Sensitive determination of hydrazine using poly(phenolphthalein), Au nanoparticles and multiwalled carbon nanotubes modified glassy carbon electrode

This study reports a detailed analysis of an electrode material containing poly(phenolphthalein), carbon nanotubes and gold nanoparticles which shows superior catalytic effect towards to hydrazine oxidation in Britton–Robinson buffer (pH 10.0). Glassy carbon electrode was modified by electropolymerization of phenolphthalein (PP) monomer (poly(PP)/GCE) and the multiwalled carbon nanotubes (MWCNTs) was dropped on the surface. This modified surface was electrodeposited with gold nanoparticles (AuNPs/CNT/poly(PP)/GCE). The fabricated electrode was analysed the determination of hydrazine using cyclic voltammetry, linear sweep voltammetry and amperometry. The peak potential of hydrazine oxidation on bare GCE, poly(PP)/GCE, CNT/GCE, CNT/poly(PP)/GCE, and AuNPs/CNT/poly(PP)/GCE were observed at 596 mV, 342 mV, 320 mV, 313 mV, and 27 mV, respectively. A shift in the overpotential to more negative direction and an enhancement in the peak current indicated that the AuNPs/CNT/poly(PP)/GC electrode presented an efficient electrocatalytic activity toward oxidation of hydrazine. Modified electrodes were characterized with High-resolution transmission electron microscopy (HRTEM), scanning electron microscopy (SEM), X-ray photoelectron spectroscopy (XPS) and electrochemical impedance spectroscopy (EIS). Amperometric current responses in the low hydrazine concentration range of 0.25–13 µM at the AuNPs/CNT/poly(PP)/GCE. The limit of detection (LOD) value was obtained to be 0.083 µM. A modified electrode was applied to naturel samples for hydrazine determination.

Metal nanoparticles have attracted considerable interest due to their unique properties such as large surface to volume ratio higher electronic and optical effects and catalytic properties as compared to the bulk equivalents [23]. It has been used many electrochemical fields due to mass transport, high catalytic activity and high effective surface area [24]. Metal nanoparticles such as palladium, gold, copper, platinum and silver have been found applications in many electrochemical fields [25].
Carbon nanotubes (CNTs) have large surface area, thermal stability, low electrical resistance, increased porosity and high electronic conductivity compared to the other carbon-based materials. They are used support materials to modify electrode surface and provide opportunities in sensor applications [26,27].
In the present study, we demonstrated a simple and versatile in situ approach for the fabrication of AuNPs/CNT/ poly(PP)/GCE. Firstly, for this prurpose, GC electrode surface was covered with polyphenolphthalein [poly(PP)] by electropolymerization of phenolphthalein monomer using cyclic voltammetry. Then, carbon nanotube suspension was injected on the modified electrode surface. Finally, Au nanoparticles were electrodeposited on the CNT/poly(PP)/GCE surface. The prepared electrode was used for the determination of hydrazine at real samples. The modified electrode surfaces were characterized by XPS, EIS, HRTEM and SEM-EDX. The obtained modified electrode had good sensitivity, selectivity, and electrocatalytic activity for hydrazine.

Chemicals and apparatus
Electrochemical measurements were carried out using Autolab 101 and 128N potentiostat electrochemical analyzer equipped with a three electrode system consisting of working electrode (glassy carbon electrode), an auxiliary platinum wire and an Ag/AgCl(sat.KCl) used as reference electrode. All electrochemical measurements performed at room temperature. Surface characterization of the modified electrodes were performed by using JEOL JEM 2100F highresolution transmission electron microscopy (HRTEM) (JEOL Ltd., Tokyo, Japan), Thermo K-Alpha-monochromated high-performance XPS spectrometer (Thermo Fisher Scientific Inc.,Waltham, MA, USA) and Zeiss Gemini 500 SEM (Carl Zeiss Microscopy GmbH, Oberkochen, Germany).
Hydrazine was purchased from Aldrich Chemical Co. Inc. (Milwaukee, WI, USA) and phenolphthalein (PP) was obtained from Honeywell Riedel-de Haën GmbH (Seelze, Germany). Ethanol was used to prepare phenolphthalein solution. Multiwalled carbon nanotubes (CNT abbreviation used instead of MWCNT in electrode name) (purity>95% diameter 110-170 nm, length 9 µm) were purchased from Aldrich Chemical Co. Inc. Acetic acid (CH 3 COOH), boric acid (H 3 BO 3 ) and phosphoric acid (H 3 PO 4 ) (all 0.04 mol L −1 ) were used to preparing Britton-Robinson (BR) buffer solution. The pH of BR buffer solution was adjusted with 3 M NaOH solution. Distilled water was used to prepare all other solutions. 2.1. Modification of electrode GCE was activated by polishing with different grade of Al 2 O 3 slurry on a synthetic cloth and rinsed with ethanol and pure water. Electrochemical polymerization of PP was carried out in 0.1 M NaNO 3 containing 1.0 mM PP solution by cyclic voltammetry. The polymerization voltammogram was obtained by repetitive 15 current-potential cycles from -1.2 V to +1.8 V. A 10 µL of multiwalled carbon nanotube suspension was injected on the poly(PP)/GCE surface and then the solvent evaporated by IR lamp exposure for 15 min. Finally, Au nano particles was deposited on the CNT/poly(PP)/GCE by cyclic voltammetry scanning between 0.60 V --0.90 V in HAuCl 4 with a scan rate of 100 mVs -1 for 10 cycles.

Sample analysis
The hydrazine content in Gediz and Karaçay Rivers waste water samples collected from industry in Manisa (Turkey) was investigated on the AuNPs/CNT/poly(PP)/GC electrode. Samples were prepared for analysis by first filtering through filter paper. The pH and conductivity of Gediz and Karaçay Rivers samples were measured to be pH 7.91, 1.46 mS and pH 7.82, 2.79 mS, respectively. 1 mL of water sample and 9 mL of pH 10.0 BR buffer solution was taken into voltammetric cell with a total volume of 10 mL. Then, determination of hydrazine was performed by using standard addition method with linear scanning voltammetry.

Electropolymerization of PP and electrodeposited of gold nanoparticle
The poly(PP) modified electrode was prepared by electrochemical polymerization of PP monomer on the GCE surface between -1.2 V and +1.8 V 200 mV s -1 in 0.1 M NaNO 3 containing of 1.0 mM PP monomer ( Figure 1A) [28]. During the electropolymerization process, in the anodic direction, two anodic peaks were observed at about +0.51 V and +1.04 V, while in the cathodic direction, two cathodic peaks were observed at about -0.5 V and +0.25 V. The observed anodic and cathodic peak currents were increased by consecutive current potential scan cycles number. In order to determine the most suitable polymerization electrolyte medium for phenolphthalein, the responses of the polymer film coated to the GCE surface in the presence of NaNO 3 in different concentrations were investigated to 1 mM hydrazine ( Figure S1). The optimum NaNO 3 concentration was chosen as 0.1 M. The number of cycles and the monomer concentration can be controlled to adjust the thickness of the polymer. The optimum cycle number in PP electrochemical polymerization process was obtained as 15 cycles. Figure 1B shows that the deposition of Au nanoparticles on the CNT/poly(PP)/GC electrode from 2.0 mM HAuCl4 subscripts solution (in 0.1 M HCl) by current potential scans from 0.60 V to -0.90 V for 10 cycles. The amount and size of Au nanoparticles deposited on the electrode surface, can be controlled by the electrodeposition process.

Surface characterization
The morphology of poly(PP)/GCE (A), poly(PP)/CNT/GCE (B), AuNPs/CNT/poly(PP)/GCE (C) and (D) was examined by SEM, as shown in Figure 2. In Figure 2A, homogeneously distributed porous polymer surface was formed as a result of the polymerization of the phenolphthalein monomer on the bare electrode surface. The morphology of poly(PP)/ CNT/GCE showed a network-like structure in Figure 2B. Nano Au particles were formed as bright, round-shaped which homogeneously distribution of Au nanoparticles on the modified electrode surface were seen in Figure 2C. The average particle size on AuNPs/CNT/poly(PP)/GCE was 200 nm. Au nanoparticles appear to be small spheres lined up on the CNT surface. Each of these spheres is composed of multiple Au particles and is lined up as stacking balls. Energy dispersive X-ray spectroscopy (EDX) study confirmed the presence Au nanoparticles decorating CNT/poly(PP)/GCE with the electrochemical deposition ( Figure 2G).
The polymer and Au nanoparticles on AuNPs/poly(PP)/CNT/GCE were also examined by transmission electron microscopy. To obtained HRTEM images, the modified electrode surface was scraped and the obtained solid was taken into microcentrifuge tubes and analysed by adding 1 mL of high purity methanol. As shown in Figures 2D-2F, HRTEM images were taken at different magnification rates. It was demonstrated that poly(PP) structures formed as round spheres on CNT ( Figures 2D and 2E). Au nanoparticles formed by combining as a ball on the modified electrode. The diameter of these ball-shaped Au nanoparticles was found to be 30 nm. At the same time, Au nanoparticles forming this sphere were presented in the form of agglomeration. The diameter of the Au nanoparticles in this agglomeration is about 3 nm. The distance between Au atom arrays in each Au nanoparticle was calculated to be 0.31 nm ( Figure 2F).
When comparing SEM and TEM images, Au nanoparticle diameter was not the same due to the different preparation of SEM and TEM sample preparation procedures. In the SEM study, the electrochemical modified electrode surface can be accessed without any additional treatment. However, in the case of TEM operation, after the modified surface prepared electrochemically with both polymer and Au nanoparticles, the content must be scraped from the electrode surface with a sharp blade and transferred to the methanol solution and then to the TEM grids. Au nanoparticles can be released into small particles during TEM sample preparation steps compared to SEM conditions. Therefore, Au nanoparticle size in SEM images is higher than TEM images. Characterizations and states of the components on the AuNPs/poly(PP)/CNT/GCE electrode surface were made using XPS spectra. Survey spectra recorded in a wide energy range, Figure 3A show the positions of the main characteristic peaks of the elements. These peaks are O1s, C1s, Au4f on the electrode surface ( Figure 3B). The C signal, which has a binding energy of approximately 285 eV, originates from the polymer and corresponds to the core level of C1s [29]. The O signal having a binding energy of approximately 535 eV corresponds to the core level of O1s [30]. Two sharp peaks of Au4f were obtained (4f7/2 and 4f5/2) and their binding energies are 84.04 and 87.88 eV, respectively. These results should be attributed to the metallic gold form (Au 0 ) [31,32].  R(R(CW)) circuit ( Figure 3D inset). Figure 3C shows that poly(PP)/GCE had a large charge transfer resistance of about 1270 ohm (Table S1). The AuNPs/CNT/poly(PP)/GCE shows small semicircle diameter than of the CNT/GCE and CNT/ poly(PP)/GCE. As a result, Au nanoparticles modified electrode has a lower interfacial charge transfer resistance.

Voltammetric behavior of hydrazine on bare and modified electrode
The electrochemical behaviour of hydrazine was investigated at the bare and modified electrodes by cyclic voltammetry. Figure 4 shows the cyclic voltammograms of bare GCE, poly(PP)/GCE, CNT/GCE, CNT/poly(PP)/GCE, AuNPs/CNT/ poly(PP)/GCE in presence of 1.0 mM hydrazine in pH 10.0 BR buffer solution. An irreversible oxidation peak was appeared at 596 mV, 320 mV, 417 mV, 313 mV, and 27 mV at the bare GCE, CNT/GCE, poly(PP)/CNT/GCE, CNT/poly(PP)/ GCE and AuNPs/CNT/poly(PP)/GCE, respectively. In terms of peak shape and location, the performance of AuNPs/CNT/ poly(PP)/GCE towards hydrazine oxidation was found very effective compared to bare and other modified electrode.
Significant peak current enhancement about 8.79 times higher than bare GCE and 5.79 times higher than poly(PP)/ GCE, 2.40 times higher than CNT/GCE and 2.36 times higher than poly(PP)/CNT/GCE and 1.64 times higher than CNT/ poly(PP)/GCE were observed while the peak potential was shifted in negative direction, indicating a positive catalytic effect on hydrazine oxidation (Table S2). The improved peak characteristics indicate that faster electron transfer can be achieved by modifying Au metal nanoparticles on the CNT/poly(PP)/GCE surface.

Optimization studies of hydrazine oxidation at AuNPs/CNT/poly(PP)/GCE
The thickness of the polymer film coated on the GCE and modified electrode surface strongly affects the oxidation behaviour of hydrazine. It can be easily controlled by changing the monomer concentration and repetitive cycle number during the electropolymerization procedure. The optimization graphics were shown in Figure 5. The effect of PP monomer concentration was investigated between 0.1 to 5.0 mmol L -1 . A higher peak current for hydrazine oxidation peak was observed at AuNPs/CNT/poly(PP)/GCE, that polymer film was prepared in the presence of 1.0 mM concentration of PP monomer. In addition, the number of polymerization cycles was examined between 5 and 15 cycles to control the effect of hydrazine on peak flow. The best result for hydrazine oxidation was determined as 10 cycles. The supporting electrolyte pH is significant parameter that can extremely change the analyte signal of the electrodes. To investigate the effect of write PH on the peak current and peak potential of hydrazine, the BR buffer in the pH 6.0-12.0 range at AuNPs/CNT/poly(PP)/GCE with cyclic voltammetry (Figures 5A and 5B). The highest current and low oxidation potential of hydrazine appeared in pH 10.0 BR buffer solution. For this reason, pH 10.0 BR buffer solution as supporting electrolyte was selected for further studies.
To investigate the catalytic effect of Au nanoparticles on hydrazine oxidation, the HAuCl 4 solution concentration was changed in the range of 0.1-5.0 mM and cyclic voltammograms were recorded in the presence of pH 10.0 BR buffer ( Figures 5C and 5D). As the Au 3+ concentration range from 0.1 mM to 2.0 mM, the hydrazine peak potential shifted towards negative potentials while the increasing of the peak current of hydrazine. Therefore, 2.0 mM HAuCl 4 concentration was chosen for preparation of Au nanoparticles in further studies.
As shown in Figures 5E and 5F, AuNPs/CNT/poly(PP)/GC electrode was prepared at different cycle numbers (5-15) from 1.0 mM HAuCl 4 solution by cyclic voltammetry. The amount of nanoparticle was adjusted by changing the repeated cycle number in electrodeposition process. The maximum peak current of hydrazine was obtained at 10 cycle Au deposition and decreased thereafter. The optimum cycle number was chosen 10.
The scan rate dependence of cyclic voltammograms for the modified electrode in pH 10.0 BR buffer solution containing 1.0 mM hydrazine was used to get the information about the diffusion controlled current ( Figure S2). As potential scan rate increased, the peak potential of hydrazine shifted to direction positive. The linearity of the current to the square root of the scan rate indicated that the current on the electrode surface was controlled by diffusion. Figure S2

The long-term stability
The performance of the five different electrodes toward the hydrazine oxidation reaction were tested by a chronoamperometry after a by holding -0.03 V constant potential at each electrode for 900 s. Figure 6A   where D is diffusion coefficients (cm 2 s -1 ), C is the bulk hydrazine concentration (mol cm -3 ), A is the electrode area (0.0707 cm 2 ), n: number of electrons. F is Faraday constant (96486 C mol -1 ). The J (current density) versus t -1/2 graph can be plotted from the Cottrell equation and the current in the electrode press is linear under diffusion-controlled conditions. ( Figure 6B and Table S3), If the diffusion coefficient for hydrazine is considered equal to 2.37 × 10 -5 cm 2 s -1 [33], the number of electrons can be calculated from the Cottrell equation, which is in good agreement with values reported in literature [34]. The number of electrons was calculated as 4.2 for the oxidation of hydrazine at the AuNPs/CNT/poly(PP)/GCE. With the oxidation of hydrazine in basic medium, nitrogen, water and four-electron are formed in the products as follows: To test the intraday repeatability of the modified electrode, five different modified electrodes were prepared by the same way and their responses toward the 1.0 mM hydrazine at pH 10.0 BR buffer were investigated by cyclic voltammetry ( Figure S3). The relative standard deviation (RSD) (n = 5) value of the hydrazine peak current was found to be 7.75% at the AuNPs/CNT/poly(PP)/GC electrode. The current change was plotted by measuring the peak current of 1.0 mM hydrazine on different days for 8 days in pH 10.0 BR buffer at AuNPs/CNT/poly(PP)/GCE.

Determination of hydrazine
LSV method was used for hydrazine determination at AuNPs/CNT/poly(PP)/GC electrode. Figure 7A shows that hydrazine oxidation peak was at applied potential -0.03 V. As the hydrazine concentrations were added to the pH 10.0BR buffer solution, the peak current of the hydrazine increased. As shown in Figure 7B, calibration graph was obtained with the peak current values in response to hydrazine concentrations in the range of 4-1000 µM. The linear regression equation is: i(μA) = 0.102 C(μM) +2.376, (R 2 : 0.9965). The limit of detection (LOD) was found to be 1.33 µM for AuNPs/CNT/ poly(PP)/GC electrode.
Amperometric responses of the hydrazine in the equal concentration range were compared to the GCE, CNT/GCE and AuNPs/CNT/poly(PP)/GCE in the Figure S4. Since the highest slope is AuNPs/CNT/poly(PP)/GCE, it has been used in chronoamperometric studies. For amperometric determination, hydrazine was added at increasing concentrations into stirred pH 10.0 BR buffer solution at an applied potential at -0.03 V and the amperometric current-time (i-t) graph was obtained. Figure 8 shows amperometric current responses in the low hydrazine concentration range of 0.25-13 µM at the AuNPs/CNT/poly(PP)/GCE. With each hydrazine addition, the current gradually increased. Since it reached saturation after a certain time, the peak current remained constant. When the peak current versus the hydrazine concentration was plotted, a linear calibration graph was obtained. The linear regression equation is: I(μA) = 0.0874 C(µM) + 0.0315, (R 2 : 0.9975). The limit of detection (LOD) was calculated to be 0.083 µM (at a signal-to-noise ratio of 3). The obtained results show that the chronoamperometric method determined lower hydrazine concentration. The LOD value obtained for hydrazine is 0.083 µM, it offers the opportunity to be determined below the WHO and EPA threshold value of 10 ppb (∼0.1 µM). The previous studies on the determination of hydrazine are given for comparison in Table 1.

Interference effect on hydrazine determination
Interferences such as K + , Cd 2+ , Ni 2+ , Co 2+ , Al 3+ , Fe 2+ , Cu 2+ , Cl -, NO 3 -, SO 4 2-, CrO 4 2-, and sodium acetate for the determination of hydrazine were investigated using amperometry technique at AuNPs/CNT/poly(PP)/GC electrode (Figure 9). The amperograms showed that 50 times excess of sodium acetate, K + , Cd 2+ , Ni 2+ , Co 2+ , Al 3+ , Fe 2+ Cl -, NO 3 -, SO 4 2-, CrO 4 2had no interference on determination of hydrazine. When Cu 2+ was added to the pH 10.0 BR buffer, the hydrazine peak current decreased and the precipitation occurred due to the basic medium at high concentrations. As a result, Cu 2+ showed significant interference in the determination of hydrazine. To eliminated Cu 2+ interference in the hydrazine signal, EDTA was added to the same medium and the linear sweep voltammograms were recorded ( Figure 9F). After the addition of EDTA, the peak current signal of hydrazine almost returned to its initial state.

Real sample analysis
To investigate for the determination of hydrazine in real samples, Gediz and Karaçay Rivers water were analyzed at the AuNPs/CNT/poly(PP)/GC electrode by amperometry. 1.0 mL of water sample and 9.0 mL of pH 10.0 BR buffer solution was taken into voltammetric cell with a total volume of 10.0 mL. Then, hydrazine determination was made by using standard addition method with linear sweep voltammetry. No hydrazine was detected in water samples taken from Karaçay and Gediz Rivers. For AuNPs/CNT/poly(PP)/GC electrode, recovery studies were investigated for Karaçay and Gediz Rivers samples for 2 different concentrations (50 and 500 μM). The results are given in Table 2.

Conclusion
In this study, we have developed a novel modified electrode for hydrazine determination. The AuNPs/CNT/poly(PP)/GCE was prepared electrochemically and this modified electrode demonstrated electrocatalytic activity towards the hydrazine oxidation in pH 10.0 BR buffer solution. The Au nanoparticle modified CNT/poly(PP)/GC electrode has shown a better stability, sensitivity and selectivity than the other electrodes. Since fabricated electrode negatively shifts the oxidation potential of hydrazine, it is also very promising for use as an anode electrode in direct hydrazine fuel cells.