3.1. Characterization and evaluation of synthesis of CuO nanoclusters on CF surface (CF/CuO)
FESEM images of anodized CF surface (Fig. 1 (A-E)) illustrate arrays of CuO nanoclusters with a large surface area which uniformly covered the macro porous surface of the CF. At higher magnifications, the diameter of the nanoclusters varies from 30.21 to 72.42 nm. The FESEM images show a uniform distribution of CuO nanoclusters on the CF surface, which ultimately leads to an increase in the surface area of CF, providing a suitable substrate for the uniform growth of Pt NPs and preventing their accumulation in the subsequent stages. EDX analysis was also performed to evaluate the purity of the CF surface (Fig. 1 (F)). As can be seen, oxygen (O) and copper (Cu) are the only two constituents of anodized CF. By examining the purity of these components (oxygen with a purity of 47.46 and copper with a purity of 52.54), it can be concluded that CuO is formed on the surface of CF without any other impurities. In all experiments, the surface area was 0.031 cm2.
3.2. Characterization and Evaluation of Synthesis of Pt Nanoparticles on CF Electrode Modified with CuO Nanoclusters (CF/CuO/Pt)
Figures 2 (A-D) shows FESEM images of the CF/CuO/Pt surface at different magnifications. As seen, the use of a three-dimensional substrate of ultra-thin CuO nanoclusters led to the uniform growth of Pt nanoparticles without agglomeration on the surface of the CF. This property will enhance the electrocatalytic activity of the nanocomposite. The FESEM images also demonstrate the spherical morphology of the synthesized Pt with an average diameter of 19.59–51.33 nm. EDX analysis was also used to investigate the purity and elemental distribution of the electrode surface after the growth of Pt nanoparticles (Fig. 2 (E)). As seen, the EDX diagram only shows the constituent elements of the CF/CuO/Pt nanocomposite, namely Pt, Cu and O. No impurities were detected in the prepared nanocomposite. According to Fig. 2 (F) and in accordance with the FESEM results, Pt NPs are evenly distributed on the electrode surface. The pattern observed in Fig. 2 (G) shows four severe grazing XRD diffraction peaks at 2θ = 41, 43, 50.5, and 74\(^\circ\), representing CuPt (006), Cu (111), Cu (200) and Cu (220) planes, respectively. Several other diffraction peaks at 2θ = 35.5, 38.5, 40, 45.5, and 48\(^\circ\) also suggest the presence of CuO (002), CuO (111), Pt (111), Pt (200), and CuPt (404) planes, respectively. The density of diffraction Pt (200) is less than the card number 01-087-0640 and has half wider peak, since Pt could be reduced and absorb hydrogen under hydrothermal conditions.
3.3. Selection of the most suitable functional monomer
According to Table S1 and Fig S2, the stability of the complexes formed in the gas phase between each of the monomers and CHO molecule follows the following order: (2-Merc)> (p-Hyd)> (p-PD)> (p-Amin). On the other hand, the influence of the solvent was also investigated as the electro-polymerization operation was performed in the liquid phase and the solvent can affect the molecular imprinting process and its efficiency. The polarizable continuum model (PCM) was implemented for molecular complex systems to estimate the interaction energies between CHO and each monomer dissolved in a specific solvent. The results are presented in Table 1. As seen, the presence of ethanol affects the bond energy between the monomer and CHO. The stability ranking of the complexes in the presence of ethanol is different from that of the gas phase. Accordingly, the stability of the complexes formed in the ethanol solvent between each of the monomers and the CHO molecule has the following order: (2-Merc) > (p-PD) > (p-Amin) > (p-Hyd). The absolute value of the bond energies for 2-Merc and p-PD are 7.482429 and 6.1613, respectively. Regarding to the proximity of the bonding energies of these two monomers and the greater stability of p-PD, the p-PD was employed in the synthesis of MIP.
Table 1
Calculated bond energies for the monomer-CHO complex in ethanol solvent.
Molecules
|
E (Hartree)
|
∆E (Hartree)
|
∆E (kJ/mol)
|
p-Phen- CHO
|
-1474.4770357
|
-0.0098187
|
-6.1613
|
p-Hyd- CHO
|
-1837.1262634
|
0.0032411
|
2.03382
|
p-Amin-CHO
|
-1817.2906653
|
-0.0066463
|
-4.17061
|
2-Merc-CHO
|
-1909.521962
|
-0.011924
|
-7.482429
|
3.4. Evaluation of synthesis of p-PD on CF/CuO/Pt nanocomposite by FT-IR
The FT-IR spectrum of p-PD synthesized by the electropolymerization process on CF/CuO/Pt/GCE is shown in Fig. 3. The following vibration peaks can be observed:
The p-PD structure has two functional groups of -NH2 in the para position. Vibration peaks related to N-H traction emerged at 3409 and 3374 cm− 1. The N-H bending peak also appeared in 798 cm− 1 while the vibration peak related to C-N-H bending was observed at 1630 cm− 1.
Existing aromatic structures indicate the presence of C-H stretching vibrations in the range of 3000–3100 cm− 1. In this sample, the C-H peak appeared at 3009 cm− 1. In-plane bending vibration peaks can be also observed at 1126 and 1086 cm− 1. The peak related to off-plane bending of C-H can be detected at 719 cm− 1.
In this sample, stretching vibrations of the C-C bond can be observed at 1516 and 1445 cm− 1. However, a C-C deformation peak was observed in this sample at 514 cm− 1.
The vibration peak corresponding to the C-N stretching vibration emerged at 1340 cm− 1. The peak due to C-N-C bending vibrations can be observed at 1263 cm− 1.
The peak due to the combination of two benzene derivatives in the p-PD structure appeared at 833 cm− 1 [18].
3.5. Evaluation of CHO and p-PD monomer bond using UV-VIS spectroscopy.
Figure 4 shows the UV-vis spectrum of CHO, p-PD, and CHO-p-PD in ethanol solvent. The peak of CHO does not appear in this wavelength range; moreover, p-PD shows two common absorption peaks at ~ 245 and 337 nm, which is consistent with previous reports [6]. The first absorption peak at ~ 272 nm can be assigned to the π-π* bond of the p-PD benzene ring; while the second absorption peak at ~ 305 nm is related to the p-π conjugate bond through sharing electrons between NH2 and the benzene ring. The CHO-p-PD bond spectrum also shows two absorption peaks at ~ 237 and 278 nm. Based on Fig. 4, the link between CHO and the p-PD functional groups doubled the density of absorption as well as the red shift (displacement of the absorption peak to shorter wavelengths); suggesting the proper binding of CHO to p-PD.
3.6. Investigation of the conductivity of different layers of CuO/Pt/MIP
The ability of a surface to conduct electricity decreases with increasing electrical resistance (R). Therefore, electrical conductivity is defined as the inverse of resistance, and in other words, the current density is directly related to the amount of electrical conductivity of the surface. The electrical conductivities of different layers of the electrode are briefly described in the Supplemental Information. The cyclic voltammograms of electrodes modified with CuO, CuO/Pt and CuO/Pt/MIP were examined as shown in Fig. S3.
3.7. The effect of CHO loading duration in MIP on biosensor response
To evaluate the optimal duration for the loading CHO molecules within the polymer sites, a certain volume of the stock CHO solution (0.8 µM) was poured into 10 ml of ethanol. Then, the modified electrode was examined at different time intervals (from 20 to 240 s). Fig. S4 shows the ratio of current density to different loading times. Accordingly, the current density rose uniformly by prolonging the loading time of CHO up to 160 s. Therefore, the retention time of 160 s was chosen as the optimal time to achieve optimal CHO loading intowithininto the MIP structure.
3.8. Biosensor behavior at different scan rates
Figure 5 (A) shows the current density vs. the applied potential while Fig. 5 (B) illustrates the current density vs the scan rates. The presence of cathodic and anodic peaks indicates the reversibility of this process. Figure 5 (B) suggests a linear relationship for the electro-oxidation activity of CHO. The electrode surface coverage (Γ*) can be calculated using the slope of this diagram and Laviron theory (Eq. 1).
$${I}_{p}=\left( \frac{{n}^{2 }{F}^{2}}{4 R T}\right).V.A.{\varGamma }^{*}$$
1
In this equation, A shows the surface area of the electrode (cm2), V is the scan rate, F denotes the Faraday constant, n is the number of electrons exchanged in the redox process (n = 1). Ip represents the density of the anodic current. R and T also respectively show the global constant of gases and ambient temperature (K). The electrode surface coverage was calculated to be 4.3 × 10− 9 mol/cm2. Figure 5 (C) shows the current density of the CHO oxidation at the surface of the modified electrode. The CHO diffusion coefficient can be calculated based on the Randles–Sevcik equation (Eq. 2).
$${I}_{P}=\left(2.68\times {10}^{5}\right) {n}^{1.5}.A. {C}_{0}. {v}^{0.5}. {D}_{0}^{0.5}$$
2
In this equation, V is the potential scanning rate (V/s), C0 represents the concentration (mol/cm3), A is the electrode surface area (cm2), Ip deotes the current density (A) and n is the number of electrons exchanged. Based on this equation, the diffusion coefficient of CHO at a concentration of 0.8 µM was calculated to be 1.49 ×10− 5 cm2/s.
3.9. Direct electrochemical oxidation of CHO at different concentrations
Figure 6 (A) shows the cyclic voltammograms of the designed biosensor at different concentrations. Accordingly, the current density peak of CHO oxidation increases by enhancing the CHO concentration. This indicates the electrooxidation process of CHO at the designed nano-catalyst surface (CuO/Pt). As the CHO concentration increased, more CHO molecules were loaded on selective MIP sites and the desired nano-catalyst surface, incrementing the current density peak of CHO oxidation. Due to the strong adsorption of O2 to CuO, oxygen can be converted into O2− and then react with H2O to produce H2O2 and hydroxyl radicals (Eqs. 3, 4, and 5). Hydroxyl radicals thus initiated the cholesterol auto-oxidation to produce oxysterols (Eq. 6).
CuO + O2 + e− → CuO + O2− (3)
O2− + H2O → H2O2 + •OH (4)
O2− + H2O2 → •OH + OH- + O2 (5)
•OH + Cholesterol → Oxysterols (6)
Platinum NPs on the CF/CuO/Pt nanocomposite can enhance the electrical conductivity of the sensor, thus facilitating the cholesterol oxidation. The linear range of the designed biosensor was from 0.4 to 6 µM. As can be seen in Fig. 6 (B) the sensitivity of the biosensor can be determined from the slope of the current calibration curve in terms of concentration logarithm, which is equal to 157.85 µAµM− 1cm− 2. Eq. 7 was used to calculate the detection limit.
(7)\(LOD=3\frac{SD}{m}\)
In which m is the slope of the calibration curve line and SD denotes the standard deviation. To calculate SD, a certain concentration of the control sample was measured under the same experimental conditions with three replications (5.79%). Accordingly, the detection limit of the designed biosensor was 0.035 µM. The sensitivity and limit of detection of the designed biosensor are compared with other sensors as listed in Table 2.
Table 2
Comparison of the response of some CHO biosensors based on different types of electrodes.
Num.
|
Electrode
|
LOD
|
Linear range
|
Sensitivity
|
Ref.
|
1
|
Cu/NiCNF/ACF/PMO
|
0.002 mg dL− 1
|
0.04–600
mg dL− 1
|
0.226 µA µM− 1 cm− 2
|
[19]
|
2
|
ChOx/Au/ZnO/CNT
|
0.1 µM
|
0.1–100 µM
|
25.89 µA µM1cm− 2
|
[20]
|
3
|
Grp/β CD/Methylene Blue
|
0.1 mM
|
|
0.01 µA mM− 1cm− 2
|
[21]
|
4
|
Pt NP/(CNT)24 bilayer
|
|
0.005–10 mM
|
8.7 µA mM− 1cm− 2
|
[10]
|
5
|
Cu2S NRS
|
0.1 µM
|
0.01–6.8 mM
|
0.101 µA µM− 1 cm− 2
|
[22]
|
6
|
Co(II)Cl2/Pt
|
2 µM
|
25–200µM
|
7.32 µA mM− 1cm− 2
|
[23]
|
7
|
Cu2O NPs/TiO2
|
0.05µM
|
24.4–622 µM
|
0.603 µA µM− 1 cm− 2
|
[24]
|
8
|
PANI/MWCNTs/C
MC/CPE
|
0.01 mM
|
0.05-5 mM
|
0.1 µA µM− 1cm− 2
|
[25]
|
9
|
β-CD/MB/Gr
|
1 µM
|
1-100 µM
|
0.01 µA µM1cm− 2
|
[26]
|
10
|
CF/CuO/Pt-p-PD-CHO/MIP
|
0.035 µM
|
0.4- 6 µM
|
157.85 µA µM− 1cm− 2
|
This work
|
3.10. Biosensor stability
Figure S5 (See Supplemental Information) shows the cyclic voltammogram obtained from 25 consecutive repetitions of the cyclic voltammetry test under the mentioned conditions. As seen, after 25 consecutive cycles, a very small decrease can be detected in the current density peak, suggesting proper strength of the surface of the modified electrode against the applied potential.
3.11. Reproducibility, stability and anti-interference
To study the biosensor reproducibility, CF/CuO/Pt/p-PD Nanocomposite attached on GCE was evaluated to measure a concentration of 0.8 µM CHO in six repetitions. RSD of the designed biosensor was calculated to be 5.86% (Fig. 7 (A)). The stability of the designed biosensor over time was also evaluated; the biosensor managed to detect 4 µM of CHO at 30-day intervals. Based on Fig. 7 (B), after 30 days, the biosensor response only decreased by 3% compared to its initial response. Figure 7 (C) shows the selectivity of the designed biosensor toward CHO in the presence of some interfering species (2 µM) such as glucose (Glu), ascorbic acid (AA), uric acid (UA), and dopamine (DA). Accordingly, the electrocatalytic behavior of the biosensor is distinctive in the presence of 4 µM of CHO and other species caused minor disturbance in the current density of CHO. Therefore, it can be concluded that the designed electrode has good selectivity and can be used to monitor actual samples. Fig. S6 demonstrates a linear relationship between the current density and 1/t2. The CHO diffusion coefficient can be determined from the slope of this diagram, which is equal to 7.8× 10− 4 cm2 s− 1.
3.13. Artificial samples study
Non-enzymatic determination of cholesterol in real samples is a complex task due to the presence of different forms of cholesterol in biological matrices. For example, biological samples (mainly blood or serum) contain not only the free form of cholesterol, but also its esterified form. Therefore, esterified cholesterol should be first converted into the free state and then oxidized. However, to evaluate the practical applications of CF/CuO/Pt/CHO-p-PD biosensor, the artificial samples were collected from three healthy subjects. Artificial samples were diluted with phosphate buffer followed by adding known concentrations of cholesterol. Cholesterol solutions of different concentrations (0.8, 2 and 4 µM) were added to samples and recovery percentages were calculated. The recovery of 104.2% demonstrates the applicability of the prepared CF/CuO/Pt/CHO-p-PD biosensor to real sample analysis.