- Email: [email protected]
Selective Determination of Chloride and Bromide Ions in Serum by Cyclic Voltammetry
ANALYTICAL BIOCHEMISTRY ARTICLE NO.
240, 109–113 (1996)
0336
Selective Determination of Chloride and Bromide Ions in Serum by Cyclic Voltammetry Kensuke Arai, Fumiyo Kusu, Naohito Noguchi, Kiyoko Takamura,1 and Hisao Osawa* School of Pharmacy, Tokyo University of Pharmacy and Life Science, 1432-1 Horinouchi, Hachioji, Tokyo 192-03, Japan; and *Nippon Filcon Co. Ltd. R&D, 2220 Ohmaru, Inagi, Tokyo 206, Japan
Received February 5, 1996
Voltammetric determination of chloride and bromide ions in serum was made using a one-body type Ag electrode, whose surface was covered with a dialysis membrane for preventing interference from proteins. This determination was based on measurement of the charge of the reduction wave of silver halide formed on the Ag electrode surface in a halide ion solution during a cathodic potential sweep. In this method, depending on the potential sweep range, the charge of the reduction wave of AgI, AgI / AgBr, or AgI / AgBr / AgCl can be measured. Here, the method was used to determine the chloride and bromide ions in artificial serum. Linear concentration ranges were 1.0 1 1002 – 2.4 1 1001 M for chloride ions and 5.0 1 1003 –1.8 1 1002 M for bromide ions. The correlation coefficient was 0.999 in each case. The relative standard deviation for chloride ion (1.0 1 1001 M) was 1.9% (n Å 10), and that for bromide ion (1.5 1 1002 M) was 2.5% (n Å 10). Bromide ions at a relatively high concentration as the component of serum had no effect on the determination of the chloride ion in the serum. This method is thus useful for the selective determination of chloride and bromide ions in serum. q 1996 Academic Press, Inc.
Variation in the chloride ion concentration in serum may serve as an index of renal diseases, adrenalism, and pneumonia and, thus, the accurate determination of this parameter is clinically important. Cathodic stripping voltammetry with a mercury (1, 2) or silver (3) electrode is sensitive to the halide ion. However, a mercury electrode is undesirable because of its toxicity. For using a silver electrode, the treatment of coexisting substances and separation of halide ions have not been 1 To whom correspondence should be addressed. Fax: /81-426-764570. E-mail: [email protected].
made. The coulometric titration method (4), the ionselective electrode (ISE)2 method (5), and Shales and Shales method (6) are presently used for determining chloride ions in serum. For routine clinical analysis, the ISE method is used most. However, the chloride ion concentration in serum becomes higher than it actually is when the serum contains bromide ions (7), as has also been pointed out in other studies (8–12). ISE response to bromide ion is 15 times more than to chloride ion (9) and, thus, even at a low concentration, bromide ions prevent an accurate determination of chloride ion. The administration of drugs containing bromide ions may cause a positive deviation in the determination of chloride. The Shales and Shales method and coulometric titration are more accurate than the ISE method for the determination of chloride ions, but erroneous data due to bromide ions are obtained. In the Shales and Shales method, bromide and iodide ions generate mercury(II) halides as do chloride ions, and in coulometric titration, bromide and iodide ions generate silver halides as do chloride ions. Thus, the selective determination of serum chloride has been a problem. The concentration of bromide in serum is an index of bromide intoxication (13) induced by inappropriate administration of bromide-containing hypnotics and sedatives. Thus, accurate bromide ion determination is clinically important. In this study, attention was directed to the separation of halide ions by voltammetry using an Ag electrode, since the reversible potentials of the silver halides differ. A selective and accurate determination method of chloride and bromide ions in serum was established, based on the measurement of cyclic voltammograms using Ag as the working electrode. The
2 Abbreviations used: ISE, ion-selective electrode; MWCO, molecular weight cutoff.
109
0003-2697/96 $18.00 Copyright q 1996 by Academic Press, Inc. All rights of reproduction in any form reserved.
AID
AB 9664
/
6m1b$$$181
07-30-96 06:42:47
abas
AP: Anal Bio
110
ARAI ET AL.
151) and xy recorder (Yokogawa Electric Corp. Type 3086 recorder) or an integrator (Jasco Corp. 807-IT). The measurement of the charge of each redox wave was made with the integrator within 1 min per cycle, and the effect of the residual current on this calculation was successfully avoided by choosing an appropriate potential range for integration. All measurements were made at 25 { 17C. The water in this study was distilled and purified by the NANO Pure II filtering system (Barnstead Co.). Sodium chloride of 99.99% purity (Manac Inc.), sodium bromide of 99.9% purity (Wako Pure Chemical Ind., Ltd.), sodium iodide of 99.9% purity (Wako Pure Chemical Ind., Ltd.), and all other chemicals were obtained commercially and used without further purification. RESULTS AND DISCUSSION
Cyclic Voltammogram of Chloride, Bromide, and Iodide Ions on the Ag Electrode
FIG. 1. (A) One-body type Ag electrode covered with the dialysis membrane and (B) electrolytic cell: (a) One-body type Ag electrode, (b) Ag disk (1 mm diameter) working electrode, (c) salt bridge, (d) saturated calomel electrode, (e) cylindrical stainless-steel tube (7 mm i.d. 1 25 mm length) counter electrode, (f) epoxy resin, (g) dialysis membrane, (h) O ring, (i) glass electrolytic cell (10 ml cell volume), (j) silicon rubber cap, (k) sample inlet, (l) electrolyte solution.
method was shown to be applicable to the separate determination of chloride and bromide ions.
When a halide ion such as a chloride, bromide, or iodide ion in a phosphate buffer solution reacts with the Ag electrode surface, a silver halide precipitate is formed on the surface to consequently produce an oxidation wave on the voltammogram. Reduction of the silver halide precipitate on the electrode surface leads to generation of a reduction current in accordance with the equations AgCl / e ` Ag / Cl0 EAg/AgCl Å 0.117 V vs SCE AgBr / e ` Ag / Br0
EXPERIMENTAL
Figure 1 shows the electrolytic cell for this study. The one-body type Ag electrode (a) consists of an Ag disk (1 mm diameter) working electrode (b), a salt bridge (c) connected to a saturated calomel reference electrode (d), and a cylindrical stainless-steel tube (7 mm i.d. 1 25 mm length) counter electrode (e). The Ag disk and salt bridge were inserted into the stainlesssteel tube and fixed in place by epoxy resin (f) (EW3LT, Sumitomo MMM). The electrode surface was covered with a dialysis membrane of cellulose ester (g) (Spectra/Por CE, Spectrum Medical Ind., molecular weight cutoff, MWCO 5000), to prevent contamination of the electrode surface by serum proteins. The membrane was fixed with O rings (h) on the edge of the electrode. The 10-ml volume glass electrolytic cell (j) was covered with a silicon rubber cap (j) with two holes as inlets for the electrode and sample (k). Cyclic voltammetry was carried out with a conventional potentiostat (Hokuto Denko Corporation HAB-
AB 9664
[2]
EAg/AgBr Å 00.011 V vs SCE
Apparatus
AID
[1]
/
6m1b$$$182
07-30-96 06:42:47
AgI / e ` Ag / I0
[3]
EAg/AgI Å 00.234 V vs SCE, where EAg/AgCl , EAg/AgBr , and EAg/AgI are reversible potentials for Eqs. [1], [2], and [3], respectively, when 5.0 1 1003 M sodium chloride, 2.0 1 1003 M sodium bromide, and 2.0 1 1003 M sodium iodide are present in the electrolyte solution. Figure 2 shows cyclic voltammograms obtained for chloride (a), bromide (b), and iodide (c) ion solutions and their mixtures (d–f) at pH 7. The potential of the Ag electrode was set at the initial potential, E1 , 00.60 V, swept linearly at a scan rate of 0.02 V/s in a positive direction, turned at the switching potential, E2 , and swept back to the final potential, E3 , of 00.60 V. The shape of the voltammogram for the solution mixture of chloride, bromide, and iodide ions thus depends on E2 . When E2 was 00.10 V, a pair of redox waves due to the redox reaction of Eq. [3] appeared on the voltammogram (Fig. 2d). When E2 was
abas
AP: Anal Bio
111
VOLTAMMETRIC DETERMINATION OF CHLORIDE AND BROMIDE IONS
(i) Cyclic voltammogram for pH 7 phosphate buffer containing halide ion was obtained with switching potential E2 of 0.10 V and total charge of the reduction waves of AgCl, AgBr, and AgI, denoted as Q(Cl, Br, I), was measured. (ii) Cyclic voltammogram was obtained with E2 of 0.02 V, and total charge of the reduction waves of AgBr and AgI, Q(Br, I), was measured. (iii) Using Q(Cl, Br, I) and Q(Br, I), charge of the reduction wave of AgCl, Q(Cl), was calculated as Q(Cl) Å Q(Cl, Br, I) 0 ArQ(Br, I),
FIG. 2. Cyclic voltammograms for solutions of (a) 5.0 1 1003 M chloride ions, (b) 2.0 1 1003 M bromide ions, and (c) 2.0 1 1003 M iodide ions, and (d–f) a solution mixture of 5.0 1 1003 M chloride ions, 2.0 1 1003 M bromide ions, and 2.0 1 1003 M iodide ions. Switching potential: (a) 0.25, (b) 0.15, (c) 00.10, (d) 00.10, (e) 0.00, (f) 0.15 V. Scan rate: 0.02 V/s.
0.00 and 0.15 V, the voltammograms in Figs. 2e and 2f were obtained, respectively. The second and third oxidation waves correspond to the formation of AgBr (Eq. [2]) and AgCl (Eq. [1]), respectively. The total charge of the reduction waves in Fig. 2e corresponds to the reduction of AgBr and AgI, and that in Fig. 2f, to AgCl, AgBr and AgI. When the sweep rate exceeded 0.05 V/s, the oxidation waves of chloride, bromide, and iodide ions ceased to separate. Voltammograms of chloride ion with the Ag electrode were obtained at various pH. At high pH, such as pH 13, the reduction wave of Ag2O, generated on the Ag surface during the anodic sweep, became superimposed on that of AgCl, but not at pH less than 8. Phosphate buffer of pH 7 was thus selected as the supporting electrolyte for halide ion determination in serum. Proteins in serum react with an Ag electrode surface. Thus in this study, the electrode surface was covered with a dialysis membrane. In consideration of the molecular size of the proteins and diffusion of halide ions to the Ag surface through the membrane, a dialysis membrane with MWCO 5000 was selected. The shape of the cyclic voltammogram was affected somewhat by the membrane, due to lowering of the diffusion rate of the halide ions.
where A is a constant. (iv) Cyclic voltammogram was obtained with E2 of 00.10 V, and charge of the reduction wave of AgI, Q(I), was measured. (v) Charge of the reduction wave of AgBr, Q(Br), was calculated by Eq. [5]. Q(Br) Å Q(Br, I) 0 BrQ(I),
Effects of Foreign Substances The dialysis membrane covering the electrode surface prevents its contamination by proteins such as
TABLE 1
Effects of Coexisting Substances on Chloride Ion Determination Concentration, Ascorbic acid Uric acid Cystine
AID
AB 9664
/
6m1b$$$182
07-30-96 06:42:47
[5]
where B is a constant. Cyclic voltammograms were obtained to estimate constants A and B using standard solution mixtures of halide ions at various concentrations. Thus, A was determined as 1.6, and B was 1.1. Good linear relationship between reduction charge and concentration of each halide ion was observed at 2.0 1 1003 –4.8 1 1002 M for chloride ions, 1.0 1 1003 –3.5 1 1003 M for bromide ions, and 8.0 1 1005 –1.0 1 1003 M for iodide ions. The correlation coefficient was 0.999 in each case. Relative standard deviation for chloride ion (1.0 1 1001 M) was 1.9% (n Å 10), and that for bromide ion (1.5 1 1002 M) was 2.5% (n Å 10).
Calibration Curves for Halide Ion Determination From charge calculated using the reduction waves of silver halide, all chloride, bromide, and iodide ions in the solution mixture could be determined as follows:
[4]
Methionine
a
2.4 1.8 1.0 3.0 2.0 6.0 8.0 3.5
1 1 1 1 1 1 1 1
M
Chloride ion found, %a
103 1002 1004 1004 1005 1005 1005 1004
Concentration of added chloride ions: 1.0 1 1001
abas
AP: Anal Bio
100.7 101.2 100.9 98.8 99.4 101.6 100.5 101.5 M.
112
ARAI ET AL. TABLE 2
Accuracy and Reproducibility of Chloride and Bromide Ion Determination in Artificial Serum Chloride ion
Bromide ion
Sample
Mean { SD, mMb
RSD, %b
Mean { SD, mMb
RSD, %b
Seruma Serum spiked with 12.0 mM bromide ion
108.0 { 1.3 109.0 { 1.9
1.4 1.7
—c 12.4 { 0.2
— 2.4
a b c
Chloride ion concentration in artificial serum: 108.0 mM. n Å 10. Bromide ion concentration in artificial serum: less than minimal detection limit, 5.0 mM.
albumin. However, ascorbic acid, uric acid, cystine, and methionine may permeate the membrane, with possibly consequent inaccurate determination of chloride. Ascorbic acid and uric acid are oxidized even at a potential more negative than 0.3 V, and cystine and methionine react with the Ag electrode in the same way as halide ions (14). The effects of coexisting substances on the determination of chloride ions are shown in Table 1. The 1.8 1 1002 M ascorbic acid and 1.8 1 1004 M methionine concentrations, which were 100 times and 10 times greater than those in serum, respectively, had no effect on the determination of chloride ions. Cystine and uric acid at serum concentrations interfered with this determination but this could have been prevented by determining chloride ions using a two-times diluted serum sample. Thus in subsequent experiments, serum was diluted five times with supporting electrolyte solution before measuring voltammograms for chloride and bromide ion determination. Selective Determination of Chloride and Bromide Ions The chloride ion concentration in the serum of healthy subjects is 9.5 1 1002 –1.1 1 1001 M. This range may vary from 5.0 1 1002 to 1.4 1 1001 M (15) according to the disease. The bromide ion concentration in the serum of healthy subjects is usually less than 1.0 1 1004 M, but in the case of bromism it may be as much
as 1.2 1 1002 M (9). Artificial serum (Control Serum I Wako, Wako Pure Chemical Ind., Ltd.) and that spiked with 1.2 1 1002 M bromide ion were used as the samples in this study. Mean values and relative standard deviation obtained by the present method are shown in Table 2. The bromide ion concentration in serum was less than the minimal detection limit, 5.0 1 1003 M, and thus could not be detected. Although bromide ions coexisted in the serum solution, selective determination of chloride and bromide ions was successfully made with high accuracy and reproducibility. Table 3 shows the values for chloride ions obtained by the Shales and Shales method and the method proposed in this study. The method by Shales and Shales for the serum containing bromide ions showed a relatively large positive error, while the present method gave a value in close agreement with that for artificial serum. By this method, bromide ion at a relatively high concentration as the component of serum had no effect on chloride ion determination in the serum, and the determination of halide ions in serum with high selectivity is thus possible. This method should thus prove useful for accurately determining chloride ions in serum containing bromide ions even at a high concentration due to bromism. The selective determination of halide ions in serum should facilitate clinical diagnosis and also be useful for diagnosing bromide and iodide intoxication. REFERENCES
TABLE 3
Determinations of Chloride Concentrations by the Shales and Shales Method and the Present Method Chloride ion concentration, mMa
1. Boll, R. G., Manning, D. L., and Menis, O. (1960) Anal. Chem. 32, 621. 2. Colovov, G., Wilson, G. S., and Moyers, J. L. (1974) Anal. Chem. 46, 1051. 3. Shain, I., and Perone, S. P. (1961) Anal. Chem. 33, 325.
Shales and Shales method
Sample Serum Serum with 12.0 mM bromide ion a
Present method
107.1 118.1
106.6 106.7
Chloride ion concentration in artificial serum: 106.0 mM.
AID
AB 9664
/
6m1b$$$182
07-30-96 06:42:47
4. Colt, E., and Nishi, H. H. (1961) Clin. Chem. 7, 285–291. 5. Koryta, J. (1990) Anal. Chim. Acta 233, 1–30. 6. Shales, O., and Shales, S. S. (1941) J. Biol. Chem. 140, 879– 884. 7. Wenk, R. E., Lustgarten, J. A., Pappas, N. J., Levy, R. I., and Lackson, R. (1976) Am. J. Clin. Pathol. 65, 49–57.
abas
AP: Anal Bio
VOLTAMMETRIC DETERMINATION OF CHLORIDE AND BROMIDE IONS 8. Elin, R. J., Robertson, E. A., and Johnson, E. (1981) Clin. Chem. 27, 778–779. 9. Kan, K., Satowa, S., Takeuchi, I., Saruta, T., and Kano, S. (1986) Int J. Clin. Pharmacol. Ther. Toxicol. 24, 399–402. 10. Pressac, M., Jardel, C., Durand, D., and Aymard, P. (1987) Clin. Chem. 33, 415–416. 11. Nagamine, Y., Hamai, Y., Chikamori, K., Kita, T., Hirota, M., Oshima, I., Yamashita, S., and Shima, K. (1988) Scand. J. Clin. Lab. Invest. 48, 177–182.
AID
AB 9664
/
6m1b$$$183
07-30-96 06:42:47
113
12. Woody, R. C., Turley, C. P., and Brewster, M. A. (1990) Ther. Drug. Monit. 12, 490–492. 13. Trump, D. L., and Hochberg, M. C. (1976) Johns Hopkins Med. J. 138, 119–123. 14. Dryhurst, G., Kadish, K. M., Scheller, F., and Renneberg, R. (1982) Biological Electrochemistry, Vol. 1, Academic Press, New York. 15. Searcy, R. L. (1969) Diagnostic Biochemistry, McGraw–Hill, New York.
abas
AP: Anal Bio
240, 109–113 (1996)
0336
Selective Determination of Chloride and Bromide Ions in Serum by Cyclic Voltammetry Kensuke Arai, Fumiyo Kusu, Naohito Noguchi, Kiyoko Takamura,1 and Hisao Osawa* School of Pharmacy, Tokyo University of Pharmacy and Life Science, 1432-1 Horinouchi, Hachioji, Tokyo 192-03, Japan; and *Nippon Filcon Co. Ltd. R&D, 2220 Ohmaru, Inagi, Tokyo 206, Japan
Received February 5, 1996
Voltammetric determination of chloride and bromide ions in serum was made using a one-body type Ag electrode, whose surface was covered with a dialysis membrane for preventing interference from proteins. This determination was based on measurement of the charge of the reduction wave of silver halide formed on the Ag electrode surface in a halide ion solution during a cathodic potential sweep. In this method, depending on the potential sweep range, the charge of the reduction wave of AgI, AgI / AgBr, or AgI / AgBr / AgCl can be measured. Here, the method was used to determine the chloride and bromide ions in artificial serum. Linear concentration ranges were 1.0 1 1002 – 2.4 1 1001 M for chloride ions and 5.0 1 1003 –1.8 1 1002 M for bromide ions. The correlation coefficient was 0.999 in each case. The relative standard deviation for chloride ion (1.0 1 1001 M) was 1.9% (n Å 10), and that for bromide ion (1.5 1 1002 M) was 2.5% (n Å 10). Bromide ions at a relatively high concentration as the component of serum had no effect on the determination of the chloride ion in the serum. This method is thus useful for the selective determination of chloride and bromide ions in serum. q 1996 Academic Press, Inc.
Variation in the chloride ion concentration in serum may serve as an index of renal diseases, adrenalism, and pneumonia and, thus, the accurate determination of this parameter is clinically important. Cathodic stripping voltammetry with a mercury (1, 2) or silver (3) electrode is sensitive to the halide ion. However, a mercury electrode is undesirable because of its toxicity. For using a silver electrode, the treatment of coexisting substances and separation of halide ions have not been 1 To whom correspondence should be addressed. Fax: /81-426-764570. E-mail: [email protected].
made. The coulometric titration method (4), the ionselective electrode (ISE)2 method (5), and Shales and Shales method (6) are presently used for determining chloride ions in serum. For routine clinical analysis, the ISE method is used most. However, the chloride ion concentration in serum becomes higher than it actually is when the serum contains bromide ions (7), as has also been pointed out in other studies (8–12). ISE response to bromide ion is 15 times more than to chloride ion (9) and, thus, even at a low concentration, bromide ions prevent an accurate determination of chloride ion. The administration of drugs containing bromide ions may cause a positive deviation in the determination of chloride. The Shales and Shales method and coulometric titration are more accurate than the ISE method for the determination of chloride ions, but erroneous data due to bromide ions are obtained. In the Shales and Shales method, bromide and iodide ions generate mercury(II) halides as do chloride ions, and in coulometric titration, bromide and iodide ions generate silver halides as do chloride ions. Thus, the selective determination of serum chloride has been a problem. The concentration of bromide in serum is an index of bromide intoxication (13) induced by inappropriate administration of bromide-containing hypnotics and sedatives. Thus, accurate bromide ion determination is clinically important. In this study, attention was directed to the separation of halide ions by voltammetry using an Ag electrode, since the reversible potentials of the silver halides differ. A selective and accurate determination method of chloride and bromide ions in serum was established, based on the measurement of cyclic voltammograms using Ag as the working electrode. The
2 Abbreviations used: ISE, ion-selective electrode; MWCO, molecular weight cutoff.
109
0003-2697/96 $18.00 Copyright q 1996 by Academic Press, Inc. All rights of reproduction in any form reserved.
AID
AB 9664
/
6m1b$$$181
07-30-96 06:42:47
abas
AP: Anal Bio
110
ARAI ET AL.
151) and xy recorder (Yokogawa Electric Corp. Type 3086 recorder) or an integrator (Jasco Corp. 807-IT). The measurement of the charge of each redox wave was made with the integrator within 1 min per cycle, and the effect of the residual current on this calculation was successfully avoided by choosing an appropriate potential range for integration. All measurements were made at 25 { 17C. The water in this study was distilled and purified by the NANO Pure II filtering system (Barnstead Co.). Sodium chloride of 99.99% purity (Manac Inc.), sodium bromide of 99.9% purity (Wako Pure Chemical Ind., Ltd.), sodium iodide of 99.9% purity (Wako Pure Chemical Ind., Ltd.), and all other chemicals were obtained commercially and used without further purification. RESULTS AND DISCUSSION
Cyclic Voltammogram of Chloride, Bromide, and Iodide Ions on the Ag Electrode
FIG. 1. (A) One-body type Ag electrode covered with the dialysis membrane and (B) electrolytic cell: (a) One-body type Ag electrode, (b) Ag disk (1 mm diameter) working electrode, (c) salt bridge, (d) saturated calomel electrode, (e) cylindrical stainless-steel tube (7 mm i.d. 1 25 mm length) counter electrode, (f) epoxy resin, (g) dialysis membrane, (h) O ring, (i) glass electrolytic cell (10 ml cell volume), (j) silicon rubber cap, (k) sample inlet, (l) electrolyte solution.
method was shown to be applicable to the separate determination of chloride and bromide ions.
When a halide ion such as a chloride, bromide, or iodide ion in a phosphate buffer solution reacts with the Ag electrode surface, a silver halide precipitate is formed on the surface to consequently produce an oxidation wave on the voltammogram. Reduction of the silver halide precipitate on the electrode surface leads to generation of a reduction current in accordance with the equations AgCl / e ` Ag / Cl0 EAg/AgCl Å 0.117 V vs SCE AgBr / e ` Ag / Br0
EXPERIMENTAL
Figure 1 shows the electrolytic cell for this study. The one-body type Ag electrode (a) consists of an Ag disk (1 mm diameter) working electrode (b), a salt bridge (c) connected to a saturated calomel reference electrode (d), and a cylindrical stainless-steel tube (7 mm i.d. 1 25 mm length) counter electrode (e). The Ag disk and salt bridge were inserted into the stainlesssteel tube and fixed in place by epoxy resin (f) (EW3LT, Sumitomo MMM). The electrode surface was covered with a dialysis membrane of cellulose ester (g) (Spectra/Por CE, Spectrum Medical Ind., molecular weight cutoff, MWCO 5000), to prevent contamination of the electrode surface by serum proteins. The membrane was fixed with O rings (h) on the edge of the electrode. The 10-ml volume glass electrolytic cell (j) was covered with a silicon rubber cap (j) with two holes as inlets for the electrode and sample (k). Cyclic voltammetry was carried out with a conventional potentiostat (Hokuto Denko Corporation HAB-
AB 9664
[2]
EAg/AgBr Å 00.011 V vs SCE
Apparatus
AID
[1]
/
6m1b$$$182
07-30-96 06:42:47
AgI / e ` Ag / I0
[3]
EAg/AgI Å 00.234 V vs SCE, where EAg/AgCl , EAg/AgBr , and EAg/AgI are reversible potentials for Eqs. [1], [2], and [3], respectively, when 5.0 1 1003 M sodium chloride, 2.0 1 1003 M sodium bromide, and 2.0 1 1003 M sodium iodide are present in the electrolyte solution. Figure 2 shows cyclic voltammograms obtained for chloride (a), bromide (b), and iodide (c) ion solutions and their mixtures (d–f) at pH 7. The potential of the Ag electrode was set at the initial potential, E1 , 00.60 V, swept linearly at a scan rate of 0.02 V/s in a positive direction, turned at the switching potential, E2 , and swept back to the final potential, E3 , of 00.60 V. The shape of the voltammogram for the solution mixture of chloride, bromide, and iodide ions thus depends on E2 . When E2 was 00.10 V, a pair of redox waves due to the redox reaction of Eq. [3] appeared on the voltammogram (Fig. 2d). When E2 was
abas
AP: Anal Bio
111
VOLTAMMETRIC DETERMINATION OF CHLORIDE AND BROMIDE IONS
(i) Cyclic voltammogram for pH 7 phosphate buffer containing halide ion was obtained with switching potential E2 of 0.10 V and total charge of the reduction waves of AgCl, AgBr, and AgI, denoted as Q(Cl, Br, I), was measured. (ii) Cyclic voltammogram was obtained with E2 of 0.02 V, and total charge of the reduction waves of AgBr and AgI, Q(Br, I), was measured. (iii) Using Q(Cl, Br, I) and Q(Br, I), charge of the reduction wave of AgCl, Q(Cl), was calculated as Q(Cl) Å Q(Cl, Br, I) 0 ArQ(Br, I),
FIG. 2. Cyclic voltammograms for solutions of (a) 5.0 1 1003 M chloride ions, (b) 2.0 1 1003 M bromide ions, and (c) 2.0 1 1003 M iodide ions, and (d–f) a solution mixture of 5.0 1 1003 M chloride ions, 2.0 1 1003 M bromide ions, and 2.0 1 1003 M iodide ions. Switching potential: (a) 0.25, (b) 0.15, (c) 00.10, (d) 00.10, (e) 0.00, (f) 0.15 V. Scan rate: 0.02 V/s.
0.00 and 0.15 V, the voltammograms in Figs. 2e and 2f were obtained, respectively. The second and third oxidation waves correspond to the formation of AgBr (Eq. [2]) and AgCl (Eq. [1]), respectively. The total charge of the reduction waves in Fig. 2e corresponds to the reduction of AgBr and AgI, and that in Fig. 2f, to AgCl, AgBr and AgI. When the sweep rate exceeded 0.05 V/s, the oxidation waves of chloride, bromide, and iodide ions ceased to separate. Voltammograms of chloride ion with the Ag electrode were obtained at various pH. At high pH, such as pH 13, the reduction wave of Ag2O, generated on the Ag surface during the anodic sweep, became superimposed on that of AgCl, but not at pH less than 8. Phosphate buffer of pH 7 was thus selected as the supporting electrolyte for halide ion determination in serum. Proteins in serum react with an Ag electrode surface. Thus in this study, the electrode surface was covered with a dialysis membrane. In consideration of the molecular size of the proteins and diffusion of halide ions to the Ag surface through the membrane, a dialysis membrane with MWCO 5000 was selected. The shape of the cyclic voltammogram was affected somewhat by the membrane, due to lowering of the diffusion rate of the halide ions.
where A is a constant. (iv) Cyclic voltammogram was obtained with E2 of 00.10 V, and charge of the reduction wave of AgI, Q(I), was measured. (v) Charge of the reduction wave of AgBr, Q(Br), was calculated by Eq. [5]. Q(Br) Å Q(Br, I) 0 BrQ(I),
Effects of Foreign Substances The dialysis membrane covering the electrode surface prevents its contamination by proteins such as
TABLE 1
Effects of Coexisting Substances on Chloride Ion Determination Concentration, Ascorbic acid Uric acid Cystine
AID
AB 9664
/
6m1b$$$182
07-30-96 06:42:47
[5]
where B is a constant. Cyclic voltammograms were obtained to estimate constants A and B using standard solution mixtures of halide ions at various concentrations. Thus, A was determined as 1.6, and B was 1.1. Good linear relationship between reduction charge and concentration of each halide ion was observed at 2.0 1 1003 –4.8 1 1002 M for chloride ions, 1.0 1 1003 –3.5 1 1003 M for bromide ions, and 8.0 1 1005 –1.0 1 1003 M for iodide ions. The correlation coefficient was 0.999 in each case. Relative standard deviation for chloride ion (1.0 1 1001 M) was 1.9% (n Å 10), and that for bromide ion (1.5 1 1002 M) was 2.5% (n Å 10).
Calibration Curves for Halide Ion Determination From charge calculated using the reduction waves of silver halide, all chloride, bromide, and iodide ions in the solution mixture could be determined as follows:
[4]
Methionine
a
2.4 1.8 1.0 3.0 2.0 6.0 8.0 3.5
1 1 1 1 1 1 1 1
M
Chloride ion found, %a
103 1002 1004 1004 1005 1005 1005 1004
Concentration of added chloride ions: 1.0 1 1001
abas
AP: Anal Bio
100.7 101.2 100.9 98.8 99.4 101.6 100.5 101.5 M.
112
ARAI ET AL. TABLE 2
Accuracy and Reproducibility of Chloride and Bromide Ion Determination in Artificial Serum Chloride ion
Bromide ion
Sample
Mean { SD, mMb
RSD, %b
Mean { SD, mMb
RSD, %b
Seruma Serum spiked with 12.0 mM bromide ion
108.0 { 1.3 109.0 { 1.9
1.4 1.7
—c 12.4 { 0.2
— 2.4
a b c
Chloride ion concentration in artificial serum: 108.0 mM. n Å 10. Bromide ion concentration in artificial serum: less than minimal detection limit, 5.0 mM.
albumin. However, ascorbic acid, uric acid, cystine, and methionine may permeate the membrane, with possibly consequent inaccurate determination of chloride. Ascorbic acid and uric acid are oxidized even at a potential more negative than 0.3 V, and cystine and methionine react with the Ag electrode in the same way as halide ions (14). The effects of coexisting substances on the determination of chloride ions are shown in Table 1. The 1.8 1 1002 M ascorbic acid and 1.8 1 1004 M methionine concentrations, which were 100 times and 10 times greater than those in serum, respectively, had no effect on the determination of chloride ions. Cystine and uric acid at serum concentrations interfered with this determination but this could have been prevented by determining chloride ions using a two-times diluted serum sample. Thus in subsequent experiments, serum was diluted five times with supporting electrolyte solution before measuring voltammograms for chloride and bromide ion determination. Selective Determination of Chloride and Bromide Ions The chloride ion concentration in the serum of healthy subjects is 9.5 1 1002 –1.1 1 1001 M. This range may vary from 5.0 1 1002 to 1.4 1 1001 M (15) according to the disease. The bromide ion concentration in the serum of healthy subjects is usually less than 1.0 1 1004 M, but in the case of bromism it may be as much
as 1.2 1 1002 M (9). Artificial serum (Control Serum I Wako, Wako Pure Chemical Ind., Ltd.) and that spiked with 1.2 1 1002 M bromide ion were used as the samples in this study. Mean values and relative standard deviation obtained by the present method are shown in Table 2. The bromide ion concentration in serum was less than the minimal detection limit, 5.0 1 1003 M, and thus could not be detected. Although bromide ions coexisted in the serum solution, selective determination of chloride and bromide ions was successfully made with high accuracy and reproducibility. Table 3 shows the values for chloride ions obtained by the Shales and Shales method and the method proposed in this study. The method by Shales and Shales for the serum containing bromide ions showed a relatively large positive error, while the present method gave a value in close agreement with that for artificial serum. By this method, bromide ion at a relatively high concentration as the component of serum had no effect on chloride ion determination in the serum, and the determination of halide ions in serum with high selectivity is thus possible. This method should thus prove useful for accurately determining chloride ions in serum containing bromide ions even at a high concentration due to bromism. The selective determination of halide ions in serum should facilitate clinical diagnosis and also be useful for diagnosing bromide and iodide intoxication. REFERENCES
TABLE 3
Determinations of Chloride Concentrations by the Shales and Shales Method and the Present Method Chloride ion concentration, mMa
1. Boll, R. G., Manning, D. L., and Menis, O. (1960) Anal. Chem. 32, 621. 2. Colovov, G., Wilson, G. S., and Moyers, J. L. (1974) Anal. Chem. 46, 1051. 3. Shain, I., and Perone, S. P. (1961) Anal. Chem. 33, 325.
Shales and Shales method
Sample Serum Serum with 12.0 mM bromide ion a
Present method
107.1 118.1
106.6 106.7
Chloride ion concentration in artificial serum: 106.0 mM.
AID
AB 9664
/
6m1b$$$182
07-30-96 06:42:47
4. Colt, E., and Nishi, H. H. (1961) Clin. Chem. 7, 285–291. 5. Koryta, J. (1990) Anal. Chim. Acta 233, 1–30. 6. Shales, O., and Shales, S. S. (1941) J. Biol. Chem. 140, 879– 884. 7. Wenk, R. E., Lustgarten, J. A., Pappas, N. J., Levy, R. I., and Lackson, R. (1976) Am. J. Clin. Pathol. 65, 49–57.
abas
AP: Anal Bio
VOLTAMMETRIC DETERMINATION OF CHLORIDE AND BROMIDE IONS 8. Elin, R. J., Robertson, E. A., and Johnson, E. (1981) Clin. Chem. 27, 778–779. 9. Kan, K., Satowa, S., Takeuchi, I., Saruta, T., and Kano, S. (1986) Int J. Clin. Pharmacol. Ther. Toxicol. 24, 399–402. 10. Pressac, M., Jardel, C., Durand, D., and Aymard, P. (1987) Clin. Chem. 33, 415–416. 11. Nagamine, Y., Hamai, Y., Chikamori, K., Kita, T., Hirota, M., Oshima, I., Yamashita, S., and Shima, K. (1988) Scand. J. Clin. Lab. Invest. 48, 177–182.
AID
AB 9664
/
6m1b$$$183
07-30-96 06:42:47
113
12. Woody, R. C., Turley, C. P., and Brewster, M. A. (1990) Ther. Drug. Monit. 12, 490–492. 13. Trump, D. L., and Hochberg, M. C. (1976) Johns Hopkins Med. J. 138, 119–123. 14. Dryhurst, G., Kadish, K. M., Scheller, F., and Renneberg, R. (1982) Biological Electrochemistry, Vol. 1, Academic Press, New York. 15. Searcy, R. L. (1969) Diagnostic Biochemistry, McGraw–Hill, New York.
abas
AP: Anal Bio