A CONDUCTIVE POLYPYRROLE BASED AMMONIUM IONSELECTIVE ELECTRODE

DO PHUC QUAN, CHU XUAN QUANG, LE THE DUAN and PHAM HUNG VIET∗Centre of Environmental Chemistry, Vietnam National University, Hanoi, Vietnam

(∗ author for correspondence, e-mail: cec@fpt.vn)

Abstract. In view of the development of miniaturized sensor arrays, a solid-contact ammoniumion selective electrode has been investigated. A conductive polypyrrole film was electrochemicallydeposited on a glassy carbon surface and used as an internal solid contact layer between the sensingmembrane and solid electrode surface. A systematic evaluation of the important parameters affectingthe electromotive force (emf) response is presented. The performances of this solid-contact sensorwere verified using a batch-mode measurement setup and a wall-jet flow cell system. The designedsensor exhibited excellent selectivity for the primary ion and a linear response over the pNH+

4 range

1–5 with a slope of 56.3 mV decade−1. The sensor has a fast response and is relatively robustness,and was also used to determine ammonium concentrations in natural waters, with promising results.

Keywords: ammonium, conductive polypyrrole, ion selective electrode, solid-contact sensor

1. Introduction

The presence of ammonium ions in environmental samples can indicate the extentof pollution and the eutrophication of natural water (Nigel, 1994; Gerald et al.,1999), and over recent years, the growing importance of controlling the levels ofenvironmental pollutants has increased interest in the development of novel sensorsfor the detection of these ions (Erkang et al., 1997; Magalhaes et al., 1997; Deviteriand Diamond, 1994; Peter et al., 1997). Ion selective electrodes (ISES) offer asimple and useful method for the direct detection of inorganic ammonium ions.They offer great advantages, which include, speed and ease of preparation, simpli-fied procedures, relatively fast response, reasonable selectivity, and a wide lineardynamic range, at a relatively low cost. Furthermore, ISE developments offer thepossibility of sensor miniaturization, based on solid-state ion sensors, which usesolvent polymeric membranes that allow the sensing liquid membrane to be caston the solid electrode surfaces and eliminates the need for an internal solution(Henry, 1987; Lemke and Cammann, 1989).

However, instability is a problem frequently encountered with solid-state ionselective electrodes. It is generally agreed that this is caused by the lack of a stableinternal reference potential at the boundary between the sensing membrane and theinner reference element. Attempts have been made to overcome this problem, not-ably by the development of solid contact membrane sensors, in which, the transfer

Environmental Monitoring and Assessment 70: 153–165, 2001.© 2001 Kluwer Academic Publishers. Printed in the Netherlands.

154 DO PHUC QUAN ET AL.

from ionic to electronic conductivity is provided by a solid contact layer having amixed ionic and electronic conductivity between the inner reference element andthe sensing membrane. Using this approach, a photo-cured ammonium electrode(Peter et al., 1997), and an ammonium-ISFET sensor (Cecilia and Jordi, 1997)based on a graphite-epoxy composite used as the conductive layer between theISFET surface and the PVC membrane, have been fabricated.

Recently, the use of conducting electroactive polymers, such as those incorpor-ating polypyrrole, polythiophene or polyaniline as the active components of am-perometric sensors have attracted much intention. These are based on the system’sability to incorporate and expel ionic species during the switching of the polymerfrom the oxidized to the reduced state (Erkang and Anhua, 1991; Omowunmi andWallace, 1994; Barisci et al., 1997). Further application of conducting polymersin potentiometric sensors involve the incorporation of an organic solvent solubleionophore in the polymer, which then allows the fabrication of wholly solid-statepotentiometric sensors (Pia et al., 1999).

In this study, it was found that adding the polypyrrole film by electropolymer-ization, as solid contact layer, significantly improves the potentiometric stabilityof solid-state potentiometric sensors, which is believed to be the result of a betterdefined interfacial potential between the sensing membrane and the solid electrodecontact. We report upon the determination of ammonium levels in a range of naturalwaters using such a solid-contact ammonium ion selective electrode.

2. Materials and Methods

2.1. MATERIALS

All reagents used were of analytical reagent grade. Standards and buffer solutionswere prepared with Milli-Q water. Pyrrole purchased from Fluka was redistilledunder vacuum prior to use and covered with aluminium foil in the refrigerator toprevent UV degradation.

Nonactin and monactin were kindly donated by Dr. Beat Muller of EAWAG,Switzerland. 2-nitrophenyl-octyl ether (2-NPOE), bis(1-butylpentyl)adipate(BBPA), high molecular weight polyvinyl chloride (PVC), potassium tetrakis(4-chlorophenyl)borate (KT4ClPB) and tetrahydrofuran (THF) were obtained fromFluka (Buchs, Switzerland).

2.2. ELECTRODE PREPARATION

2.2.1. Preparation of a Conventional Internal Electrolyte ElectrodeThe membrane components in Table I (200 mg in total) were dissolved in 6 mL offresh the distilled THF. This solution was placed in a glass ring of 24 mm i.d. rest-ing on a glass plate. After solvent evaporation overnight, the resulting membranewas peeled off the glass mould and discs of 7 mm i.d. were cut out, and mounted

A CONDUCTIVE POLYPYRROLE BASED AMMONIUM ION SELECTIVE ELECTRODE 155

TABLE I

Composition of the ammonium solvent polymeric membrane mixturesprepared in this study

Membrane composition Membranes

AM1 AM2 AM3 AM4

(w./w. %)

Nonactin/monactin 3 3 3 3

2-Nitrophenyl-octyl ether 67 – – –

Dioctyl sebacate – 67 – –

Bis(1-butylpentyl) adipate – – 67 67

Potassium tetrakis(4-chlorophenyl) – – – 1

borate in a 70% molar to the inophore

High molecular weight PVC 30 30 30 30

in Philips IS 561 electrode bodies (Eindhoven, The Netherlands) for electromotiveforce (EMF) measurements in batch-mode setup. A 10 mM solution of NH4Cl wasused as the internal filling solution.

2.2.2. Fabrication of the Solid-Contact Ion Selective ElectrodePolypyrrole (PPy) synthesis and characterization were performed in a conventionalthree electrode system comprising, the 2 mm i.d. glassy carbon disc (GC) workingelectrode (6.1204.110 GC, Metrohm, Switzerland), a platinum wire gauze auxili-ary electrode and an Ag/AgCl (3 M NaCl) reference electrode, against which allpotentials were measured. A laboratory-made microgalvanostat was used for theelectropolymerization. Chronopotentiograms were recorded using this instrumentwith data acquisition system support. Cyclic voltammetric measurements of con-ductive polymer were performed using a PC-controlled system for Voltammetry(Model 757 VA Computrace, Metrohm, Switzerland).

Before polymerization, the surface of the glassy carbon working electrode waspolished on a polishing cloth with alumina slurry (0.05 µm) and then cleaned withdouble distilled water and finally in a water-filled ultrasonic bath for 30 sec. Thepolypyrrole film was prepared by anodic galvanostatic electropolymerization of thepyrrole monomer from aqueous solution (0.5 M) onto the electrode surface. Thecounterion solution used for polymerization contained 1 M of chloride. Solutionwas deoxygenated with nitrogen for 5 min to remove any trace of oxygen fromthe solution, prior to the polymer synthesis. A current density of 2 mA cm−2 for150 sec was used to achieve electropolymerisation.

156 DO PHUC QUAN ET AL.

Scheme 1.

The polymeric membrane-coated modified electrode was a solution of a mixtureof ammonium membrane AM4 solution (30 µL) was applied directly on the top ofthe polypyrrole film, then dried for 2 hr under a gently nitrogen atmosphere, andconditioned overnight in 10 mM of the selected primary ion solution prior to anymeasurement.

2.3. POTENTIOMETRIC MEASUREMENTS

Batch-mode potentiometric measurements were made while stirring at a constantrate and with the electrodes immersed to the same depth in the solution. The calib-ration curve was obtained by a standard addition method, involving the additionsof 10−6 to 10−1 M of the primary ion.

The potentials were measured against a double junction Ag/AgCl referenceelectrode (Orion 90-02-00) using a 692 pH/Ions meter (Metrohm, Switzerland),the accuracy of the potential measurements were +0.1 mV.

Flow injection potentiometric measurement was performed with a simple flowmanifold including a four channel peristaltic pump (Ismatec - Switzerland), a lowpressure six way injection valve (5020 Rheodyne) with 100 µL sample loop and awall-jet flow cell (Metrohm, Switzerland). The carrier eluent in the flow injectionexperiments contained 3 mM of sodium acetate and 2 mM of sodium chlorideincluding 1 µM of ammonium, to provide base line stability.

All measurements were carried out at room temperature.

3. Results and Discussion

3.1. CYCLIC VOLTAMMETRY MEASUREMENTS

A polypyrrole film was deposited on electrode surface by oxidation of monomer

A CONDUCTIVE POLYPYRROLE BASED AMMONIUM ION SELECTIVE ELECTRODE 157

Figure 1. Cyclic voltammetry of a PPy/Cl electrode in 0.1 M of different cation solutions (a) NH+4 ;

(b) K+; (c) Na+; (d) Li+ at scan rate 100 mV sec−1.

from an aqueous solution containing the appropriate counterion as supporting elec-trolyte. Chronopotentiograms were recorded during film growth. Relatively con-stant potential was observed throughout the polymerization, indicating the forma-tion of a conductive polymer layer. Since the polypyrrole film plays the role of thesolid contact layer in the ion selective electrode, the investigation of the incorpor-ation process of ions of interest into the polymer film is of great importance to thedevelopment process.

Counterion injection and release to accompany the redox cycling of electro-polymerized polypyrrole films in electrolytes is well known. Recent studies (Lienet al., 1991; John and Wallace, 1993) of the doping-dedoping process occurring atpolypyrrole have reported that, at least in some instances, two distinct processesoccur, as shown in Scheme 1, one involving anion transport and the other cationtransport. The degree to which each of these processes occur depends upon thenature of the polymer materials used and the mobilities of the corresponding anionand cation in solution and through the polymer.

In the present study, cyclic voltammograms recorded after the electropolymer-ization of the polypyrrole indicated that polymer PPy/Cl was conductive in the

158 DO PHUC QUAN ET AL.

Figure 2. Cyclic voltammetry of a PPy/Cl electrode in 0.1 M of different cation solutions (a) NH+4 ;

(b) Ba2+; (c) Ca2+; (d) Mg2+ at scan rate 100 mV sec−1.

counterion solution. In order to investigate the incorporation of cation into thepolypyrrole film, cyclic voltammogram measurements of the PPy/Cl electrode insolutions of different cations were taken. The most striking feature of these voltam-mograms was the fact that the second reduction wave which occurs at a morenegative potential was present in the univalent cation solutions (Figure 1). Figure 1also indicates that the second reduction response moved in the positive directionwith a decrease in the solvated cation size according to Li+ > Na+ > K+ > NH+

4 .The absence of a second reduction response in the divalent cation solutions (Fig-ure 2) suggests that polymer was not reduced to the extent that occurred in theunivalent cation solutions. This was confirmed by the fact that the subsequentoxidation peaks were smaller in solutions containing the divalent cations.

3.2. POTENTIOMETRIC RESPONSE CHARACTERISTICS OF ELECTRODES

Principle electrode characteristics, such as a Nerntian slope, dynamic linear range,detection limit and selectivity, depend on the composition of the ion selectivemembrane. One of the first tasks in a development of this type involves the determ-

A CONDUCTIVE POLYPYRROLE BASED AMMONIUM ION SELECTIVE ELECTRODE 159

TABLE II

Determination of ammonium in natural water samples using the designedsolid-contact ammonium selective electrode as compared with the ammo-nia gas electrode. Real samples were prefiltered through filter membrane of0.45 µm pore size

Sample place Solid-contact NH+4 NH3 gas electrode

selective liquid (Orion Inc., U.S.A.)

membrane electrode

Ammonium (NH+4 ) in groundwater samples, mg L−1 (SRD)

Phap Van 13.9 (0.88%) 13.4 (0.95%)

Ha Dinh 10.2 (0.87%) 9.5 (0.92%)

Yen Phu 4.7 (0.79%) 4.4 (0.86%)

Tuong Mai 6.9 (0.82%) 6.4 (0.89%)

Ammonium (NH+4 ) in river and ponds water samples, mg L−1 (SRD)

Thanh Tri fish pond 1 3.8 (1.45%) 3.2 (1.49%)

Thanh Tri fish pond 2 5.1 (1.46%) 4.6 (1.52%)

Kim Nguu River 5.5 (1.53%) 5.0 (1.62%)

To Lich River 5.2 (1.51%) 4.8 (1.58%)

Lu River 5.0 (1.48%) 4.3 (1.53%)

ination of the best composition for the preparation of a solid-contact ammoniumion selective electrode. The optimal composition of the ion selective membrane forthe ammonium was determined in several experiments by varying the nature andthe percentage of various liquid membrane plasticizers, results are summarized inTable II.

Three ammonium membrane electrodes employing three different plasticizerswere prepared. The electrode responses observed for the AM1, AM2 and AM3ISEs were both Nernstian in character over range between 0.01 and 100 mM inpure ammonium chloride solutions. However, the AM2 and AM3 ISE gave super-ior response characteristics with a Nernstian slope of 57.6 and 58.3 mV dec−1,respectively. Furthermore, this study also indicated that anion interference at con-centrations >100 mM, for the liquid membrane – based ISE, are more evident for apolar plasticizer, such as NPOE. The severity of the anion interference is also de-pendent on the lipophilicity of the anion, that is, the greater the lipophilicity of theanion, the greater the penetration of the anion into the liquid membrane, and con-sequently, the greater the interference exerted upon the cation selective membraneat higher concentrations. Therefore, in order to reduce anion interference at higherconcentrations, a lipophilic salt, such as potassium tetrakis(4-chlorophenyl) borate,

160 DO PHUC QUAN ET AL.

Figure 3. Charge transfer process occurring in (A) a conventional internal electrolyte selectiveelectrode; and (B) in a solid-contact conducting polymer based electrode.

was added to the membrane. In the following section, we describe a solid-contactsensor that was fabricated with the optimal AM4 mixture composition.

Figure 3 compares the charge transfer process occurring in a conventional in-ternal electrolyte selective electrode with that in a solid-contact conducting poly-mer based electrode. Comparisons of the charge transfer between the sensing mem-brane and inner element of both ammonium electrodes showed similar trends.The potential response of the solid-contact electrode was very stable in the meas-ured ammonium concentration range. Calibration curves for a conventional internalelectrolyte electrode and the solid-contact electrode in ammonium solutions, areshown in Figure 4. This result demonstrates that the response was linear over theinvestigated range of 10−5 to 10−1 M L−1 of ammonium with a slope of 56.3 mVdec−1.

Response time was also investigated, because it is a very important factor interms of the practical use of solid-contact sensors. The time taken for the bothelectrodes to attain 90% of the steady-state response was typically a few seconds,suggesting that this electrode is ideal for flow injection measurement.

A CONDUCTIVE POLYPYRROLE BASED AMMONIUM ION SELECTIVE ELECTRODE 161

Figure 4. Comparison of ammonium ion concentration calibration curves for the internal electro-lyte ammonium ISE and the solid-contact ammonium electrode in 2 mM NaCl, to adjust the ionicstrength.

The selectivity of solvent polymeric membrane electrodes in the presence of in-terfering cations were determined by the separateg solution method (SSM) (Morf,1981) and calculated using Equation (1).

log Kpoti,j = (Ej − E)

S+

(1 − zi

zj

)lg ai , (1)

where Kpoti,j is the selectivity coefficient, i is the primary ion (NH+

4 ), j is the in-terfering ion, E is the measured potential (mV), S is the Nernstian slope factor(mV/decade), z is the electrical charge of the ion, and a is the activity calculatedfrom the activity coefficients. The selectivity coefficient values obtained by theseparate solution method are given in Figure 5 and are compared to the selectivitycoefficients reported for ammonium electrodes in previous studies (Thomas et al.,1988). Comparison of the selectivity coefficients of interfering ions showed thatthe potassium ion is the most notable interference ion of the ions tested.

The solid-contact ammonium ion selective electrode AM4SCS was mounted inthe FIA system described in the Experimental Section. Typical FIA signals ob-

162 DO PHUC QUAN ET AL.

Figure 5. Selectivity coefficients, logKpotNH4,j

for the ammonium selective electrodes made during thiswork, compared to a previously reported ammonium ISE. Results were obtained by the separate solu-tion methods in Tris buffered solutions of 0.1 M chloride salts at pH 7.1. AM (1–4): four conventionalinternal electrolyte ammonium selective electrodes; ISE∗: ammonium ion selective microelectrode(Thomas et al., 1988); and AM4SCS: solid-contact ammonium ion selective electrode.

tained for series injection of different ammonium solutions are shown in Figure 6,which demonstrates the high reproducibility of the observed peak heights. Thepotentiometric response was linear over the range investigated i.e., between 10−5

and 10−1 M L−1 of ammonium.

3.3. WATER ANALYSIS

Finally the analytical performance of the solid-contact ammonium ion selectiveelectrode was tested in terms of the determination of ammonium in natural water.The samples of natural water used were representative complexes of real matricescontaining high concentrations of inorganic and organic substrates. Moreover, theconcentration of ammonium in these samples represented an important qualitylevel determining factor. In order to verify the analytical results obtained using this

A CONDUCTIVE POLYPYRROLE BASED AMMONIUM ION SELECTIVE ELECTRODE 163

Figure 6. Potentiometric responses of solid-contact ammonium selective electrode in an FIA system.Flow injection conditions were as reported in the Experimental Section.

electrode, the same samples were analyzed using an ammonia gas sensor, and theresults obtained using the two techniques were found to be close agreement. Theseresults demonstrate clearly the benefits of the features of the ammonium solid-contact electrode in a flow injection system for the determination of ammoniumlevels in real complex matrices.

4. Conclusions

The solid-contact ion selective electrode, based on a conductive polypyrrole mem-brane, showed good sensitivity in the linear pNH+

4 range 1–5 with a Nernstianslope of 56.3 mV dec−1. This sensor exhibited a fast response (5 sec) and relat-ively high robustness in both batch-mode measurements and flow injection sys-tems, using a wall-jet flow cell. It is hoped that this study will contribute to thedesign and construction of an ammonium selective solid-contact electrode based

164 DO PHUC QUAN ET AL.

on integrated planar sensor technology. Further investigations to improve the elec-trode, involving lowering the detection limit and lengthening the life-time of thesolid-contact sensor will be conducted in our laboratory.

The use of this conceptual approach for development of solid-contact selectivesensors for other relevant environmental ions, such as nitrite and potassium remainschallenging work.

Acknowledgements

This work was supported by the Vietnam National University, Hanoi (GrantQGTD.99.02) and partly supported by the SDC (Swiss Agency for Developmentand Cooperation, Switzerland) in framework of the co-operation Project ESTNVbetween CEC, VNU Hanoi and EAWAG, Switzerland in the field of Environ-mental Chemistry. Do Phuc Quan express his deep gratitude to Prof. Dr. GordonG. Wallace, Director of the Intelligent Polymer Research Institute (University ofWollongong, Australia) for the effective support during his former academic stayin the Gordon G. Wallace’s Laboratory and for many challenging discussions.The authors also express their thankfulness to thank Dr. Beat Müller (EAWAG,Switzerland) for his kind donation of the nonactin/monactin ionophore.

References

Barisci, J. N., Wallace, G. G. and Clarke, A.: 1997, ‘Amperometric detection of electroinactive anionsusing conducting polymer electrodes subsequent to chromatographic separation’, Electroanalysis9(6), 461.

Beck, F. and Oberst, M.: 1987, ‘Electrodeposition and cycling of polypyrrole’, Markromol. Chem.Macromol. Symp. 8, 97–125.

Buhrer, T., Gehrig, P. and Simon, W.: 1988, ‘Neutral-carrier-based ion-selective microelectrodesdesign and application. A review’, Anal. Sci. 4, December.

Jimenez, C. and Bartoli, J.: 1997, ‘Development of an ion-sensitive field effect transistor based onPVC membrane technology with improved long-term stability’, Electroanalysis 7(4), 316.

Deviteri, F. J. S. and Diamond, D.: 1994, ‘Ammonium detection using an ion-selective electrodearray in flow-injection’, Electroanalysis 6(1), 9.

Erkang, W., Zhu, L., Ma, L. and Patel, H.: 1997, ‘Optical sensors for sodium, potassium and am-monium ions based on lipophilic fluorescein anionic dye and neutral carriers’, Anal. ChimicaActa 357(1–2), 85–90.

Erkang, W. and Anhua, L.: 1991, ‘Polyaniline chemically modified electrode for detection of anionsin flow injection analysis and ion chromatography’, Anal. Chimica Acta 252, 53.

Freiser, H.: 1987, ‘Coated wire ion selective electrodes and their application to environmentalproblems’, Pure and Appl. Chem. 59(4), 539–544.

John, R. and Wallace, G. G.: 1993, ‘Doping-dedoping of polypyrrole: a study using current-measuring and resistance-measuring techniques’, J. Electroanal. Chem. 354, 145–160.

Gardner, W. and Bartlett, P. N.: 1995, ‘Application of conducting polymer technology in microsys-tems’, Sensors and Actuators A 51, 57–66.

Lemke, U. and Cammann, K.: 1989, ‘Coated film electrodes’, Fresenius Z Anal. Chem. 335, 852–854.

A CONDUCTIVE POLYPYRROLE BASED AMMONIUM ION SELECTIVE ELECTRODE 165

Lien, M., Smyrl, W. H. and Morita, M.: 1991, ‘Cation and anion insertion in separate processes inpoly(pyrrole) composite films’, J. Electroanal. Chem. 309, 333–340.

Magalhaes, J. M. C. S., Cespedes, F., Alegret, S. and Machado, A. A. S. C.: 1997, ‘Study of thetemperature behaviour of all-solid-state nonactin ammonium electrodes with PVC membraneapplied to graphite-epoxy supports of varied composition’, Anal. Chimica Acta 355(2–3), 241–247.

Morf, W. E.: 1981, The Principles of Ion-Selective Electrode and Membrane Transport, Elsevier,Amsterdam.

Nigel, B.: 1994, Environmental Chemistry, Wuerz Publishing, Winnipeg, pp. 221.Omowunmi, A. S. and Wallace, G. G.: 1994, ‘Detection of electroinactive ions using conducting

polymer microelectrodes’, Electroanalysis 6, 860.Alexander, P. W., Dimitrakopolos, T. and Brynn Hibbert, D.: 1997, ‘Photo-cured ammonium and

hydrogen ion selective coated-wire electrodes used simultaneously in a portable battery-poweredflow injection analyzer’, Electroanalysis 9(17), 1331–1336.

Schuytema, G. S. and Nebeker, V. A.: 1999, ‘Comparative toxicity of ammonium and nitrate com-pounds to pacific treefrog and African clawed frog tadpoles’, Environ. Toxicol. and Chem. 18(10),2251–2257.

Sjöberg, P., Bobadca, J., Lewenstam, A. and Ivaska, A.: 1999, ‘All-solid-state chloride-selectiveelectrode based on poly(3-octylthiophene) and tridodecylmethylammonium chloride’, Elec-troanalysis 11(10–11), 821–824.