Article pubs.acs.org/ac
Application of Nanostructured TCNQ to Potentiometric Ion-Selective K+ and Na+ Electrodes Beata Paczosa-Bator,* Magdalena Pięk, and Robert Piech AGH-UST University of Science and Technology, Faculty of Material Science and Ceramics, Mickiewicza 30, PL-30059 Cracow, Poland ABSTRACT: A new type of potentiometric solid-state ionselective electrode (SS-ISE) has been fabricated with an intermediate layer made of 7,7,8,8-tetracyanoquinodimethane (TCNQ) or its ion-radical salts and an ionophore-based ionselective membrane. To show the influence of the TCNQ layer on electrode selectivity, sodium- and potassium-sensitive membranes were applied. A good Nernstian response with a slope of 59.24 mV/ dec in the range from 10−6.5 to 10−1 M KCl and 58.68 mV/dec (10−6 to 10−1 M NaCl) was observed. The influence of an interfacial water film was assessed by an aqueous-layer test performed during potentiometric measurements. The stability of the electrical potential of the new solid-contact electrodes was tested by performing current-reversal chronopotentiometry, and the capacitance of the electrodes is 132 μF or 154 μF for K+ and Na+ solid-contact electrodes. These properties confirmed the analytical applicability of TCNQ-based SC-ISEs and should allow the development of a new solid-state ion sensor group. In this paper we present analytical characteristics of the first potentiometric solid-state electrode with TCNQ or the ion radical salt of TCNQ (potassium or sodium) and a polymeric K+- and Na+-selective membrane with significant long-term potential stability, potential repeatability, and improved detection limit. These electrodes exhibit excellent Nernstian response without light or oxygen sensitivity.
O
rganic materials have attracted significant attention over the past five decades because of their interesting and unusual electrical, magnetic, and optical properties,1 rendering them promising for utilization in industrial and technological applications. One of the most popular organic compounds whose materials have excellent electrical performances is 7,7,8,8-tetracyanoquinodimethane (TCNQ), discovered in 1960.2 Owing to high electron affinity, TCNQ can form organic charge-transfer complexes and ion-radical salts of both simple and complex composition. TCNQ salts such as lithium, sodium, or potassium can be prepared by any of the usual methods, e.g., that proposed by Melby.3 TCNQ-based materials are used nowadays in organic solar cells,4 transistors,5 or light-emission diodes.6 Already TCNQ has been successfully used in electrochemical sensors as voltammetric and potentiometric ion-selective electrode (ISE) active material.7,8 In a voltammetric approach the TCNQmodified electrode operates under dynamic kinetically controlled conditions9 whereas in potentiometric detection the signal is generated by charge separation at the TCNQ radical ion salt interface due to the selective partitioning of the ionic species between the TCNQ layer and the primary ion solution. Potentiometric ISEs with a solid-contact configuration (SCISEs) have made very rapid progress in recent decades. Different novel materials have been introduced between the electrode support and the ion-selective polymeric membrane to improve the potential stability and reproducibility, limit the formation of the interfacial water film, and reduce the sensitivity to oxygen or light of SC-ISEs.10−20 The detection limits of SCISEs have been also improved. There is a large range of potentiometric sensors that are capable for direct measurements in the nanomolar concentration range.21−23 © XXXX American Chemical Society
■
EXPERIMENTAL SECTION Materials. 7,7,8,8-Tetracyanoquinodimethane 98% (TCNQ), potassium ionophore I (valinomycin), sodium ionophore IV, bis(1-butylpentyl) adipate (BBPA), tris(2ethylhexyl) phosphate (TEHP), potassium tetrakis(4chlorophenyl)borate (KTpClPB), o-nitrophenyl octyl ether (o-NPOE), poly(vinyl chloride) (PVC) of high molecular weight, acetone, acetonitrile, and tetrahydrofuran were the selectophore reagents obtained from Sigma-Aldrich. All other chemicals were of analytical-reagent grade. Distilled and deionized water was used to prepare the aqueous solutions. Electrode Preparation. The drop-casting method was used to obtain the modified glassy carbon disc (GCD) electrodes used for ion sensing. First, the GCD electrodes (Mineral, Poland) were polished with 0.3 μm alumina powder, rinsed with water, and finally cleaned ultrasonically with water and methanol. Then was added the intermediate layer based on TCNQ or the potassium (sodium) salt of TCNQ, and then the electrode was covered with the polymeric membrane sensitive for K+ or Na+. Figure 1 shows schematically the structure of the Received: September 19, 2014 Accepted: December 31, 2014
A
DOI: 10.1021/ac503521t Anal. Chem. XXXX, XXX, XXX−XXX
Article
Analytical Chemistry
Five identical electrodes were prepared and examined. All the electrodes were separately conditioned and stored. Potentiometric Measurements. The potentials were measured using an EMF 16-channel meter (Lawson Labs, Inc., Malvern, PA). The reference electrode was an Ag/AgCl electrode with 3 M KCl solution in a bridge cell (type 6.0733.100 ΩMetrohm, Switzerland) or an Ag/AgCl/3 M KCl (type 6.0729.100 ΩMetrohm, Switzerland) with a 1 M lithium acetate salt bridge. Chronopotentiometry and Cyclic Voltammetry. The chronopotentiometry and cyclic voltammetry measurements were performed with the use of an Autolab General Purpose Electrochemical System (AUT32N.FRA2-AUTOLAB, Eco Chemie, The Netherlands) connected to a conventional, three-electrode cell. The Ag/AgCl/3 M KCl electrode (type 6.0733.100 ΩMetrohm, Switzerland) was used as a reference, and a glassy carbon rod was used as an auxiliary electrode. The electrolyte solution was purged with nitrogen for 10 min prior to each measurement. Scanning Electron Microscopy. The morphologies and chemical analysis of the chemically synthesized K(TCNQ) or Na(TCNQ) and TCNQ layer before and after electrochemical synthesis of the potassium or sodium salt were examined using a scanning electron microscope (SEM), model LEO 1530, from LEO Electron Microscopy Ltd. equipped with the Image and X-ray Analysis System, model Vantage, from ThermoNoran for energy-dispersed X-ray spectroscopy (EDS).
Figure 1. Schematic representation of the potentiometric potassiumand sodium-selective electrode.
TCNQ-, K(TCNQ)-, and Na(TCNQ)-modified polymeric ion-selective electrodes. In the case of the TCNQ-modified electrodes (GCD/TCNQ) a 25 μL solution containing 1 mg of TCNQ dispersed in 1 mL of acetone was placed on top of the GCD electrodes. To obtain modified GCD electrodes with the potassium or sodium salt of TCNQ (GCD/K(TCNQ) and GCD/Na(TCNQ)), chemical or electrochemical synthesis of K(TCNQ) and Na(TCNQ) was performed. The first group of electrodes (GCD/K(TCNQ)I and GCD/Na(TCNQ)I) was prepared by covering GCD electrodes with a 25 μL coating solution containing 1 mg of chemically prepared K(TCNQ) or Na(TCNQ) dispersed in 1 mL of acetone. Chemical synthesis of K(TCNQ) and Na(TCNQ) was carried out according to the methods of Melby.3 Potassium or sodium iodide was present in a 2:1 molar ratio to discourage the back reaction. KI or NaI dissolved in methanol was added to a hot solution of TCNQ in acetonitrile. The reaction mixture was washed with acetonitrile. The second group of K- or Na(TCNQ)-modified GCD electrodes (GCD/K(TCNQ)II and GCD/Na(TCNQ)II) were prepared by drop casting a 25 μL solution containing 1 mg of TCNQ dissolved in 1 mL of acetone, and after allowing the solvent to evaporate, electrochemical synthesis of K(TCNQ) or Na(TCNQ) was performed. Electrochemical reduction of solid TCNQ in the presence of potassium or sodium ions was performed in 0.1 M KCl or NaCl (pH = 6.8) by cycling the potential at a scan rate of 20 mV s−1 between +0.45 V and −0.1 V (vs Ag/AgCl/3 M KCl) during 50 cycles. To prepare the potassium-selective electrodes, solid-contact layers (GCD/TCNQ and GCD/K(TCNQ)) were subsequently coated with a 30 μL THF solution containing 1.1% (w/w) valinomycin, 0.25% (w/w) KTpClPB, 65.65% (w/w) oNPOE, and 33% (w/w) PVC. The Na-SCISEs were obtained by twice adding a 15 μL THF solution containing 2.9% (w/w) sodium ionophore IV, 0.4% (w/w) KTpClPB, 65.6% (w/w) (BBPA), 2.1% (w/w) TEHP, and 29% (w/w) PVC on top of the GCD/TCNQ and GCD/Na(TCNQ) electrodes. The coated disc electrodes (GCD/K+-ISM and GCD/Na+ISM) were prepared by covering bare GCD electrodes with the above-mentioned potassium or sodium-selective membrane. After the electrodes were covered with the ion-selective membrane (ISM), they were left in the air for 24 h to ensure complete evaporation of THF. Afterward the potassiumselective electrodes were conditioned in aqueous 0.001 M KCl solutions, and sodium-selective electrodes were conditioned in 0.001 M NaCl, respectively, for 24 h. The conditioning step was also repeated before every measurement.
■
RESULTS AND DISCUSSION Figure 2 shows exemplary cyclic voltammograms obtained at a scan rate of 20 mV s−1 over the potential range of +0.45 V and −0.1 V with the bare GCD, GCD/TCNQ, and GCD/ K(TCNQ)I electrodes when placed in a solution of 0.1 M KCl. As expected, potassium or sodium ions do not undergo any redox processes in the potential window used (Figure 2 inserts); therefore, the electrochemical activity of the GCD/ TCNQ electrode seen in Figure 2 is attributed to the reduction of TCNQ to TCNQ− and the introduction of ions into the crystal lattice. The difference in the shape of the oxidation and reduction processes is attributed to the higher conductivity of K(TCNQ) compared with TCNQ.24,25 The characteristic current loops in the cyclic voltammograms (Figure 2a insert) observed only in the case of the TCNQ-modified electrode and the large peak-to-peak separation (ΔEp = Epox − Epred = 250 mV) are accompanied by crystal fragmentation and rearrangement during the initial cycles of potential and are characteristic for an electrochemically irreversible TCNQ-K(TCNQ) (or TCNQ-Na(TCNQ)) phase conversion governed by nucleation−growth kinetics.24,25 In EDS analysis of the TCNQ, K(TCNQ), and Na(TCNQ), the elements C and N were clearly identified. In the case of the chemically and electrochemically synthesized K(TCNQ) and Na(TCNQ) samples, an additional K (or Na, respectively) peak appeared in the EDS spectrum that represents the K(TCNQ) molecule (or Na(TCNQ)). Exemplary SEM images showing the morphology of the TCNQ and chemically and electrochemically synthesized K(TCNQ) are presented in Figure 3. Figure 3a shows a typical rhombus-shaped crystal of TCNQ. The SEM image of chemically synthesized K(TCNQ) in Figure 3b shows a characteristic needle-shaped crystal. Figure 3c shows that electrochemical reduction of TCNQ resulted in needle-shaped B
DOI: 10.1021/ac503521t Anal. Chem. XXXX, XXX, XXX−XXX
Article
Analytical Chemistry
sodium-selective electrodes were as follows: 58.72 (58.70) mV/decade (10−6 to 10−1 M NaCl) for the TCNQ/Na+-ISM, 57.30 (57.15) mV/decade (10−5 to 10−1 M NaCl) for the Na (TCNQ)I/Na+-ISM, and 58.65 (58.62) mV/decade (10−5.8 to 10−1 M NaCl) for the Na (TCNQ)II/Na+-ISM. With the use of the same experimental conditions, the electrodes based only on the layer of TCNQ and TCNQ potassium or sodium salts showed calibration curves with the slope significantly lower than the theoretical value (see Table 1). Typical potentiometric responses of TCNQ and K(TCNQ)- or Na(TCNQ)-modified electrodes are shown in Figure 6. As the TCNQ-based GCD electrode behavior is similar to that recorded for a bare GCD electrode (dotted line in Figure 6), it is suggested that some EMF increase observed above 10−3 M KCl or NaCl is attributed to the redox response of electrodes. The fabricated electrodes show a very stable response over time. Even after a long conditioning time in 0.001 M KCl or NaCl (4−6 weeks), the electrodes still showed a linear response in the same range of the K+ and Na+ activity. Figure 7 shows the exemplary log aK+ dependence concerning the following electrodes: GCD/TCNQ/K+-ISM (i), GCD/K(TCNQ)II/K+ISM (ii), GCD/K(TCNQ)I/K+-ISM (iii), and GCD/K+-ISM (iv) obtained by recording the EMF. The reproducibility of the basic analytical parameters was also studied. For this purpose, five identical electrodes of each type were prepared by applying the same conditions. The obtained data are collected in Table 1. The detection limit was calculated as the activity of K+ at the point of intersection of the two slope lines.26 Interestingly, in the case of the SC-ISEs sensors based on the TCNQ layers, a very wide range of linear response was found with the detection limit of 10−6.9 M KCl for the GCD/TCNQ/K+-ISM electrode and 10−6.3 M NaCl for the GCD/TCNQ/Na+-ISM electrode. The obtained values are better than those for carbon black (10−6.4 M KCl and 10−5.1 M NaCl),18,27 graphene (10−5 M KCl),17 or colloid-imprinted mesoporous carbon (10−5.4 M KCl)20 as solid contact. The potential drift of the GCD/TCNQ/K+-ISM electrode over 172 h was only 11.1 ± 1.5 μV/h. Concerning the K(TCNQ)-modified PVC, the potential drift over 172 h was 18.0 ± 2.0 μV/h (GCD/K(TCNQ)I/K+-ISM) and 15.0 ± 1.7 μV/h (GCD/K(TCNQ)II/K+-ISM). The obtained values are similar to those observed for carbon materials used as the intermediate layer.11,19 In the case of sodium-selective electrodes, the potential drift was 9.2 ± 1.1 μV/h (GCD/TCNQ/Na+-ISM), 15.0 ± 1.3 μV/ h (GCD/Na(TCNQ)I/Na+-ISM), and 11.2 ± 1.2 μV/h (GCD/Na(TCNQ)II/Na+-ISM). The potentiometric selectivity coefficients Kijpot, where i is the primary ion (K+ or Na+) and j is the interfering ion, were obtained from the response potentials in an alkali metal or alkaline earth metal chloride solutions using the separate solution method (SSM) according to the traditional procedure28 (n = 3). Exemplary potentiometric selectivity coefficient values observed for the studied electrodes are presented in Figure 8. The values of Kijpot obtained for the coated with ISM GCD electrodes are presented in gray color. In the case of the TCNQ-based SC-ISEs sensors no significant changes in potentiometric selectivity coefficients were observed when compared to those received for the GCD-ISM electrodes. However, the presence of the K(TCNQ) and Na(TCNQ) layers in the GCD/K(TCNQ)/K+-ISM and GCD/Na(TCNQ)/Na+-ISM electrodes influenced on the Kij values
Figure 2. Cyclic voltammograms obtained for cycles of the potential over the range of +0.45 V and −0.1 V at (a) a TCNQ-modified GCD electrode and (b) a GCD electrode with chemically synthesized K(TCNQ).
microcrystals. Similar results were obtained in the case of Na(TCNQ) samples. The potentiometric response of the potassium and sodium sensors was studied in the concentration range 10−7 to 10−1 M of the main ion (K+ or Na+, respectively). The response reversibility of the presented electrodes recorded in the main ion solutions was very high and is shown in Figure 4. The response times of the TCNQ-modified electrodes for the progressive additions of different amounts of potassium or sodium ions were similar to those obtained for sensors based on carbon materials12−15,17−19 or conducting polymers16 and equals 4−5 s even at a low concentration of potassium or sodium ions. The calibration curve slope values measured for the developed electrodes with TCNQ, K(TCNQ), and Na(TCNQ) used as an intermediate layer are close to the Nernstian value. Figure 5 shows the dependency of log aK+ and log aNa+ on electromotive force (EMF) for all the studied solidcontact K+-ISM and Na+-ISM electrodes. The measurements were started at 10−7 M KCl (or NaCl) and continued successively up to 10−1 M KCl (or NaCl) and then back to 10−7 M KCl (or NaCl). The slopes of the linear part of the calibration curves of the potassium-selective electrodes were as follows: 59.20 (59.19) mV/decade (10−6.5 to 10−1 M KCl) for the TCNQ/K+-ISM, 57.81 (57.78) mV/ decade (10−5.5 to 10−1 M KCl) for the K(TCNQ)I/K+-ISM, and 58.60 (58.58) mV/decade (10−6 to 10−1 M KCl) for the K(TCNQ)II/K+-ISM. The slope values obtained for the C
DOI: 10.1021/ac503521t Anal. Chem. XXXX, XXX, XXX−XXX
Article
Analytical Chemistry
Figure 3. SEM images of (a) TCNQ and K(TCNQ) crystals which were prepared by (b) chemical and (c) electrochemical reduction of TCNQ, respectively.
the Na(TCNQ)/Na+-ISM electrodes showed improved selectivity for potassium ions. The redox sensitivity measurements were carried out for the TCNQ-modified electrodes covered with K+-ISM and Na+ISM, as well as for the electrodes based only on the TCNQ layers. The solution used contained FeCl3 and FeCl2 redox couple at a total concentration 1 mM with the log of Fe3+/Fe2+ ratio equal to −1, −0.5, 0, 0.5, and 1 and constant ionic background of 0.1 M KCl (or NaCl) (Figure 9). The GCD/ TCNQ, GCD/Na(TCNQ), and GCD/K(TCNQ) electrodes show a clear redox response. However, after covering the TCNQ, Na(TCNQ), or K(TCNQ) layer with PVC ionselective membrane, no redox sensitivity was observed. Similar behavior was observed earlier for the electrodes with graphene, carbon black, or PtNP intermediate layer.10,12−19 The light sensitivity of the developed electrodes was also studied. The potentials of the each electrode were measured continuously during the light sensitivity test. The electrodes were kept in the dark before starting the experiment. The following measuring sequence was used in the test: darkness (no room light during 5 min), room light (5 min), and then a return to darkness (no room light during 5 min again). EMF of the cell was recorded in 0.01 M KCl solution. As illustrated in
Figure 4. Potentiometric response of the TCNQ-modified PVC electrodes (i: TCNQ/K+-ISM, ii: K(TCNQ)II/K+-ISM, iii: K(TCNQ)I/K+-ISM) and the coated disc electrode (iv) vs time determined in the KCl solutions.
especially for those determined for potassium and sodium ions. Interesting, the application of potassium salt of TCNQ resulted in lowering the selectivity toward sodium ions and vice versa D
DOI: 10.1021/ac503521t Anal. Chem. XXXX, XXX, XXX−XXX
Article
Analytical Chemistry
Figure 5. EMF dependence on K+ and Na+ activities for (a) the TCNQ-modified PVC electrode, (b) the chemically synthesized K(TCNQ) or Na(TCNQ)-modified PVC electrode, and (c) the electrochemically synthesized K(TCNQ) or Na(TCNQ)-modified PVC electrode after 72 h of conditioning in a solution of 10−3 M KCl or 10−3 M NaCl.
M NaCl. Finally, the solution was changed again to 0.01 M KCl (see the example of the GCD/TCNQ/K+-ISM electrode in Figure 11). The dynamic EMF response was analyzed in terms of the potential drifts upon the primary ions being replaced by the interfering ions. Interestingly, the TCNQ-, Na(TCNQ)-, and K(TCNQ)-modified PVC electrodes are stable upon sample changing despite being in contact with the conditioning solution for a longer period of time (1 month). This proves that the water layer was reduced in the developed electrodes due to the hydrophobic character of TCNQ.30 Current-reversal chronopotentiometry was used to evaluate the electric capacity of the solid contact, the potential stability, and resistance of the developed electrodes according to the reported protocol.31 Figure 12 shows a typical change in the potential vs time recorded for the studied electrodes when a constant current of +1 nA was applied to the working electrodes for 60 s followed by a current of −1 nA applied for the same length of time. The changes in the polarization of the electrode were repeated three times as shown in Figure 12. For comparison, the GCD/TCNQ and coated disc electrodes were also tested. The potential jump observed in the response after current polarity change (ΔEdc) was used to calculate the total resistance of the electrode (Rtotal = ΔEdc/2I, where the I equals 1 nA). The potential drift of the electrodes was derived from the dEdc/dt ratio, and the capacity of the sensors was calculated from the equation dEdc/dt = I/C. All parameters determined as the average of three measurements are summarized in Table 2. The resistance of the TCNQ- or K,Na(TCNQ)-modified GCD electrodes generally decreased after covering with a PCV ion-selective membrane. On the other hand, the coated disctype electrode (GCD/ISM) resistance is considerably reduced after application of the TCNQ layer. This indicates that potassium or sodium ion transport is facilitated when both layers are applied in the sensor. The lowest value of Rtotal was observed in the case of electrodes with the chemically prepared K,Na(TCNQ). The obtained for developed electrode potential drift values are little higher than those obtained from the PtNPs-VXCbased sensors (4.6 μV/s)19 or the sensors with colloidimprinted mesoporous carbon as solid contact (1.0 μV/s)20 and lower than those obtained from the graphene-based sensors (12 μV/s),17 SC-ISEs with nanotubes (17 μV/s),14 and coated disc electrodes developed under similar conditions (see Table 2). The capacity of the GCD/TCNQ/K+-ISM electrode is calculated to be 132 μF and GCD/Na(TCNQ)/Na+-ISM electrode −154 μF. These values are much better than those measured for CPs-CNT (83 μF14), graphene (83 μF17), Printex XE-2 carbon black (51 μF18), or platinum nanoparticle (82 μF10) based solid-contact electrodes and worse than 1.0 mF calculated for sensors with colloid-imprinted mesoporous carbon.20
the potassium electrode example in Figure 10, no significant potential drift was observed during measurement. The presence of an unfavorable water layer between the ionselective membrane and the solid contact was tested by recording the dependency of EMF vs time when the primary ions are substituted by interfering ions, according to Fibbioli et al.29 The studied electrodes were initially conditioned in the primary ion solution (0.01 M KCl). After 3 weeks, the solution was changed to 0.01 M NaCl and then to 0.01 M KCl and 0.1
CONCLUSIONS Potentiometric solid-state sensors based on conducting organic material (TCNQ) and its cation-radical salts were synthesized and studied as K+- and Na+-selective electrodes. These novel sensors, prepared in a simple drop-casting method, showed a good Nernstian response to potassium or sodium in a wide linear range of primary ions, and the best resulting detection limit for the proposed potassium-selective electrode is 10−6.9 M and for the Na+-selective electrode 10−6.3 M. The presence of
■
E
DOI: 10.1021/ac503521t Anal. Chem. XXXX, XXX, XXX−XXX
Article
Analytical Chemistry Table 1. Reproducibility of the Basic Analytical Parameters of Different Potassium and Sodium Sensors type of electrode GCD/: Potassium Sensors TCNQ/K+-ISM K(TCNQ)I/K+-ISM K(TCNQ)II/K+-ISM K+-ISM TCNQ K(TCNQ)I K(TCNQ)II Sodium Sensors TCNQ/Na+-ISM K(TCNQ)I/Na+-ISM K(TCNQ)II/Na+-ISM Na+-ISM TCNQ Na(TCNQ)I Na(TCNQ)II
slope (mV/dec) ± SD (n = 5)
E0 (mV) ± SD (n = 5)
detection limit (M) ± SD (n = 5)
linear range (M)
59.24 57.62 58.58 57.84 25.42 42.14 36.84
± ± ± ± ± ± ±
0.08 0.43 0.18 1.10 1.55 1.23 1.45
398.8 340.4 379.1 322.7 308.4 148.4 182.3
± ± ± ± ± ± ±
1.4 6.6 5.7 28.1 11.0 12.2 12.4
10−6.9±0.1 10−5.9±0.1 10−6.5±0.1 10−5.5±0.4 10−3.3±0.9 10−4.2±0.7 10−3.8±0.5
10−6.5−10−1 10−5.5−10−1 10−6−10−1 10−5−10−1 10−3−10−1 10−4−10−1 10−3.4−10−1
58.68 57.10 58.78 57.12 32.58 44.82 39.73
± ± ± ± ± ± ±
0.20 0.72 0.41 1.31 1.02 1.71 1.82
221.4 156.6 192.2 138.8 360.3 240.7 297.8
± ± ± ± ± ± ±
1.2 5.8 5.1 31.0 10.0 11.4 10.2
10−6.3±0.1 10−5.7±0.1 10−6.1±0.1 10−5.5±0.4 10−3.3±0.9 10−4.2±0.7 10−3.8±0.5
10−6−10−1 10−5−10−1 10−5.8-10−1 10−5−10−1 10−3−10−1 10−4−10−1 10−3.8−10−1
Figure 8. Comparison of the potentiometric selectivity coefficients of the proposed TCNQ- and K,Na(TCNQ)-modified PVC electrodes. Figure 6. Exemplary potentiometric response recorded in KCl (black lines) and NaCl (red lines) for the bare GCD and TCNQ-, Na(TCNQ)-, and K(TCNQ)-modified GCD electrodes.
Figure 7. EMF dependence on K+ activities for i: GCD/TCNQ/K+ISM, ii: GCD/K(TCNQ)II/K+-ISM, iii: GCD/K(TCNQ)I/K+-ISM and GCD/K+-ISM with mean EMF values with standard deviation (±SD) recorded over 14 calibrations performed during 1 month.
Figure 9. Exemplary redox sensitivity of the bare GCD (×, black line), GCD/TCNQ (□, orange line), and GCD/K(TCNQ)II (○, green line) electrodes before and after being covered with potassiumselective membrane (GCD/TCNQ/K+-ISM (□, red line) and GCD/ K(TCNQ)II/K+-ISM (○, violet line)).
TCNQ significantly decreases the membrane resistance and improves the analytical parameters of sensors such as the linear range of potentiometric response, the long-term potential stability, and the potential repeatability. If the TCNQ is used as F
DOI: 10.1021/ac503521t Anal. Chem. XXXX, XXX, XXX−XXX
Article
Analytical Chemistry
Table 2. Electrical Parameters of Different Potassium and Sodium Sensors Studied type of electrode GCD/: Potassium Sensors TCNQ/K+-ISM K(TCNQ)I/K+-ISM K(TCNQ)II/K+-ISM K+-ISM TCNQ K(TCNQ)I K(TCNQ)II Sodium Sensors TCNQ/Na+-ISM K(TCNQ)I/Na+-ISM K(TCNQ)II/Na+-ISM Na+-ISM TCNQ Na(TCNQ)I Na(TCNQ)II
Figure 10. Light sensitivity of studied potassium-selective electrodes.
Rtotal (kΩ)
C (μF)
dEdc/dt (μV/s)
333 253 292 1483 807 423 601
± ± ± ± ± ± ±
5 8 7 15 9 9 10
7.56 18.6 8.54 1534 234 402 320
± ± ± ± ± ± ±
0.26 0.4 0.30 40 5 6 5
132 53.7 117 0.65 4.27 2.49 3.13
± ± ± ± ± ± ±
2 3.0 3 0.02 0.09 0.07 0.07
425 302 381 1302 809 398 702
± ± ± ± ± ± ±
6 9 9 13 8 10 10
6.49 15.1 7.58 1098 220 575 453
± ± ± ± ± ± ±
0.20 0.4 0.21 34 5 5 4
154 66.2 132 0.91 4.55 1.74 2.21
± ± ± ± ± ± ±
2 2.6 2 0.03 0.10 0.09 0.05
the solid-contact layer was detected. These results show that TCNQ and its ion-radical salts are suitable in solid-contact ISEs.
■
AUTHOR INFORMATION
Corresponding Author
*E-mail: [email protected]; tel: +480126175021; fax: +480126341201. Notes
The authors declare no competing financial interest.
■ ■
ACKNOWLEDGMENTS This work was supported by NCBiR (no. LIDER/31/7/L-2/ 10/NCBiR/2011).
Figure 11. Exemplary water layer test performed for the GCD/ TCNQ/K+-ISM electrode in 10−2 M KCl, 10−2 M NaCl, 10−2 M KCl, 10−1 M NaCl, and again 10−2 M KCl.
REFERENCES
(1) Starodub, V. A.; Starodub, T. N. Russ. Chem. Rev. 2014, 83, 391− 438. (2) Acker, D. S.; Harder, R. J.; Hertler, W. R.; Mahler, W.; Melby, L. R.; Benson, R. E.; Mochel, W. E. J. Am. Chem. Soc. 1960, 82, 6408− 6409. (3) Melby, L. R.; Harder, R. J.; Hertler, W. R.; Mahler, W.; Benson, R. E.; Mochel, W. E. J. Am. Chem. Soc. 1962, 84, 3374. (4) Michinobu, T.; Satoh, N.; Cai, J.; Li, Y.; Han, L. J. Mater. Chem. C 2014, 2, 3367−3372. (5) Nawata, T.; Yamakado, H.; Uno, K.; Tanaka, I. Phys. Status Solidi C 2013, 11, 1636−1639. (6) Grover, R.; Srivastava, R.; Kamalasanan, M. N.; Mehta, D. S. J. Lumin. 2014, 146, 53−56. (7) Harris, A. R.; Zhang, J.; Cattrall, R. W.; Bond, A. M. Anal. Methods 2013, 5, 3840−3852. (8) Sharp, M.; Johansson, G. Anal. Chim. Acta 1971, 54, 13−21. (9) Wooster, T. J.; Bond, A. M. Analyst 2003, 123, 1386−1390. (10) Paczosa-Bator, B.; Cabaj, L.; Piech, R.; Skupień, K. Analyst 2012, 22, 5272−5277. (11) Wardak, C.; Lenik, J. Sens. Actuators, B 2013, 189, 52−59. (12) Mousavi, Z.; Teter, A.; Lewenstam, A.; Maj-Zurawska, M.; Ivaska, A.; Bobacka, J. Electroanalysis 2011, 23, 1352−1358. (13) Lai, C.; Fierke, M. A.; Stein, A.; Bühlmann, P. Anal. Chem. 2007, 79, 4621−4626. (14) Crespo, G. A.; Macho, S.; Rius, F. X. Anal. Chem. 2008, 80, 1316−1322. (15) Zhu, J.; Qin, Y.; Zhang, Y. Electrochem. Commun. 2009, 11, 1684−1687.
Figure 12. Chronopotentiograms for the TCNQ-modified PVC electrode (i), the electrochemically synthesized K(TCNQ)-modified PVC electrode (ii), the chemical synthesized K(TCNQ)-modified PVC electrode (iii), the coated disc electrode, and the TCNQmodified disk electrode.
the radical salt, it may also affect the potentiometric selectivity of the electrodes. As TCNQ is classified as hydrophobic, no significant water layer between the ion-selective membrane and G
DOI: 10.1021/ac503521t Anal. Chem. XXXX, XXX, XXX−XXX
Article
Analytical Chemistry (16) Mousavi, Z.; Bobacka, J.; Lewenstam, A.; Ivaska, A. J. Electroanal. Chem. 2009, 633, 246−252. (17) Li, F.; Ye, J.; Zhou, M.; Gan, S.; Zhang, Q.; Han, D.; Niu, L. Analyst 2012, 137, 618−623. (18) Paczosa-Bator, B. Talanta 2012, 93, 424−427. (19) Paczosa-Bator, B.; Cabaj, L.; Piech, R.; Skupień, K. Anal. Chem. 2013, 85, 10255−10261. (20) Hu, J.; Zou, U.; Stein, A.; Bühlmann, P. Anal. Chem. 2014, 86, 7111−7118. (21) Sutter, J.; Radu, A.; Peper, S.; Bakker, E.; Pretsch, E. Anal. Chim. Acta 2004, 523, 53−59. (22) Michalska, A.; Konopka, A.; Maj-Zurawska, M. Anal. Chem. 2003, 75, 141−144. (23) Sutter, J.; Lindner, E.; Gyurcsányi, R. E.; Pretsch, E. Anal. Bioanal. Chem. 2004, 380, 7−14. (24) Nafady, A.; Al-Qahtani, N. J.; Al-Farhan, K. A.; Bhargava, S.; Bond, A. M. J. Solid State Electrochem. 2014, 18, 851−859. (25) Harris, A.; Zhang, J.; Cattrall, R. W.; Bond, A. M. Anal. Methods 2013, 5, 3840−3852. (26) Buck, R. P.; Lindner, E. Pure Appl. Chem. 1994, 66, 2527−2536. (27) Paczosa-Bator, B.; Cabaj, L.; Raś, M.; Baś, B.; Piech, R. Int. J. Electrochem. Sci. 2014, 9, 2816−2823. (28) Bakker, E.; Pretsch, E.; Bühlmann, P. Anal. Chem. 2000, 72, 1127−1133. (29) Fibbioli, M.; Morf, W. E.; Badertscher, M.; de Rooij, N. F.; Pretsch, E. Electroanalysis 2000, 12, 1286−1292. (30) Liljeroth, P.; Quinn, B. M.; Kontturi, K. Electrochem. Commun. 2002, 4, 255−259. (31) Bobacka, J. Anal. Chem. 1999, 71, 4932−4937.
H
DOI: 10.1021/ac503521t Anal. Chem. XXXX, XXX, XXX−XXX
Application of Nanostructured TCNQ to Potentiometric Ion-Selective K+ and Na+ Electrodes Beata Paczosa-Bator,* Magdalena Pięk, and Robert Piech AGH-UST University of Science and Technology, Faculty of Material Science and Ceramics, Mickiewicza 30, PL-30059 Cracow, Poland ABSTRACT: A new type of potentiometric solid-state ionselective electrode (SS-ISE) has been fabricated with an intermediate layer made of 7,7,8,8-tetracyanoquinodimethane (TCNQ) or its ion-radical salts and an ionophore-based ionselective membrane. To show the influence of the TCNQ layer on electrode selectivity, sodium- and potassium-sensitive membranes were applied. A good Nernstian response with a slope of 59.24 mV/ dec in the range from 10−6.5 to 10−1 M KCl and 58.68 mV/dec (10−6 to 10−1 M NaCl) was observed. The influence of an interfacial water film was assessed by an aqueous-layer test performed during potentiometric measurements. The stability of the electrical potential of the new solid-contact electrodes was tested by performing current-reversal chronopotentiometry, and the capacitance of the electrodes is 132 μF or 154 μF for K+ and Na+ solid-contact electrodes. These properties confirmed the analytical applicability of TCNQ-based SC-ISEs and should allow the development of a new solid-state ion sensor group. In this paper we present analytical characteristics of the first potentiometric solid-state electrode with TCNQ or the ion radical salt of TCNQ (potassium or sodium) and a polymeric K+- and Na+-selective membrane with significant long-term potential stability, potential repeatability, and improved detection limit. These electrodes exhibit excellent Nernstian response without light or oxygen sensitivity.
O
rganic materials have attracted significant attention over the past five decades because of their interesting and unusual electrical, magnetic, and optical properties,1 rendering them promising for utilization in industrial and technological applications. One of the most popular organic compounds whose materials have excellent electrical performances is 7,7,8,8-tetracyanoquinodimethane (TCNQ), discovered in 1960.2 Owing to high electron affinity, TCNQ can form organic charge-transfer complexes and ion-radical salts of both simple and complex composition. TCNQ salts such as lithium, sodium, or potassium can be prepared by any of the usual methods, e.g., that proposed by Melby.3 TCNQ-based materials are used nowadays in organic solar cells,4 transistors,5 or light-emission diodes.6 Already TCNQ has been successfully used in electrochemical sensors as voltammetric and potentiometric ion-selective electrode (ISE) active material.7,8 In a voltammetric approach the TCNQmodified electrode operates under dynamic kinetically controlled conditions9 whereas in potentiometric detection the signal is generated by charge separation at the TCNQ radical ion salt interface due to the selective partitioning of the ionic species between the TCNQ layer and the primary ion solution. Potentiometric ISEs with a solid-contact configuration (SCISEs) have made very rapid progress in recent decades. Different novel materials have been introduced between the electrode support and the ion-selective polymeric membrane to improve the potential stability and reproducibility, limit the formation of the interfacial water film, and reduce the sensitivity to oxygen or light of SC-ISEs.10−20 The detection limits of SCISEs have been also improved. There is a large range of potentiometric sensors that are capable for direct measurements in the nanomolar concentration range.21−23 © XXXX American Chemical Society
■
EXPERIMENTAL SECTION Materials. 7,7,8,8-Tetracyanoquinodimethane 98% (TCNQ), potassium ionophore I (valinomycin), sodium ionophore IV, bis(1-butylpentyl) adipate (BBPA), tris(2ethylhexyl) phosphate (TEHP), potassium tetrakis(4chlorophenyl)borate (KTpClPB), o-nitrophenyl octyl ether (o-NPOE), poly(vinyl chloride) (PVC) of high molecular weight, acetone, acetonitrile, and tetrahydrofuran were the selectophore reagents obtained from Sigma-Aldrich. All other chemicals were of analytical-reagent grade. Distilled and deionized water was used to prepare the aqueous solutions. Electrode Preparation. The drop-casting method was used to obtain the modified glassy carbon disc (GCD) electrodes used for ion sensing. First, the GCD electrodes (Mineral, Poland) were polished with 0.3 μm alumina powder, rinsed with water, and finally cleaned ultrasonically with water and methanol. Then was added the intermediate layer based on TCNQ or the potassium (sodium) salt of TCNQ, and then the electrode was covered with the polymeric membrane sensitive for K+ or Na+. Figure 1 shows schematically the structure of the Received: September 19, 2014 Accepted: December 31, 2014
A
DOI: 10.1021/ac503521t Anal. Chem. XXXX, XXX, XXX−XXX
Article
Analytical Chemistry
Five identical electrodes were prepared and examined. All the electrodes were separately conditioned and stored. Potentiometric Measurements. The potentials were measured using an EMF 16-channel meter (Lawson Labs, Inc., Malvern, PA). The reference electrode was an Ag/AgCl electrode with 3 M KCl solution in a bridge cell (type 6.0733.100 ΩMetrohm, Switzerland) or an Ag/AgCl/3 M KCl (type 6.0729.100 ΩMetrohm, Switzerland) with a 1 M lithium acetate salt bridge. Chronopotentiometry and Cyclic Voltammetry. The chronopotentiometry and cyclic voltammetry measurements were performed with the use of an Autolab General Purpose Electrochemical System (AUT32N.FRA2-AUTOLAB, Eco Chemie, The Netherlands) connected to a conventional, three-electrode cell. The Ag/AgCl/3 M KCl electrode (type 6.0733.100 ΩMetrohm, Switzerland) was used as a reference, and a glassy carbon rod was used as an auxiliary electrode. The electrolyte solution was purged with nitrogen for 10 min prior to each measurement. Scanning Electron Microscopy. The morphologies and chemical analysis of the chemically synthesized K(TCNQ) or Na(TCNQ) and TCNQ layer before and after electrochemical synthesis of the potassium or sodium salt were examined using a scanning electron microscope (SEM), model LEO 1530, from LEO Electron Microscopy Ltd. equipped with the Image and X-ray Analysis System, model Vantage, from ThermoNoran for energy-dispersed X-ray spectroscopy (EDS).
Figure 1. Schematic representation of the potentiometric potassiumand sodium-selective electrode.
TCNQ-, K(TCNQ)-, and Na(TCNQ)-modified polymeric ion-selective electrodes. In the case of the TCNQ-modified electrodes (GCD/TCNQ) a 25 μL solution containing 1 mg of TCNQ dispersed in 1 mL of acetone was placed on top of the GCD electrodes. To obtain modified GCD electrodes with the potassium or sodium salt of TCNQ (GCD/K(TCNQ) and GCD/Na(TCNQ)), chemical or electrochemical synthesis of K(TCNQ) and Na(TCNQ) was performed. The first group of electrodes (GCD/K(TCNQ)I and GCD/Na(TCNQ)I) was prepared by covering GCD electrodes with a 25 μL coating solution containing 1 mg of chemically prepared K(TCNQ) or Na(TCNQ) dispersed in 1 mL of acetone. Chemical synthesis of K(TCNQ) and Na(TCNQ) was carried out according to the methods of Melby.3 Potassium or sodium iodide was present in a 2:1 molar ratio to discourage the back reaction. KI or NaI dissolved in methanol was added to a hot solution of TCNQ in acetonitrile. The reaction mixture was washed with acetonitrile. The second group of K- or Na(TCNQ)-modified GCD electrodes (GCD/K(TCNQ)II and GCD/Na(TCNQ)II) were prepared by drop casting a 25 μL solution containing 1 mg of TCNQ dissolved in 1 mL of acetone, and after allowing the solvent to evaporate, electrochemical synthesis of K(TCNQ) or Na(TCNQ) was performed. Electrochemical reduction of solid TCNQ in the presence of potassium or sodium ions was performed in 0.1 M KCl or NaCl (pH = 6.8) by cycling the potential at a scan rate of 20 mV s−1 between +0.45 V and −0.1 V (vs Ag/AgCl/3 M KCl) during 50 cycles. To prepare the potassium-selective electrodes, solid-contact layers (GCD/TCNQ and GCD/K(TCNQ)) were subsequently coated with a 30 μL THF solution containing 1.1% (w/w) valinomycin, 0.25% (w/w) KTpClPB, 65.65% (w/w) oNPOE, and 33% (w/w) PVC. The Na-SCISEs were obtained by twice adding a 15 μL THF solution containing 2.9% (w/w) sodium ionophore IV, 0.4% (w/w) KTpClPB, 65.6% (w/w) (BBPA), 2.1% (w/w) TEHP, and 29% (w/w) PVC on top of the GCD/TCNQ and GCD/Na(TCNQ) electrodes. The coated disc electrodes (GCD/K+-ISM and GCD/Na+ISM) were prepared by covering bare GCD electrodes with the above-mentioned potassium or sodium-selective membrane. After the electrodes were covered with the ion-selective membrane (ISM), they were left in the air for 24 h to ensure complete evaporation of THF. Afterward the potassiumselective electrodes were conditioned in aqueous 0.001 M KCl solutions, and sodium-selective electrodes were conditioned in 0.001 M NaCl, respectively, for 24 h. The conditioning step was also repeated before every measurement.
■
RESULTS AND DISCUSSION Figure 2 shows exemplary cyclic voltammograms obtained at a scan rate of 20 mV s−1 over the potential range of +0.45 V and −0.1 V with the bare GCD, GCD/TCNQ, and GCD/ K(TCNQ)I electrodes when placed in a solution of 0.1 M KCl. As expected, potassium or sodium ions do not undergo any redox processes in the potential window used (Figure 2 inserts); therefore, the electrochemical activity of the GCD/ TCNQ electrode seen in Figure 2 is attributed to the reduction of TCNQ to TCNQ− and the introduction of ions into the crystal lattice. The difference in the shape of the oxidation and reduction processes is attributed to the higher conductivity of K(TCNQ) compared with TCNQ.24,25 The characteristic current loops in the cyclic voltammograms (Figure 2a insert) observed only in the case of the TCNQ-modified electrode and the large peak-to-peak separation (ΔEp = Epox − Epred = 250 mV) are accompanied by crystal fragmentation and rearrangement during the initial cycles of potential and are characteristic for an electrochemically irreversible TCNQ-K(TCNQ) (or TCNQ-Na(TCNQ)) phase conversion governed by nucleation−growth kinetics.24,25 In EDS analysis of the TCNQ, K(TCNQ), and Na(TCNQ), the elements C and N were clearly identified. In the case of the chemically and electrochemically synthesized K(TCNQ) and Na(TCNQ) samples, an additional K (or Na, respectively) peak appeared in the EDS spectrum that represents the K(TCNQ) molecule (or Na(TCNQ)). Exemplary SEM images showing the morphology of the TCNQ and chemically and electrochemically synthesized K(TCNQ) are presented in Figure 3. Figure 3a shows a typical rhombus-shaped crystal of TCNQ. The SEM image of chemically synthesized K(TCNQ) in Figure 3b shows a characteristic needle-shaped crystal. Figure 3c shows that electrochemical reduction of TCNQ resulted in needle-shaped B
DOI: 10.1021/ac503521t Anal. Chem. XXXX, XXX, XXX−XXX
Article
Analytical Chemistry
sodium-selective electrodes were as follows: 58.72 (58.70) mV/decade (10−6 to 10−1 M NaCl) for the TCNQ/Na+-ISM, 57.30 (57.15) mV/decade (10−5 to 10−1 M NaCl) for the Na (TCNQ)I/Na+-ISM, and 58.65 (58.62) mV/decade (10−5.8 to 10−1 M NaCl) for the Na (TCNQ)II/Na+-ISM. With the use of the same experimental conditions, the electrodes based only on the layer of TCNQ and TCNQ potassium or sodium salts showed calibration curves with the slope significantly lower than the theoretical value (see Table 1). Typical potentiometric responses of TCNQ and K(TCNQ)- or Na(TCNQ)-modified electrodes are shown in Figure 6. As the TCNQ-based GCD electrode behavior is similar to that recorded for a bare GCD electrode (dotted line in Figure 6), it is suggested that some EMF increase observed above 10−3 M KCl or NaCl is attributed to the redox response of electrodes. The fabricated electrodes show a very stable response over time. Even after a long conditioning time in 0.001 M KCl or NaCl (4−6 weeks), the electrodes still showed a linear response in the same range of the K+ and Na+ activity. Figure 7 shows the exemplary log aK+ dependence concerning the following electrodes: GCD/TCNQ/K+-ISM (i), GCD/K(TCNQ)II/K+ISM (ii), GCD/K(TCNQ)I/K+-ISM (iii), and GCD/K+-ISM (iv) obtained by recording the EMF. The reproducibility of the basic analytical parameters was also studied. For this purpose, five identical electrodes of each type were prepared by applying the same conditions. The obtained data are collected in Table 1. The detection limit was calculated as the activity of K+ at the point of intersection of the two slope lines.26 Interestingly, in the case of the SC-ISEs sensors based on the TCNQ layers, a very wide range of linear response was found with the detection limit of 10−6.9 M KCl for the GCD/TCNQ/K+-ISM electrode and 10−6.3 M NaCl for the GCD/TCNQ/Na+-ISM electrode. The obtained values are better than those for carbon black (10−6.4 M KCl and 10−5.1 M NaCl),18,27 graphene (10−5 M KCl),17 or colloid-imprinted mesoporous carbon (10−5.4 M KCl)20 as solid contact. The potential drift of the GCD/TCNQ/K+-ISM electrode over 172 h was only 11.1 ± 1.5 μV/h. Concerning the K(TCNQ)-modified PVC, the potential drift over 172 h was 18.0 ± 2.0 μV/h (GCD/K(TCNQ)I/K+-ISM) and 15.0 ± 1.7 μV/h (GCD/K(TCNQ)II/K+-ISM). The obtained values are similar to those observed for carbon materials used as the intermediate layer.11,19 In the case of sodium-selective electrodes, the potential drift was 9.2 ± 1.1 μV/h (GCD/TCNQ/Na+-ISM), 15.0 ± 1.3 μV/ h (GCD/Na(TCNQ)I/Na+-ISM), and 11.2 ± 1.2 μV/h (GCD/Na(TCNQ)II/Na+-ISM). The potentiometric selectivity coefficients Kijpot, where i is the primary ion (K+ or Na+) and j is the interfering ion, were obtained from the response potentials in an alkali metal or alkaline earth metal chloride solutions using the separate solution method (SSM) according to the traditional procedure28 (n = 3). Exemplary potentiometric selectivity coefficient values observed for the studied electrodes are presented in Figure 8. The values of Kijpot obtained for the coated with ISM GCD electrodes are presented in gray color. In the case of the TCNQ-based SC-ISEs sensors no significant changes in potentiometric selectivity coefficients were observed when compared to those received for the GCD-ISM electrodes. However, the presence of the K(TCNQ) and Na(TCNQ) layers in the GCD/K(TCNQ)/K+-ISM and GCD/Na(TCNQ)/Na+-ISM electrodes influenced on the Kij values
Figure 2. Cyclic voltammograms obtained for cycles of the potential over the range of +0.45 V and −0.1 V at (a) a TCNQ-modified GCD electrode and (b) a GCD electrode with chemically synthesized K(TCNQ).
microcrystals. Similar results were obtained in the case of Na(TCNQ) samples. The potentiometric response of the potassium and sodium sensors was studied in the concentration range 10−7 to 10−1 M of the main ion (K+ or Na+, respectively). The response reversibility of the presented electrodes recorded in the main ion solutions was very high and is shown in Figure 4. The response times of the TCNQ-modified electrodes for the progressive additions of different amounts of potassium or sodium ions were similar to those obtained for sensors based on carbon materials12−15,17−19 or conducting polymers16 and equals 4−5 s even at a low concentration of potassium or sodium ions. The calibration curve slope values measured for the developed electrodes with TCNQ, K(TCNQ), and Na(TCNQ) used as an intermediate layer are close to the Nernstian value. Figure 5 shows the dependency of log aK+ and log aNa+ on electromotive force (EMF) for all the studied solidcontact K+-ISM and Na+-ISM electrodes. The measurements were started at 10−7 M KCl (or NaCl) and continued successively up to 10−1 M KCl (or NaCl) and then back to 10−7 M KCl (or NaCl). The slopes of the linear part of the calibration curves of the potassium-selective electrodes were as follows: 59.20 (59.19) mV/decade (10−6.5 to 10−1 M KCl) for the TCNQ/K+-ISM, 57.81 (57.78) mV/ decade (10−5.5 to 10−1 M KCl) for the K(TCNQ)I/K+-ISM, and 58.60 (58.58) mV/decade (10−6 to 10−1 M KCl) for the K(TCNQ)II/K+-ISM. The slope values obtained for the C
DOI: 10.1021/ac503521t Anal. Chem. XXXX, XXX, XXX−XXX
Article
Analytical Chemistry
Figure 3. SEM images of (a) TCNQ and K(TCNQ) crystals which were prepared by (b) chemical and (c) electrochemical reduction of TCNQ, respectively.
the Na(TCNQ)/Na+-ISM electrodes showed improved selectivity for potassium ions. The redox sensitivity measurements were carried out for the TCNQ-modified electrodes covered with K+-ISM and Na+ISM, as well as for the electrodes based only on the TCNQ layers. The solution used contained FeCl3 and FeCl2 redox couple at a total concentration 1 mM with the log of Fe3+/Fe2+ ratio equal to −1, −0.5, 0, 0.5, and 1 and constant ionic background of 0.1 M KCl (or NaCl) (Figure 9). The GCD/ TCNQ, GCD/Na(TCNQ), and GCD/K(TCNQ) electrodes show a clear redox response. However, after covering the TCNQ, Na(TCNQ), or K(TCNQ) layer with PVC ionselective membrane, no redox sensitivity was observed. Similar behavior was observed earlier for the electrodes with graphene, carbon black, or PtNP intermediate layer.10,12−19 The light sensitivity of the developed electrodes was also studied. The potentials of the each electrode were measured continuously during the light sensitivity test. The electrodes were kept in the dark before starting the experiment. The following measuring sequence was used in the test: darkness (no room light during 5 min), room light (5 min), and then a return to darkness (no room light during 5 min again). EMF of the cell was recorded in 0.01 M KCl solution. As illustrated in
Figure 4. Potentiometric response of the TCNQ-modified PVC electrodes (i: TCNQ/K+-ISM, ii: K(TCNQ)II/K+-ISM, iii: K(TCNQ)I/K+-ISM) and the coated disc electrode (iv) vs time determined in the KCl solutions.
especially for those determined for potassium and sodium ions. Interesting, the application of potassium salt of TCNQ resulted in lowering the selectivity toward sodium ions and vice versa D
DOI: 10.1021/ac503521t Anal. Chem. XXXX, XXX, XXX−XXX
Article
Analytical Chemistry
Figure 5. EMF dependence on K+ and Na+ activities for (a) the TCNQ-modified PVC electrode, (b) the chemically synthesized K(TCNQ) or Na(TCNQ)-modified PVC electrode, and (c) the electrochemically synthesized K(TCNQ) or Na(TCNQ)-modified PVC electrode after 72 h of conditioning in a solution of 10−3 M KCl or 10−3 M NaCl.
M NaCl. Finally, the solution was changed again to 0.01 M KCl (see the example of the GCD/TCNQ/K+-ISM electrode in Figure 11). The dynamic EMF response was analyzed in terms of the potential drifts upon the primary ions being replaced by the interfering ions. Interestingly, the TCNQ-, Na(TCNQ)-, and K(TCNQ)-modified PVC electrodes are stable upon sample changing despite being in contact with the conditioning solution for a longer period of time (1 month). This proves that the water layer was reduced in the developed electrodes due to the hydrophobic character of TCNQ.30 Current-reversal chronopotentiometry was used to evaluate the electric capacity of the solid contact, the potential stability, and resistance of the developed electrodes according to the reported protocol.31 Figure 12 shows a typical change in the potential vs time recorded for the studied electrodes when a constant current of +1 nA was applied to the working electrodes for 60 s followed by a current of −1 nA applied for the same length of time. The changes in the polarization of the electrode were repeated three times as shown in Figure 12. For comparison, the GCD/TCNQ and coated disc electrodes were also tested. The potential jump observed in the response after current polarity change (ΔEdc) was used to calculate the total resistance of the electrode (Rtotal = ΔEdc/2I, where the I equals 1 nA). The potential drift of the electrodes was derived from the dEdc/dt ratio, and the capacity of the sensors was calculated from the equation dEdc/dt = I/C. All parameters determined as the average of three measurements are summarized in Table 2. The resistance of the TCNQ- or K,Na(TCNQ)-modified GCD electrodes generally decreased after covering with a PCV ion-selective membrane. On the other hand, the coated disctype electrode (GCD/ISM) resistance is considerably reduced after application of the TCNQ layer. This indicates that potassium or sodium ion transport is facilitated when both layers are applied in the sensor. The lowest value of Rtotal was observed in the case of electrodes with the chemically prepared K,Na(TCNQ). The obtained for developed electrode potential drift values are little higher than those obtained from the PtNPs-VXCbased sensors (4.6 μV/s)19 or the sensors with colloidimprinted mesoporous carbon as solid contact (1.0 μV/s)20 and lower than those obtained from the graphene-based sensors (12 μV/s),17 SC-ISEs with nanotubes (17 μV/s),14 and coated disc electrodes developed under similar conditions (see Table 2). The capacity of the GCD/TCNQ/K+-ISM electrode is calculated to be 132 μF and GCD/Na(TCNQ)/Na+-ISM electrode −154 μF. These values are much better than those measured for CPs-CNT (83 μF14), graphene (83 μF17), Printex XE-2 carbon black (51 μF18), or platinum nanoparticle (82 μF10) based solid-contact electrodes and worse than 1.0 mF calculated for sensors with colloid-imprinted mesoporous carbon.20
the potassium electrode example in Figure 10, no significant potential drift was observed during measurement. The presence of an unfavorable water layer between the ionselective membrane and the solid contact was tested by recording the dependency of EMF vs time when the primary ions are substituted by interfering ions, according to Fibbioli et al.29 The studied electrodes were initially conditioned in the primary ion solution (0.01 M KCl). After 3 weeks, the solution was changed to 0.01 M NaCl and then to 0.01 M KCl and 0.1
CONCLUSIONS Potentiometric solid-state sensors based on conducting organic material (TCNQ) and its cation-radical salts were synthesized and studied as K+- and Na+-selective electrodes. These novel sensors, prepared in a simple drop-casting method, showed a good Nernstian response to potassium or sodium in a wide linear range of primary ions, and the best resulting detection limit for the proposed potassium-selective electrode is 10−6.9 M and for the Na+-selective electrode 10−6.3 M. The presence of
■
E
DOI: 10.1021/ac503521t Anal. Chem. XXXX, XXX, XXX−XXX
Article
Analytical Chemistry Table 1. Reproducibility of the Basic Analytical Parameters of Different Potassium and Sodium Sensors type of electrode GCD/: Potassium Sensors TCNQ/K+-ISM K(TCNQ)I/K+-ISM K(TCNQ)II/K+-ISM K+-ISM TCNQ K(TCNQ)I K(TCNQ)II Sodium Sensors TCNQ/Na+-ISM K(TCNQ)I/Na+-ISM K(TCNQ)II/Na+-ISM Na+-ISM TCNQ Na(TCNQ)I Na(TCNQ)II
slope (mV/dec) ± SD (n = 5)
E0 (mV) ± SD (n = 5)
detection limit (M) ± SD (n = 5)
linear range (M)
59.24 57.62 58.58 57.84 25.42 42.14 36.84
± ± ± ± ± ± ±
0.08 0.43 0.18 1.10 1.55 1.23 1.45
398.8 340.4 379.1 322.7 308.4 148.4 182.3
± ± ± ± ± ± ±
1.4 6.6 5.7 28.1 11.0 12.2 12.4
10−6.9±0.1 10−5.9±0.1 10−6.5±0.1 10−5.5±0.4 10−3.3±0.9 10−4.2±0.7 10−3.8±0.5
10−6.5−10−1 10−5.5−10−1 10−6−10−1 10−5−10−1 10−3−10−1 10−4−10−1 10−3.4−10−1
58.68 57.10 58.78 57.12 32.58 44.82 39.73
± ± ± ± ± ± ±
0.20 0.72 0.41 1.31 1.02 1.71 1.82
221.4 156.6 192.2 138.8 360.3 240.7 297.8
± ± ± ± ± ± ±
1.2 5.8 5.1 31.0 10.0 11.4 10.2
10−6.3±0.1 10−5.7±0.1 10−6.1±0.1 10−5.5±0.4 10−3.3±0.9 10−4.2±0.7 10−3.8±0.5
10−6−10−1 10−5−10−1 10−5.8-10−1 10−5−10−1 10−3−10−1 10−4−10−1 10−3.8−10−1
Figure 8. Comparison of the potentiometric selectivity coefficients of the proposed TCNQ- and K,Na(TCNQ)-modified PVC electrodes. Figure 6. Exemplary potentiometric response recorded in KCl (black lines) and NaCl (red lines) for the bare GCD and TCNQ-, Na(TCNQ)-, and K(TCNQ)-modified GCD electrodes.
Figure 7. EMF dependence on K+ activities for i: GCD/TCNQ/K+ISM, ii: GCD/K(TCNQ)II/K+-ISM, iii: GCD/K(TCNQ)I/K+-ISM and GCD/K+-ISM with mean EMF values with standard deviation (±SD) recorded over 14 calibrations performed during 1 month.
Figure 9. Exemplary redox sensitivity of the bare GCD (×, black line), GCD/TCNQ (□, orange line), and GCD/K(TCNQ)II (○, green line) electrodes before and after being covered with potassiumselective membrane (GCD/TCNQ/K+-ISM (□, red line) and GCD/ K(TCNQ)II/K+-ISM (○, violet line)).
TCNQ significantly decreases the membrane resistance and improves the analytical parameters of sensors such as the linear range of potentiometric response, the long-term potential stability, and the potential repeatability. If the TCNQ is used as F
DOI: 10.1021/ac503521t Anal. Chem. XXXX, XXX, XXX−XXX
Article
Analytical Chemistry
Table 2. Electrical Parameters of Different Potassium and Sodium Sensors Studied type of electrode GCD/: Potassium Sensors TCNQ/K+-ISM K(TCNQ)I/K+-ISM K(TCNQ)II/K+-ISM K+-ISM TCNQ K(TCNQ)I K(TCNQ)II Sodium Sensors TCNQ/Na+-ISM K(TCNQ)I/Na+-ISM K(TCNQ)II/Na+-ISM Na+-ISM TCNQ Na(TCNQ)I Na(TCNQ)II
Figure 10. Light sensitivity of studied potassium-selective electrodes.
Rtotal (kΩ)
C (μF)
dEdc/dt (μV/s)
333 253 292 1483 807 423 601
± ± ± ± ± ± ±
5 8 7 15 9 9 10
7.56 18.6 8.54 1534 234 402 320
± ± ± ± ± ± ±
0.26 0.4 0.30 40 5 6 5
132 53.7 117 0.65 4.27 2.49 3.13
± ± ± ± ± ± ±
2 3.0 3 0.02 0.09 0.07 0.07
425 302 381 1302 809 398 702
± ± ± ± ± ± ±
6 9 9 13 8 10 10
6.49 15.1 7.58 1098 220 575 453
± ± ± ± ± ± ±
0.20 0.4 0.21 34 5 5 4
154 66.2 132 0.91 4.55 1.74 2.21
± ± ± ± ± ± ±
2 2.6 2 0.03 0.10 0.09 0.05
the solid-contact layer was detected. These results show that TCNQ and its ion-radical salts are suitable in solid-contact ISEs.
■
AUTHOR INFORMATION
Corresponding Author
*E-mail: [email protected]; tel: +480126175021; fax: +480126341201. Notes
The authors declare no competing financial interest.
■ ■
ACKNOWLEDGMENTS This work was supported by NCBiR (no. LIDER/31/7/L-2/ 10/NCBiR/2011).
Figure 11. Exemplary water layer test performed for the GCD/ TCNQ/K+-ISM electrode in 10−2 M KCl, 10−2 M NaCl, 10−2 M KCl, 10−1 M NaCl, and again 10−2 M KCl.
REFERENCES
(1) Starodub, V. A.; Starodub, T. N. Russ. Chem. Rev. 2014, 83, 391− 438. (2) Acker, D. S.; Harder, R. J.; Hertler, W. R.; Mahler, W.; Melby, L. R.; Benson, R. E.; Mochel, W. E. J. Am. Chem. Soc. 1960, 82, 6408− 6409. (3) Melby, L. R.; Harder, R. J.; Hertler, W. R.; Mahler, W.; Benson, R. E.; Mochel, W. E. J. Am. Chem. Soc. 1962, 84, 3374. (4) Michinobu, T.; Satoh, N.; Cai, J.; Li, Y.; Han, L. J. Mater. Chem. C 2014, 2, 3367−3372. (5) Nawata, T.; Yamakado, H.; Uno, K.; Tanaka, I. Phys. Status Solidi C 2013, 11, 1636−1639. (6) Grover, R.; Srivastava, R.; Kamalasanan, M. N.; Mehta, D. S. J. Lumin. 2014, 146, 53−56. (7) Harris, A. R.; Zhang, J.; Cattrall, R. W.; Bond, A. M. Anal. Methods 2013, 5, 3840−3852. (8) Sharp, M.; Johansson, G. Anal. Chim. Acta 1971, 54, 13−21. (9) Wooster, T. J.; Bond, A. M. Analyst 2003, 123, 1386−1390. (10) Paczosa-Bator, B.; Cabaj, L.; Piech, R.; Skupień, K. Analyst 2012, 22, 5272−5277. (11) Wardak, C.; Lenik, J. Sens. Actuators, B 2013, 189, 52−59. (12) Mousavi, Z.; Teter, A.; Lewenstam, A.; Maj-Zurawska, M.; Ivaska, A.; Bobacka, J. Electroanalysis 2011, 23, 1352−1358. (13) Lai, C.; Fierke, M. A.; Stein, A.; Bühlmann, P. Anal. Chem. 2007, 79, 4621−4626. (14) Crespo, G. A.; Macho, S.; Rius, F. X. Anal. Chem. 2008, 80, 1316−1322. (15) Zhu, J.; Qin, Y.; Zhang, Y. Electrochem. Commun. 2009, 11, 1684−1687.
Figure 12. Chronopotentiograms for the TCNQ-modified PVC electrode (i), the electrochemically synthesized K(TCNQ)-modified PVC electrode (ii), the chemical synthesized K(TCNQ)-modified PVC electrode (iii), the coated disc electrode, and the TCNQmodified disk electrode.
the radical salt, it may also affect the potentiometric selectivity of the electrodes. As TCNQ is classified as hydrophobic, no significant water layer between the ion-selective membrane and G
DOI: 10.1021/ac503521t Anal. Chem. XXXX, XXX, XXX−XXX
Article
Analytical Chemistry (16) Mousavi, Z.; Bobacka, J.; Lewenstam, A.; Ivaska, A. J. Electroanal. Chem. 2009, 633, 246−252. (17) Li, F.; Ye, J.; Zhou, M.; Gan, S.; Zhang, Q.; Han, D.; Niu, L. Analyst 2012, 137, 618−623. (18) Paczosa-Bator, B. Talanta 2012, 93, 424−427. (19) Paczosa-Bator, B.; Cabaj, L.; Piech, R.; Skupień, K. Anal. Chem. 2013, 85, 10255−10261. (20) Hu, J.; Zou, U.; Stein, A.; Bühlmann, P. Anal. Chem. 2014, 86, 7111−7118. (21) Sutter, J.; Radu, A.; Peper, S.; Bakker, E.; Pretsch, E. Anal. Chim. Acta 2004, 523, 53−59. (22) Michalska, A.; Konopka, A.; Maj-Zurawska, M. Anal. Chem. 2003, 75, 141−144. (23) Sutter, J.; Lindner, E.; Gyurcsányi, R. E.; Pretsch, E. Anal. Bioanal. Chem. 2004, 380, 7−14. (24) Nafady, A.; Al-Qahtani, N. J.; Al-Farhan, K. A.; Bhargava, S.; Bond, A. M. J. Solid State Electrochem. 2014, 18, 851−859. (25) Harris, A.; Zhang, J.; Cattrall, R. W.; Bond, A. M. Anal. Methods 2013, 5, 3840−3852. (26) Buck, R. P.; Lindner, E. Pure Appl. Chem. 1994, 66, 2527−2536. (27) Paczosa-Bator, B.; Cabaj, L.; Raś, M.; Baś, B.; Piech, R. Int. J. Electrochem. Sci. 2014, 9, 2816−2823. (28) Bakker, E.; Pretsch, E.; Bühlmann, P. Anal. Chem. 2000, 72, 1127−1133. (29) Fibbioli, M.; Morf, W. E.; Badertscher, M.; de Rooij, N. F.; Pretsch, E. Electroanalysis 2000, 12, 1286−1292. (30) Liljeroth, P.; Quinn, B. M.; Kontturi, K. Electrochem. Commun. 2002, 4, 255−259. (31) Bobacka, J. Anal. Chem. 1999, 71, 4932−4937.
H
DOI: 10.1021/ac503521t Anal. Chem. XXXX, XXX, XXX−XXX
