Determination of copper at ng levels by in-line preconcentration and flow-injection analysis coupled with flame atomic-absorption spectrometry

Talonro,Vol. 38, No. 7, pp. 753-759, 1991 Printed in Great Britain. All rights reserved

0039-9140/9153.00+ 0.00 Pergamon Press plc

DETERMINATION OF COPPER AT ng LEVELS BY IN-LINE PRECONCENTRATION AND FLOW-INJECTION ANALYSIS COUPLED WITH FLAME ATOMIC-ABSORPTION SPECTROMETRY RAJESH F+UROHITand SUREKHA DEW* Department of Chemistry, Faculty of Science, M.S. University of Baroda, Baroda 390002, India (Received 2 May 1989. Revised 19 December

1990. Accepted I February 1991)

Sammary--Oxine/formaldehyde/resorcinol and oxine/formaldehyde/hydroquinone resins have been synthesized and their physicochemical properties studied. Conditions were optimized for the preconcentration of copper by batch extraction and column chromatography with the resins. A flow-injection analysis (FIA) manifold was constructed for the determination of copper at ng levels by preconcentration on microcolumns containing the resins, stripping, and atomic-absorption spectrometry. For batch preconcentration a pH of about 2.5-3 was optimal whereas in the FIA system a broader pH range (~2-3.5) could be used. Separations of binary mixtures of Cu(II) with Ni(II) or Pb(I1) at pg/ml level did not show any cross-contamination. In the FIA, a 2 cm long column and 2 ml/min flow-rate were adequate for quantitative uptake of copper; 50 ~1 of O.lM hydrochloric acid quantitatively eluted the copper.

There is a need for improved automatic/ semiautomatic methods for the determination of heavy metal ions in water. Flame atomicabsorption spectrometry (AAS) lacks sensitivity in detection at pug/l. levels. Electrothermalatomization and hydride-generation AAS etc. are sensitive but suffer from matrix interferences. Flow-injection analysis (FIA) has contributed’-5 to the development of several semiautomatic methods of preconcentration and determination of metal ions. These have employed activated alumina,4 oxine immobilized on silica gel, 3,6,7Chelex 100 *J resin 12296 and tri(pyridylmethyl)ethylenediamine.8 Chelating resins containing oxine groups are synthesized either from oxine, formaldehyde and resorcinol by condensation9s” or by diazotization of a poly(aminostyrene) resin followed by coupling to oxine.” The resins have low metal-exchange capacities and slow exchange rates. Their properties were improved by Pennington and Williams’* and by Parrish and Stevenson’3 by controlling the water regain and the curing conditions, but the samples required storage in moist conditions. Vernon and Nyo,14 and Parrish,i5 have reported that the resins are unstable in >2M hydrochloric acid at room temperature. We have attempted to improve the *Author for correspondence.

kinetics of metal exchange and the stabilities of the oxine-containing resins by using hydroquinone in place of resorcinol in the condensation reaction. A new synthetic route has also been established for the synthesis of oxine/ resorcinol/formaldehyde resins. Both resins are useful in the semiautomatic FIA method for the determination of copper at ng levels. Conditions have been optimized for determination of copper by (i) flow injection, (ii) a batch method and (iii) column chromatography. EXPERIMENTAL Synthesis of resins Oxine/formaldehyde/resorcinol (8HOQFR) resin was made by refluxing 0.1 mole of oxine in 50 ml of dimethylformamide (DMF) with 0.25 mole of formaldehyde for 20 min. A further 0.1 mole of resorcinol in 50 ml of DMF was added to the refluxing mixture. A reddish gel was formed within 10 min. Refluxing was continued for a further 30 min. The gel was collected, washed with DMF and methanol, and cured at 80” (open cure). The resin to be used for further studies was ground to 60-80 mesh size. Oxine/ formaldehyde/hydroquinone (8HOQFHQ) resin could not be synthesized by this method, so a base-catalysed process was used instead. Oxine (0.1 mole) in finely powdered form was stirred into 20 ml of 2M sodium hydroxide. To this

753

RAJESH PUROI-IIT and SUREKHADEW

754

suspension, 0.25 mole of formaldehyde was added with stirring and heating in a water-bath. Hydroquinone (0.1 mole) in 20 ml of 2M sodium hydroxide was added to the oxineformaldehyde mixture with stirring. The condensation reaction was continued until a black gel was formed, and this was open cured at 80”. The 60-80 mesh resin was washed thoroughly with water, converted into H+-form and used for further studies. The resins were characterized, and used for batch and column chromatographic studies of copper exchange. The optimal pH for copper uptake was found by following the literature procedureI and the copper left in solution was determined complexometrically.” The copper taken up by the resin was eluted and similarly determined. The rate constant and activation energy for formation of the chelate were obtained by observing the exchange kinetics at various temperatures. The mode of diffusion of metal ions through the solution towards the resin was determined by the interruption tesP’* and was confirmed by the method reported by Nativ et aLI During the kinetics study by the interruption test the resin beads were removed from the solution for a brief period of time (10 min) and were then re-immersed. The plots of % exchange vs. time give the nature of diffusion of the ionsI The efficiency of the eluents was tested by the batch method. Chromatographic separations

Chromatographic columns, 19 cm long, i.d. 7 mm, were prepared with the synthesized resins. Binary and ternary mixtures (25 ml) of copper, cobalt, lead and nickel (each 400 pg/ml) were passed through columns of 8HOQFR and 8HOQFHQ at pH 2 or 3 and a flow-rate of 1 ml/min, followed by washing with water. The

2 Fig. 1. FIA manifold: 1, two-way stopcock; 2, peristaltic pump; 3, injection valve; 4, column containing resin; 5, detector (AAS),

metal ions were eluted with 3M hydrochloric acid, 0. IM hydrochloric acid, 1M acetic acid and 0.1 M hydrochloric acid respectively. Characterization of the columns

The void volume of the columns was determined by the method given by Helfferich.16 The break-through capacities and column capacities were determined as described by Incz&dy.*O Determination of copper at ng level

A Varian model AA 775 atomic-absorption spectrometer with a chart recorder was used for determination of copper at 324.8 nm. The operating conditions were monochromator band-pass 1 nm, lamp current 2.5 mA, sample uptake rate 1 ml/min, air pressure 39 psig, acetylene pressure 9.8 psig. The FIA manifold was constructed as shown in Fig. 1. A Gilson Minipuls peristaltic pump was used along with a rotary injection valve (Rheodyne RH 5020). Microcolumns (4 cm long, 2 mm bore) containing an oxine resin were connected in the manifold with 0.5 mm bore Teflon tubing. Acetate solutions (0.2M) adjusted to pH 2 and 3 with a small amount of hydrochloric acid were used as carrier streams. Standard copper solutions were passed through the column at 1 ml/min flow-rate. Copper was eluted from the column by injection of 50 ,ul of OSM hydrochloric acid and determined by flame AAS. The column was further washed

Table 1. Phvsicochemical _ nronerties of the resins _ 8HQQFR 8HQQFHQ Properties Moisture content. % Density, g/cm’ Mesh size Sodium exchange capacity, t$ for sodium, min t,,* for copper, min t,,2 for nickel, min r,,* for lead, min r,,z for zinc, min Kd for nickel (pH 6) Kd for copper @H 3) K,,for lead @H 4) /cd for zinc (pH 6)

3 mmole /g

620 4.6 7 15 5 15 120 7.33 19.30 15.55 13.62

*tt,,2= time required for 50% exchange of the metal ion.

8 1.5 60-80 5.0 40 30 8 10 8 8.25 19.21 12.69 9.98

Determination

of copper at ng levels

155

Table 2. Comparison of resin properties 8HOQFR

8HOQFHQ

5M 4M 2M 0.7* 2.32’ 1.72t 6hr 7 min

Stability towards acid (maximum acidity) Copper capacity, mmole /g t,,* for sodium

5M y 1.747 40 min

Reference Present work 14 15 14 15 Present work 14 Present work

*At pH 5. tAt pH 3.

with two 50 ~1 injections of acid and passage of buffer solution through the system. A threeway stopcock was used to control the flow of standard solution and buffer. RESULTS AND DISCUSSION

The physicochemical properties of the resins synthesized (8HOQFHQ and 8HOQFR) are listed in Table 1. The resins are insoluble in most organic solvents, and in acids and alkalis of higher concentrations. They are thermally stable up to 300”. The infrared spectra show a broad band at 3400-3200 cm-’ for polymeric -OH stretching, a sharp but weak band for the aromatic tertiary C-N vibration at 1370 cm-‘, a methylene group bending vibration at 1400 cm-’ and a stretch at 1500 cm-’ indicating aromaticity. A weak band at 1600 cm-’ indicates the presence of C=N stretching. Unlike the resins reported earlier,14.” our 8HOQFR resin did not show instability in 2-4M acids, and although it had the same resorcinol/oxine/ formaldehyde ratio as the earlier resins, its

water content was lower. The exchange capacity for copper was higher (1.72 mmole/g at pH 3, againstI 0.7 at pH 5). The copper capacity of our 8HOQFHQ was also high, 1.74 mmole/g at pH 3. The t,,2 values for exchange of Cu(II), Ni(II), Zn(I1) and Pb(I1) are quite low, except for Zn(I1) on SHOQFR (Table 1). Our resins also showed greater stability towards acids and faster equilibration rates than the resins reported earlier (Table 2). The interruption testI indicated that the exchange of metal ions with these resins is governed by diffusion within the particles. The interruption gives time for the concentration gradients in the beads to level out, and in particle-controlled diffusion, the rate immediately after re-immersion is greater than that prior to the interruption. In film-controlled diffusion, there is no concentration gradient within the bead, and the diffusion rate depends on the concentration gradient in the film. The interruption does not affect the gradient and hence has no effect on the rate.

2.0 r 8HOQFHQ

8 HOQFR -

-

Q PH

PH

lC~(ll),~Ni(ll).

@Pb(ll).

oZn(lll

Fig. 2. Effect of pH on metal exchange (batch method): l Cu(II), 0 Ni(II), 0 Pb(II), 0 Zn(II); O.lM metal ion.

RAJESHF?JROHITand SUREKHA DEVI

756

8 HOQFR

8 HOQFHQ

20

0

0 05

0 15

0

025 concentration,

lCu(ll). Fig. 3. Effect of metal ion concentration

q

Ni(ll),

on exchange

The effect of pH on the metal exchange capacities is shown in Fig. 2. The selectivity, based on the distribution coefficients, was observed to be Cu > Zn > Ni > Pb at 0.1 M metal concentration, for both the resins. The uptake of metal ions by the resins increases (Fig. 3) and the distribution coefficient decreases with increasing metal ion concentration. For elution of the chelated metal ions, O.l-5M acids, sodium chloride, sodium citrate, sodium tartrate, potassium thiocyanate, thiourea and 5-50% w/w perchloric acid were tested. Copper, zinc and nickel were quantitatively eluted with 3, 0.2 and O.lM hydrochloric acid respectively, and Pb(I1) with 1M acetic acid. Conditions for separations of Cu-Ni, Cu-Pb, and Cu-Ni-Co can be predicted from the pH and elution studies. Column chromatographic separations

Mixtures (25 ml) of Cu-Ni, Cu-Pb and Cu-Ni-Co in 1: 1 and 1: 1: 1 proportions, prepared from 400 pug/ml metal ion solutions and adjusted to pH 3, were passed through the column at a flow-rate of 1 ml/min. The column was washed with water, then the metal ions were eluted with appropriate reagents. As shown in Fig. 4, separation of copper from nickel with 8HOQFR is incomplete, but is satisfactory with 8HOQFHQ. The kinetics in column processes is faster than for the batch process, because the resin surface is constantly coming into contact with fresh mobile phase. This also indicates that the mechanism of the column process is different from that of the batch process.

005

015

0.25

M

@Pb(ll).

capacity:

oZn(ll)

0

Cu(II), 0 Ni(II), 0 Pb(II), 0 Zn(I1).

When the ternary mixture of Cu-NiCo is adjusted to pH 2 and passed through the column, nickel and cobalt are detected in the effluent, but copper is retained, and can then be eluted with 3M hydrochloric acid. However, if the mixture is at pH 3, the cobalt and part of the nickel pass through the column; the copper and the rest of nickel are retained on the column and can then be eluted by gradient elution. The separation shows cross-contamination to some extent. Lead and copper can be completely separated by consecutive elution with 2M acetic acid and 3M hydrochloric acid respectively. Thermodynamics

and rate of copper exchange

The rates for uptake of copper by the resins at three different temperatures were calculated from the equations for a first order reaction -;

=kc;

-log(a

-f)=&

where a is the initial concentration of metal ion and f is the concentration left in solution after time t. Therefore (a -f) gives the uptake of metal ion by the resin. A plot of log (a -f) vs. t gave straight lines passing through the origin, for 8HOQFR, the slopes giving the values of k. However, the plots for 8HOQFHQ were not linear, and a mirror method2’s2*was used. The rate constant was calculated from the initial rate of the reaction, the appropriate part of the plot being identified by placing a mirror normal to the curve, and moving it until the reflection was a smooth continuation of the plot. The tangent was then drawn perpendicular to the mirror and

5

0

20-

40

0

2Q

40

8 WQFHQ

-

8 HQQFR

s

*

2

0

20-

40-

fb)

ml of eluent

40

+----iMCH$xxs+--3Mnc~-

80

-

8 HOQFHQ

ii

5 ._ v 3

0

20

40

40

t-0

c

lM”CC+3M

fcf

ml of eluent

O.IMHCI4-3MHCI-----

80

8 HOQFHQ

HCl”-----

8 HQQFR

separations. (a) l cu(II) and 0 WI), (b) 0 Cu(II) and 0 Pb(II), (c) l Cu(II), 0 Ni(I1) and 0 Co(H). Concentration of each metal ion in mixture, 400 Pg/ml, flow-rate 1 ml/min, pH 3, total volume of mixture 25 ml.

(a)

ml of eluent

*IMHCI-+s33M”CI

OIMHCl+3MHCI~

c-

I-

Fig. 4. ehromatographic

:.

5 i= 3

G

5

40

RAJESH F?JROHIT and

758

SUREKHADEW

Table 3. Rate constant and activation energy data

0 6 HOOFR 40

Rate constant k, set-’ Temperature “C

8HOQFR SHQQFHQ

30 40 50 Activation energy, kcallmole

1.8 2.8 4.6

1.2 2.0 3.6

9.15

9.86

E E

t

.

%

6 HOPFHQ

x

z

.a

r

2c

x

P Q

IO t 0:

from the slope of the line the rate constant (k) was calculated. The activation energy of complex formation was determined by using the Arrhenius equation

and a plot of log k us. l/T is given in Table 3. The activation energy is higher for 8HOQFHQ than for 8HOQFR, which indicates a slower exchange rate for copper on the former and is in agreement with the t,,* values (Table 1). Flow -injection system

Copper at ng level was determined with the FIA manifold shown in Fig. 1. When copper was preconcentrated on the chelating-resin microcolumn by passage of 5 ml of 1 x IO-‘M copper solution, the column effluent did not show the presence of copper when continuously analysed by AAS, indicating quantitative chelation of copper on the microcolumn. The column was then washed with buffer, and the chelated copper was eluted by injection of 50 ~1 of OSM hydrochloric acid to ensure quantitative elution. The dispersion due to the FIA system was calculated from the AAS signals for direct nebulization of a 50 ,ug/ml copper solution and for 50 ~1 of the same solution injected into the FIA system, with and without introduction of the present microcolumn. It was found that use of the microcolumn did not affect the peak height.

flow

rate. ml/min

Fig. 6. Effect of flow-rate in FIA analysis: pH 3, column length 2 cm, Cu(I1) 5 ml, 10e6M.

The dispersion was found to be 1.5. The sharp peak for the elution of copper from the microcolumn indicates faster kinetics. Continuous elution of copper with 0.5M hydrochloric acid also gives a narrow peak, of comparable height, indicating that the analyte is concentrated in a relatively narrow zone of the eluent. The dispersion of the system depends on the volume of sample injected. Conventional flame AAS gives pg/ml detection levels whereas the FIA system with a microcolumn for preconcentration gives ng/ml levels. The FIA system was optimized by varying the pH of the copper solution (Fig. 5). The optimum pH was lower than that for the batch system (Fig. 2). Tris, phosphate and acetate

0 EHOPFR .

20

0

I 2

I 4

BHOQFHO

I 6

PH

Fig. 5. Effect of pH in FIA method: flow-rate 2 ml/min, column length 2 cm, Cu(II) 5 ml, IO-‘M.

scan Fig. 7. Calibration peaks: 10 ml of 20, 40, 60, 80 and 100 ng/ml Cu solution, 20 ~1 of 0.5M HCl injected.

Determination

of copper at ng levels

buffers (0.2M) were tested as carriers in the FIA system. With phosphate buffer the system was less critically sensitive to variation in pH. The effect of column length was studied with a carrier stream flow-rate of 2 ml/min. A 2 cm column (id. 2 mm) is sufficient for preconcentration. A column length greater than 6 cm increases the dispersion coefIicient to - 15%. Increasing the flow-rate decreases the peak height and broadens the width of the peak (Fig. 6). A 2 ml/min flow-rate is recommended. Elution of copper was quantitative with 20 ~1 of > 0.1 M hydrochloric acid, but lower concentrations gave only 20-30% elution. To ensure complete elution 50 ~1 of 0.5M acid is recommended. A calibration graph for copper was obtained by passing 10 ml of each standard solution through the FI system and eluting the copper with 50 ~1 of 0.5M hydrochloric acid. Each data point was measured in triplicate. The regression coefficients for the 8HOQFR and 8HOQFHQ plots were 0.9917 and 0.9940 respectively. The lower limit of detection, based on the copper concentration equivalent to three times the standard deviation of the blank signal for the carrier stream, was 5 ng/ml with preconcentration from 15 ml of solution. REFERENCES

9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21.

1. J. RtiElca

and E. H. Hansen, Flow Injection Analysif, 1st Ed., Wiley-Interscience, New York, 1981.

22.

759

S. Olsen, L. C. R. Pesaenda, J. RtIZiEka and E. H. Hansen, Analyst, 1983, II, 905. F. Malamas, M. Bengtsson and G. Johansson, Anal. C/rim. Acta, 1984, 160, 1. Y. Zhang, P. Riby, A. G. Cox, C. W. McLeod, A. R. Date and Y. Y. Cheung, Analyst, 1988, 113, 125. S. D. Hartenstein, J. Rscka and G. D. Christian, Anal. Chem., 1985, 57,21. 2. Fang, S. Xu and S. Zhang, Anal. Chim. Acta, 1984, 164,41. M. A. Marshall and H. A. Mottola, Anal. Chem., 1985, 57, 129. M. Bengtsson, F. Malamas, A. Torstensson, 0. Regnell and G. Johansson, Mikrochim. Acta, 1985 III, 209. H. Lillin, Angew. Chem., 1954, 66, 649. J. R. Parrish, Chem. ind., 1955, 386. Idem, ibid., 1956, 137. L. D. Pennington and M. B. Williams, Znd. Eng. Chem., 1959, 51, 159. J. R. Parrish and R. Stevenson, Anal. Chim. Acta, 1974, 70, 189. F. Vernon and K. M. Nyo, ibid., 1977, 93, 203. J. R. Parrish, Anal. Chem., 1982, 54, 1890. F. Helfferich, Ion Exchange, p. 256, McGraw-Hill, New York, 1962. A. I. Vogel, Text Book o~~a~t~tative Inorganic Analysis, 3rd Ed., pp. 435, 441. ELBS, London, 1978. Y. R. E. Kressman and J. A. Kitchener, Disc. Faraday sot., 1949, 7, 90. M. Nativ, S. Goldstein and G. Schmuckler, J. Inorg. Nucl. Chem., 1975, 37, 1951. J. Inczedy, Analytical Applications of Ion Exchangers, Pergamon Press, London, 1966. A. A. Frost and R. G. Pearson, Kinetics and ~e~han~srn, p. 46. Wiley, New York, 1961. H. Diehl, Tahmta, 1989, 36, 799.