- Email: [email protected]
Use of ethylene-vinyl-acetate as a new membrane matrix for calcium ion-selective electrode preparation
Talaata,Vol. 38. No. 8, pp. 929-935, 1991 Printed in Great Britain. All rights reserved
0039-9 MO/9 I s3.00 + 0.00 Copyright 0 1991 Fkrgamon Press plc
USE OF ETHYLENE-VINYL-ACETATE AS A NEW MEMBRANE MATRIX FOR CALCIUM ION-SELECTIVE ELECTRODE PREPARATION A. EL-JAMMAL,A. A. BOUKLOUZEand G. J. PA~UARCHE Institut de Pharmacie, CP 205/6, Universite Libre de Bruxelles, Boulevard du Triomphe, 1050 Bruxelles, Belgium
G. D. CHRISTIAN* Department of Chemistry, BG-10, University of Washington, Seattle, WA 98195, USA (Received 28 November 1990. Revised 12 February 1991. Accepted 20 February 1991)
Summary-A new polymer matrix based membrane electrode with an ion-exchanger responding to calcium was constructed by dissolving the copolymer ethylene-vinyl-acetate together with the ionexchanger in chloroform in the presence of a mixture of dioctylphthalate-nitrobcnzene as plasticizer. The ion-exchanger used as the electroactive component was calcium didecyl phosphate in di-(n-octylphenyl) phosphonate (Orion). This electrode exhibited near-Nemstian response over the concentration range lo-‘4 x 10m6M calcium. The pH did not affect the electrode performance within the range 8-l 1. Response time varied from 15 to 120 se-cand the lifetime exceeded six months. The membrane is subject to static charge buildup, but this is avoided by controlling the level of dryness of the membrane. Selectivity coefficients determined for both monovalent and divalent cations showed negligible interference by most of these ions. The electrode was applied successfully to the determination of calcium in commercial mineral waters.
The use of ion-selective membrane electrodes is finding increasing application in pharmaceutical analysis. ‘-3The development of sensors that are easy to use and maintain, fast-responding, with low cost materials have boosted the use of potentiometric methods. Traditional methods for analysis utilizing spectrophotometry,4 polarchromatography6 and enzymatic ography,’ measurements’ that may be characterized in some cases by low detection limits and high selectivity may possess limitations such as toxicity of reagents, high cost and complexity of the apparatus or the procedure. Ion-selective polymer-membrane electrodes based on ion-pair or chelating ligand complexes entrapped in the polymer, suffer from problems such as interferences and leaching of the the active component and the plasticizer&” from the polymer matrix. The leaching, which causes the deterioration of the response of the electrodes and leads to limited electrode lifetime, depends primarily on the nature (the lipophilicity) of the softener and the membrane active component. I1 In order to minimize the problems of interference, the use of alternatives *Author for correspondence. TAL 3816-l
to poly(viny1 chloride) (PVC) as the polymer matrix have been suggested.’ However, depending on the polymer used, the electrodes may exhibit poorer performance. Hence, the use of VAGH copolymer (hydrolyzed vinyl chloride/ vinyl acetate) for calcium electrodes offers no advantages,’ while poly(viny1 isobutyl ether) exhibits the same behavior as a PVC electrode containing the same sensing ingredients.r2 In this paper, we describe the use of ethylenevinyl-acetate copolymer (EjVAC) as the matrix membrane for construction of ion-selective polymer-membrane electrodes. The use of this copolymer in an ISE has not previously been reported. The structure is as follows: -(-CH,--CH,-),--(-CH,-CH-),--
E/VAC
L L -0 L H3 A new method for the assembly of this electrode was developed. This new membrane has been characterized for use in construction of a calcium ion-selective electrode. The proposed electrode exhibits high selectivity for cakium 929
A. EL-JAMMAL et al.
930
over many cations. The main advantages and disadvantages of the electrode are compared with the ISE, previously developed for the determination of calcium. EXPERIMENTAL
Reagents
All reagents were of analytical-reagent grade and were used without further purification. E/VAC, 40% weight of the vinyl acetate monomer in the copolymer, (LEVAPREN 400, m < n) Bayer was kindly supplied by Professor G. Geuskens (ULB). Sources of all other reagents were as follows: tetrahydrofuran, chloroform and dioctyl phthalate from Janssen and nitrobenzene from Merck. All solutions were prepared in Milli-Q water. Stock solutions, lo- ‘M, were stored at room temperature and standard solutions were prepared from stock solutions by sequential dilution with Milli-Q water.
Fig. 1. Schematic diagram of the electrode: (1) coaxial cable; (2) glass tube, (3) cable junction; (4) paratilm; (5) micropipette tip; (6) glass tube; (7) lo-‘MCaCl,; (8) Ag/AgCl; (9) E/VAC membrane.
Preparation of the electrode
A commercially available solution of Orion 92-20-02 calcium didecyl phosphate in di(n-octylphenyl) phosphonate was used as the electroactive material for the calcium ISE. Membranes were prepared by dissolving the electroactive material (100 ~1) and about 350 mg of E/VAC in 4 ml of an appropriate solvent (chloroform) at about 60” in a water bath. To this solution is added 1 ml of both dioctyl phthalate and nitrobenzene as plasticizer and the mixture was exposed to ultrasound to insure good dispersion of electroactive material in the matrix. Ag/AgCl
lo-‘M CaCl,
Membrane E/VAC
Potential measurements
Potentiometric measurements were carried out at room temperature with a Tacussel Ion processor II. A Tacussel 6000 pH-meter with a Tacussel glass electrode was used to measure pH. A saturated calomel electrode was employed as the reference electrode in both measurements. Electrode response was examined by immersing the membrane electrode in 10 ml of buffered solution together with the double junction saturated calomel reference electrode. The cell used to investigate the characteristics of the electrode may be represented as follows:
/ Buffer 1 KC1 / Hg2C12/Hg Sample solution Ca2+ (xM), pH (y) bridge saturated
A simplified scheme of the electrode is shown in Fig. 1. A micropipette tip was used as the electrode body. The end of the tip was dipped in the polymer mixture for a few seconds and then it was left in air for three hours. Longer drying times should not be used, in order to minimize static electricity build up. Calcium chloride solution, lo-‘M, was used as the internal reference solution, and the internal reference electrode was an Ag-AgCl electrode. The electrode was activated before use by soaking it overnight in a lo-‘M solution of calcium chloride and it was stored in the same solution when not in use.
The calibration curves were generated by addition of a O.l-ml aliquot of 10-6-10-‘M aqueous primary ion solution to 10 ml of the stirred solution. The potential was recorded when the reading became stable (*O. 1 mV/ 10 set) after each addition. The data were plotted as observed potential vs. the logarithm of the concentration of the primary ion. Concentration of calcium in commercial mineral water were then determined in a similar manner by means of a calibration graph, and the values obtained were checked by standard addition.
931
Calcium ion-selective electrode preparation RESULTS AND DISCUSSION
Membrane materials It has been suggested that polymers with low glass transition temperatures (Tg) are suitable matrices for polymer membrane electrodes.13 Our results agree with this hypothesis since polystyrene (Tg = 100”) tested in our laboratory resulted in electrodes with no response, while good results have been found for the measurement of both organic (methixene, promethazine and some local anesthetics) and inorganic (calcium) cations by using E/VAC matrix whose glass transition temperature is below room temperature (Tg = - 50”). This study reports the results for the measurement of calcium ions. An advantage of E/VAC is the wide range of solvents that can be used to dissolve it, which allowed the use of alternatives to tetrahydrofuran and cyclohexanone commonly used with PVC. Six solvents are able to dissolve E/VAC: carbon tetrachloride, chloroform, dichloromethane, p-xylene, tetrahydrofuran and toluene. From a response-stability point of view, the use of E/VAC instead of PVC causes an excessive susceptibility of the electrode to static electricity if allowed to become too hard, but this problem can be avoided by controlling the level of dryness of the membrane. Excellent potential stability could be achieved by limiting the dryness to an extent that provides for minimum mechanical strength of the membrane in the ISE, accomplished by allowing to air dry for three hours only. The technique described for the assembly of the electrode, in which the sensor is inserted into the end of a micropipette tip, demonstrates the use of a simple and inexpensive material for the body of the electrode and insures good protection of the sensor against solution turbulence.
I i’ \ \\\
Izomv
x + .
\
x
x\*, \’. +\+_:-+ +
. \ l\ \
.\ L l,,
\
I
I
I
I 3
l-0,.
‘7 .-m-m
I 5
I 7
pCa’+ Fig. 2. Dependence of the Ca*+ E/VAC membrane electrode response on the calcium ion-exchanger content per 6 ml of polymer mixture. (0) 20 ~1 (a) SO~1 (0) 100 PI(@) 200 ~1 (+ ) 500 pl ( x ) 1 ml. pH 9.4, Beckman Buffer.
tration decade and a correlation coefficient of 0.9996. The range of linear response for the first three electrodes (Orion sensor reagent > 0.1 ml) can be extended to lo-‘M by a much longer period of activation or after several uses of the electrode. An electrode prepared under the condition of 100 pl of sensor reagent without any other softener gives no response at all. Therefore, we selected the mixture dioctylphthalate nitrobenzene as plasticizer for our calcium ISE. The performance of the electrode with time (Fig. 3)
Response characteristics of the electrode Several parameters were investigated for optimal performance of the electrode. Typical calibration curves obtained for electrodes containing different amounts of calcium ionexchanger, after one night of conditioning in O.lM calcium chloride are shown in Fig. 2. The optimal tested amount of the Orion sensor reagent appears to be 0.1 ml per 6 ml of polymer mixture, for maximum potential response to calcium ion. The electrode demonstrates, under these conditions, a linear response in the lo-‘lo-‘M range, with a useful range extended to 5 x 10e6M, a slope of about 26 mV per concen-
pCa’+ Fig. 3. Calibration plots for CaZ+ E/VAC membrane electrodes in buffered solution, pH 9.4. (0) electrode one week old. (0) electrode six months old.
A. EL-JAMMAL et
932
al.
Table 1. Characteristics of polymer membrane Ca *+ electrodes based on Orion 92-20-02 sensor Electrode type
Lower limit of response (M) Detection limit Slope (mv/decade) pH range
E/VAC electrode (this study)
PVC electrode* (this study)
PVC electrode (reported)
4 x 10-e 1 x lo-6t 26 ca. 8-11
3 x 10-G 2 x lo-6t 24 ca. g-11
5 x lo-JS.11; 2 x 10-q 5 x 10-65 3W 7.5-105; 5-9.511
*Electrode prepared with the following composition: to 300 mg PVC dissolved in 4 ml THF is added 100 pl Orion sensor, I ml DOP, and 1 ml nitrobenzene. tDetermined by S log 2. #Griffiths et a1.‘4 $Lima and Machado.ls l/Moody and Thomas.16
was evaluated from calibration curves obtained over six months. In fact, measurements with this electrode prior to any activation step showed a short range of linear response and long response time. Figure 3 was evaluated from calibration curves obtained over six months. During the first week of conditioning, a large variation in the potential was observed. However, after this stage of stabilization, the change in the potential was about 9 mV over six months of use (Fig. 3, linear portions). After six months of use, the lower limit of linear response was 4 x 10e6M. And the detection limit (Fig. 3, lowest curve), taken as the calcium concentration at which the potential deviated by S log 2 from the extrapolation of the linear portions of the calibration curve, was 10e6M. The major effect of conditioning was to lower the response time of the electrode. The response time depended on the concentration of the test solution. For electrodes conditioned for one week, fast response (cu. 15 set for 95% response) was observed with solutions > 10m3M while in more dilute solutions, as much as 2 min or more were required for 95% response. Table 1 summarizes the characteristics of the E/VAC membrane calcium electrode compared with those of a PVC membrane electrode of similar composition prepared in our laboratory and those reported for PVC membrane calcium electrodes based on the same ion-exchanger. The E/VAC membrane electrode shows moreor-less the same performance as the PVC electrode prepared in this study. The limit of linear response is lower than that of other reported PVC electrodes. The slope found here is less than that reported for PVC based electrodes.‘4-‘6 While we cannot assume a lifetime as long as the eighteen months of the conventional electrode,16 the calibration parameters found for our electrode after six months of continuous use
indicate a long durability of this electrode. It is evident from Table 1, when comparison is made between E/VAC and PVC electrodes prepared under the same conditions, that E/VAC is as useful as a PVC like matrix membrane for ion-selective electrode construction. EfSect of pH and foreign cations
The pH dependence of the electrode potential was tested over the range 4-l 1 for 10-4-10-‘M calcium chloride (Fig. 4). The pH values of the solutions, prepared in a Beckman buffer at pH 4.9, were adjusted by addition of small volumes of 10M sodium hydroxide solution. The well known dip in the potential-pH curves which characterizes calcium ISEs based on calcium bis-dialkylphosphate type sensor”.‘* appears for the present membrane below pH 7. The plots in Fig. 4 indicate that at pH 8-l 1, the potentials are constant. Potentiometric measurements were therefore done at pH about 9.4. We note that at this pH, the electrode exhibits the same
Fig. 4. pH-effect on the Ca*+ E/VAC membrane electrode response at four different concentrations of calcium chloride.
Calcium ion-selective electrode preparation
characteristics by changing the bulfer from a Beckmann buffer to an acetate buffer. In order to investigate the selectivity of this new membrane, the response of the electrode was examined in the presence of various foreign cations. Potentiometric selectivity coefficients J&+,x2+ given in Table 1 were evaluated by both the separate solution method and the mixed solution method following the equation
Kca2+, xz+= [~a2+1/[x1+12’z where z is the charge on the foreign ion, [x”‘] is its concentration and [Ca*+] is the concentration of calcium that provides equal response to that of the interfering ion under investigation. According to the separate solution method, the electrode potentials were measured for solutions containing only interfering ion. Then the calcium ion concentration that produced the same potential as the foreign ion at [x2+] was determined graphically from the calibration curve established in the absence of any foreign ion. In the mixed solution case, calibration curves were established by measuring potentials of the solutions containing a constant concentration of interfering ion and varying the concentration of calcium from 10-6-10-‘M. The concentration at which the potential deviates by S log 2 from the extrapolation of the linear
933
portion of the curve was considered as [Ca*+]. If such deviation caused by the presence of the interfering ion did not begin at a concentration of calcium above 4 x 10e6M for the E/VAC electrode and 3 x 10e6M for the PVC electrode, which are the Nemstian limits in a pure solution of calcium, we have considered KRca2+,xz+ as < 10e4. The cations investigated with the concentrations stated in Table 2 are completely soluble in water at the working pH (9.4) of the electrode. The selectivity coefficients listed in Table 2 show that the E/VAC membrane calcium electrode has a high selectivity for calcium over many cations, especially relative to Na+, K+ and Mg2+, which are usually present in real samples. Only Mn*+ ions interfere significantly. Despite the small selectivity coefficient for Hg+, the calibration curve obtained in the presence of 10e3M Hg*+ shows a strong deviation from linearity between 10-l and lo-*M calcium. However, the electrode is not affected by immersing it in a solution of Hg2+ since complete recovery of the calcium response is observed after washing the electrode, without a need for reconditioning. Generally speaking there is good agreement between the selectivity coefficient values of our E/VAC electrode and those published’“” for PVC electrodes, especially when the mixed solution method was applied.
Table 2. Selectivity coefficients for E/VAC membrane Ca 2+ electrode based on Orion 92-20-02 sensor reagent, compared with PVC membrane electrodes Kc.z+.x2+ (separate solution method)
K, z;.xz+ (mixed solution method) Interferent specie Ix)’ Li+ Na+ Fe:+ Hg2+ cu2+ Ni2+ Pb2+ cos+ Sn2+ Zn2+ Mg2+ Ba2+ Sr2+ Cd2+ Mn2+
Present study E/VAC
PVC
< 10-4
-
<10-4
6 x lo-) 1 x 10-T 10 4 x 10-Z 2.5 x lO-2 -
2 x 10-l
Reported values? PVC 1.2$ 1.6 x 10-3§; 6.7 x lO-sll 5.1 x 10-q; 2.2 x 10-q
Present study EIVAC
Reported values?
PVC
PVC
6.1 x IO”11 5.5 x IO-311
q
2.1 x 10-34 G x 8x 4x 4.6 x -
IO-‘11;8 x lo-‘9 lo-‘& 4.5 x 10-311 10-4§; 2.3 x lO-3ll IO-‘1
*Concentrations of interfering ions were IO-‘M, except for Fe, Pb, Sn and Cd, which were 10e4M. Walues obtained under similar conditions of interfering ion concentration and calculation procedure as used in this study. IGriffiths er aLI4 #Lima and Machado.15 l/Moody and ThomasI # Moody and Thomas.”
Calcium ion-selective electrode preparation 8. J. D. R. Thomas, Anal. Chim. Acta, 1986, 180, 289. 9. R. S. Lawton and A. M. Yacynych, ibid., 1984, 160, 149. 10. J. D. R. Thomas, J. Chem. Sot., Faraday Trans., 1986, 82, 1135. 11. U. Gesch and W. Simon, Anal. Chem., 1980, S2, 692. 12. 0. F. Schafer, Anal. Chim. Acta, 1976, 87, 495. 13. U. Fiedler and J. RiZZka, ibid., 1973, 67, 179. 14. G. H. Griffiths, G. J. Moody and J. D. R. Thomas, Analyst, 1972, 97, 420.
935
15. J. L. F. C. Lima and A. A. S. C. Machado, ibid., 1986, 111, 799. 16. G. J. Moody, R. 8. Oke and J. D. R. Thomas, ibid.,
1970, 95, 910. 17. G. J. Moody and J. D. R. Thomas, in Chemical Sensors, T. E. Edmonds (ed.), p. 83. London, 1988. 18. P. L. Bailey, Analysis with Ion-Selective Electrodes, 2nd Ed., p. 124. Heyden, London, 1980. 19. A. A. Bouklouze, A. El-Jammal, G. J. Patriarche and G. D. Christian, unpublished work.
0039-9 MO/9 I s3.00 + 0.00 Copyright 0 1991 Fkrgamon Press plc
USE OF ETHYLENE-VINYL-ACETATE AS A NEW MEMBRANE MATRIX FOR CALCIUM ION-SELECTIVE ELECTRODE PREPARATION A. EL-JAMMAL,A. A. BOUKLOUZEand G. J. PA~UARCHE Institut de Pharmacie, CP 205/6, Universite Libre de Bruxelles, Boulevard du Triomphe, 1050 Bruxelles, Belgium
G. D. CHRISTIAN* Department of Chemistry, BG-10, University of Washington, Seattle, WA 98195, USA (Received 28 November 1990. Revised 12 February 1991. Accepted 20 February 1991)
Summary-A new polymer matrix based membrane electrode with an ion-exchanger responding to calcium was constructed by dissolving the copolymer ethylene-vinyl-acetate together with the ionexchanger in chloroform in the presence of a mixture of dioctylphthalate-nitrobcnzene as plasticizer. The ion-exchanger used as the electroactive component was calcium didecyl phosphate in di-(n-octylphenyl) phosphonate (Orion). This electrode exhibited near-Nemstian response over the concentration range lo-‘4 x 10m6M calcium. The pH did not affect the electrode performance within the range 8-l 1. Response time varied from 15 to 120 se-cand the lifetime exceeded six months. The membrane is subject to static charge buildup, but this is avoided by controlling the level of dryness of the membrane. Selectivity coefficients determined for both monovalent and divalent cations showed negligible interference by most of these ions. The electrode was applied successfully to the determination of calcium in commercial mineral waters.
The use of ion-selective membrane electrodes is finding increasing application in pharmaceutical analysis. ‘-3The development of sensors that are easy to use and maintain, fast-responding, with low cost materials have boosted the use of potentiometric methods. Traditional methods for analysis utilizing spectrophotometry,4 polarchromatography6 and enzymatic ography,’ measurements’ that may be characterized in some cases by low detection limits and high selectivity may possess limitations such as toxicity of reagents, high cost and complexity of the apparatus or the procedure. Ion-selective polymer-membrane electrodes based on ion-pair or chelating ligand complexes entrapped in the polymer, suffer from problems such as interferences and leaching of the the active component and the plasticizer&” from the polymer matrix. The leaching, which causes the deterioration of the response of the electrodes and leads to limited electrode lifetime, depends primarily on the nature (the lipophilicity) of the softener and the membrane active component. I1 In order to minimize the problems of interference, the use of alternatives *Author for correspondence. TAL 3816-l
to poly(viny1 chloride) (PVC) as the polymer matrix have been suggested.’ However, depending on the polymer used, the electrodes may exhibit poorer performance. Hence, the use of VAGH copolymer (hydrolyzed vinyl chloride/ vinyl acetate) for calcium electrodes offers no advantages,’ while poly(viny1 isobutyl ether) exhibits the same behavior as a PVC electrode containing the same sensing ingredients.r2 In this paper, we describe the use of ethylenevinyl-acetate copolymer (EjVAC) as the matrix membrane for construction of ion-selective polymer-membrane electrodes. The use of this copolymer in an ISE has not previously been reported. The structure is as follows: -(-CH,--CH,-),--(-CH,-CH-),--
E/VAC
L L -0 L H3 A new method for the assembly of this electrode was developed. This new membrane has been characterized for use in construction of a calcium ion-selective electrode. The proposed electrode exhibits high selectivity for cakium 929
A. EL-JAMMAL et al.
930
over many cations. The main advantages and disadvantages of the electrode are compared with the ISE, previously developed for the determination of calcium. EXPERIMENTAL
Reagents
All reagents were of analytical-reagent grade and were used without further purification. E/VAC, 40% weight of the vinyl acetate monomer in the copolymer, (LEVAPREN 400, m < n) Bayer was kindly supplied by Professor G. Geuskens (ULB). Sources of all other reagents were as follows: tetrahydrofuran, chloroform and dioctyl phthalate from Janssen and nitrobenzene from Merck. All solutions were prepared in Milli-Q water. Stock solutions, lo- ‘M, were stored at room temperature and standard solutions were prepared from stock solutions by sequential dilution with Milli-Q water.
Fig. 1. Schematic diagram of the electrode: (1) coaxial cable; (2) glass tube, (3) cable junction; (4) paratilm; (5) micropipette tip; (6) glass tube; (7) lo-‘MCaCl,; (8) Ag/AgCl; (9) E/VAC membrane.
Preparation of the electrode
A commercially available solution of Orion 92-20-02 calcium didecyl phosphate in di(n-octylphenyl) phosphonate was used as the electroactive material for the calcium ISE. Membranes were prepared by dissolving the electroactive material (100 ~1) and about 350 mg of E/VAC in 4 ml of an appropriate solvent (chloroform) at about 60” in a water bath. To this solution is added 1 ml of both dioctyl phthalate and nitrobenzene as plasticizer and the mixture was exposed to ultrasound to insure good dispersion of electroactive material in the matrix. Ag/AgCl
lo-‘M CaCl,
Membrane E/VAC
Potential measurements
Potentiometric measurements were carried out at room temperature with a Tacussel Ion processor II. A Tacussel 6000 pH-meter with a Tacussel glass electrode was used to measure pH. A saturated calomel electrode was employed as the reference electrode in both measurements. Electrode response was examined by immersing the membrane electrode in 10 ml of buffered solution together with the double junction saturated calomel reference electrode. The cell used to investigate the characteristics of the electrode may be represented as follows:
/ Buffer 1 KC1 / Hg2C12/Hg Sample solution Ca2+ (xM), pH (y) bridge saturated
A simplified scheme of the electrode is shown in Fig. 1. A micropipette tip was used as the electrode body. The end of the tip was dipped in the polymer mixture for a few seconds and then it was left in air for three hours. Longer drying times should not be used, in order to minimize static electricity build up. Calcium chloride solution, lo-‘M, was used as the internal reference solution, and the internal reference electrode was an Ag-AgCl electrode. The electrode was activated before use by soaking it overnight in a lo-‘M solution of calcium chloride and it was stored in the same solution when not in use.
The calibration curves were generated by addition of a O.l-ml aliquot of 10-6-10-‘M aqueous primary ion solution to 10 ml of the stirred solution. The potential was recorded when the reading became stable (*O. 1 mV/ 10 set) after each addition. The data were plotted as observed potential vs. the logarithm of the concentration of the primary ion. Concentration of calcium in commercial mineral water were then determined in a similar manner by means of a calibration graph, and the values obtained were checked by standard addition.
931
Calcium ion-selective electrode preparation RESULTS AND DISCUSSION
Membrane materials It has been suggested that polymers with low glass transition temperatures (Tg) are suitable matrices for polymer membrane electrodes.13 Our results agree with this hypothesis since polystyrene (Tg = 100”) tested in our laboratory resulted in electrodes with no response, while good results have been found for the measurement of both organic (methixene, promethazine and some local anesthetics) and inorganic (calcium) cations by using E/VAC matrix whose glass transition temperature is below room temperature (Tg = - 50”). This study reports the results for the measurement of calcium ions. An advantage of E/VAC is the wide range of solvents that can be used to dissolve it, which allowed the use of alternatives to tetrahydrofuran and cyclohexanone commonly used with PVC. Six solvents are able to dissolve E/VAC: carbon tetrachloride, chloroform, dichloromethane, p-xylene, tetrahydrofuran and toluene. From a response-stability point of view, the use of E/VAC instead of PVC causes an excessive susceptibility of the electrode to static electricity if allowed to become too hard, but this problem can be avoided by controlling the level of dryness of the membrane. Excellent potential stability could be achieved by limiting the dryness to an extent that provides for minimum mechanical strength of the membrane in the ISE, accomplished by allowing to air dry for three hours only. The technique described for the assembly of the electrode, in which the sensor is inserted into the end of a micropipette tip, demonstrates the use of a simple and inexpensive material for the body of the electrode and insures good protection of the sensor against solution turbulence.
I i’ \ \\\
Izomv
x + .
\
x
x\*, \’. +\+_:-+ +
. \ l\ \
.\ L l,,
\
I
I
I
I 3
l-0,.
‘7 .-m-m
I 5
I 7
pCa’+ Fig. 2. Dependence of the Ca*+ E/VAC membrane electrode response on the calcium ion-exchanger content per 6 ml of polymer mixture. (0) 20 ~1 (a) SO~1 (0) 100 PI(@) 200 ~1 (+ ) 500 pl ( x ) 1 ml. pH 9.4, Beckman Buffer.
tration decade and a correlation coefficient of 0.9996. The range of linear response for the first three electrodes (Orion sensor reagent > 0.1 ml) can be extended to lo-‘M by a much longer period of activation or after several uses of the electrode. An electrode prepared under the condition of 100 pl of sensor reagent without any other softener gives no response at all. Therefore, we selected the mixture dioctylphthalate nitrobenzene as plasticizer for our calcium ISE. The performance of the electrode with time (Fig. 3)
Response characteristics of the electrode Several parameters were investigated for optimal performance of the electrode. Typical calibration curves obtained for electrodes containing different amounts of calcium ionexchanger, after one night of conditioning in O.lM calcium chloride are shown in Fig. 2. The optimal tested amount of the Orion sensor reagent appears to be 0.1 ml per 6 ml of polymer mixture, for maximum potential response to calcium ion. The electrode demonstrates, under these conditions, a linear response in the lo-‘lo-‘M range, with a useful range extended to 5 x 10e6M, a slope of about 26 mV per concen-
pCa’+ Fig. 3. Calibration plots for CaZ+ E/VAC membrane electrodes in buffered solution, pH 9.4. (0) electrode one week old. (0) electrode six months old.
A. EL-JAMMAL et
932
al.
Table 1. Characteristics of polymer membrane Ca *+ electrodes based on Orion 92-20-02 sensor Electrode type
Lower limit of response (M) Detection limit Slope (mv/decade) pH range
E/VAC electrode (this study)
PVC electrode* (this study)
PVC electrode (reported)
4 x 10-e 1 x lo-6t 26 ca. 8-11
3 x 10-G 2 x lo-6t 24 ca. g-11
5 x lo-JS.11; 2 x 10-q 5 x 10-65 3W 7.5-105; 5-9.511
*Electrode prepared with the following composition: to 300 mg PVC dissolved in 4 ml THF is added 100 pl Orion sensor, I ml DOP, and 1 ml nitrobenzene. tDetermined by S log 2. #Griffiths et a1.‘4 $Lima and Machado.ls l/Moody and Thomas.16
was evaluated from calibration curves obtained over six months. In fact, measurements with this electrode prior to any activation step showed a short range of linear response and long response time. Figure 3 was evaluated from calibration curves obtained over six months. During the first week of conditioning, a large variation in the potential was observed. However, after this stage of stabilization, the change in the potential was about 9 mV over six months of use (Fig. 3, linear portions). After six months of use, the lower limit of linear response was 4 x 10e6M. And the detection limit (Fig. 3, lowest curve), taken as the calcium concentration at which the potential deviated by S log 2 from the extrapolation of the linear portions of the calibration curve, was 10e6M. The major effect of conditioning was to lower the response time of the electrode. The response time depended on the concentration of the test solution. For electrodes conditioned for one week, fast response (cu. 15 set for 95% response) was observed with solutions > 10m3M while in more dilute solutions, as much as 2 min or more were required for 95% response. Table 1 summarizes the characteristics of the E/VAC membrane calcium electrode compared with those of a PVC membrane electrode of similar composition prepared in our laboratory and those reported for PVC membrane calcium electrodes based on the same ion-exchanger. The E/VAC membrane electrode shows moreor-less the same performance as the PVC electrode prepared in this study. The limit of linear response is lower than that of other reported PVC electrodes. The slope found here is less than that reported for PVC based electrodes.‘4-‘6 While we cannot assume a lifetime as long as the eighteen months of the conventional electrode,16 the calibration parameters found for our electrode after six months of continuous use
indicate a long durability of this electrode. It is evident from Table 1, when comparison is made between E/VAC and PVC electrodes prepared under the same conditions, that E/VAC is as useful as a PVC like matrix membrane for ion-selective electrode construction. EfSect of pH and foreign cations
The pH dependence of the electrode potential was tested over the range 4-l 1 for 10-4-10-‘M calcium chloride (Fig. 4). The pH values of the solutions, prepared in a Beckman buffer at pH 4.9, were adjusted by addition of small volumes of 10M sodium hydroxide solution. The well known dip in the potential-pH curves which characterizes calcium ISEs based on calcium bis-dialkylphosphate type sensor”.‘* appears for the present membrane below pH 7. The plots in Fig. 4 indicate that at pH 8-l 1, the potentials are constant. Potentiometric measurements were therefore done at pH about 9.4. We note that at this pH, the electrode exhibits the same
Fig. 4. pH-effect on the Ca*+ E/VAC membrane electrode response at four different concentrations of calcium chloride.
Calcium ion-selective electrode preparation
characteristics by changing the bulfer from a Beckmann buffer to an acetate buffer. In order to investigate the selectivity of this new membrane, the response of the electrode was examined in the presence of various foreign cations. Potentiometric selectivity coefficients J&+,x2+ given in Table 1 were evaluated by both the separate solution method and the mixed solution method following the equation
Kca2+, xz+= [~a2+1/[x1+12’z where z is the charge on the foreign ion, [x”‘] is its concentration and [Ca*+] is the concentration of calcium that provides equal response to that of the interfering ion under investigation. According to the separate solution method, the electrode potentials were measured for solutions containing only interfering ion. Then the calcium ion concentration that produced the same potential as the foreign ion at [x2+] was determined graphically from the calibration curve established in the absence of any foreign ion. In the mixed solution case, calibration curves were established by measuring potentials of the solutions containing a constant concentration of interfering ion and varying the concentration of calcium from 10-6-10-‘M. The concentration at which the potential deviates by S log 2 from the extrapolation of the linear
933
portion of the curve was considered as [Ca*+]. If such deviation caused by the presence of the interfering ion did not begin at a concentration of calcium above 4 x 10e6M for the E/VAC electrode and 3 x 10e6M for the PVC electrode, which are the Nemstian limits in a pure solution of calcium, we have considered KRca2+,xz+ as < 10e4. The cations investigated with the concentrations stated in Table 2 are completely soluble in water at the working pH (9.4) of the electrode. The selectivity coefficients listed in Table 2 show that the E/VAC membrane calcium electrode has a high selectivity for calcium over many cations, especially relative to Na+, K+ and Mg2+, which are usually present in real samples. Only Mn*+ ions interfere significantly. Despite the small selectivity coefficient for Hg+, the calibration curve obtained in the presence of 10e3M Hg*+ shows a strong deviation from linearity between 10-l and lo-*M calcium. However, the electrode is not affected by immersing it in a solution of Hg2+ since complete recovery of the calcium response is observed after washing the electrode, without a need for reconditioning. Generally speaking there is good agreement between the selectivity coefficient values of our E/VAC electrode and those published’“” for PVC electrodes, especially when the mixed solution method was applied.
Table 2. Selectivity coefficients for E/VAC membrane Ca 2+ electrode based on Orion 92-20-02 sensor reagent, compared with PVC membrane electrodes Kc.z+.x2+ (separate solution method)
K, z;.xz+ (mixed solution method) Interferent specie Ix)’ Li+ Na+ Fe:+ Hg2+ cu2+ Ni2+ Pb2+ cos+ Sn2+ Zn2+ Mg2+ Ba2+ Sr2+ Cd2+ Mn2+
Present study E/VAC
PVC
< 10-4
-
<10-4
6 x lo-) 1 x 10-T 10 4 x 10-Z 2.5 x lO-2 -
2 x 10-l
Reported values? PVC 1.2$ 1.6 x 10-3§; 6.7 x lO-sll 5.1 x 10-q; 2.2 x 10-q
Present study EIVAC
Reported values?
PVC
PVC
6.1 x IO”11 5.5 x IO-311
q
2.1 x 10-34 G x 8x 4x 4.6 x -
IO-‘11;8 x lo-‘9 lo-‘& 4.5 x 10-311 10-4§; 2.3 x lO-3ll IO-‘1
*Concentrations of interfering ions were IO-‘M, except for Fe, Pb, Sn and Cd, which were 10e4M. Walues obtained under similar conditions of interfering ion concentration and calculation procedure as used in this study. IGriffiths er aLI4 #Lima and Machado.15 l/Moody and ThomasI # Moody and Thomas.”
Calcium ion-selective electrode preparation 8. J. D. R. Thomas, Anal. Chim. Acta, 1986, 180, 289. 9. R. S. Lawton and A. M. Yacynych, ibid., 1984, 160, 149. 10. J. D. R. Thomas, J. Chem. Sot., Faraday Trans., 1986, 82, 1135. 11. U. Gesch and W. Simon, Anal. Chem., 1980, S2, 692. 12. 0. F. Schafer, Anal. Chim. Acta, 1976, 87, 495. 13. U. Fiedler and J. RiZZka, ibid., 1973, 67, 179. 14. G. H. Griffiths, G. J. Moody and J. D. R. Thomas, Analyst, 1972, 97, 420.
935
15. J. L. F. C. Lima and A. A. S. C. Machado, ibid., 1986, 111, 799. 16. G. J. Moody, R. 8. Oke and J. D. R. Thomas, ibid.,
1970, 95, 910. 17. G. J. Moody and J. D. R. Thomas, in Chemical Sensors, T. E. Edmonds (ed.), p. 83. London, 1988. 18. P. L. Bailey, Analysis with Ion-Selective Electrodes, 2nd Ed., p. 124. Heyden, London, 1980. 19. A. A. Bouklouze, A. El-Jammal, G. J. Patriarche and G. D. Christian, unpublished work.