Gold nanoparticles/tetraaminophenyl porphyrin functionalized multiwalled carbon nanotubes nanocomposites modified glassy carbon electrode for the simultaneous determination of p-acetaminophen and p-aminophenol

2.1 Reagents and apparatus

4-Acetamidophenol(AMP) and 4-aminophenol(AP) were purchased from Aladdin. MWCNTs were purchased from Shenzhen Nanotech Port Co., Ltd. Disodium hydrogen phosphate dodecahydrate (Na₂HPO₄·12H₂O) and sodium dihydrogen phosphate dihydrate (NaH₂PO₄·2H₂O) were obtained from Yantai Shuangshuang Chemical Co., Ltd. All other reagents were of analytical grade and double distilled water was used throughout the experiments. Paracetamol Dispersible Tablets were purchased from local drugstore and water sample was collected in Yellow River (Lanzhou).

All electrochemical experiments were carried out on a CHI660 B electrochemical workstation (Shanghai, China) equipped with a three-electrode cell with a glassy carbon electrode (GCE, 3 mm diameter) or a modified GCE as the working electrode, a saturated calomel electrode (SCE) as the reference electrode and a platinum sheet as the counter electrode. Scanning electron microscopy (SEM) images were determined with a JSM−6701 F field emission scanning electron microscope (Japanese Electron Optics Company). Fourier transform infrared spectra (FTIR) of the samples were recorded in the spectral range of 4000 - 400 cm^-1 using the Nicolet 6700 Infrared Spectrometer (Thermo Scientific Co., USA). Elemental chemical states on the surface of the materials were measured by ESCALAB 250Xi X-ray photoelectron spectroscopy (XPS, Thermo, USA).

2.2 Preparation of tetraaminophenyl porphyrin and acryl chloride-functionalized CNTs

Tetraaminophenyl porphyrin (TAPP) was synthesized via a literature procedure (Bettelheim et al., 1987). The MWCNTs were purified by refluxing the as-received MWCNTs in concentrated nitric acid for 36 h at 80 °C (Pumera and Iwai, 2009). Carboxyl-functionalized CNTs (CNTs-COOH) were prepared by treating the purified MWCNTs with a concentrated nitric/sulfuric acid (1:3, v/v) mixture solution at 50 °C for 6 h. When cooled to room temperature, the suspension was filtered, washed with water several times until the aqueous solution reached neutral pH, and then dried under vacuum at 50 °C (Zhang et al., 2015). Acryl chloride-functionalized CNTs (CNTs-COCl) were prepared by suspending 100 mg of CNTs-COOH into a mixture of thionyl chloride (SOCl₂) and dimethylformamide (DMF) (20:1, v/v) at 70 °C for 24 h. The suspension cooled to room temperature was centrifuged and washed with anhydrous tetrahydrofuran several times and dried to obtain anhydrous CNTs-COCl (Shen et al., 2007).

2.3 Synthesis of carbon nanotubes/tetraaminophenyl porphyrin nanocomposites

The nanocomposites were prepared using a modified reference method (Gromov et al., 2015). Typical preparation was as follows: 10 mg of TAPP and 20 mg of CNTs-COCl were added to 20 mL of anhydrous DMF solution. The mixture was reacted at 100 °C under nitrogen atmosphere for 96 h, and then treated using 20 mL of ammonia at 70 °C for 24 h. Finally, the resulted CNTs-CONH-TAPP nanocomposite dispersion was filtrated, washed with ultrapure water and dried in a vacuum at 50 °C overnight.

2.4 Fabrication of the AuNPs/CNTs-CONH-TAPP/GCE

Prior to modification, the glassy carbon electrode (GCE) was polished with 0.05 μm alumina powder slurries to get a mirror-like surface, then sonicated sequentially in ethanol/ultrapure water mixture(1:1, v/v) and double distilled water, and allowed to dry in N₂ atmosphere at room temperature. Afterwards, 2 mg of CNTs-CONH-TAPP nanocomposites was dispersed in 1 mL of DMF solution. 6 μL of the CNTs-CONH-TAPP nanocomposite suspension was dropped on the surface of the bare glassy carbon electrode and dried under an infrared lamp for 10 min. To electrochemically deposit AuNPs on the surface of the CNTs-CONH-TAPP/GCE, the CNTs-CONH-TAPP/GCE was immersed in 0.2 g L⁻¹ HAuCl₄ solution and treated at the constant potential of −0.2 V for 90 s. Finally, the electrode was rinsed with distilled water and dried under an infrared lamp for later use. The whole preparation process is shown in Scheme 1.

Close

Scheme 1

Preparation procedure of the modified electrode.

Export to PPT

2.5 Electrochemical measurements

The AuNPs/CNTs-CONH-TAPP/GCE was used as working electrode in a three-electrode electrochemical cell. The experiments were carried out by studying the cyclic voltammetric behavior of the p-acetaminophen and p-aminophenol at a potential range of from −0.2 to 0.6 V.

The DPV was performed at potential range of from 0.0 V to 0.6 V, with a pulse width of 0.05 s, pulse period of 0.5 s and potential increment of 4 mV.

3.1 Characterization of the electrode materials and the modified electrodes

IR spectra of the prepared CNTs-COOH(a), CNTs-COCl(b), TAPP(c) and CNTs-CONH-TAPP(d) are shown in Fig. 1. CNTs-COOH showed an obvious —OH stretching vibration band at 3363 cm⁻¹ and a C⚌O stretching vibration band at 1709 cm⁻¹ (curve a). CNTs-COCl showed a sharp C⚌O stretching band at 1701 cm⁻¹ and a weak C—Cl stretching band at 623 cm⁻¹ respectively (curve b). TAPP showed characteristic absorption bands at 1612 cm⁻¹, 1506 cm⁻¹, 1450 cm⁻¹, respectively assigned to benzene ring and pyrrole ring skeleton vibration peaks, and the bands at 3448 cm⁻¹ and 3355 cm⁻¹ ascribed to primary amine stretching vibration peaks (curve c). After reaction of CNTs-COCl and TAPP to form CNTs-CONH-TAPP, the stretching vibration of the -CONH- at 3444 cm⁻¹, C-N stretching vibration at 1247 cm⁻¹ and -NH- bending vibration at 687 cm⁻¹ appeared (curve d). These results indicated that the CNTs-CONH-TAPP nanocomposites had been successfully fabricated.

Close

Fig. 1

FT-IR spectra of (a) TAPP, (b)CNTs-COOH, (c) CNTs-COCl and (d) CNTs-CONH-TAPP.

Export to PPT

Fig. 2 shows typical XPS survey spectra and high-resolution elemental scan of N1s of the prepared CNTs- CONH-TAPP. As shown in Fig. 2A, in XPS survey spectra appeared C1s, N1s and O1 characteristic peaks at 284.88 eV, 399.44 eV and 532.86 eV, respectively. The high resolution spectra of N1 s showed that there were four nitrogen-containing functional groups in CNTs-CONH-TAPP, which were —C⚌N (400.44 eV), ⚌CN— (400.31 eV) in the porphyrin ring with similar area, -CONH-(399.94 eV) and —NH₂ (399.44 eV), respectively (Fig. 2B). -CONH- binding energy was larger than the NH₂ binding energy, mainly because N in —CONH- was adjacent to the electron withdrawing group. In addition, the area ratio of the two peaks is about 1:1, indicating that the two amino groups in the tetraaminophenyl porphyrin were involved in the amidation reaction. The results demonstrated that CNTs-COCl reacted successfully with tetraaminophenyl porphyrin, which was consistent with IR results.

Close

Fig. 2

XPS survey spectra of CNTs- CONH-TAPP and high resolution elemental scan of N1s.

Export to PPT

The SEM images of the surface of CNTs-CONH-TAPP/GCE and AuNPs/CNTs- CONH-TAPP/GCE were shown in Fig. 3. The SEM image of the CNTs-CONH- TAPP/GCE showed a three-dimensional network structure and this network structure played an important role in electron transfer. In addition, the porous network structure made the electrode surface area and the target molecular adsorption capacity greatly increase to further improve the sensitivity of the determination (Fig. 3A). π-π interaction between CNTs and porphyrin molecules made the composites uniformly distributed and there was not obvious agglomeration of CNTs or porphyrin molecules. Fig. 3B shows that uniform size of AuNPs are well-distributedly dispersed on the walls of the CNTs-CONH-TAPP.

Close

Fig. 3

SEM images of CNTs-CONH-TAPP/GCE (A) and AuNPs/CNTs-CONH-TAPP/GCE (B).

Export to PPT

3.2 Electrochemical properties of AuNPs/CNTs-CONH-TAPP/GCE

The electrochemical characteristics of the different modified electrodes were investigated by cyclic voltammetry in 0.1 mol L⁻¹ KCl solution containing 1 mmol L⁻¹ Fe(CN)₆^3−/4− at a scan rate of 100 mV s⁻¹. Fig. 4A shows the CVs of GCE(a), TAPP/GCE(b), CNTs-CONH-TAPP/GCE(c) and AuNPs/CNTs-CONH-TAPP/GCE(d). As shown in curve a, a couple of weak redox peaks were observed on the bare GCE. After the electrode was covered by TAPP(b), the redox peaks completely disappeared, indicating that TAPP completely blocked interfacial electron transfer. However, when it was modified with the CNTs-CONH-TAPP nanocomposites (curve c), the currents of redox peaks increased obviously, suggesting that the CNTs-CONH-TAPP nanocomposites could improve the electron transfer on the surface of CNTs-CONH-TAPP/GCE. What’s more, when AuNPs were deposited on the surface of CNTs-CONH-TAPP/GCE(curve d), the redox currents further enhanced.

Close

Fig. 4

(A) CVs of bare GCE (a), TAPP/GCE (b), CNTs-CONH-TAPP/GCE (c) and AuNPs/CNTs-CONH-TAPP/GCE (d) in 0.1 mol L⁻¹ KCl containing 1 mmol L⁻¹ Fe(CN)₆^3−/4−. (B) EIS of bare GCE (a), CNTs-CONH-TAPP/GCE (b) and AuNPs/CNTs- CONH-TAPP/GCE (c) in 0.1 mol L⁻¹ KCl containing 5 mmol L⁻¹ [Fe(CN)₆]^3−/4−. Inset: Amplification of high frequency range.

Export to PPT

Electrochemical impedance spectroscopy (EIS) is usually used to characterize the features of a surface to allow the understanding of chemical transformations and processes associated with the conductive electrode surface. The Nyquist plots were made in 0.1 mol L⁻¹ KCl solution containing 5 mmol L⁻¹ Fe(CN)₆^3−/4− at 100,000–0.01 Hz. In the Nyquist diagram (Fig. 4B), the semicircle diameter of EIS in high frequency range is equal to R_et. The small semicircle at high frequencies and linear part at low frequencies could be observed for bare GCE. The semicircle diameter of the CNTs-CONH-TAPP nanocomposites modified electrode was much smaller than that of bare GCE. This result showed the smart conductivity of the CNTs-CONH-TAPP nanocomposites, which reduced the electron transfer resistance of the GCE. After gold nanoparticles were deposited on the CNTs-CONH-TAPP film surface, it was found clearly that the electron transfer resistance decreased significantly due to the superior conductivity. The above resulted could clearly suggest the success of the assembly of the electrode.

3.3 Electrochemical behaviors of AMP and AP at different modified electrodes

The electrochemical behaviors of AMP and AP were investigated by cyclic voltammetry (CV). In contrast, CV curves of AuNPs/CNTs-CONH-TAPP/GCE in 0.1 mol L⁻¹ PBS (pH = 7.0) in the presence/absence of AMP and AP at 100 mV s⁻¹ were shown in Fig. 5A. In blank 0.1 mol L⁻¹ PBS (pH = 7.0), there was not any redox peak appeared (curve a). When AP was added (curve b), a pair of well-defined redox peaks were observed at 0.032 and 0.108 V respectively, attributing to the electrochemical redox behavior of AP, while the modified electrode was immersed in AMP solution, a pair of well-defined redox peaks appeared at 0.255 and 0.400 V respectively resulted from redox of AMP (curve c). Notably, the modified electrode immersed in the mixture solution of AP and AMP exhibited two couples of well-defined reduction and oxidation peaks respectively at 0.037 V and 0.108 V for AP and 0.240 V and 0.410 V for AMP. Their formal potentials (E^o′), calculated as the midpoint of anodic and cathodic peak potentials, were 73 mV and 325 mV corresponding to AP and AMP, respectively. Since the oxidation peak of AMP was shifted to a more positive potential, distinct redox peaks were acquired for AMP. The difference between the formal potentials of AMP and AP was found to be 252 mV. Hence, the highly selective simultaneous determination of AMP and AP in PBS buffer became possible with the AuNPs/CNTs-CONH-TAPP/GCE.

Close

Fig. 5

(A) CVs of AuNPs/CNTs-CONH-TAPP/GCE in 0.1 mol L⁻¹ PBS (pH = 7.0) solutions without AMP and AP (a), with AP (b), AMP (c) and AP + AMP (d), respectively, at a scan rate of 100 mV s⁻¹; (B) CVs of bare GCE (a), TAPP/GCE (b), CNTs-CONH-TAPP/GCE (c) and AuNPs/CNTs-CONH-TAPP/GCE (d) in 0.1 mol L⁻¹ PBS (pH = 7.0) containing 0.03 mmol L⁻¹ AP and AMP.

Export to PPT

Fig. 5B shows the cyclic voltammograms of the bare GCE(a), TAPP/GCE(b), CNTs-CONH-TAPP/GCE(c) and AuNPs/CNTs-CONH-TAPP/GCE(d) in 0.1 mol L⁻¹ PBS (pH = 7.0) containing 0.03 mmol L⁻¹ AMP and AP. There were a pair of weak and broad redox peaks for AP and no reduction peak appeared for AMP on the bare GCE. TAPP modified glassy carbon electrode showed no redox response. But when CNTs-CONH-TAPP was dropped on the surface of bare GCE electrode, the redox peak currents of AMP and AP obviously increased. The enhancement in peak currents might be ascribed to the high surface area, excellent electrocatalytic activity and good conductivity of the CNTs-CONH-TAPP. Moreover, deposition of AuNPs further enhanced the current responses, indicating that AuNPs showed a good electrocatalytic activity towards redox reaction of AMP and AP on the modified electrode. Two couples of well-separated and strong quasi-reversible redox peaks promised a potential for chemically sensing determination of AMP and AP simultaneously. Curve b did not show any redox response for AMP and AP, which indicated introduction CNTs into TAPP played an important role.

3.4 Optimization of the experimental conditions

Influence of the dripping amount of CNTs-CONH-TAPP of 2 mg mL⁻¹ on the oxidation peak current of 0.03 mmol L⁻¹ AP and AMP were investigated by CV and the results were showed in Fig. S1. The oxidation peak currents of AP and AMP increased obviously with the amount of CNTs-CONH-TAPP suspension from 3.0 to 6.0 μl and then decreased when it further increased. The reason might be that too much amount of CNTs-CONH-TAPP would hinder the electron delivery and the resulting film was easy to fall off. Thus, 6.0 μl was used in further experiment.

In order to acquire the optimal electrochemical performance of the modified electrode, deposition conditions of gold nanoparticles were optimized. Figs. S2–S4 showed that when concentration of HAuCl₄ solution, deposition potential and deposition time were 0.2 g L⁻¹, −0.2 V and 90 s respectively, the oxidation peak currents of the two target molecules were the highest. When the deposition time was too long or the concentration of HAuCl₄ was too high, the gold nanoparticles would agglomerate so that the effective surface area of the composite would decrease and the current response would reduce.

The effect of pH on the electrochemical behavior of AMP and AP on the modified electrode in 0.1 mol L⁻¹ PBS over the range from 5.5 to 8.5 by DPV (Fig. S5). The oxidation currents gradually increased with increasing pH from 5.5 to 7.0 then decreased. So pH 7.0 was chosen for the subsequent analytical experiments. The oxidation peak potentials (Epa) of AMP and AP shifted linearly toward more negative potential with the pH increasing from 5.5 to 8.5, indicating that protons took part in the electrode reaction. The relationships between the peak potentials and pH could be expressed by the equations: E_pa = 0.7777–0.0515pH (R² = 0.9934) for AMP and E_pa = −0.0572pH + 0.5397 (R² = 0.9969) for AP, respectively. The absolute values of the slopes were close to the theoretical value of 59 mV/pH, indicating that the number of protons was equal to the number of the transferred electrons in the processes (Liu et al., 2011).

The kinetics of the electrode reaction was investigated by studying the effects of the scan rate on the redox peak potentials and currents of AMP and AP at the AuNPs/CNTs-CONH-TAPP/GCE. With the increase of the scan rate, the redox peak current increased gradually and the anode peak and cathode peak shifted to more positive and negative potentials, respectively. This showed that the responses of AMP and AP to the modified electrode were quasi-reversible process. As shown in Fig. S6, the redox peak currents of AMP and AP were linearly proportional to the scan rate (ν) from 40 to 140 mV s⁻¹ with regression equations: I_pa,AMP (μA) = −10.0213–0.4723ν (mV s⁻¹) (R² = 0.9942), I_pc,AMP(μA) = 8.9086 + 0.2966ν (mV s⁻¹) (R² = 0.9932) and I_pa,AP(μA) = −3.5505–0.4161ν (mV s⁻¹) (R² = 0.9971), I_pc,AP(μA) = 4.2329 + 0.4089ν (mV s⁻¹) (R² = 0.9971), respectively, indicating that the electrochemical process of AMP and AP were adsorption-controlled.

Moreover, as shown in Fig. S7, the anode and cathode peak potentials of AMP and AP had liner relationships with the natural logarithm of scan rates (lnν). The regression equations were expressed as E_pa,AMP(V) = 0.3505 + 0.0289lnν (mV s⁻¹) (R² = 0.9925); E_pc,AMP(V) = 0.4318–0.0284lnν (mV s⁻¹) (R² = 0.9948) and E_pa,AP(V) = 0.0626 + 0.0224lnν (mV s⁻¹) (R² = 1.0000); E_pc,AP(V) = 0.1740–0.0184lnν (mV s⁻¹) (R² = 1.0000), respectively. According to the Laviron’s model (Laviron,1974; Laviron,1979), the electrochemical parameters (n, ks and α) can be calculated by the following formulas:

Close

Scheme 2

The electrode reaction mechanisms of AMP and AP.

Export to PPT

The effects of accumulation potential and accumulation time on the peak currents AMP and AP were also demonstrated by DPV. As shown in Fig. S8A, the oxidation peak currents increased gradually and then decreased from −0.1 to 0.30 V. The most sensitive response was found at 0.1 V. According to the Fig. S8B, when the accumulation time varied from 70 s to 110 s, the oxidation peak currents of AMP and AP increased gradually at first and then decreased, and the maximum value was obtained for 100 s. Therefore, 0.1 V and 100 s were chosen as the optimal accumulation potential and time for simultaneous determination of AMP and AP.

3.5 Chronocoulometry

The electrochemical effective surface areas (A) of bare GCE (a), CNTs-CONH -TAPP/GCE(b) and AuNPs/CNTs-CONH-TAPP/GCE(c) can be determined by chronocoulometry according to Anson equation (Anson, 1964): $$\text{Q}=2{\text{nFAcD}}^{1/2}{\text{t}}^{1/2}/{\mathit{pi}}^{1/2}+{\text{Q}}_{\text{dl}}+{\text{Q}}_{\text{ads}}$$ where Q_ads is Faradic charge, Q_dl is double layer charge which can be eliminated by background subtraction, D is diffusion coefficient, c is concentration of the substrate and A is surface area of the working electrode. Other symbols have their usual meanings. The experiment was firstly carried out in 1 mmol L⁻¹ Fe(CN)₆^3−/4− solution containing 0.1 mol L⁻¹ KCl. As shown in Fig. S9(A, B), the linear regression equations of Q - t^1/2 curves on bare GCE(a), CNTs-CONH -TAPP/GCE(b) and AuNPs/CNTs-CONH -TAPP/GCE(c) were Q(mC) = 0.0098t^1/2 + 0.0109 (R² = 0.9926), Q(mC) = 0.0462t^1/2 + 0.1621 (R² = 0.9918) and Q(mC) = 0.0633t^1/2 + 0.1500 (R² = 0.9911), respectively. Based on the slopes of the equations, A was calculated to be 0.0326cm² for bare GCE, 0.1535 cm² for CNTs-CONH-TAPP/GCE and 0.2104cm² for AuNPs/CNTs-CONH-TAPP/GCE. These results indicated that the electrode effective surface area was increased obviously after modification, which could increase the electrochemical active sites, enhance the electrochemical response and decrease the detection limit.

Furthermore, chronocoulometry was also used to determine the saturation absorption capacity for AMP and AP at AuNPs/CNTs-CONH-TAPP/GCE surface in 0.1 mol L⁻¹ PBS (pH = 7.0). As can be seen in Fig. S9(C,D), bigger intercepts and almost same slope were obtained from curves b and c compared with a, which further meant that the oxidation processes of AMP and AP were mainly controlled by adsorption (Ye et al., 2014). The plots of Q versus t^1/2 showed good linear relationships and the regression equations could be expressed as Q(mC) = 0.0357t^1/2 + 0.1274 (R² = 0.9924) (in absence of AMP and AP), Q(mC) = 0.0459t^1/2 + 0.1869 (R² = 0.9924) (in presence of 0.03 mmol L⁻¹ AMP) and Q(mC) = 0.0394t^1/2 + 0.1703 (R² = 0.9917) (in presence of 0.03 mmol L⁻¹ AP), respectively. Q_ads was calculated to be 0.0595mC and 0.0429mC for AMP and AP. As n = 2, A = 0.2104cm², according to the Laviron equation Q_ads = nFAΓ^∗, the maximum adsorption capacity (Γ^∗) of AMP and AP could be obtained as 1.47 × 10⁻⁶ and 1.06 × ⁻⁶ mol cm⁻².

3.6 Individual and simultaneous determination of AMP and AP using AuNPs/CNTs -CONH-TAPP/GCE

DPV was performed to investigate the relationship between the peak current and concentration of AMP and AP. The individual determination of AMP or AP in their mixtures was made when the concentration of one species changed while another remained constant. As shown in Fig. 6A, with increasing of AMP concentration its oxidation peak current increased but that of AP almost did not change when concentration of AP was kept 28 μmol L⁻¹. I_pa was linearly proportional to AMP concentration in the range of 4.5 to 500 μmol L⁻¹ with a linear equation I(μA) = −44.8496–0.03774 c (μmol L⁻¹) (R² = 0.9950) and the limit of detection of 0.44 μmol L⁻¹ (S/N = 3), which was comparable to that at C-Ni/GCE (0.6 μmol L⁻¹) (Wang et al., 2007), GR/GCE (0.6 μmol L⁻¹) (Kang et al., 2010), MWCNT-PDDA-PSS/GE (0.5 μmol L⁻¹) or even lower than that at SWCNTs/nafion modified nitrocellulose/AuNP-PGA/GCE (15.0 μmol L⁻¹) (Lee et al., 2016), AuNP-PGA/SWCNE (1.18 μmol L⁻¹) (Bui et al., 2012), HRP-PEI-SWCNT/GCE (1.36 μmol L⁻¹) (Tertiş et al., 2013), CuO-NiO/GR/GCE (1.33 μmol L⁻¹) (Liu et al., 2016), fCNT-PMG-CE (9.3 μmol L⁻¹) and PMG-fCNT-CE (4.3 μmol L⁻¹) (Barsan et al., 2015), HRP-Ppy-SWCNT/SPE (8.09 μmol L⁻¹) (Tertiş et al., 2013). Similarly, as shown in Fig. 6B, the peak current of AP linearly increased with its concentration increasing from 0.08 to 60 μmol L⁻¹ in presence of 180 μmol L⁻¹ AMP with a linear equation I(μA) = −40.5489–0.278 c (μmol L⁻¹) (R² = 0.9953) and the detection limit of 0.025 μmol L⁻¹ (S/N = 3), which was lower than that at carbon electrode (1.4 μmol L⁻¹) (Chu et al., 2008), p[NVCzVBSA1]/CFME (1.0 μmol L⁻¹) (Jamal et al., 2004), G-PANI/CPE coupled with droplet-based microfluidic sensor (15.67 μmol L⁻¹) (Rattanarat et al., 2016), or close to that at SWNTs/POAPE (0.06 μmol L⁻¹) (Wang Z. et al., 2009)), graphene–chitosan/GCE (0.057 μmol L⁻¹) (Yin et al., 2010), SPCE (0.0767 μmol L⁻¹) (Lamas-Ardisana et al., 2008). As a method for simultaneous determination of p-acetaminophen and p-aminophenol, the detection limit of the proposed method was also relatively lower than or close to those of some electrochemical methods reported previously, such as DPV with PEDOT/GCE (0.4 μmol L⁻¹ for AMP and 1.2 μmol L⁻¹ for AP) (Mehretie et al., 2011), DPV with CILE (0.5 μmol L⁻¹ for AMP and 0.10 μmol L⁻¹ for AP) (Safavi et al., 2008), DPV with Au/Pd/rGO/GCE (0.30 μmol L⁻¹ for AMP and 0.12 μmol L⁻¹ for AP) (Wang et al., 2017), DPV with GR–PANI/GCE (0.065 μmol L⁻¹ for AP, not reported for AMP) (Fan et al., 2011) and DPV with RGO–TiN/GCE (0.02 μmol L⁻¹ for AMP and 0.013 μmol L⁻¹ for AP) (Kong et al., 2015), and other methods, such as CE with amperometric detection (1.4 μmol L⁻¹ for AMP and AP) (Chu et al., 2008), spectrophotometry (1.2 μmol L⁻¹ for AMP and AP) (Cekic et al., 2005) and CZE equipped with a diode array detector (0.028 μmol L⁻¹ for AMP and 0.10 μmol L⁻¹ for AP) (Pérez-Ruiz et al., 2005). Moreover, the simultaneous determination of AMP and AP was investigated by changing their concentrations simultaneously. As shown in Fig. 7, the peak currents of AMP and AP were linear with increasing AMP and AP concentrations and corresponding linear equations were I(μA) = −61.1655–0.0544 c (μmol L⁻¹) (R² = 0.9960) and I(μA) = −46.8382–0.3201c (μmol L⁻¹) (R² = 0.9945), respectively. And the linear ranges for AMP and AP were 50 to 348 μmol L⁻¹ and 4 to 42 μmol L⁻¹, respectively. It can be seen that individual or simultaneous determination of AMP and AP on AuNPs/CNTs-CONH-TAPP/GCE could be achieved with high sensitivity and selectivity. These results demonstrated that AuNPs/CNTs-CONH-TAPP/GCE were a promising candidate for the simultaneous determination of AMP and AP. In order to show the real application possibilities of the proposed method for determining of AMP and AP, the detection (S/N = 3) and quantification (S/N = 10) limits have also been calculated in real matrix (commercially available Paracetamol Dispersible Tablets and Yellow River water) and it was found that the sensitivity was acceptable (Table S1).

Close

Fig. 6

DPVs at AuNPs/CNTs-CONH-TAPP/GCE modified electrode in 0.1 mol L⁻¹ PBS (pH = 7.0) containing 28 µmol L⁻¹ AP and different concentrations of AMP: 4.5, 45.2, 70, 86, 92, 100, 120, 200, 252, 300, 400, 500 µmol L⁻¹ (from up to down) (A) and containing 180 µmol L⁻¹ AMP and different concentrations of AP: 0.8, 3, 8, 12, 16, 18, 20, 25, 30, 35, 40, 50, 60 µmol L⁻¹ (from up to down) (B). Inset: Calibration plots of the anodic peak current versus concentrations of AMP and AP.

Export to PPT

Close

Fig. 7

DPVs at AuNPs/CNTs-CONH-TAPP/GCE modified electrode in 0.1 mol L⁻¹ PBS (pH = 7.0) containing different concentrations of AMP: 50, 70, 80, 120, 148, 180, 220, 250, 300, 348 μmol L⁻¹ and AP: 4, 6, 10, 14, 18, 22, 26, 30, 36, 42 μmol L⁻¹ respectively (from a to g). Inset: Calibration plots of the anodic peak current versus concentration of AMP and AP.

Export to PPT

3.7 Reproducibility, stability and interferences

The reproducibility of the AuNPs/CNTs-CONH-TAPP/GCE was estimated by fabricating four electrodes independently and using them to determine 120 μmol L⁻¹ AMP and 35 μmol L⁻¹ AP. The relative standard deviation (RSD) was 1.93% and 3.37%, respectively, indicating an excellent reproducibility. Repeatability of the modified electrode was investigated and the relative standard deviation (RSD) of the sensor response to 120 μmol L⁻¹ AMP and 35 μmol L⁻¹ AP was 3.3% and 2.2% for ten successive measurements. After each measurement the modified electrode was immersed 0.1 mol L⁻¹ blank PBS buffer and was CV scanned for 10 cycles to remove adsorbed contaminant. The result indicated the sensor had a good repeatability.

The anti-interference ability of the AuNPs/CNTs-CONH-TAPP/GCE was evaluated by adding potential interfering substances into 0.1 mol L⁻¹ PBS (pH = 7.0) containing 120 μmol L⁻¹ AMP and 35 μmol L⁻¹ AP. The results illustrated that 50-fold K⁺, Ca²⁺, Mg²⁺, Cl⁻, glucose, citric acid, L-ascorbic acid, phenol, hydroquinone, dopamine, salicylic acid and caffeine did not interfere the detection of AMP and AP (Fig. S10). These results demonstrated that the modified electrode had a good anti-interference ability toward the determination of AMP and AP. The good selectivity was attributed to π-π interaction between the porphyrin ring and the analyte molecules and hydrogen bond interaction between hydroxyl and amino groups of AMP and AP molecules and carbonyl groups of the functionalized carbon nanotubes.

3.8 Application in real samples analysis

In order to evaluate the feasibility of the modified electrode, the proposed method was employed for assaying AMP and AP in Paracetamol Dispersible Tablets and Yellow River water under the optimal conditions with standard addition method. Prior to determination, four tablets of Paracetamol Dispersible Tablets were finely pulverized in a mortar and accurately weighed. The sample was extracted with 60 mL of anhydrous ethanol for 1 h in an ultrasonic bath, and then it was filtered and the filtrate was collected and diluted to 100 mL with deionized water. After the above sample solution was diluted 250 times with 0.1 mol L⁻¹ PBS (pH = 7.0), the resulting solution was transferred to electrolytic cell for the determination using DPV. Water sample was first filtrated with a 0.45 μmol L⁻¹ membrane filter and then injected in 0.1 mol L⁻¹ PBS (pH = 7.0) solution for DPV measurement. The results were presented in Table 1. It can be seen that AMP and AP in mixtures could be satisfactorily detected with the recoveries in the range from 97.8% to 109.1% for AMP, and from 94.4% to 103.8% for AP, additionally, the RSD (n = 3) was less than 5.0%.

The results show that tablet matrix does not have any interference on the simultaneous determination of the analytes. The amount of AMP in each pharmaceutical tablet was found to be 104.4 mg with 4.4% error, showing a good agreement with the content of AMP given by the manufacturer.

These results indicated that the proposed electrode could be efficiently applied to AMP and AP detection in real samples.

As seen in Table 1, p-aminophenol was not found in any of the pharmaceutical samples. The amount of p-aminophenol determined by The United States, British and Chinese Pharmacopoeias is limited to 0.005% (w/w) in paracetamol containing products. The limits of p-aminophenol may change depending on the formulation and dosage. p-Aminophenol is limited to 0.1% (w/w) in the monograph of paracetamol tablets present in British Pharmacopeia (Sornchaithawatwong et al., 2010). In this way, the proposed sensor can achieve the limit concentration of p-aminophenol in a pharmaceutical tablets of 400 mg using appropriate dilutions for sample preparation.

AMP and AP were not detected in Yellow River water sample, demonstrating that AMP and AP contents in the tested sample was below the detection limit of the present method. Notably, HPLC with porous graphitized carbon (PGC) column (Monsera and Darghouth, 2002) also did not find their existence.