A disposable single-use optical sensor for potassium determination based on neutral ionophore

Sensors and Actuators B 88 (2003) 217±222

Short technical note

A disposable single-use optical sensor for potassium determination based on neutral ionophore L.F. CapitaÂn-Vallvey*, M.D. FernaÂndez Ramos, Muneer Al-Natsheh Department of Analytical Chemistry, University of Granada, Granada 18071, Spain Received 12 February 2002; received in revised form 30 September 2002; accepted 4 October 2002

Abstract A disposable single-use optical sensor to determine potassium based on an ion-exchange mechanism is described. The test strip is formed by a circular sensing ®lm zone 6 mm in diameter and 4.7 mm in thickness that contains all the reagents necessary to produce a selective response to potassium on a polyester sheet. The sensing zone is formed by a plasticised PVC that incorporates the cation-selective neutral ionophore dibenzo-18-crown-6, lipophilised Nile Blue, and a lipophilic salt. At pH 9.0, the absorbance response of the test strip at 660 nm shows a good correlation with the theoretical behavior. All experimental variables that in¯uence response, especially in terms of selectivity and response time, have been studied. The sensor responded linearly in activities in the range of 0.0125 and 76.8 mM. The detection limit is 0.0125 mM, the reproducibility intermembrane, at a medium level of the range, is 3.4%, as R.S.D. of log aK‡ and the intramembrane, 3.0%. The procedure was applied to the determination of potassium in different human plasma samples, pharmaceutical compounds and seawater samples, validating results against a reference procedure. # 2002 Elsevier Science B.V. All rights reserved. Keywords: Potassium determination; Optical test strip; Single-use optical sensor; Neutral ionophore; Chromoionophore; Clinical analysis

1. Introduction Potassium ion is routinely measured by ion selective electrodes [1] or by ¯ame photometry. Several methods to determine potassium based on sensors have been proposed, such as optical sensors which measure radiation absorption, like those based on hexaester of calix[6]arene as ionophore and ETH 5294 as chromoionohore [2] or on 2,4,6,20 ,40 ,60 hexanitrodiphenylamine covalently bonded to transparent triacetylcellulose ®lm. In the last case, a pre-treatment step is required [3]. Other sensors, however, some of which operate with ®bre optics [4], use ¯uorescence detection [5±10]. Electric measurements such as potenciometric membranes using rifamycin as ionophore [11] or electroactive nickel(II) hexacyanoferrate thin ®lms [12] have also been used. Test methods are specially adapted and integrated analytical methods, both reactions and processes, and consist of a chemical, biochemical or immunological reaction, along with a system for evaluation, that are well ®tted each to *

Corresponding author. Tel.: ‡34-958-24-84-36; fax: ‡34-958-24-33-28. E-mail address: [email protected] (L.F. CapitaÂn-Vallvey).

other. These quick tests are prepared for concrete problems, such as serum or water, and can carried out either in solution or in solid phase [13]. Dry phase reagent strips are very popular in the ®eld of clinical diagnosis and are available for the determination of potassium ion in serum or plasma. Usually the test means are composed of an ionophore and a chromoionophore incorporated into a non-porous, non-polar carrier matrix or by a homogeneous hydrophobic composition containing the ionophore, chromoionophore and a buffer to maintain pH incorporated into a porous carrier matrix. There are a large number of patents for this type of test strip and for commercial systems to determine alkaline ions in body ¯uids, among which it is possible to cite the well-known systems from Miles Inc. based on [14±18] or by Abbott Laboratories [19]. As an antecedent to the method proposed here, a singleuse disposable optical sensor for potassium based on photocrosslinked decyl methacrylate ®lms on a glass support can be cited; although in this case the ®lms need previous conditioning and storage in a TRIS buffer [20]. In this paper, we propose a new method to determine potassium by a single-use disposable optical sensor that measures by transmission and uses polyester to support a microzone that contains the necessary chemistries, except

0925-4005/02/$ ± see front matter # 2002 Elsevier Science B.V. All rights reserved. PII: S 0 9 2 5 - 4 0 0 5 ( 0 2 ) 0 0 3 2 7 - 1

218

L.F. CapitaÂn-Vallvey et al. / Sensors and Actuators B 88 (2003) 217±222

for the buffer, to selectively and quantitatively respond to potassium by modifying the color of the microzone. The test strip for potassium proposed here is based on the use of ionophore and chromoionophore and works by ionexchange mechanism [21]. Of the different ionophores existing for potassium, we have selected the dibenzo-18-crown-6-ether, which was previously used in the design of an ion-selective electrode [22], because it is selective enough for our intended application as well as inexpensive, making it an interesting alternative to the traditionally used valinomycin [9,10,23]. In fact, if we compare the stability constants of the complexes of alkaline and alkaline-earth metals for the above dibenzo18-crown-6-ether, the selectivity sequence found is K > Na > Rb > Cs @ Li and Ba > Sr @ Ca [24], which indicates the selectivity of this ionophore for potassium. 2. Experimental 2.1. Apparatus and software To perform the absorbance measurements, a Perkin-Elmer Lambda 2 (Norwalk, CT, USA) spectrometer interfaced to an IBM SX-486 microcomputer was used. To obtain the absorbance measurements, the sensor was placed along with the corresponding blank sensor into two similar homemade laboratory membrane holders [25]. The acquisition and manipulation of the spectral data was carried out by means of the Pecss software package supplied by Perkin-Elmer, and the later statistical calculations were carried out with the Statgraphics software package (Manugistics Inc. and Statistical Graphics Corporation, USA, 1992), ver.6.0 STSC Inc. Statistical Graphics Corporations, USA, 1993 and Graphmatica for Win 32 ver. 1:60d, 1998 edited by Hertzer and adapted by Garrido. 3. Reagents and materials Potassium chloride stock solution (1000 mg l 1) was prepared in water from potassium chloride (Aldrich, Steinheim, Germany). Solutions of lower concentration were prepared by dilution with water. Stock solutions of 1000 mg l 1 of the following ions were also used: Mg(II), Ca(II), Na(I) and Li(I) as chloride all supplied by Merck (Darmstadt, Germany). pH 9.0 Tris(hydroxymethyl) aminomethane (TRIS) 0.2 M buffer solution supplied by Sigma (Sigma±Aldrich QuõÂmica S.A., Madrid, Spain). For preparing the optode ®lms poly(vinylchloride) (PVC; high molecular weight), dibenzo18-crown-6-ether, tributylphosphate (TBP), Tris(2-ethylhexyl)phosphate (TEHP), 2nitrophenyloctylether (NPOE), bis(2-ethylhexyl) sebacate (DOS) and tetrahydrofuran (THF) were purchased from Sigma and sodium tetrakis [3,5-bis(tri¯uoromethyl)phenyl]

borate (NaTm(CF3)PB) and potassium tetrakis (4-chlorophenyl)borate (KTpClPB) were purchased from Fluka (Fluka, Madrid, Spain). The lypophilised Nile Blue (1,2benzo-7-(diethylamino)-3-(octadecanoylimino) phenoxazine) was synthesized, puri®ed and identi®ed by us according to [22]. Sheets of Mylar type polyester (Goodfellow, Cambridge, UK) were used as support. All chemicals used were of analytical-reagent grade. Reverse-osmosis type quality water (Milli-RO 12 plus Milli-Q station from Millipore) was used throughout. Human serum samples were from the Virgen de las Nieves Hospital in Granada, (Spain) and the Potasion1 600 pharmaceutical product from Synthelabo Pharma, S.A., Madrid (Spain). 3.1. Membrane preparation Mixtures for the preparation of potassium-sensitive membranes were made from a batch of 26.0 mg (28.0 wt.%) of PVC, 63.0 mg (68.5 wt.%) of TBP, 0.8 mg (0.87 wt.%) of dibenzo-18-crown-6-ether, 1.3 mg (1.4 wt.%) of lypophilised Nile Blue and 1.1 mg (1.2 wt.%) of KTPClPB. The membrane components were dissolved in 1.5 ml of freshly distilled THF. Using a homemade spin-on device [26], the single-use membranes were cast by placing 20 ml of the mixture on a 14 mm  40 mm  0:5 mm thick polyester sheet. The physical characteristics of the sensing zone were as follows: solid and homogeneous 6 mm 1 circular ®lm, transparent red color, well adhered to the solid support. The resulting sensing layer was calculated to have a thickness of about 4.7 mm. The concentration of the ionophore, chromoionophore and anionic sites in the dry, thin ®lm that was obtained was calculated to be 24.1 mmol kg 1 in all components. 3.2. Procedure for samples and standards An aliquot of aqueous standard solution containing between 0.0131 and 100 mM (0.0125 and 76.8 mM in activities) of potassium was placed in a 50 ml ¯ask and 5 ml of pH 9.0 TRIS 0.2 M buffer solution was added with water used to raise the solution to the mark. Then 10 ml of the above solution was placed in a 10 cm  1:5 cm polyethylene plastic tube and a disposable sensor was introduced for 5 min into the tube without shaking. Next, the membrane was pulled out of the solution, wiped to remove any solution droplets and its absorbance measured at 660 nm. In the case of the water samples, 10 ml of water was introduced in a polyethylene tube together with 1 ml of pH 9.0, 0.2 M TRIS buffer operating as described above. 3.3. Calculations The Ke value to obtain an optimal ®t of the experimental data was calculated according to the methodology previously proposed by us [27], according to which we take values in the maximum slope zone of the calibration

L.F. CapitaÂn-Vallvey et al. / Sensors and Actuators B 88 (2003) 217±222

curves, which provides us with more information and less error. To obtain the degree of protonation 1 a of chromoionophore from the spectra we use the maximum and minimum absorbance values (AHC‡ and AC) from the spectra of equilibrated test strips with 10 2 M HCl and 10 2 M sodium hydroxide, respectively. But since the chromoionophore cannot be fully protonated (1 a  0:9) at the working pH (9.0), we de®ne an effective a value, aeff, in order to use a measurement which gives us the degree of protonation with respect to the pH 9.0 buffer solutions. Thus, we calculate the maximum and minimum absorbance values, Abuffer and AC, equilibrating sensors with 0.2 M TRIS buffer and 10 2 M NaOH. Activities were calculated according to the Debye HuÈckel formalism [28]. 4. Results and discussion This single-use sensor is based on a cation exchange system between a complete organic phase of plasticized PVC membrane and the aqueous phase as described in [30]. It should be brie¯y noted (see [27]) that the membrane contains the K‡ selective ionophore L, the chromoionophore C selective for the hydrogen ions, giving both the positively ‡ charged species in the membrane phase, KL‡ p and HC , where p is the stoichiometry of the complexes formed. An alkaline salt of a highly lipophilic anion R is also incorporated into the membrane for reasons of electroneutrality. In the sensing zone of the test strip in contact with an aqueous solution containing potassium ions, the following ion-exchange equilibrium holds, if the ionophore is assumed to form a 1:1 complex with potassium: ‡ pL ‡ HC‡ ‡ R ‡ K‡ $ C ‡ KL‡ p ‡R ‡H

resulting in a mass transfer of potassium ions into the bulk of the organic membrane phase, characterised by the constant Ke. Thus, the activities of K‡ and H‡ in the sample determine their concentrations in the membrane. By using the absorbance of the protonated form of chromoionophore, which is the measurable species in membrane phase, the degree of protonation 1 a is obtained. As we indicated in [27] we used an effective a value, aeff. The ion activities ratio in aqueous phase is related to the equilibrium constant Ke and a through the response function:   aH‡ aeff Ke ‰…Ro …1 aeff †Co †Š aK ‡ ˆ (1) 1 aeff ‰Io p…Ro …1 aeff †Co †Šp where Io, Co and Ro are the analytical concentrations of ionophore, chromoionophore and lipophilic anion, respectively. We ignore the activity coef®cients in the membrane phase. To optically determine potassium activities, the pH value is adjusted with an appropriate buffer solution in order to keep the pH value constant.

219

5. Optimisation of sensor response 5.1. Experimental parameters The parameters that can potentially in¯uence the optical response can be classi®ed in two groups: (a) parameters related to the design and composition of the membrane, and (b) parameters related to the equilibration process between the sample solution and the sensor. 5.1.1. Membrane composition The optimisation of the composition of the sensor is carried out in two different stages: ®rst the optimal proportions of the ionophore, lipophilic anion and cromoionophore are established; to do this, the proportions that result in a higher selectivity in the membrane in the face of interferent species are established. Once the proportions of these components are ®xed, the quantity of PVC and the type and quantity of the plasticiser are optimized, via a study of the kinetic of the process. For the lipophilic anion, we tested NaTm(CF3)PB and KTpClPB observing that the ®rst compound did not provide a response, leading us to select KTpClPB. As a plasticiser we tested TBP, DOS, DOP, NPOE and TEHP. The only response was obtained using NPOE, which was, therefore, selected. Due to the considerable in¯uence that the proportion of lipophilic anions present in the membrane exercise over the selectivity [29], the optimal percentage to reach maximum potassium selectivity was established. To do this, the selectivity coef®cient KKj [27,30] for Li(I), Na(I), Mg(II), and Ca(II) was determined, using the separate solutions method [31]. For this, the quantity of the cromoionophore was kept constant while the molar ratio of lipophilic anion to ionophore was varied from 50 to 150%, because the membrane changes color above 150% and does not respond. Fig. 1 shows that the best selectivity is obtained by using a 1:1 ratio between lipophilic anion to a ligand and Table 1 shows the selectivity coef®cients for this ratio. To optimise the amounts of polymer PVC and plasticizer NPOE, we studied the response time of sensors containing different amounts ranging from 40 to 80 wt.% in NPOE. For these experiments, the sensor was placed on one side of a

Fig. 1. Selectivity (SSM) variation of the sensor as a function of lipophilic anion/ionophore ratio; …a† ˆ K; …b† ˆ Na; …c† ˆ Ca; …d† ˆ Mg; …e† ˆ Li.

220

L.F. CapitaÂn-Vallvey et al. / Sensors and Actuators B 88 (2003) 217±222

Table 1 Selectivity coefficients KKj of potassium test strip for foreign ions at pH 9.0 Foreign cations Na(I) Li(I) Mg(II) Ca(II)

log KKj 1.638 3.340 2.128 2.010

10 mm path-length optical cell, and 5 ml of 4:07  10 2 M in activities of K(I) solution was added. Amounts <58% do not give off a good response. The response times for reaching equilibrium were: 3.5 min for 58%, 2.5 min for 68.5%, 2 min for 69.5% and 1.5 min for 80%. Despite the fact that the rate increased with the amount of NPOE, a percentage of 68.5% was selected as optimal, since higher percentages worsened the physical properties; speci®cally, the membrane is less homogeneous and the drying time increases. These results were corroborated by studying the response time of membranes containing different percentages of PVC, in which it was found that 28% of PVC was the most suitable. 5.2. Reaction parameters The pH, reaction time and potassium activity are the parameters that should in¯uence the sensor response to potassium. To test the in¯uence of pH we studied the response of the sensor to different activities of potassium at different pH values (Fig. 2). We selected 9.0, adjusted with 0.2 M TRIS buffer, as the working pH with the goal of gaining selectivity. The response time, which was studied above, compares favourably with the theoretically expected value [24,27,28] for the diffusion of potassium within a 4.7 mm thick plasticised PVC membrane. The disposable sensor based on dibenzo-18-crown-6ether ionophore responds to potassium activities between 9:51  10 8 and 0.0768 M (10 7 and 0.1 M in concentrations) at pH 9.0 (Fig. 3). In order to check the possible stoichiometry we de®ned the summation of residual squares between experimental points and theoretical curves for different stoichiometries (from Eq. (1)). The results obtained (5:8  10 3 (P ˆ 1); 2:7  10 2 (P ˆ 0:5) and 1:7  10 1 (P ˆ 2)) suggest a 1:1 stoichiometry potassium:ionophore [31], which agrees with the observations in Fig. 3 which

Fig. 2. Effect of pH on the sensor response. (a) pH ˆ 5:5; (b) pH ˆ 6:0; (c) pH ˆ 7:0; (d) pH ˆ 7:8; (e) pH ˆ 8:5; (f) pH ˆ 9:5; (g) pH ˆ 9:0; (h) pH ˆ 10.

Fig. 3. The degree of protonation of the chromoionophore (1 aeff ) as a function of the free potassium activity for (a) P ˆ 0:5; (b) P ˆ 1; (c) P ˆ 2.

show clearly that the experimental points adapt themselves for a value of P ˆ 1. Additionally, the ®t by least-squares of the experimental points in the linear maximum slope zone (seven different concentration levels and nine replicates of each one) of the calibration curve to the mathematical model indicated by Eq. (1) makes it possible for us to calculate for log Ke the value 5.894. 5.3. Analytical parameters As discussed in [28], as a measuring range we used the linear relationship in the middle of the sigmoidal response function de®ned by means a lack-of-®t test and as the detection limit the intersection of the linear calibration function de®ned above and a linear function adjusted in the minimal slope zone (background), as shown in Fig. 4. In order to determine the range and the detection limit, we prepared two series of standards; one in the maximum slope zone (seven standards, nine replicates each one), between 9:51  10 6 and 7:68  10 2 in activities (1  10 5 and 0.1 M in concentration), and another in the minimum slope zone of lower activity (5 standards, 9 replicates each one), between 9:51  10 8 and 9:51  10 6 M in activities (10 7 and 1  10 5 M in concentration). The linearity of both series was tested by applying the lack-of-®t test [32] and the linear functions obtained were: aeff ˆ 0:9820 0:0057 log aK‡ =aH‡

Fig. 4. Calibration plot for free potassium activity.

L.F. CapitaÂn-Vallvey et al. / Sensors and Actuators B 88 (2003) 217±222

and aeff ˆ 1:8803 0:2341 log aK‡ =aH‡ , respectively. The interception of both functions gives us a LD of 0.0125 mM. The upper limit of the measuring range was obtained from the intercept of the linear calibration function with the axis of abscise; the value obtained is 98.8 mM. Thus, the measuring range for potassium determination is between 0.0125 and 98.8 mM, both expressed in activities. However, in practice we prefer to use the value in activities of 76.8 mM (100 mM in concentration) as the upper limit of the measuring range because it is the higher tested value. The precision using the same disposable sensor, expressed as relative standard deviation (R.S.D.), was obtained at three activity levels of potassium, namely 4:73  10 4, 4:58  10 3 and 3:25  10 2 M (5  10 4 , 5  10 3 and 5  10 2 M in concentrations) and 10 replicates of each one, and has values of 2.8, 3.0 and 22.9% for log aK‡ . The repeatability using different sensors was determined as above, working at the same three activity levels of potassium and 10 replicates of each one, and has values of 3.3, 3.4 and 7.7% for log aK‡ . The high R.S.D. value found using the same test strip for high activity levels made us think that the system may not be completely reversible. This is corroborated if we observe the tendency of the signal obtained for each of the 10 measured replicates using the same membrane, which can be expressed through the representation slope of 1 aeff versus the number of replicates. The values found are 3:19  10 3 (5  10 4 M), 7:55  10 3 (5  10 3 M), and 1:29  10 2 (5  10 2 M), which suggests a certain irreversible behaviour. However, this is not a problem since they are used in a disposable way. Table 2 shows these and other analytical parameters.

221

Table 3 Determination of potassium in different types of humam serum, pharmaceutical product and seawater samples using AAS as a reference methoda Matrix

Sensor method (mg/L)

S

Reference method (mg/L)

S

Pval (%)

Serum pool I Serum pool II Serum pool III Pharmaceutical productb Seawater (AlmunÄecar, Granada) Seawater (SalobrenÄa, Granada)

123.0 124.3 108.0 592.5

1.4 2.4 2.0 27.5

126.0 123.5 104.0 589.5

1.4 0.7 2.0 8.5

16.8 69.5 10.5 52.6

347

20.5

378.8

2.5

15.5

286

16.0

288.6

0.6

72.8

a

Thrsee replicate samples in all cases. Potassium cloride 76% with technology MICROCAPS1 from EURAND. Results in mg/capsule. b

6. Application of the method

of human serum from the Virgen de las Nieves Hospital in Granada, pharmaceutical products and seawater samples. With the seawater, it was possible to apply the procedure directly without needing to dilute the sample, since the elevated ionic strength of the sample does not affect the measured signal. Table 3 shows the results obtained using the test strip proposed here with an atomic absorption procedure [32] used as a reference method. Table 3 also includes the mean values from three determinations of each sample, standard deviations of these measurements and the probability value (Pval) of the test used for the comparison of the measurements obtained for both methods. As can be seen, the results obtained for both methods are statistically similar.

In order to assess the usefulness of the proposed method for the determination of potassium, it was applied to samples

7. Conclusion

Table 2 Analytical figures of merit Parameter Intercept Slope Probability level % (lack-of-fit test) Linear range (mM) Detection limit (mM)

Value; S 1.880; 0.09085 0.234; 0.039 29.96 0.0125±76.8 0.0125

Intrasensor: %R.S.D. (mM)a 0.473b 4.58b 32.5b

2.8 3.0 22.9

Intersensor: %R.S.D. (mM)a 0.473b 4.58b 32.5b

3.3 3.4 7.7

a b

R.S.D. for log aK‡ . Activities tested; S: standard deviation.

We have designed an optical test strip based on a neutral ionophore which measures by transmission and offers quite good repeatability and a short response time and which can be considered as an inexpensive alternative for obtaining analytical information. The procedure has suf®cient selectivity for the determination of potassium in samples of human serum, pharmaceutical products and seawater, without previous treatment of the sample. Only needing buffering, this strip offers good results, with an accuracy, precision and cost that make it useful for routine analysis with portable equipment. Acknowledgements The authors are grateful to the Ministerio de Educacion y Cultura, DireccioÂn General de EnsenÄanza Superior (Spain) for ®nancial support (Project No. PB98-1302).

222

L.F. CapitaÂn-Vallvey et al. / Sensors and Actuators B 88 (2003) 217±222

References [1] J. Bobacka, A. Ivaska, A. Lewenstam, Plasticizer-free all-solid-state potassium-selective electrode based on poly(3-octylthiophene) and valinomycin, Anal. Chim. Acta 385 (1999) 195±202. [2] W.H. Chan, A.W.M. Lee, D.W.J. Kwong, W.L. Tam, K.M. Wang, Potassium ion-selective optrode based on calix(6)arenehexaester and application in human serum, Analyst 121 (1996) 531±534. [3] A. Safavi, M. Pakniat, Dipicrylamine-modifiel triacetylcellulose membrane for optical pH and potassium ion measurement, Anal. Chim. Acta 335 (1996) 227±233. [4] I. Koronczi, J. Reichert, G. Heinzmann, H.J. Ache, Development of a submicron optochemical potassium sensor with enhanced stability due to internal reference, Sens. Actuators B 51 (1998) 188±195. [5] H. He, H. Li, G. Mohr, B. Kovacs, T. Werner, O.S. Wolfbeis, Novel type of ion selective fluorosensor based on the inner filter effecct: an optrode for potassium, Anal. Chem. 65 (1993) 123±127. [6] M.R. Shortree, S. Dourado, R. Kopelman, Development of fluorescent optical potassium-selective ion sensor with ratiometric response for intracellular applications, Sens. Actuators B 38 (1997) 8±12. [7] O.S. Wolfbeis, B. Schaffar, Optical sensors: a ion-selective optrode for potasssium, Anal. Chim. Acta 198 (1987) 1±12. [8] C. Krause, T. Werner, C. Huber, I. Klimant, O.S. Wolfbeis, Luminiscence decay time-based determination of potassium ions, Anal. Chem. 70 (1998) 3983±3985. [9] C. Krause, T. Werner, C. Huber, O.S. Wolfbeis, pH insensitive ion selective optode: a coextraction-based sensor for potassions ions, Anal. Chem. 71 (1999) 1544±1548. [10] J.N. Roe, A.S. Verkman, Analyst, Fibre optic sensor for the determination of potassium using fluorescence energy transfer 115 (1990) 353±358. [11] H.S.M. Saad, M.H. Wagiha, A.H. Othman, A novel potassium ion membrane sensor based on rifamycin neutral ionophore, Talanta 44 (1997) 1087±1094. [12] R.J. Mortimer, P.J.S. Barbeira, A.F.B. Sene, N.R. Stradiotto, Potenciometric determination of potassium cations using a nickel(II)hexacyanoferrate modified electrode, Talanta 49 (1999) 271±275. [13] Y. Zolotov, Ann. Chim, Analytical tools for everybody: express test method 87 (1997) 285±295. [14] S.C. Charlton, R.L. Fleming, A.L.Y. Lau, Compound useful in detecting ion and method of preparing it, US Patent 4,540,520 (1985). [15] K.F. Yip, S.C Charlton, Compound useful in detecting ion and method of preparing it, US Patent 4,552,697 (1985). [16] S.C. Charlton, P. Hemmes, A.L.Y. Lau, Ion test means having a hydrophilic carrier matrix, US Patent 4,649,123 (1987). [17] S.C. Charlton, R.L. Fleming, P. Hemmes, A.L.Y. Lau, Unified test means for ion determination, US Patent 4,645,744 (1987). [18] M.L. Gantzer, P. Hemmes, D. Wong, Ion test having a porous carrier matrix, US Patent 4,670,218 (1987). [19] S.-T. Wong, Determination of alkali metals, US Patent No. 4,639,424 (1987). [20] T.M. Ambrose, M.E. Meyerhoff, Optical ion sensing with immobilized thin films of photocrosslinked decylmethacrylate, Anal. Chim. Acta 378 (1999) 119±126.

[21] S.S.S. Tan, P.C. Hauser, K. Hauser, K. Wang, K. Fluri, K. Seiler, B. Rusterholz, G. Suter, M. KruÈttli, U.E. Spichiger, W. Simon, Synthesis and characterization of neutral hydrogen ion-selective chromoionophores for use in bulk optodes, Anal. Chim. Acta 278 (1993) 211± 225. [22] P.C. Pandey, R. Prakash, Polyindole modified potassium ion-sensor using dibenzo-18-crown-6 mediated PVC matrix membrane, Sens. Actuators B 46 (1998) 61±65. [23] S. Komada, J. Arakawa, M. Seyama, T. Osaka, I. Satoh, S. Nakamura, Flow injection anaÂlisis of potassium using an all-solidstate potassium selective electrode detector, Talanta 46 (1998) 1293± 1297. [24] Y.A. Zolotov, Macrocyclic Compounds in Analytical Chemistry, vol. 143, Wiley, New York, 1997, p. 90. [25] L.F. CapitaÂn Vallvey, E. Arroyo, C. Berenguer, M.D. FernaÂndez Ramos, R. Avidad, Single-use optical sensor for the determination of iron in water and white winw, Fresenius J. Anal. Chem. 369 (2001) 139±144. [26] K. Seiler, W. Simon, Theorical aspects of bulk optode membranes, Anal. Chim. Acta 266 (1992) 73±87. [27] L.F. CapitaÂn Vallvey, M.D. FernaÂndez Ramos, P. Alvarez de Cienfuegos GaÂlvez, Optical test strip for calcium determination based on a neutral ionophore, Anal. Chim. Acta 451 (2002) 231±241. [28] P.C. Meier, Two-parameter debye-huÈckel approximation for the evaluation of mean activity coefficients of 109 electrolites, Anal. Chim. Acta 136 (1982) 363±368. [29] W. Zhang, L. Jenny, U. Spichiger, A comparison of neutral Mg(II)selective ionophores in solvent polymeric membranes: complex stoichiometry and lipophicity, Anal. Sci. 16 (2000) 11±18. [30] E. Bakker, P. BuÈhlmann, E. Pretsch, Carrier-based ion selective electrodes and bulk optodes. 1. General characteristics, Chem. Rev. 97 (1997) 3083±3132. [31] Analytical Methods Committee, Analyst, 119 (1994) 2363±2369. [32] M.A.H. Franson (Ed.) Standard Methods for the Examination of Water and Wasterwater, APHA-AWWA-WPCF, 15th ed., 1980, pp. 147.

Biographies L.F. CapitaÂn-Vallvey, PhD in chemistry since 1976, professor in analytical chemistry at the Department of Analytical Chemistry in the University of Granada (Spain). Fields of interest: optical sensors; gas sensors; test strips; hand-held dedicated devices; cultural heritage. M.D. FernaÂndez Ramos, PhD in chemistry since 1997, associate professor at the Department of Analytical Chemistry in the University of Granada (Spain). Fields of interest: disposable single use optical sensors; ionophores; sensors based in molecular imprinted polymers. Muneer Al-Natsheh, BSc in chemistry by Qatar University (Qutar) in 1993, MSc in analytical chemistry by the University of Mousel (Iraq) in 1996. Scholarship given by Agencia EspanÄola de CooperacioÂn con el Mundo Arabe (AECI) for PhD studies in the Department of Analytical Chemistry of the University of Grenada (Spain). Fields of interest: single use optical sensors.