- Email: [email protected]
Spectrophotometric determination of trace amounts of selenium with catalytic reduction of bromate by hydrazine in hydrochloric acid media
Tahta, Vol. 39, No. 8, pp. 993-996, 1992 Printed in Great Britain. All rights reserved
0039-9140/92 ss.00 + 0.00 Copyright 0 1992 Pergamon Press Ltd
SPECTROPHOTOMETRIC DETERMINATION OF TRACE AMOUNTS OF SELENIUM WITH CATALYTIC REDUCTION OF BROMATE BY HYDRAZINE IN HYDROCHLORIC ACID MEDIA A. AFKHAMI,A. SAFAVI*and A. MASSOUMI Department of Chemistry, College of Sciences, Shiraz University, Shiraz, Iran (Received 7 June 1991. Revised 24 September 1991. Accepted 26 September 1991)
Sammary-A method is presented for the determination of selenium, based on the catalytic effect of selenium(W) on the reduction reaction of BrO; by N,H,.ZHCl. The decolourization of Methyl Orange by the reaction products was used to monitor the reaction spectrophotometrically at 525 nm. This method is precise, highly sensitive, simple, rapid, widely applicable and selective for the determination of selenium(IV) and total selenium. The variables which affected the reaction rate were fully investigated and the optimum conditions were established. Selenium, as low as 1 ng/ml, can be determined by this method. The relative standard deviation of 20 ng of selenium was 0.94% (N = 10). The method was applied to the determination of Se(W) in a health-care product.
Selenium is becoming increasingly important from a toxicological and physiological point of view.’ Selenium is an essential micro-nutrient for animals although excesses have been known to cause toxicity. On the other hand selenium deficiency syndromes have also been reported. It has been reported that selenium acts as an antidote for mercury, cadmium, arsenic and other elements2 On the basis of this background, the determination of micro amounts of selenium is becoming increasingly important. Determination of selenium as hydrogen selenide by atomicabsorption spectrometry is subject to many interferences3 such as Ag, Cu, Ni, Pd, Pt, Rh, Ru and Sn, and it is necessary to ensure that the matrix being analysed does not interfere. An alternative method for determination of traces of selenium is based on the spectrophotometric or fluorimetric measurement of piazselenol formed by the reaction of selenium(W) with diaminonaphthalene.4 These methods are tedious, time consuming and the reagents used are toxic and relatively unstable. Catalytic kinetic methods, which are simple, sensitive and selective for many elements,’ have also been applied to the determination of this element, and some of them are described below. Selenium(W) has been determined by using its
catalytic effect on the reduction of methylene blue by sodium sulphide;* this method had a detection limit of 0.05 pg. Another catalytic method for the determination of selenium is based on the reduction of 1,4,6,1 l-tetraazanaphthacene by glyoxal and hypophosphorous acid and has a detection limit of 0.03 ,ug of Se.g This method suffers from a non-linear calibration and has the drawback of requiring a reagent that is not commercially available. A third method is based on the catalytic effect of selenium on the reduction of tetranitro blue tetrazolium by dithiothreitol.‘O Recently, we reported a highly sensitive and selective method for the determination of selenium based on its catalytic effect on the reduction reaction of resazurin by sulphide.” This paper describes the development of a new method for the determination of selenium(W) based on its catalytic effect on the reduction reaction of bromate by hydrazine dihydrochloride. This method which is very rapid, simple, highly selective and sensitive, uses only readily available reagents. EXPERIMENTAL
Reagents Demineralized triply-distilled water was used throughout. Selenite standard solutions were prepared by dissolving 1.40 g of selenium dioxide (Merck) in water and diluting to the
*Author for correspondence. 993
A. APKHAMI et al.
994
mark in a one-litre standard flask. This solution was standardized iodimetrically.i2 Working solutions were prepared by diluting the stock solution with water. A OSM solution of hydrazine dihydrochloride was prepared by dissolving 13.12 g of N2H,. 2HCl (BDH) in water and diluting to the mark in a 250-ml standard flask. Bromate (0.24M) was prepared by dissolving 10.02 g of potassium bromate (Merck) in water and diluting to the mark in a 250-ml standard flask. Methyl Orange, 0.1% was prepared by dissolving 0.1 g of Methyl Orange (Merck) in water and diluting to the mark in a loo-ml standard flask. A pH = 1 buffer solution (Merck) of glytine-hydrochloric acid was used. Apparatus
A Perkin-Elmer model 35-spectrophotometer with a l-cm glass cell was used for absorbance measurements. Procedure
All the measurements were performed at 25.0 + 0.1”. Into a lo-ml standard flask introduce a suitable aliquot of the sample solution containing 0.05-8.0 pg of selenium(IV) and 1.5 ml of buffer solution. Add 2 ml of 0.5M hydrazine dihydrochloride followed by 0.2 ml of 0.1% methyl orange solution and 1 ml of 0.24M bromate, then dilute the solution to the mark with water and mix well. Transfer a portion of the solution to a glass cell within 30 set and measure the absorbance at 525 nm. Find the initial reaction rate (AA/AT) during the first three minutes after starting the reaction. RESULTS AND DISCUSSION
Selenium(IV) has a catalytic effect on the reduction reaction of bromate by hydrazine dihydrochloride. The steps involved in this reaction are:
0.50
1
0.40 t
& 0.30 S m 0.20 0.10 I
,
I
I
0.50
1.00
Iso
2.00
0
2.50
3.00
PH
Fig. 1. Effect of pH on the rate of (m) catalysed and (0) uncatalysed reaction.
hydra&e in the medium slows down step (1) which is fairly fast in its absence or when the medium is very acidic. I3 Methyl Orange reacts with the products of the reaction (bromine and chlorine) and is decolourized. This reaction can be monitored spectrophotometrically by measuring the decrease in absorbance versus time for the first three minutes of the reaction. Eflects of variables
The graph of the decrease in absorbance versus time at 525 nm was linear during the first three minutes and the slope (tan 01 = dA/dt) was used as a measure of the initial reaction rate. The reaction rate of both catalyzed and uncatalyzed reactions increased with decreasing pH. Figure 1 shows the influence on the catalyzed and uncatalyzed reaction of pH in the range l&2.5. Samples with pH lower than 1 were not tested because the rate of the uncatalyzed reaction is too fast at these pH values. Therefore, pH 1 was selected for routine works. The influence of the concentrations of bromate and hydrazine dihydrochloride on the reaction rate was also studied. The reaction rate increases as the concentration of these two reagents is increased. On the other hand a high increase in the reaction rate causes perturbation in the absorbance reading by the appearance
2BrOr + lOCl- + 12H+sBr,
+ 5Cl2 + 6H2O
(~10~)
(1)
2C1, + (NzHg)*+eN2 + 4Cl- + 6H+ (fast)
(2)
2Br, + (N2H6)2+eN2 + 4Br- + 6H+ (fast)
(3)
The first step of the reaction is slow and the second and third are fast. Selenium(IV) acts as a catalyst for the first step. The presence of
Table 1. Accuracy and precision of the recommended wxedure Amount of Se(W) taken, ng/ml 20 50
300 500
Relative error, % 1.4 1.8
Relative standard deviation, % (n = 10) 0.94 0.83 0.53 0.10
Spectrophotometric
determination of trace amounts of Se
Table 2. Effect of diverse species on the determination of 0.5 &ml se(lv) Tolerance limit
rktbl
IOll
Th(Iv), se(vI), W(H). LalIII).+ O&ii), do@), I&@), kg@), (Xv),+ MowI), W(vI), Ce(HI), KO), NaO. TV), Uq+ , Zn(II), NKII), Fe(III),* Ca(II), Mg(II), As(V), Te(vI), AsQII), Cd(H), Co(III), Al(III), PO:-, CH,OO, F-, SO;, NO; I-, IO,, BrH&+, Cu2+* $Id(II)t
100
30 10 2 n.nsA4
*After addition of 0.5 ml of O.lM EDTA. tAfier addition of 1 ml of saturated dimethyl-glyoxime solution.
of nitrogen bubbles. Therefore the optimum concentration of these two reagents is the concentration at which the reaction has a high reaction rate and N, bubbling does not perturb the absorbance reading. These concentrations were found to be 0.10 and 0.024M for N,H,+2HCl and BrO; , respectively. Increasing temperature increased the reaction rate of both the catalyzed and uncatalyzed reactions, but at high temperatures N2 gas bubbles are formed. For this reason and also for simplicity 25” & 0.1 was selected for routine work. Ionic strength up to 0.81M had no effect on the reaction rate. Analytical parameters The calibration graph was obtained under the experimental conditions chosen. A plot of initial reaction rates as a function of selenium(W) concentration is linear in the range 5-800 ng/ml, with the equation:
tan o! = 1.482 x 10m4C + 0.0806 with r = 0.9994, where C is the concentration of selenium(IV) in ng/ml. The limit of detection, defined as the average of the blank value plus three times its standard deviation, was 1 ng/ml. Table 1 shows the accuracy and precision of the recommended procedure. Eflects of foreign species
The effects of various cations and anions on the determination of 0.5 pug/ml selenium(W) were studied. The results are summarized in
995
Table 2. The tolerance limit was defined as the concentration of added ion causing less than 3% relative error. Most ions did not interfere with the determination even when they are present at concentrations 200 times as great as that of selenium(W). Positive interferences were observed from lanthanum(III), copper( palladium(II), vanadium(III), tellerium(IV) and bromide because they could also catalyse the reaction. Negative interferences were observed from iodate and iodide, because they inhibit the indicator reaction. H&+ interferes by precipitation with the reagents. Iron(II1) and cerium(IV) interfere by changing the colour of the solution probably by complexation with Methyl Orange. The interfering effect of copper(I1) in concentrations less than 2 pg/ml and palladium(I1) in concentrations less than 2 pg/ml can be removed by the addition of EDTA and dimethylglyoxime, respectively. However, higher concentrations of copper(I1) can be removed by extracting its complex with dimethylglyoxime into chloroform, and higher concentrations of palladium(I1) can be removed by extracting its complex with oxine into chloroform. The interfering effect of vanadium(II1) can also be removed by extracting its complex with oxine into chloroform. Determination product
of selenium
in a health-care
In order to test the described method, selenium was determined in a health-care product (a shampoo for the treatment of dandruff). Sample dissolution was carried out by the procedure described by Belarra et aLI Approximately 1 g of sample was weighed into a loo-ml Kjeldahl flask. Concentrated sulphuric acid (1 ml) was added and the mixture was heated to fuming for 15 min. The solution was allowed to cool and 5 ml of 30% w/v hydrogen peroxide was added. The mixture was boiled vigorously to eliminate excess of hydrogen peroxide and the flask was allowed to cool. The mixture was diluted to 1.0 litre with triply distilled water. A 0.5ml volume of this solution was taken for determination of selenium(W) as in the recommended procedure. The results showed the presence of 13.45 St 0.42 g/l. selenium (n = 7). (Manufacturer’s value is given as 13.81 g/l. selenium as selenium sulfide.)
996
A. ApKHMnet al. CONCLUSIONS
This method can be used to determine selenium as low as 1.0 ng/ml without the need for any preconcentration step. The method is very simnle and more selective than the other kinetic methods reported previously or hydride generation atomic-absorption spectrometry which is subject to many interferences. Also, the limit of detection obtained by the proposed method (1 .O ng/ml) is lower than that obtained by hydride generation atomic-absorption spectroscopy (5 ng/ml).3 Since the method only responds to selenium(IV) species, the determination of selenium(IV) in the presence of selenium(W) is possible. Acknowledgement-The authors wish to express their gratitude to Shiraz University Research Council for their support of this work.
REFERENCES H. RobbemhtandR.VanGrieken, Takmta, 1982,29,823. K. Toei and Y. Shimoishi, ibid., 1981, 2S, 867. A. E. Smith, Andpt, 1975, 100, 300. M. W. Brown and J. H. Watkinson, Anal. Chim. Acta, 1977, 89, 35. 5. Y. Shim&hi and K. Toei, ibid., 1978, 100,65. 6. N. D. Michie, E. J. Dixon and N. G. Bunton, J. Assoc. Anal. Chetn., 1978, 61,48. 7. K. B. Yatsimirskii, Kinetic Methodp of Analysis, Pergamon Press, Oxford, 1966. 8. P. W. West and T. V. Ramakrishna. Anal. Ckem., 1968, 48,966. 9. T. Kawashima and M. Tanaka, Anal. Chim. Acta, 1968, 48, 143. 10. W. C. Hawkes, ibid., 1986, 183, 197. 11. A. Safavi, A. Afkhami and A. Massoumi, ibid., 1990, 232, 351. 12. A. I. Vogel, A Text Book of Quantitative Inorganic Analysis, 3rd Ed., Longman, London, 1961. 13. P. Linares, M. D. Luque de Castro and M. Valcarcel, Analyst, 1986, 111, 1405. 14. M. A. Belarra, F. Gallarta, J. M. Anxano and J. R. Castillo, J. Anal. At. S’cr., 1986, 1, 141. l. 2. 3. 4.
0039-9140/92 ss.00 + 0.00 Copyright 0 1992 Pergamon Press Ltd
SPECTROPHOTOMETRIC DETERMINATION OF TRACE AMOUNTS OF SELENIUM WITH CATALYTIC REDUCTION OF BROMATE BY HYDRAZINE IN HYDROCHLORIC ACID MEDIA A. AFKHAMI,A. SAFAVI*and A. MASSOUMI Department of Chemistry, College of Sciences, Shiraz University, Shiraz, Iran (Received 7 June 1991. Revised 24 September 1991. Accepted 26 September 1991)
Sammary-A method is presented for the determination of selenium, based on the catalytic effect of selenium(W) on the reduction reaction of BrO; by N,H,.ZHCl. The decolourization of Methyl Orange by the reaction products was used to monitor the reaction spectrophotometrically at 525 nm. This method is precise, highly sensitive, simple, rapid, widely applicable and selective for the determination of selenium(IV) and total selenium. The variables which affected the reaction rate were fully investigated and the optimum conditions were established. Selenium, as low as 1 ng/ml, can be determined by this method. The relative standard deviation of 20 ng of selenium was 0.94% (N = 10). The method was applied to the determination of Se(W) in a health-care product.
Selenium is becoming increasingly important from a toxicological and physiological point of view.’ Selenium is an essential micro-nutrient for animals although excesses have been known to cause toxicity. On the other hand selenium deficiency syndromes have also been reported. It has been reported that selenium acts as an antidote for mercury, cadmium, arsenic and other elements2 On the basis of this background, the determination of micro amounts of selenium is becoming increasingly important. Determination of selenium as hydrogen selenide by atomicabsorption spectrometry is subject to many interferences3 such as Ag, Cu, Ni, Pd, Pt, Rh, Ru and Sn, and it is necessary to ensure that the matrix being analysed does not interfere. An alternative method for determination of traces of selenium is based on the spectrophotometric or fluorimetric measurement of piazselenol formed by the reaction of selenium(W) with diaminonaphthalene.4 These methods are tedious, time consuming and the reagents used are toxic and relatively unstable. Catalytic kinetic methods, which are simple, sensitive and selective for many elements,’ have also been applied to the determination of this element, and some of them are described below. Selenium(W) has been determined by using its
catalytic effect on the reduction of methylene blue by sodium sulphide;* this method had a detection limit of 0.05 pg. Another catalytic method for the determination of selenium is based on the reduction of 1,4,6,1 l-tetraazanaphthacene by glyoxal and hypophosphorous acid and has a detection limit of 0.03 ,ug of Se.g This method suffers from a non-linear calibration and has the drawback of requiring a reagent that is not commercially available. A third method is based on the catalytic effect of selenium on the reduction of tetranitro blue tetrazolium by dithiothreitol.‘O Recently, we reported a highly sensitive and selective method for the determination of selenium based on its catalytic effect on the reduction reaction of resazurin by sulphide.” This paper describes the development of a new method for the determination of selenium(W) based on its catalytic effect on the reduction reaction of bromate by hydrazine dihydrochloride. This method which is very rapid, simple, highly selective and sensitive, uses only readily available reagents. EXPERIMENTAL
Reagents Demineralized triply-distilled water was used throughout. Selenite standard solutions were prepared by dissolving 1.40 g of selenium dioxide (Merck) in water and diluting to the
*Author for correspondence. 993
A. APKHAMI et al.
994
mark in a one-litre standard flask. This solution was standardized iodimetrically.i2 Working solutions were prepared by diluting the stock solution with water. A OSM solution of hydrazine dihydrochloride was prepared by dissolving 13.12 g of N2H,. 2HCl (BDH) in water and diluting to the mark in a 250-ml standard flask. Bromate (0.24M) was prepared by dissolving 10.02 g of potassium bromate (Merck) in water and diluting to the mark in a 250-ml standard flask. Methyl Orange, 0.1% was prepared by dissolving 0.1 g of Methyl Orange (Merck) in water and diluting to the mark in a loo-ml standard flask. A pH = 1 buffer solution (Merck) of glytine-hydrochloric acid was used. Apparatus
A Perkin-Elmer model 35-spectrophotometer with a l-cm glass cell was used for absorbance measurements. Procedure
All the measurements were performed at 25.0 + 0.1”. Into a lo-ml standard flask introduce a suitable aliquot of the sample solution containing 0.05-8.0 pg of selenium(IV) and 1.5 ml of buffer solution. Add 2 ml of 0.5M hydrazine dihydrochloride followed by 0.2 ml of 0.1% methyl orange solution and 1 ml of 0.24M bromate, then dilute the solution to the mark with water and mix well. Transfer a portion of the solution to a glass cell within 30 set and measure the absorbance at 525 nm. Find the initial reaction rate (AA/AT) during the first three minutes after starting the reaction. RESULTS AND DISCUSSION
Selenium(IV) has a catalytic effect on the reduction reaction of bromate by hydrazine dihydrochloride. The steps involved in this reaction are:
0.50
1
0.40 t
& 0.30 S m 0.20 0.10 I
,
I
I
0.50
1.00
Iso
2.00
0
2.50
3.00
PH
Fig. 1. Effect of pH on the rate of (m) catalysed and (0) uncatalysed reaction.
hydra&e in the medium slows down step (1) which is fairly fast in its absence or when the medium is very acidic. I3 Methyl Orange reacts with the products of the reaction (bromine and chlorine) and is decolourized. This reaction can be monitored spectrophotometrically by measuring the decrease in absorbance versus time for the first three minutes of the reaction. Eflects of variables
The graph of the decrease in absorbance versus time at 525 nm was linear during the first three minutes and the slope (tan 01 = dA/dt) was used as a measure of the initial reaction rate. The reaction rate of both catalyzed and uncatalyzed reactions increased with decreasing pH. Figure 1 shows the influence on the catalyzed and uncatalyzed reaction of pH in the range l&2.5. Samples with pH lower than 1 were not tested because the rate of the uncatalyzed reaction is too fast at these pH values. Therefore, pH 1 was selected for routine works. The influence of the concentrations of bromate and hydrazine dihydrochloride on the reaction rate was also studied. The reaction rate increases as the concentration of these two reagents is increased. On the other hand a high increase in the reaction rate causes perturbation in the absorbance reading by the appearance
2BrOr + lOCl- + 12H+sBr,
+ 5Cl2 + 6H2O
(~10~)
(1)
2C1, + (NzHg)*+eN2 + 4Cl- + 6H+ (fast)
(2)
2Br, + (N2H6)2+eN2 + 4Br- + 6H+ (fast)
(3)
The first step of the reaction is slow and the second and third are fast. Selenium(IV) acts as a catalyst for the first step. The presence of
Table 1. Accuracy and precision of the recommended wxedure Amount of Se(W) taken, ng/ml 20 50
300 500
Relative error, % 1.4 1.8
Relative standard deviation, % (n = 10) 0.94 0.83 0.53 0.10
Spectrophotometric
determination of trace amounts of Se
Table 2. Effect of diverse species on the determination of 0.5 &ml se(lv) Tolerance limit
rktbl
IOll
Th(Iv), se(vI), W(H). LalIII).+ O&ii), do@), I&@), kg@), (Xv),+ MowI), W(vI), Ce(HI), KO), NaO. TV), Uq+ , Zn(II), NKII), Fe(III),* Ca(II), Mg(II), As(V), Te(vI), AsQII), Cd(H), Co(III), Al(III), PO:-, CH,OO, F-, SO;, NO; I-, IO,, BrH&+, Cu2+* $Id(II)t
100
30 10 2 n.nsA4
*After addition of 0.5 ml of O.lM EDTA. tAfier addition of 1 ml of saturated dimethyl-glyoxime solution.
of nitrogen bubbles. Therefore the optimum concentration of these two reagents is the concentration at which the reaction has a high reaction rate and N, bubbling does not perturb the absorbance reading. These concentrations were found to be 0.10 and 0.024M for N,H,+2HCl and BrO; , respectively. Increasing temperature increased the reaction rate of both the catalyzed and uncatalyzed reactions, but at high temperatures N2 gas bubbles are formed. For this reason and also for simplicity 25” & 0.1 was selected for routine work. Ionic strength up to 0.81M had no effect on the reaction rate. Analytical parameters The calibration graph was obtained under the experimental conditions chosen. A plot of initial reaction rates as a function of selenium(W) concentration is linear in the range 5-800 ng/ml, with the equation:
tan o! = 1.482 x 10m4C + 0.0806 with r = 0.9994, where C is the concentration of selenium(IV) in ng/ml. The limit of detection, defined as the average of the blank value plus three times its standard deviation, was 1 ng/ml. Table 1 shows the accuracy and precision of the recommended procedure. Eflects of foreign species
The effects of various cations and anions on the determination of 0.5 pug/ml selenium(W) were studied. The results are summarized in
995
Table 2. The tolerance limit was defined as the concentration of added ion causing less than 3% relative error. Most ions did not interfere with the determination even when they are present at concentrations 200 times as great as that of selenium(W). Positive interferences were observed from lanthanum(III), copper( palladium(II), vanadium(III), tellerium(IV) and bromide because they could also catalyse the reaction. Negative interferences were observed from iodate and iodide, because they inhibit the indicator reaction. H&+ interferes by precipitation with the reagents. Iron(II1) and cerium(IV) interfere by changing the colour of the solution probably by complexation with Methyl Orange. The interfering effect of copper(I1) in concentrations less than 2 pg/ml and palladium(I1) in concentrations less than 2 pg/ml can be removed by the addition of EDTA and dimethylglyoxime, respectively. However, higher concentrations of copper(I1) can be removed by extracting its complex with dimethylglyoxime into chloroform, and higher concentrations of palladium(I1) can be removed by extracting its complex with oxine into chloroform. The interfering effect of vanadium(II1) can also be removed by extracting its complex with oxine into chloroform. Determination product
of selenium
in a health-care
In order to test the described method, selenium was determined in a health-care product (a shampoo for the treatment of dandruff). Sample dissolution was carried out by the procedure described by Belarra et aLI Approximately 1 g of sample was weighed into a loo-ml Kjeldahl flask. Concentrated sulphuric acid (1 ml) was added and the mixture was heated to fuming for 15 min. The solution was allowed to cool and 5 ml of 30% w/v hydrogen peroxide was added. The mixture was boiled vigorously to eliminate excess of hydrogen peroxide and the flask was allowed to cool. The mixture was diluted to 1.0 litre with triply distilled water. A 0.5ml volume of this solution was taken for determination of selenium(W) as in the recommended procedure. The results showed the presence of 13.45 St 0.42 g/l. selenium (n = 7). (Manufacturer’s value is given as 13.81 g/l. selenium as selenium sulfide.)
996
A. ApKHMnet al. CONCLUSIONS
This method can be used to determine selenium as low as 1.0 ng/ml without the need for any preconcentration step. The method is very simnle and more selective than the other kinetic methods reported previously or hydride generation atomic-absorption spectrometry which is subject to many interferences. Also, the limit of detection obtained by the proposed method (1 .O ng/ml) is lower than that obtained by hydride generation atomic-absorption spectroscopy (5 ng/ml).3 Since the method only responds to selenium(IV) species, the determination of selenium(IV) in the presence of selenium(W) is possible. Acknowledgement-The authors wish to express their gratitude to Shiraz University Research Council for their support of this work.
REFERENCES H. RobbemhtandR.VanGrieken, Takmta, 1982,29,823. K. Toei and Y. Shimoishi, ibid., 1981, 2S, 867. A. E. Smith, Andpt, 1975, 100, 300. M. W. Brown and J. H. Watkinson, Anal. Chim. Acta, 1977, 89, 35. 5. Y. Shim&hi and K. Toei, ibid., 1978, 100,65. 6. N. D. Michie, E. J. Dixon and N. G. Bunton, J. Assoc. Anal. Chetn., 1978, 61,48. 7. K. B. Yatsimirskii, Kinetic Methodp of Analysis, Pergamon Press, Oxford, 1966. 8. P. W. West and T. V. Ramakrishna. Anal. Ckem., 1968, 48,966. 9. T. Kawashima and M. Tanaka, Anal. Chim. Acta, 1968, 48, 143. 10. W. C. Hawkes, ibid., 1986, 183, 197. 11. A. Safavi, A. Afkhami and A. Massoumi, ibid., 1990, 232, 351. 12. A. I. Vogel, A Text Book of Quantitative Inorganic Analysis, 3rd Ed., Longman, London, 1961. 13. P. Linares, M. D. Luque de Castro and M. Valcarcel, Analyst, 1986, 111, 1405. 14. M. A. Belarra, F. Gallarta, J. M. Anxano and J. R. Castillo, J. Anal. At. S’cr., 1986, 1, 141. l. 2. 3. 4.