Gold Plate Electrodes Functionalized by Multiwall Carbon Nanotube Film for Potentiometric Thallium(I) Detection

Solid-contact potentiometric ion-selective electrodes (SC-ISEs) for thallium determination have been designed using multiwall carbon nanotubes (MWCNTs) as the ion-to-electron transducer. Dispersed MWCNTs were drop-casted over a gold plate electrode. Two different crown ethers were used in the sensing membrane for the recognition of thallium (I). Sensorsbased on dibenzo-18-crown-6 (DB18C6) as a neutral carrier and NaTPB as an anionic additive exhibited a near Nernstian response of 57.3 mV/decade towards Tl+ ions over the activity range 4.5 × 10−6–7.0 × 10−4 M, with a limit of detection of 3.2 × 10−7 M. The time required to achieve 95% of the steadyequilibrium potential was <10 s. The complex formation constant (log βML) between dibenzo-18-crown-6 and thallium (I) (i.e., 5.99) was measured using the sandwich membrane technique. The potential response was pH independent over the range 3.0–9.5. The introduction of MWCNTs as an electron-ion-transducer layer between gold plate and the sensing membrane lead to a smaller membrane resistance and a large double layer capacitance, which was proven using impedance spectra and chronopotentiometry (i.e., 114.9 ± 12 kΩ, 52.1 ± 3.3 pF, 200 ± 13.2 kΩ, and 50 ± 4.2 µF). Additionally, reduction ofthe water layer between the sensing membrane and the underlying conductor wastested. Thus, it is clear that MWCNTs can be used as a transducing layer in SC-ISEs. The proposed sensor was introduced as an indicator electrode for potentiometric titration of single and ternary mixtures of I−, Br−, and S2− anions.


Introduction
Natural processes and human activity increase the concentration of metals in the environment. This poses a threat to human and other living organisms. One of the most significantpollutants that isintroduced and affects the environment is thallium (Tl), which is produced as waste from lead and coal combustion [1]. Thallium is the most dangerous pollutant, with a higher toxicity to the biosphere than Cd, Cu, Hg, and Pb [2]. It is used in the manufacture of low-temperature thermometers, optical lenses, semiconductors, jewelry, and as a catalyst in certain alloys [3]. Thallium (I) sulfate is also used as substrates were also reported to exhibit higher E • reproducibility than those based on glassy carbon substrates [43].Therefore, careful consideration should be taken inthe choice of the electron-conducting substrate (i.e., glassy carbon, Au, and Pt) when designing solid-contact ISEs.
In this work, miniaturized solid-contact potentiometric ion-selective electrodes (SC-ISEs) for Tl + determination are presented for the first time. The sensor membranes are based on the use of dibenzo-18-crown-6 (DB18C6) as a neutral carrier. Multiwall carbon nanotubes (MWCNTs) are used as a solid contact material on a gold film (i.e., electron-conducting substrate). The constructed Tl + -SC-ISEs exhibit high potential stability and small membrane impedance. This simple fabrication of the electrode design permits the fabrication of these ISEs by a non-specialist. The developed SC/Tl + -ISEs are fully characterized and tested in potentiometric titration of single and ternary mixtures of I − , Br − , and S 2− ions. As a consequence of the excellent analytical performance, we conclude that this technology may become promising for the design of in-situ Tl + sensing probes.

Equipment
Potential measurements were performed with an Orion (Cambridge, MA, USA) Model 720/SA pH/mV meter at 25 ± 1 • C using the gold plate thallium polymeric membrane sensor in conjunction with Ag/AgCl double-junction reference electrode (Orion 90-20) filled with 1 M CH 3 COOLi. Chronopotentiometry and electrochemical impedance spectroscopy (EIS) measurements were carried out using a potentiostat/galvanostat (Autolab Model 2000, Metrohom Instruments, Herisau, Switzerland). A three-electrode configuration cell containing a silver/silver chloride (3 M KCl) reference electrode and an auxiliary electrode made from platinum wire was employed. The impedance spectra were measured and recorded at open-circuit potential in a 0.01 M NaClO 4 solution with an excitation amplitude of 10 mV and a frequency range of 100 kHz-0.1 Hz.

Materials and Reagents
All chemicals and reagents used were of analytical grade and were used as received and prepared with de-ionized water. For membrane preparation, high molecular weight poly (vinyl chloride) (PVC), O-nitrophenyloctyl ether (o-NPOE), Tetrahydrofurane (THF), and sodium tetraphenylborate (NaTPB) were used as received from Fluka or Merck. Multiwall carbon nanotubes (MWCNTs) were purchased from (EPRI, cairo, Egypt). Sodium sulphide (Na 2 S), sodium iodide (NaI), sodium bromide (NaBr), and Trizma ® buffer were purchased from Sigma Aldrich. A 0.1 M stock solution of TlNO 3 was freshly prepared.

Preparation of Gold Plate Electrodes
Gold plate electrodes (GPEs) witha gold diameter of 4 mm and a purity of 99.999% were sonicated for 10 min in acetone then cleaned by isopropanol and ethanol, each for 10 min. De-ionized water was then used to rinse the electrodes for 5 min, which were then left for 2 hto dry. Commercial multiwalled carbon nanotubes (MWCNTs) with a diameter of 1-20 nm were sonicated for 30 min with tetrahydrofuran (THF) in order to get a good dispersion for the carbon nanotubes (CNTs) [44,45]. Using a micropipette, droplets of the CNTs solution with THF were dropped over the gold plate electrode. As a result, CNTs were uniformly deposited over the gold plate electrode. It was then left for 2 h for drying by evaporation of tetrahydrofuran from the electrode.

Sensor Construction
The ion selective membranes were prepared by dissolving appropriate amounts of ionophores, NaTPB as an anionic additive, o-NPOE as a plasticizer, and PVC in THF (3 mL) as recommended bythe Internatioal Union of pure Applied chemistry (IUPAC) [46]. 50 µL of the homogenous membrane cocktail was drop-casted on the gold plate electrode directly, or on the CNTs deposited on the gold plate electrode, using a micropipette and then leftfor 2 h for drying by the evaporation of tetrahydrofuran as shown in Figure 1. Thereafter, the sensors were conditioned by soaking them in a 0.01M aqueous solution of TlNO 3 for 12 h. bythe Internatioal Union of pure Applied chemistry (IUPAC) [46]. 50 µL of the homogenous membrane cocktail was drop-casted on the gold plate electrode directly, or on the CNTs deposited on the gold plate electrode, using a micropipette and then leftfor 2 h for drying by the evaporation of tetrahydrofuran as shown in Figure 1. Thereafter, the sensors were conditioned by soaking them in a 0.01M aqueous solution of TlNO3 for 12 h.

Potential Measurements
The proposed ISE was calibrated using a standard two electrode (working and reference) cell configuration, using a gold plate electrode and a CNT/gold plate electrode in conjunction with an Ag/AgCl reference electrode. The cell solution contentsof 10.0 mL of 0.01 M Trizma buffer (pH ~6) and (0.5-1.0) mL aliquots of a 1.0 × 10 -1 -1.0 × 10 -7 M from an aqueous solution of Tl + were sequentially added. After potential stabilization, the readings were plotted as a function of log aTl+. Estimation of Tl + concentration was done using the calibration plot.

Complex Formation Constants Measurements
According to the sandwich membrane method, experiments were carried out [47,48]. The PVC/plasticizer (1:2) membranes containing 80 mmol/kg of ionophore and 56 mmol/kg of NaTPB (total membrane mass 100 mg) were cast. Additionally, membranes without ionophore were prepared by the same constituents. A 50 µL of the membrane cocktail was drop-casted from both types of membrane and then conditioned overnight in a 0.01 M solution of the appropriate salt (Tl + , K + , Pb 2+ , Na + , NH4 + ). The membrane with the ionophore faced to the sample solution. The membranes were then instantly placed into a salt solution that was identical to the conditioning of the membrane. The mean potential was recorded for the last minute of a 10 min measurement period in the test solution. In a separate measurement, the potential of an electrode with the membrane containing no ionophore was measured in the same solution. The formation constant (logβML) was calculated according to Equation(1) [47,48]: where n is the complex stoichiometry, LT is the total concentration of ionophore, RT is the concentration of the lipophilic ionic site additives, and R, T, and F are the gas constant, the absolute temperature, and the Faraday constant, ZIcharge of the tested ion. The membrane potential EM is determined by subtracting the cell potential for the membrane without ionophore from that of the sandwich membrane.

Potential Measurements
The proposed ISE was calibrated using a standard two electrode (working and reference) cell configuration, using a gold plate electrode and a CNT/gold plate electrode in conjunction with an Ag/AgCl reference electrode. The cell solution contentsof 10.0 mL of 0.01 M Trizma buffer (pH~6) and (0.5-1.0) mL aliquots of a 1.0 × 10 -1 -1.0 × 10 -7 M from an aqueous solution of Tl + were sequentially added. After potential stabilization, the readings were plotted as a function of log a Tl+ . Estimation of Tl + concentration was done using the calibration plot.

Complex Formation Constants Measurements
According to the sandwich membrane method, experiments were carried out [47,48]. The PVC/plasticizer (1:2) membranes containing 80 mmol/kg of ionophore and 56 mmol/kg of NaTPB (total membrane mass 100 mg) were cast. Additionally, membranes without ionophore were prepared by the same constituents. A 50 µL of the membrane cocktail was drop-casted from both types of membrane and then conditioned overnight in a 0.01 M solution of the appropriate salt (Tl + , K + , Pb 2+ , Na + , NH 4 + ).
The membrane with the ionophore faced to the sample solution. The membranes were then instantly placed into a salt solution that was identical to the conditioning of the membrane. The mean potential was recorded for the last minute of a 10 min measurement period in the test solution. In a separate measurement, the potential of an electrode with the membrane containing no ionophore was measured in the same solution. The formation constant (logβ ML ) was calculated according to Equation (1) [47,48]: where n is the complex stoichiometry, L T is the total concentration of ionophore, R T is the concentration of the lipophilic ionic site additives, and R, T, and F are the gas constant, the absolute temperature, and the Faraday constant, Z I charge of the tested ion. The membrane potential E M is determined by subtracting the cell potential for the membrane without ionophore from that of the sandwich membrane.

Performace Characterisics of the Proposed Sensors
Attempts were made to use two different crown ethers as neutral carriers for the Tl + ion in the PVC matrix membrane sensors. For ISEs based on neutral carriers, the potentiometric selectivity depends mainly on the ability of the neutral carrier to extract the desired ion [49,50]. According to the previously reported studies for solvent extraction [51], some crown ether derivatives revealed good complexing ability towards the Tl + ion. In this work, dibenzo-18-crown-6 (DB18C6) was tested as a neutral carriers for designing a miniaturized SC/Tl + -ISEs. MWCNTs were used as an intermediate layer between the sensing membrane and Au, which acts as an electron conductor substrate. The membrane composition was 32.0 wt.% PVC, 65.0 wt.% plasticizer, and 3.0 wt.% ionophore. Potentiometric characteristics of sensors based on DB18C6 revealed a strong response towards Tl + ions. Results from replicate studies showed sub-Nernstian slopes of 32.7 ± 2.5 mV/decade, with detection limits of 4.0 × 10 −6 M, respectively. The addition of 1.5 wt.% Na-TPB enhanced the sub-Nernstian calibration slope from 32.7 ± 2.5 to a Nernstian slope of 57.3 ± 1.6 mV/decade and lowered the detection limit from 4.0 × 10 −6 M to 3.2 × 10 −7 M. The addition of NaTPB as lipophilic anions to the membrane improved its performance characteristics, because it stabilizes membrane operating conditions, decreases membrane resistance, and significantly reduces response times. The calibration plots of these ISEs are shown in Figure 2, and their performance characteristics are presented in Table 1. The potential responses revealed by the proposed sensors can be attributed to the selective interaction of Tl + ions with DB18C6 and are affected by: (i) The ionic diameter of the crown ionophore; (ii) the ionic size of the thallium (I) ion; and (iii) the spatial configuration of this ionophore. The ionic size of Tl + is 1.73 Å, while the ionic diameters of both 18-crown-6 varies from 2.6-3.2 Å [52].  The response time of the sensor was also tested by measuring the time required to achieve a 95% steady potential for all concentration ranges (10 −6 -10 −2 M TlNO 3 solutions). Fairly short response times of 5 s for [Tl + ] > 10 −5 M and 10 s for [Tl + ] ≤ 10 −5 M were obtained ( Figure 2). A drop in the sensitivity was observed after 1 week of reasonably stable behavior.
The pH dependence of the DB18C6+TPB membrane based sensor was tested using two Tl + concentrations (i.e., 10 −3 and 10 −4 M). A stable potential response was noticed over the pH range 3-9.5, as shown in Figure 3. A pH adjustment was carried out using dilute LiOH and HNO 3 . Below pH 3, the potential drift observed is attributed to either the response of SC/Tl + -ISEs to H + ions, or it may be due to the protonation of the ionophore itself in the membrane. The observed potential drop above pH 9.5 is attributed to the formation of TlOH precipitate.
Potentiometric selectivity coefficients (log Pot Tl+, j ) for the sensor were evaluated using the modified separate solution method (MSSM) [53]. Before measuring the potential response of all tested ions, the membrane had never been in contact with the primary ions. In order of decreasing discrimination, we first started with the most discriminated ions. The highest measured concentrations (10 −1 M) were used for the estimation of the selectivity values [53]. The activities of both Tl+ and the interfering ions were calculated by the extended Deby-Huckel equation. The results are shown in Figure 4. The selectivity order of the sensor was found to be: It can be seen that the sensor reveals enhanced selectivity towards Tl + ions over many of the ions under investigation, so the designed sensor is appropriate for the intended estimations.
Nanomaterials 2019, 9, x FOR PEER REVIEW 6 of 14 tested ions, the membrane had never been in contact with the primary ions. In order of decreasing discrimination, we first started with the most discriminated ions. The highest measured concentrations (10 −1 M) were used for the estimation of the selectivity values [53]. The activities of both Tl+ and the interfering ions were calculated by the extended Deby-Huckel equation. The results are shown in Figure 4. The selectivity order of the sensor was found to be: It can be seen that the sensor reveals enhanced selectivity towards Tl + ions over many of the ions under investigation,so the designed sensor is appropriate for the intended estimations.    tested ions, the membrane had never been in contact with the primary ions. In order of decreasing discrimination, we first started with the most discriminated ions. The highest measured concentrations (10 −1 M) were used for the estimation of the selectivity values [53]. The activities of both Tl+ and the interfering ions were calculated by the extended Deby-Huckel equation. The results are shown in Figure 4. The selectivity order of the sensor was found to be: Tl + > K + > Pb 2+ > NH4 + > Li + > Cu 2+ ≈ Sr 2+ ≈ Ca 2+ > Ni 2+ ≈ Na + > Zn 2+ ≈ Ag + ≈ Co 2+ ≈ Ba + > Mg 2+ ≈ Fe 2+ > Cd 2+ . It can be seen that the sensor reveals enhanced selectivity towards Tl + ions over many of the ions under investigation,so the designed sensor is appropriate for the intended estimations.    tested ions, the membrane had never been in contact with the primary ions. In order of decreasing discrimination, we first started with the most discriminated ions. The highest measured concentrations (10 −1 M) were used for the estimation of the selectivity values [53]. The activities of both Tl+ and the interfering ions were calculated by the extended Deby-Huckel equation. The results are shown in Figure 4. The selectivity order of the sensor was found to be: Tl + > K + > Pb 2+ > NH4 + > Li + > Cu 2+ ≈ Sr 2+ ≈ Ca 2+ > Ni 2+ ≈ Na + > Zn 2+ ≈ Ag + ≈ Co 2+ ≈ Ba + > Mg 2+ ≈ Fe 2+ > Cd 2+ . It can be seen that the sensor reveals enhanced selectivity towards Tl + ions over many of the ions under investigation,so the designed sensor is appropriate for the intended estimations.

Formation Constantβ ML
The potentiometric selectivity of the polymeric membranes is related to the differences in the lipophilicity of the ions, the membrane composition, and the stability constant [54]. As the last factor is typically the most important one, several methods suitable for the estimation of stability constant values in the membrane phase were proposed [47,48]. Among them, the sandwich membrane method seems to be the most appropriate due to its simplicity and reliability. In this work, various cations and dibenzo-18-crown-6 formation constants were determined. The obtained results are presented in Table 2. It was found that the Tl + ion can form the strongest complex with dibenzo-18-crown-6. The complex stability constants, log β ML , measured for other cations, are significantly lower compared to thallium, while the values of log β ML for K + and Pb 2+ are nearly the same.

Water Layer Test
The water layer, which is formed between the underlying conductor and the ISM, leads to some failures, the most important of which are responsive hysteresis, potential instability, and mechanical problems [19,24]. A simple protocol was presented byPretsch et al. [55] in order to confirm the presence of the water layer between the sensing membrane and the electron conducting substrate. The possible drift by replacing the primary ions solution can be observed with discriminated interfering ions, and the reverse is true in the presence of this layer of water. When interfering ions are replaced by primary ions, or vice versa, either a possible negative or positive potential driftis obtained. This is due to the transfer of the corresponding ions from the samples spread over the ISM, which leads to a change in the composition of the water layer [19,24].
Herein, tests for the reduction of the water layer after the insertion of the lipophilic MWCNT layers were carried out. Sequential immersions of the gold plate electrode in the presence and absence of the MWCNT layer in 1.0 × 10 −2 M Trizma buffer (pH~6) solution was done for the first hour then in a 1.

Chronopotentiometric Test
At a constant current of ±1 nA, chronopotentiograms were recorded to verify the potential stability of the developed GPE/MWCNTs, and the results are clearly shown in Figure 6. As can be seen in the case of GPE in the absence of MWCNTs, there is a large potential drift of up to 100±0.01 µV/s (n = 3). Nevertheless, in the case of GPE/MWCNTs, a much smaller potential drift value of 16 ± 0.02 µV/s (n = 3) is recorded for the SC-ISE. Due to the high double layer capacitance of MWCNTs, this is likely to result in potential stability. Using the method provided by Bobacka [56], the membrane resistance and capacitance of the ISE can be calculated. According to the fundamental equations:∆E/∆t=i/C and ∆E =I R, in which ∆E, ∆t, I, C, and R represent the change of potential, variation of time, applied current, capacitance, and bulk membrane resistance, respectively, the capacitances in the absence and presence of MWCNTs were calculated to be 23.8±5.1 and 57.8±1.3 µF, respectively. The resistances of the membrane in the absence and presence of the MWCNT layer were 393±9.1 and 100±6.3 kΩ, respectively. From the results shown above, we can confirm the relationship between the potential stability and the presence of the MWCNT layer as a solid contact.

Chronopotentiometric Test
At a constant current of ±1 nA, chronopotentiograms were recorded to verify the potential stability of the developed GPE/MWCNTs, and the results are clearly shown in Figure 6. As can be seen in the case of GPE in the absence of MWCNTs, there is a large potential drift of up to 100 ± 0.01 µV/s (n = 3). Nevertheless, in the case of GPE/MWCNTs, a much smaller potential drift value of 16 ± 0.02 µV/s (n = 3) is recorded for the SC-ISE. Due to the high double layer capacitance of MWCNTs, this is likely to result in potential stability. Using the method provided by Bobacka [56], the membrane resistance and capacitance of the ISE can be calculated. According to the fundamental equations: ∆E/∆t = i/C and ∆E = I R, in which ∆E, ∆t, I, C, and R represent the change of potential, variation of time, applied current, capacitance, and bulk membrane resistance, respectively, the capacitances in the absence and presence of MWCNTs were calculated to be 23.8 ± 5.1 and 57.8 ± 1.3 µF, respectively. The resistances of the membrane in the absence and presence of the MWCNT layer were 393 ± 9.1 and 100 ± 6.3 kΩ, respectively. From the results shown above, we can confirm the relationship between the potential stability and the presence of the MWCNT layer as a solid contact.

Chronopotentiometric Test
At a constant current of ±1 nA, chronopotentiograms were recorded to verify the potential stability of the developed GPE/MWCNTs, and the results are clearly shown in Figure 6. As can be seen in the case of GPE in the absence of MWCNTs, there is a large potential drift of up to 100±0.01 µV/s (n = 3). Nevertheless, in the case of GPE/MWCNTs, a much smaller potential drift value of 16 ± 0.02 µV/s (n = 3) is recorded for the SC-ISE. Due to the high double layer capacitance of MWCNTs, this is likely to result in potential stability. Using the method provided by Bobacka [56], the membrane resistance and capacitance of the ISE can be calculated. According to the fundamental equations:∆E/∆t=i/C and ∆E =I R, in which ∆E, ∆t, I, C, and R represent the change of potential, variation of time, applied current, capacitance, and bulk membrane resistance, respectively, the capacitances in the absence and presence of MWCNTs were calculated to be 23.8±5.1 and 57.8±1.3 µF, respectively. The resistances of the membrane in the absence and presence of the MWCNT layer were 393±9.1 and 100±6.3 kΩ, respectively. From the results shown above, we can confirm the relationship between the potential stability and the presence of the MWCNT layer as a solid contact.

Electrochemical Impedance Spectroscopy (EIS) Measurements
The electrochemical impedance spectrum for gold plate electrode (GPE) in the presence and absence of MWCNTs as an ion-to-electron transducer were measured by EIS measurements. The obtained EIS plots are shown in Figure 7. The bulk membrane resistance, together with contact resistance between the underlying conductor and the ISM, is represented by the diameter of the semicircle at the high frequency region, represented by R bc , in parallel with the combined geometric capacitance (C g ) [57]. While at the low-frequency region, characteristics are attributed to the double-layer capacitance (C d ) coupled with a charge-transfer resistance (R ct ) at the interface between the ISM and the underlying conductor [57]. After inserting the MWCNT layer, a small R bc can be obtained for the GPE/MWCNTs (R bc = 114.9 ± 12 kΩ) when compared with the GPE in the absence of MWCNTs (R bc = 418 ± 11.2 kΩ). The geometric capacitances (C g ) for GPE in the absence and presence of MWCNTs are 52.1 ± 3.3 and 141.8 ± 7.4 pF. At the low frequency region, the charge-transfer resistance (R ct ) and double-layer capacitance (C dl ) for GPE in the absence and presence of MWCNTs areR ct = 454.5 ± 13.2 kΩ, Cd = 22 ± 2.1 µF and R ct = 200 ± 13.2 kΩ, C d = 50 ± 4.2 µF, respectively. As a result, the presence of MWCNTs as an ion-to-electron transducer offers higher potential stability. This is because of the presence of the MWCNT layers, which facilitate the charge transport between interfaces in the SC-ISEs.

Electrochemical Impedance Spectroscopy (EIS) Measurements
The electrochemical impedance spectrum for gold plate electrode (GPE) in the presence and absence of MWCNTs as an ion-to-electron transducer were measured by EIS measurements. The obtained EIS plots are shown in Figure 7. The bulk membrane resistance, together with contact resistance between the underlying conductor and the ISM, is represented by the diameter of the semicircle at the high frequency region, represented by Rbc, in parallel with the combined geometric capacitance (Cg) [57]. While at the low-frequency region, characteristics are attributed to the double-layer capacitance (Cd) coupled with a charge-transfer resistance (Rct) at the interface between the ISM and the underlying conductor [57]. After inserting the MWCNT layer, a small Rbc can be obtained for the GPE/MWCNTs (Rbc=114.9±12 kΩ) when compared with the GPE in the absence of MWCNTs (Rbc=418±11.2 kΩ). The geometric capacitances (Cg) for GPE in the absence and presence of MWCNTs are 52.1±3.3 and 141.8±7.4 pF. At the low frequency region, the charge-transfer resistance (Rct) and double-layer capacitance (Cdl) for GPE in the absence and presence of MWCNTs areRct= 454.5±13.2 kΩ, Cd = 22±2.1 µF and Rct = 200±13.2 kΩ, Cd = 50±4.2 µF, respectively. As a result, the presence of MWCNTs as an ion-to-electron transducer offers higher potential stability. This is because of the presence of the MWCNT layers, which facilitate the charge transport between interfaces in the SC-ISEs.

Analytical Application
The proposed sensor was successfully applied as an indicator electrode in the potentiometric precipitation titration of the Na2S, NaBr, and NaI solutions with TlNO3. Typical results for the titration of 50.0 mL of 2 × 10 −3 M S 2− , 1 × 10 −3 M Br − , and I −-solutions of each of them singly and ternary against 0.01 M TlNO3 are shown in Figure8a-d. It is clear that the concentration of sulfide, bromide, and iodide ions in solution can be accurately determined from the resulting neat titration curve providing a sharp end point.The solubility product of each salt is responsible for the obtianed inflection break, and is the controller for the order and sloping of the potential jump.

Analytical Application
The proposed sensor was successfully applied as an indicator electrode in the potentiometric precipitation titration of the Na 2 S, NaBr, and NaI solutions with

Conclusions
In summary, we presented here for the first time a solid-contact Tl + -ISE based on lipophilic multi-walled carbon nanotubes (MWCNTs) as an ion-to-electron transducer over a gold substrate with dibenzo18-crown-6 as a neutral carrier. The optimized formulation of the membrane resulted in a linear concentration range of 4.5 × 10 −6 -7.0 × 10 −4 M with a slope of 57.30 mV/decade and a limit of detection of 3.2 × 10 −7 M. The electrode revealed a stable potential over the pH range of 3.0-9.5. The introduction of a MWCNT layer as a high double layer capacitance to the gold plate electrode (GPE), offers the advantages of low electrical resistance and high response stability. The advantages and limitations of many of the previously suggested potentiometric solid-contact thallium sensors are given in Table 3. for comparison. It can be seen that the sensors suggested in the present work have several inherent advantages over many of those previously described. The proposed solid contact electrode is used successfully as an indicator electrode in potentiometric precipitation titrations for singly and ternary mixtures of sulfide, bromide, and iodide anions.

Conclusions
In summary, we presented here for the first time a solid-contact Tl + -ISE based on lipophilic multi-walled carbon nanotubes (MWCNTs) as an ion-to-electron transducer over a gold substrate with dibenzo18-crown-6 as a neutral carrier. The optimized formulation of the membrane resulted in a linear concentration range of 4.5 × 10 −6 -7.0 × 10 −4 M with a slope of 57.30 mV/decade and a limit of detection of 3.2 × 10 −7 M. The electrode revealed a stable potential over the pH range of 3.0-9.5. The introduction of a MWCNT layer as a high double layer capacitance to the gold plate electrode (GPE), offers the advantages of low electrical resistance and high response stability. The advantages and limitations of many of the previously suggested potentiometric solid-contact thallium sensors are given in Table 3. for comparison. It can be seen that the sensors suggested in the present work have several inherent advantages over many of those previously described. The proposed solid contact electrode is used successfully as an indicator electrode in potentiometric precipitation titrations for singly and ternary mixtures of sulfide, bromide, and iodide anions.