Disposable Multi-Walled Carbon Nanotubes-Based Plasticizer-Free Solid-Contact Pb2+-Selective Electrodes with a Sub-PPB Detection Limit

Potentiometric plasticizer-free solid-contact Pb2+-selective electrodes based on copolymer methyl methacrylate-n-butyl acrylate (MMA-BA) as membrane matrix and multi-walled carbon nanotubes (MWCNTs) as intermediate ion-to-electron transducing layer have been developed. The disposable electrodes were prepared by drop-casting the copolymer membrane onto a layer of MWCNTs, which deposited on golden disk electrodes. The obtained electrodes exhibited a sub-ppb level detection limit of 10−10 mol·L−1. The proposed electrodes demonstrated a Nernstian slope of 29.1 ± 0.5 mV/decade in the linear range from 2.0 × 10−10 to 1.5 × 10−3 mol·L−1. No interference from gases (O2 and CO2) or water films was observed. The electrochemical impedance spectroscopy of the fabricated electrodes was compared to that of plasticizer-free Pb2+-selective electrodes without MWCNTs as intermediated layers. The plasticizer-free MWCNTs-based Pb2+-selective electrodes can provide a promising platform for Pb(II) detection in environmental and clinical application.


Introduction
Potentiometry with ion-selective electrodes (ISEs) is attractive for practical applications in many fields, for example, medical diagnosis and environmental monitoring [1][2][3]. The detection limit for most traditional liquid-contact electrodes based on plasticized poly(vinyl chloride) membrane was limited to micromolar range mainly due to the primary ion leaching from the inner filling solutions [4,5]. Conventional Pb 2+ -selective liquid-contact electrodes were usually prepared by incorporating a disk of elastic membrane into Philips-type electrode bodies and adding primary ions in the inner filling solutions. One solution is the addition of complexing agent (e.g., EDTA or NTA) [6], interfering ions (Et 4 NNO 3 ), [7] or ion-exchange resins (Dowex C−350) [8] to precisely control the primary ion activity in the inner filling solutions. For example, the Pb 2+ -selective liquid-contact electrodes with ethylenediaminetetraacetic acid disodium salt Na 2 (EDTA)-buffered inner solution exhibited a picomolar detection limit [9]. This type of Pb 2+ -selective liquid-contact electrodes were made of membranes glued to PVC tubing and the inner filling solutions (10 −3 mol·L −1 Pb(NO 3 ) 2 and 5 × 10 −2 mol·L −1 Na 2 (EDTA)) in contact with the reference element (Ag/AgCl in 3 mol·L −1 KCl) evaporated thoroughly at ambient temperature. Before measurements, they were conditioned in 10 −3 mol·L −1 Pb 2+ solution overnight and then 10 −9 mol·L −1 Pb 2+ solution for 2 days (background: 10 −4 mol·L −1 HNO 3 ). Then, potentials were recorded continuously with increasing Pb 2+ concentrations from lower to higher concentrations by a 16-channel interface.
In this work, MWCNTs-based Pb 2+ -selective SC-ISEs with a trace level analysis at sub-ppb concentrations are presented. The proposed potentiometric sensor is designed based on the following principles: (a) Reducing the diffusion coefficient by the employment of plasticizer-free copolymer methacrylate-n-butyl acrylate (MMA-BA), (b) the utilization of hydrophobic MWCNTs as a transducing layer deposited by solution-casting, (c) the dispersion of MWCNTs in plasticizer instead of surfactants to reduce the surfactants' interference. To the best of our knowledge, this is the first report on MWCNTs-based Pb 2+ -selective SC-ISEs with copolymer MMA-BA as membrane matrix. The obtained sensor was investigated by electrochemical impedance spectroscopy (EIS). The possibility of the formation of a thin water film at the interface was probed with a potentiometric water layer test. The influences of gases on the potential stability were also studied. The results suggest that the proposed sensor is promising for environmental and clinical Pb(II) determination.

Polymer Preparation and Characterization
The copolymers composed of MMA and BA were synthesized via thermally initiated free radical solution polymerization according to previous literature (see the scheme in Figure 1) [25]. Firstly, to remove the inhibitors, monomers MMA and BA were washed with a caustic solution (containing 5% (w/v) NaOH and 20% NaCl) in a 1:5 (monomer/caustic solution) ratio and water. The initiator AIBN was recrystallized from methanol and dried before use. Secondly, calculated amounts of monomers were added to 100 mL of dry ethyl acetate. The solution was degassed for 20 min by bubbling with nitrogen before the addition of 2,2 -azobisisobutyronitrile (AIBN). About 10 mg of AIBN was used for the polymerization. The homogeneous solution was continuously stirred and maintained at 85 • C for 16 h under an atmosphere of nitrogen. Thirdly, after the reaction was complete, the solvent was evaporated, and the precipitate was redissolved in 10 mL of 1,4-dioxane resulting in a gelatinous solution. Then, the gelatinous solution was added dropwise to 500 mL of DI water under vigorous stirring. The collected white precipitate was dissolved in 50 mL of methylene chloride, which was dried thoroughly and filtered by anhydrous Na 2 SO 4 . Finally, the transparent copolymer was obtained by evaporation and dried under vacuum at least for 2 days. The glass-transition temperature (T g ) was detected by the differential scanning calorimeter (Diamond DSC, PerkinElmer, Waltham, MA, USA) and heated scanning at a rate of 10 • C/min. The T g was then determined from DSC thermogram in Figure 2. The relative molecular mass of the copolymer was measured by gel permeation chromatography (GPC, Waters1515) with THF as solvent, as shown in Table 1.

Electrode Fabrication
Golden disk electrodes (Au, inner diameter (ID) = 2 mm, outside diameter (OD) = 6.35 mm) were applied for the fabrication of the proposed electrodes. They were polished with 0.3 µm alumina suspensions, rinsed with DI water, sonicated with ethanol and DI water separately, and finally dried under nitrogen. The cleaned electrodes were then tightly inserted into a piece of PVC tube (1 cm long, 5 mm ID and 8 mm OD) at the distal end.
The ion-selective sensing membrane cocktail (total mass 100 mg) was prepared by dissolving lead ionophore IV (2.0 wt %), NaTFPB (1.0 wt %), and MMA-BA (97 wt %) in 1 mL THF. The intermediate layer was prepared by dissolving 0.15 mg MWCNTs and 5 mg NPOE in 1 mL THF, and the mixture was sonicated for at least 20 min to obtain a uniform suspension, in a similar way to previous literature [36].
For MWCNTs-modified SC-ISEs, 100 µL of the MWCNTs suspension was drop-casted on the bare golden disk electrodes. The MWCNTs coatings were left to dry thoroughly in a desiccator. For MWCNTs-modified Pb 2+ -selective SC-ISEs, 100 µL of the ion-selective sensing membrane cocktail was evenly drop-casted on the MWCNTs-modified SC-ISEs in a desiccator. After the solvent evaporation, the fabricated electrodes were conditioned in 10 −5 mol·L −1 Pb 2+ solution for 2 days and then 10 −10 mol·L −1 Pb 2+ solution for 1 day before measurements. All the Pb(NO 3 ) 2 solutions had the same background of 10 −4 mol·L −1 HNO 3 (pH = 3.8) in which Pb 2+ is the predominating form of lead [37].

Apparatus and Measurements
The potentiometric responses were measured with a 16−channel interface (Lawson Labs, Inc.) controlled by a PCI−6281 data acquisition board and LabView 8.5 software (National Instruments, Austin, TX, USA). A double-junction Ag/AgCl/3 mol·L −1 KCl reference electrode containing 1 mol·L −1 CH 3 COOLi bridge electrolyte by Metrohm Ion Meter (Switzerland) was used. Different amounts of lead ions in the concentration range from 2.0 × 10 −12 to 1.5 × 10 −3 mol·L −1 were added progressively to 1.0 L of 10 −4 mol·L −1 HNO 3 solution in a crystallizing dish (200 mm). Before measurements, the crystallizing dish was washed with 10 −1 mol·L −1 HNO 3 solutions and pretreated overnight in 10 −4 mol·L −1 HNO 3 under magnetic stirring. The stability in time was measured by recording potentials of the developed electrodes consecutively under magnetic stirring. The activities of the ions were based on the activity coefficients, which were calculated according to the extended Debye−Hückel equation [38]. All the SC-ISEs' potential results were the average of sets of at least three membranes, which were performed in laboratory ambient temperature.
The electrochemical impedance spectroscopy (EIS) measurements were performed in 1.5 × 10 −3 mol·L −1 Pb(NO 3 ) 2 solution at room temperature, within the frequency range between 0.01 Hz to100 kHz using 100 mV amplitude at 0.2 V. All measurements were performed with a CHI 760D electrochemical workstation (Shanghai Chenhua Apparatus Corporation, Shanghai, China) with a Ag/AgCl/3 mol·L −1 KCl as reference electrode and a platinum as counter electrode.

Characterization of the Copolymer
This work reports the first plasticizer-free MWCNTs-based Pb 2+ -selective SC-ISEs based on the copolymer MMA-BA. Former literature pointed out that copolymer with T g between −20 to −44 • C had the proper physical and mechanical property for the ion-sensing membranes and functionality of an ionophore when incorporated into the membranes [39,40]. Since a low T g of the copolymer is critical to the functional polymeric ion-sensing membranes, the Fox equation was utilized to calculate the approximate T g of the copolymer based on the weight fractions and T g of the respective monomers (T g (polyMMA) = 378 K, T g (polyBA) = 218 K) [40]. In this study, to obtain a low T g below −20 • C, the calculated weight fraction of MMA-BA is about 1:3. As can been seen from Figure 2, the fabricated MMA-BA shows a low experimental T g of −25 • C. The resulting product also has a polydispersity of 1.57 and M w of about 15,487 (Table 1), giving an elastic and tough film, which correlates well with former reports [25,40]. The results indicate that it may have the right characteristics to function as Pb 2+ -selective membrane without plasticizer. Since the native anionic sites in the membrane matrix can lead to a Nernstian response even in the absence of anionic additives NaTFPB [25], the potentiometric response of the membranes made of MMA-BA and lead ionophore IV was studied without NaTFPB. The blank copolymer membranes showed no response to ions, which indicates there is few ionic impurities. Then, the copolymer MMA-BA was applied as membrane matrix for the fabrication of MWCNTs-based Pb 2+ -selective SC-ISEs. Our previous work demonstrates that the existence of surfactants deteriorated the sensitivity of electrodes [36]. Thus, to avoid the potential interference from surfactants in the intermediate layer, MWCNTs were suspended in plasticizer NPOE in the aid of sonication. Subsequently, the obtained plasticizer-free MWCNTs-based Pb 2+ -selective SC-ISEs were characterized in terms of potentiometric response, impedance measurements, and so on.

Potentiometric Behavior
The potentiometric response of the plasticizer-free MWCNTs-based Pb 2+ -selective SC-ISEs (Au/MWCNTs/(MMA-BA)-Pb 2+ -ISEs) was recorded in the Pb 2+ concentration range from 2.0 × 10 −12 to 1.5 × 10 −3 mol·L −1 . The proposed electrodes showed a Nernstian response of 29.1 ± 0.5 mV/decade over a linear range from 2.0 × 10 −10 to 1.5 × 10 −3 mol·L −1 , as shown in Figure 3. A sub-ppb detection limit of 10 −10 mol·L −1 is observed, which is calculated as the intersection of the two slopes ( Figure 3). Table 2 displays the response characteristics and sensor construction of Au/MWCNTs/(MMA-BA)-Pb 2+ -ISEs in comparison with those of reported Pb 2+ -selective SC-ISEs with lead ionophore IV. As can been seen from Table 2, the proposed Au/MWCNTs/(MMA-BA)-Pb 2+ -ISEs show the lowest detection limit so far, down to 0.1 ppb for Pb 2+ among available Pb 2+ -selective membrane with lead ionophore IV. Additionally, as shown in Figure 4, the developed electrodes exhibit fast response time of less than 30 s with a drift below 4 µV/s, which is much smaller than that of Au/POT/(MMA-DMA)-Pb 2+ -ISEs (0.4 mV/min) [18]. sites in the membrane matrix can lead to a Nernstian response even in the absence of anionic additives NaTFPB [25], the potentiometric response of the membranes made of MMA-BA and lead ionophore IV was studied without NaTFPB. The blank copolymer membranes showed no response to ions, which indicates there is few ionic impurities. Then, the copolymer MMA-BA was applied as membrane matrix for the fabrication of MWCNTs-based Pb 2+ -selective SC-ISEs. Our previous work demonstrates that the existence of surfactants deteriorated the sensitivity of electrodes [36]. Thus, to avoid the potential interference from surfactants in the intermediate layer, MWCNTs were suspended in plasticizer NPOE in the aid of sonication. Subsequently, the obtained plasticizer-free MWCNTs-based Pb 2+ -selective SC-ISEs were characterized in terms of potentiometric response, impedance measurements, and so on.

Impedance Measurements
Impedance measurements were performed to evaluate the electrochemical properties of the proposed electrodes. Figure 6  Impedance measurements were performed to evaluate the electrochemical properties of the proposed electrodes. Figure 6 compares the EIS spectra of the plasticizer-free MWCNTs-based Pb 2+ -selective SC-ISEs (Au/MWCNTs/(MMA-BA)-Pb 2+ -ISEs, circle) and plasticizer-free Pb 2+ -selective electrodes in the absence of the MWCNTs layer (Au/(MMA-BA)-Pb 2+ -ISEs, triangle). The

Influence of Oxygen and Carbon Dioxide
The importance of a MWCNTs solid-contact layer is demonstrated by Figure 7. Interferences from O 2 and CO 2 have been reported from several SC-ISEs where gases can easily permeate through the polymeric membrane and cause disturbances at the surface of the Au substrate [15,16]. More specifically, O 2 can form an oxygen half-cell affecting the phase boundary potential, while CO 2 can change the local pH at the electrode surface [49]. Therefore, the effects of O 2 and CO 2 on the potential stability of the Au/MWCNTs/(MMA-BA)-Pb 2+ -ISEs were investigated. The gas concentrations (O 2 or CO 2 ) were adjusted by bubbling these gases or Ar through the Pb(NO 3 ) 2 solutions (1.5 × 10 −3 mol·L −1 ). As exhibited in Figure 7, the Au/MWCNTs/(MMA-BA)-Pb 2+ -ISEs display good potential stability when exposed to O 2 or CO 2 . The outcome suggests that gases barely reach into the surface of the metal contact, which is probably due to the hydrophobicity of MWCNTs [36].

Influence of Oxygen and Carbon Dioxide
The importance of a MWCNTs solid-contact layer is demonstrated by Figure 7. Interferences from O2 and CO2 have been reported from several SC-ISEs where gases can easily permeate through the polymeric membrane and cause disturbances at the surface of the Au substrate [15,16]. More specifically, O2 can form an oxygen half-cell affecting the phase boundary potential, while CO2 can change the local pH at the electrode surface [49]. Therefore, the effects of O2 and CO2 on the potential stability of the Au/MWCNTs/(MMA-BA)-Pb 2+ -ISEs were investigated. The gas concentrations (O2 or CO2) were adjusted by bubbling these gases or Ar through the Pb(NO3)2 solutions (1.5 × 10 −3 mol•L −1 ). As exhibited in Figure 7, the Au/MWCNTs/(MMA-BA)-Pb 2+ -ISEs display good potential stability when exposed to O2 or CO2. The outcome suggests that gases barely reach into the surface of the metal contact, which is probably due to the hydrophobicity of MWCNTs [36].

Potentiometric Water Layer Test
The potential water film at the ion-sensing membrane/electron conductor interface acts as a localized microscopic water pool in which primary ions may accumulate [50]. The leaching of

Potentiometric Water Layer Test
The potential water film at the ion-sensing membrane/electron conductor interface acts as a localized microscopic water pool in which primary ions may accumulate [50]. The leaching of primary ions into the sample during measurements can result in poor lower detection limit. Thus, potentiometric water layer test was carried out for the plasticizer-free MWCNTs-based Pb 2+ -selective SC-ISEs. As indicated in Figure 8, the proposed electrodes were firstly conditioned in the primary ion solution of 1.5 mmol·L −1 Pb(NO 3 ) 2 . A stable potential for about 3.7 h was initially observed in Figure 8. After the primary ion solution was replaced with a discriminated interfering ion solution of 1.5 mmol·L −1 CaCl 2 , the immediate large potential shift was recorded. This phase boundary potential change corresponds well to the high selectivity behavior of plasticizer-free MWCNTs-based Pb 2+ -selective SC-ISEs (Table 3). After the CaCl 2 solution was successively changed by the initial primary ion solutions, the stable potential response for nearly 17 h revealed the elimination of the undesirable water layer. primary ions into the sample during measurements can result in poor lower detection limit. Thus, potentiometric water layer test was carried out for the plasticizer-free MWCNTs-based Pb 2+ -selective SC-ISEs. As indicated in Figure 8, the proposed electrodes were firstly conditioned in the primary ion solution of 1.5 mmol·L −1 Pb(NO3)2. A stable potential for about 3.7 h was initially observed in Figure 8. After the primary ion solution was replaced with a discriminated interfering ion solution of 1.5 mmol·L −1 CaCl2, the immediate large potential shift was recorded. This phase boundary potential change corresponds well to the high selectivity behavior of plasticizer-free MWCNTs-based Pb 2+ -selective SC-ISEs (Table 3). After the CaCl2 solution was successively changed by the initial primary ion solutions, the stable potential response for nearly 17 h revealed the elimination of the undesirable water layer.

Conclusions
This work demonstrates for the first time that a 0.1 ppb limit of detection for lead(II) was achieved by the disposable plasticizer-free Pb 2+ -selective SC-ISEs based on the copolymer MMA-BA as membrane matrix and MWCNTs as a conducting layer. With good physical and mechanical properties, the copolymer MMA-BA is suitable for the fabrication of plasticizer-free Pb 2+ -selective SC-ISEs. The obtained electrodes show a Nernstian response of 29.1 ± 0.5 mV/decade within the concentration range from 2.0 × 10 −10 to 1.5 × 10 −3 mol·L −1 Pb 2+ solution. Additionally, with high bulk capacitance and double layer capacitance, the proposed electrodes showed great potential stability due to the introduction of the MWCNTs layer. Moreover, the plasticizer-free MWCNTs-based Pb 2+ -selective SC-ISEs exhibited no obvious potential drift when exposed to O2 and CO2. The potentiometric water layer test confirms the absence of water films between the ion-selective membrane and the inner electron conductor. This work indicates that potentiometric solid-contact ion-selective electrodes for lead(II) detection has reached a performance well comparable to most advance methods.

Conclusions
This work demonstrates for the first time that a 0.1 ppb limit of detection for lead(II) was achieved by the disposable plasticizer-free Pb 2+ -selective SC-ISEs based on the copolymer MMA-BA as membrane matrix and MWCNTs as a conducting layer. With good physical and mechanical properties, the copolymer MMA-BA is suitable for the fabrication of plasticizer-free Pb 2+ -selective SC-ISEs. The obtained electrodes show a Nernstian response of 29.1 ± 0.5 mV/decade within the concentration range from 2.0 × 10 −10 to 1.5 × 10 −3 mol·L −1 Pb 2+ solution. Additionally, with high bulk capacitance and double layer capacitance, the proposed electrodes showed great potential stability due to the introduction of the MWCNTs layer. Moreover, the plasticizer-free MWCNTs-based Pb 2+ -selective SC-ISEs exhibited no obvious potential drift when exposed to O 2 and CO 2 . The potentiometric water layer test confirms the absence of water films between the ion-selective membrane and the inner electron conductor. This work indicates that potentiometric solid-contact ion-selective electrodes for lead(II) detection has reached a performance well comparable to most advance methods.