ANALYST NOVEMBER 1986 VOL. 111 1231 Electrodes in Series. Simultaneous Flow Injection Determination of Chloride and pH with Ion-selective Electrodes Jacobus F. van Staden Department of Chemistry University of Pretoria Pretoria 0002 South Africa Chloride and pH can be determined simultaneously by flow injection potentiometry with a series electrode arrangement at a rate of 60 samples per hour with a standard deviation of 1.5% for chloride and 1% for pH. Samples (30 pl) are injected into a carrier buffer solution (pH 7.6) containing 0.5 mol dm-3 of potassium nitrate and mol dm-3 of sodium dihydrogen phosphate as an ionic strength adjustment buffer. The sample -buffer zone formed is transported through a laboratory-made coated tubular solid-state chloride-selective electrode via a glass membrane micro pH combination electrode on to the reference electrode (for chloride).The method is suitable for the determination of chloride in the range 20-5000 mg dm-3 and a pH range from 3 to 10. Keywords Chloride determination; pH determination; simultaneous determination; flow injection analysis; ion-selective electrodes The use of electrochemical detection in flowing streams has grown rapidly in recent years.1-5 There are many examples of potentiometric detection with glass pH electrodes69 and also with other membrane electrodes.1,4>5JG16 However in most of the methods reported the emphasis was placed on the determination of a single species in a sample. A survey of the recent literature indicated a lack of procedures describing the simultaneous determination of more than one species in the same sample,17 which could make the flow injection analysis (FIA) concept more attractive for routine laboratories.By using the cascade electrode arrangement RfiiiCka et al. 1 were able to determine sodium and potassium simultaneously in blood serum. In this configuration the stream of electrolyte cascaded from the potassium to the sodium electrode and then to a constant-level reservoir in which the reference electrode was located. Mascini and Palleschils used the same principle for the simultaneous determination of glucose and urea in serum samples but in their arrangement the cascade elec-trodes were placed opposite each other with the stream of electrolyte cascading between the two electrodes. Virtanenlg developed a method for the simultaneous determination of potassium sodium calcium and chloride in serum by placing four cascade ion-selective electrodes sequentially in an FIA system with the reference electrode downstream.The errors in the analysis of samples due to mutual interferences from some species in the determination were corrected with the aid of regression coefficients which were determined by measure-ments of known mixtures. The same idea of electrodes in series was implemented by Hansen et aZ.20 for the simul-taneous determination of pH and calcium in serum. In their system the carrier buffer was pumped via a flow-through capillary glass electrode (G299 Radiometer) to a cascade calcium-selective electrode. The reference electrode located at the bottom of the calcium electrode received both the sample flow after impact with the calcium sensor and an additional reagent stream of buffer.The additional buffer stream was added to dilute the sample in such a way that the composition of the solution surrounding the reference elec-trode almost remained constant. Flow-through tubular arrangements for ion-selective elec-trodes have been reported recently. 1 2 - 1 4 ~ 6 ~ 2 2 In this geo-metric mode the sample solution is channelled through the tubular configuration across the sensing membrane surface in a kind of open path. The incorporation of a tubular ion-selective electrode into the conduits of a flow injection system seems an ideal design as the hydrodynamic flow conditions can be kept constant throughout the flow system.This approach opens new dimensions in the development of the simultaneous determination of more than one species in a single sample by placing tubular electrodes in series into the conduits of a flow injection system. This paper describes the exploitation of this concept in a study of the simultaneous determination of chloride and pH in a single sample. Experimental Reagents and Solutions All reagents were prepared from analytical-reagent grade chemicals unless otherwise specified. Doubly distilled de-ionised water was used throughout. The water was tested beforehand for traces of chloride. All solutions were de-gassed before measurements by the use of a water vacuum pump. Ionic strength adjustment buffer reagent (ISA B ) Dissolve 2.76 g of sodium dihydrogen phosphate monohy-drate in 1500 cm3 of distilled water in a 2-dm3 calibrated flask, add 101.11 g of potassium nitrate and swirl the flask gently until the solid has completely dissolved.Dilute the solution quantitatively to 1900 cm3 with distilled water. Adjust the pH carefully to 7.6 by adding sodium hydroxide or nitric acid. Dilute this solution to 2 dm3 with distilled water. For 0.1 and 1.0 mol dm-3 potassium nitrate solution 20.222 and 202.22 g of potassium nitrate are used respectively. Buffer and chloride standard solutions Dissolve 32.9680 g of dried sodium chloride in 2 dm3 of distilled water to give a stock solution with a chloride concentration of 10000 mg dm-3. Combined working solu-tions containing chloride in the range 20-5000 mg dm-3 and a pH range of 3-10 are prepared by suitable dilution of the chloride stock solution and as described in Tables 10.25, 10.37 10.43 10.45 and 10.5 in the book by Perrin and Dempsey ,23 respectively.Apparatus Electrodes Tubular flow-through chloride-selective cell. The construc-tion preparation and conditioning of the coated tubular flow-through solid-state chloride-selective membrane was similar to the procedure previously described. l 1232 ANALYST NOVEMBER 1986 VOL. 111 Micro combination p H electrode. A Schott micro pH combination electrode with a Type N60 cylindrical glass membrane Ag - AgCl internal reference elements zero potential pH value = 7 platinum junction glass resistance ( R ) at 25 "C = 600 MB shaft diameter 3 mm and an operating range of pH 0-14 in the temperature range 20-80 "C was used.The final arrangement of the electrodes in the flow system is shown in Fig. 1. The carrier solution containing the sample is first channelled through the tubular chloride-selective elec-trode then via the glass membrane on to the reference electrode. Flow system The electrodes were incorporated into the conduits of a flow injection system as shown schematically in Fig. 2. The injection valve system was a Carle microvolume two-position sampling valve (Carle Catalogue No. 2014) with two identical sample loops each having a volume of 30 1.11. The sampling unit (Cenco) was used together with a Cenco peristaltic pump that supplied a constant stream of samples to the sampling valve system.The valve system was actuated on a time basis that was correlated with the sampler unit. A 60-s cycle Micro - pH combination electrode Electrode holder- - '' ~ ' li I I l l I Shielded cable to ionalyzer I l l ' Glass membrane-i-) ' ; i Coated insoluble Ag metal inorganic salt Fig. 1. silver chloride electrode surface is ca. 2 mm i.d. and 5 mm long Electrodes arrangement into the flow system. The active Sampler sampling time was used giving the system a capacity of about 60 samples h-1; the sampling valve was actuated every 58 s. The carrier and reagent streams were pumped with a Cenco peristaltic pump operating at 10 rev. min-1. Tygon tubing (0.51 mm i.d.) was used to construct the manifold. Mixing coils were made by winding appropriate lengths of the Tygon tubing on Perspex rods (15 mm 0.d.).The carrier stream, containing a solution of the ionic strength adjustment buffer reagent was pumped at a constant flow-rate of 3.9 cm3 min-1. To avoid the slight pulsation originating at the peristaltic pump and also sample plug pulsation pulse suppressors were used. For plug pulsation 105 cm x 0.51 mm i.d. transmission tubing was incorporated just after the sampling valve. Pump pulsation was avoided by the incorporation of 200 cm X 0.51 mm i.d. transmission tubing as a pulse suppressor between the pump and the sampling valve. The potentials were measured at room temperature with an Orion Research (Model 901) microprocessor Ionalyzer. The detector output was recorded with a two-channel Cenco recorder (Catalogue No.34195-041). The constructed flow-through tubular indicator electrodes were used in conjunction with an Orion 90-02 double-junction reference electrode with 10% potassium nitrate as the outer chamber filling solution. Results and Discussion Two typical recorder output series taken at a rate of 60 determinations per hour are given in Figs. 3 and 4. The results were obtained by injecting 30 1-11 of mixed sodium chloride pH standard solutions into pH 7.6 sodium dihydrogen phosphate (10-2 mol dm-3) ionic strength adjustment buffer carrier solutions. The experimental conditions for both series were the same except for a change from 0.1 mol dm-3 potassium nitrate solution in Fig. 3 to 1.0 mol dm-3 potassium nitrate solution in Fig. 4 in the ionic strength adjustment carrier solution.No difficulty was experienced in the determination of pH in both series provided that the precautions as previously described by RfiiiEka and Hansen4.24 were taken [Figs. 3(b) and 4(b)]. Although R8iiCka and Hansen4.24 used sodium chloride as the ionic strength adjustment solution, which is not possible in this instance the results indicated that potassium nitrate was also suitable for this purpose. The incorporation of the tubular chloride-selective electrode before the glass membrane (pH) electrode shows no signifi-cant influence on the peak shape. This means that the effect of the incorporated tubular system on the sample - buffer zone is negligible which proves the assumption that the hydrody-namic flow conditions can be kept constant through the microprocessor - - .- . x { T w o - p e n C recorder J Reference electrode -=? iona I yzer \ r combination electrode cm3min Pert sta I t ic Fig. 2. rate 60 h-1 Manifold and flow diagram of the FIA system. Valve loop size 30 p1; sampler 60 s; wash 0 s; valve actuation at 58 s; samplin ANALYST NOVEMBER 1986 VOL. 111 1233 ( b ) pH 9.9 3.2 L I I ( C) pH = 8.6 B D F H AI cl EIG~I nnnnnnrinn BDF H pH = 6.1 Fig. 3. Typical strip-chart recording for the simultaneous flow injection analysis of chloride and pH with the FIA s stem of Fig. 2 using 0.1 rnol dm- KNO, 10-2mol dm-3 NaH2P04 &H adjusted to 7.6) ionic strength adjustment buffer carrier and reagent solutions. (a) Recorder trace for a series of standard chloride solutions each solution in triplicate.Recorder paper speed 1 mm rnin-'; recorder range 20 mV. Chloride concentrations A 5000; B 4000; C 3000; D, 2000; E 1000; F 500; G 250; H 100; I 80; J 60; K 40; and L 20 m dm-3 (p.p.m.). (b) Recorder trace for a series of standard pH sofutions each solution in triplicate. Recorder paper speed 1 mm min-'; recorder range 50 mV. Numerals on calibration peaks refer to pH values. (c) Recorder trace demonstrating the influence of chloride concentration on the sensitivity of the pH readout of a pH 6.1 and 8.6 solution each chloride interference injected in triplicate. Recorder paper speed 1 mm min-1; recorder range 50 mV. Chloride concentrations in pure buffer solutions A 0; B 100; C 250; D 500; E 1000; F 2000; G 3000; H 4000; and I 5000 mg dm-3 (p.p.m.) 9.9 7.6 111.I 3.2 II" AI cl EIGI I B DF H DH = 6.1 Fig. 4. Typical strip-chart recording for the simultaneous flow injection analysis of chloride and pH with the FIA system of Fig. 2 and 1.0 rnol dm-3 KN03 10-2 mol dm-3 NaH2P04 (pH adjusted to 7.6), ionic strength adjustment buffer carrier and reagent solutions. (a) Recorder trace for a series of standard chloride solutions each solution injected in triplicate. Recorder paper speed 1 mm min-l; recorder ran e 20 mV. Chloride concentrations A 5000; B 4000; C, 3000; D 2008; E 1000; F 500; G 250; H 100; I 80; J,.60; K 40; and L 20 mg dm-3 (p.p.m. j. (bj Recorder trace for a series of standard pH solutions each solution injected in triplicate. Recorder paper speed 1 mm min-1; recorder range 50 mV.Numerals on calibration peaks refer to pH values. ( c ) Recorder trace demonstrating the influence of chloride concentration on the sensitivity of the pH readout of a pH 6.1 and 8.6 solution each chloride interference injected in triplicate. Recorder paper speed 1 mm min-'; recorder range 50 mV. Chloride concentrations in pure buffer solutions A 0, B 100; C 250; D 500; E 1000; F 2000; G 3000; H 4000; and I 5000 mg dm-3 (p.p.m.) A I Fig. 5. Recorder trace for a series of standard chloride solutions with the FIA system of Fig. 2 using 0.5 mol dm-3 KN03 and 10-2 rnol dm-3 NaH2P04 (pH adjusted to 7.6) ionic strength adjustment buffer carrier and reagent solutions. Each solution injected in triplicate. Recorder paper speed 1 mm min-l; recorder range 20 mV.Chloride concentrations A 5000; B 4000; C 3000; D 2000; E 1000; F 500; G 250; H 100; I 80; J 60; K 40; and L 20 mg dm-3 (p.p.m.) tubular chloride-selective electrode. The calibration graph is still linear between pH 3 and 10 and the precision is good (the coefficients of variation were less than 1% for ten injections). Figs. 3(c) and 4(c) demonstrate the influence of 0-5000 mg dm-3 of chloride on the sensitivity of the pH readout. It is clear from the results obtained that a variation of chloride content between the two limits has no effect on the pH results. It was obvious from preliminary experiments that the flow injection conditions for the determination of chloride pre-viously described16 were not suitable for use in the simul-taneous determination of chloride and pH.The results indicated that the combination of dihydrogen phosphate with potassium nitrate carrier solution has an effect on the tubular chloride-selective electrode system. A mixture containing 1 .O mol dm-3 of potassium nitrate and 10-2 rnol dm-3 of sodium dihydrogen phosphate at pH 7.6 as a carrier solution tends to give a positive base-line drift [Fig. 4(a)]. This effect became large when 30-p1 standard solutions containing less than 250 mg dm-3 of chloride were injected. A decrease of the potassium nitrate strength to 0.1 rnol dm-3 in the pH 7.6 ionic strength adjustment buffer carrier solution resulted in a less stable measuring system for chloride [Fig. 3(a)] with a lower precision at the same time.Optimum working conditions in terms of stability and precision (the coefficients of variation were less than 1.5% for ten injections) are obtained by using 0.5 rnol dm-3 of potassium nitrate together with 10-2 mol dm-3 of sodium dihydrogen phosphate at a pH of 7.6 as carrier solution (Fig. 5 ) for the chloride determination which also suited the simultaneous determination of both com-ponents very well. Conclusion The simultaneous determination of chloride and pH with a flow injection system at a sampling rate of about 60 samples per hour and a standard deviation of less than 1.5% for chloride and 1% for pH is made possible by using electrodes in series in the procedure described in this paper 1234 ANALYST NOVEMBER 1986 VOL. 111 The author thanks the Council for Scientific and Industrial Research Pretoria and the University of Pretoria for financial support.He also thanks Miss M. L. Aveling for assistance in performing some of the experiments. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. References RfiiiEka J. Hansen E. H. and Zagatto E. A. Anal. Chim. Acta 1977 88 1. Pungor E. Fehkr Zs. Nagy G. Toth K. Horvai G. and Gratzl M. Anal. Chim. Acta 1979 109 1. Toth K. Nagy G. Fehkr Zs. Horvai G. and Pungor E., Anal. Chim. Acta 1980 114 45. Rfiiitka J. and Hansen E. H. “Flow Injection Analysis,” Wiley Chichester 1981. “Flow Injection Analysis Bibliography,” Tecator AB Hoga-nas Sweden 1985. Rfiiitka J. Hansen E. H. and Ghose A. K. Anal. Chem., 1979 51 199. Zagatto A. E. Reis B.F. Bergamin F” H. and Krug F. J., Anal. Chim. Acta 1979 109 45. Astrom O. Anal. Chim. Acta 1979 105 67. Basson W. D. and Van Staden J. F. Lab. Pract. 1980 29, 632. Hansen E. H. Krug F. J . Ghose A. K. and RfiiiEka J . , Analyst 1977 102 714. Hansen E. H. Ghose A. K. and RfiiiEka J . Analyst 1977, 102 705. Van der Linden W. E. and Oostervink R. Anal. Chim. Acta, 1978 101 419. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. Meyerhoff M. E. and Kovach P. M. J . Chem. Educ. 1983, 60 766. Frend A. J. Moody G. J. Thomas J . D. R. and Birch B. J., Analyst 1983 108 1357. A1 Hitti I. K. and Thomas J. D. R. Anal. Lett. 1985 18, 975. Van Staden J. F. Anal. Chim. Acta 1986 179 407. Luque de Castro M. D. and ValcBrcel M. Analyst 1984, 109 413. Mascini M. and Palleschi G. Anal. Chim. Acta 1983 145, 213. Virtanen R. in Pungor E. Editor “Ion-selective Electrodes 3. Proceedings of the Third Symposium Matrafiired Hungary, October 1980; Analytical Symposium Series,” Volume 8, Elsevier Amsterdam 1981 p. 375. Hansen E. H. RfiiiEka J. and Ghose A. K . Anal. Chim. Acta 1978 100 151. Mascini M. and Palleschi G. Anal. Chim. Acta 1978 100, 215. Alegret S. Alonso J. Bartroli J. Paulis J. M. Lima, J. L. F. C. and Machado A. A. S. C. Anal. Chim. Acta 1984, 164 147. Perrin D. D. and Dempsey B. “Buffers for pH and Metal Ion Control,” Chapman and Hall London 1974. Rfiiitka J . and Hansen E. H. Anal. Chim. Acta 1984,161, 1. Paper A61162 Received May 23rd 1986 Accepted June 25th 198