Although hydrazine is used in some fields,especially in industry, and plays an important role in human life, it is dangerousfor the environment and hazardous to health ^{[1, 2]}. Hydrazine is volatile, toxicand easily absorbed by oral, dermal, or inhalation routes ^[3]. Because of this,the measurement of hydrazine in the environment is important. Hydrazine can bemeasured using several methods, including spectrophotometry, chemiluminescence,ion-exclusion chromatography, high- performance liquid chromatography, amperometryand voltammetry ^{[4, 5, 6, 7, 8, 9, 10, 11]}. Electrochemicalmethods are very simple, sensitive, and effective for detecting differentspecies. However, many species have high oxidation potential and low current onthe surface of unmodified electrodes. Chemically modified electrodes increasethe rate of electron transfer by reducing the overvoltage for the reaction^{[12, 13]}. The oxidation of hydrazine at carbon paste electrodes (CPE) has a highoverpotential and an extremely low oxidation current. Although the oxidation ofhydrazine can be easily carried out on gold and platinum electrodes, these twometals are very expensive ^[14].

It is well known that phenol is highlydangerous to the skin, eyes, and mucous membranes in humans after shortinhalation or dermal exposure, and is quite toxic via oral exposure. Prolongedexposure to phenol can have several serious effects on human health such asanorexia, progressive weight loss, diarrhea, vertigo, excessive salivation,darkening of the urine, and blood ^[15].Therefore, the detection of phenol andphenol derivatives is also very important. Phenol and phenol derivatives havebeen measured using various methods, including spectrophotometry,high-performance liquid chromatography, amperometry, enzyme sensing, andvoltammetry ^{[16, 17, 18, 19, 20]}. Although,there are numerous reports on the individual determination of hydrazine andphenol using voltammetric methods, but few researchers have reported the simultaneousdetermination of hydrazine and phenol ^[21].

Great attention has been paid to the use ofmagnetic beads (MBs) in electrochemistry in recent years ^{[22, 23, 24]}. Magnetic nanoparticles are usuallyused owing to their super strong paramagnetic properties, low toxicity, andeasy preparation. Iron oxide nanoparticles (Fe₃O₄NPs) areone of the most important components for fabricating magnetic beads possessinghigh surface area. Owing to their good properties, magnetic nanoparticles havebeen used to modify the surface of electrodes ^[25]. At the same time, Fe₃O₄nanoparticles have been applied as catalysts in electrochemical methods becausethey increase the electrode surface area, electrical conductivity, and electrontransfer kinetics for many species ^{[26, 27]}.

Self-assembly procedures are useful for modifyingsurfaces, and have been recently applied in various fields such as surfaceprotection, molecular electronics, and fabrication of sensors and biosensors^{[28, 29]}. Self-assembled monolayers (SAM) are an elegant way to modify an electrodesurface, and can be formed by spontaneous adsorption of self-assembling moleculeson solid metals such as gold, silver, iron, nickel, and platinum ^{[29, 30, 31, 32]}. Until now only a few studies have beenreported on the self-assembly of Fe₃O₄NPs as amodification for the determination of chemical species. Herein, we report forthe first time a novel electrochemical sensor for hydrazine determination basedon self-assembled Fe₃O₄NPs as modifier and carbon pasteelectrode as the platform.

The exploitation of electrocatalytic processes at the surface of modified electrodeshas been investigated to increase the oxidation current and minimize theoxidation overvoltage of target species. In ourprevious work in electrocatalysis, a modified electrode (MBCPE/DPSPP/RGO/Fe₃O₄NPs) was designed for the determination of hydrazine and simultaneousdetermination of hydrazine and hydroxylamine ^[33]. In this work, 2-(3,4-dihydroxyphenyl)benzothiazole (DPB), an organo- sulfur compound, was used as a modifier. Themain aim was to achieve electrocatalytic oxidation of hydrazine and simultaneousdetermination of hydrazine and phenol using a self- assembling modifier (DPB)applied to the surface of Fe₃O₄ nanoparticles. Firstly, amagnet bar carbon paste electrode (MBCPE) was fabricated ^[34] and Fe₃O₄NPs/DPBwas adsorbed onto the electrode surface to prepare a MBCPE/Fe₃O₄NPs/DPBelectrochemical sensor. A schematic representation of the MBCPE/Fe₃O₄NPs/DPBbased electrochemical hydrazine sensor is shown in Scheme 1. As mentionedabove, the magnetic bar was inserted into the CPE to attract the magneticmaterial to the electrode surface, a method that can lead to a great enhancementof effective surface area with a good stability and without any loss ofmagnetic nanoparticles from the electrode surface during electrochemical tests.The obtained results revealed that the anodic peak current for the MBCPE/Fe₃O₄NPs/DPB sensor was proportional to the hydrazine concentration in two ranges(0.1-0.4 and 0.7-12.0 µmol/L), and the obtained detection limit was 18.0 nmol/L.Under optimum conditions, some kinetic parameters for hydrazine oxidation, suchas the diffusion coefficient and electron transfer coefficient, were determinedusing electrochemical methods.

Scheme 1. Schematic representation of the electrochemical detection of hydrazine based on the MBCPE/Fe3O4NPs/DPB electrode.

The oxidation potentials of hydrazine and phenolon a bare carbon paste electrode are close to each other and their differentialpulse voltammograms (DPV) overlap. The DPV of a mixture of these analytes atthe proposed electrode (MBCPE/ Fe₃O₄NPs/DPB) showed apeak potential difference (380.0 mV). This is sufficient to performsimultaneous determination of hydrazine and phenol, as conducted in thisresearch. Additionally, the fabricated electrochemical sensor (MBCPE/ Fe₃O₄NPs/DBP)was applied to the determination of hydrazine in tablet samples, and thesimultaneous determination of hydrazine and phenol in water samples.

The DPB was synthesized according to a previouslyreported method ^[35]. The chemical structure of DPB is shown in the Scheme 1.Other materials and chemicals were purchased from Merck and used as received.The electrochemical measurements were performed with an Autolab potentiostat/ galvanostat(PGSTAT-302 N, Eco Chemie, Netherlands). The experimental conditions werecontrolled with Nova 1.7 software on a PC. The working, counter, and referenceelectrodes were the magnetic bar carbon paste/Fe₃O₄NPs/DPBelectrode, a platinum electrode, and an Ag/AgCl (sat.), KCl (3 mol/L) electrode,respectively. All potentials in this research are reported with respect to thisreference electrode. Metrohm model 691 pH/mV meters were used for pHmeasurements. FT-IR experiments were performed using a spectrometer (VERTEX 70,Bruker Optics, Ettlingen, Germany) equipped with a deuterated triglycinesulfate (DTGS) detector. Scanning electron microscopy (SEM) was performed witha TESCAN VEGA3 instrument. The absorption spectra of the samples were obtainedusing an Analytik Jena SPECORD-205 spectrophotometer in the range of 190-1000nm and at 1 nm resolution.

Fe₃O₄ nanoparticles weresynthesized in accordance with the method described in Ref. ^[36]. Briefly, 5.17g FeCl₃ and 2.02 g FeSO₄·7H₂O were dissolvedin 1 L deionized water under ultrasonication. Subsequently, a NaOH solution (1mol/L) was added dropwise into the mixture over 50 min under ultrasonicationand nitrogen atmosphere, using a burette. The colour of the resulting mixturechanged from yellow to black ^[36]. The formed precipitate wascollected with a permanent magnet and washed five times with distilled water.

Carbon paste was prepared by hand mixing 0.5 ggraphite powder and three drops of paraffin (Dc 350, Merck) using a mortar andpestle. This paste was then packed into the end of a glass tube (ca. 6 mm i.d.and 10 cm long). A magnetic bar was inserted into the tube to be coated withthe carbon paste to provide a magnetic field. An electrical contact was made bypushing a conductive copper wire through the glass tube to connect with themagnetic bar ^[34]. For the attraction of Fe₃O₄NPs/DPBonto the MBCPE electrode surface, a dispersion of Fe₃O₄NPs/DPB(1 mg/mL) was dropped onto the MBCPE electrode. After washing the electrodesurface with deionized water, the fabrication of the electrochemical sensor(MBCPE/ Fe₃O₄NPs/DPB) was complete.

A self-assembled monolayer of DPB was formed onthe Fe₃O₄ nanoparticles by mixing 5 mmol DPB and 1 mg Fe₃O₄NPsin 1 mL ethanol solution for 5 h at room temperature. The mixture was placedonto the magnet bar electrode and then rinsed several times with distilledwater until all unadsorbed modifier had been washed away. The FT-IR spectrum ofthe Fe₃O₄NPs (Fig. 1(a), curve (1)) showed a weak peak at3500-3300 cm^-1, corresponding to the O-H stretch of water moleculesaccompanying the Fe₃O₄NPs ^[37]. Also shown in Fig. 1(a)curve (2), the FT-IR spectrum of DPB exhibited ʋ = 3478 (OH), 1623(C=N), 1589 and 1470 (C=C, aromatic), 1364, 1264 (C=S), 1200 (C-O), 1175, 1071,1055, 765, 721 (=C-H) cm^{-1 [35]}. The FT-IR spectrum of Fe₃O₄NPs/DPB(curve (3)) was similar to that of DPB, demonstrating that DPB molecules weresuccessfully adsorbed on the nanoparticles. All of the DPB FT-IR peaks wereseen in the Fe₃O₄NPs/DPB spectrum except ʋ = 1264(C=S) cm^-1, which indicated that the DPB molecules were connected tothe Fe₃O₄NPs via the benzothiazole group. A peak (O-H, ʋ= 3300 cm^-1) indicating the presence of free OH groups was alsoobserved, meaning that the hydroxyl groups of DPB in Fe₃O₄NPs/DPBwere free for an electrocatalytic role.

Fig. 1. (a) FT-IR spectra and (b) UV-Vis spectra of (1) Fe3O4NPs, (2) DPB, and (3) Fe3O4NPs/DPB.

Fig. 1(b) reveals the UV-Vis absorbance spectra ofthe bare Fe₃O₄NPs, DPB, and Fe₃O₄NPswith self-assembled DPB. The Fe₃O₄NPs displayed a weakabsorption band at 300 nm (curve (1)) ^[38]. DPB displayed an absorption band at360 nm (curve (2)) ^[39], while a sharp peak at 420 nm was observed in thespectrum of Fe₃O₄NPs/DPB which was related to presence ofDPB on the Fe₃O₄NPs surface. Moreover, the red shift ofthe spectrum confirmed that the DPB molecules were self- assembled on the Fe₃O₄NPs.

The surface of the MBCPE/Fe₃O₄NPs/DPBelectrode was studied by SEM, cyclic voltammetry, and electrochemical impedancespectroscopy. The morphology of bare MBCPE, MBCPE/Fe₃O₄NPsand MBCPE/Fe₃O₄NPs/DPB was studied by SEM (Fig. 2). Asshown in Fig. 2(a), the bare MBCPE had an amorphous and flat structure. Fig. 2(b)shows that the Fe₃O₄NPs-modified electrode had a highersurface area than the bare electrode because of the large amount of sphericalshapes on its surface. Fig. 2(c) reveals that the Fe₃O₄NPs/DPBsurface was not rougher than that of the Fe₃O₄NPs. Inaddition, the conjoined spherical-like shapes observable in Fig. 2(c) areprobably indicative of the self-assembly of the DPB molecules on the surface ofthe Fe₃O₄NPs.

Fig. 2. SEM images of (a) MBCPE, (b) MBCPE/Fe3O4NPs, and (c) MBCPE/Fe3O4NPs/DPB electrodes.

The surface of different electrodes wasinvestigated by cyclic voltammetry in a solution containing the redox couple K₄Fe(CN)₆(0.5 mmol/L)/K₃Fe(CN)₆ (0.5 mmol/L) and KCl (0.1 mol/L).Fig. 3(a) shows the cyclic voltammograms of the three prepared electrodes,MBCPE, MBCPE/Fe₃O₄NPs, and MBCPE/Fe₃O₄NPs/DPBin the solution. Curve (1) in Fig. 3(a) reveals the peak current oxidation of[Fe(CN)₆]^3-/4- at the bare MBCPE. Placing Fe₃O₄NPson the surface of the electrode (curve (2)) led to an increase in the peakcurrent of the [Fe(CN)₆]^3-/4- redox couple, owing to thegreatly increased electrode surface area. After adsorption of DPB on theMBCPE/Fe₃O₄NPs electrode surface (curve (3)), the [Fe(CN)₆]^3-/4-peak current decreased owing to reductions in the effective area and number ofactive sites for electron transfer. The surface area of the MBCPE/Fe₃O₄NPselectrode was estimated to be 0.52 cm² from cyclic voltammograms ofK₃[Fe(CN)₆] (1.0 mmol/L) at various scan rates using theRandles-Sevcik equation ^[40].

Fig. 3. (a) Cyclic voltammograms of (1) MBCPE, (2) MBCPE/Fe3O4NPs, and (3) MBCPE/Fe3O4NPs/DPB electrode in KCl solution (0.1 mol/L) containing the redox couples of K4Fe(CN)6 (0.5 mmol/L) and K3Fe(CN)6 (0.5 mmol/L), scan rate 100 mV/s. (b) Nyquist plots of (1) MBCPE, (2) MBCPE/Fe3O4NPs, and (3) MBCPE/Fe3O4NPs/DPB. EIS conditions: ac potential amplitude of 5 mV, frequency range of 10 kHz to 0.1 Hz. Inset: The applied equivalent circuit to model impedance data in the presence of redox couples [Fe(CN)6]3−/4−. (c) Variation of anodic peak current versus concentration of Fe3O4NPs in KCl solution (0.1 mol/L) containing the redox couples of K4Fe(CN)6 (0.5 mmol/L) and K3Fe(CN)6 (0.5 mmol/L). (d) Variation of anodic peak current versus distance of magnetic bar of electrode surface in KCl solution (0.1 mol/L) containing the redox couples of K4Fe(CN)6 (0.5 mmol/L) and K3Fe(CN)6 (0.5 mmol/L).

The electrode fabrication process was alsocharacterized by EIS. Fig. 3(b) shows a comparison of the EIS signals of theMBCPE, MBCPE/Fe₃O₄NPs, and MBCPE/Fe₃O₄NPs/DPBelectrodes in KCl solution (0.1 mol/L) containing the [Fe(CN)₆]^3-/4-redox couple. The charge transfer resistance of the bare MBCPE was R_ct= 1.64 kΩ. When the Fe₃O₄ nanoparticles were addedto the surface of the electrode the conductance surface increased, whichdecreased its charge transfer resistance (R_ct = 0.42 kΩ;curve (2)). In contrast, when the electrode surface was covered by DPB, itselectron transfer resistance increased (R_ct = 1.76 kΩ, curve (3)).Thus, there was a good agreement between the results of the CV and EIScharacterizations of the electrode fabrication. The inset of Fig. 3(b) showsthe Randles equivalent circuit applied for analyzing the Nyquist plots. In thisequivalent circuit, R1 is the electrolyte resistance (for [Fe(CN)₆]^3-/4-(0.5 mmol/L) and KCl (0.1 mol/L)), W is the Warburg impedance resulting fromthe diffusion of ions, CPE is the constant phase element, and R2 (R_ct)is the electron transfer resistance.

The optimization of the concentration of Fe₃O₄NPssuspension applied to the electrode surface was performed (Fig. 3(c)). As shown in Fig. 3(c), ∆I (∆I= I_final - I_initial, where I_finaland I_initial are the [Fe(CN)₆]^3-/4-peak current in the presence and absence of Fe₃O₄NPs onthe surface of the MBCPE, respectively, in 0.1 mol/L KCl) was enhanced byincreasing the concentration of the Fe₃O₄NPs suspensionfrom 0.2 to 1.0 mg/mL, beyond which it leveled off. These resultsindicated that increasing the concentration of the Fe₃O₄NPssuspension to 1.0 mg/mL allowed the total surface of the MBCPE electrode tobecome saturated with the NPs, and that increases beyond this value this had noeffective influence on the surface of the designed electrode. According tothese observations, 1.0 mg/mL was chosen as theoptimum concentration of Fe₃O₄NPs for application to theelectrode surface.

Another parameter which must be optimized in thiswork is the distance of the magnet bar from the surface of the preparedelectrode, because of its significant effect on the attraction of the magneticnanoparticles onto the surface of electrode. As Fig. 3(d) indicates, with increasingdistance of the magnetic bar from the electrode surface (from 0.2 to 0.5 cm), ∆I(∆I = I_MBCPE - I_CPE, where I_MBCPEand I_CPE are the peak current measured at CPE and MBCPE,respectively, in a solution containing [Fe(CN)₆]^3-/4-(0.5 mmol/L) and KCl (0.1 mol/L)) was approximately constant until a certainpoint, and then decreased. This result can be explained by a decreasing amountof Fe₃O₄NPs attracted to the electrode surfacewhen the distance was increased to more than 0.5 cm, so 0.5 cm was selected asthe optimum distance of the magnetic bar from the electrode surface.

Cyclic voltammetry was used to evaluate theelectrocatalytic oxidation of hydrazine in the presence of the DPB modifier.The electrochemical parameters of DPB are reported in Ref. ^[35]. Cyclicvoltammograms were recorded at the MBCPE, MBCPE/Fe₃O₄NPs,and MBCPE/Fe₃O₄NPs/DPB electrodes in a phosphate buffersolution (pH = 7.0). The typical processes associated with the electrochemicaloxidation of hydrazine are illustrated in Fig. 4. Curves (1) and (2) in Fig. 4show cyclic voltammograms of a phosphate buffer solution (0.1 mol/L; pH = 7.0)at the MBCPE in the absence and presence of hydrazine (10 µmol/L),respectively. Curve (3) shows the cyclic voltammogram at MBCPE/Fe₃O₄NPsin presence of hydrazine. Curves (4) and (5) exhibit the cyclic voltammogramsat MBCPE/ Fe₃O₄NPs/DPB in the absence and presence ofhydrazine, respectively. Comparison of curves (1) and (2) reveals that hydrazinewas oxidized at 0.70 V on the MBCPE electrode in the absence of modifier. Theoxidation current of hydrazine was enhanced in the presence of Fe₃O₄nanoparticles. In the presence of DPB, the hydrazine was oxidized at 0.15 V; apotential about 550 mV more negative than that observed in absence of DPB. Asthis figure shows, hydrazine was more easily oxidized in the presence of DPB,and the current at MBCPE/Fe₃O₄NPs/ DPB was enhanced byalmost two times in the presence of hydrazine. These results confirm thatelectrocatalytic behavior was obtained for the oxidation of hydrazine in thepresence of DPB as a heterogeneous modifier on the surface of MBCPE/ Fe₃O₄NPs.

Fig. 4. Cyclic voltammograms of phosphate buffer solution (0.1 mol/L; pH = 7.0) at scan rate 20 mV/s. (1) MBCPE in absence of hydrazine; (2) MBCPE in presence of hydrazine (10 µmol/L); (3) MBCPE/Fe3O4NPs in presence of hydrazine; (4) MBCPE/Fe3O4NPs/DPB in absence of hydrazine; (5) MBCPE/Fe3O4NPs/DPB in present of hydrazine.

The influence of variables including pH, DPBconcentration, and time for self-assembly was studied. As previously explained,a self-assembled monolayer of DPB was formed on the Fe₃O₄nanoparticles when different amounts of DPB and 1.0 mg of Fe₃O₄NPswere mixed in 1.0 mL ethanol solution for 5 h at room temperature. Theinfluence of the concentration of DPB in the Fe₃O₄NPsuspension (1.0 mg/mL) on the oxidation peak current of hydrazine was studiedfor mass ratios of 1:1 to 7:1 for DPB/Fe₃O₄NPs at a scanrate of 20 mV/s. This experiment was carried out for three differentconcentrations of hydrazine (0.1, 0.2, and 0.3 µmol/L). The results showed thatincreasing the concentration of DPB in the Fe₃O₄NPssuspension up to a mass ratio of 3:1 increased the measured peak current.Further increases in the ratio of DPB had no effect on the magnitude of thehydrazine peak current. This result revealed that the total surface of the Fe₃O₄NPswas saturated with DPB at a DPB/ Fe₃O₄NPs mass ratio of3:1. Therefore, mass ratio of 3:1 for DPB: Fe₃O₄NPs wasselected as the optimal mass ratio.

The optimum time for the self-assembly of DPB onthe Fe₃O₄NPs was also examined. As shown in Fig. 5(a), asthe time allowed for the self-assembly of DPB on the Fe₃O₄NPswas increased up to 5 h, ∆I (∆I = I_final - I_initial,where I_final and I_initial are the DPB peakcurrent with and without hydrazine, respectively, in a buffer solution (0.1 mol/L))first increased and then slightly decreased. Thus, the total surface of the Fe₃O₄NPsappeared to be saturated with DPB within 5 h, and additional time did notinfluence the amount of DPB self-assembled on the Fe₃O₄NPs.Therefore, 5 h was selected as the optimal period for the self-assembly of DPBon the Fe₃O₄NPs.

Fig. 5. (a) Variation of anodic peak current of DPB versus time of self-assembles in phosphate buffer solution (pH = 7.0); (b) Variation of anodic peak current of DPB versus pH in presence of hydrazine at phosphate buffer solution.

The effect of pH on the oxidation peak ofhydrazine at the MBCPE/Fe₃O₄NPs/DPB electrode wasinvestigated by cyclic voltammetry in buffer solution (0.1 mol/L) at various pHfrom 2.0 to 12.0. The pH was found to affects electrocatalytic oxidation of DPBin the presence of hydrazine. It can be seen in Fig. 5(b) that ∆Iincreased from pH = 2.0 to 7.0, and then decreased after pH = 7.0 (7.0-12.0).Therefore, pH = 7.0 was selected as the optimum pH for the followingexperiments.

The effect of scan rate on the electrocatalyticoxidation of hydrazine at MBCPE/Fe₃O₄NPs/DPB wasinvestigated by cyclic voltammetry. Fig. 6(a) shows the cyclic voltammograms ofMBCPE/Fe₃O₄NPs/DPB at various scan rates obtained in aphosphate buffer solution (0.1 mol/L; pH = 7.0) containing hydrazine (4 µmol/L). Fig. 6(b) shows a selected region of theTafel plot for the cyclic voltammogram obtained at a scan rate of 20 mV/s. Thehydrazine oxidation current was found to increase linearly with the square rootof scan rate (Fig. 6(c)), indicating that the reaction was mass transfercontrolled. Also, a plot of normalized current (I_p/ν^1/2)versus sweep rate (Fig. 6(d)) revealed that increasing ν caused thenormalized currents (I_p/ν^1/2) to decreaseuntil a certain level was reached. This confirms that the oxidation ofhydrazine at MBCPE/ Fe₃O₄NPs/DPB had an EC' catalytic mechanism.Fig. 6(e) reveals the Tafel plot drawn from the data of the rising part of theplotted current-voltage curve obtained at a scan rate of 20 mV/s. The slope ofthe Tafel plot was calculated to be 0.204 V/decade. The charge transfercoefficient between hydrazine and MBCPE/Fe₃O₄NPs/DPB wascalculated to be α = 0.69 using the slope of Tafel plot, the equation2.3RT/n_α(1-a)F, and assumingn_α = 1.

Fig. 6. (a) Cyclic votammograms of MBCPE/Fe3O4NPs/DPB in phosphate buffer (0.1 mol/L; pH = 7.0) containing hydrazine (4.0 µmol/L) at different scan rates. The curves from down to up correspond to 20, 40, 60, 80, 100 , 150, 200 and 300 mV/s scan rates, respectively. (b) The selected region for plotting Tafel diagram in scan rate 20 mV/s for electrocatalytic oxidation of hydrazine at MBCPE/Fe3O4NPs/DPB. (c) Variation of the electrocatalytic peak currents vs. ν1/2. (d) Variation of the scan rate-normalized peak current (Ip/ν1/2) versus scan rate (ν). (e) Tafel plot derived from the rising part of current potential curve recorded at scan rate of 20 mV/s.

The chronoamperometric behavior of MBCPE/Fe₃O₄NPsin a phosphate buffer solution containing DPB was examined in the absence andpresence of hydrazine. Chronoamperometric measurements of variousconcentrations of hydrazine solutions were carried out by setting the potentialof working electrode at 250 mV versus the Ag/AgCl electrode ^[41]. The obtainedchronoamperograms are depicted in Fig. 7(a). The diffusion coefficient ofhydrazine was determined from the results. In the presence of hydrazine at anexperimental time of 0-20 s, the modifier was oxidized and the rate ofelectrocatalyzed hydrazine oxidation increased with the diffusion of hydrazinefrom the bulk to the modifier on the electrode surface, thus increasingoxidation current. The diffusion coefficient (D) of hydrazine can beevaluated using the Cottrell equation, which describes relationship betweendiffusion coefficient and the hydrazine bulk concentration ^{[42, 43]}.

Fig. 7. (a) Chronoamperogrames of MBCPE/Fe3O4NPs/DPB in phosphate buffer solution (0.1 mol/L) containing 0, 1, 2, 3 and 4 µmol/L of hydrazine (from down to up). (b) Plot of I vs. t −1/2 obtained from chronoamperometric experiments. (c) The slopes of the resulting straight line of (b) vs. hydrazine concentration.

where D and Care the diffusion coefficient (cm²/s) and the bulk concentration(mol/cm³) of hydrazine, respectively. A is the effectiveelectrode area (A = 0.52 cm²) and n is the number ofelectrons transferred (n = 4) in the oxidation reaction. Fig. 7(a) showsthat the level of Cottrell current increased with the hydrazine concentration.Fig. 7(b) shows the plot of I vs. t ^-1/2.The slopes of the straight lines from the Fig. 7(b) are plotted versushydrazine concentration in Fig. 7(c). From the slope of this line, the diffusioncoefficient of hydrazine was calculated to be 7.19 × 10^-6 cm²/s,comparable with the previously reported values of 8.2 × 10^-6 cm²/s^[14] and 4.5 × 10^-6 cm²/s ^[12].

The chronoamperometric method was also applied toevaluate the catalytic rate constant, k (L/(mol·s)), of the reactionbetween hydrazine and DPB using the Galus method ^[44].

where I_cis the catalytic current of DPB in the presence of hydrazine at MBCPE/Fe₃O₄NPs,I_l is the limited current in the absence of hydrazine, and γ= kCt (where C is the bulk concentration of hydrazine) is theargument of the error function. In cases where γ is more than 2, theerror function is almost equal to 1 and the above equation can be reduced to:

where k, C,and t are the catalytic rate constant (L/(mol·s)), catalystconcentration (mol/L), and time elapsed, respectively. I_c andI_l are the currents of MBCPE/Fe₃O₄NPs/DPBin a buffer solution (pH = 7.0) in the presence and absence of hydrazine,respectively. The values of I_c/I_l versus t^½were plotted, and from the slope of this plot, the average value of kwas calculated to be 5.1 × 10³ L/(mol·s).

The following processes are suggested for themechanism of the electrocatalyzed hydrazine oxidation reaction. (1) Firstly,DPB is oxidized on the surface of MBCPE/Fe₃O₄NPs. (2) Electrocatalyticreaction between the oxidized form of DPB and hydrazine takes place on thesurface of the electrode, increasing the anodic peak current. Fig. 8(a) shows cyclic voltammograms ofMBCPE/Fe₃O₄NPs/DPB in the presence of different concentrationsof hydrazine. With increasing hydrazine concentration, the oxidation peakcurrent increased while the cathodic peak current decreased. These resultsconfirm an electrocatalytic behavior for hydrazine oxidation on the surface ofMBCPE/Fe₃O₄NPs/DPB via an EC' mechanism. This mechanismis shown in Fig. 8(b). Hydrazine is oxidized in the catalytic chemical reaction(C') by the oxidized form of DPB, which is produced via an electrochemicalreaction (E). Hydrazine in aqueous solutions can be oxidized with the loss offour electrons and the production of nitrogen gas ^{[45, 46]}.

Fig. 8. (a) CVs of MBCPE/Fe3O4NPs/DPB in phosphate buffer solution (0.1 mol/L; pH = 7.0) at scan rate 10 mV/s in 0, 0.2, 0.4, 0.6 and 0.8 µmol/L hydrazine from bottom to top, respectively. (b) The electrocatalytic reaction mechanism for hydrazine on the surface of MBCPE/Fe3O4NPs/DPB.

Differential pulse voltammetry (DPV) is the mostwidely applicable technique in electrochemistry owing to its high sensitivityand low charging current contribution to the background current, so thistechnique was used to determine the limit and the linear range of hydrazinedetection at the present electrodes. The effect of hydrazine concentration onthe DPV at the electrodes is presented in Fig. 9(a). As Figs. 9(b) and (c)show, the plot of peak current vs. hydrazine concentration exhibited two linearranges, the first 0.1-0.4 μmol/L with a correlation coefficient of 0.9995 andthe second 0.7-12.0 μmol/L with a correlation coefficient of 0.9705. Based on 3s_b/m(where s_b is the standard deviation of a blank solution and mis the slope of the calibration curve (Fig. 9(b)), the detection limit was determinedto be 18.0 nmol/L. The limit of detection, linear range, optimal pH for thecatalytic effect, and peak potential shift of a number of previously reportedmodified electrodes against that of the proposed electrode for thedetermination of hydrazine are listed in Table 1 ^{[8, 31, 32, 47, 48, 49, 50]}. According toTable 1, the detection limit of the proposed electrode sensor was lower thanthat of all other methods except for that of Ref. 48. The very low detectionlimit for determination of hydrazine obtained in this work arose from thepresence of the magnetic bar in the CPE electrode, which caused the electrodesurface to adsorb a large amount of Fe₃O₄NPs/DPB andthereby greatly increased the effective area of the electrode.

Fig. 9. (a) Differential pulse voltammograms of MBCPE/Fe3O4NPs/DPB in phosphate buffer solution (0.1 mol/L; pH = 7.0) containing different concentrations of hydrazine. (b) Calibration curve in the range of 0.1 to 0.4 µmol/L. (c) Calibration curve in the range of 0.7 to 12.0 µmol/L.

Table 1 Comparison of the efficiency of some modified electrodes used for determination of hydrazine. Method Electrode Peak potential shift (mV) pH Dynamic range (μmol/L) Detection limit (μmol/L) Ref. Amp WO3 700 7.0 100−1000 144 [49] CV PANI/Pd0 — 7.0 0.3−12000 0.05 [50] CV PEDOP/MWCNT-Pd — 7.4 0.1−5000 0.04 [1, 2, 3] DPV p-APMWCNTPE 600 7.0 0.5−175 0.3 [51] Amp AG/Au-NPs 400 7.0 0.002−936 0.00057 [52] DPV BFT-CNT-GCE 170 8.0 0.5−700.0 0.033 [1] DPV Co(II)BBAEDI-MWCNT-CPE 500 7.0 0.3−70 0.1 [2] DPV MBCPE/ Fe3O4/DPB 550 7.0 0.1−12 0.018 This work CV = Cyclic voltammetry; DPV = Differential pulse voltammetry; Amp = Amperometry; p-APMWCNTPE = p-Aminophenol-multiwall carbon nanotube-paste electrode; BFT = 1-Benzyl-4-ferrocenyl-1H-[1, 2, 3]-triazole; PEDOP = Poly-ethylendioxy pyrrole; CNT = Carbon Nano tube; CPE = Carbon paste electrode; AG = Activated graphite.

Table 1 Comparison of the efficiency of some modified electrodes used for determination of hydrazine.

Several cations and anions including Na⁺,Cs⁺, NH₄⁺, Mg²⁺, Ca²⁺,Ba²⁺, Cd²⁺, Zn²⁺, Pb²⁺, Mn²⁺,Cu²⁺, C₂O₄^2-, NO₃^-,CH₃COO^-, Br^-, F^-, I^-were checked for interference with the hydrazine oxidation reaction. Toinvestigate the effect of these interferents, the concentration of hydrazinewas set to 1.0 μmol/L under optimum conditions. Up to a molar ratio of 1000,none of the investigated anions and cations had any interference effect on thehydrazine oxidation peak current. The checking of compounds such ashydroxylamine that may be present in water matrix was also carried out, and itwas found that hydroxylamine had a serious interference effect.

Hydrazine and phenol are two important hazardouscompounds that exist together in water samples. Owing to small differencebetween the oxidation potentials of hydrazine and phenol, the simultaneousdetermination of these two components using electrochemical methods is difficultbut vital. In this work, the simultaneous determination of hydrazine and phenolwas successfully achieved using the proposed chemically modified (Fe₃O₄NPs/DPB)electrochemical sensors.

DPV was used for the simultaneous determination ofhydrazine and phenol at MBCPE/Fe₃O₄NPs/DPB, because ofthe better selectivity and lower detection limit of this technique comparedwith those of other electrochemical methods. At first, the intermolecularinteraction between the two compounds was investigated at MBCPE/Fe₃O₄NPs/DPBand then the concentration of phenol was varied while the hydrazine concentrationwas kept constant. A hydrazine concentration of 0.3 μmol/L, the peak current ofphenol oxidation was proportional to its concentration from 100-470.0 μmol/L (∆I(μA) = 0.0061C_phenol (μmol/L) + 0.655 and R²= 0.98) and there was no change in the peak current of hydrazine. Therefore,increasing the concentration of phenol did not affect the hydrazine peakcurrent (Figure not shown). Similar experiments showed that increasing theconcentration of hydrazine (∆I (μA) = 4.38C_hydrazine (μmol/L)+ 11.44 and R² = 0.975) did not affect the peak current ofphenol at the MBCPE/Fe₃O₄NPs/DPB electrode. Thus, the twoanalytes did not exhibit any intermolecular interaction and simultaneous determinationof these two analytes without interference was possible at MBCPE/Fe₃O₄NPs/DPBsensor. Differential pulse voltammograms obtained for different concentrationsof hydrazine and phenol are plotted in Fig. 10(a). The oxidation peak currentsof hydrazine and phenol increased with their respective concentrations. Figs.10(b) and (c) exhibit the dependence of the peak current on the concentrationof hydrazine (∆I (μA) = 4.407C_hydrazine (μmol/L) +12.45, R² = 0.972 and ∆I (μA) = 16.5C_hydrazine(μmol/L) - 1.55, R² = 0.985) and phenol (∆I (μA)= 0.0062C_phenol (μmol/L) + 8.74, R² =0.991), respectively. It is notable that the sensitivity of the modifiedelectrode to hydrazine was approximately the same in the absence and presenceof phenol, which indicates that the oxidation of hydrazine and phenol atMBCPE/Fe₃O₄NPs/DPB were independent processes. From thecalibration graph (Fig. 10), the lower detection limit of phenol (3s_b/m)was found to be 24.3 μmol/L.

Fig. 10. (a) Differential pulse voltammograms of MBCPE/Fe3O4NPs/DPB in phosphate buffer solution (0.1 mol/L; pH = 7.0) containing different concentrations of hydrazine and phenol. (1)−(9) Mixed solutions of 0.2+100, 0.3+130, 0.4+160, 0.7+210, 1+290, 2+340, 3+390, and 5+420 respectively, in which the values are the concentrations of hydrazine and phenol in µmol/L. (b) Plot of the peak currents as a function of hydrazine concentration (in the linear range of 0.2−5 &mimicro;mol/L of hydrazine). (c) Plot of the peak currents as a function of phenol concentration of 100−470 µmol/L.

For repeatability investigations of the fabricatedhydrazine sensor, three parallel MBCPE/Fe₃O₄NPs/DPBsensors were constructed and then examined using hydrazine (4 μmol/L). Thethree independent hydrazine sensors were produced on the same day and allsolutions used, such as the Fe₃O₄NPs and DPB suspensions,were identical. An acceptable relative standard deviation of 3.8% (n =3) was obtained for ΔI (where ∆I = I_final -I_initial and I_final and I_initialare DPB peak currents in buffer solution (0.1 mol/L) at pH = 7.0 with andwithout hydrazine, respectively), indicating that this electrochemical sensor(MBCPE/Fe₃O₄NPs/DPB) had a satisfactory repeatability.The stability of the proposed modified electrode was also investigated. Afterstorage of the modified electrode in phosphate buffer solution (0.1 mol/L; pH = 7.0)at 25 °C for 4 d, the response of electrode retained about 94% of its primary response,a suitable stability.

As explained, hydrazine in the environment posessome risks to human life. Although the determination of hydrazine with thedesigned MBCPE/Fe₃O₄NPs/DPB sensor was found to be applicable underlaboratory conditions, an attempt was made to assess the practical analyticalperformance of the method with water samples. Five solutions containing appropriateamounts of hydrazine were prepared and analyzed by the proposed method.According to Table 2, an average recovery of 99% was obtained, which indicatesthat the typical constituents of water did not significantly interfere with thedetermination of hydrazine.

Table 2 Determination of hydrazine in water samples using proposed sensor (n = 5). Sample Hydrazine added (μmol/L) Hydrazine found (μmol/L) Recovery (%) Drinking water Drinking water Drinking water Drinking water River water* River water* River water* — 1 3 6 9 10 12 <Detection limit 1.03 (±0.08) 3.0 (±0.2) 6.1 (±0.5) 8.9 (±0.2) 10.5 (±0.1) 12 (±0.2) — 101 100 101 98 104 100 * Zayandeh-Rood in Isfahan city.

Table 2 Determination of hydrazine in water samples using proposed sensor (n = 5).

To ensure the accuracy of this method for thepractical measurement of hydrazine, the hydrazine concentration of auxiliarycooling water from Yazd Power Generation Company was analyzed by the proposedmethod, and the results were compared with those obtained using aspectrophotometric method. The hydrazine concentration of the water sample wasfound to be 150.2 (±3.8) ppb (n = 4) using the proposed method and 146.8(±3.1) ppb (n = 4) using the selected spectrophotometric method ^[51],respectively. A t-test ^[52] was carried out and t_exp wasfound to be 1.42, which is less than the critical t (2.45)within a 95% confidence level. The t-test data show that there was nosystematic error between the results obtained by the two methods. Therefore,the present electrochemical method seems to be promising for the determinationof hydrazine in water samples. The standard addition method was also used todetermine hydrazine and phenol concentrations in water samples by DPV. Theobtained results are provided in Table 3. The recovery percentages in thistable show that the proposed sensor (MBCPE/Fe₃O₄NPs/DPB)is applicable for the simultaneous determination of hydrazine and phenol inreal water samples.

Table 3 Determination of hydrazine and phenol in water sample using proposed sensor(n = 5). Sample Added (μmol/L) Found (μmol/L) Recovery (%) Hydrazine 0.00 0.20 0.30 0.40 0.70 1.00 — 0.19 0.31 0.41 0.69 1.05 — 95 103 102 98 105 Phenol 0.00 100 130 160 210 290 — 103 129 162 208 294 — 103 99 101 99 101

Table 3 Determination of hydrazine and phenol in water sample using proposed sensor

In the present study, robust new magnetic barcarbon paste electrode modified with Fe₃O₄NPs/DPB wasconstructed as an electrochemical sensor for determination of hydrazine. Theelectrochemical oxidation of hydrazine at the proposed electrode was studied bycyclic voltammetry, differential pulse voltammetry, and chronoamperometry.Compared with the response observed for unmodified MBCPE, the electrochemicalsensitivity of hydrazine at the modified electrode MBCPE/Fe₃O₄NPs/DPBwas dramatically improved. Therefore, and owing to its unique structure, thefabricated sensor (MBCPE/Fe₃O₄NPs/DPB) has a number ofadvantages over the non-modified electrode (MBCPE), including as high conductivityand fast electron transfer. Briefly, kinetic parameters such as the diffusioncoefficient and the electron transfer coefficient of hydrazine were determinedby electrochemical methods under optimum conditions. The results obtained werecomparable with those reported in the literature for similar approaches. Themethod was also successfully used for the determination of hydrazine in watersamples. The fabricated sensor was very rapid, reproducible, stable, selective,and sensitive for determination of hydrazine. Generally, this electrochemicalsensor is capable of simultaneous determination of hydrazine and phenol inwater samples. The present work may lead to the fabrication of new electrochemicalsensors based on the application of nano-materials and modifiers to variouselectrode surfaces for the determination of different compounds, especiallytoxic species.

We gratefully acknowledge the support of this workby Yazd University research council.