Water cience and Technology, 2003, Vol. 47, No. 2, pp. 135-140 New chalcogenide glass chemical sensors for 2- and dissolved H 2 monitoring M. Miloshova, D. Baltes and E. Bychkov LPCA, UMR CNR 8101, Université du Littoral, F-59140 Dunkerque, France (E-mail: milochov@univ-littoral.fr; bychkov@univ-littoral.fr) Abstract A non-optimised treatment of wastewaters containing organic and biological substances is very often accompanied by an accidental emanation of hydrogen sulphide H 2 and therefore leads (i) to an undesirable odour in the vicinity of water treatment plants, and (ii) to a potential hazard for neighbouring population. Fast, sensitive and reliable monitoring devices are hence of significant importance. Chalcogenide and chalcohalide glasses are new promising membrane materials for detection of heavy metal ions and toxic anions and particularly well adapted for continuous in situ monitoring and industrial process control. In the present paper, we will discuss analytical characteristics of new chalcogenide glass chemical sensors for detection of 2- and dissolved H 2, which allow reliable process control to be carried out at natural ph of wastewaters. Keywords Chemical sensors, direct in situ monitoring of dissolved H 2, 2- ion detection INTRODUCTION Continuous in situ measurements of sulphur species in a broad ph range are extremely important for (i) waste water treatment in order to prevent an accidental release of hydrogen sulphide, (ii) sulphide determination in natural waters for medical purposes (therapeutic mud-bath, etc.), (iii) hydrogen sulphide detection in sea water (pollution control, ocean research). Commercially available sulphide ion sensors with polycrystalline Ag 2 membranes can hardly be used in this operation mode because they exhibit a stable response only in strongly basic solutions (ph 12-14), and therefore a preliminary probe treatment with ph adjustment and anti-oxidising buffer addition is needed to obtain reliable results (Camman and Galster, 1996). Chalcogenide and chalcohalide glasses are new promising membrane materials for detection of heavy metal ions and toxic anions (Vlasov and Bychkov, 1987; Bychkov, 1995; Miloshova et al., 1999). Compared to commercial ion sensors, they exhibit a lower detection limit, better selectivity and enhanced chemical durability and thus are particularly well adapted for continuous in situ monitoring and industrial process control. In addition, a nearly unlimited ability of glasses to be doped or modified provides a unique possibility to improve the sensor performance by optimising the chalcogenide glass membrane composition to a particular species and/or process. In the present contribution, we will discuss analytical characteristics of new chalcogenide glass chemical sensors for detection of 2- and dissolved H 2, which allow reliable process control to be carried out at natural ph of wastewaters. EXPERIMENTAL DETAIL Two kinds of experimental difficulties should be taken into account when working with sulphides in aquatic media: (a) chemical equilibrium between different sulphide species depending on ph, and (b) redox processes. A two-step dissociation of the dissolved hydrogen sulphide is described by the following equations and equilibrium constants (Rodier, 1984):
H 2 (aq) H - + H +, K 1 = 1.1 10-7 (1) H - 2- + H +, K 2 = 1.3 10-13 (2) The respective fractions of H 2 (aq), H - and 2- are given in Fig. 1 as a function of ph. Below ph 5, the dissolved molecular form of hydrogen sulphide is mainly present which also increases a possibility of H 2 release. The dominant species at 7 ph 13 is H -, whereas the 2- ions appear in a significant proportion only above ph 12. It means actually that except for the solutions of strong basicity, the 2- ions represent a very small part (from 10-20 at ph 0 to 10-3 at ph 10) of the total dissolved sulphides. Chemical equilibrium of different sulphide forms 1.0 0.8 H 2 H - 2-0.6 0.4 0.2 0.0 4 6 8 10 12 14 ph Fig. 1. Chemical equilibrium between different sulphide species as a function of ph. All sulphide species are stable only in reducing environment when the redox potential E h < 0. In oxidising medium or in the presence of oxygen, the (II) species are oxidised to elemental sulphur and finally to O 2-4. Therefore, a special care was taken to prevent the sulphide oxidation. Before preparing the calibration solutions, deionised water was boiled and degassed to remove the dissolved oxygen and CO 2. Anti-oxidising buffer solutions containing sodium salicylate and ascorbic acid were used as supporting electrolytes. The prepared concentrated 0.1 M Na 2 stock solution at ph > 13 was stored under argon maximum for one week and then the preparation procedure was repeated. The diluted Na 2 solutions were prepared just before calibration measurements and used within 1-2 h of preparation in order to minimise losses due to aerial oxidation. The following electrochemical cell was used for potentiometric measurements: Ag,AgCl KCl KNO 3 Na 2 calibration solution ulphide sensor, (3) which included commercial silver/silver chloride reference electrode and a set of sulphide ion
sensors of different types. Potentiometric measurements were carried out using a high input impedance voltmeter at a temperature of 22 2 0 C. The concentration range of calibration measurements was typically between 10-6 and 10-2 M Na 2. The selectivity coefficients were determined by the mixed solution method (IUPAC, 1970) using the calibration solutions with a constant interfering ion content (1 and/or 0.1 M) and variable Na 2 concentration. Three types of sulphide ion sensors were studied: (1) chemical sensors with polycrystalline Ag 2 membranes, (2) chalcogenide glass sensors of type A, and (3) chalcogenide glass sensors of type B. The glass membrane compositions will not be disclosed caused by future patent applications. REULT AND DICUION ensitivity Typical response of new chalcogenide glass and commercial Ag 2 membrane sensors in sodium sulphide calibration solutions prepared using the anti-oxidising buffer at neutral ph is shown in Fig. 2. Chalcogenide glass sensors exhibit a stable Nernstian response with the slope of -28 mv/decade in the concentration range between 10-6 and 3 10-3 M Na 2 and then a step-like decrease of the sensor potential by 250 mv with an apparent slope of -700 mv/decade. This sudden change is caused by an abrupt increase of ph in the calibration solutions from ph 5 to ph 11 with increasing Na 2 content (Fig. 3). In contrast, the Ag 2 membrane sulphide sensor shows a Nernstian response only between 3 10-5 and 3 10-3 M Na 2. At lower sodium sulphide concentrations, the response is instable, and the sensor potentials appear to be much higher then expected. -100-200 -300 Ag 2 Nernstian lope -28 mv/decade -500 Type B -600-700 uper-nernstian lope -700 mv/decade Fig. 2. Response of chalcogenide glass and Ag 2 membrane sensors in Na 2 solutions.
ph ph of Na 2 Calibration olutions 11 10 9 8 7 6 5 4 Fig. 3. ph of Na 2 calibration solutions. ulphide ion responses at constant ph are shown in Figs. 4 and 5 for the chalcogenide glass and Ag 2 membrane sensors, respectively. In this case, the response of the sensors is monotonic without any unexpected changes in the sensor potential. One observes a systematic decrease of the potential with increasing ph, which means that the potential-generating ions are indeed the 2- ionic species, first in accordance with the Nernst equation E = const - 2.303RT/2F log a 2-, (4) where the theoretical slope is 2.303RT/z i F = -29.6 mv/decade at 25 0 C for the 2- anions (z i = -2) and also with Eqs. (1) and (2). In the latter case, the increasing basicity is accompanied by an increase in the 2- ion activity a 2-, and thus leads to the decrease of the sensor potential.
E (mv) vs. Ag/AgCl Type B Chalcogenide Glass ensor -200-300 ph 4-500 ph 7-600 ph 11-700 Fig. 4. Response of a type B chalcogenide glass sensor in Na 2 solutions at constant ph. Ag 2 Membrane ensor -300 ph 4-500 ph 7-600 -700 ph 11 Fig. 5. Response of a Ag 2 membrane sensor in Na 2 solutions at constant ph. Chalcogenide glass sensors are extremely sensitive to the 2- ions. The low detection limit, calculated from the experimental data using Eqs. (1) and (2), appears to be lower than 10-17 M 2-. It should also be noted that the Ag 2 membrane sensors exhibit less reproducible and considerably less stable results compared to the chalcogenide glass sensors especially at ph 7.
electivity The developed sensors are highly selective in the presence of usual in wastewater anions. Chloride, nitrate and sulphate ions do not show any notable effect on the 2- response. Typical sensor behaviour in Na 2 /1 M NaCl solutions is shown in Fig. 6. In fact, there is no difference in the sensor sensitivity between pure sodium sulphide and mixed sodium sulphide/sodium chloride solutions. The apparent selectivity coefficient K Na2,Cl -, calculated using the Nicolsky-Eisenman equation (Eisenman, 1967), E = E 0 + RT/z i F ln (a i + K ij a j ), (5) where a i and a j are the activities of primary and interfering ions, respectively, is hence less than 10-6 (the lower is the K ij value, the better is the sensor selectivity). Basically, the real selectivity of the sensors is even much better than the above estimation because the selectivity coefficients K 2-,X z- are several orders of magnitude lower than K Na2,X z-. Type B ensor -200-250 1 M NaCl / Na 2-300 -350 Na 2-450 K Na2,Cl - < 10-6 Fig. 6. Chalcogenide glass sensor response in pure Na 2 and mixed Na 2 /1 M NaCl solutions. ome interfering influence was found for carbonate ions (Fig. 7) (the apparent selectivity coefficient K Na2,CO3 2-10 -3 ), but the observed effect seems to be negligible in real wastewaters with a much lower carbonate concentration.
Type B ensor -550-600 1 M Na 2 CO 3 / Na 2-650 Na 2 (ph 11) K Na2,CO 3 2- = 2 x 10-4 -700 Fig. 7. Chalcogenide glass sensor response in pure Na 2 (ph 11) and mixed Na 2 /1 M Na 2 CO 3 solutions. The apparent selectivity coefficients for the investigated sensors are summarised in Table 1. The results show that the type B chalcogenide glass sensors exhibit a slightly better selectivity compared to the other sensor types. Table 1. Apparent selectivity coefficients K Na2,X z- for the investigated sulphide ion sensors. ensor Cl - NO 3 - electivity coefficients O 4 2- CO 3 2- Ag 2 1 10-5 1 10-5 4 10-5 6 10-4 Type A 2 10-6 2 10-5 1 10-5 1 10-2 Type B < 1 10-6 2 10-5 1 10-5 3 10-4 Direct measurements of the dissolved hydrogen sulphide Chemical equilibrium involving dissolved H 2, H - and 2- species, Eqs. (1) and (2), means that the potentiometric measurements in the following electrochemical cell (Midgley and Torrance, 1984): ph electrode solution 2- sensor (6) will give the activity of dissolved hydrogen sulphide a H2 E = E H2 - (RT/2F) ln a H2, (7) where E H2 is the standard potential of the H 2 formation. Measurements in the Na 2 solutions using electrochemical cell (6) show that the response of chalcogenide glass sensors corresponds to
Eq. (7). The response is linear and stable over the entire composition range at neutral ph (Fig. 8) and suggests promising applications of chalcogenide glass sensors for this type of analysis. Type B ensor -450-500 -550 ph electrode solution 2- sensor Fig. 8. Response of a chalcogenide glass sensor to dissolved H 2 using electrochemical cell (6). CONCLUION Chalcogenide glasses are promising membrane materials for sulphide ion detection in a broad ph range and exhibit (i) better sensitivity, (ii) enhanced selectivity, and (iii) response stability at neutral ph compared to commercial sulphide ion sensors. Chalcogenide glass sulphide sensors can be used for direct measurements of the dissolved hydrogen sulphide in aquatic media. REFERENCE Bychkov E. (1995). pectroscopic studies of chalcogenide glass membranes of chemical sensors: Local structure and ionic response. ensors and Actuators, B27(1-3), 351-359. Camman K. and Galster H. (1996). Das Arbeit mit ionenselektiven Elektroden. pringer, Heidelberg. Eisenman G. (1967). The origin of the glass electrode potential. In: Glass Electrodes for Hydrogen and Other Cations. Principles and Practice, G. Eisenman (ed.), Marcel Dekker, New York, pp. 133-173. IUPAC Recommendations for Nomenclature of Ion-elective Electrodes (1970). Pure Appl. Chem. 48(1), 127. Midgley D. and Torrance K. (1984). Potentiometric Water Analysis. John Wiley & ons, New York. Miloshova M., Bychkov E., Tsegelnik V., trykanov V., Klewe-Nebenius H., Bruns M., Hoffmann
W., Papet P., arradin J., Pradel A. and Ribes M. (1999). Tracer and surface spectroscopy studies of sensitivity mechanism of mercury ion chalcogenide glass sensors. ensors and Actuators, B57(1-3), 171-178. Rodier J. (1984). Analyses de l Eau. Dunod, Paris. Vlasov Yu. G. and Bychkov E. A. (1987). Ion-selective chalcogenide glass electrodes. Ion-elective Electrode Rev., 9(1), 5-93.