MWCNT Modified Composite Pencil Graphite Electrodes Fabricated by Direct Dripping and Electrophoretic Deposition Methods: A Comparison Study

The CPG leads were 6H, 4H, 2H, HB, 2B, 4B, 6B and 8B of Mars Lumograph 100, Staedtler, Germany. The MWCNT (8 to15 nm OD, 3 to 5 nm ID, 50 μm length, purity >95%) was purchased from Chengdu Organic Chemicals Co. Ltd., Chinese Academy of Sciences, China. Araldite^@ epoxy resin was from Ciba-Geigy, Switzerland. Organic solvents were obtained from various sources. Scopoletin was from Sigma, USA. Water was ultra-pure (18.2 MΩ cm) from Direct – Q3 of Millipore, USA.

Scopoletin was used as standard to determine total coumarin in dry fruits of A. sessiliflorus. For this purpose, stock solutions of scopoletin were prepared in ethanol and stored at −20°C. A small amount was taken and left to equilibrate at room temperature before analysis. The stock solution was diluted with 0.1 M citric-Na₂HPO₄ buffer (pH 6.0) prior to use.

The dry fruits of A. sessiliflorus were pulverized in a disintegrator and then sieved through 100 meshes (powder: 44.42%). The powder was then kept and sealed at 4°C in brown vials. A 10 g of the powder was soaked in 100 mL ethanol for 24 h in the dark to extract coumarin. The extract obtained was then filtered and later subjected for the experiment.

Electrochemical measurements were carried out on potentiostat/galvanostat Model 273A complete with Power Suite program (EG&G, Princeton Applied Research, USA). A conventional three-electrode electrochemical cell comprising a platinum wire as auxiliary electrode, a c-MWCNT/CPGE as working electrode and an Ag/AgCl (sat. NaCl) as reference electrode, was used. Electrochemical impedance spectroscopy (EIS) measurements were carried out using frequency response analyser FRD 100 (EG&G, Princeton Applied Research, USA) at frequency range of 100 kHz to 100 mHz and AC voltage amplitude 5 mV. The result of EIS was analyzed using the ZSimpWin 3.22 software. The morphologies of the electrodes were characterized by LEO SUPRA 55VD ultra-high resolution analytical field emission scanning electron microscopy (FESEM) of ZEISS, Germany. The UV-Visible spectrophotometer Lambda 35 (Perkin Elmer, USA) with 1 cm quartz cell was used to determine scopoletin in the comparative study. The XRD profiles of pristine MWCNT and c-MWCNT were obtained at high-resolution using X-ray diffractometer model PANalytical (X'pert Pro, Netherlands) equipped with monochromatic high-intensity Cu Kα radiation (λ = 1.54056 Å, 40 kV, 30 mA) filter in the 2θ range 10–90°.

The pristine MWCNT was purified by sonication in a 6 mol L⁻¹ HCl for 4 hours. Later, 50 mg of purified MWCNT was dispersed in a 40 mL mixture of H₂SO₄/HNO₃ (v: v = 3:1) prior to further sonication in a water bath for 7 hours at 40°C. After cooling at ambient temperature, the MWCNT was then centrifuged and washed to neutral. This pretreatment was subsequently repeated over again until all acids were removed. The final solid was then dried to a constant weight and characterized by FESEM and XRD. For a particular reason, DM was prepared using 1.0 mg mL⁻¹ dry c-MWCNT in DMF whilst EPD was prepared a 0.5 mg mL⁻¹ dry c-MWCNT dispersed in MeCN. Both suspensions were ready to use prior to 20 minutes sonication.

The pencil lead was obtained by completely removing the outer wooden sheath of the pencil using a sharp knife. It was then cut at 6.0 cm length before use as CPGE. The CPGE was cleaned in an ultrasonic bath first in acetone and then in water for a period of 10 minutes each. A tip of disposable polypropylene pipette was cut to a suitable size to fit the CPGE and then sealed with epoxy resin. The surface of CPGE disk was polished on a polishing paper until a mirror like surface appeared. The CPGE disk was later polished in 0.05 μm alumina slurry before rinsing thoroughly with water and finally, sonicated in ethanol and water for 5 minutes each. For conventional method, the disk CPGE was first dried in a desiccator prior to modification with c-MWCNT. The DM experiment was carried out by dripping a 2 μL of c–MWCNT/DMF solution (1.0 mg mL⁻¹) onto the surface of CPGE and dried at room temperature. For each series of EPD experiments, a CPGE as anode and a platinum as cathode, were used. Then c-MWCNT was deposited onto the CPGE disk by applying a constant voltage 25 V, with deposition time 1 min, and 2 cm distance between electrodes. After EPD, electrodes were dried upright in a desiccator for 24 h prior to be used.

The cyclic voltammetry of 5 mmol L⁻¹ K₄[Fe(CN)₆] in 1.0 mol L⁻¹ KCl on the modified electrodes were carried out at potential range of −0.2 V to 0.8 V. The EIS measurement was immediately performed after cyclic voltammetry. Amplitude of 5 mV rms sinusoidal signal was applied. The modulus and phase were plotted over a frequency range of 100 kHz to 100 mHz. The FESEM with an acceleration voltage of 15 kV was used to characterize the surface morphologies of the disks.

The CPGE at different hardness as substrate electrodes have been reported in our previous work.²⁹ The results show that as a bare electrode, the soft B group CPGE provides higher current density, which is essential for quantitative analysis. While the H group CPGE with less graphite and smaller diameter provides the better reversibility, which is essential for qualitative analysis. Among all, the 2H CPGE in the rigid region and 6B CPGE in the soft region are better electrodes compared with others. Thus, they were selected as the better substrate electrodes for further study.

According to the good crystallinity, oxygen-containing surface functional groups and high mesoporous 3D structure, c-MWCNT has been chosen as modifier of various kinds of electrodes.^9–29 The XRD patterns of pristine MWCNT and c-MWCNT are illustrated in Fig. 1. Both pristine MWCNT and c-MWCNT show peaks at 25.2 {002} and 43.5 {100}, which indicates the graphite lamellar structure of CNTs is not changed by oxidation. But the increasing peak shows the increasing of crystallinity after modification. This is also revealed by the FESEM characterization of c-MWCNT with more purity, better array and larger surface area. Thus, coating of c-MWCNT onto CPGE has effectively improved the electron transfer rate and reversibility by conventional DM.²⁹ However, this procedure raises its disadvantages, such as non-uniform CNTs film and uncontrollable film thickness. Thus EPD is proposed to improve these drawbacks.

Zoom In Zoom Out Reset image size

The EPD method is favorable to be used as it is friendlier to a variety of substrates and applications. This process is suitable for flat, cylindrical or even complicated shapes. Furthermore, the thickness and morphology of a deposited film can be easily controlled through adjustment of the deposition time and applied potential. In a typical EPD process, the negative charged acid-treated CNTs^40,41 are attracted and then deposited onto the anode (CPGE) by a DC electric field. The electrophoretic motion of particles during EPD results in the accumulation of particles and the formation of a homogeneous and rigid deposit at the relevant electrode. Compared with many colloidal processes, suspensions with relatively small particle sizes are better dispersed and preferred in this process.

In this study, parameters such as deposition time, deposition voltage and solvent effect have been selected to control the film thickness and surface morphology. The charge transfer resistance (R_ct) is obtained by fitting the impedance of the electrodes with the proposed modified Randles' equivalent circuit (Fig. 2A). Results present that even though the bare 2H (3118 Ωcm²) and 6B (1011 Ωcm²) show higher R_ct (Table I), the R_ct is much decreased by EPD of c-MWCNT. It seems that from 10 s to 10 min, R_ct reaches to a minimal value (2H 9.54 Ωcm² and 6B 5.64 Ωcm²) at 1 min, and then stable until the duration time reaches 3 min. A longer period of coating will increase R_ct and worsened adhering. At constant voltage from 2.5 V to 25 V, R_ct decreases constantly and the minimum value is shown at 20 V and become stable afterward (Fig. 2B (a) and insert). The Bode Plot (Fig. 2B (b)) also shows that the electrode starts to increase relaxation from 2.5 V (−65.5° at 1.08 Hz) and exhibits larger relaxation (−80.77° at 0.1 Hz) at 20 V, then the value becomes stable and changed slightly at 25 V (−80.79° at 0.1 Hz) with more refined and narrower distribution. This result indicates that a more homogeneous coating of c-MWNCT has been produced at higher voltages and deposition voltage of 20 V to 25 V is viable to be applied. Moreover, the effect of EPD on bare CPGE at high potential (25 V) has been carried out in blank solution. Result in Fig. 2B also shows that the EPD does not adversely affect the surface and electrocatalysis properties of CPGE. This is due to DC electric field only and very small current in the EPD process. Moreover, many dispersing solvents are also under consideration with R_ct of c-MWCNT/2H in the order of acetone/ethanol (v:v = 1:2) (28.7 Ωcm²) > THF (17.77 Ωcm²) > DMF (16.7 Ωcm²) > H₂O (14.12 Ωcm²) > DMSO (14.06 Ωcm²) > MeCN (9.54 Ωcm²). Influence of R_ct of c-MWCNT/6B with c-MWCNT dispersed in many solvents is also demonstrated the similar result as c-MWCNT/2H. Therefore, the EPD of a compact CNTs film can be obtained at a constant voltage of 20–25 V and deposition time of 1–3 min in MeCN.

Zoom In Zoom Out Reset image size

The features of surface-modified electrodes (bare, DM and EPD modified) with different hardness were characterized by EIS using 5 mmol L⁻¹ K₄[Fe(CN)₆] in 1.0 mol L⁻¹ KCl as a redox probe. Fig. 2A shows the proposed Randles' equivalent circuit fit the impedance of the electrodes satisfactorily. The values of the main parameters of equivalent circuit elements of EIS spectra (Fig. 3) are presented in Table I. In general, the unmodified CPGE is a pure resistor and only exhibits a non-ideal capacitive behavior at higher frequency range (Fig. 3a). Thus, due to the higher R_ct of bare CPGE, the non-ideal capacitive behavior is found, especially, on 6H with the highest R_ct 5117 Ω cm². Resistivities on 2H and 6B show that the electron transfer processes at the interfacial layer are facile at the 2H (k_app = 5.42 × 10⁻⁷ cm s⁻¹) for the rigid CPGE and 6B (k_app = 7.43 × 10⁻⁷ cm s⁻¹) for the soft CPGE (Table I). Hence, this brings to presumption that the 2H CPGE of the rigid region and 6B CPGE of the soft region of pencil hardness are the better substrate electrodes.

Zoom In Zoom Out Reset image size

Fig. 3b indicates the c-MWCNT/CPGE via DM has a better capacitive behavior over the whole frequency range and the imaginary part of impedance reach lower values than the bare CPGE. Though the decrease in resistance it is obvious of all the modified electrodes, the value seems less ideal to the EPD due to the imperfect coverage of c-MWCNT film. The c-MWCNT/CPGE by EPD (Fig. 3c) shows that regardless of the hardness, R_ct of all the CPGE has decreased and the R_ct for c-MWCNT/6B reaches to a minimum value of 5.64 Ωcm² with k_app of 1.3 × 10⁻⁴ cm s⁻¹. This result also indicates a more compact and homogenous has been produced by EPD.

The EIS results also closely agree with the CV data (Fig. 3d) obtained. It shows that the reversibility, stability and reproducibility of bare CPGE have been enhanced after c-MWCNT deposition. The gradual increase of peak currents (I_p) implies that the catalytic affects as well as kinetic properties of the electrode. The ΔE_p values for c-MWCNT/CPGE prepared by DM and EPD have converged to the theoretical value (59 mV) according to the Nernst Equation as compared to the bare CPGE (lowest in 2H 41.03 mV and highest in 6H 194.51 mV). A gradually increase in I_pa and decrease in ΔE_p also indicates the faster electron transfer rate of the c-MWCNT/CPGE by EPD.

The FESEM images (Fig. 4) of the electrodes indicate that the DM coating only modifies part of the surface, but the surface coverage is more compact and homogeneous with the EPD method. It is also evidence that for both methods entangled CNT mesh is formed on 2H surface (Fig. 4c and 4e) and a much better orientation is found on 6B (Fig. 4d and 4f). As a result, the available pores may have provided a multitude of pathways for electrons to pass through in a randomly organized MWCNT either by the sidewalls or through the open ends.

Zoom In Zoom Out Reset image size

The 6B CPGE is then selected for further quantitative study and many parameters haven been compared with conventional electrodes. CV (Fig. 5a) of the electrodes shows that the coating of c-MWCNT has improved the electron transfer process which leads to increase in I_pa. Results are also presented in Bode Plots. Fig. 5b exhibits a single phase angle maxima at all electrodes corresponding to one relaxation process of the electrode/bulk solution interface. The c-MWCNT modified electrodes demonstrate better capacitive behavior with a shifting and narrower distribution peak in lower frequency. Thus, the maximum at 0.25 Hz, −80.02° for c-MWCNT/6B via EPD indicates a more homogeneous coating of c-MWCNT. It also shows that the electron transfer process of the c-MWCNT/CPGE (DM or EPD) is much superior to the bare CPGE, carbon paste (CPE) and glassy carbon electrode (GCE). On the other hand, the c-MWCNT/6B via EPD has the highest conductivity compared to others in the Bode magnitude plots (Fig. 5c). These results further agree with the CV in Fig. 5a.

Zoom In Zoom Out Reset image size

The CV of 10 mg mL⁻¹ scopoletin in 0.1 M citric-Na₂HPO₄ buffer at the surface of c-MWCNT/CPGE is then investigated. The voltammetric behavior of scopoletin on EPD modified 6B electrode (EPD-CNTs/6B) in the pH range of 2.0 to 8.0 is examined. Fig. 6 shows that with the increase of pH value from pH 4.0 to 7.0, anodic peak current, I_pa, is stable but anodic peak potential, E_pa, has shifted toward negative. As pH 6.0 has the best peak separations, ΔE_p, and reasonable, I_pa, it is then selected for further study.

Zoom In Zoom Out Reset image size

Other electrodes are also applied for comparative study. It is shown from Fig. 7A that the I_pa of scopoletin is enhanced as and when the c-MWCNT/6B via EPD (I_pa 66.7 μA, E_pa 557 mV) is used. For comparison, the results obtained are GCE (I_pa 1.6 μA, E_pa 622 mV), bare 6B (I_pa 5.4 μA, E_pa 622 mV) and DM-CNTs/6B (I_pa 9.6 μA, E_pa 528 mV). Fig. 7B shows the linear regression equations peak currents and scan rates of scopoletin on bare 6B, DM-CNTs/6B, and EPD-CNTs/6B. This demonstrates that the current is controlled by diffusion process and fast charge-transfer kinetics occurs on EPD-CNTs/6B. At optimum conditions, parameters like calibration curve and reproducibility have been studied by differential pulse voltammetry (DPV). Fig. 8A Shows a linear regression equations of I_pa (μA) = 4.206 C (mg L⁻¹) +18.436 (γ = 0.9968) is obtained with limit of detection (LOD) of 0.062 mg L⁻¹ (S/N = 3). As scopoletin absorbs strongly at 382.70 nm in 0.1 M citric-Na₂HPO₄ buffer (pH 8.0) and produces blue fluorescence, UV-vis is then used as comparison (Fig. 8B). The calibration curve is obtained with a linear regression equations of A = 0.0821 C (mg L⁻¹) + 0.0448 (γ = 0.9986) and LOD of 0.48 mg L⁻¹.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

The measurement by c-MWCNT/6B can be repeated several times in a blank solution of 10 mg L⁻¹ scopoletin. The electrode has shown a good operational stability with relative standard deviation (RSD) of I_pa 4.2%. After repetitive use (over 30 times) the response of the electrode has only decreased slightly, which indicates a good storage stability of the c-MWCNT/6B prepared through EPD method.

The c-MWCNT/6B is investigated for the determination of total coumarin in dry fruits of A. sessiliflorus. This is carried out by adding scopoletin as the standard. Results from the proposed method using EPD-CNTs/6B as the electrode and UV-Vis method are shown in Table II. The proposed method shows relatively LOD and better recovery than the UV-Vis method. This means the proposed method has good sensitity and reproducibility. The amount of hyperin found in the A. sessiliflorus powder is 1.89 mg g⁻¹, and with recoveries ranging between 97.3% and 101.2%. This indicates a successful application of the proposed method for determination of coumarin.