Instrument-Free Control of the Standard Potential of Potentiometric

Oct 4, 2014 - Solid-Contact Ion-Selective Electrodes by Short-Circuiting with a ... Three different types of ion-selective membranes (ISMs) are studie...
0 downloads 0 Views 2MB Size
Download PDF
Article pubs.acs.org/ac

Instrument-Free Control of the Standard Potential of Potentiometric Solid-Contact Ion-Selective Electrodes by Short-Circuiting with a Conventional Reference Electrode Ulriika Vanamo and Johan Bobacka* Laboratory of Analytical Chemistry, Process Chemistry Centre, Åbo Akademi University, Biskopsgatan 8, 20500 Turku-Åbo, Finland ABSTRACT: A simple, instrument-free method to control the standard potential (E°) of potentiometric solid-contact ion-selective electrodes (SCISE) is described. In this method, the electrode potential of a SC-ISE is reset by short-circuiting the electrode with a metallic wire to a conventional Ag/ AgCl/3 M KCl reference electrode (RE) in a solution containing primary ions. The method is studied experimentally for SC-ISEs where the conducting polymer poly(3,4-ethylenedioxythiophene) doped with the bulky anion poly(sodium 4-styrenesulfonate), PEDOT(PSS), is used as the solid contact. Three different types of ion-selective membranes (ISMs) are studied: two potassium-selective membranes, with and without the lipohilic additive tetradodecylammonium tetrakis(4-chlorophenyl)borate (ETH-500) and a cation-sensitive membrane without an ionophore. When the SC-ISE is short-circuited with the RE, the PEDOT(PSS) layer is oxidized or reduced, thereby shifting the potential of the SC-ISE to the proximity of the potential of the RE so that the potential difference between these two electrodes becomes zero or close to zero. The slope of the calibration curve is preserved after the short-circuit treatment of the SC-ISEs. The short-circuiting method is an important step toward calibration-free potentiometric analysis.

P

for potential stability should naturally be adjusted to match the needs of a particular application, as there can be significant differences in short-, middle- and long-term stabilities1 (e.g., a single-shot pH sensor for medical use can be allowed a smaller time window for stability than one inserted in a mote in the Atlantic meant to operate for months or years). There have been different approaches striving for calibrationfree realizations ranging from long-term use devices (not disposable or partly disposable) based on optical sensing for environmental monitoring of CO2 (e.g.,13), health-care purposes for observing oxygen saturation in arterial blood,15 concentration of blood gases and pH16 to symmetrical potentiometric measuring systems for sodium and potassium in blood serum.17 Ion selective membranes9,12 and solid electrolyte sensors18 in the coulometric measurement mode as well as stopped-flow anodic stripping coulometry on thin gold electrodes19 have been suggested as means of achieving calibration-free operation. A calibration-free method to measure soil pH directly on site20 and the interesting approach of using an internal standard iontopohoresis device for calibration-free glucose measurements have been reported.21 Low-cost and disposability have been addressed for calibration-free optical pH sensors,8 planar potentiometric pH sensor chips based on nanoporous platinum,17 lead selective sensors for environmental monitoring14 and for all plastic

otential stability and reliability are key properties of potentiometric sensors.1 A traditional method to guarantee the validity and repeatability of the analytical results is to calibrate frequently, as is done in clinical analysis on a daily basis.2,3 Major application fields of potentiometric sensors, such as environmental and clinical analysis, would clearly benefit from a calibration-free mode of operation. Clinical analysis is heading toward portable point-of-care and home diagnostics,1,4−7 where the end-user is not necessarily a chemist4 accustomed to performing calibrations or willing to use time for it,8 making easy-to-use devices providing fast answers desirable. In environmental analysis, the direction is toward autonomous sensing motes in remote locations,3,9,10 for which calibration of the system requires integration of reagents, standards, waste storage and valves for fluid handling, which is expensive4,11 and impractical.2,12 Decreasing the need for calibration procedures could be the answer for both simplifying the devices and reducing their costs.4 In reality, shortcomings in sensor-to-sensor reproducibility and signal stability in time still necessitate calibrations.13 Thus, there is a balance to be considered between the validity of the analysis results (with the trouble of making calibration) and optimal, simple and low-cost realization of application setups.2,4 Alternatively, new types of solutions to address drift and irreproducibility should be found. Also, disposability of sensors calls for calibration-free arrangements8 or at least very simple and fast (e.g., one point) calibrations14 in which case sensor-to-sensor reproducibility is crucial.8,11,13 When it comes to drifts, the requirements © XXXX American Chemical Society

Received: April 22, 2014 Accepted: October 4, 2014

A

dx.doi.org/10.1021/ac501464s | Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry

Article

Figure 1. Schematic illustration of short-circuiting of (a) a SC-ISE and a conventional RE, (b) two SC-ISEs with each other and (c) two similar SCISEs with a conventional RE. Open-circuit measurement setup and components of the SC-ISE are shown schematically in panel d.

Cu2+-,22 Ca2+- and Pb2+-selective sensors,23 all of which (except the optical one) contain also a reference electrode in the solid state. The use of paper as a sensor substrate and even ionselective electrode (ISE) substrate is also an attractive approach.24 A key requirement for calibration-free measurements with potentiometric sensors is the reproducibility of the standard potential (E°),25 which still, in solid-contact ion-selective electrodes (SC-ISEs), often poses some difficulties26,27 despite its evident importance.1,10,28 This requirement has been recently addressed by introducing a well-controlled redox buffer layer between the ion-selective membrane and electrical contact29 and is also at the heart of our current work presented in this paper. The oxidation state of the conducting polymer (CP) film influences its ionic content and vice versa, and together, they determine the potentiometric response and standard potential of the film.30,31 Thus, pretreatment of CP films by applying a potential to convert all studied films to the same oxidation state has been found to improve the reproducibility of the standard potential of bare CP films in potentiometric stability measurements32 In our previous work, we reported on the possibility to adjust the redox-state of a CP-film covered with an ion-selective membrane in a solid-contact ISE by applying a nanoamp-range current or a potential to control the position of the calibration slope.33 Here we continue that work by introducing a simplified method to control E° of an ISE with a CP as the solid contact. Instead of using an external power supply (potentiostat) to adjust the redox-state of the conducting polymer, the SC-ISE is simply connected with a metallic wire (short-circuited) to a conventional reference electrode (RE) with internal filling solution and a large redox capacitance, in an electrolyte containing primary ions. This corresponds to a situation where 0 V is applied to an ISE with a potentiostat. Electrical current is thus flowing between the electrodes until their potential difference becomes close to zero (0 V). This is expected to influence the redox-state and ion-content of the CP solid contact under the ISM but leave the RE uninfluenced, as it is designed to tolerate passage of small currents by definition.34 Interestingly, we found in the literature that short-circuiting of two Ag/AgCl quasi-reference electrodes in order to render them symmetric and thus decreasing the deviation of their standard potentials to less than 30 μV was used by Rumpf et al.25 already in the early 1990s. To study this approach for SC-ISEs, we use poly(3,4ethylenedioxythiophene) doped with the bulky polyanion poly(sodium 4-styrenesulfonalte), PEDOT(PSS) as a solid contact, and three different types of ion-selective membranes

(ISM) to cover it: two potassium-selective membranes, with and without the lipohilic additive tetradodecylammonium tetrakis(4-chlorophenyl)borate (ETH-500) and a cationsensitive membrane without ionophore. A conventional macroscopic Ag/AgCl/3 M KCl electrode with or without a salt bridge was used as the reference electrode.

■

EXPERIMENTAL SECTION Chemicals. Valinomycin, potassium tetrakis(4chlorophenyl)borate (KTpClB), potassium tetrakis[3,5-bis(trifluoromethyl)phenyl]borate (KTFPB), tetradodecylammonium tetrakis(4-chlorophenyl)borate (ETH-500), 2-nitrophenyl octyl ether (o-NPOE), bis(2-ethylhexyl)sebacate (DOS), poly(vinyl chloride) (PVC) and tetrahydrofuran (THF) were purchased from Fluka and were of Selectophore purity grade. Poly(sodium 4-styrenesulfonate) (NaPSS, Mw ∼ 70 000), 3,4ethylenedioxythiophene (EDOT, 97%) and KCl (≥99.5%) were from Sigma-Aldrich. Deionized water (ELGA Purelab Ultra) was used throughout the experiments. Electrodes. Electrodes used in this study were prepared as described in our previous work.33,35 Disk-shaped working electrodes were manufactured by fitting a glassy carbon (GC) rod (Sigradur G, HTW Hochtemperatur−Werkstoffe GmbH, D-86672, Thierhaupten, Germany) in a PVC cylinder. The GC disk electrodes were carefully polished with 0.3 μm Al2O3 powder and cleaned chemically by immersion in 1 M HNO3 followed by ultrasonic cleaning in ethanol and in deionized water. Polymerization of PEDOT(PSS) films on GC disk electrodes (area = 0.07 cm2) was done galvanostatically with an Autolab general purpose electrochemical system (PGSTAT20, FRA2, AUTOLAB, Eco Chemie, B.V., The Netherlands) in a conventional three-electrode cell by applying a 0.014 mA current (0.2 mA/cm2) for 714 s in aqueous 0.1 M NaPSS + 0.01 M EDOT solution that was stirred overnight to guarantee proper dissolution of the monomer prior to electropolymerization.35 The solution was deaerated with nitrogen. A Metrohm double junction Ag/AgCl/3 M KCl reference with a 0.1 M NaPSS bridge was used for electropolymerization of PEDOT(PSS) films, and the counter electrode was a GC rod. After polymerization, the GC/PEDOT(PSS) electrodes were conditioned in 0.1 M KCl for 1 day, after which they were rinsed with water and let dry in air before drop-casting of the membrane cocktail. Three different membrane cocktails were prepared and dissolved in THF (15% dry content), resulting in the following membrane compositions (after evaporation of THF): (i) a potassium selective membrane without ETH-500 (1% valinomycin, 0.5%, KTFPB, 33.3% PVC, 65.2% DOS), (ii) a potassium selective membrane with ETH-500 (1% valinomycin, 0.5% KTFPB, 1% ETH-500, 32.5% PVC and 65% DOS) and B

dx.doi.org/10.1021/ac501464s | Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry

Article

induced by short-circuiting the SC-ISE with a conventional reference electrode (RE). Any open-circuit potential difference between the SC-ISE and the RE is thus expected to result in a current flow between the two electrodes under short-circuit, i.e., reduction/oxidation of the CP film (eq 1) as well as reduction/ oxidation of the RE (in this case, AgCl + e− ↔ Ag + Cl−), until equilibrium is reached and the potential difference between the short-circuited electrodes approaches zero. As long as the redox capacitance of the RE is much larger than the redox capacitance of the solid contact of the SC-ISE, the potential of the RE will remain practically constant during short-circuiting. The principle is demonstrated in practice in Figure 2, where three K+-selective SC-ISEs of type SC-ISE(i) are first calibrated

(iii) a cation-sensitive membrane without ionophore (1% KTpClPB, 33% PVC and 66% o-NPOE). Two to three replicate GC/PEDOT(PSS) electrodes were covered with each membrane by drop-casting 100 μL of the prepared cocktail on the electrode surface. THF was left to evaporate, resulting in SC-ISEs with membrane compositions (i), (ii) and (iii), which will be called SC-ISE(i), SC-ISE(ii) and SC-ISE(iii), respectively. The SC-ISEs were conditioned in 0.1 M KCl before and always in between measurements. Measurements. Potentiometric calibrations were done with a multichannel meter (Lawson EMF16 interface potentiometer, Lawson Laboratories, Inc.) using a Metrohm double junction Ag/AgCl/3 M KCl/1 M LiAc as a reference. Another similar double junction RE was connected as the working electrode in one of the measurement channels in all calibrations in order to verify proper functioning of the main reference electrode. Short-circuiting treatment was carried out with a simple metallic wire and crocodile clips that connected the studied SC-ISE to a conventional reference electrode in a solution of chosen KCl concentration (Figure 1a). Shortcircuiting the studied SC-ISE to another similar ISE (Figure 1b) or to one or two other similar SC-ISEs and a conventional RE simultaneously (Figure 1c) was also studied. The duration of short-circuiting was varied from overnight to several days, and calibrations (from 10−1 to 10−8 and back to 10−1 M KCl) of the studied electrodes were performed before and after each short-circuit treatment. Each calibration was started by first measuring the open-circuit potential (OCP) in 0.1 M KCl for 30 min, after which each proceeding solution was measured for 5 min before taking a data point for the calibration curve. The very first calibration was done before any short-circuit treatment. Electrochemical impedance spectroscopy (EIS) measurements were done with the same Autolab as described above using a conventional three-electrode cell with a GC rod counter electrode and an Ag/AgCl/3 M KCl reference electrode in 0.1 M KCl. The frequency range used was 100 kHz−10 mHz and the excitation amplitude was 10 mV. All EIS measurements were done in duplicate and for two identical electrodes in order to have information on reproducibility.

Figure 2. (a) Calibration slopes of three SC-ISE(i)s before any treatment (●) and after short-circuiting them together with each other and with a conventional Ag/AgCl/3MKCl RE in 0.1 M KCl for 3 days (Δ). (b) Calibration curves of the SC-ISE(i) (marked with an arrow in panel a) after short-circuiting with a conventional Ag/AgCl/3 M KCl reference electrode in 0.1 M KCl for 3 days (*) an additional overnight (○), and after 3 days at open-circuit followed by shortcircuiting again in 0.1 M KCl for 3 days (■).

without any short-circuiting treatment (filled circles), and then short-circuited in 0.1 M KCl simultaneously with each other and with a conventional macroscopic Ag/AgCl/3 M KCl RE (as shown schematically in Figure 1c) for 3 days followed by calibration (open triangles) in open-circuit conditions. It can be seen that originally all three SC-ISE(i)s have similar calibration slopes (from 10−1 M to 10−5 M KCl; average 59.3 ± 1.3 mV/ dec) but varying standard potential E° values. After the SCISE(i)s were short-circuited with each other and with the conventional RE in 0.1 M KCl for 3 days, the potential values for all three SC-ISE(i)s became close to equal. The slopes of the calibration curves remained practically unchanged after short-circuiting (from 10−1 M to 10−5 M KCl; average 58.2 ± 0.8 mV/dec). The results presented in Figure 2a reveal, however, that even if short-circuiting improved the similarity of electrode responses of the three SC-ISE(i)s with each other, the calibration curves after short-circuiting still do not superimpose perfectly, which may be attributed to slightly varying membrane resistances of individual electrodes due to manual fabrication. Nevertheless, when the short-circuit treatment for a single SC-ISE(i) (indicated with an arrow in Figure 2a) is repeated, it can be seen (Figure 2b) that its response is highly reproducible. Three calibration curves for the chosen SC-ISE(i) in Figure 2b are measured after short-circuiting with an Ag/AgCl/3 M KCl RE in 0.1 M KCl first for 3 days (*), then for an additional overnight (○) and after letting it first relax disconnected (open-circuit) in 0.1 M KCl for 3 days and then short-circuiting it again for 3 days with the same RE (■). In this case, the curves superimpose perfectly within the concentration range

■

RESULTS AND DISCUSSION The standard potential of SC-ISEs, where a conducting polymer (CP) is used as an ion-to-electron transducer (solid contact) and coated by an ion-selective membrane, can be adjusted electrochemically by applying a potential or current that shifts the redox state of the CP layer, as reported recently in our previous work.33 The shift of the redox state of the CP is shown in eq 1: CP+A−(film) + B+(ISM) + e− ↔ CP0A−B+(film)

(1)

where the subscript “film” denotes the conducting polymer phase (here PEDOT) and ISM the ion-selective membrane phase. A− represents a bulky counteranion (PSS− in this case) and B+ represents a small mobile cation (here K+). When the reaction (eq 1) occurs from left to right, the film is becoming more reduced; when the reaction occurs from right to left, the film is becoming more oxidized. In this work, instead of applying a potential with a potentiostat to the SC-ISE, the driving force for reduction/ oxidiation of the conducting polymer film under the ISM (and thus for changing the standard potential of the electrode) is C

dx.doi.org/10.1021/ac501464s | Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry

Article

from 10−1 to 10−5 M KCl. This behavior is representative for all three SC-ISE(i)s shown in Figure 2a. The adjusted calibration curves of SC-ISE(i)s shown in Figure 2 did not shift to 0 mV in 0.1 M KCl, which may be related to the high membrane resistance of tens of MΩ for the ion-selective membrane used in SC-ISE(i). Short-circuiting SCISE(ii) with a lower membrane resistance of only a few MΩ with the RE in 0.1 M KCl for overnight shifts its potential closer to zero (0 V) compared to SC-ISE(i) (Figure 3).

where the potential of two cation-sensitive SC-ISE(iii)s is measured in 0.1 M KCl for 30 min before and after shortcircuiting them together in 0.1 M KCl (shown schematically in Figure 1b) overnight or for 3 days. It can be seen that the potentials of the two short-circuited SC-ISE(iii)s become close to equal already after overnight short-circuit treatment (Figure 4, curves 2). Oxidation of the PEDOT(PSS) film (reaction in eq 1 from right to left) shifting the potential of SC-ISE(iii) in the positive direction seems to be favored compared to reduction, as the potential of SCISE(iii)a is increased to a larger extent than the potential of SCISE(iii)b is decreased. When short-circuiting is disconnected and the potential is measured, both electrodes drift slightly toward their original potentials (−54 μV/min and +68 μV/min after overnight short-circuiting). When the short-circuiting is carried on further, the potentials of both SC-ISE(iii) electrodes continue to increase (Figure 4, curves 3). After disconnection and letting the electrodes relax for 3 days, the potentials of the two SC-ISE(iii)s return toward their original values (Figure 4, dashed lines). These results show that the potentials of two SCISEs can be equalized via short-circuiting. However, because the redox capacitance of the solid contacts in the two SC-ISEs is comparable, short-circuiting changes the potentials of both SC-ISEs (Figure 4). It should be noted that the membrane resistance of SC-ISE(iii) is only a few hundred kilo-ohms, and the ionic contents and redox state of the CP film can thus easily be changed due to the fast ion transport through the membrane. Figure 5a shows representative results for the SC-ISE(iii) measured in 0.1 M KCl for 30 min after short-circuiting it with

Figure 3. Calibration curves of a SC-ISE (ii) before any shortcircuiting treatment (▼), after short-circuiting with an Ag/AgCl/3 M KCl reference electrode in 0.1 M KCl overnight (●), for 4 days (×) and after disconnecting the electrodes for several days and shortcircuiting it again with RE for 3 days (+).

Carrying out the short-circuiting for several days further decreases the potential difference with the conventional RE (Figure 3). As the oxidation/reduction of the CP film proceeds as a result of short-circuiting, the potential difference between the short-circuited electrodes decreases accompanied by a decrease in the driving force for the redox reaction. The adjustment of the last millivolts before reaching the equilibrium potential is very slow, as the current flowing is only in the order of few hundred picoamps. However, short-circuit treatment for a few days results in a highly reproducible response for SCISE(ii), as shown for one representative SC-ISE(ii) in Figure 3. It should be pointed out that the only way to control the standard potential in a reproducible manner with this method is to wait for equilibrium in the short-circuiting process. The process of equilibration slows down toward the end and must be given enough time in order to achieve a reproducible state. If the short-circuiting is done between two similar SC-ISEs, their resistances and redox capacitances are of similar magnitude and they both will be influenced by the drive to diminish the potential difference, unlike the situation with large-capacitance conventional RE. This is shown in Figure 4

Figure 5. (a) Open-circuit potential of a cation-sensitive SC-ISE(iii) measured in 0.1 M KCl vs Ag/AgCl/3 M KCl/1 M LiAc after shortcircuiting with a conventional Ag/AgCl/3 M KCl RE in 10−1 M KCl for 3 days, with Ag/AgCl/3 M KCl/0.1 M KCl in 10−2 M KCl for 3 days and in 10−3 for overnight and with Ag/AgCl/3 M KCl/1 M LiAc in 10−4 M KCl for overnight. (b) Calibration curves measured after the short-circuit treatments. The KCl concentrations used during shortcircuiting are indicated in the panels.

a conventional RE in 10−1, 10−2, 10−3 and 10−4 M KCl, respectively. The corresponding calibration curves measured after short-circuit treatments are shown in Figure 5b. No corrections were made for slight differences in the liquidjunction potentials of the reference electrodes (REs) used during short-circuiting (Figure 5). The results presented in Figure 5 illustrate that if a SC-ISE is short-circuited with a RE in 0.1 M KCl, the resulting opencircuit potential is close to 0 V in 0.1 M KCl. After short-circuit treatment in 10−2 M KCl, the open-circuit potential is close to 0 V in 10−2 M KCl, and so on for 10−3 and 10−4 M KCl. Correspondingly, SC-ISE(i) and SC-ISE(ii) were shortcircuited with a conventional RE in 10−1, 10−2 and 10−3 M KCl and calibrated afterward in open-circuit conditions. The results for SC-ISE(ii) are shown in Figure 6. The slopes for all

Figure 4. Open-circuit potential of two similar SC-ISE(iii)s (black and gray lines, respectively) measured in 0.1 M KCl before any treatment (1), after short-circuiting them with each other in 0.1 M KCl overnight (2) or for 4 days (3) and after disconnecting the short-circuit wire and letting them relax for 3 days (4) - dashed lines. D

dx.doi.org/10.1021/ac501464s | Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry

Article

spectra show a semicircle that can be attributed to the bulk resistance and geometric capacitance of the ion-selective membrane and a small “tail” from the solid-contact layer, in good agreement with earlier results for similar SC-ISEs.35 The diameter of the semicircle (i.e., the resistance of the ionselective membrane) increased during conditioning in 0.1 M KCl, independent of the short-circuit treatment (Figure 7). Such changes in bulk membrane resistance have been observed earlier for plasticized PVC membranes.36 Importatly, the EIS results show that short-circuiting did not cause any changes in the impedance spectra. These results confirm that shortcircuiting for several days has no adverse effects on the electrochemical properties of the SC-ISEs used. There is great promise in this simple method to control the E° of SC-ISEs, but also some challenges that should be considered. To obtain a reproducible potential, the conditions during short-circuit treatment should be carefully controlled. The concentration of the short-circuiting electrolyte, determining the position of zero potential, may change due to evaporation during long treatments that are required to reach equilibrium (up to 3 days) and, therefore, proper sealing of the system is required. Also, when short-circuiting is carried out with a conventional RE, the traditional problems related to evaporation and leakage of the internal filling solution must be addressed. The drift toward the original potential after shortcircuiting should be defined for specific short-circuiting conditions and it should be taken into account when performing analysis with SC-ISEs whose E° has been reset by short-circuiting. The potential drifts of the studied SC-ISEs in an open-circuit directly after short-circuiting were typically 0.1− 0.4 mV/min, which should be taken into account in analytical applications. Such a small drift may be acceptable when the analysis is performed directly after short-circuiting and in potentiometric measurements done for screening purposes. In applications where the sensor is used repeatedly for long times, short-circuiting should be done between measurements, but this would require a means of supplying a solution with known concentration of primary ion (calibration solution) during short-circuiting. In this case, short-circuiting would resemble a one-point calibration, the advantage of short-circuiting being that it is really instrument free. The short-circuiting approach would be ideally suited for single-use sensors that can be kept short-circuited in a sealed package containing a solution with a known concentration of primary ion. After the package is opened (and simultaneously the short-circuit opened), the sensor should be used immediately. This is an application where the short-circuiting method would be most useful and powerful. However, before this can become a reality, there is still a need to evaluate the impact of short-circuiting on SC-ISEs and reference electrodes over long times (months−years).

Figure 6. (a) Open-circuit potentials in 0.1 M KCl and (b) calibration curves of SC-ISE (ii) after short-circuiting it with an Ag/AgCl/3 M KCl reference electrode in 0.1 M KCl for 3 days (■), with an Ag/ AgCl/3 M KCl/0.1 M KCl in 10−2 M KCl for 3 days (Δ) and with an Ag/AgCl/3 M KCl/0.1 M KCl in 10−3 M KCl for 3 days (●). The KCl concentrations used during short-circuiting are indicated in the panels.

SC-ISEs remained unchanged. The experiments in Figures 5 and 6 were done with two replica electrodes, and the results shown are representative. The same experiment was done also with SC-ISE(i), and the trend in behavior was the same for all studied SC-ISEs. All results indicate that the potential of SCISEs short-circuited with a conventional RE approaches zero in the short-circuiting electrolyte, and the magnitude and direction of potential adjustment is defined by the initial open-circuit potential (OCP) at the short-circuiting electrolyte−ion-selective membrane interface vs the RE. In the short-circuit cases discussed in this work, the reference electrode is typically a single-junction Ag/AgCl/3 M KCl or a double-junction RE with an additional salt bridge. All the SCISEs studied have higher original potentials in 0.1 M KCl compared to these REs, which means that short-circuiting them in 0.1 M KCl induces reduction of the PEDOT(PSS) film. If short-circuiting is done in more diluted solutions, the shift from the original potential of the electrode is smaller, and in some cases (e.g., for SC-ISE(iii) in Figure 5), the original potentials in 10−3 and 10−4 M KCl are lower than that of the RE, which means that the short-circuiting induces oxidation of the PEDOT(PSS) film. Electrochemical impedance spectroscopy (EIS) measurements were done for all types of SC-ISEs studied in this work. Impedance spectra recorded for SC-ISE(ii) at open-circuit potentials after different conditioning times with and without short-circuiting are illustrated in Figure 7. The impedance

■

CONCLUSIONS A simple, instrument-free approach to control the standard potential (E°) of SC-ISEs with a conducting polymer as a solid contact is presented. By short-circuiting this type of SC-ISE to a conventional RE, the potential of the SC-ISE approaches the potential of the RE due to reduction/oxidation of the conducting polymer used as a solid contact, involving transport of charge-balancing ions to/from the conducting polymer film until the potential difference between the RE and SC-ISE is close to zero. The kinetics of the equilibration process at shortcircuit depends on the properties of the ion-selective

Figure 7. Electrochemical impedance spectrum recorded for SCISE(ii) at open-circuit potential after 5 days (○) and 14 days (●) conditioning in 0.1 M KCl and at 0 mV after short-circuiting with RE overnight (Δ) and for 3 days (□), when electrode had been conditioned for 15 and 17 days, respectively. E

dx.doi.org/10.1021/ac501464s | Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry

Article

(18) Schelter, M.; Zosel, J.; Oelssner, W.; Guth, U.; Mertig, M. Sens. Actuators, B 2013, 187, 209−214. (19) Marei, M. M.; Roussel, T. J.; Keynton, R. S.; Baldwin, R. P. Anal. Chim. Acta 2013, 803, 47−55. (20) Kahlert, H.; Steinhardt, T.; Behnert, J.; Scholz, F. Electroanalysis 2004, 16, 2058−2064. (21) Sieg, A.; Guy, R. H.; Begoña Delgado-Charro, M. Biophys. J. 2004, 87, 3344−3350. (22) Michalska, A.; Ocypa, M.; Maksymiuk, K. Electroanalysis 2005, 17, 327−333. (23) Kisiel, A.; Michalska, A.; Maksymiuk, K. Bioelectrochemistry 2007, 71, 75−80. (24) Novell, M.; Parrilla, M.; Crespo, G. A.; Rius, F. X.; Andrade, F. J. Anal. Chem. 2012, 84, 4695−4702. (25) Rumpf, G.; Spichiger-Keller, U.; Buehler, H.; Simon, W. Anal. Sci. 1992, 8, 553−559. (26) Michalska, A. Electroanalysis 2012, 24, 1253−1265. (27) Bobacka, J.; Ivaska, A.; Lewenstam, A. Chem. Rev. 2008, 108, 329−351. (28) Ivanova, N. M.; Levin, M. B.; Mikhelson, K. N. Russ. Chem. Bull. 2012, 61, 926−936. (29) Zou, X. U.; Cheong, J. H.; Taitt, B. J.; Buhlmann, P. Anal. Chem. 2013, 85, 9350−9355. (30) Lewenstam, A.; Bobacka, J.; Ivaska, A. J. Electroanal. Chem. 1994, 368, 23−31. (31) Bobacka, J.; Gao, Z.; Ivaska, A.; Lewenstam, A. J. Electroanal. Chem. 1994, 368, 33−41. (32) Lindfors, T. J. Solid State Electrochem. 2009, 13, 77−89. (33) Vanamo, U.; Bobacka, J. Electrochim. Acta 2014, 122, 316−321. (34) Oldham, K. B.; Myland, J. C. In Fundamentals of Electrochemical Science; Academic Press, Inc.: San Diego, CA, 1994; pp 115, 190. (35) Bobacka, J. Anal. Chem. 1999, 71, 4932−4937. (36) Horvai, G.; Gráf, E.; Tóth, K.; Pungor, E.; Buck, R. P. Anal. Chem. 1986, 58, 2735−2740.

membrane, but if continued until equilibrium is reached, it is reproducible for each electrode. The concentration of primary ion in the solution where short-circuiting is performed determines the equilibrium potential. The method presented here offers a very promising way to control E° of SC-ISEs and a possibility to simplify calibrations in the future. In principle, the short-circuiting method should work for any solid-contact ISE where the capacitance of the ion-to-electron transducer is much smaller than that of the reference electrode. The beauty of the method is its simplicity. In particular, the short-circuiting method opens up new possibilities to control the initial potential of single-use potentiometric sensors, which is an important step toward calibration-free potentiometric analysis. This may accelerate further development and applications of disposable potentiometric sensors in the future.

■

AUTHOR INFORMATION

Corresponding Author

*J. Bobacka. Tel.: +358 2 215 3246. E-mail: johan.bobacka@ abo.fi. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS Financial support from the Graduate School of Chemical Sensors and Microanalytical Systems (CHEMSEM) and Åbo Akademi University is gratefully acknowledged.

■

REFERENCES

(1) Lindner, E.; Gyurcsanyi, R. E. J. Solid State Electrochem. 2009, 13, 51−68. (2) Bakker, E.; Bhakthavatsalam, V.; Gemene, K. L. Talanta 2008, 75, 629−635. (3) Diamond, D. Anal. Chem. 2004, 76, 278A−286A. (4) Radu, A.; Anastasova-Ivanova, S.; Paczosa-Bator, B.; Danielewski, M.; Bobacka, J.; Lewenstam, A.; Diamond, D. Anal. Methods 2010, 2, 1490−1498. (5) Bakker, E.; Chumbimuni-Torres, K. J. Braz. Chem. Soc. 2008, 19, 621−629. (6) Lewenstam, A. In Comprehensive Analytical Chemistry Vol. 49; Alegeret,S., Merkoci, A., Eds.; Elsevier:Amsterdam, The Netherlands and Oxford, U. K., 2007; pp 5−24. (7) Wang, J. Chem. Rev. 2008, 108, 814−825. (8) Ge, X.; Kostov, Y.; Tolosa, L.; Rao, G. Anal. Chim. Acta 2012, 734, 79−87. (9) Grygolowicz-Pawlak, E.; Bakker, E. Anal. Chem. 2010, 82, 4537− 4542. (10) McElhoney, K.; O’Neil, G. D.; Chaniotakis, N. A.; Kounaves, S. P. Electroanalysis 2012, 24, 2071−2078. (11) Radu, A.; Anastasova, S.; Fay, C.; Diamond, D.; Bobacka, J.; Lewenstam, A. Sensors, 2010 IEEE, Kona, HI, November 1−4, 2010; pp 1487−1490. (12) Grygolowicz-Pawlak, E.; Bakker, E. Electrochem. Commun. 2010, 12, 1195−1198. (13) DeGrandpre, M. D.; Baehr, M. M.; Hammar, T. R. Anal. Chem. 1999, 71, 1152−1159. (14) Anastasova, S.; Radu, A.; Matzeu, G.; Zuliani, C.; Mattinen, U.; Bobacka, J.; Diamond, D. Electrochim. Acta 2012, 73, 93−97. (15) Reddy, K. A.; George, B.; Mohan, N. M.; Kumar, V. J. Instrum. Meas., IEEE Trans. 2009, 58, 1699−1705. (16) Boalth, N.; Wandrup, J.; Larsson, L.; Frischauf, P. A.; Lundsgaard, F. C.; Andersen, W. L.; Jensen, N.; Singer, R.; Troldborg, C. P.; Lunding, G. Clin. Chim. Acta 2001, 307, 225−233. (17) Noh, J.; Park, S.; Kim, H. C.; Chung, T. D. Bull. Korean Chem. Soc. 2010, 31, 3128−3132. F

dx.doi.org/10.1021/ac501464s | Anal. Chem. XXXX, XXX, XXX−XXX