3.1. Fabrication and chemical analysis of the Cyt.c/CNT electrode
For applications within blood vessels as a smart stent, the electrode should be lightweight, flexible, and sufficiently small to avoid blood flow obstruction [19–22, 27]. Hence, the Cyt.c and CNT sheets were fabricated as threads using the biscrolling technique, which enables efficient electron and ionic conduction pathways, high cytochrome c loading, and large surface area (Fig. 1). The electrode fabricated through the biscrolling method operates with EDLC and pseudo-capacitive materials, demonstrating excellent electrochemical performance [28, 29]. The fabricated Cyt.c/CNT electrode is ultimately designed to be integrated into a smart stent system for sensing oxidative stress and subsequently storing energy through an EBFC. The cross section of the fabricated electrode was analyzed using scanning electron microscopy (SEM) and energy-dispersive spectroscopy (EDS) mapping images (Fig. S1). The mapping images show the distributions of C (white), O (green), N (blue), S (yellow), and Fe (pink) atoms, corresponding to the atomic composition of cytochrome c, throughout the CNT yarn (Fig. 2a). This indicates that the Cyt.c is uniformly distributed within the CNT yarns, creating an interconnected network through the close contact between the Cyt.c and CNT bundles. Figure 2b-e shows results of X-ray photoelectron spectroscopy (XPS), supporting the molecular integration between Cyt.c and CNTs. In the case of CNT, the asymmetric shape of the XPS spectrum is inclined toward high binding energies due to the presence of sp2 and sp3 carbon bonds (Fig. 2b). The XPS analysis of Cyt.c/CNT shows C-C, C-O, C-N, and COO bonds in the C 1s spectrum (Fig. 2c), R-NH2 bond in the N 1s spectrum (Fig. 2d), and peak corresponding to the double bond of C and O in the O 1s spectrum (Fig. 2e). This results indicate the prominent adsorption of Cyt.c onto CNT during biscrolling process and cyclic training. Remarkably, after the repetitive training of the Cyt.c/CNT electrode, the overall binding ratio of cytochrome c increased, which indicates a higher adsorption of Cyt.c with CNT (Fig. S2). Therefore, electrodes that have undergone training are expected to have high performances as an energy storage device.
3.2. Electrochemical sensing of hydrogen peroxide (H2O2) by the Cyt.c/CNT electrode
Under a reduction potential applied to the electrode comprising Cyt.c and CNT, Cyt.c (Fe3+) undergoes a reduction to Cyt.c (Fe2+) [26]. The reduced form of Cyt C (Fe2+) on the electrode surface then reduces, approaching H2O2 to H2O, while it oxidizes back to Cyt.c (Fe3+) simultaneously [26]. The electrocatalytic reaction can be described as
\(Cyt.c Heme \left({Fe}^{3+}\right)+{e}^{-} \leftrightarrow Cyt.c Heme \left({Fe}^{2+}\right),\) \(Cyt.c Heme \left({Fe}^{2+}\right)+{2H}^{+}+{H}_{2}{O}_{2} \leftrightarrow Cyt.c Heme \left({Fe}^{3+}\right)+2{H}_{2}O.\)
Figure 3a shows the electrocatalytic reaction of the Cyt.c/CNT electrode in response to the addition of H2O2 in a phosphate-buffered solution (PBS), a representative bio-simulated solution. Consequently, the cathode peak current exhibited a linear correlation with the increasing H2O2 concentration. The Cyt.c/CNT electrode exhibited a linear calibration plot with an excellent sensitivity (49.02 µAµM− 1cm− 2) and high correlation coefficient of 0.9962 in a measurement of the H2O2 concentration against the current response (Fig. 3b). Figure 3c shows the amperometric response observed when H2O2 (2 µM) was periodically added every 5 s to PBS at an applied potential of -0.5 V. Consequently, the current response to H2O2 remained constant at approximately 2.55 µA when each addition of H2O2 occurred. The steady-state current is reached within 6 seconds, indicating a short response time for the sensor and the potential for reversible cyclic behavior. These findings demonstrate the successful immobilization of cytochrome c with CNTs on the electrode, which showcases its potential application in hydrogen peroxide sensing.
To further investigate the specificity, we performed a comparative analysis of the electrode's responses to several biomolecules potentially found in the physiological environment, including dopamine (DA), ascorbic acid (AA), uric acid (UA), cysteine (Cys), and glucose (Glc). As shown in Fig. 3d, the Cyt.c/CNT electrode induced a discernible current response to hydrogen peroxide, while no significant response to DA, AA, UA, Cys, or Glc was observed. This result indicates that the Cyt.c/CNT electrode exhibits a highly selective response to H2O2, which demonstrates its ability to discriminate against other coexisting analytes.
Notably, the real-time monitoring of cellular production of H2O2 has a significant value in comprehending the biological effects of this important ROS. Thus, we attempted to perform a real-time monitoring of H2O2 levels in living cells. As shown in Fig. 3e, the addition of AA to the extracellular fluid can induce the production of H2O2 in living cells. AA in the extracellular fluid can oxidize to form an ascorbate radical (Asc•), which acts as a reducing agent by donating an electron to a protein-bound iron (Fe3+), converting it to its ferrous form (Fe2+) [30, 31]. The reduced form of iron (Fe2+) can then donate electrons to oxygen, leading to the ROS formation including superoxide (O2−) and ultimately hydrogen peroxide. However, H2O2 formed in the extracellular fluid is rapidly degraded by the enzyme catalase present in the plasma. Indeed, the Cyt.c/CNT electrode rapidly exhibited concentration-dependent currents in real time upon exposure to AA (2, 5, 10 µM) in HEK-293T cells (Fig. 3f). In contrast, when cell-free solutions or HEK-293T cells treated with catalase (200 µg mL− 1) were tested, no significant currents were observed. This suggests that the observed currents in the presence of HEK-293T cells treated with AA were specifically attributed to the presence of H2O2 derived from living cells (Fig. 2f). These findings indicate that the Cyt.c/CNT electrode has the potential to serve as an electrochemical biosensor for a real-time in-vivo detection of H2O2.
3.3. Electrochemical performance of the Cyt.c/CNT electrode as an energy storage device
The biscrolling technique was employed to apply cytochrome c onto a bare CNT, enhancing the adsorption between cytochrome c and CNT through the training process [29]. Consequently, the energy storage performance of the Cyt.c/CNT electrode steadily improved during repeated cycles, ultimately reaching an impressive capacitance of 257.95 mF/cm2 (Fig. 4a). To assess the performance of implantable Cyt.c/CNT electrodes, cyclic voltammetry measurements were performed on bare CNT and Cyt.c/CNT electrodes in various biofluids including PBS, Hank's balanced salt solution (HBSS), Dulbecco's modified Eagle's medium (DMEM), and horse serum. As a result, the Cyt.c/CNT electrodes exhibited consistent performances across different biofluids, with the capacitance of the Cyt.c-containing electrode approximately 10 times that of the bare CNT (2 mF cm− 2). Figure 4b presents the CV graph of the Cyt.c/CNT electrode as a function of the scan rate in the range of 10 to 100 mV/s. At a scan rate of 10 mV/s, the Cyt.c/CNT electrode exhibited maximum capacitances of 257.95 mF/cm2 and 79.33 mF/cm³ in terms of area and volume, respectively. Additionally, the energy density and power density are 91.71 µWh/cm2 and 2.06 mW/cm2, respectively (Fig. 4c). The galvanic charge/discharge curve of the Cyt.c/CNT electrode was measured from a current density of 0.3 mA/cm² to 1 mA/cm² in PBS (Fig. 4d). The capacitance calculated was 236.83 mF/cm2 and the shape showed a triangular pattern, showing that the Cyt.c/CNT electrode behaves as a pseudocapacitive material.
The Nyquist curve of the Cyt.c/CNT electrode exhibited an initial equivalent series resistance of 61 Ω cm− 1, a typical characteristic observed for supercapacitors (Fig. 4e). For stability testing, we conducted a repeated charging/discharging on the Cyt.c/CNT electrode for 10,000 cycles using various biofluids. At a scan rate of 100 mV/s, the Cyt.c/CNT electrode exhibited a capacitance retention of approximately 96.45% over 10,000 cycles in different fluids, including PBS, HBSS, DMEM, and horse serum, which indicates the electrode's stability and durability for long-term applications in various biological environments (Fig. 4f).
To demonstrate an in-vivo charging system in a bioelectrolyte, the Cyt.c/CNT supercapacitor was connected to an EBFC to convert chemical energy from glucose into electrical energy [24, 32]. As shown in the Fig. S3, the polarization curves indicate that the anode electrode with 0.1M glucose concentration and the cathode electrode under oxygen condition generate higher current density than those observed in PBS solution under ambient condition. This supercapacitor/EBFC hybrid system enabled a continuous charging through biocatalytic energy conversion. When the supercapacitor is charging by EBFC, the wired glucose oxidase (GOX) of the EBFC anode oxidizes glucose to gluconolactone, releasing two electrons that are passed to the supercapacitor. In the EBFC cathode, bilirubin oxidase (BOD) catalyzes the reduction of oxygen to water. The electrons are sourced from the supercapacitor (Fig. 4g). Remarkably, in the presence of a glucose concentration of 0.1 M, the EBFC-Cyt.c/CNT supercapacitor hybrid system exhibited continuous recharging capabilities through biocatalytic energy conversion. Fig.. 4h indicates the voltage charged to the supercapacitor by the EBFC that generates an open-circuit voltage (OCV) of 0.6 V in the PBS with glucose. The supercapacitor connected to EBFC was charged to 0.6 V, equal to the OCV of the EBFC. It was stabilized at 0.42 V due to phenomena such as voltage recovery and self-discharge caused by the low-power charging of EBFC. The ability to continuously recharge the supercapacitor in the presence of glucose highlights the potential of this hybrid system for self-powered bioelectronic devices or other applications that require a long-term and sustainable energy supply within biofluid environments for smart stents.