3.1. Characterization of Nanozyme-Based Cathode (Pt-Pd-Nps/PGE) and Its Electrochemical Response to ORR
In the first stage of the studies, the characterization of the cathode was carried out, and its electrocatalytic activity with the reduction of O2 was examined. It is well known that biocathodes consisting of enzymes such as bilirubin oxidase and laccase have been extensively used in developing EFCs due to their direct electron transfer ability and selectivity [26, 29–31]. These properties allow them to operate without a separation membrane to prevent cross-reactions at the electrodes [26]. Enzyme-mimicking nanomaterials called nanozymes have recently been introduced to biosensor and fuel cell studies [32–34]. In this study, PGEs modified with Pt-Pd bimetallic nanoparticles used as nanozyme have been proposed as a new cathode material in biofuel cell studies for the first time.
First, PtNps, PdNps and Pt-PdNps were modified on PGE using electrochemical methods. SEM images and EDX spectra were then recorded for unmodified and modified PGEs. SEM images (Fig. 1) show that the layered structures of unmodified PGE are homogeneously covered with nanoparticles obtained from monometal and bimetal deposition. According to EDX results (Fig. S2), it is expected to see %76.93 C, %17.68 O, %0.68 Mg, %1.53 Al, % 0.25 Ca, %2 Fe and %2.75 Si in unmodified PGE because pencil leads consist of clay and graphite [35]. In addition, 13.38% Pt in PtNps, 8.52% Pd in PdNps, and 4.27% Pt and 16.47% Pd in Pt-PdNps were found, which reflects that monometallic and bimetallic nanoparticles were successfully attached to the PGE surface.
EIS is widely used to understand electrode surface changes and resistance properties. The EIS curve consists of two main parts: the semicircle and the linear portion. These parts provide detailed information on the impedance changes of an electrode surface, which is valuable for understanding modifications of electrode surfaces and their use in electrochemical sensing applications [36]. The linear portion of impedance spectra corresponds to the processes of electron transfer that are limited by diffusion. The semicircle is closely related to high frequencies, and its diameter provides information about the charge transfer resistance (Rct), which controls the electron transfer kinetics of the redox probe at the electrode's surface [36]. Thus, unmodified and modified PGEs were also electrochemically characterized by evaluating their CVs and EIS to understand electrode surface changes and surface resistance properties. EIS spectra of unmodified Pt-Nps/PGE, Pd-Nps/PGE, and PtPd-NPs/PGE were recorded in 10.0 mM [Fe(CN)6]3-/4- containing 0.10 M KCl. The Rct values obtained from Nyquist curves were obtained as 110, 37, 18 and 8 ohms for unmodified PGE, Pd-Nps/PGE, Pt-Nps/PGE, and PtPd-NPs/PGE, respectively (Fig. S3A). This is attributed to the fact that the metal nanoparticles speed up the transfer of electrons on the surface of the PGE. CVs recorded in the redox probe for each electrode agree well with the EIS measurements. It can be seen from CVs that the highest peak currents of the redox probe were obtained from Pt-PdNps/PGE, while the lowest peak currents were obtained from unmodified PGE (Fig. S3B).
To compare the electrocatalytic activities of the electrodes (bimetallic NPs-modified PGE), ORR was performed by using unmodified PGE and monometallic modified PGEs (PtNps/PGE and PdNps/PGE) together with Pt-PdNps/PGE as cathodes. Figure 2 shows the CVs and LSVs of these electrodes in the argon and oxygen-saturated 0.1 M KCl containing 0.1 M PBS at pH 7.0. It can be clearly seen that a remarkable enhancement in the electrocatalytic reduction of O2 was observed at Pd-PtNps/PGE compared to unmodified PGE and the monometallic NPs-modified electrode. On the other hand, ORR has a high overpotential at unmodified PGE, and the current starting at 0 mV did not take the form of a peak but gradually increased by shifting potential in a more negative direction. ORR current of PtNps/PGE, PdNps/PGE, and Pt-PdNps/PGE were found to be 280, 340, and 570 µA at -50, + 50, and + 90 mV vs. Ag/AgCl, respectively. It was concluded that Pt-PdNps/PGE is the most suitable electrode for the cathode in glucose/O2 type EBFC because bimetallic nanoparticle-modified PGE has the highest peak current at a more positive voltage in the reduction potential compared to monometallic-modified PGEs.
3.2. Characterization and Electrochemical Behavior of Bioanode (MWCNT-Fc-GOx/GCE)
Direct electron transfer becomes difficult for GOx since the electroactive group is in the inner parts of the enzyme [3]. To overcome this situation, electrodes are modified with redox mediators such as ferrocene derivatives and osmium redox polymers [17, 19, 21]. Redox mediators facilitate electron transfer rates between substrate and electrode, resulting in an enhancement in electron transfer at the bioanaode surface and in output power. In this study, ferrocene was immobilized with GOx onto MWCNT/GCE to obtain an effective bioanode.
To obtain the optimum MWCNT-Fc-GOx/GCE, which gives the most effective response to glucose oxidation, the GCE surface was optimized by adjusting the different concentrations of MWCNT, Fc and GOx during the electrode preparation step. To determine the optimum electrode composition, CVs of each electrode prepared with different ratios of modification reagents were recorded in the absence and presence of 5.0 mM glucose. Fig. S4. shows the curves of the concentration of modification reagents versus the electrocatalytic oxidation peak current of glucose. The maximum peak current for the electrocatalytic oxidation of glucose was achieved when MWCNT, Fc, and GOx (concentrations of 5.0 mg/mL, 25.0 mM, and 20.0 mg/mL, respectively) were immobilized at GCE surface.
The obtained optimum MWCNT-Fc-GOx/GCE was characterized by recording EIS and CVs in the 5.0 mM Fe(CN)64-/3- redox probe containing 0.10 M KCl. The Nyquist curves were used to estimate the Rct, which was then used to evaluate the insulating properties of the electrode/electrolyte interface. Fig. S5A shows that the Rct values were found to be 1610 (Fig. S5A/c), 140 (Fig. S5A/b), and 580 (Fig. S5A/c) ohms for unmodified GCE, MWCNT-Fc-GCE and MWCNT-Fc-GOx-GCE, respectively. Rct value of unmodified GCE (1610 ohm) was remarkably decreased after the modification of both MWCNT and Fc (140 ohm) due to high surface area and conductivity of MWCNT and electrostatic interaction between positively charged Fc and negatively charged redox probe. These results indicate that the conductivity of the MWCNT-Fc/GCE, which is directly related to its electron transfer ability, is higher than that of the unmodified GCE. When the surface of MWCNT-Fc/GCE is immobilized with GOx, Rct is increased to 580 ohms due to steric hindrance between the redox probe and large GOx. In addition, there is good agreement between the CV results given in Fig. S5B and the EIS measurements. CVs from Fig S5B show that the highest peak currents of the redox probe were obtained from MWCNT-Fc/GCE, while the lowest peak currents were obtained from MWCNT-Fc-GOx/GCE. The information obtained from the voltammograms and impedance spectra indicates that MWCNT and Fc play an important role in increasing the electron transfer and conductivity properties between GCE and the Fe(CN)63-/4- redox probe. Finally, the modification procedure with MWCNT-Fc and GOx was successful.
GOx is often used as a biocatalyst to produce biosensors and EBFCs. To evaluate the efficacy of MWCNT-Fc-GOx/GCE in catalyzing glucose oxidation, a series of comparative CVs were recorded in the absence and presence of glucose in a 0.1 M PBS with pH 7.0 containing 0.1 M KCl at a scan rate of 20 mVs-1. It is expected that CV recorded by an unmodified GCE showed no anodic peaks attributed to glucose oxidation (Fig. 3A). Figure 3B/a shows that the response of MWCNT-Fc-GOx/GCE in the absence of glucose. A well-defined redox couple with a formal potential of 300 mV indicates high activity and shows a characteristic, reversible redox peak behavior of Fc, in which a single electron transfer is involved for the electro-oxidation peak current of Fc. Also, electrocatalytic oxidation of glucose was observed at the GOx immobilized on the MWCNT-Fc/GCE surface (Fig. 3B/b). The anodic peak of the MWCNT-Fc-GOx/GCE remarkably increased in the presence of 10 mM of glucose, while in the presence of the same glucose concentration, the cathodic peaks slightly decreased (Fig. 3B/b). Therefore, the significant increase in the anodic peak of the modified electrodes when exposed to glucose demonstrates the excellent oxidative electrocatalytic activity of the GOx with electrocatalyst (Fc) towards glucose. Considering that the GCE area, which is 0.07 cm², results in a high current density, fast charge transfer on the GCE proves its suitability for electrode modification with GOx, Fc, and MWCNT. CVs of MWCNT-Fc-GOx/GCE were also measured at different glucose concentrations ranging from 0 to 20 mM (Fig. 3C). The electrodes exhibited a higher oxidative current response with increasing glucose concentrations. They showed the same onset potential (approximately − 0.1 V vs. Ag/Ag+) in the CV test, which is dependent on the activation of the glucose oxidation reaction. The successful immobilization of enzymes on the electrode surface and electron transfer between the enzyme and electrode via Fc mediator is suggested to have been achieved for Fc-MWCNTs modified electrodes.
The electrocatalytic glucose oxidation at the enzymatic electrode (MWCNT-Fc-GOx/GCE) was also investigated by chronoamperometry (CA). To assess the effect of applied potential on the response to glucose, CAs were recorded for MWCNT-Fc-GOx/GCE at various potentials ranging from 0.20 to 0.5 V. This was achieved by adding 1 mM glucose successively into a continuously stirred 10 mL solution of 0.10 M pH 7.0 PBS containing 0.1 M KCl. Results show that maximum peak current was obtained when the applied potential was fixed 0.3 V vs Ag/AgCl. CAs of MWCNT-GOx/GCE and MWCNT-Fc-GOx/GCE at 0.30 V were recorded based on the successive addition of glucose in the range from 1 to 10 mM. The CAs in Fig. 4A indicate a steady-state current was recorded within approximately 5 seconds of adding glucose. Additionally, it can be observed that these steady-state currents increased as the glucose concentration increased. The MWCNT-GOx/GCE (as shown in Fig. 4A, red line) does not respond to glucose. However, the MWCNT-Fc-GOx/GCE electrode responds well to glucose. The signal recorded by the modified GCE (Fig. 4, black line) was higher than that of the MWCNT-GOx/GCE. The calibration plot suggests a linear relationship between current and glucose concentration, which varied from 1 to 10 mM during the study of MWCNT-Fc-GOx/GCE (Fig. 4B). To calculate the limit of detection (LOD), limit of quantification (LOQ) and sensitivity values of the proposed sensor, the lowest concentration that gives a amperometric response was recorded ten times. The LOD and LOQ were calculated according to the equation of 3xSD/m and 10xSD/m,(“SD”: the standard deviation for peak currents and “m”: the slope calculated from the calibration curve) and determined as 0.261 mM and 0.87 mM, respectively. Also sensitivity of the proposed sensor were calculated 25.71 µA/mM.
During the final step of anode characterization, the influence of potentially interfering compounds, such as other monosaccharides (galactose, mannose, fructose, maltose), disaccharides (sucrose), uric acid (UA), dopamine (DA), and ascorbic acid (AA), on the response to glucose was examined at MWCNT-Fc-GOx/GCE. For this purpose, the steady-state background current was stabilized under the optimized conditions. The CAs of MWCNT-Fc-GOx/GCE were recorded by adding known volumes of stock solutions of glucose and interfering compounds to an electrochemical cell. The concentrations of glucose and each interfering compound were fixed at the same concentration. Table S1 demonstrates the response of interference molecules toward glucose. The results show that the designed anode electrode was minimally affected by fructose, sucrose, maltose, mannose, galactose, glutamic acid, lactose, and sodium lactate while highly affected by UA, AA, and DA. It is well known that these substances exhibit positive interference effects by oxidizing at high anodic potentials and enhancing the glucose biosensor's signal. Our previous studies showed that a pre-oxidant (NaBiO3) was successfully used to minimize the interference of these substances [37]. Thus, solutions, including glucose and interference substances, were passed from NaBiO3 filled in an injector several times. After that, the interference effects of UA, DA, and AA were significantly reduced by their pre-oxidation with NaBiO3, while glucose was not oxidized by NaBiO3. The results indicate that MWCNT-Fc-GOx/GCE demonstrates a high selectivity towards glucose.
3.3. Fuel Cell Performance and Real Sample Analysis
Glucose was chosen as a biofuel model in our BFC development because it is one of the significant components of human biofluids, including sweat. The fuel cell performance of the system was evaluated by monitoring power output and current density dependent on OCP in air-saturated PBS containing 0.10 M KCl and various concentrations of glucose in the range from 1 to 10 mM. The maximum current density of 460 µA/cm2 and a maximum power output of 32.05 µW/cm2 were obtained at a voltage of 0.285 V (Fig. 5). The high current and power density were achieved due to the customized composite electrodes in both the anode and cathode. These electrodes possess desirable electrical conductivity and electrocatalytic properties. The hybrid MWCNT-Fc enables an efficient transfer of electrons from the anode to the cathode, significantly increasing the power output. Figure 5B demonstrates a positive correlation between glucose concentration and power density within the 0–10 mM glucose range. The response generated by the BFC was observed with successive additions of glucose in the range of 0–10 mM. The calibration plot for the self-generated current signal demonstrates linearity up to 10 mM glucose. This observation provides strong evidence for the ability of the BFC to detect glucose levels in this range.
Table 1. compares the performance of the designed glucose-based EBFC with the previously reported EBFCs. As can be seen from Table 1, the proposed glucose-based EBFC demonstrates better performance than some other previously reported EBFCs in terms of power density values. Furthermore, the designed electrode for the cathode has never been used previously for ORR. PGEs have several benefits, including disposability, affordability, mechanical rigidity, simplicity, commercial availability, and ease of surface modification [38]. Also, nanozymes have many advantages such as diverse enzyme-mimicking activities, low cost, high stability, robustness, unique surface chemistry, ease of surface tunability and biocompatibility. Therefore, our cathode design brings the advantages of PGE and nanozymes together. The mechanism of Pt-Pd-NpsPGE boosting the electrochemical reduction of oxygen can be attributed to the synergistic effect of Pt and Pd nanoparticles, which makes oxygen reduction at the cathode surface easier. Additionally, the cathode electrode can reduce oxygen to water without the need for enzymes such as BOD and Lc. The investigation was conducted to determine the performance of the designed EBFC with artificial serum samples. To attain the desired outcome, the glucose levels in artificial serum samples were measured by the LSV method, and the voltammograms for each sample were recorded under optimized conditions. At 0.280 V, a maximum current density of 290 µA/cm2 and a maximum power output of 11.35 µW/cm2 was achieved using 10 mM glucose (Fig. S6). This study has shown that designed EBFC can be effectively used for glucose determination in real and artificial samples and suggests their potential as simple to configure energy-sustaining and non-invasive glucose monitoring devices.
Table 1
Comparison of the analytical performance of proposed self-powered EBFC with some previously reported EBFCs in the literature.
BFC/analyte | Anode | Cathode | OCP (mV) | Power Density (mW/cm2) | Current Density (mA/cm2) | Reference |
Glucose/O2 | GDH/poly(TBO)/GCE | GOx/HRP/MWCNT/GE | 650.0 | 31.3 | 165.9 | [39] |
Glucose/O2 | GRE/5AP/GOx | GRE/HRP/GOx | 450.0 | 3.5 | 22.0 | [40] |
Glucose/O2 | PGE/MWCNT/GOx | PGE/MWCNT/laccase | 433.0 | 18.0 | 60.0 | [41] |
Glucose/O2 | PPD/(AuNPs)PPCA‒GOx | GRE/PB-PPCA/PPCA‒GOx | 640.0 | 10.94 | 60.52 | [5] |
Glucose/O2 | GDH-MB | GOx-hemin | 500.0 | 18.0 | 230.0 | [42] |
Glucose/O2 | PPy–GOD–CNT | PPy–CNT–laccase | - | 37.4 | - | [43] |
Glucose/O2 | MWCNT-PCA/Ca2+/GOx | rGC | 140.0 | 6.25 | - | [44] |
Glucose/O2 | Fc-Go/GC | BOD-ABTS2−/fBP | 200.0 | 26 | - | [45] |
Glucose/O2 | MWCNT-Fc-GOx/GCE | PtPd-Nps/PGE | 285.0 | 32.05 | 460.0 | In this work |
(rGC: Reduced graphene oxide-ceria, MWCNT-PCA: Multiwalled carbon nanotube-pyrene carboxylic acid, Fc: Ferrocene, GC: Glassy Carbon Electrode, BOD: Bilirubin Oxidase, ABTS: 2,2′-azinobis(3-ethylbenzothiazoline-6-sulfonate) diammonium salt, fBP: functionalized bucky paper, Pt-Pd-Nps: Pt-Pd Nanoparticles, PPy: Poly-Pyrrole, Box: Bilirubin Oxidase, GDH: Glucose Dehydrogenase, MB: Methylene blue, PPD: poly(1,10-phenanthroline-5,6-dione), PB: Prussian blue, PPCA: poly(pyrrole-2-carboxylic acid), MWCNT: Multi-walled carbon nanotube, HRP: Horseradish peroxidase) |