Laser Ablation Inductively Coupled Plasma Mass Spectrometry

It was found that the contents and the distribution of Pb2+ and Na+ ions ...... Crank, J. The Mathematics of Diffusion; Oxford University Press: New Y...
0 downloads 0 Views 148KB Size
Download PDF
Anal. Chem. 2006, 78, 5584-5589

Laser Ablation Inductively Coupled Plasma Mass Spectrometry Assisted Insight into Ion-Selective Membranes Agata Michalska,* Marcin Wojciechowski, Barbara Wagner, Ewa Bulska, and Krzysztof Maksymiuk

Department of Chemistry, Warsaw University, Pasteura 1, 02-093 Warsaw, Poland

Laser ablation inductively coupled plasma mass spectrometry was used to evaluate ion depth profiles across ion-selective membranes. Advantageously, this approach does not require incorporation of additional components (e.g., chromoionophore) in the membrane composition, as compared to that used in typical potentiometric applications. Moreover, comparison of the distribution of ions in differently pretreated membranes is possible. Concentration profiles of primary and interfering agent (Na+) ions were recorded, for example, of Pb2+-selective poly(vinyl chloride)-based membranes. It was found that the contents and the distribution of Pb2+ and Na+ ions across the membrane is strongly dependent on the composition of the solutions to which both sides of the membrane are exposed during preconditioning and on the plasticizer included in the membrane formulation. Typical plasticizers, bis(2-ethylhexyl sebacate) (DOS) and the more polar 2-nitrophenyl octyl ether (o-NPOE), were used. It was found that faster ion transport occurs for o-NPOE, and the membrane saturation with Pb2+ ions was achieved within less than 20 h for a 400-µm-thick membrane. In the case of the less polar plasticizer DOS, due to slower rate of ion transport, even after 20 h, the Pb2+ concentration gradients were still visible within the membrane. On the basis of concentration profiles, primary ion diffusion coefficients in both membranes were calculated, and the value obtained for o-NPOE containing membrane was found to be ∼2 times higher than for its DOS-plasticized counterpart. Knowledge about ion distribution and transport through an ion-selective membrane (ISM) as well as their changes in time is extremely important for both understanding the mechanism of ion-selective electrodes’ (ISEs) operations and improving their analytical parameters. However, imaging the diffusion of ions through ISM is not an easy task. Various methods have been proposed to study ion transport through the membrane, including application of stacks of relatively thin membranes, which were analyzed separately.1 Later, electrochemical methods were used for this purpose, including scanning electrochemical microscopy to follow time-dependent building of a diffusion layer at the (1) Nahir, T. M.; Buck, R. P. Helv. Chim. Acta 1993, 76, 407-415.

5584 Analytical Chemistry, Vol. 78, No. 15, August 1, 2006

aqueous phase boundary of ISE.2 Despite great advantages of these methods, the information that is obtained is related mostly to the membrane bulk because the variation between particular membrane regions is more difficult to follow with electrochemical techniques.3,4 Much effort has been put forth to evaluate surfacerelated phenomena by FT-IR ATR (e.g., see ref 5) and the second harmonic generation.6 Pretsch, Lindner, Bakker, and co-workers have used VIS spectroscopic methods to follow concentration profiles in ion-selective membranes.7-12 These methods offer many advantages in visualizing ion concentration profiles; however, they can be applied only for membranes containing chromoionophores. Moreover, this approach requires dedicated, customized setup. Inductively coupled plasma mass spectrometry (ICPMS) has been gaining popularity due to its multielement capability, high sensitivity, and wide dynamic range.13-15 For this reason, ICPMS is a method that can be used as a reference for potentiometric analysis.16 In this paper, we describe the first attempt to use ICPMS coupled with a laser ablation (LA) sample introduction system for imaging ion (total) concentration profiles across the membranes used in ISEs. To achieve this, the membrane (after conditioning/measurement under typical potentiometric conditions) is exposed to the laser beam, and at each pulse, a minute amount of the solid (membrane) is evaporated, transferred by carrier gas to the ICPMS, and analyzed. Depending on the (2) Gyurcsa´nyi, R. E.; Pergel, E˙ .; Nagy, R.; Kapui, B.; Lan, B. T. T.; Toth, K.; Bitter, I.; Lindner, E. Anal. Chem. 2001 73, 2104-2111. (3) Horvai, G.; Gra´f, E.; To´th, K.; Pungor, E.; Buck, R. P. Anal. Chem. 1986, 58, 2735-2740. (4) To´th, K.; Gra´f, E.; Horvai, G.; Pungor, E.; Buck, R. P. Anal. Chem. 1986, 58, 2741-2744. (5) Kellner, R.; Fischbo ¨ck, G.; Go¨tzinger, G.; Pungor, E.; To´th, K.; Po´los, L.; Lindner, E. Fresenius’ J. Anal. Chem. 1985, 322, 151-156. (6) Tohda, K.; Umezawa, Y.; Yoshiyagawa, S.; Hashimoto, S.; Kawasaki, M. Anal. Chem. 1995, 67, 570-577. (7) Schneider, B.; Zwickl, T.; Federer, B.; Pretsch, E.; Lindner, E. Anal. Chem. 1996, 68, 4342-4350. (8) Lindner, E.; Zwickl, T.; Bakker, E.; Lan, B. T. T.; To´th, K.; Pretsch, E. Anal. Chem. 1998, 70, 1176-1181. (9) Gyurcsa´nyi, R.; Lindner, E. Anal.Chem. 2002, 74, 4060-4068. (10) Gyurcsa´nyi, R.; Lindner, E. Anal.Chem. 2005, 77, 2132-2139. (11) Heng, L. Y.; To´th, K.; Hall, E. A. H. Talanta 2004, 63, 73-87. (12) Long, R.; Bakker, E. Anal. Chem. 2004, 511, 91-95. (13) Moens, L.; Jakubowski, N. Anal. Chem. 1998, 70, 251A. (14) Hattendorf, B.; Latkoczy, Ch.; Gu ¨ nther, D. Anal. Chem. 2003, 75, 341A347A. (15) Begerow. J.; Turfeld, M.; Dunemann, L. J. Anal. At. Spectrom. 2000, 15, 347-352. (16) Ceresa, A.; Bakker, E.; Hattendorf, B.; Gu ¨ nther, D.; Pretsch, E. Anal. Chem. 2001, 73, 343-351. 10.1021/ac0605243 CCC: $33.50

© 2006 American Chemical Society Published on Web 07/01/2006

exposure time and the energy of the laser beam, microsamples of materials from different depths can be evaporated; thus, results can be transformed to element (ion) content vs depth (distance from the membrane surface) relation. A significant advantage of this approach is that membranes of typical composition, that is, as used for potentiometric measurement, with no modifications or additives can be used. The aim of this work was to evaluate the advantages of using LA ICPMS for the investigation of ion transfer through ionselective, solvent, polymeric membranes. As an example, poly(vinyl chloride) (PVC)-based, lead-selective membranes were used. The main interest was focused on the influence of the composition of the contacting solution on the concentration profiles of lead and sodium (a typical interfering agent) ions in the membranes. Special attention was paid to solutions of composition similar to that applied to achieve low detection limit ISEs. To the best of our knowledge, LA ICPMS studies of migration/diffusion processes in ISE membranes have not yet been described in the literature. The studies were also focused on the influence of the plasticizer included in the membrane formulation. Since the plasticizer can affect ion binding and its transport rate, it can significantly affect the analytical parameters of the sensor.17,18 Therefore, the influence of plasticizer polarity on ion concentration profiles and their changes over time were studied, and as a consequence, the diffusion coefficients of ions in the membrane were evaluated. EXPERIMENTAL SECTION Apparatus. An inductively coupled plasma mass spectrometer ELAN 9000 (Perkin-Elmer, Germany) equipped with the laser ablation system LSX-200+ (CETAC, Omaha, NE) was used. The LSX-200+ combines a stable, environmentally sealed, 266-nm UV laser (Nd:YAG, solid state, Q-switched) with a high sampling efficiency, variable 1-20-Hz pulse repetition rate, and maximum energy up to 6 mJ/pulse. Samples were inserted in a cell on an X-Y-Z-translation stage. The exact position of the sample was observed with a CCD camera as a viewing system under PC control. The LSX-200+ is centrally controlled by CETAC Windows software that allows selective ablation of chosen areas of investigated samples. The applied laser energy was 3.2 mJ; if not stated otherwise, the repetition rate was 5 Hz; and the spot size was 100 µm. The changes in distribution of elements within the ionselective membrane were followed, and their amounts in different membranes were compared. The quantitative analysis of the membranes’ components was not intended. Reagents. Tetrahydrofuran (THF), poly(vinyl chloride), 2-nitrophenyl octyl ether (o-NPOE), bis(2-ethylhexyl sebacate) (DOS), sodium tetrakis[3,5-bis(trifluoromethyl)phenyl]borate (NaTFPB), lead ionophore IV (tert-butylcalix[4]arene-tetrakis(N,N-dimethylthioacetamide), and calcium ionophore N,N-dicyclohexyl-N′,N′dioctadecyl-3-oxapentanediamide were from Fluka AG (Buchs, Switzerland). Doubly distilled and freshly deionized water (resistance 18.2 MΩcm, Milli-Q Plus, Millipore, Austria) was used throughout this work. All salts used were of analytical grade and were obtained from POCh (Gliwice, Poland). (17) Bakker, E.; Bu ¨ hlmann, P.; Pretsch, E. Chem. Rev. 1997, 97, 3038-3132. (18) Bedlechowicz, I.; Maj-Z˙ urawska, M.; Sokalski, T.; Hulanicki, A. J. Electroanal. Chem. 2002, 537, 111-118.

Ion-Selective Membranes. Lead-selective membrane contained (in wt %) 0.6% lead ionophore IV, 0.5% NaTFPB, 66.3% o-NPOE, and 32.6% PVC; or 0.6% lead ionophore IV, 0.6% NaTFPB, 63.4% DOS, and 35.4% PVC. A total 400 mg of membrane components was dissolved in 4 mL of THF. This cocktail was poured into a glass ring, i.d. 25 mm. After evaporation of the solvent, disks of 7-mm diameter were cut out and glued into PVC tubes with PVC slurry. If not stated otherwise, lead-selective membranes were used. In a control experiment, calcium-selective membranes of the above given composition, with an equivalent molar portion of Ca ionophore applied instead of Pb, were used. To avoid errors related to uncontrolled membrane thickness variation (within the membranes cut out form one disk), the thickness of each membrane was determined after the LA ICPMS experiment using micrometer caliper, and in all cases, the thicknesses of the membranes used were close to 400 µm. To mimic conditions prevalent in normal ion-selective electrode solutions, the membranes had contact with solution either of the same or of different composition (these were applied inside and outside of the above-described element); in a manner similar to how it is done in routine procedures involving ion-selective electrodes. As a buffer of constant and low activity of Pb2+, a mixture of 0.1 M citric acid with sodium hydroxide/sodium chloride, and 0.1 M citric acid, pH ) 6.4, was used. The total concentration of sodium ions was 0.3 M; the solution was spiked with Pb(NO3)2 to yield a calculated lead ion activity, a, for Pb2+ of 1.5 × 10-11 M. For comparison, membranes newly prepared or conditioned in distilled water were also analyzed. The changes in distribution of isotopes 208Pb and 23Na across the membrane (if not otherwise stated, from the outer to the inner surface of the membrane) were followed. The obtained PVC tubes with ion-selective glued membrane, if not stated otherwise, were conditioned for 20 h, rinsed well with water (both outside and inside), tissue-dried, then exposed to the laser beam. The evaporated material was then analyzed by ICPMS. If the second or second and third conditioning solutions were applied, after removing the element from the primary conditioning solution, the outside of the PVC tubes with the ion-selective glued membrane was well rinsed, tissue-dried, and placed in the next conditioning solution for exactly 1 h. Finally, the PVC tubes with ion-selective glued membrane were rinsed well with water (both outside and inside), tissue-dried, then immediately analyzed. Each LA ICPMS analysis was performed in duplicate (for at least two different points on the ISM surface). RESULTS AND DISCUSSION By executing laser ablation from the surface and subsurface domain of the ion-selective membrane material, changes in lead and sodium contents in tested membranes can be followed. Laser ablation parameters: laser energy (3.2 mJ), spot size (100 µm), and the repetition rate (5 Hz) were optimized to ensure the reproducibility of the results as well as the resolution, allowing the visualization of the element concentration profile within the ion-selective membrane. With spot size smaller that 50 µm, the intensity of the signal was too low, but spot sizes above 100 µm do not allow the resolution necessary to track the element contents changes within the membrane. Considering the repetition rate, a rate higher than 5 Hz causes very fast evaporation of the Analytical Chemistry, Vol. 78, No. 15, August 1, 2006

5585

Table 1. Estimated Mean Intensity of the Signal Representative for Each Membrane Sample (cps) for 208Pb and 23Na for Different Plasticizers (DOS or o-NPOE) and Conditioning Solutions DOS

o-NPOE

conditioning solution

Pb

Na

Pb

Na

as prepared membrane water 0.01 M Pb(NO3)2 0.1 M NaCl Pb2+ ion buffer

200 200 1.6 × 106 3 ×103 300

1 × 105 1 × 105 1.3 × 104 7 × 104 105

200 400 1.8 × 106 9 × 103 650

1.4 × 105 1.4 × 105 1.3 × 104 8 × 104 1.6 × 105

membrane material, resulting again in poor resolution when tracking the signal changes. This was unfavorable for studying ion diffusion/migration within the membrane. To evaluate the influence of the crater formation when a laser beam penetrates the membrane during the following shots, the profiles were registered for the asymmetrically conditioned membrane starting from opposites sides of the membrane. Indeed, a small difference in recorded profiles for the distribution of 208Pb and 23Na was registered, which was due to the different shape of the crater. However, this difference was negligible with respect to the general overview of the profile of the element of interest. Sodium ions were introduced to the phase with liphophilic salt (NaTFPB); alternatively, they can originate from solutions (NaCl or applied buffer). Lead ions originated only from conditioning solutions. Therefore, it was of great interest to follow simultaneously, that is, in the same subportion of the evaporated membrane, the signal for lead and sodium isotopes. For as-prepared membranes, an increase in the 208Pb signal intensity was observed when the very first portion of the membrane was evaporated. This occurred for both types of membranes, either with DOS or o-NPOE plasticizers. However, it was difficult to evaluate quantitatively the amount of lead present in the as-prepared membrane. Therefore, the as-prepared membranes were considered blank samples to which others were compared. Analysis of membranes that were not in contact with any solution revealed high intensities of sodium signals, in the range of 105 cps. (Table 1). Symmetrically Conditioned Membranes. The tested membranes were symmetrically conditioned (applying the same solution both inside and outside the PVC tube closed from one end with the ISM) overnight, applying one of the following solutions: water, 0.1 M NaCl, 10-2 M Pb(NO3)2, or Pb2+ in ion buffer. As expected for the equilibrated membranes, no concentration gradients were observed in the membrane, regardless if it was plasticized with DOS or with o-NPOE. Interestingly, significant differences in lead and sodium contents were revealed. Obtained results are collected in Table 1. Membranes conditioned in water contain, similar to unconditioned ones, a very little amount of lead (signal intensity 200 and 400 cps for DOS- and o-NPOE-plasticized membranes, respectively), whereas the intensity of sodium signals was close to 105 cps both for DOS- and o-NPOE-plasticized membranes, (Table 1). Thus, sodium contents in tested membranes were not altered in the course of conditioning in water, as compared to nontreated membranes. For samples conditioned in 10-2 M Pb(NO)3, a significant decrease in the intenisity of the sodium signals was observed: 5586

Analytical Chemistry, Vol. 78, No. 15, August 1, 2006

for both types of membranes, it was close to 104 cps, pointing to a decrease in the sodium content, as compared to as-prepared/ water-conditioned membranes. As expected, a significant increase in the intensities of lead was observed: they were equal to 1.6 × 106 and 1.8 × 106 cps for DOS- and o-NPOE-plasticized membranes, respectively. Higher Pb2+ contents in o-NPOE-plasticized membrane confirms better suitability of this plasticizer in divalent ion-selective membranes.19 Sodium signal intensities obtained for membranes conditioned in 0.1 M NaCl were similar for both types of membranes (close to 104). Surprisingly, conditioning in 0.1 M NaCl results in a relatively high lead signal intensity; however, these are 2-3 orders of magnitude lower as compared to membranes conditioned in Pb(NO3)2. The significant difference in Pb contents between membranes containing different plasticizers was revealed. The Pb content was 3 times higher for the o-NPOE-containing membrane, as compared to the counterpart with DOS. Taking into account that water-conditioned membranes contained significantly less lead (Table 1), it seems that lead ion impurities present in analytical grade NaCl were selectively accumulated in the ionophore-containing membrane during conditioning. This statement is well supported by the results of ICPMS analysis of the NaCl solution. The lead content of 0.1 M NaCl was close to 4 × 10-7 M, whereas the lead signal registered for water was at the blank (carrier gas) level. In a control experiment for membranes with Ca ionophore conditioned in 0.1 M NaCl, similar lead signal intensities, close to 500 cps, were obtained for both o-NPOE- and DOS-plasticized selective membranes. Sodium signal intensities obtained after conditioning in 0.1 M NaCl were equal to 5 × 104 cps and 2 × 104 cps for o-NPOE- and DOS-plasticized Ca-selective membranes, respectively. These results confirm that for lead-selective membranes, Pb ionophore is responsible for selective accumulation of Pb from the solution. For membranes conditioned in Pb2+ ions buffer, relatively high sodium signal intensities were recorded, close to 105 cps. In this case, despite constant and low activity of lead ions in the conditioning solution, lead signal intensities obtained were equivalent to those for water-conditioned membranes. The comparison of sodium contents in the membranes conditioned in lead ion buffer with those pretreated in Pb(NO3)2 suggests that the former membranes were not saturated with lead. Moreover, sodium contents in the membranes, after contact with lead ions buffer, compared to nonconditioned/water-conditioned membranes, was unaffected within the range of experimental error. Interestingly, comparison of results obtained for these membranes and for those conditioned in NaCl solution (containing 4 × 10-7 M of lead) points out that there is a threshold value for the concentration of free lead ions in solution, not the total lead content, that results in accumulation of lead in the membrane. Asymmetrically Conditioned Membranes. A different picture was obtained for changes in lead and sodium contents in the membranes conditioned in sample solutions of composition different from that of the internal solution (asymmetrically conditioned membranes). Figure 1 presents lead and sodium signals changes recorded for DOS- and o-NPOE-plasticized membranes that after equilibra(19) Armstrong, R. D.; Horvai, G. Electrochim. Acta 1990, 35, 1-7.

Figure 1. Intensity of the measured signal as a function of laser ablation penetration depth obtained for tested membranes conditioned overnight in lead ion buffer (symmetrically), then for 1 h in 0.01 M Pb(NO3)2 (sample solution) using lead ion buffer (internal solution): DOS-plasticized (thick lines) and o-NPOE-plasticized (thin lines). Red lines correspond to the sodium signal; black lines, to the lead signal. Note the difference in the scales. Inset: Lead signal intensities measured as a function of laser ablation penetration depth obtained for tested membranes conditioned overnight in 0.01 M Pb(NO3)2 (sample solution) and lead ion buffer (internal solution): DOS-plasticized (thick line) and o-NPOE-plasticized (thin line).

tion for 20 h with buffer solution of constant and low activity of lead ions (symmetrically) were transferred for 1 h to 0.01 M Pb(NO3)2 (external) solution (asymmetric arrangement). For DOSplasticized membrane, the highest intensities of the lead signal were recorded quite close to the surface that had been in contact with Pb(NO3)2 solution. However, while moving deeper into the membrane, toward the internal solution, a rapid decrease in lead signal intensities was recorded. The region of high lead content was reached in less than 20% of membrane thickness. Also in this region, when moving into the membrane bulk, a gradual increase in the intensity of the sodium signal was observed. Then, the intensity of the sodium signal does not change much until the region very close to the internal solution is reached. Sodium contents in the bulk of the DOS-plasticized membrane was close to that of this type of membrane conditioned symmetrically in lead ion buffer. This suggests that starting from ∼20-25% of the membrane thickness from the sample side, not much change was introduced by contact of the membrane with Pb(NO3)2 solution. Interestingly, lead contents in the o-NPOE-plasticized membrane did not change monotonically. Close to the surface that was in contact with Pb(NO3)2 solution, the lead content was relatively low, and it increased while moving into the membrane bulk. For o-NPOE-plasticized membranes, a maximum lead signal intensity was observed deeper in the membrane, as compared to its DOSplasticized counterpart. Starting from ∼30% of o-NPOE-plasticized membrane thickness, the intensity of the lead signal decreases, to reach a level similar to that recorded for the DOS-plasticized membrane at the internal solution side. Sodium signal intensities decreased from values typical for a o-NPOE-plasticized membrane in contact with lead ion buffer, prevailing at the internal side of the membrane, to significantly lower values at the opposite side of the membrane. These results obtained for membranes that were in contact with primary ion solution for a relatively short time (although comparable to the duration of a potentiometric measurement)

point to a different mobility of Pb(II) ions in the membrane, higher in the o-NPOE-plasticized membrane as compared to its DOScontaining counterpart (see Figure 1). These results are in agreement with conclusions based on potentiometric studies,18 pointing to more effective ion transport through o-NPOEplasticized membranes. Duration of the conditioning step should obviously affect the concentration profiles. After overnight conditioning of the membranes in 0.01 M Pb(NO3)2 (external solution), with the same internal solution as above, the Pb(II) concentration gradients were significantly reduced; however, the influence of the kind of plasticizer is pronounced (see Figure 1 inset). In the o-NPOEcontaining membrane, uniform Pb distribution was recorded (5 × 106 cps), but in the DOS-plasticized membrane, the Pb-related intensity decreased from 8 × 106 (outer solution side) to 4 × 106 cps (inner solution side), confirming the role of the time scale of the experiment and higher mobility of Pb(II) in the presence of the more polar plasticizer. Taking into consideration that membranes symmetrically conditioned in NaCl or in buffer solution of constant and low activity of lead ions had similar sodium contents, it was interesting to see how the composition of the membranes conditioned in NaCl is affected upon 1 h contact with 0.01 M Pb(NO3)2 solution (see Figure 2). The obtained results, lead and sodium intensities, were pretty similar to those presented in Figure 1, with the exception of the maximal values of lead signal intensities recorded for the DOS-plasticized membrane. It should be noted that total lead contents of either DOS- or o-NPOE-plasticized membranes (area below the lead signal) conditioned before contact with Pb(NO3)2 solution in NaCl or in buffer were similar. Surprisingly, from the elemental composition point of view, there is not much difference between conditioning in low-activity primary ion or interfering agent solution shortly after the membrane comes to contact with primary ion; however, it is influenced by the plasticizer that is applied. Analytical Chemistry, Vol. 78, No. 15, August 1, 2006

5587

Figure 2. Intensity of the measured signal as a function of laser ablation penetration depth obtained for tested membranes conditioned overnight in 0.1 M NaCl (symmetrically), then for 1 h in 0.01 M Pb(NO3)2 (sample solution) using 0.1 M NaCl (internal solution): DOS-plasticized (thick lines) and o-NPOE-plasticized (thin lines). Red lines correspond to the sodium signal; black lines, to the lead signal. Note the difference in the scales.

Figure 3. Intensity of the measured signal as a function of laser ablation penetration depth obtained for tested membranes conditioned overnight in lead ion buffer (symmetrically), then for 1 h in 0.01 M Pb(NO3)2 (sample solution) using lead ion buffer (internal solution), and for another 1 h in lead ion buffer (sample solution) using lead ion buffer (internal solution): DOS-plasticized (thick lines) and o-NPOE-plasticized (thin lines). Red lines correspond to the sodium signal; black lines, to the lead signal. Note the difference in the scales.

Figure 3 presents the responses of membranes that after symmetrical conditioning in the buffer of constant and low activity of lead ions were placed in 0.01 M Pb(NO3)2 sample solution for 1 h and then were transferred again to the buffer for another 1 h. The aim of this experiment was to see if the content or distribution of lead and sodium changes and if the initial state of the membranes symmetrically conditioned in the buffer can be restored. Both lead and sodium signal intensities vs distance from the sample side of the DOS-plasticized membrane surface are not much different from those recorded for the same type of membrane after contact with the primary ion solution (without further conditioning), except for the lower maximal value of the lead signal intensity. However, for the o-NPOE-plasticized membrane, the maximum of the lead signal intensity decreased and moved deeper into the membrane. The curve of intensity vs 5588 Analytical Chemistry, Vol. 78, No. 15, August 1, 2006

distance became more flattened, pointing to progressing uniform redistribution of lead in the membrane (Figure 3). This proves that although the membrane is no longer in contact with the primary ions, incorporated Pb2+ ions were able to continue diffusion uniformly into the membrane bulk. The sodium gradient in the membrane, observed as a gradual increase of the sodium signal intensity from the sample to the internal solution side of the membrane, was not affected, as compared to the sample that was not transferred into the buffer solution after contact with the primary ions. Results presented in Figure 3 clearly point out that despite the membranes’ contact with buffer of constant and low activity of Pb2+, the initial state of the samples, as after symmetrical conditioning in this buffer, cannot be restored. Evaluation of Lead Apparent Diffusion Coefficient. On the basis of determined concentration profiles of lead ions in the

Figure 4. Pb concentration profiles (solid lines) calculated from eq 1 for t ) 1 h and increasing D values (arrow) equal to 1 × 10-8, 2 × 10-8, 2.5 × 10-8, 3 × 10-8, 5 × 10-8, and 6.5 × 10-8 cm2 s-1. x/d is the ratio of the distance from the outer solution/membrane interface to the membrane thickness. Experimental points for DOS- (b) and o-NPOE-plasticized (O) membranes are included; for better clarity, only some experimental data from curves in Figure 1 were taken. For the membrane with o-NPOE, the maximal signal intensity was taken from extrapolation of signal intensity for x/d > 0.4 to x/d ) 0.

membranes and their changes in time, diffusion coefficients of these ions could be calculated. Reported values of diffusion coefficients of free ionophore molecules and complexed cations in membranes containing 33% PVC are similar, close to 10-8 cm2 × s-1.19 Precise determination of diffusion coefficients of ions is not an easy task because ion fluxes are affected not only by concentration gradients but also by potential changes within the membrane. Therefore, numerical resolution of the coupled NernstPlanck-Poisson equation system is required to obtain exact data.20 The procedure simplifies considerably if potential change influence within the membrane is neglected and, thus, equations of diffusion can be applied. This assumption has been used previously by Pretsch, Lindner, and Bakker and co-workers7,21 to analyze spectrometric data related to a protonated/nonprotonated chromoionophore system in a membrane. In this case, an equation corresponding to diffusion in a thin layer system can be successfully applied.22 This equation, adapted to the present case conditions, is

( ) (

)

nπx x 2 ∞ I(l) cos(nπ) Dn2π2t sin I(x, t) ) I(l) + exp l π n)1 n l l2

∑

(1) where I(x, t) is the signal intensity of lead in the membrane at distance x (from the membrane/internal solution interface) for time t. I(l) is the intensity in the membrane next to the membrane/outer solution interface (contacting with high Pb2+ concentration solution), for time t > 0 and x ) l (l is the membrane thickness); D is the diffusion coefficient. Figure 4 presents signal intensity calculated from eq 1 as a (20) Sokalski, T.; Lingenfelter, P.; Lewenstam, A. J. Phys. Chem. B 2003, 107, 2443-2452. (21) Long, R.; Bakker, E. Electroanalysis 2003, 15, 1261-1269. (22) Crank, J. The Mathematics of Diffusion; Oxford University Press: New York, 1993. (23) Iglehart, M. L.; Buck, R. P.; Pungor, E. Anal. Chem. 1988, 60, 290-295.

function of distance from the outer solution/membrane interface for various D values and t ) 1 h. Experimental data obtained for DOS- and o-NPOE-plasticized membranes (from Figure 1) recorded after symmetrical conditioning in Pb2+ ions buffer and then 1 h conditioning in 0.01 M Pb(NO3)2 were included. On the basis of this comparison, for a DOS-plasticized membrane, the estimated D value is 2.5 ((0.5) × 10-8 cm2 s-1 However, for the o-NPOE-plasticized membrane, the situation is more complicated due to nonmonotonic intensity changes. Because the membrane contains similar amounts of ionophore and lipohilic salt (no ionophore excess), the nonmonotonic change can result from ionophore depletion close to the membrane/outer solution interface. Just after immersion in 0.01 M Pb(NO3)2 solution, vigorous extraction of Pb2+ ions to the membrane occurs, and this front of high Pb(II) (complexed) concentration moves inside the membrane. However, for longer times, the accessible ionophore concentration near the interface decreases, and the amount of incorporated lead is lower, resulting in a lower intensity next to the interface. Therefore, we used eq 1, taking intensity values for the percent of membrane thickness higher than 50%, that is, far beyond the maximum area, where the ionophore depletion effect is not significant. The value of D determined for the o-NPOE-plasticized membrane was equal to 5.5 (( 1) × 10-8 cm2 s-1. Higher uncertainty of D determination results from a lower reproducibility of the signal intensity and a lower influence of a higher D value on the shape of the calculated concentration profiles. Nevertheless, the obtained value points out that the diffusion coefficient in a o-NPOE-containing membrane is ∼2 times higher than for the membrane with the less polar DOS. These results and the difference related to the plasticizer applied are consistent with those presented elsewhere, for example, for a K+ complex with valinomycin.23 CONCLUSIONS To the best knowledge of the authors, the advantages of the ICPMS technique with laser ablation were for the first time used to visualize transport of ions in ion-selective, solvent polymeric membranes. Thus, without compromising membrane composition, data related to the transport of primary and interfering agent ions can be obtained. Herein, we studied membranes that were pretreated in the way corresponding predominantly to working conditions of internal solution electrodes, applying ion buffers of constant and low activity of primary ions, that is, typical for lowactivity potentiometric measurements. Obtained results point out that from a membrane composition point of view, there is not much difference between membranes conditioned in the interfering agent solution only and those conditioned in a buffer of constant and low activity of the primary ion. Upon transfer to a primary ion solution, saturation with primary ions occurs. This effect is significantly quicker for o-NPOE-, as compared to DOSplasticized membranes. ACKNOWLEDGMENT Financial support of the research project 3 T09A 017 27 in the years 2004-2007, from KBN, Poland, is gratefully acknowledged. AC0605243 Analytical Chemistry, Vol. 78, No. 15, August 1, 2006

5589