Mercury; Freconcentration;

Separation

1. Introduction Polymeric supports loaded with the versatile chelating reagent dithizone were prepared by deposition of dithizone [l-3] on various polymeric materials. They were used for the preconcentration of mercury from seawater. But the recovery of mercury was poor. Garcia et al. [4] made comparative study of silica, Amberlite XAD2 and Amberlite XAD7 loaded with diethyldithiocarbamate, pyrrolidine- 1-yldithioformate and ditbizone for on-line preconcentration of mercury. They found that in the case of a dithizone loaded support, preconcentration of mercury is pH dependent. For the estimation and separation of mercury from waste waters, different chelating
* Corresponding author.

resins [5] containing dithiocarbamate, methyl thiourea, thiol or dithiocarbamate, iminodiacetic acid [6] and (2-hydroxypropyl)-picolyl amine [7] have been reported. Braun and Farag [8] used dithizone-loaded polyurethane foam for the preconcentration of mercury. Griesbach and Lieser [9] have reported fifteen sorbents, including one loaded with dithizone, but the properties and sorption conditions of the products are not reported. Uchiumi et al. [lo] have used porous polymeric resin loaded with dithizone, thionalide and bismuth01 II for the removal of As(III), Sb(II1) and Cu(I1) ions from industrial waste water. Chwastowska and Kosiarska [ 1 l] have reported the anchoring of dithizone to a diazotised styrene-divinyl benzene support. This chelating resin was further used for the sorption of Ag(I), Cd(II), Cu(II), Pb(II), Ni(II), Co(I1) and Zn(I1) at

1381-.5148/96/$15.00 Copyright 0 1996 Elsevier Science B.V. All rights reserved. PliS1381-5148(96)00029-6

R. Shah. S. Devi/Reactive

& Functional Polymers 31 (1996) I-9

static and dynamic conditions. The method was reported to be successful for the separation of copper and lead. In this paper are reported details on anchoring of dithizone on 4-vinyl pyridine-divinyl benzene copolymer. The chelating polymeric matrix is further used for the preconcentration and separation of mercury from zinc, cadmium and palladium. 2. Experimental

2.1. Reagents
All the reagents used were of AR grade. Double-distilled deionised water was used throughout the experiments. 2.2. Synthesis of sorbent Dithizone anchored poly(viny1 pyridine) resin was synthesised from chloromethylated copolymer of 4-vinyl pyridine-divinyl benzene. Polymeric support was synthesised in the following three steps. 2.2.1. Synthesis of poly(viny1 pyridine) A cross-linked polymer of 4-vinyl pyridine and divinyl benzene was prepared through suspension polymerisation following the standard procedure. In a three-necked round-bottom flask, 90 g of monomer and 10 g of 50% divinyl benzene were mixed. To this 0.4 g of a,cr azobisisobutyronitrile as free radical initiator was added. The mixture was stirred at constant temperature of 65-70C for 20 min. At the first appearance of milkiness, which indicates the beginning of polymerisation, the mixture was poured rapidly into 600 cm3 of 20% sodium chloride solution containing 6.66 g of sodium dodecyl sulphate. The suspension was heated to 80C with efficient stirring to get the desired bead size. The reaction was further continued for 8 h to ensure complete polymerisation. The polymer was then filtered, washed with water to remove soap, with dimethyl formamide to remove unreacted monomer and finally again with deionised water before drying.

2.2.2. Chloromethylation of cross-linked poly(viny1pyridine) In a l-l three-necked round-bottom flask, 50 g of cross-linked poly(viny1 pyridine), 30 g of 88% paraformaldehyde, 5 g of AlC13, 300 cm3 of 1,2-dichloroethane and 25 g of acetic anhydride were mixed and heated to -40C. Hydrochloric acid was bubbled through the reaction mixture and was heated to -60C for 25 h with continuous stirring. The colourless chloromethylated poly(viny1 pyridine) [12] formed was washed with 150 cm3 aqueous 1 : 1 dioxane, followed by a mixture of 100 cm3 aqueous 1: 1 dioxane and 10 cm3 concentrated HCl. Finally, the product was washed with 150 cm3 methanol and conductivity water to make it free from chloride and methanol. The product was dried at 70C for S-10 h and the content of chlorine was determined by Volhard method [ 131. s 2.2.3. Anchoring of dithizone to the chloromethylated poly(viny1pyridine) To the 30 g of chloromethylated poly(viny1 pyridine) in 100 cm3 of dimethyl formamide (DMF) held in a 250-cm3 Erlenmeyer flask 10 g of dehydrodithizone, an oxidised product of dithizone [14], dissolved in 50 cm3 of DMF was added. The flask was covered with aluminium foil and was shaken mechanically at room temperature for 15 days to get orange resin. The resin was filtered and washed according to the literature procedure [ 151. The wet resin was mixed with a solution of 100 cm3 of 0.5 M ascorbic acid and 50 cm3 of 1 M sodium hydroxide and was shaken for 30 min to convert dehydrodithizone to dithizone. The brown-coloured product was filtered and washed with 50 cm3 each of water, 0.01 M NaOH, water, 0.01 M HCl and, finally, with water. The resin was stored in a brown glass bottle. The physico-chemical properties of the chelating resin were studied according to the procedure reported earlier [ 161. The sorption of mercury was studied under static (batch) and dynamic (column) conditions. To determine the maximum chelating capacity

R. Shah, S. Devi /Reactive

& Functional Polymers 31 (1996) I-9

and stability of the sorbent, conditions such as pH, metal ion concentration, equilibration time and eluting agents were optimised at static conditions. Contact time (flow rate) was optimised at dynamic conditions by using columns of different sizes. The extent of mercury exchange was determined volumetrically [ 131 by mass balance in the supematant liquid as well as after elution from the resin. 2.3. Total sorption capacity The mercury total exchange capacity of the resin was determined by conditioning of -0.1 g of the resin sample with 0.2 M acetate buffer at pH 3.0 for 24 h. The resin was then equilibrated with 25.0 cm3 of 1.0 mg cme3 mercury solution for 24 h. The metal exchange capacity (Q) and its distribution coefficient (&) were calculated as Q = and & = (Co;
A (CO CA) x 100

without interruption pause and with interruption pause of 10 min (see Fig. 3 below). 2.5. Thermodynamics of exchange process

The determination of exchange rate constant k and activation energy AE was based on the kinetics of exchange of mercury ions at different temperatures: 30,35,45 and 50C. 2.6. Desorption studies Quantitative stripping of mercury from the dithizone-loaded sorbent was carried out by using various acids and complexing agents of different concentrations. 2.7. Chromatographic separations

W x % solid : w v

where, CO is the initial concentration of metal ion; CA the equilibrium concentration of metal ion; V the volume of the solution; and W the weight of the resin sample. 2.4. Exchange kinetics

A compact column of -15.0 cm length and 1.O cm I.D. was prepared with 3.0 g of resin. The column was equilibrated at pH 3.0 using 0.2 M acetate buffer solutions. Interstitial volume [ 161 (Vi cm3), break-through and column capacities [17] were determined as per the reported processes. Separations of binary mixtures of Hg(I1) and Cd(II), Hg(I1) and Zn(II), Hg(I1) and Pd(I1) and ternary mixtures of Hg(II), Cd(II), Pd(I1) and Hg(II), Cd(II), Zn(I1) were attempted. 3. Results and discussion

3.1. Structure of chelating resin

The rate of exchange of mercury in 0.2 M acetate buffer of pH 3.0 was determined by contacting -0.1 g of sorbent with 25.0 cm3 solution of 1.0 mg cme3 mercury for various time intervals. The results are presented as % exchange versus time. From the plot, the time required for 50% exchange (tl/z) was calculated. The mode of diffusion of metal ion through solution towards the resin was examined through the interruption test [ 171 and was confirmed by the method reported by Nativ et al. [ 183. The nature of the diffusion process was determined by plotting the metal exchange capacity (mmol g- versus time ) For the anchoring of dithizone to chloromethylated poly(viny1 pyridine), dehydrodithizone, an oxidised form of dithizone was used. Because of the enhanced nucleophilic properties of the mesoionic compound, displacement reactions on the chloromethylated group of the reactive polymer proceed successfully, resulting in a support carrying sulfur-bonded dehydrodithizone, denoted as P-D,. Opening of the heterocycle was achieved by treating the orange resin, P-D,, with a neutralised solution of ascorbic acid, resulting in a support carrying S-bonded dithizone, denoted as

4
~1~6~5

R. Shah, S. Devi/Reactive

& Functional Polymers 31 (1996) I-9

@-CH,Cl

%C N:;(

_,

@-CH2-S-C

C@

N--N\C6H5 chlommethybted Poly~vl ylpyr,d;nel DZ

P -HDZ

Fig. 1. Synthesis

of the sorbent.

Fig. 2. Effect of pH. Conditions: tration, 25 cm3 reaction volume.

0.1 g resin, 0.12 M Hg concen-

P-HD,. The reactions involved in the preparation of chelating resin are given in Fig. 1. 3.2. Characterisation of polymeric support

Table 2 Physico-chemical Property

properties

of the resin

Resin beads of 20-60 mesh size were used throughout the experiments. The resin was thoroughly analysed by spectral analysis, thermal techniques and elemental analysis (Table 1). The intermediate chloromethylated poly(viny1 pyridine) showed 1.4 mmol g- chlorine content. The extent of dithiozone anchored on chloromethylated intermediate was found to be -1.1 mmol g-l. This was confirmed from the results obtained in elemental analysis of the chloromethylated poly(viny1 pyridine) and dithizone loaded resin. The presence of dithizone was further confirmed through FTIR analysis of the chloromethylated poly(viny1 pyridine) and dithizone loaded resin. The appearance of the IR absorption band at 1400-1450 cm- characteristic for a C=S group linked to nitrogen in the IR spectrum of copolymer loaded with dithizone,

Moisture content (g g-l) True density (g cmm3) Void volume (cm3 g- ) Hg(II) exchange capacity at pH 3 (mmol gg ) t1/2 0.12 M mercury exchange (min) for Ko (distribution coefficient) for mercury at pH 3

0.047 1.24 0.72 0.72 20.00 340.00

Table 1 Elemental Element

is taken as the proof for the presence of the dithizone in the support. The resin showed good thermal and chemical stability up to 200C and 6 M acids. In the presence of alkalies, the colour of the resin became intense. The colour change is reversible, but the resin loses its exchange capacity. The physico-chemical properties of the dithizone-loaded support are given in Table 2. A sorbent used for more than 20 times in succession under static conditions did not show any decrease in sorption efficiency. The effect of pH on mercury exchange is illustrated in Fig. 2. It was observed that mercury exchange capacity is maximum at pH 3.0 in 0.2 M acetate buffers. 3.3. Efect of mercury ion concentration equilibrium sorption on

analysis

of products Content (%)

synthesised

Chloromethylated copolymer (resin) C H N Cl 60.95 6.85 1.27 4.97

Dithizone resin 12.65 6.94 13.39 -

loaded

In the study of the effect of mercury ion concentration on the exchange capacity of the resin, it was observed that initially the exchange capacity increases with increasing concentration of

R. Shah, S. Devi/Reactive

& Functional Polymers 31 (1996) l-9

0.4

0.08

0.16 (t.4)

Time

(min)

co

Fig. 5. Nativ plot for the diffusion s Fig. 3. Effect of mercury ion concentration on equilibrium sorp tion. Conditions: 0.1 g resin, pH 3.0, 25 cm3 reaction volume, equilibration time 24 h.

process.

mercury (Fig. 3). However, at 0.12 M initial mercury concentration the resin attains saturation, exhibiting 0.72 mM g-t exchange capacity. 3.4. Rate of mercury ion exchange From the kinetic study (Fig. 4), it was observed that 9 h contacting time is required for 100% exchange of mercury. However, the time required for the 50% exchange of mercury, (t1/2) was observed to be 20.0 min. For the study of the exchange mechanism an interruption test was carried out as described by Helfferich [16]. According to the shell progression mechanism, exchange of metal ions is governed by either particle diffusion or film diffusion. From the results obtained in the interruption test (Fig. 4), it was

100

86

.//f-----

.O

observed that mercury exchange is a particle diffusion process. Because 10.0 min interruption pause during the test allows time for the concentration gradients developed in the beads to level out, resulting in a increase in exchange rate after reimmersion. In film diffusion, no concentration gradient exists in the beads and hence the interruption does not affect the concentration gradient across the beads, resulting into no effect on exchange rate. The particle diffusion is also confirmed with the mathematical model proposed by Nativ et al. [ 181 for a film diffusion process. A plotoftimeversus[1-3(1-x)2~3+2(1-x)] was constructed (Fig. 5) where x is the fraction conversion of the resin, which is the ratio between the measured and maximum exchange capacity. According to this model a straight line passing through the origin confirms the shell progression mechanism controlled by film diffusion through the reacted layer. However, we are getting a plot with the intercept on the x-axis, indicating that it is not a film diffusion but a particle diffusion process. For an ion exchange process, the exchange reaction can be written as mRA + B+ + &zB + mA+ where I?A indicates solid phase having mA exchangeable ions and B+ is counter exchangeable ion in the solution phase. In the present study, charges on A and B are 1 and 2. i.e., m = 2 and n = 1. Adsorption of ions from solution to solid is governed by a Langmuir adsorption isotherm. This can be applied to the ion

B
(minuteS\ lfme

(hours)

Fig. 4. Kinetic study. Symbols: e-e, rate of exchange; O-G, rate of exchange without interruption; and o- - -0, rate of exchange with interruption. Conditions: 0.1 g resin, 0.15 M Hg concentration, 25 cm3 reaction volume.

R. Shah, S. Devi/Reactive

& Functional Polymers 31 (1996) 1-9 Table 3 Exchange Time (min) 10 20 30 60 120 180 Rate constant (X 10-3 SC )

exchange process with specific assumptions. The first assumption is that the selectivity coefficient or separation factor remains unchanged or little changed over the exchange involved. Secondly, for homovalent and heterovalent exchange, the concentration of one of the ions or of the adsorbed solution is constant. Based on these assumptions Misak [ 191 has modified the Langmuir adsorption isotherm for the ion exchange process. Accordingly the surface exchange process can be governed by $+A(l-;)+~ where qA is the amount of metal sorbed (mmol), CA is equilibrium concentration of metal ion in the solution, Ca is initial concentration of metal ion, Q is the maximum exchange capacity, a is the separation factor and the values of m and n are 2 and 1, respectively. From a plot of CA/qA versus CA (Fig. 6) for the rate of exchange, cxand Q were calculated from the slope and intercept. The maximum exchange capacity Q was found to be 0.603 mm01 g-l, which is in good agreement with the experimental value (0.72 mmol g-l). From the results for Q and CX, selectivity coefficient K was the calculated by using the equation [ 191

kinetics and rate constants log(a - f) at 30C -1.041 -1.11 -1.18 -1.29 -1.49 - 1.61 8.23 35C - 1.06 -1.12 -1.20 - 1.33 -1.52 -1.70 8.53 45C -1.09 -1.16 - 1.25 - 1.37 -1.59 -1.86 9.91 50C -1.10 -1.18 -I .27 -1.41 - 1.64 ;; .;Bs

100% exchange for 30 min to 12 h, K values were obtained in the range of 8 x lo-* to 20 x lo-*. 3.5. Thermodynamics and rate of exchange

Reaction rates for the exchange of mercury on the resin were determined at four different temperatures. The rate constant k (for the initial exchange) was calculated using equations for the first-order reaction: -dc/dt = kc and - log(a - f) = kt/2.303

For the given system, cz and Q will remain constant. But the selectivity coefficient will vary with varying concentration of metal ions. For various CA and QA values ranging from 57.4% to

2.0

*SI.O

where is the initial concentration of metal ion a and is the concentration of ions exchanged f on the resin. Data for the rate model is given in Table 3. It was observed that the plot of -log(a f) versus t is a straight line not passing through the origin. From the slope of the straight line, the rate constant k was calculated (Table 3). Where a straight line is not obtained, a mirror method [20] was used to calculate the rate constant from the initial rate of the exchange reaction. From the log k versus l/T plot (Fig. 7), the activation energy required for complex formation was determined using the Arrhenius equation. The calculated activation energy AE was observed to be 1.04 x lo4 J mol-' . A comparatively higher activation energy was observed for the exchange of mercury with the resin. 3.4. Desorption of mercury from the resin Desorption of mercury from the resin was tried with various acids of different strengths.

Fig. 6. Modified Langmuir

adsorption

isotherm for ion exchange.

R. Shah, S. Devi/Reactive

& Functional Polymers 31 (1996) 1-9

Table 5 Column characteristics Column length (cm) Interstitial volume (VO, cm3) Break-through capacity (mmol) Total capacity (mmol) Degree of utilization Ratio of break-through capacity 4.6 2.1 0.372 1.02 0.364 and total capacity.

Fig. 7. Arrhenius

plot.

Table 4 Effect of eluting agents Sr. no. 1 Reagents M HCl 0.1 0.5 1.o M H2S04 1.0 2.0 3.0 M thiourea 3.0 M HNO, 0.1 0.5 1.0 2.0 3.0 MHNOs+%N&N03(1:1) 0.1 + 0.5 0.1 + 1.0 0.1 + 2.0 0.2 + 2.0 1.0 + 2.0 used % Elution

10.52 7.70 5.10

70.25 74.30 92.00

84.60

55.40 63.15 14.40 91.65 91.50

62.15 76.91 99.98 99.98 99.98

It was observed that HN03 of higher strength and a 1 : 1 mixture of 0.1 M HNOs and 2.0% NH4NOs were useful for quantitative desorption of mercury. Percentage desorptions with various eluting reagents of different strengths are given in Table 4. 3.7. Chromatographic separations

From the results obtained for the effect of pH on the metal exchange and the elution study,

it was observed that mercury can be separated from many other elements due to the difference in pH of chelation and eluting reagents. Hence the synthetic mixtures (each containing 10.0 mg of individual metal in 30 cm3 of buffer solution) (i) mercury and cadmium, (ii) mercury and palladium, (iii) mercury and zinc, (iv) mercury, cadmium and palladium, (v) mercury, cadmium and zinc were passed through the column at a flow rate of 1 cm3 min- The column effluents . were analysed for every metal ion in the mixture. The properties of the column under study are given in Table 5. For the separation of binary mixtures of mercury and cadmium or zinc, mixtures were passed through the column at pH 3 using 0.2 M acetate buffer at a flow rate of 1 cm3 min- Column . effluent showed the presence of cadmium or zinc, as in acetate buffer exchange of cadmium and zinc does not take place. On the other hand, mercury was not found in the column effluents, indicating quantitative exchange in acetate buffer of pH 3. The desorption of mercury was carried out using a 1 : 1 mixture of 0.5 M nitric acid and 2.0% ammonium nitrate. Quantitative desorption of mercury did not take place even at higher concentrations of nitric acid or any other acids as well as ammonium nitrate alone. But a combination of HN03 and NHdNOs was effective for quantitative elution (Table 4). The representative separation pattern for mercury from cadmium and from zinc is given in Figs. 8 and 9. No cross-contamination was observed at any level. In the separation of mercury and palladium, mixtures were passed through the column at pH 3 using 0.3 M acetate buffer. At this condition,

R. Shah, S. Devi/Reactive & Functional Polymers 31 (1996) l-9

I 1

k------AI-------AZ+

V&me

effluent (cm31

Fig. 8. Separation of Hg and Cd. . =Hg; o =Cd. Ai =acetate buffer of pH 3. AZ = 0.5 M HNO3 + 2% NH4N03 (1: 1).

40

4FA2 -i

In the separation of ternary mixtures of mercury, cadmium and zinc, mixtures were passed through the column at pH 3. Here mercury was retained on the column, whereas zinc and cadmium were found in the column effluent. The mercury was eluted with a 1 : 1 mixture of 0.5 M nitric acid and 2.0% ammonium nitrate. The column effluent containing zinc and cadmium was passed through the column at pH 5 of 0.2 M citrate phosphate buffer. Both zinc and cadmium were retained on the column quantitatively. The zinc was further eluted with 0.1 M nitric acid and cadmium was eluted with 0.5 M acetic acid. The representative separation pattern for Hg, Cd and Zn is given in Fig. 11. In the separation of mercury, cadmium and palladium, 1 : 1 : 1 mixtures were passed through the column at pH 3 of 0.2 M acetate buffer. Here cadmium was not retained on the column and

40 Volume

00 effluent

km3

120

Fig. 9. Separation of Hg and Zn. ?? =Hg; o =Zn. Al = acetate buffer of pH 3. AZ = 0.5 M HNO3 + 2% NH4N03 1: 1).

40:
Volume effluentId)

both mercury and palladium are retained on the column. Mercury was eluted with a 1 : 1 mixture of 0.5 M nitric acid and 2.0% ammonium nitrate and palladium was eluted with 0.2 M HCl + 1.O% thiourea solution. Only HCl was found to be inefficient for the elution of palladium. Thiourea alone could give only 90% elution. However, the mixture of HCl and thiourea resulted into quantitative elution of palladium (Fig. 10).

Fig. 11. Separation of Hg, Cd and Zn. . =Hg; O=Zn. A2~0.5 M HN03+2% NH.+NO3 (1:l). HN03. As = 0.5 M CH3COOH.

o =Cd; and A4=0.1 M

40

A2 --A4---+-+

30

A2.A3

Volume

effluent

lcm51

Volume

effluent

(cm 1

Fig. 10. Separation of Hg and Pd. . =Hg; A =Pd. A2 =0.5 M HNO3 +2% N&N03 (1 : 1). A3 =0.2 M HCl+ 1% thiourea (1: 1).

Fig. 12. Separation of Hg, Cd and Pd. . = Hg; o = Cd; and A = Pd. Al=acetate bufferofpH3(1:1).A2=0.5MHNO3+NbN03 (1 : 1). A3 =0.2 M HCI + 1% thiourea (1 : 1).

R. Shah, S. Devi/Reactive

& Functional Polymers 31 (1996) l-9 (Cl. 210-638, (81 [9] [IO] B 01 D 13/00) 1988, Appl. 937,950

9 (1986)

was found in the column effluent, whereas mercury and palladium were retained on the column. Selective elution of mercury and palladium was done as described earlier using a 1 : 1 mixture of 0.5 M HN03 and 2.0% NbNOs and a 1: 1 mixture of 0.2 M HCl and 1% thiourea, respectively (Fig. 12). No cross-contamination was observed in any of the separations under study. References
PI A. Chow and D. Buksak, Can. J. Chem., 5 (1975) 132. PI Y.K. Lee, K.J. Whang and K. Ueno, Talanta, 22 (1975) 535. [31 A.G. Howard and M.H. Abbab-Zavar, Talanta, 26 (1979)
895. 141 M.F. Garcia, R.P. Garcia, N.B. Garcia and A.S. Medel, Talanta, 41 (1994) 1833. 151 A. Lezzi, A. Roggero, and S. Cabianco (Eniricerche S.P.A.), Eur. Pat. Appl. EP 547,720 (C1.C 08 F8/34), 1993. I61 N.S.C. Becker and R.J. Eldridge (CSIRO, Clayton 3168 Australia), React. Polym., 21 (1993) 5-14. [71 R.R. Grinstead (Dow Chemical Co.), U.S. US 4,741,831

[ 111
[12] [13] [14] [15] [16] [17] [18]

[ 191
[20]

19 PP. T. Braun and A.B. Farag, Anal. Chim. Acta, 69 (1974) 85. M. Griesbach and K.H. Lieser (Riedel de Haen A.-G.), Eur. Pat. Appl. 14,222 (Cl. 08 F8/30), 1979. A. Uchiumi, S. Tokunaga, T. Eyama and H. Shibayama (Natl. Inst. Master Chem. Res., Tsukuba, Japan 305). Int. Conf. Process. Mater. Prop., 1 (1993) 101-104. J. Chwastowska and E. Kosiarska, Talanta, 35 (1988) 439. J.D. Spivack and S. Vauey, US Pat. 3281505, 1966. A.I. Vogel. Handbook of Quantitative Inorganic Analysis. Longman, London, 1978. H.M.N.H. Irving, A.M. Kiwan, D.C. Rupainwar and S.S. Sahena, Anal. Chim. Acta, 56 (1971) 205. M. Grote and A. Kettrup, Anal. Chim. Acta, 172 (1985) 223-239. F. Helfferich. Ion Exchange. McGraw-Hill, New York, 1962. J. Inczedy. Analytical Applications of Ion Exchangers. Pergamon, Oxford, 1966. M. Nativ, S. Goldstein and G. Schmuckler, J. Inorg. Nucl. Chem., 37 (1975) 1951. N.Z. Misak, React. Polym., 21 (1993) 53-64. A.A. Frost and R.G Pearson. Kinetics and Mechanism. Wiley, New York, 196 1.