Phthalocyanine Doped Metal Oxide Nanoparticles on Multiwalled Carbon Nanotubes Platform for the detection of Dopamine

The electrocatalytic properties of metal oxides (MO = Fe3O4, ZnO) nanoparticles doped phthalocyanine (Pc) and functionalized MWCNTs, decorated on glassy carbon electrode (GCE) was investigated. Successful synthesis of the metal oxide nanoparticles and the MO/Pc/MWCNT composite were confirmed using UV-Vis, EDX, XRD and TEM techniques. Successful modification of GCE with the MO and their composite was also confirmed using cyclic voltammetry (CV) technique. GCE-MWCNT/ZnO/29H,31H-Pc was the best electrode towards DA detection with very low detection limit (0.75 μM) which compared favourably with literature, good sensitivity (1.45 μA/μM), resistance to electrode fouling, and excellent ability to detect DA without interference from AA signal. Electrocatalytic oxidation of DA on GCE-MWCNT/ZnO/29H,31H-Pc electrode was diffusion controlled but characterized with some adsorption of electro-oxidation reaction intermediates products. The fabricated sensors are easy to prepare, cost effective and can be applied for real sample analysis of dopamine in drug composition. The good electrocatalytic properties of 29H,31H-Pc and 2,3-Nc were related to their (quantum chemically derived) frontier molecular orbital energies and global electronegativities. The better performance of 29H,31H-Pc than 2,3-Nc in aiding electrochemical oxidation of DA might be due to its better electron accepting ability, which is inferred from its lower ELUMO and higher χ.

other reagents are of analytical grade and obtained from Sigma-Aldrich, Merck chemicals and LABCHEM respectively. Ultra-pure water of resistivity 18.2 MΩ was obtained from a Milli-Q Water System (Millipore Corp., Bedford, MA, USA) and was used throughout for the preparation of solutions. A phosphate buffer solution (PBS) of 7.0 was prepared with appropriate amounts of NaH 2 PO 4 · 2H 2 O, Na 2 HPO 4 · 2H 2 O, and H 3 PO 4 , and adjusted with 0.1 M H 3 PO 4 or NaOH. Prepared solutions were purged with pure nitrogen to eliminate oxygen and prevent any form of external oxidation during every electrochemical experiment.

Synthesis of Zinc Oxide (ZnO) Nanoparticles. Sodium hydroxide (NaOH) was dissolved in deionised
water to a concentration of 1.0 M and the resulting solution was heated, under constant stirring, to the temperature of 70 °C. After achieving this temperature, a solution of 0.5 M Zn(NO 3 ) 2 · 6H 2 O was added slowly (drop wise for 60 minutes) into the NaOH aqueous solution under continuous stirring. In this procedure the reaction temperature was constantly maintained at 70 °C. The suspension formed with the dropping of 0.5 M Zn(NO 3 ) 2 6H 2 O solution to the alkaline aqueous solution was kept stirred for two hours at 70 °C. The material formed was filtered and washed several times with deionised water. The washed sample was dried at 65 °C in the oven for 24 hours to obtain the dry powder 32 .  (50.8 mL), diluted to a total volume of 800 mL was used. Fe 3+ and SO 3 2− were mixed; the colour of the solution changed from light yellow to red, indicating formation of a complex ion. The solution was quickly poured into the diluted ammonia solution under vigorous stirring, and the colour changed from red to yellow again. A black precipitate was formed and stirring continued for 30 minutes. After the reaction, the beaker containing the suspension was placed on a permanent magnet. The supernatant liquid was decanted and fresh deionised water was added into the beaker. This procedure was repeated several times until most of the ions in the suspensions were removed. The dry powder was obtained by filtration and drying at room temperature 33 .

Equipment and Procedure. The Transmission electron microscopy (TEM) analysis was performed using
Tecnai G2 Spirit FEI (USA). UV-vis analyses were carried out using UV-visible spectrophotometer (Agilent Technology, Cary series UV-vis spectrometer, USA). Electrochemical experiments were carried out using an Autolab Potentiostat PGSTAT 302 (Eco Chemie, Utrecht, and The Netherlands) driven by the GPES software version 4.9. Electrochemical impedance spectroscopy (EIS) measurements were performed with Autolab Frequency Response Analyser (FRA) software between 10 kHz and 1 Hz using a 5 mV rms sinusoidal modulation with the solution of the analyte at their respective peak potential of oxidation (vs. Ag|AgCl in sat' d KCl). A Ag|AgCl in saturated KCl and platinum wire were used as reference and counter electrodes respectively. A bench top Crison pH meter, Basic 20+ model, was used for pH measurements. All experiments were performed at 25 ± 1 °C while the solutions were de-aerated before every electrochemical experiment.
Electrode Modification Procedure. Electrode modification was carried out using the drop-dry method.
The glassy carbon electrode was cleaned by gentle polishing in aqueous slurry of alumina nanopowder on a silicon carbide-emery paper followed by a mirror finish on a Buehler felt pad. The electrode was further sonicated in double distilled water, and then absolute ethanol for 2 min to remove residual alumina particles that were trapped on the surface and dried at room temperature. Separate suspensions of metal oxide (MO) nanoparticles, phthalocyanines (Pc), MWCNT and MWCNT/MO/Pc hybrids were prepared in 1 mL of DMF and sonicated as described above. 10 μ L drops of the prepared suspensions were dropped on the bare GCE and dried in an oven at 50 °C Preparation of the Dopamine Hydrochloride Injection Solution. 2 mL of the drug (dopamine hydrochloride injection-Dopamine HCl-Fresenius ® ) sample was diluted to 100 mL with distilled de-ionized water. 2 mL of the diluted solution was pipette into six 50 mL volumetric flask and all except one were spiked with different concentration of standard dopamine solution, and made to volume with phosphate buffer pH 7.0. The concentration of each test aliquot solution was determined using square wave voltammetry. Two different injections from the same batch were analysed using the same procedure. The experiment was repeated 3 times for each sample.

Results and Discussion
UV-Vis, EDX, XRD and TEM Characterization. UV-Vis Results. Figure 1 shows the comparative UV-Vis absorption spectra for the metal oxide (MO) nanoparticles/phthalocyanines hybrids in DMF solution, confirming the formation of Fe 3 O 4 /2,3-Nc, Fe 3 O 4 /29H,31H-Pc, ZnO/2,3-Nc and ZnO/29H,31H-Pc composites. Phthalocyanines are organic semiconductors with important electrical properties applicable to sensor system. The UV-Vis spectrum of phthalocyanines originates from molecular orbitals within the aromatic 18π electron system 34 and they have been considered as electrophotographic materials due to their absorption ability in the ultraviolet and visible region 35 . The evidence for the successful doping of 29H,31H-Pc and 2,3-Nc with Fe 3 O 4 and ZnO is demonstrated by the change in the absorption bands as shown in the Figure. For example, Fe 3 O 4 and ZnO nanoparticles showed absorption peaks at 397 nm and 370 nm (B band) respectively in the region of the visible spectrum which are in the range of 300 to 400 nm for transition metal oxide as reported in literature 36,37 . However after the phthalocyanine were doped with the nanoparticles there was a blue shift (386 nm) in absorption peak confirming formation of Fe 3 O 4 /29H,31H-Pc composite.
Appearance of a new peak at 665 nm and a red shift in absorbance peak at 376 nm (B-band) confirmed the formation of ZnO/29H,31H-Pc composite. Fe 3 O 4 /2,3-Nc and ZnO/2,3-Nc showed absorptions peaks at 369 nm and 362 nm respectively, which are lower compared with 397 nm and 370 nm for Fe 3 O 4 and ZnO nanoparticles alone further confirming the successful formation of Fe 3 O 4 /2,3-Nc and ZnO/2,3-Nc composite. Broad absorption band for these two composites were also observed at 725 nm and 730 nm (Q-band) respectively. These characterized peaks of the Q-band have generally been interpreted in terms of the π -π * transitions of the π -electron on the phthalocyanine macrocycle 38,39 . Since phthalocyanines are found to be relatively good electron donors, Spadavecchia 40 mentioned that the dispersion forces between adjacent molecules of phthalocyanine are produced by the delocalized π -electron system constructing a highly polarizable electronic cloud. As a result, low ionization energy favours the charge transfer interactions when electron acceptor molecules are adsorbed, hence the difference in the absorption spectra come from an increase of the density in the electronic levels available for π -π * transition. These electron cloud and charge transfer behaviour of Pc could also be responsible for the successful formation of the different MO/Pc composite synthesized in this study probably due to ionic and intermolecular interaction between the MO nanoparticles and the Pc molecules.
EDX and XRD Results. The EDX spectra shown in Fig. 2 indicate the elemental composition of hybrids formed when phthalocyanines doped with metal oxides (Fe 3 O 4 , ZnO) were supported on MWCNTs. The presence of MWCNTs is confirmed by a carbon and sulfur, since the MWCNTs were purified and functionalized by acid mixture. The appearance of sulfur is due to sulfuric acid used during functionalization of MWCNTs. The presence of phthalocyanines is confirmed by a small peak of nitrogen while the zinc, oxygen and iron peaks further confirms the successful doping of Fe 3 O 4 and ZnO. These results illustrates that MWCNTs was successfully modified with metal oxides doped phthalocyanine.  [41][42][43] . The average crystallite size obtained was 23.8 nm which was comparable to that obtained from TEM analysis. The diffraction peak at 2θ of 35.6° corresponds to the spinel phase of the Fe 3 O 4 nanoparticles and the broadening of the diffracting bands indicates small particle size and ultrafine nanoparticles 37,43 . These data was in agreement with the XRD patterns of Fe 3 O 4 nanoparticles reported in literature, therefore it confirms that the synthesized nanoparticles were pure Fe 3 O 4 nanoparticles with spinel structure 41,44 .  According to literature the peaks for (100), (002) and (101) indices illustrates hexagonal ZnO of wurtzite structure 46,48,49 . The average crystallite size obtained was 14.2 nm which was comparable to that obtained from TEM analysis. Figure 3(c) shows the XRD pattern of functionalised and pristine MWCNTs. The three diffraction peaks observed at pristine MWCNTs also appeared at functionalized MWCNTs with enhanced intensity, indicating that MWCNTs were successfully functionalized with carboxyl groups, and also the MWCNTs structure was not destroyed. The diffraction peaks occurred at 2θ of 26.12°, 43.62° and 52.49° which corresponds to indices of (002), (100) and (101) respectively as reported in literature 13,50 . The broad peak (002) in the XRD pattern indicates the graphitic structure of CNTs 51 which is the ordered arrangement of the concentric cylinders of graphitic carbon 52 . Therefore this confirmed the hexagonal structure of MWCNTs 50 . The TEM images showed that Fe 3 O 4 NPs were similar in shape, approximately spherical and they appeared to be almost uniformly distributed and less aggregated with the average diameter of 12 nm. The particle size was found to range from 6-23.7 nm. ZnO NPs were shaped like small rods and were aggregated with an average diameter of 26.1 nm. The aggregation might be due to the large surface area and high surface energy of ZnO NPs, which probably took place during the drying process 53 . The length size of the nanorods ranged from 20-78.6 nm. 2,3-Nc (c) and 29H,31H-Pc (d) images showed an irregularly shaped, small rod-like structures, or clustered flakes, none-uniform in both size and shape with a rough surface of 2,3-Nc (c) and a smooth surface of 29H,31H-Pc (d). The TEM image of MWCNTs (e) showed a tangled-net like structure of long tubes with a rather smooth surface. The presence of dark round spots show the cap ends of the tubes that were cut opened by acid treatment; they result from the oxidation of C-O, O-H and C-O groups to form COOH groups on the end caps of the CNTs. The MWCNTs showed a diameter of 8.95 nm.
The TEM images of MWCNT/ZnO/2,3-Nc (Fig. 5a) and MWCNT/ZnO/29H,31H-Pc ( Fig. 5c) showed that ZnO nanoparticles and phthalocyanines combined into larger aggregates that were scattered and poorly dispersed on the MWCNTs for MWCNT/ZnO/2,3-Nc surface (Fig. 5a), and smaller particles were found distributed along the tubes for MWCNT/ZnO/29H,31H-Pc (  NPs alone showed a drastic decrease in current response, this could be due to the formation of passive oxide layers of ZnO and Fe 3 O 4 NPs, which were obstructive to the electron transfer activity 54 . Pairs of well-defined redox peaks were observed at I (0.2858 V), I I (0.2174 V) and I (0.2834 V), I I (0.2199 V) for Fig. 6a,c respectively. Similar pairs of redox peaks were observed in about the same redox potentials in Fig. 7(a,c). These redox peaks represent the [Fe(CN) 6 ] 4− /[Fe(CN) 6 ] 3− redox process. Two additional peaks were discovered at II (0.8009 V) and III (0.05386 V) (Fig. 6c), these peaks might be associated with the oxidation of the Fe (II) to Fe (III) in Fe 3 O 4 and reduction of Fe (III) to Fe (0). Similarly, two additional anodic peaks were detected on Fig. 7(a,c) at II (0.8864 V) and II (0.8766 V) respectively, which might be due to the electrochemical oxidation Zn (II) to Zn (III). These two peaks showed a high oxidation current that almost suppressed the anodic oxidation peak of [Fe(CN) 6   (0.8009 V) (Fig. 6d). The peaks at around 0.26 and 0.3 correspond to the oxidation of Fe (II) to Fe (III), while the one around 0.8 V can be attributed to the oxidation of Fe (III) to FeOOH. Similar oxidation peaks were observed around 0.8-0.9 V (Fig. 7b,d) for ZnO oxidation. Zang 55 mentioned that the electrochemical redox reaction of regular structured ZnO in batteries is not reversible and suggested that the reversibility of the redox reaction observed in PBS electrolyte might be attributed to the unique nanostructure of ZnO NPs fabricated. It is evident from the results in both electrolytes that there is combining cooperative effect between MO, Pc/Nc, and MWCNT that leads to enhance electron transport and high current response of the nanocomposite modified electrodes. This can mainly be due to the great surface area of MWCNT and Pc, and electrical conductivity of the MO and MWCNT NPs.
Electrochemical Impedance Spectroscopy Studies of Modified Electrodes. The electrochemical impedance spectroscopy is a useful technique to investigate the electrode/solution interface properties. Figure 8 shows the Nyquist plots of the bare and modified electrodes in [Fe(CN) 6 ] 4− /[Fe(CN) 6 ] 3− redox probe at a fixed potential of 0.22 V (vs Ag/AgCl, Sat' d KCl). The EIS data was recorded in the frequency range between 1 Hz to 10 kHz and the impedance parameters obtained are summarized in Table 1. Figure 8(e) represents the circuit used in the EIS fitting and it consists of solution resistance (R s ), constant phase element (CPE) which relates to the porous nature of the electrode, charge transfer resistance (R ct ) and double layer capacitance (C dl ). The chi-square values, which are used to judge the goodness of EIS fittings and appropriateness of the adopted equivalent circuit and the percentage errors (in parentheses) associated with each circuit parameter are also listed in Table 1. The small chi-square values and percentage errors (Table 1) indicate that the experimental data are well fitted by the equivalent circuit. From the results obtained GCE-Fe 3 O 4 and GCE-ZnO in Fig. 8 showed a semicircle domain. According to literature such behaviour demonstrates a charge transfer limited process 52 and the diameter of the semicircle is equal to the charge transfer resistance 56 . Referring to impedance data in Table 1  and naphthalocyanine which accounts for the different rates of charge transfer and the other factor might be attributed to the axial ligands and substituents on the ring 59 . Therefore the oxidation of metal phthalocyanines may occur both at the metal and at the ring depending on the energies and proximity of the metal d and phthalocyanine ring orbitals 60 . These results are in good agreement with results obtained during CV studies.

Electrocatalytic and Electroanalysis Studies. Electrocatalytic Oxidation of Dopamine (DA). A series
of CV experiments were conducted to investigate the electrocatalytic oxidation of DA in pH 7.2 PBS solution containing 0.085 mM of DA using the bare GCE and modified electrodes. Figure 9 presents the oxidation potentials and the comparative current response of the GCE electrode modified with Fe 3 O 4 nanoparticles in the analyte. Except for Fig. 9(a), other figures present the voltammograms for the DA oxidation process after the background current subtraction. The voltammetric parameters such as I pa and E pa are summarized in Table 2. From the results obtained, broad ill-defined oxidation peaks of DA were observed for bare GCE, GCE-ZnO, and GCE-29H,31H-Pc. It is also interesting to note that the GCE-Fe 3 O 4 , GCE-ZnO and GCE-29H,31H-Pc gave low current response compared with the bare GCE. This may suggest that the Fe 3 O 4 and ZnO nanoparticles film acted as passive layer obstructing the flow of current due to DA oxidation. However, an enhanced DA oxidation current response at lower oxidation potential was observed at the GCE-MWCNT/Fe 3 O 4 /2,3-Nc, and GCE-MWCNT/Fe 3 O 4 /29H,31H-Pc electrode (Fig. 9, Table 2). Similar trend was obtained for the ZnO nanoparticles GCE modified electrode where GCE-MWCNT/ZnO/2,3-Nc and GCE-MWCNT/ZnO/29H,31H-Pc electrodes gave the highest DA current at lower DA oxidation potential compared with the bare and GCE-MWCNT electrodes (Fig. 9, Table 2). The current was found to be 10 times higher for GCE-MWCNT/  electrodes stability during repeated cycles. These results suggest that there is some level of adsorption of DA at electrodes surface 65 . The adsorptive nature of electrodes might be due to the porous structured film of the CNTs on the electrodes 65 . The relative standard deviation (R.S.D.) of the fabricated sensors response to DA was 7.32%, 5.9%, 3.4% and 6.2% for MWCNT/Fe 3 O 4 /2,3-Nc, MWCNT/Fe 3 O 4 /29H,31H-Pc, MWCNT/ZnO/2,3-Nc and MWCNT/ZnO/29H,31H-Pc modified electrodes respectively, indicating that these electrodes are relatively stable and are not subjected to obvious surface fouling towards DA determination. The second reduction peak (II) was observed in Fig. 10(d), this peak might be attributed to the reduction of ZnOOH to Zn(OH) 2 and ZnO respectively. GCE-MWCNT/ZnO/2,3-Nc showed a broad oxidation peak of DA as compared to other electrodes, however it gave the lowest RSD value indicating that it was the better performing and most stable electrode. The electrode can be used for the analysis of DA after storage in a refrigerator for up to two weeks without a significant change in its response.
The effect of scan rate (ν). Cyclic voltammetry was used to study the scan rate The plots of peak current vs square root of scan rate showed a linear relationship. The oxidation peak currents were simultaneously increasing with increasing scan rate suggesting diffusion controlled process. Similarly, the anodic peak current (I pa ) was directly proportional to the square root of scan rate (ν 1/2 ) (Figs 11 and 12) but with negative intercept suggesting possibility of some adsorption intermediate in the process 65 . GCE-MWCNT/ZnO/2,3-Nc (Fig. 12a), GCE-MWCNT/ZnO/29H,31H-Pc (Fig. 12c) detect the presence of DA up to 300 mV s −1 , from 350-1000 mV s −1 the oxidation peak current of DA became invisible. This might be due to electrode fouling effect. It could also suggest that the catalyst film on the electrode surface gets saturated and hence cannot detect the analyte over multiple cycles. Similar phenomenon have been observed by other workers and was attributed to factors such as method of electrode preparation, electrode conformation, the catalyst-analyte interaction, and porosity which play a significant role in the adsorption of the analyte on the catalyst film 65 .
The surface coverage (Γ ) of DA on the surface of the modified electrode was calculated from the plot of peak current, I p versus the square root of scan rate using Laviron equation 66 : (1)    68 . In addition the data obtained from plots of peak potential (E p ) versus the log of scan rate (log ν ) showed a shift or an increase of peak potential as the log of scan rate increases. This variation of peak potential indicated that DA was oxidized by means of an adsorption process and the process was irreversible 69 .
Concentration study. It is known that oxidation current depends on the concentration of the analyte, therefore because of the advantage of its good sensitivity, DPV was used to study the effect  71 mentioned that sensitivity has often been mistakenly referred to as the ability to achieve low limits of detection, meaning that the more sensitive an electrode is, the low limit of detection can be achieved. He defined sensitivity as the change in signal versus the change in concentration of the analyte, or it can be defined as the slope of the calibration curve in an analysis. The sensitivity of a sensor depends widely on the reactive surface area of a sensor and the catalytic material used to modify an electrode 72 . In fabrication of a sensor, a limit of detection is more important to investigate the performance of a sensor. A sensor with a high sensitivity won't necessarily equate to a low limit of detection, meaning a highly sensitive sensor may have a high background noise level which might not be a problem at higher concentrations, but at lower concentrations excessive noise can hinder the good measurements of detection limits 72 . Therefore other than increasing sensitivity, a low detection limit can be achieved by reducing matrix interference (increasing selectivity) and high background noise 71 .   Fig. 15, two well separated oxidation peaks of AA and DA were observed for all the four electrodes. It was observed that up to 30-fold excess of AA there was no significant interference between AA and DA peaks separations. Oxidation peak potentials and the difference between the two peak potentials are summarized in Table 3. The difference between oxidation peak potentials for all the four electrodes was high enough to   DA is positively charged 61,62,64 , therefore the substantial shift in peak potentials could be related to the electrostatic repulsion between negatively charged AA, the electron reach oxygen atom of the MO nanoparticles and the negatively charged COO − functionalized MWCNT. It can be seen that the anodic peak signals of AA and DA were independent from each other; therefore the modified electrodes were able to adequately identify the two analytes. The height and amplitude of the peak corresponding to DA also increase proportionally with the DA concentration. In fact, in all the concentrations of the DA studied there was no detectable interference of the AA, the signal of the AA was about 200 mV away from that of the DA. However MWCNT/Fe 3 O 4 /2,3-Nc and MWCNT/ZnO/2,3-Nc modified electrodes showed better results in terms of peak separation. This might be due to the extension of π -electron systems of Ncs which facilitates its improved repulsion of the interfering species as compared to Pcs analogues 73,74 .  Table 4.
Relating electrocatalytic properties of modified electrodes with quantum chemical descriptors of 29H,31H-Pc and 2,3-Nc. Isaacs et al. 75 related electrocatalytic and redox activities of Co-Pc and Co-Nc to experimentally observed electronic spectra of the complexes. In their work, the authors reported nearly the same values of lowest unoccupied molecular orbital energy (E LUMO ) for the two complexes, higher value of highest occupied molecular orbital energy (E HOMO ) and lower value of HOMO-LUMO energy gap (∆ E L-H ) for Co-Nc compared to Co-Pc. Similar results have also been reported earlier based on experimental and theoretical semiempirical calculations 76 . These and our previous knowledge of frontier molecular orbital (FMO) energies and electron density distributions of Pcs and Ncs motivated us to explore possible correlations between theoretical quantum chemical parameters of 29H,31H-Pc and 2,3-Nc and the performance of our fabricated electrodes. Even though the composite electrode systems of the form GCE-MWCNT/metal oxide/macrocycle used in the experimental sections could not be modeled in the present work, attempt was made to correlate the observed good electrocatalytic behavior of the electrodes with molecular structure and FMO indices of the utilized macrocycles (29H,31H-Pc and 2,3-Nc). Figure 17 shows the optimized molecular structures, HOMO and LUMO electron density isosurfaces of 29H,31H-Pc and 2,3-Nc. The apparent planar geometries, widely delocalized HOMO and LUMO electron densities, and presence of σ and π orbitals in the FMOs are favourable to electron-transport processes between these macrocycles, the metal oxides (Fe 3 O 4 and ZnO) used in the composite electrode systems and the analyte (DA). Computational details and extensive explanations of other quantum chemical parameters on 29H,31H-Pc and 2,3-Nc including Mulliken atomic charges are contained in our previous work 77 . Table 5 shows selected quantum chemical descriptors of 29H,31H-Pc and 2,3-Nc obtained using the B3LYP/6-31G(d,p) model chemistry and reported elsewhere 77 . The results revealed high values of E HOMO and low values of E LUMO for the compounds. Generally, a high value of E HOMO connotes high tendency of a molecule to donate its HOMO electron to a suitable accepting orbital while a low value of E LUMO implies good propensity of a molecule to accept electron into its LUMO from a suitable donor orbital, and vice versa. The results in Table 5 support better electron transport properties of 29H,31H-Pc and 2,3-Nc and their good electrocatalytic behaviours. It was also observed that 29H,31H-Pc is more disposed to electron acceptance than 2,3-Nc based its lower value of E LUMO and higher value of electronegativity (χ ).
If it is assumed that oxidation of DA to dopamine-o-quinone occurred according to the equation 78 : then, 29H,31H-Pc with better electron accepting ability (lower E LUMO and higher χ ) may show better electrocatalytic property than 2,3-Nc towards DA oxidation. In other words, an electrode system containing a better electron accepting specie as a catalyst may show better detection of DA. This assumption is in agreement with our observed (experimental) LOD values in which electrode systems containing 29H,31H-Pc exhibited lower LODs than those containing 2,3-Nc.

Conclusions
The  response is associated with the excellent strong super paramagnetic properties of Fe 3 O 4 and high electrical conductivity of MWCNTs, ZnO and phthalocyanines. All the four electrodes showed a good stability with low current drop (< 10%) towards DA oxidation. GCE-MWCNT/ZnO/29H,31H-Pc was the best electrodes towards DA detection with very low detection limit that compared with literature, good sensitivity, resistance to electrode fouling, and excellent ability to detect DA without interference from AA signal. Electrocatalytic oxidation of DA  on GCE-MWCNT/ZnO/29H,31H-Pc electrode was diffusion controlled but characterized with some adsorption of electro-oxidation reaction intermediates products. The MWCNT/ZnO/29H,31H-Pc and MWCNT/ Fe 3 O 4 /29H,31H-Pc GC modified electrodes has also proven to be a potential sensor for dopamine detection in real sample analysis. The relatively high-lying highest occupied molecular orbitals and low-lying lowest unoccupied molecular orbitals of 29H,31H-Pc and 2,3-Nc may be partly responsible for their great electrocatalytic activities. The better sensing properties of 29H,31H-Pc containing modified electrodes towards DA might be connected with the better electron-accepting ability of 29H,31H-Pc (compared to 2,3-Nc).  Table 5. Quantum chemical parameters for 29H,31H-Pc and 2,3-Nc 77 .