Anal. Chem. 2000, 72, 956-962
Salicylate-Selective Electrodes Based on Al(III) and Sn(IV) Salophens Said Shahrokhian, Mohammad K. Amini,* Reza Kia, and Shahram Tangestaninejad
Department of Chemistry, University of Isfahan, Isfahan 81744, Iran
New salicylate-selective electrodes based on aluminum(III) and tin(IV) salophens are described. The electrodes were prepared by incorporating the ionophores into plasticized poly(vinyl chloride) (PVC) membranes, which were directly coated on the surface of graphite electrodes. These novel electrodes display high selectivity for salicylate with respect to many common inorganic and organic anions. The influence of membrane composition and pH and the effect of lipophilic cationic and anionic additives on the response properties of the electrodes were investigated. The electrode based on aluminum salophen, with 32% PVC, 65.8% plasticizer, and 2.2% ionophore, shows the best potentiometric response characteristics and displays a linear log [Sal-] vs EMF response over the concentration range 1 × 10-6-1 × 10-1 M in phosphate buffer solutions of pH 7.0, with a Nernstian slope of -59.2 mV/decade of salicylate concentration. Highest selectivity was observed for the membrane incorporating 38.8% PVC, 57.3% plasticizer, 2.6% Sn(salophen), and 1.3% sodium tetraphenylborate. The electrodes exhibit fast response times and micromolar detection limits ( ∼1 × 10-6 M salicylate) and could be used over a wide pH range of 3-8. Applications of the electrodes for determination of salicylate in pharmaceutical preparations and biological samples are reported. Potentiometric detection based on ion-selective electrodes (ISEs), as a simple method, offers great advantages such as speed and ease of preparation and procedures, relatively fast response, reasonable selectivity, wide linear dynamic range, and low cost. These characteristics have inevitably led to sensors for several ionic species, and the list of available electrodes has grown substantially over the past years.1 A very interesting development of potentiometric sensors is in the construction of electrodes that respond selectively to biological compounds. The wide use of ISEs in routine chemical and biochemical analysis has been accompanied by a search for ionophores that can chemically recognize specific ions and offer either new or improved selectivities for different ions. In many instances in the literature, the membranes used in anion-selective sensors are based on quaternary ammonium or phosphonium salts.2-5 Selectivity of these classical ion exchangers depends on (1) Bu ¨ hlmann, P.; Pretsh, E.; Bakker, E. Chem. Rev. 1998, 98, 1593-1687. (2) Wegmann, D.; Weiss, H.; Ammann, D.; Morf, W. E.; Pretsch, E.; Sugahara, K.; Simon, W. Microchim. Acta 1984, 3, 1-16. (3) Arnold, M. A.; Solsky, R. L. Anal. Chem. 1986, 58, 84R-101R.
956 Analytical Chemistry, Vol. 72, No. 5, March 1, 2000
the lipophilicity of the anions, with more lipophilic anions responding the best. One of the most important recognition elements that can be utilized in the development of ISEs involves specific metalligand interactions.6 For the development of a truly anion-selective electrode, a strong interaction between the ionophore and the anions is required in order to complex anions in a selective fashion. Complexes of several metal ions with different ligands, such as phthalocyanines,7,8 porphyrins,9,10 and metallocenes,11 as ionophores for anions, have been observed to show such specific metal-ligand interactions and induce anion selectivities in the membranes that differ somewhat from the classical anion exchangers. Potentiometric response of the membranes doped with these complexes is believed to be based on the coordination of analyte anions as an axial ligand to the metal center of the carrier molecule. The purpose of the present work has been the development of salicylate-selective electrodes based on poly(vinyl chloride) (PVC) membranes of aluminum and tin salophens, coated on the surface of graphite electrodes. Salicylic acid is one of the common metabolites of acetylsalicylic acid (aspirin), which is widely used as an analgesic and inflammatory agent and, recently, also as a preventive of heart attacks.12 It is easily hydrolyzed to salicylic acid, which circulates in blood in the ionized form. Monitoring the plasma salicylate concentration is important for controlling the dose and frequency of aspirin administration. The free acid is widely used as an antiseptic and a preservative for foods. In recent years, the electrochemical properties and preparations of several new salicylate-selective membrane electrodes have been reported by using a variety of ion carriers.13-18 The development of a (4) Yu, R. Q. Ion-Sel. Electrode Rev. 1986, 3, 153-171. (5) Wotring, V. J.; Johnson, D. M.; Bachas, L. G. Anal. Chem. 1990, 62, 2, 1506-1510. (6) Hutchins, R. S.; Bachas, L. G. Anal. Chem. 1995, 67, 1654-1660. (7) Li, J. Z.; Wu, X. C.; Yuan, R.; Lin, H. G.; Yu, R. Q. Analyst 1994, 119, 13631368. (8) Nakamura, T.; Hayashi, C.; Ogawara, T. Bull. Chem. Soc. Jpn. 1996, 69, 1555-1559. (9) Bakker, E.; Malinowska, E.; Schiller, R. D.; Meyerhoff, M. E. Talanta 1994, 41, 881-890. (10) Chaniotakis, N. A.; Park, S. B.; Meyerhoff, M. E. Anal. Chem. 1989, 61, 566-570. (11) Hisamoto, H.; Siswanta, D.; Nishihara, H.; Suzuki, K. Anal. Chim. Acta 1995, 304, 171-176. (12) Reiman, A. N. Engl. J. Med. 1988, 318, 245-246. (13) Chang, Q.; Meyerhoff, M. E. Anal. Chim. Acta 1986, 186, 81-90. (14) Hassan, S. S. M.; Hamada, M. A. Analyst 1988, 113, 1709-1713. (15) Hutchins, R. S.; Bansal, P.; Molina, P.; Alajarin, M.; Vidal, A.; Bachas, L. G. Anal. Chem. 1997, 62, 1273-1278. (16) Li, Z. Q.; Song, X. P.; Shen, G. L.; Yu, R. Q. Anal. Lett. 1998, 31, 14731486. 10.1021/ac990749w CCC: $19.00
© 2000 American Chemical Society Published on Web 02/03/2000
suitable salicylate-selective sensor, would enable the detection of “free” salicylate in samples and could aid researchers in studying the pharmacological role of this drug.10 The proposed membrane electrodes based on metallosalophens display a low detection limit and high selectivity and sensitivity to salicylate determination and is promising for measurement of salicylate in biological samples. EXPERIMENTAL SECTION Reagents. Poly(vinyl chloride) of high relative molecular weight, dioctyl phthalate (DOP), trioctylmethylammonium chloride (TOMACl), and sodium tetraphenylborate (NaTPB) were used as received from Aldrich. Tetrahydrofuran (THF) and all other chemicals were of the highest purity available from Merck and were used without further purification, except THF, which was distilled before use. All aqueous solutions were prepared with deionized, distilled water. The complexes, N,N′-bis(salicylidene)-1,2-phenylenediaminoaluminum(III) nitrate [Al(salophen)], and its corresponding tin(IV) chloride [Sn(salophen)] (general structural formula I), were
I
synthesized by modification of the reported procedures,19-22 as follows: 1,2-Phenylenediamine, 0.54 g (5.0 mmol), was dissolved in 25 mL of ethanol and then transferred into a 250-mL threenecked flask. Under reflux, 1.22 g (10.0 mmol) of salicylaldehyde in 25 mL of ethanol was added dropwise to the flask. The stirred mixture was kept reacting for 45 min, under reflux, and then cooled to room temperature. The solid mass was filtered, and the product, N,N′-bis(salicylidene)-1,2-phenylenediamine (salophenH2), was recrystallized from ethanol and vacuum-dried for 12 h: 1H NMR δ 6.8-7.6 (m, 12H, phenyl), 8.8 (s, 2H, -CHdN-), 13.0 (s, 2H, OH); IR (ν, cm-1) 1620 s, 1590 s, 1550 s, 1475 s, 1450 s, 1290 s, 1210 m, 1150 m, 1120 w, 1020 m, 930 m, 830 m, 750 s, and a very broad band at 3000 cm-1. To a 1.58-g (5.0 mmol) sample of salophenH2 in 50 mL of ethanol was added 1.13 g (5.0 mmol) of SnCl2‚2H2O or 1.88 g (5.0 mmol) of Al(NO3)3‚9H2O in 30 mL of ethanol, drop by drop. After being stirred for 1 h at room temperature, the mixture was refluxed for 2 h and cooled to room temperature, and the bright yellow Al(salophen) or the orange product of Sn(salophen) was filtered, washed with ethanol and hot water several times, and dried under vacuum at 100 °C. There is no obvious frequency corresponding to ν(OH) in the IR spectra of the complexes as in the case of the free ligand (17) Li, J. Z.; Pang, X. Y.; Gao, D.; Yu, R. Q. Talanta 1995, 42, 1775-1781. (18) Katsu, T.; Mori, Y. Talanta 1996, 43, 755-759. (19) Atwood, D. A.; Jegier, J. A.; Rutherford, D. Inorg. Chem. 1996, 35, 63-70. (20) Mao, L.; Tian, Y.; Shi, G.; Liu, H.; Jin, L.; Yamamoto, K.; Tao, S.; Jin, J. Anal. Lett. 1998, 31, 1991-2007. (21) Marzilli, L. G.; Marzilli, P. A.; Halpern, J. J. Am. Chem. Soc. 1971, 93, 13741378. (22) Chen, D.; Martell, A. E. Inorg. Chem. 1987, 26, 1026-1030.
spectrum in which hydrogen bonding occurs. The resonance of OH protons, which appears at δ 13.0 in the free salophenH2 spectrum and is broad owing to intramolecular exchange, is absent in spectra of the complexes. A stock solution of salicylate, 0.1 M, was prepared by dissolving 1.601 g of sodium salicylate in 100 mL of water. Solutions of interferences, 0.1 M, were prepared by dissolving the appropriate amount of each compound, usually their sodium salts, in water. Working solutions were prepared by successive dilutions with water. All of the working solutions were buffered at pH 7.0, using 0.05 M phosphate buffer solution. For the preparation of pharmaceutical samples, tablets of three different samples of aspirin were finely powdered. A precisely weighed portion of each sample was refluxed with ∼50 mL of 0.5 M NaOH for 1 h. After being filtered, the solution was diluted to 250 mL in a volumetric flask and used for the determination of salicylate content by potentiometric and spectrophotometric methods. Preparation of Electrodes. Electrodes were prepared from spectroscopic-grade graphite rods (3 mm diameter and 10 mm long). The graphite rod was sealed into the end of a PVC tube of about the same diameter by epoxy cement. Electrical contact was made by attaching a shielded copper wire, via silver-loaded epoxy cement, into the back of the graphite rod. The working surface of the electrode was polished with fine alumina slurries on a polishing cloth, ultrasonicated in distilled water, and allowed to dry. Membrane solutions were prepared by dissolving varying amounts of the ion active phase (aluminum or tin salophen), together with appropriate amounts of DOP (as plasticizer) and PVC, to give a total mass of 200 mg, in ∼5 mL of THF. NaTPB and TOMACl as lipophilic anionic and cationic additives were also incorporated in some of the mixtures. Membrane compositions are listed in Table 1. The polished graphite electrode was dipped into the membrane solution several times, and the solvent was evaporated each time. A membrane was formed on the graphite surface, and the electrode was allowed to set overnight. Potential Measurements. The electrodes were equilibrated for 24 h in 0.05 M salicylate solution. The potential measurements were carried out at 25 ( 1 °C with a digital pH/mV meter (Jenway model 3305) by setting up the following cell assembly:
Hg, Hg2Cl2(S), KCl (satd) | sample solution | membrane | graphite surface The pH of the sample solutions was monitored simultaneously with a conventional glass pH electrode. Calibration curves were constructed by plotting the potential, E, vs the logarithm of the concentration of salicylate at constant pH. RESULTS AND DISCUSSION Potentiometric Response Characteristics of Electrodes. The potentiometric response characteristics of the plasticized PVCbased membrane electrodes, incorporating Al(salophen) or Sn(salophen) ionophores, toward salicylate, in buffer solutions of pH 7.0, are shown in Figures 1 and 2, respectively. The electrodes based on both ionophores generated stable potential responses in a solution containing salicylate after conditioning for ∼24 h. Analytical Chemistry, Vol. 72, No. 5, March 1, 2000
957
Table 1. Variation of Potentiometric Response Characteristics of Electrodes Based on Aluminum and Tin Salophens with Different Membrane Compositions, Measured in Phosphate Buffer Solutions (pH 7.0) % (w/w) of various components electrode
PVC
DOP
A B C D E F
31.9 32.0 31.8 38.9 38.0 37.9
62.6 65.8 66.8 58.5 58.2 58.2
G H I J K
33.1 32.0 39.2 38.8 38.5
64.8 63.0 58.2 57.3 57.6
Al(salophen) 5.5 2.2 1.4 2.6 2.7 2.9
TOMACl
NaBPh4
slope (mV/decade)
linear range (M) 10-1
detection limit (M)
-56.0 -59.2 -50.2 -58.9
× 1× 1 × 10-6-1 × 10-1 1 × 10-5-1 × 10-1 5 × 10-6-1 × 10-1
1 × 10-6 1 × 10-6 2 × 10-6 2.5 × 10-6
1.0
-58.9
1 × 10-6-1 × 10-1
1 × 10-6
1.1
-59.0 -58.3 -58.7 -57.0 -57.2
1 × 10-5-1 × 10-1 1 × 10-4-1 × 10-1 5 × 10-5-1 × 10-1 1 × 10-5-1 × 10-1 1 × 10-6-1 × 10-1
4 × 10-6 1 × 10-4 3 × 10-6 1 × 10-5 8 × 10-7
1.1
10-6-1
Sn(salophen) 2.1 5.0 2.6 2.6 2.8
1.3
Figure 1. Potentiometric response of different electrodes based on Al(salophen). Compositions of electrodes are summarized in Table 1.
Figure 2. Potentiometric response of different electrodes based on Sn(salophen). Compositions of electrodes are summarized in Table 1.
After such a treatment, a Nernstian response was obtained for some of the electrodes in salicylate solutions of different concentrations, and the slope remained almost constant. Some of the electrodes displayed remarkable selectivity for salicylate over most common anions. It is well known that the sensitivity, linear dynamic range, and selectivity obtained for a given sensor depend significantly on the membrane composition. The effect of membrane composition on the electrode performance was studied by varying the ratio of the membrane active phase, PVC, and DOP. Additionally, we prepared membranes with TOMACl and NaTPB as cationic and anionic additives. Table 1 presents the compositions of several typical membranes, along with their other characteristics. The potential response of all the membrane sensors was studied in the concentration range 1 × 10-7-1 × 10-1 M salicylate. As can be seen in Figures 1 and 2 and Table 1, an increase in the concentration of the membrane active phase, up to ∼2.2% for both ionophores, is beneficial for obtaining any electrode slope closer to the theoretical value. Further increase in the concentration level
of the ion active phase in both membranes results in electrodes that display somewhat smaller slopes. The incorporation of TOMACl, as a cationic additive to the membranes, had no significant effect on the sensitivity of the Al(salophen)-based electrode, but decreased the slope of the calibration graph for the Sn(salophen) electrodes, and, as will be discussed later in this paper, worsened the selectivities of the membranes based on both ionophores. The influence of the lipophilic anionic additive on the potentiometric response of the electrodes based on both ionophores was also investigated. Without this additive, the membranes already exhibited Nernstian responses toward salicylate (electrodes B and G). In the case of the Al(salophen) membrane, addition of NaTPB dramatically worsened the response of the electrode (electrode E), so that the electrode no longer responded to anions. For the membrane containing Sn(salophen), in the presence of NaTPB (electrode J), the slope of the calibration plot was slightly decreased, but as will be seen later in this paper, the selectivity was greatly improved.
958 Analytical Chemistry, Vol. 72, No. 5, March 1, 2000
Figure 4. pH response of Al(salophen)-based membrane electrode (A) in the absence and (B) in the presence of 0.01 M salicylate. Figure 3. Calibration plots of electrodes B and K.
The potentiometric responses of the sensor systems, incorporating Al(salophen) and Sn(salophen) membranes, were studied in the salicylate concentration range of 1 × 10-7-1 × 10-1 M at pH 7.0 and 25 °C. The calibration curves for these systems in the linear region are shown in Figure 3. It can be seen that electrode B, with 2.2% Al(salophen), shows a linear potentiometric response over a wide concentration range of 1 × 10-6-1 × 10-1 M salicylate, with a Nernstian slope of -59.2 mV/decade of concentration. For electrode K, with Sn(salophen) ionophore, about the same linear range with a slope of -57.2 mV/decade was observed. The linear range of the electrode response for these sensors is about five decades of concentration, which is wider than those reported for the highly selective salicylate electrodes.10,17,23,24 The limits of detection, defined as the concentration of salicylate obtained when extrapolating the linear region of the calibration curve to the baseline potential, for different electrodes are given in Table 1. As can be seen, electrodes B and K show detection limits of about 1 × 10-6 and 8 × 10-7 M, respectively. The limit of detection obtained in this work is several orders of magnitude below the recommended therapeutic level of salicylate in biological fluids (e.g., whole blood, plasma, serum and urine, ∼1.5 × 10-4 M). Consequently, these samples can be substantially diluted with a suitable buffer solution to minimize variation in sample pH and probably decrease the effect of interferences. The electrode also covers the range of both acute and chronic cases of salicylate intoxication. For analytical applications, the response time of a sensor is an important factor. The response time of the electrodes, tested by measuring the time required to achieve a steady-state potential (within (1 mV), was within 20-30 s for solutions with salicylate levels in the range 1 × 10-6-1 × 10-1 M, depending on the stirring efficiency. The stability and reproducibility of the electrodes were also tested. The potentials remained almost constant for ∼10 min (drift e1 mV), after which a very slow divergence was observed. (23) Liu, D.; Chen, W. C.; Shen, G. L.; Yu, R. Q. Analyst 1996, 121, 1495-1499. (24) Li, Z. Q.; Yuan, R.; Ying, M.; Song, Y. Q.; Shen, G. L.; Yu, R. Q. Talanta 1998, 46, 943-950.
The standard deviation of the potential readings for electrodes B and K in 1 × 10-2 and 1 × 10-3 M salicylate solutions over periods of 20 min and 2 h were about 0.6 and 1.0 mV, respectively. The detection systems are very stable and can be used over a period of 1 month, without observing a considerable change in their response characteristics. The slopes of the calibration graphs were reproducible to within 1.2 mV/decade over a period of 1 month. The pH Response of the Electrodes. The potentiometric response of both membrane electrodes was found to be sensitive to pH changes. Figure 4A shows a typical pH response profile for the Al(salophen)-doped membrane, evaluated by titrating 0.1 M phosphoric acid solution with sodium hydroxide solution. Such a potentiometric pH response can be ascribed to the coordination of water molecules as axial ligands to the central metals,19 i.e., formation of [Al(salophen)(H2O)2]+X- and [Sn(salophen)(H2O)(X-)]+X- complexes. Once associated, the water can readily lose a proton to the solution, yielding hydroxide-coordinated metal complexes of the following type: [Al(salophen)(H2O)(OH-)] and [Sn(salophen)(OH-)X-]. This results in a response analogous to that resulting from anion coordination at the same axial site. The mechanism of pH response appears to be similar in nature to those found with the membranes doped with manganese(III) and tin(IV) porphyrins and phthalocyanines.10,17,25,26 However, when those same salophen-based membranes are conditioned in 0.05 M sodium salicylate solution, the response of the electrode is hardly affected by the change in pH in the range of about 3-8.5 for both membranes (Figure 4B). The behavior of the electrodes at high pH can be explained in terms of the increased interference from hydroxide ions. Protonation of salicylate anion may be the reason for potential changes at low pH values. The working pH range over which the electrodes can be used (3-8) covers the physiological conditions (pH 7.2-7.6). The optimum working pH for the most selective salicylate electrodes are 1-2 pH units below the physiological conditions.10,16,17,23,24 (25) Chaniotakis, N. A.; Chasser, A. M.; Meyerhoff, M. E.; Groves, J. T. Anal. Chem. 1988, 60, 185-188. (26) Ma, S. C.; Chaniotakis, N. A.; Meyerhoff, M. E. Anal. Chem. 1988, 60, 2293-2299.
Analytical Chemistry, Vol. 72, No. 5, March 1, 2000
959
Table 2. Potentiometric Selectivity Coefficients of the Al(salophen) and Sn(salophen) Membrane Electrodesa
Table 3. Results of the Analysis of Pharmaceutical Preparations and an Urine Sample
log Ksal,x of different membranes interfering ion, x -
N3 BrHCO3ClCNFINO3NO2C2O42ClO4HPO42SO42SCNmaleate acetate benzoate tartrate adipate citrate phthalate 3-nitrophthalate hydroxideb
potentiometry
B
F
G
J
K
-4.0 -4.0 -4.1 -4.3 -4.0 -4.3 -2.9 -3.8 -4.2 -5.3 -2.3 -5.7 -5.6 -2.0 -4.9 -4.0 -4.0 -5.1 -5.1 -5.6 -4.6 -4.8 1.0
-4.2 -3.9 -3.8 -4.4 -4.3 -4.5 -1.3 -2.9 -4.2 -5.4 -0.6 -5.5 -5.0 -0.5 -4.4 -3.6 -3.5 -4.2 -4.3 -4.9 -3.5 -4.0 1.0
-3.5 -3.5 -3.5 -3.8 -3.4 -4.0 -2.9 -3.8 -3.8 -4.4 -2.2 -5.1 -5.0 -1.7 -4.4 -3.5 -3.7 -4.5 -4.4 -5.0 -4.0 -4.3 1.3
-4.1 -4.2 -3.7 -4.2 -4.0 -4.2 -3.8 -4.1 -3.9 -4.6 -2.6 -5.3 -5.4 -2.2 -5.6 -3.9 -4.5 -5.3 -5.3 -5.8 -5.3 -5.2 1.5
-3.2 -3.1 -3.4 -3.7 -3.5 -4.0 -1.4 -3.2 -4.0 -4.1 -1.4 -5.0 -4.7 -1.0 -4.1 -3.2 -3.3 -4.1 -4.2 -4.6 -3.4 -4.0 0.8
a All values obtained in 0.05 M phosphate buffer solution, pH 7.0, at an anion concentration of 0.01 M. b The selectivity for salicylate relative to hydroxide were obtained at pH 8.4 using the fixed interference method.
Selectivity. The most important characteristic of any ionsensitive sensor is its relative response for the primary ion over other ions present in solution, which is expressed in terms of the potentiometric selectivity coefficients. Potentiometric selectivity coefficients (KSal,X), describing the preference by the membrane for an interfering ion X- relative to salicylate, were determined in 0.01 M solutions of the corresponding sodium salts by the separate solution method, except for hydroxide interference, which was estimated by the fixed interference method.27 Concentrations were used rather than activities due to uncertainties in estimating the ionic strength of the background electrolyte. The potentiometric ion selectivity coefficients of different electrodes, based on Al(salophen) and Sn(salophen) membranes, are summarized in Table 2. The selectivity coefficient patterns clearly indicate that the electrodes are highly selective to salicylate over a number of other anions. A typical selectivity pattern for a series of anions presented by electrode B is as follows: salicylate . thiocyanate > perchlorate > iodide > nitrate > benzoate > acetate > bromide - cyanide - azide > carbonate > nitrite > chloride > fluoride > phthalate > nitrophthalate > maleate > adipate > tartrate > oxalate > sulfate > citrate > phosphate. Selectivities over the organic and inorganic anions studied in this paper are better than those of other salicylate sensors based on enzyme electrodes,28 Nitron,14 ion-exchanger membranes,10,17,29 and membranes based on several metal complexes.11,13,15 The selectivities for salicylate over several anions, including azide, carbonate, cyanide, fluoride, (27) Guilbault, G. G.; Durst, R. A.; Frant, M. S.; Freiser, H.; Hansen, E. H.; Light, T. S.; Pungor, E.; Rechnitz, G. A.; Rice, N. M.; Rohm, T. J.; Simon, W.; Thomas, J. D. R. Pure Appl. Chem. 1976, 48, 127-132. (28) Fonong, T.; Rechnitz, C. A. Anal. Chim. Acta 1984, 158, 357-362. (29) Midgley, D. Anal. Chim. Acta 1986, 182, 91-101.
960 Analytical Chemistry, Vol. 72, No. 5, March 1, 2000
sample 1 2 3
label value 100 325 500
electrode B
Aspirin Content 100.4 311.4 468.0
electrode J
colorimetry
(mg/Tablet)a 98.1 309.6 459.0
94.1 296.6 446.4
Salicylate Content in Urine Sample (mg/100 mL)a 7.45 7.40 7.04 a
Average of three determinations.
oxalate, phosphate, sulfate, maleate, tartrate, adipate, phthalate, and nitrophthalate, have not been reported for the most selective salicylate electrodes.10,16,17,23,24 The report on these anions by Hassan and Hamada14 shows selectivity coefficients that are generally higher than those of electrode J by ∼2 orders of magnitude. The selectivity coefficients reported for bromide,16 chloride,16,17,23 perchlorate,10,16,17,23 thiocyanate,10,17,24 and acetate16 are superior to the proposed electrodes, but for the other anions, the selectivities observed in this report are comparable to or better than those reported for the most selective salicylate electrodes. The electrode based on Al(salophen) without additive (electrode B) is generally more selective to salicylate, compared to that based on the same electrode with 1% cationic additive (electrode F). In the case of Sn(salophen) ionophore, although the selectivities in the absence of the anionic additive are generally worse than those of the Al(salophen) (electrode B), the presence of 1.3% of NaTPB in the Sn(salophen) membrane results in better selectivities over aluminum ionophore. The KSal,OH values, given in Table 3, were obtained by the fixed interference method using the data shown in Figures 1 and 2, which were in good agreement with those obtained by the fixed parent ion using Figure 4. From the data given in Figure 4B, it can be seen that the interference of hydroxide ion becomes significant at pH values above 8.4, and therefore, the interference from hydroxide is negligible at the working pH of 7.0. Meyerhoff and co-workers10 also obtained the selectivity coefficient for the salicylate relative to hydroxide ion, from the calibration plot at pH 7.2. They reported a log KSal,OH value of 4.2 at this pH, but the interference at pH 5.5 was negligible. Mechanism of Salicylate Response and Selectivity. There have been several reports on the use of quaternary ammonium salts for the development of salicylate-selective electrodes.10,11,30,31 Although these electrodes offer a fairly good discrimination of hydrophilic anions, their preferred anions are perchlorate, thiocyanate, and iodide, respectively. The Al(salophen)- and Sn(salophen)-doped membrane electrodes displayed substantially improved selectivity for salicylate over several other anions. Most of these anions would be expected to interfere seriously with classical anion-exchanger-type membrane electrodes. The high selectivity for salicylate over perchlorate, thiocyanate, and iodide clearly deviates from the conventional Hofmeister anion response pattern and suggests that salicylate is directly interacting with the (30) Choi, K. K.; Fung, K. W. Anal. Chim. Acta 1982, 138, 385-391. (31) Mitsana-Papazoglou, A.; Diamandis, E. P.; Hadjiioannou, T. P. Anal. Chim. Acta 1984, 159, 393-399.
Table 4. Composition of Simulated Serum compound D,L-alanine L-arginine L-aspartic
acid
L-cysteine
glycine L-histidine L-lysine
concn (mM)
compound
concn (mM)
0.41 0.21 0.88 0.071 0.14 0.14 0.20
D,L-methionine
0.034 0.16 0.16 0.085 8.0 88.8 0.17
L-phenylalanine L-serine D,L-tryptophan
NaHCO3 NaCl citric acid
metallosalophens as an axial ligand. The strength of this interaction, relative to other anions, dictates the observed selectivity pattern of the electrodes. The hard metal centers in both ionophores endow a strong oxophilic character on the complexes and interact strongly with the carboxylate group of salicylate, which acts as a hard base. As can be seen in Table 2, the membranes containing TOMACl (electrodes F and K) generally show less selectivity to salicylate in comparison to the electrodes without the cationic additive (electrodes B and G). Such a difference is more pronounced for the highly lipophilic ClO4-, SCN-, and I- anions. The anion responses observed for the electrodes containing TOMACl may stem from the ion-exchange process that occurs to some extent at the membrane surface. Recently, there have been several reports on the effect of anionic additives on the potentiometric selectivity of the membrane electrodes.9,17,32 These reports demonstrated that for ionophores that operate via a charged carrier mechanism, incorporation of a certain ratio of fixed anionic sites into the membranes improves the selectivity of electrodes. The effect of NaTPB on the selectivity of Al(salophen)- and Sn(salophen)-doped membranes was studied. Addition of NaTPB to the Al(salophen)-based membrane worsened the performance of the electrode, so that the electrode was no longer selective to salicylate (electrode E). However, the selectivity of the Sn(salophen)-based electrode was considerably improved in the presence of the anionic additive (electrode J). As can be seen from the selectivity coefficients given in Table 2, electrode J with 1.3 wt % NaTPB (∼39 mol % relative to the ionophore) shows the highest selectivity among the different compositions studied for both ionophores. It is interesting to note that, although the Al(salophen) membrane (electrode B) generally shows somewhat better selectivities over the corresponding electrode based on Sn(salophen), the reverse is true in the presence of the anionic additive. According to the theoretical models that predict the effect of lipophilic anionic additives on the selectivity of the membrane electrodes,9,32 the above studies involving these effects suggest that Al(salophen) ionophore acts on the basis of a neutral carrier mechanism but Sn(salophen) ionophore behaves as a charged carrier in the membranes. Previous studies have shown that aluminum salen19 and tin porphyrin26 exist as [Al(L)(H2O)2]+ and [Sn(L)(H2O)(X-)]+ species in aqueous solutions and that H2O groups occupy the axial position of an octahedral coordination environment around the central metals. On the basis of these reports, and the results of this study, i.e., (1) deteriorating effect of NaTPB on Al(salophen) (32) Schaller, U.; Bakker, E.; Spichiger, U. E.; Pretsch, E. Anal. Chem. 1994, 66, 391-398.
Table 5. Determination of Salicylate Added to Synthetic Serum Sample electrode B
electrode J
amt added (mmol/L)
amt (mmol/L)
% recovery
amt founda (mmol/L)
% recovery
0.990 1.48 1.96 2.91 3.65 5.66
0.993 1.43 1.89 2.71 3.74 5.37
100.3 96.6 96.4 93.2 97.2 94.9
1.02 1.52 2.01 2.87 3.94 5.86
103.0 102.7 102.6 98.6 102.3 103.5
a
founda
Average of three determinations.
and its beneficial effect on Sn(salophen) membranes, which suggest that these compounds act as electrically neutral and charged ionophores, respectively, and (2) the acid-base titration curve (Figure 4A), which suggests the pH-dependent deprotonation of the axial water molecules leaving hydroxide ions, the following mechanisms may be proposed for the potentiometric response of these membranes toward salicylate: H2O
-H+
[Al(salophen)X-]0 y\z [Al(salophen)(H2O)2]+X- y\z +sal-
[Al(salophen)(H2O)(OH-)]0 y\z neutral ionophore [Al(salophen)(sal-)(OH-)]-M+ H2O
[Sn(salophen)(X-)2]0 y\z -H+
[Sn(salophen)(H2O)X-]+X- y\z charged ionophore [Sn(salophen)(OH-)X-]0 +sal-
[Sn(salophen)(H2O)X-]+X- y\z charged ionophore [Sn(salophen)(sal-)X-]0
Analytical Applications. With the wide use of aspirin, there is an increasing need for measurement of “free” salicylate in biological samples such as urine and serum. The high degree of salicylate selectivity exhibited by these metallosalophens makes them potentially useful for monitoring concentration levels of “free” salicylate in biological samples. For therapeutic drug monitoring, determining the free concentration of a drug is claimed to be physiologically more relevant than determining its total concentration.10 In this regard, experiments were performed to determine the feasibility of using these electrodes to measure salicylate in human urine samples. The urine sample was diluted 1:10 with phosphate buffer, pH 7.0. The calibration plot, obtained under the same experimental conditions, was employed to evaluate the concentration of salicylate in urine sample. The results obtained by the potentiometric method are in good agreement with those of the spectrophotometric procedure,33 as shown in Table 3. (33) Trinder, P. Biochem. J. 1954, 57, 301-303.
Analytical Chemistry, Vol. 72, No. 5, March 1, 2000
961
Experiments were conducted to measure the concentration of salicylate added to a synthetic serum sample, the composition of which is listed in Table 4. The concentration of each component was chosen to match its normal level in human serum.34 The results of salicylate determination, summarized in Table 5, show an average recovery of 96.4% for electrode B and 102.1% for electrode J, indicating that the constituents of the synthetic serum sample do not significantly interfere with the detection of salicylate. These electrodes seemed to provide an alternative device for the direct determination of salicylate in biological samples. The proposed electrodes were used for the determination of salicylate content of the hydrolyzed pharmaceutical preparations. Tablets of different samples of aspirin were treated according to the procedure described above, and the resulting solutions were used for the determination of salicylate content by potentiometric and spectrophotometric methods. The results obtained by the electrode method (Table 3) are in good agreement with those obtained by the spectrophotometric method.35 (34) Pau, C. P.; Rechnitz, G. A. Anal. Chim. Acta 1984, 160, 141-147. (35) British Pharmacopoeia; University Press: Cambridge, 1993; Vol. II.
962
Analytical Chemistry, Vol. 72, No. 5, March 1, 2000
CONCLUSIONS The results of this study show that the potentiometric method based on Al(salophen) and Sn(salophen) membranes coated on graphite electrodes may provide an attractive alternative for the determination of salicylate. The proposed sensors are very easy to prepare and show high sensitivity and wide dynamic range. High selectivity, very low detection limit, and rapid response make these electrodes suitable for measuring the concentration of salicylate in a wide variety of samples, including pharmaceutical preparations and biological samples, without the need for preconcentration or pretreatment steps and without significant interaction from other anionic species present in the samples. ACKNOWLEDGMENT The authors express their appreciation to the University of Isfahan Graduate Office for financial support of this work.
Received for review July 8, 1999. Accepted November 19, 1999. AC990749W
