ANALYST JANUARY 1984 VOL. 109 15 Studies with lmmobilised Chemical Reagents Using a Flow-cell for the Development of Chemically Sensitive Fibre-optic Devices Gordon F. Kirkbright Ramaies Narayanaswamy and Neal A. Welti Department of Instrumentation and Analytical Science UMlST Manchester M60 7QD UK Results of studies using a flow-cell and immobilised indicator dye reagents for the development of a chemically sensitive fibre-optic system for measurement of pH are presented. The indicator dye reagents are immobilised on a cross-linked styrene - divinyl benzene polymer matrix and reflectance measurements are made as a function of pH using a bifurcated fibre-optic system. Keywords Flow-cell; immobilised chemical reagents; chemically sensitive pH fibre-optic device Recently the development of chemically sensitive fibre-optic devices has been the subject of considerable interest.The techniques employed have normally been based on the immobilisation of chemical reagents that allow a colorimetric or fluorimetric indication of the chemical change in a solution. Fibre-optic pH sensors based on the use of classical indicator dye optical absorption have been developed for physiological and other uses. 1-3 Optical waveguides coated with chemical reagents which utilise changes in absorption of the reagent material have been employed to sense ammonia in the vapour phase."5 Fibre-optic sensors using changes in fluorescence have been reported for the measurement of pH,6 aluminium7 and glucose.* Chemically sensitive fibre-optic devices have several advan-tages over conventional potentiometric or voltammetric elec-trode sensors.One of the most important advantages is their inherent safety in operation due to their non-electrical nature. Other advantages of such sensors include their small size and flexibility capability of remote operation in hostile environ-ments rugged construction and reliability. Our present interest is in the development of fibre-optic transduction systems based on immobilised chemical reagents for the measurements of pH gases and metals for both industrial on-line applications and environmental and clinical purposes. In this paper we report our preliminary studies using a flow-cell to assist the development of chemically sensitive fibre-optic probes for measurement of pH.A reliable small sensor based on immobilisation of bromothymol blue on XAD-2 polymer has been developed as a result of these studies; the performance characteristics of this sensor and the development of a simple dedicated pH meter into which this is incorporated are to be described elsewhere. Experimental Instrumentation The arrangement of the flow-cell used in the study of the immobilisation of indicator dyes and its associated instrumen-tation is shown in Fig. 1. Radiation from a tungsten halogen lamp (12 V 55 W) was modulated using a rotating sector and is passed through a bifurcated fibre-optic. The fibre-optic employed was a bundle containing 16 polymer fibres (Quan-tum Jump Ltd. Liverpool) with a total diameter of ca. 1 mm. The sensing end of the fibre-optic was located in the flow-cell (Fig.2). Solutions of known pH were passed through the flow-cell at a constant rate using a peristaltic pump (Cole-Palmer Instrument Co. Model 7567-10). The reflectance signal through the fibre-optic was measured at a given wavelength using a grating monochromator (Varian Techtron AA4 Monochromator Australia) and a photomultiplier tube (Hamamatsu R213). The reflectance spectra were recorded on a chart recorder (JJ X - Y plotter Model PL51). The pH of Source Rotating sector t t Reservoir chromator t 1 I I Reservoir parameter recorder (P.M.T.) meter con t ro I Fig. 1. flow-cell Schematic diagram of instrumentation employed with the t lmmobilised ii disc 7dicator t Direction of flow Fig.2. Flow-cell assembly with immobilised indicator dye and fibre-opti 16 ANALYST. JANUARY 1984. VOL. 109 the test solutions was measured with a Model ECM201 pH meter (EDT Research London). The pH of the medium was altered by the addition of suitable aliquots of 1 M hydrochloric acid and sodium hydroxide solutions. Measurements were made by attaching the input and the reflectance output arms of the bifurcated fibre-optic to the source lens housing and monochromator respectively by means of light-tight fittings. Reagents Indicator dyes and chemical reagents used for the preparation of buffer solutions were purchased from BDH Chemicals Ltd. and XAD-2 polymer beads (a cross-linked polymer of styrene and divinylbenzene) also obtained from BDH Chem-icals Ltd. were used as the polymer support.Analytical-reagent grade methanol was employed. Immobilisation Procedure XAD-2 polymer beads were washed thoroughly with distilled water and then with acetone. The polymer was dried and stored. About 1 g of the polymer beads was placed in 10 ml of 0.1% indicator dye solution in methanol and left to stand for 4 h. The polymer with adsorbed indicator dye was washed with distilled water and then placed in the flow-cell using the pump. The polymer beads surrounded the fibre-optic in the flow-cell and were packed tightly so that they did not move. A constant flow-rate of liquid through the flow-cell containing immobi-lised indicator dye was maintained throughout the experi-ments; the flow-rate employed was 0.10 1 min-1. Reflectance Measurements The optimum wavelengths for measurements of the reflec-tance signal from the immobilised indicator dye may be determined from the reflectance spectra obtained.Fig. 3 shows the reflectance spectra recorded from the immobilised bromothymol blue indicator in the flow-cell at different pH values. The wavelength at which reflectance measurements were made was chosen as that wavelength at which a large change in the reflectance signal with pH was observed. Accordingly a wavelength of 593 nm was used for measuring reflectance signal from the immobilised bromothymol blue indicator. It was also found that this wavelength (593 nm) was also suitable for measurement of the reflectance signal for the other immobilised indicator dyes studied. The reflectance signals reported here are expressed as a percentage of reflectance and no reference wavelength to provide a ratio-metric measurement was employed.Results and Discussion Response Versus pH The reflectance signal measured from the six indicator dyes examined after immobilisation on the cross-lin ked XAD-2 polymer beads varies with pH as shown in Fig. 4. The approximate linear pH ranges and estimated indicator con-stants (pKI,) of the immobilised indicator dyes are presented in Table 1. It is noted that the indicator dyes studied all gave a linear response within a region of ca. 2 pH units. The response was observed to be reproducible rapid and reversible with changes in pH. It was observed that all the dyes except phenolphthalein were firmly retained in the polymer matrix and were not extracted by the aqueous solutions employed.The indicator phenolphthalein was extracted from the poly-mer matrix by alkaline solutions (methyl orange indicator dye was similarly extracted from the polymer matrix by acid solutions) and would not be suitable for use as a durable pH sensor in the immobilised state. Effect of Temperature and Ionic Strength The temperature coefficient of the pH response of immobili-sed bromothymol blue and thymolphthalein in the flow-cell was measured by recording the response of the indicator dyes while varying the temperature of the buffer solution between 25 and 45 "C. The temperature coefficient of the immobilised bromothymol blue and thymolphthalein expressed as change Table 1. Approximate linear pH ranges and estimated indicator constants of immobilised indicator dyes Apparent indicator constant Bromophenol blue .. 3.0-5.0 (3.0-4.6) 3.7 (4.1) Bromocresol purple . . 5.0-7.0 (5.2-6.8) 5.8 (6.1) Bromothymol blue . . 7.0-9.0 (6.0-7.6) 7.7 (7.1) Thymol blue . . . . 9.0-1 1 .O (8.0-9.6) 9.6 (8.9) Phenolphthalein . . . . 9.6-11.0 (8.3-10.0) 10.2 (9.6) Thymolphthalein . . . . 10.5-12.0 (8.310.5) ll.0(9.3) Indicator pH range* (PK,") * 9 t * Values in parentheses are from reference 9 in aqueous solution. 1. Approximate pKrn from this work. 100 350 500 650 Wavelengthlnm 800 0 -I-.- I I 1 I 2 4 6 8 10 12 PH Fig. 3 Reflectance spectra obtained from immobilised bromothymol blue in the flow-cell at a pH of (A) 2.77 (B) 7.07 (C) 8.19 (D) 9.18 and (E) 11.17 Fig.4. Reflectance signal as a function of pH for the immobilised (1) bromophenol blue (2) bromocresol purple (3) bromothymol blue, (4) thymol blue (5) phenolphthalein and (6) thymolphthalei ANALYST JANUARY 1984 VOL. 109 17 x A O B o c I I 1-I 7.0 8.0 9.0 10.0 PH Fig. 5. Effect of (A) 0.1 M (B) 0.5 M and ( C ) 1 .O M sodium chloride in the buffer solution on the response (reflectance vs. pH) for immobilised bromothymol blue in the flow-cell of pH per “C temperature change was found to be 0.013 k 0.003 and 0.015 k 0.003 respectively over this temperature range. The effect of variation of ionic strength on the response of the indicator dye bromothymol blue was investigated using a series of buffer solutions containing 0.1,0.5 and 1.0 M sodium chloride to vary the ionic strength; it was observed that pH response of the bromothymol blue indicator was independent of variation in ionic strength over this range as shown in Fig.5. Conclusions This study demonstrates the potential feasibility of chemical pH sensors based on immobilised indicator dye reagents. The indicator dyes studied present some useful pH ranges for measurements. The response characteristics yield rapid, reversible sensors that are independent of variations in ionic strength of sample solutions. Using the results of this study, we have developed a miniature self-contained fibre-optic pH probe for use in static or flowing solution sample streams. The fibre-optic probe has important advantages over an electrode in safety reliability applicability and cost. The performance characteristics of the fibre-optic probe are at present under investigation and will be reported in a later paper. 1. 2. 3. 4. 5. 6. 7. 8. 9. References Peterson J. I. Goldstein S. R. Fitzgerald R. V. and Buckhold D. K . Anal. Chem. 1980 52 864. Goldstein S. R. Peterson J. I. and Fitzgerald R. V. 1. Biomech. Eng. 1980 102 141. Peterson J. I . and Goldstein S. R. Diabetes Care 1982 5 , 272. Smock P. L. Orofino T. A. Wooten G. W. and Spencer, W. S . Anal. Chem. 1979 51 505. Giuliani J. F. Wohltjen H. and Jarvis N. L. Opt. Lett., 1983 8 54. Saari L. A. and Seitz W. R. Anal. Chem. 1982,54 821. Saari L. A. and Seitz W. R. Anal. Chem. 1983 55 667. Schultz J. S. Mansouri S. and Goldstein I. J. Diabetes Care 1982,5,245. Vogel A. I . “A Textbook of Quantitative Analysis,” Long-mans London 1979 p. 241. Paper A31248 Received August 8th 1983 Accepted September 23rd 198