Fresenius J Anal Chem (1998) 360 : 755–758
© Springer-Verlag 1998
CONFERENCE CONTRIBUTION
A. Trost · M. Borchardt · K. Cammann · C. Dumschat
Sample preparation system for ion-selective electrodes
Received: 4 August 1997 / Revised: 13 October 1997 / Accepted: 21 October 1997
Abstract The sample preparation of real world water samples for potentiometric determinations is often time and labor consuming. Therefore sample preparation systems were developed for the determination of water hardness and nitrate concentrations in drinking water and well water. The systems allow the adjustment of the ionic strength and removal of interfering ions. The sample preparation agents required are fixed by using a disposable sample preparation disk. The system was tested by analyzing real water samples and the results were compared to those obtained by using reference methods.
Introduction Two facts complicate the analysis using potentiometric ion-selective electrodes (ISE). Firstly, because of their repeated employment commercially available electrodes often need a time and labor consuming maintenance. Secondly, no ISE is known which allows the determination of solely one ion species. Therefore, in most cases a sample preparation step prior to the quantification is needed to remove interfering ions. Besides, the ionic strength has to be adjusted because of its great influence on the measurement of ion activity. The first problem was solved by using disposable electrodes. Using the patented Double Matrix Membranetechnology (DMM-technology) disposable electrodes have been recently developed [1]. The special merits of theses electrodes are their extreme low costs and their good reliability in potentiometric measurements. Unfortunately the second problem, the influence of interfering ions, is not solved yet. Many efforts have been made to achieve improved selectivity coefficients [2–4]. Nevertheless, especially the various matrices of environmental samples badly influence the determination em-
A. Trost (Y) · M. Borchardt · K. Cammann · C. Dumschat ICB – Institut für Chemo- und Biosensorik, Mendelstraße 7, D-48149 Münster, Germany
ploying potentiometric ISE. Therefore a sample preparation is essential to obtain highly reproducible results. In this paper we present a simple and reliable sample preparation method, which distinctly reduces time and labor. In the case of two important analytes, i.e. water hardness and nitrate, sample preparation systems have been developed, which allow the adjustment of the ion strength and simultaneously the removal of interfering ions.
Experimental Sensors. ISEs were constructed using DMM-technology, which has been described before [5, 6]. The ingredients of the ion-selective polymer membranes are given in Table 1. As reference electrode a disposable Ag/AgCl-electrode is used. The structure of the electrodes is outlined in Fig. 1. To allow the attachment of the sample preparation system to the sensor element a side by side arrangement of the reference and the ion-selective electrode was chosen. Before the sensors were calibrated they were conditioned for 12 h in a 0.1 mol/L solution of the measuring ion. Selectivity coefficients, slopes and detection limits were determined according to IUPAC recommendations [7, 8]. Table 1 Contents of the used ion-selective membranes (% (W/W); total amount: 300 mg) Water hardness membrane [12, 13] Ionophore 1% Magnesium Ionophore II ETH 4030 (FLUKA, Germany) Additive 70 mol% Potassium tetrakis-(4-chlorophenyl) borate (FLUKA, Germany) Plasticizer 33% Chloroparaffin (FLUKA, Germany) 33% Nitrophenyloctyl ether Polymer 33% Poly(vinyl chloride) high molecular weight (FLUKA, Germany) Solvent 2 mL Tetrahydrofuran (MERCK, Germany) 1 mL Cyclohexanone (MERCK, Germany)
Nitrate-selective membrane [12] 6% Tridodecylammonium nitrate (FLUKA, Germany)
65% 2-Nitrophenyloctyl ether (FLUKA, Germany) 29% Poly(vinyl chloride) high molecular weight (FLUKA, Germany) 2 mL Tetrahydrofuran (MERCK, Germany) 1 mL Cyclohexanone (MERCK, Germany)
756 Analysis of drinking water and well water samples. Analysis is performed according to the following procedure. First the sensor has to be calibrated followed by the attachment of the sample preparation system using a holding device, which is located on the sensor element. A 1 mL aliquot of the real world sample is applied into the sample chamber of the sample preparation system. While the sample is passing the PV-DISK by gravity flow all required reactions take place. The ionic strength is adjusted and interfering ions are removed. For the water hardness determination the PVdisk was made out of Schwarzband filter paper (Schleicher und Schuell, Germany). As sample preparation agent 50 mg tris(hydroxymethyl)aminomethane (Fluka, Germany) and 15 mg KCl (Fluka, Germany) were used. In the case of sample preparation for nitrate determination 6 mg AgF was used for ionic strength adjustment and for the removal of chloride. The AgF supply was fixed between Blauband filter paper (Schleicher und Schuell, Germany). The prepared sample enters the measuring chamber, and the potentiometric determination can be performed. Fig. 1 Schematic view of the nitrate selective electrode, the reference electrode and the sensor arrangement
Fig. 2 Schematic view of the sample preparation disk (PV-disk)
Reference method for water hardness determination. According to the German standard instruction DIN 38406 total water hardness of real world samples was determined by complexometric titration using EDTA (Na-ethylenediaminetetraacetic acid, Fluka Germany). For the determination of the permanent water hardness the sample was brought to boil, filtered and the loss of water was refilled using double-distilled water. The contents of calcium and magnesium ions were obtained according to DIN 38406. Reference method for nitrate determination. Nitrate concentrations were determined using Dionex Ion-chromatography DX 300 (Germany) equipped with Dionex advanced gradient pump, autosampler (Spark Holland Basic Marathon) and PED (Puls Electrochemical Detector, Dionex). Separation was achieved on column IONPAC AS12A (4 mm) by using a Dionex Anionmicromembransuppressor (supressor solution: 12.5 mmol/L H2SO4). As eluent a solution consisting of 2.7 mmol/L Na2CO3 and 0.3 mmol/L NaHCO3 was used (flow: 1.5 mL/min); run time: 13 min.
Results and discussion
Fig. 3 Schematic view of the sample preparation system (IPS)
Sample preparation system. The heart of the sample preparation system (Integriertes Probenvorbereitungssystem IPS) is the sample preparation disk (Probenvorbereitungs-Disk, PV-DISK), shown in Fig. 2. The sample preparation agents are deposited between two filter papers, which are closed using heating sealing film rings. For each application the composition of sample preparation chemicals was optimized with respect to ionic strength adjustment and removal of interfering ions. The PV-disk is fastened in front of the sensor by using the sample preparation system shown in Fig. 3. This system can be used for several times and for different sample preparations. Only the PV-disk needs to be changed after each analysis.
The main topic of interest was to enable real world samples with their extreme varying matrices to be determined easily and rapidly by using potentiometric test strips. First the sensors were tested in real world samples in a batch measurement procedure using standard sample preparation methods. Prior to measurements of water hardness, the ionic strength of the samples was adjusted by using KCl ([KCl] = 0.05 mol/L) and the pH was brought to a value of 8 by using tris(hydroxymethyl)aminomethane. This pH adjustment is necessary, because of the interfering effects of hydrogenium ions towards the determination of water hardness. Furthermore, the samples were analyzed using reference methods. The results of both water hardness determinations are shown in Fig. 4. Figure 4a shows that the use of sensors yields to lower water hardness values in all samples compared to the results obtained by using DIN 38 406 method. These results are not surprising, because ISEs are able to measure only free ions in a solution, whereas the temporary water hardness is bound in a complicated equilibrium with hydrogen carbonate [9, 10]. Therefore this part of the water hardness can not be determined using ion-selective electrodes. To verify this assumption the permanent water hardness of the boiled samples was determined by using the reference
757
water hardness [°d]
a
sample
water hardness [°d]
b
plained. The chloride ions have to be eliminated before a reliable determination can be performed. To achieve this, several methods were investigated with test solutions containing chloride and nitrate in concentrations of 0.1 mol/L. To remove chloride out of the test solution several silver salts were used for precipitation, also the removal by ion-exchangers was investigated. The best results were obtained by adding AgF to the test solution. The ion-chromatographic investigation of the test solution has shown that chloride was completely removed from the solution while no effects on the concentrations of nitrate could be observed. Summing up these investigations lead to the optimal composition of sample preparation agents. For nitrate determination AgF is used as agent for the removal of chloride and, due to its good solubility, also for ion strength adjustment. The water hardness determination was performed by using KCl for ionic strength adjustment and tris(hydroxymethyl)aminomethane for pH adjustment. After these optimization steps, the selected agents were integrated into the developed PV-disk. This disk allows the cheap and simple attachment of the chemicals. The PV-disk consists of two ring-shaped pieces of filter paper which are encapsulated by using two heat sealing film
sample
Fig. 4 a, b Water hardness determination of real world samples after sample preparation (N = 3); a) Determination of the total water hardness b) Determination of the permanent water hardness
Fig. 5 Determination of chloride in real world samples; sensor determination without sample preparation
method and compared to the sensor results (Fig. 4b). As expected these measurements show very good agreement. Prior to the determination of nitrate in the real world samples, the ionic strength was adjusted by using KF ([KF] = 0.05 mol/L). The results are compared to those obtained by ion-chromatography (Fig. 5). In all experiments the sensor measurements resulted in contents higher than those obtained from the reference method. This fact elucidates the necessity to remove interfering ions. The ion-chromatographic investigations have shown that the samples also include chloride and thus the high results when using ion-selective electrodes can be ex-
Fig. 6 Determination of real world samples using sample preparation system (IPS) and sensors compared to results obtained by reference methods (ion-chromatography)
758 Table 2 Main characteristics of the sensors developed (N = 5)
Sensor
Slope
Limit of detection
Selectivity coefficient
Water hardness Nitrate
(27.0 ± 1.0) mV/AD (–57.7 ± 0.3) mV/AD
(1.0 ± 0.1) × 10–5 mol/L (1.0 ± 0.1) × 10–5 mol/L
log K (Ca2+-H+) = 2.1 ± 0.1 log K (NO3–-CI– = –1.6 ± 0.1
rings as shown in Fig. 2. The use of filter paper has several advantages:
Conclusion
• The filter paper is cheap and therefore the disposable use of theses systems can be performed. • The top filter paper leads to an even distribution of the sample solution. Furthermore, it is possible to remove turbidness probably disturbing from the sample solution. • The second layer of filter paper has a very important merit. This paper has the function to remove the appearing precipitates, e.g. the separation of AgCl, out of the sample solution, to avoid interfering effects from these precipitates.
The sample preparation system can completely replace the externe sample pre-treatment of real world drinking water and well water samples for the determination of water hardness and nitrate. The construction of the PV-disk allows a multifarious employment, because every solid sample preparation agent can be integrated. Therefore we consider to employ this technique to solve various problems in sample preparation for potentiometric determinations. Furthermore, this system will be used for other detection method like amperometric and voltammetric measurements.
The PV-disk is integrated in the sample preparation system as shown in Fig. 3. This system can be reused, only the PV-disk needs to be replaced after each analysis. The sample preparation agents were fixed between the filter papers in sufficient supplies. Using solutions which contain measuring and interfering ions in concentrations equivalent to the German standard concentrations, the break- through volumes were determined [11]. These investigations have shown that the break-through volume can be expected to be 5 mL. Higher concentrations of interfering ions lead to lower break-through volumes. The sample preparation sensor system was now used to analyze real world sample solutions which had previously been determined in batch procedures. The results are compared to the reference methods as shown in Fig. 6. It is obvious that the results of water hardness and nitrate analysis show very good agreement to the reference methods. The influence of chloride towards the nitrate determination was eliminated completely, also that which is caused by an acidic pH value in water hardness measurement. The analysis time did not exceed 2 min in all cases. The standard deviations of the determinations using the sample preparation sensor system are comparable to those obtained by using batch procedures. Furthermore time and labor consumption was reduced considerably. Above all the determination of water hardness and nitrate in drinking water and well water can now be performed easily and reliably.
Acknowledgements Financial support of the Deutsche Forschungsgemeinschaft is gratefully acknowledged.
References 1. Dumschat C, Borchardt M, Diekmann C, Hepke J, Cammann K, Knoll M (1994) Fresenius J Anal Chem 348:553–555 2. Hu Z, Qi D (1991) Anal Chim Acta 248:177–181 3. Spichiger UE (1993) Electroanalysis 5:739–745 4. Altura BT (1994) Scand J Clin Lab Invest 54 Suppl 217:21–36 5. Diekmann C, Dumschat C, Cammann K, Knoll M (1995) Sensors and Actuators B 24–25:276–278 6. Borchardt M, Diekmann C, Dumschat, Cammann K, Knoll M (1994) Talanta 41:1025–1028 7. Den Boef G, Hulanicki A (1983) Pure & Appl Chem 55: 553–556 8. Giulbault GG (1981) Pure & Appl Chem 53:1907–1912 9. Instruction Manual (1990) ORION Modell 93-92, Divalent Cation Electrode 10. Gmelin Handbook of Inorganic Chemistry (1973) Syst.-Nr. 14, C3 11. Hein H, Schwedt G (1995) Richt- und Grenzwerte Luft Wasser Boden Abfall Chemikalien. Vogel, Würzburg, Germany 12. Selectophore, Ionophores for Ion-selective Electrodes and Optrodes, FLUKA Chemica, Germany 13. Maj-Zurawska M, Rouilly M, Morf WE, Simon W (1989) Anal Chim Acta 218 : 47–56