A facile one-step electrochemical preparation of graphene–Pd nanocomposite as a catalyst for hydrogen evolution reaction

INTRODUCTION

Hydrogen is considered a promising source of green energy owing to its high energy density, environmental friendliness, safety, and cleanliness [1, 2]. To facilitate the production of hydrogen, cathode electrode materials with favorable properties such as high electrocatalytic activity (to reduce overpotential), a sizable active surface area, moderate electrochemical stability, strong electrical conductivity, and ease of use are required [3]. Although Pt or Pt based catalysts are effective HER catalysts, their high cost and scarcity limit their widespread application [4]. On the other hand, Pd metal can serve as a good substitute for Pt due to its relative abundance in the Earth's crust and lower price [5]. Graphene, a two-dimensional honeycomb lattice of carbon atoms, is a fundamental structural component of several carbon allotropes, such as graphite, carbon nanotubes, and fullerenes [6]. Due to its large surface area, aspect ratio, tensile strength, thermal and electrical conductivity, flexibility, and transparency, graphene is also regarded as the best carbon nanofiller when compared to other traditional nanofillers [7, 8].

Metal nanoparticles can be supported by using graphene or reduced graphene oxide (rGO). Due to their size-dependent properties, metal NPs are of tremendous scientific interest, and their efficacy in catalysis can be improved by carefully regulating their size and shape [9-11]. During the synthesis of metal nanoparticles (MNPs), the uncontrolled development and agglomeration of large particles can result in the loss of their dispersibility and other beneficial features [12]. Therefore, the synthesis of MNPs with controllable size, homogeneous morphology, good crystallinity, and high dispersion is crucial. Reduced graphene oxide (rGO) is expected to be the best catalyst carrier for this purpose because it provides a good 2D support for nucleating and anchoring metal NPs on the edges and surface [13-15]. Due to its crinkly structure and the oxygen atoms, it managed to hold onto its film and decrease NP aggregation. Electrodeposition is the most reliable and controllable technology for the synthesis of metal NPs due to its straightforward process, capacity to regulate the size and shape of the nanoparticles, and its relatively short completion time [16-18]. GO can be reduced through both chemical and electrochemical processes. Chemical reducing agents, such as hydrazine hydrate [19], sodium borohydride [20], nascent hydrogen [21], sulfur-containing compounds [22], and ascorbic acid [23], are frequently used in the production of graphene. However, the use of reducing agents poses several drawbacks such as prolonged thermal process for the deoxygenation, toxicity of the chemical reductants, and extremely harsh reaction conditions. Furthermore, such agents may adversely impact the fundamental properties of graphene and result in time-consuming and complicated processes. Moreover, some epoxide groups on GO cannot be entirely reduced using these procedures, which would diminish the effectiveness of electron transport and weaken the electrochemical activity. Therefore, electrochemical procedures have been carried out without the use of any reducing chemicals. These methods have recently gained significant attention because of their quick, environmentally friendly, controlled, strong, and effective way of removing oxygenation fault spots from GO and enhancing their electrical characteristics.

In this study, a one-step and simple electrochemical technique was employed to prepare rGO and PdNPs on CPE. The resulting nanocomposites (Pd-rGO) were analyzed using scanning electron microscopy (SEM), cyclic voltammetry (CV), and chronocoulometry (CC) techniques. The density and morphology of PdNPs on the surface of graphene appear to be markedly different from those that were simply deposited on the bare CPE surface (Pd/CPE). Moreover, the created Pd-rGO nanocomposites exhibit an effective electrocatalytic activity on HER. Along with having high electrocatalytic activity, Pd-rGO is also more stable in acidic environments than Pd/CPE. This is caused by the interpenetrating network that rGO creates to immobilize PdNPs.

EXPERIMENTAL

Reagents and apparatus

A modified Hummer's process was used for making the graphene oxide (GO), and PdCl₂ was extracted from Fluka. All other compounds were used without purification since they were all of analytical grade. Deionized water was used to make all stock and buffer solutions, and experiments were conducted at room temperature. An Autolab type III was employed to carry out electrodeposition, CV, and CC, utilizing GPES version 4.9 operating system. In this investigation, a three-electrode setup was employed, consisting of an Ag|AgCl| 3 M KCl reference electrode, a Pt wire as a counter electrode, and a bare or modified CPE as the working electrode. A Teflon cylinder (with a 2.0 mm inside diameter) filled with carbon paste served as the body of the working electrode, and electrical contact was established by inserting a stainless steel rod into the Teflon tube containing the carbon paste. The morphology of the nanoparticle-modified electrodes was analyzed using a Leica Cambridge model S 360 SEM with a 25 keV accelerating voltage. Ultra-pure N₂ was utilized to deaerate the solutions for at least 15 min before each run, and all tests were conducted under a N₂ atmosphere after degassing the solutions.

Electrode preparation

After preparing the carbon paste by combining graphite powder and paraffin oil (70:30 w/w), a portion of it was firmly packed into the cavity (with a 2.0 mm diameter) of a Teflon tube. To establish electrical connection, a stainless steel rod was inserted into the tube containing the carbon paste. Polishing paper was used to polish and smooth the surface of the electrodes.

A yellow-brown dispersion was created by ultrasonically preparing 1 mg mL^-1 GO aqueous solution for 60 minutes. To create the GO/CPE, by casting the CPE surface with 10 µL of GO suspension and allowing it to be air-dried.

The electrochemically reduced GO (rGO) modified CPE (rGO/CPE) was synthesized by repeatedly scanning a CV (10 cycles) from 0 to -1.5 V at a scan rate of 50 mV s^-1.

For the PdNPs modified rGO/CPE (rGO-Pd/CPE), the electrochemical codeposition was performed in a dearated PBS solution containing 1mM PdCl₂. The electrochemical deposition of PdNPs modified CPE (Pd/CPE) was accomplished in a dearated PBS containing 1mM PdCl₂.

RESULTS AND DISCUSSION

SEM image

Fig. 1 illustrates the surface morphology of the rGO-Pd and Pd on the CPE. As shown in Fig. 1A, rGO serves as a scaffold wrinkled sheet, anchoring the nucleation and loading of PdNPs, and preventing their aggregation. Conversely, PdNPs tends to accumulate on bare CPE, as depicted in Fig.1B. Compared to the plain CPE, PdNPs generate a structure on rGO that is relatively homogeneous, highly distributed, and of high density.

Electrochemical characterization of different modified electrode

Fig. 2 displays the CVs of electrochemical reduction of several customized electrodes in a phosphate buffer solution (pH=4) at a scan rate of 50 mV s^-1. The electrochemical reduction of GO on CPE is depicted in Fig. 2a. The appearance of a substantial cathodic peak current at approximately -1.2 V demonstrates the rapid and irreversible electrochemical reduction of oxygen-containing groups on GO [24-26]. In each subsequent cycle, the reduction peak significantly diminishes until it vanishes entirely in the fifth cycle [26-30].  As shown in Figs. 2b and 2c, PdCl₂ undergoes electrochemical reduction, and GO and PdCl₂ are simultaneously deposited on CPE. Fig. 2c displays clearly a distinct CV compared to that in Figs. 2b and 2a. Although the cathodic peak at -1.2 V persists due to the GO reduction, the reductive currents are noticeably greater than those in either the GO or Pd reduction, indicating the co-reduction of the GO and Pd. The reductive currents that show a decrease in Pd are also present in  subsequent cycles, especially after the fifth cycle.

The electrochemical properties of the unmodified and modified electrodes were examined in a 5 mM solution of K₃Fe(CN)₆,. Since the surface chemistry of carbon-based electrodes is sensitive to the redox probe Fe(CN)₆^3-/4- [31], it was utilized to assess the graphene electrode's ability to transfer charges.

The CVs of both unmodified and modified electrodes in a solution containing 5mM K₃Fe(CN)₆ and 0.1 M KCl are shown in Fig. 3. The results indicate that the rGO-modified CPE exhibited a significantly lower peak potential separation and a higher peak current compared to the unmodified CPE. This suggests that the reactive edge defects on graphene and the increased electroactive surface area of electrode 20 treated with rGO led to an accelerated electron transport [32, 33]. In addition, the improved current of rGO-Pd/CPE, compared to rGO/CPE, serves as evidence of PdNPs' function in facilitating electron transport and boosting the electroactive surface area. The high-density loading of PdNPs resulted from the electrochemical co-reduction of Pd and GO.

Effect of the thickness of the graphene-Pd nanocomposite film

One of the most important controlling factors in the experiment was the thickness of the graphene and Pd nanocomposite layer. Both GO and PdCl₂ were reduced during the cathodic sweep of CV. In Fig. 4 illustrates a link between the proton reduction current at -0.6 V and the number of cycles (N). As demonstrated, a significant increase in the proton reduction current was observed by increasing N up to 10. However, as the loading amount of deposits was increased further, the current began to decline, indicating that the nanocomposite film had become thick and the electron transfer rate had decreased. As a result, ten cycles were deemed to be the optimal number of prospective cycles for the remaining experiments.

Real surface area determination

It is crucial to understand the actual surface area of the electrode, which typically surpasses the geometric area, while studying solid electrodes and electrocatalysis. To this end, Trasatti and Petrii [34] go into great length regarding the methods for ascertaining the actual surface area. In the present investigation, the charges of the oxide reduction peaks are a viable measure to determine the actual surface area [35]. Based on the following equation, the actual surface area of the electrodes enhanced with nanoparticles was determined:

S=Q₀/q₀

Where Q₀ is the surface charge that can be calculated from the region under the oxygen desorption CV trace, and q₀ represents the charge necessary for the desorption of an oxygen monolayer on a smooth Au or Pd surface. The value of q₀ for the smooth Pd electrode is 424 cm^-2 [36].

The actual surface area (S) for the Pd/CPE and rGO-Pd/CPE were calculated to be  0.07 cm² and 0.25 cm², respectively. Based on the findings, the high specific surface area of rGO significantly contributed to the effective surface area of rGO-Pd/CPE, which is 3.6 times more than that of Pd/CPE.

Electrocatalytic activity of the different modified electrodes for HER

Fig. 5 illustrates the comparative CVs of various electrodes in 0.5 M H₂SO₄ for HER. It can be seen that all of the modified electrodes exhibit electrocatalytic activity for HER when compared to bare electrodes. The reduction potentials of hydrogen generation were reduced to lower potentials than those of rGO/CPE. Moreover, both Pd/CPE and rGO-Pd/CPE exhibited strong current at the same potential. The findings indicate that the maximum electrocatalytic activity for HER is shown by rGO-Pd/CPE among the various electrodes. Thus, the electrochemical activity for HER can be significantly increased by adding Pd to rGO. Table 1 shows the potential of individual electrodes measured at a current value of -0.5 mA (E_-0.5), as well as the magnitude of the current ratio of modified to naked electrodes at switching potential (E=-0.6 mV). The data presented in this table demonstrate that the values of E_-0.5 significantly drop from bare to modified electrodes. Notably, the hybrid catalyst has the lowest overpotential in comparison to other electrodes, suggesting better electrocatalytic activity for HER. At -0.6 mV, the I_m/I_b values for CPE-rGo, CPE-Pd, and CPE-rGoPd were approximately 2, 8, and 20 times greater than those for the bare electrode, respectively. Additionally, as shown in Fig. 5, it is obvious that the onset potential of HER at rGo-Pd is higher than that of the other electrodes, indicating its superior electrocatalytic activity towards HER.

Kinetic parameters and mechanism

To assess the electrocatalytic activities of the electrodes, steady-state polarization curves of the HER at bare and modified electrodes were determined. The related Tafel plots (log J vs. η) can be seen in Fig. 6. A scan rate of 2 mV s^-1 was used to record the polarization curves in 0.5 M H₂SO₄.

As is shown in Fig. 6, the curves follow a typical Tafel behavior based on the following equation:

where b is the Tafel slope,  is the overpotential, and J is the steady-state current density and Jo is the exchange current density. Tafel plot extrapolation to the intercept yields log. The graph's slopes can be used to determine the Jo and Tafel slope.

As can be seen, rGO-Pd/CPE increased electrocatalytic activity in the following order: Pd/CPE>rGO/CPE>CPE. This demonstrates an excellent agreement between the findings of both investigations and CV experiments. Further, the rGO-Pd/CPE has the strongest electrocatalytic impact for HER.

Table 2 lists the Tafel parameters for HER on unmodified and modified CPEs. As can be seen, the rGO-Pd/CPE electrode exhibits stronger catalytic efficacy towards HER than the other modified electrodes, despite having a significantly higher J_o (0.25×10^-3 mA cm^-2).

The HER in acidic media has three potential reaction stages, as reported by study [37]. The rate-limiting step, also known as rate determining step or rds, determines the Tafel slope, which is an inherent feature of a catalyst.

To elucidate the basic procedures, it is necessary to determine and interpret the Tafel slope. For multi-step reactions, the value of the charge-transfer coefficient depends on the rds [38]. According to the general model for the HER mechanism, when the value of α is 0.5 and the Tafel slope shifts to 120 mV dec^-1, the Volmer reaction controls the HER. In this case, the α is approximately 0.5 on the modified electrodes and their Tafel slopes are near 120 mv dec^-1 (see Table 2). Therefore, it can be concluded that the Volmer step must regulate the HER on the modified electrodes.

Stability and renewability studies

The stability of rGO-Pd/CPE and Pd/CPE was assessed by using CA and CV techniques. To explore their stability, 100 repetitive CV cycles of the rGO-Pd/CPE and Pd/CPE were recorded by repeatedly scanning in a solution of 0.5 M H₂SO₄ from 0.0 to -0.5 V at a scan rate of 50 mV s^-1. The results demonstrate that, as opposed to Pd/CPE, rGO-Pd/CPE significantly reduces the electrocatalytic current of HER at switching potential. Specifically, after 50 cycles, the current diminished by up to 6% and 51%, and after 100 cycles up to 18% and 68% of the initial current value for rGO-Pd/CPE and Pd/CPE. This indicates that the graphene-Pd hybrid has good stability towards HER. It should be noted, however, that variations in voltamogarm resulting from subsequent cycles are not implausible [39].   

CONCLUSION

A one-step, facile, and rapid electrochemical method was used in the co-deposition of GO and PdNPs onto CPE. The resulting rGO-Pd/CPE was characterized using SEM, CV and CC techniques. PdNPs exhibited a fairly homogenous and high dispersed structure with high density on the rGO compared to the bare CPE. The rGO-Pd/CPE showed superior electrocatalytic activity toward HER compared to other electrodes. The kinetic parameters, including transfer coefficient and exchange current density, as well as the HER mechanism were determined from the Tafel plot. Furthermore, the rGO-Pd/CPE was renewable and demonstrated good stability for HER.

ACKNOWLEDGMENT

The authors express their gratitude to Shiraz University Research Council for the support of this work.

CONFLICT OF INTEREST

The authors declare no conflicts of interest.