Electro-oxidation of pyrene on glassy carbon electrode modified with fMWCNTs/CuO nanocomposite

The electrochemical oxidation of pyrene, a well-known polycyclic aromatic hydrocarbon, was investigated using a glassy carbon electrode (GCE) modified with nanocomposite of copper oxide nanoparticles incorporated functionalized multi-walled carbon nanotubes (fMWCNTs). The catalytic copper oxide nanoparticles (CuONPs) synthesized through a chemical co-precipitation method was combined with the highly electrically conductive functionalized multi-walled carbon nanotubes using a simple and efficient method. Several analytical techniques were employed in characterizing the nanomaterials namely: the scanning electron microscopy (SEM), energy-dispersive x-ray spectroscopy (EDS), transmission electron microscopy (TEM), Fourier-transform infrared (FTIR) spectroscopy, x-ray diffraction (XRD), and the ultraviolet–visible (UV–vis) spectroscopy, to validate the authenticity of the synthesis. The electrochemical behaviour of the proposed electrode was investigated in 10 mM [Fe(CN)6]3-/4- via electrochemical impedance spectroscopy (EIS) and cyclic voltammetry (CV), revealing the highest current response and lowest charge transfer resistance at the hybrid nanocomposite modified electrode (GCE/fMWCNTs/CuO NPs) in comparison with the other electrodes studied in this work (GCE, GCE/CuO NPs, and GCE/fMWCNTs. The electrocatalytic efficacy of the electrodes towards pyrene oxidation was also evaluated, with a similarly outstanding increment in the oxidation peak current response and highly reduced resistance to charge transfer at the nanocomposite-modified glassy carbon electrode. This enhanced electrocatalytic activity facilitated the transport of electrons between the pyrene molecules and the nanocomposite-modified electrode which is attributable to the synergy between the functionalized multi-walled carbon nanotubes and the copper oxide nanoparticles. The low detection limit of 1.30 μM within the linear range (1.2–23.1 μM) demonstrated by the sensor indicates its high sensitivity and potential for environmental based analytical applications such as pyrene detection.


Introduction
Carcinogenicity, mutagenicity, and immune suppression are some of the many adverse health effects attributed to the product of incomplete combustion of organic materials -polycyclic aromatic hydrocarbons (PAHs) [1,2]; hence the need for timely detection and easy monitoring.
Polycyclic aromatic hydrocarbons (PAHs), a large class of ubiquitously persistent organic environmental pollutants [2] comprise two or more fused benzene rings in different configurational arrangements [3].The four-ringed model PAH, pyrene, found in petroleum and coal tar, majorly formed from the incomplete combustion of oil, oil products, and organic compounds; and a by-product of oil spills and seeps [4], accounts for a significant part of the total PAHs found in contaminated sites and is one of the most widely spread [5,6].
Commonly found in aquatic ecosystems, contaminated water or seafood from contaminated waters with higher detected concentrations at creosotes and pyrogenic discharge sites [7], it has been used as a PAH contamination indicator in waste monitoring [5] and is considered a priority pollutant by the United States Environmental Protection Agency (USEPA) [8][9][10].
The preferred PAH detection methods are the chromatographic techniques: gas-chromatography coupled with mass spectrometry and ultra-high-performance liquid chromatography (UHPLC) with UV-vis absorbance or fluorescence detection for accuracy, sensitivity, and reliability [9][10][11][12]; but the cost, time consumption, and complexity of these methods make for the difficulty of use for on-site monitoring [11].Alternative approaches such as the electroanalytical method have emerged stemming from the demand for simplicity, sensitivity, and portability for the rapid and accurate detection of these contaminants [13].
A degree of success has been achieved in the electrochemical detection of PAHs; with a need to explore more platforms for improved sensitivity and selectivity [14].Electrochemical sensor performance is highly dependent on electrode materials, and surface-modified electrodes have demonstrated significantly improved sensitivity over unmodified ones due to the intrinsic properties of the modifiers [15].Several nanomaterials are being explored as electrode modifiers in the production of electrochemical PAH sensors for the real-time detection [16] as they enhance sensitivity and selectivity due to increased conductivity and electroactive surface area [17].
Multi-walled carbon nanotubes have been extensively used as electrode modifiers for the fabrication of electrochemical sensor systems due to their large surface area, attractive electron transfer abilities, inertness, high conductivity, chemical specificity, stiffness and excellent strength, and efficient catalytic activities [15,18].Modification by substitutionally doping the nanotube lattice's parent structure have been proven to further increase the nanotubes' electrical conductivity [18], resistance to surface fouling and electrochemical reactivity with analytes.
Semiconducting metal oxide (MO) nanoparticles are also widely explored by researchers due to their significant properties such as large active surface area, high porosity, and electrical conductivity.The synergy between metal oxide nanoparticles and multi-walled carbon nanotube enhances electrochemical activities significantly [19]; as a result, the integration of copper oxide nanoparticle (a good sensing p-type semiconductor with narrow band gap and excellent electrocatalytic activity extensively used in heterogeneous catalysts for hydrocarbons oxidation) [20] with MWCNTs can boost sensor performance by enhancing the charge transfer between the analytes and the support matrices [21,22].
In a previous study [23], the effect of nanoleaves of CuO incorporated on GCE modified with MWCNTs for folic acid's electrochemical oxidation was examined.With a 15.2 nM limit of detection, and a 3.35 μA/μM sensitivity within a linear range (0.01 to 0.9 μM), the modified electrode, CuONs/MWCNTs/GCE exhibited excellent electrocatalytic and electrochemical activity, favouring the oxidation of folic acid by lowering the it's oxidation over potential.Similarly, in another report [24], GCE modified with nanocomposite of CuO nanoparticles and functionalized carbon nanotubes (CuO/fCNTs/GCE) was designed as an electrochemical sensor for rutin detection in Cinnamomum camphora (L.)J. Presl leaves.The electrochemical characteristics of the designed sensor, examined using differential pulse voltammetry, revealed enhanced selectivity and sensitivity in the electrochemical response towards rutin detection at a linear range (10-200 M), with an 11 nM detection limit and a 0.06087 A/M sensitivity.
Prior studies have demonstrated that the hybrid nanocomposite comprising fMWCNTs and CuO NPs has a potential for enhanced electrochemical activity and fast electron transport.Consequently, this work studies the electrochemical properties of glassy carbon electrode (GCE) modified with a nanocomposite of copper oxide nanoparticles (CuO NPs) and functionalised multi-walled carbon nanotubes (fMWCNTs) and its electrocatalytic activity as a highly sensitive and selective electrochemical sensor in pyrene detection.The fabricated hybrid-nanocomposite modified electrode displayed very good sensitivity towards the electrochemical oxidation of pyrene.All of the reagents were analytically graded and used devoid of extra purification.Distilled water was used for each solution preparation.The ferrocyanide-ferricyanide redox couple was prepared in phosphate buffer solution (PBS; 0.1 M) at pH 7.0 with KCl (0.1 M) supporting electrolyte using potassium ferricyanide

Experimental
), potassium chloride (KCl), disodium hydrogen phosphate (Na 2 HPO 4 ) and sodium dihydrogen phosphate (NaH 2 PO 4 ).The pyrene was dissolved in acetonitrile.The glassy carbon electrode (GCE; diameter 3 mm) purchased from CH Instrument USA and the modified glassy carbon electrodes were each employed as working electrode respectively.
Nitrogen-saturated ferrocyanide-ferricyanide redox couple was used for the electrochemical characterization while the electroanalysis were performed in an acetonitrile-water (80:20) mixture with 0.1 M lithium perchlorate (LiClO 4 ) supporting electrolyte.

Instrumentation
The AUTOLAB PGSTAT302N potentiostat -galvanostat (Eco Chemie, Utrecht, the Netherlands) operating on a Nova software version 2.1.6with the conventional three-electrode system comprising a 0.1 M KCl saturated (Ag|AgCl) reference electrode, glassy carbon as well as modified glassy carbon working electrodes, and platinum disk counter electrode was used to perform all electrochemical investigations.The EIS experiments were regulated by the AUTOLAB's frequency response analyzer (FRA32M) with Randles current model of NOVA used for the data fittings.Dry samples of the nanomaterials were used in studying the material's functional groups at a wavenumber range (4000-400 cm −1 ) on an Opus Alpha-P FT-IR spectrometer obtained from Bruker optics Incorporation, Billerica, MA, USA.The ultraviolet-visible spectroscopy analysis, wavelength ranged (200-900 nm) was conducted on a Spectroquant Prove 600 UV/VIS spectrophotometer by Merck KGaA, Darmstadt, Germany.The material crystalline structures were studied with powder x-ray diffractograms recorded at room temperature over a 2θ Braggs angle ranged (0-90°) using the advanced x-ray diffractometer, Bruker D8 (Bruker -AXS, Karlsruhe, Germany).The surface morphology accompanied by the energy dispersive x-ray studies were conducted on a Quanta FEG-250 field emission gun scanning electron microscope, FEI (ThermoFisher Scientific, Waltham, MA, USA).TEM micrographs obtained from a G2 F20 field emission gun transmission electron microscope, FEI Tecnai (Thermo Fisher Scientific, USA) were used for analysing the material internal morphology while distribution of particle size was analysed using the image J software.The experiments were all conducted at ambient temperature.

Fabrication of the electrochemical sensor 2.3.1. Preparation of CuO nanoparticles
Copper sulfate precursor salt and sodium hydroxide reducing agent were used in copper oxide nanoparticles' (CuO NPs) chemical co-precipitation synthesis as proposed in a previous work [25].100 ml of 0.2 M CuSO 4 was prepared in distilled water, heated under constant magnetic stirring and maintained at 50 °C.1.0 M NaOH was subsequently prepared and added in drops into the copper sulphate solution until a 1:1 volume ratio was reached maintaining a constant stirring of about 500 rpm throughout the experiment.A gradual change of the solution's color from blue to brown with brown precipitate was formed after which the reaction condition was maintained for about 30 mins.Thereafter, the filtration of the precipitate was carried out with repeated rinsing in distilled water and ethanol until neutral.It was then oven dried for 5 h at 80 °C, calcinated for 3 h at 300 °C and characterized using a variety of analytical techniques.

Multi-walled carbon nanotubes treatment
The purification and functionalization of the nanotubes were carried out as reported in previous literature [26,27].100 mg pristine MWCNTs slurry obtained from refluxing in 2.6 M HNO 3 for 48 h was diluted in water, filtered and thoroughly rinsed in water (distilled) to achieve a neutral pH, then oven-dried at 40 degrees Celsius.Subsequently, the residue was ultrasonicated in a 100 ml combination of concentrated H 2 SO 4 and HNO 3 (3:1 v/ v) ratio at 40 degrees Celsius for 24 h, filtered, thoroughly washed in distilled water and dried.The obtained residue was then stirred in a mixture of concentrated H 2 SO 4 and 30% aqeous H 2 O 2 (4:1 v/v) for 4 h at 70 °C, filtered, thoroughly washed in distilled water until obtaining a neutral filtrate pH, then oven-dried at 70 °C overnight and stored in an air-tight container.

Preparation of fMWCNTs/ CuO NPs nanocomposite
A similar but slightly modified method described in literature [28] was used in the preparation of the fMWCNTs/CuO NPs nanocomposite.Copper oxide nanoparticles in the appropriate quantity was added to distilled water, agitated, and ammonia was added in drops while stirring continuously until a clear blue solution was visible.An appropriate amount of fMWCNTs was subsequently added to the solution while agitating at 160 °C for 1 h.The resulting mixture was maintained static for a further hour after cooling to ambient temperature.Filtering of the nanocomposite precipitate was carried out with repeated neutralization using distilled water and alcohol.Oven drying at 80 °C for 12 h and calcination at 300 °C was also carried out.

Electrode pre-treatment
A mirror finish GCE surface was achieved by polishing using a slurry of 10 μm aluminum oxide on a silky pad.The electrode was washed, subjected to consecutive ultrasonic treatments in distilled water, ethanol, and distilled water, and then allowed to air out at ambient temperature [29].

The modified glassy carbon electrode electrochemical sensor fabrications
The drop-casting technique was used for the fabrication of the fMWCNTs/CuONPs modified GCE. 2 mg fMWCNTs/copper oxide nanocomposite, f-MWCNTs, and copper oxide nanoparticles were each dissolved in an ideal amount of dimethylformamide and ultrasonically processed for 24 h.A thin layer of these suspensions, about 3 μl, was each applied to the bare GCE's surface and allowed drying for 3 mins in an oven set to 40 °C [29] resulting in a modified glassy carbon electrode -GCE/fMWCNTs/CuO NPs, the GCE/fMWCNTs, and the GCE /CuONPs.

Fabrication of the electrochemical sensor
The glassy carbon electrodes (modified and unmodified) were characterized electrochemically in a solution of 10 mM [Fe(CN) 6 ] 3-/4-prepared in PBS (0.1 M) with KCl (0.1 M) supporting electrolyte at pH 7 by means of the cyclic voltammetry (CV) technique at a 25 mVs −1 scan rate over a (−0.5 V to +0.9 V) potential range and the electrochemical impedance spectroscopy (EIS) technique.The EIS sinusoidal modulation was set at a frequency range 10 -1 to 10 5 Hz and the oxidation potential of the analyte solutions (Versus Ag|AgCl in saturated KCl).With the use of the CV, EIS, and SWV techniques, the electrochemical oxidation of pyrene in a 0.40 mM solution at pH 8 was investigated.The impedance experiment data were fitted using the current Randles' NOVA model.

Nanomaterials characterization 3.1.1. SEM studies
The nanomaterials' surface morphology were examined using the scanning electron microscope (SEM).Figures 1(a)-(d) depicts micrographs of the synthesized CuO nanoparticles, pMWCNTs, fMWCNTs, and fMWCNTs/CuO NPs, respectively.The CuO nanostructure in figure 1(a) presents as agglomerated clusters of rod-like CuO nanosheets similar to literature [30,31], while the pMWCNTs in figure 1(b) revealed the characteristic tube-like morphology typical of multi-walled carbon nanotubes.Increased surface roughness, defects and etched cavities can be observed on the tube walls of the fMWCNTs (figure 1(c)) as a result of the functionalization process with minimal damage to the graphitic carbon evident in the retained fMWCNTs' tubular structure.The CuO nanorod clusters can clearly be seen decorated on the fMWCNTs' surface confirming that the synthesis of the nanocomposite was successful (figure 1(d)).

Transmission electron microscopy studies
To further investigate the nanomaterials' structure, examinations using transmission electron microscopy were conducted.Micrographs revealing the surface morphology of the CuO nanoparticles, functionalized MWCNTs, and fMWCNTs/ CuO NPs nanocomposite are displayed in figures 2(a)-(c) respectively.Similar to the SEM, the CuO nanostructure (figure 2(a)) exhibits a sheet-like morphology comprising an array of agglomerated nanorods consisting many small particles as previously reported in the literature [32,33].The nanorods were probably formed through the assembly of the nanoparticles [34].
Figure 2(b) displays the cylindrical tube-like morphology typical of fMWCNTs, while figure 2(c) for the nanocomposite, features a dispersion of the copper oxide nanostructure on the surface of the nanotubes.

FTIR studies
The Fourier transform infrared transmission spectra which reveals the functional groups present in CuO nanoparticles, fMWCNTs, and fMWCNTs/CuO NPs nanocomposites, recorded at ambient temperature over the 4000-400 cm −1 range are displayed in figure 3. Two intense absorption bands in the region below 1000 cm −1 were observed in the copper oxide nanoparticles spectra at 504 cm −1 and 609 cm −1 [35,36] indicating n(Cu-O) modes, and confirming the formation of pure monoclinic CuO crystals.Similar bonds were observed in the nanocomposite with a slight blue shift to 496 cm −1 and 599 cm −1 .This blue shift of the Cu-O stretching vibration bands in the nanocomposite and its slightly reduced intensity suggests that the oxide nanoparticles have been incorporated into the fMWCNTs' lattice.Other absorption bands observed in the fMWCNTs, the fMWCNTs/CuO NPs nanocomposite and their assignments are presented in table 1.

UV-vis spectroscopic studies
The nanomaterials' optical characterisics evaluated using the UV-vis spectroscopy revealed strong absorption bands in the CuO NPs spectra (figure 4) at an absorption maximum (λ max = 300 nm) showing the characteristic band of pure copper oxide similar to that obtained in an earlier report [47], confirming the formation of the     nanoparticles.The fMWCNTs (figure 4) revealed intense absorption bands at 296 nm due to the attached carboxylic (-COOH groups) on the MWCNTs' surface [48].A subtle red change in the nanocomposite's (fMWCNTs/ CuONPs) absorption bands to 298 nm (figure 4) suggests nanoparticles' entrapment in the carbon nanotube.The change in the nanocomposite's absorption band indicates the emergence of a new substance that is distinct from its constituent parts.

EDS studies
The elemental composition and material purity of the synthesized nanomaterials studied using the EDS, with spectra illustrated in figure 5, reveals peaks indicating that the synthesized CuO nanoparticles (figure 5(a)) comprises copper (Cu) and oxygen (O) elements while the nanocomposite (figure 5(b)) comprises copper (Cu), Carbon (C) and oxygen (O) elements, confirming the successful synthesis of the nanomaterials.The results show that no contaminants were present in the synthesized CuO nanoparticles and nanocomposite; similar findings have also been published elsewhere [49].The material elemental composition is displayed in the inset tables.reactions of [Fe(CN) 6 ] 3-/4-and a second pair between +0.72 and +0.67 V due to CuO redox reactions similarly reported in a previous literature [61].The anodic to cathodic peak current response ratio (I p a , /I p c , ) for the bare GCE and GCE /CuO NPs were < 1 while that of the GCE/fMWCNTs and GCE/fMWCNTs/CuO NPs were > 1 (table 2) suggesting quasi-reversible reaction process at the electrodes.ΔE p value for reversible one-electron redox processes is ideally (ΔEp ≈ 57 mV, at 298 K) and are scan rate independent.Processes with ΔE p value larger than 57 mV and dependent on scan rate are said to be quasi-reversible or irreversible systems [62].As seen in table 2, the ΔE p values of all the four electrodes are greater than 57 mV, also indicating that the redox process at the electrode surfaces were quasi-reversible.The electrodes' effective surface area were estimated to be 0.096 cm 2 , 0.047 cm 2 , 0.296 cm 2 , and 0.473 cm 2 for the bare GCE, GCE/CuO NPs, GCE/ fMWCNTs, and GCE/fMWCNTs/CuO NPs respectively using the Randles-Ševc´ik quasi-reversible system equation (2) at 298 K [63], where I p quasi is the quasireversible system's forward peak current (Amperes), A , eff is the effective electrode surface area (cm 2 ), n, number of electrons involved in the charge transfer steps, C, the bulk solution concentration (moIcm −3 ), D, the [Fe(CN) 6 ] 3-/4-diffusion coefficient (7.6 × 10 −6 cm 2 s −1 ), and v, the potential scan rate (V/s).
The obtained results establish the nanocomposite modified electrode (GCE/fMWCNTs/CuO NPs) as the electrode with the largest active surface area in comparison with the other electrodes investigated.This validates the fact that modifying the electrode with the nanocomposite increased the electrodes' active surface area, facilitating faster electron transport.Further research was done on the electrodes' electron transfer properties using electrochemical impedance spectroscopy.The figure 7(b) displays the electrodes' fitted Nyquist plots in 10 mM [Fe(CN) 6 ] 3-/4-between a 10 5 Hz to - 10 1 Hz frequency range at a +0.2 V potential.The charge transfer resistance (R ct ) which equates the Nyquist plot's semicircle diameter was found to be the lowest at the GCE/fMWCNTs/CuO NPs (figure 7(b); table 2) with the results corroborating the cyclic voltammetry results.The electrical circuits corresponding to the impedance spectra experimental data fittings are displayed in the figure 7(b) inset, where the R , ct R , s W , Q, and C dl represent the resistant charge transfer, the solution resistance, the warburg impedance, the constant phase element and the double layer capacitance respectively.The least resistance to transport of electrons was observed at the fMWCNTs/ CuO NPs nanocomposite modified electrode (9.87 Ω).The huge reduction in its interfacial electron transfer resistance in comparison with the bare GCE (253.69Ω) reveals the conducting nature of the fMWCNTs and the enhancement that the integration with CuO NPs provides, making for easy and faster transportation of electrons between the surface of the electrode and the electrolyte interface.The heterogeneous rate constant, (k 0 ) of the electrodes which indicates the intrinsic rate of electron transfer between the electrode and the analyte was evaluated using the equations (3)-( 4) [64].Electron transfer resistance, From the exchange current, i , 0 (A), obtained from equation (3), the heterogeneous electron transfer rate constant, k 0 were deduced from equation (4) The heterogeneous electron transfer rate constant (k 0 ) was determined as ( based on the kinetics of electrons [64,65].The nanocomposite modified electrode had the highest value suggesting that electron Table 2. Summary of data obtained from the investigation of modified and unmodified glassy carbon electrodes in 10 mM [Fe(CN) 6 ] 3-/4-, where E p,a is the potential of the anodic peak current, I p,a and E p,c is the potential of the cathodic peak current, I p,c .

Electrodes
E p,a E p,c I p,a I p,c (I p,a /I p,c ) ΔE p (V) A (cm 2 ) R ct (Ω) k 0 (× 10 -3 cm s transport proceeded faster on its surface.This implies that the synthesized fMWCNTs/CuO NPs nanocomposite is a better electron mediator in contrast to the other electrodes, and would thus be further explored in electrochemical sensing.

Scan rate studies
The impact of varying the CV scan rate over a 25-1000 mV s −1 range in 10 mM redox probe, [Fe(CN) 6 ] 3-/4- (pH 7) for insights into the electron transfer process occurring on the GCE/fMWCNTs/CuO NPs electrode surface was examined.With increasing scan rates ( ) v as presented in the cyclic voltammograms in figures 8(a), a corresponding increment in the redox peak currents with positive and negative shifts in the anodic and cathodic peak potentials, E p a , and E , p c , respectively, was detected.A linear dependence was established between the redox peak currents, I p a , and I p c , and the scan rates' square root ( ) / v 1 2 for the whole range of scan rates investigated (figure 8(b)).Equations ( 5) and (6) display the linear regression equations: The regression values with a 0.99 correlation coefficient suggests a diffusion controlled electron transfer reaction process on the GCE/fMWCNTs/CuO NPs surface.A graph of the anodic peak current's logarithm as a function of the scan rate's logarithm was evaluated to further substantiate the proposed reaction mechanism on the GCE/fMWCNTs/CuO NPs surface.Theoretically, pure diffusional and adsorption processes exhibit gradients of 0.5 and 1, respectively; however, gradients having values between 0.5 and 1 signify mixed diffusionadsorption.Processes with gradient values between 0.75 and 1.0 are categorized as pure adsorption controlled, gradient values between 0.6 and 0.75 are categorized as mixed diffusion-adsorption controlled, and gradient values between 0.2 and 0.6 are categorized as diffusion controlled [66][67][68].A straight line with a 0.58 slope resulted from the graph of ( (7), figure 8(c)) confirming diffusion controlled process for the electrochemical reaction of [Fe(CN) 6 ] 3-/4-on the GCE/fMWCNTs/CuO NPs surface.The peak potentials were linearly dependent on scan rates ( ) v for faster scan rates (200-1000 mVs −1 ), obeying Laviron's equation [69] (figure 8(d)).The coefficient of charge transfer a and the rate constant of apparent charge transfer, k s can be deduced from slopes obtained from the linear regression equations ( 8) and (9) derived from the linear portions of the peak potentials' plot (E p ) against the scan rate's logarithm (figure 8(d)) by comparing with equations (10) and (11) [70]: and, Where E 0 , is the standard potential, E p a , the anodic peak potential, E , , the cathodic peak potential, R, the gas constant (8.314Jmol −1 K −1 ), a , a the anodic transfer coefficient, a , c the cathodic transfer coefficient, a n , the number of electron involved in rate limiting step, DE, the anodic and cathodic peak-to-peak potential separation respectively, ( ) T, the temperature 298K , F, the Faraday's constant (96500 C mol −1 ), and v is the scan rate.
The gradient of the linear portion of the anodic peak potential on the graph can be deduced as while the gradient of the linear portion of the cathodic peak potential on the graph can be deduced as a when the equations (8) and (9) are combined with the equations (10) and (11), respectively [69] .Accordingly, the values a a and a c were deduced as 0.67 and 0.33, respectively, while the heterogeneous rate of electron transfer, k s in the redox probe was estimated to be 0.16 cms −1 at 25 mVs −1 using equation (12) [70,71]: In accordance with Laviron's equation, the Tafel value (b) in equation (13) deduced from equations (8) and (9) as 0.36 V/dec surpassed the expected standard (0.118 V/dec) value for one-electron processes, indicating the electrode surface's adsorption of reactants or intermediates.This type of adsorption has been linked to the strong binding force between the electrode modifier and the analyte [72].

Effect of the electrolyte's pH
The supporting electrolyte pH's influence on the electrocatalytic oxidation of pyrene was examined in a 0.4 mM pyrene solution over a 3-9 pH range using cyclic voltammetry at a 25 mV/s scan rate in acetonitrile/water (80:20) with 0.1 M LiClO 4 supporting electrolyte.This was done to achieve the best pH possible, which would favour the electrocatalytic oxidation of pyrene.As can be seen in figures 9(a) and (b), the oxidation current response was optimal at pH 8 suggesting better sensitivity, subsequently, further investigations were carried out at pH 8.The observed variations in the anodic peak potentials further suggests electron and proton transfer.

The electrocatalytic response of the electrodes to pyrene oxidation
The response of the unmodified and modified GCE to the electrocatalytic oxidation of 0.4 mM pyrene in acetonitrile/water (80:20) with 0.1 M LiClO 4 supporting electrolyte were examined using the cyclic voltammetry (CV) at 25 mV/s scan rate.Pyrene was oxidized on all electrodes as seen in the cyclic and square wave voltammograms depicted in figures 10(a) and (c) respectively.Both the cyclic and square wave voltammograms revealed two irreversible oxidation peaks for 0.4 mM pyrene on the GCE/fMWCNTs/CuO NPs, GCE/CuO NPs and the bare GCE but a single irreversible anodic peak on the GCE/fMWCNTs.The GCE/ fMWCNTs/CuO NPs displayed well resolved irreversible oxidation peaks in addition to presenting the highest current response and an anodic peak potential that was lower than the bare's (table 3), indicating exceptional electrocatalytic activity.The strong interaction between the highly conducting fMWCNTs and CuO nanoparticles and the large surface area they present, facilitated electron transport at the electrode, enhancing the oxidation of pyrene.The anodic CV current responses of the electrodes increased sequentially in the order: Bare GCE (32.50 μA, 46.30 μA) < GCE/CuO NPs (76.32 μA, 356.41 μA) < GCE/fMWCNTs (670.17 μA) < GCE/fMWCNTs/CuO NPs (465.73 μA, 997.62 μA) as reported in table 3.
The first stage of the electrochemical oxidation process of pyrene is predicted to be the formation of pyrene cation radicals from the direct transfer of one electron from the adsorbed PAH to the electrode, followed by the  formation of hydroxyquinones, dihydroxyquinones, and finally pyrene-1,6-dione and pyrene-1,8-dione [26,73].This is typical of aromatic hydrocarbons which are oxidized by losing an electron from the highest occupied p-molecular orbital, to produce p-delocalised radical cations.The electron transfer process is usually an irreversible reaction in acetonitrile, with the radical-anion reacting rapidly with extraneous nucleophiles and residual water at the carbon atoms that has the highest positive charge density.Another irreversible oxidation wave also occurs at a more positive potential as a result of the production of a transient dication [74].This accounts for the two irreversible oxidation peaks observed in the voltammograms.The first irreversible oxidation peak being that of the cation-radical species while the second is for the transient dication.
The detection of two oxidation peaks due to the formation of cation-radicals from the electrochemical oxidation of pyrene has been observed in a previous work [75].Also, the two isomers; 1, 6-pyrenedione and 1, 8-pyrenedione were reportedly formed during the electrochemical oxidation of pyrene in previous studies [76].A single irreversible oxidation peak was observed on the GCE/fMWCNTs which could be as a result of the compression of the potential difference between the two waves, merging to produce one voltammetric wave as Scheme 1. Proposed pyrene metabolism to 1-hydroxypyrene and its subsequent oxidation to pyrene-1, 6 -and 1, 8 -diones via 1, 6and 1, 8-dihydroxypyrenes [26,[78][79][80].
Table 3. Overview of the cyclic and square wave voltammetry data obtained from the investigation of unmodified and modified glassy carbon electrodes in 0.4 mM pyrene, where E p,aI is the potential of the first anodic peak current, I p,aI and E p,aII is the potential of the second anodic peak current, I p,aII .
CV Peak potentials (V) can be observed in the peak's broadness [77].A possible mechanism for the electro-oxidation of pyrene is outlined in scheme 1.
The electrodes' resistance to charge transfer (Rct) investigated by EIS have their Nyquist plots represented in figure 10(b).The electrical equivalent circuits modelled for the Nyquist plots fitting are inset as figure 10(c) for the bare GCE and GCE/CuO NPs and figure 10(d) for the GCE/fMWCNTs and GCE/fMWCNTs/CuO NPs with Rct values of 10.42 kΩ, 5.98 kΩ, 91.48 Ω and 46.60 Ω for the bare GCE, GCE/CuO NPs, GCE/fMWCNTs and GCE/fMWCNTs/CuO NPs respectively.The lowest Rct value with a huge reduction to the bare's was recorded at the GCE/fMWCNTs/CuO NPs.The Rct value of the GCE/fMWCNTs clearly reveals fMWCNTs' highly conductive nature, its integration with CuO nanoparticle equally shows an Rct value that is almost halved, signifying better interfacial electron exchange and enhanced electrocatalytic activity.From equations (3)-( 4), the exchange current density, o j , was determined to be: 0.95 mAcm , 2 and ) -1.17 mAcm 2 while the heterogeneous electron transfer rate constant, k , 0 was calculated as: ( ´--2.68 10 cm s , ´--9.50 10 cm s , ´--9.84 10 cm s , for the unmodified GCE, GCE/CuO NPs, GCE/fMWCNTs and GCE/fMWCNTs/CuO NPs respectively.The higher k 0 value obtained for the GCE/fMWCNTs/CuO NPs exceeds the value expected for irreversible processes signifying faster transfer of electrons.The values obtained positions GCE/fMWCNTs/CuO NPs as a better electrocatalyst than the other electrodes explored, enhancing the electro-oxidation of pyrene. 3.5.Effect of scan rate variation at the GCE/fMWCNTs/CuO NPs Scan rates of the reaction at GCE/fMWCNTs/CuO NPs, varied from 10 mV s −1 to 120 mV s −1 between a +0.60 to +1.6 V potential window in 0.4 mM pyrene in acetonitrile/water (80:20) mixture including lithium perchlorate (LiClO 4 ; 0.1 M) supporting electrolyte (pH 8) were recorded using cyclic voltammetry as shown in figure 11(a).Increasing anodic peak current responses, I paI and I p aII , with positive shifts in the peak potentials, E paI and E p aII , were observed corresponding to increasing scan rates.The oxidation peak occurring at E paI was more prominent at the lower scan rates as shown in figure 11(a), but gradually diminished in intensity as the oxidation peak at E , p aII , became increasingly prominent at the higher scanning rates confirming that rapid chemical reactions ensue the initial electron transfer.A plot of the anodic peak current responses, I paI and I p aII , versus the scan rate's square root, u (figure 11(b)) suggests a diffusion-controlled process with the regression equations ( 14) and (15) showing the linear relation between the current responses and the scan rate's square root:  16) and (17) were both found to be were found to be 0.23 and 0.29 Vdec −1 respectively, surpassing the anticipated standard (0.118 Vdec −1 ) value for a one-electron process.
These high Tafel values are possibly due to reactions taking place inside the porous electrode structure, and, or electrode surface adsorbed reactants or intermediates [65].The ( ) a a n 1 values were determined to be 0.518 and 0.409 from the slope in equations ( 16) and (17) using equation (11).From the irreversible oxidation wave-shape in equation (18) [63], ( a -1 ) was deduced as 0.465 and 0.22, a was subsequently evaluated as 0.54 and 0.78; where / E p 2 (mV) is the potential at which the current is halved relative to its peak value.The number of electrons transferred at GCE/fMWCNTs/CuO NPs in the electrochemical oxidation of pyrene in (80:20) mixture of acetonitrile and water with LiClO 4 (0.1 M) was calculated as 1.1 for E p aI , and 1.9 for E p aII , respectively.This confirms the initial one electron removal step predicted for the initial stage in the electrochemical oxidation of most aromatic hydrocarbons proposed in previous studies [81,82].Also, the electron transfer rate constant ( k s ), estimated from equation (19) [69,83], was determined to be The electrode's performance in comparison to previously published modified electrodes for pyrene detection are presented in table 4. Our findings proposes that GCE/fMWCNTs/CuO NPs is a suitable electrode for micromolar electrochemical detection of pyrene.

Interference study
Since polycyclic aromatic hydrocarbons typically appear in mixtures, the sensor's selectivity for pyrene detection was assessed in the presence of another polycyclic aromatic hydrocarbon, anthracene.The investigation performed using square wave voltammetry revealed good electrode sensitivity and selectivity regardless of the presence of anthracene with well-resolved peaks for anthracene at the potential (1.02 V) and and pyrene at the potential (1.16 V and 1.27 V), as depicted in figure 13(a).The anthracene concentration was kept constant (at 0.04 mM) while the pyrene concentration was increased sequentially at a 0.01 mM increment.The anodic current response to pyrene oxidation was observed increasing as the concentration of pyrene increased with a shift in poential to the less positive, while the anodic current was reducing till it became constant, suggesting that the electrode has the capacity for pyrene detection in the presence of other PAHs, especially anthracene (figure 13(b)).

Reproducibility
The reproducibility of the electrode in 0.4 mM pyrene in acetonitrile/water (80:20) with 0.1 mM lithium perchlorate (LiClO 4 ) at pH 8 was also examined using cyclic voltammetry.At a 25 mV/s scan rate, the relative standard deviation (RSD) for three different electrode modifications was 7.97% (graph not shown), demonstrating that the proposed sensor is reasonably reliable and reproducible.

Conclusion
The electrochemical behaviour of glassy carbon electrode (GCE), modified with copper oxide (CuO) nanoparticle and multi-walled carbon nanotubes (fMWCNTs) nanocomposite in pyrene has been described in this work.The remarkably increased current response and reduced charge transfer resistance of the modified electrode depicts synergistic effect of the metal oxide and the fMWCNTs, significantly enhancing the electrode's conductivity.The electro-oxidation of pyrene at the GCE/fMWCNTs/CuO NPs was achieved at a detection limit of 1.30 μM demonstrating the sensors' potential for environmental analytical applications.

3 for
the unmodified GCE, GCE/CuO NPs, GCE/ fMWCNTs, and GCE/fMWCNTs/CuO NPs respectively.The results obtained correspond to quasi-reversible behaviour, which are estimated to be in the range of ( )

Figure 9 .
Figure 9. (a) Cyclic voltammograms of 0.4 mM Pyrene at pH ranged 3-9 (25 mVs −1 scan rate) and (b) plot of the corresponding anodic peak current and anodic peak potential against the pH.

Figure 10 .
Figure 10.(a) Cyclic voltammogram at 25 mVs −1 scan rate, (b) Electrochemical impedance spectroscopy nyquist plot at 1.2 V Inset: EIS experimental data equivalent electrical circuit fit for: (c) bare GCE and GCE/CuO NPs (d) CE/fMWCNTs and GCE/fMWCNTs/ ZnO NPs and (e) Square wave voltammogram of the modified and unmodified glassy carbon electrodes in 0.40 mM pyrene.
studies Using the square wave voltammetry (SWV) technique, the effect of concentration on pyrene's electrochemical oxidation at the nanocomposite-modified electrode was examined.The GCE/fMWCNTs/CuO NPs' response to pyrene oxidation on gradually increasing its concentration from 1.2 μM to 23.1 μM in (80:20) mixture of acetonitrile and water with LiClO 4 (0.1 M) was recorded in figure12(a).With increasing pyrene concentration, an increase in peak current linearly related by the regression equation: (figure12(b)).The relationship between the gradient of the calibration curve (m) and the relative standard deviation of the intercept (d) in equation (20) was used to determine the limit of detection (LoD) of the electrode; determined as 1.30 μM with a sensitivity of 2.03 μA/μM.

Table 1 .
An overview of the FTIR absorption bands for CuO nanoparticles, fMWCNTs and fMWCNTs/ CuO NPs.
[52]sing electrochemical impedance spectroscopy and cyclic voltammetry at a 25 mV/s scan rate.The anodic (I p,a ) and cathodic (I p,c ) peak current responses resulting from redox processes in [Fe(CN) 6 ] 3-/4-at 25 mV/s scan rate are well-defined as shown in the cyclic voltammograms of the electrodes (figure7(a); table1).The GCE/CuO NPs' anodic peak current response at 54.11 μA was observed much lower than the response of the bare GCE at 110.32 μA possibly due to the thickness of the CuO nanoparticle film or a layer formation on the electrode's surface, obstructing the necessary electrochemical reaction active sites.A similar occurrence of a CuO nanoparticle modified GCE with reduced current response in comparison to the bare GCE's in the redox probe's presence has been previously reported[52].The nanocomposite modified electrode (GCE/fMWCNTs/CuO NPs) had the highest anodic peak current response (546.24μA) with a 395% increment than that of the bare GCE (110.32 μA).This remarkably improved response is attributable to the combined excellent electron transport facilitation, large specific [58][59][60][56][57]' electrochemical characterisationProbe of the electrochemical properties of the unmodified and modified glassy carbon electrodes (GCE/CuO NPs, GCE/fMWCNTs, and GCE/fMWCNTs/CuO NPs) were conducted in 10 mM [Fe(CN) 6 ] 3-/4-(pH 7) prepared with KCl (0.1 surface area, great chemical stability and high electrical conductivity of carbon nanotubes[53][54][55][56][57]with the high specific surface area, good electrochemical properties, chemical stability and high electroncommunication features of copper-oxide nanoparticles[58][59][60]; promoting [Fe(CN) 6 ] 3-/4-accumulation on the electrode's surface.A reduced peak-to-peak potential separation, ΔEp, of the nanocomposite modified electrode (table 2) in contrast to the bare GCE's, also signifies that the barrier to electron transfer has been lowered, suggesting faster electron transport.The GCE/fMWCNTs/CuO NPs cyclic voltammogram (figure7(a)) reveals a pair of anodic and cathodic peak between +0.30V and +0.19 V due to the redox