1 Introduction

Platinum nanomaterials are widely used as electrocatalysts in polymer electrolyte membrane fuel cells ^[1-3]. Nevertheless, one major technical challenge of Pt nanoparticles is that these zero-dimensional materials have frequently large number of lattice boundaries and defect sites, resulting in low catalyst activities ^[4]. Also, the expensiveness of Pt hinders its wide application in industries ^[3,5]. Pd-based nanostructured materials can be a promising alternative to Pt, due to its lower cost, its relatively high catalytic activity and its similar intrinsic properties to Pt such as lattice parameters and energy band structures ^[6,7].

Palladium nanoparticles (Pd NPs) play a significant role in the catalysis of a variety of organic reactions including hydrogenation, C–C coupling reactions and oxygen reduction in fuel cells ^[8,9]. It has been demonstrated that substantially smaller, uniform and well-dispersed nanoparticles are required for high-yield catalytic applications ^[10-12]. Meanwhile, due to large surface energy and electrostatic attractive force of the Pd NPs, they are easily to aggregate into bigger clusters, hence decreased catalytic activities ^[13]. Therefore, carbon-based materials such as carbon nanotubes (CNTs) and graphite are often used as supporting materials for catalytic metal nanoparticles because of their fascinating structure and electrical/mechanical properties ^[3,14-16]. In order to fix Pd NPs on the substrates, carbon materials shall be functionalized because of their hydrophobic and inert surface. However, chemical treatments on carbon materials often involve harsh experimental conditions, which may destroy the mechanical and electronic properties of CNTs and introduce defects.

Recently, biological supramolecules have attracted dramatic attention as template materials for the controlled synthesis of metallic Pd NPs as a consequence of their well-defined structures ^[17-19]. For example, in their pioneering work, Nguyen et al. fabricated Pd clusters along DNA molecules by reducing PdO precipitates with gaseous hydrogen ^[17]. Compared with DNA molecules, amyloid nanofibrils are much stable and rigid, because their internal structures comprise a cross-β sheet core with richer hydrogen bonds formed between the amino and carbonyl groups of the polypeptide chain ^[20]. Importantly, amyloid fibrils are rich in functional groups that can specifically interact with various nanoparticles. Therefore it provides another green alternative as scaffolds for nanostructures assembly ^[21,22].

In our previous work, we prepared “amyloid fibrils-Pd chip” by alternatively depositing Pd NPs and amyloid fibrils on a solid substrate for catalyzing Suzuki Coupling reactions ^[22,23]. In this paper, we describe a facile and mild approach to fabricate Pd-insulin catalyst with controlled Pd NPs size. The Pd-insulin catalyst demonstrated very nice electrocatalytic activities towards the reduction of H₂O₂.

2 Experimental

3 Results and Discussion

The INSAFs was obtained by incubating insulin peptides solution (1.0 mg/mL, pH 1.6) at 80°C for two days. The PdO NPs were obtained by aging the Na₂PdCl₄ solution (5 mM) at 25°C for two days ^[23]. After mixing with the as-prepared INSAFs, the PdO NPs were adsorbed onto the INSAFs (Fig.1a and 1b) and gradually reduce to metallic palladium probably due to the existence of Cys groups on fibrils surface as confirmed by XPS measurement (Fig.1c). Interestingly, two shoulder peaks can be seen clearly at 335.6 and 340.8 eV, which are very close to the reported values of Pd(0) species ^[17,24]. The light shift of Pd may be caused by the interaction of Pd NPs with fibrils. It is worthy to stress that, unlike previous reports ^[25,26], no external reduce agency such as NaBH₄ was introduced in our system. Here, we attribute the reduction of palladium precursor into Pd (0) to the presence of Cys residue along INSAFs surface as suggested by Kurouski et al. ^[27]. Several groups reported that palladium precursors could self-mineralize on Cys mutant TMV ^[28,29].

Fig. 1.Full-screen View

AFM images of pure INSAFs (a) and Pd NPs-coated INSAFs (b); and XPS spectra of Pd/INSAFs (c), the peaks at 335.6 and 340.8 eV suggest the presence of metallic Pd.

Quantitative detection of H₂O₂ is of great importance in recent years as it is increasingly used in food, chemical and biochemical industries. Direct electrochemical detection of H₂O₂ has been achieved via its reduction at a variety of electrode materials. Some metal or metal oxide-modified electrodes have exhibited high electrocatalytic activities towards the reduction of H₂O₂. Here, the electrocatalytic activity of Pd/INSAFs was investigated by CV. Details of experiments were described in the experimental section. Briefly, the obtained Pd/INSAFs were adsorbed on the Au-coated GCE. The modified electrodes were immersed in 0.1 M PBS, and the potential was scanned from −0.2 to 0.8 V at a scan rate of 100 mV/s to obtain CV curves. As shown in Fig. 2(a), no apparent reduction current could be observed in the presence of 0.4 mM of H₂O₂ for the INSAFs-GCE, while commercial Pd/C demonstrated a remarkable increased activity in comparison with INSAFs-GCE (Fig. 2b). For the Pd/INSAFs-GCE, we observed a strong and sharp reduction peak in the absence of H₂O₂ between −0.1~ +0.1 mV (Fig. 2c), which can be attributed to the reduction of palladium oxide ^[30]. Interestingly, with 0.4 mM H₂O₂, an obvious catalytic reduction peak was seen in even stronger intensity. The peak position of the CV curve shows characteristics similar to those of commercial Pd/C, indicating that Pd/INSAFs were effectively loaded on the GCE surface. The remarkable increase of peak intensity suggests a faster electron transfer rate and higher electrocatalytic activity toward the reduction of H₂O₂ at the Pd/INSAFs ^[30]. We further estimated the electrochemical active surface area (ECSA), which provides important information regarding the number of available active sites. As reported in Ref. [31, 32], the electrochemical active sites increase with the ECSA. The ECSA was calculated by ECSA =Q_o/0.424, where Q_o is the coulombic charge for the reduction of PdO, i.e. the surface charge obtained from the area under the CV trace, and 0.424 mC·cm⁻² charge is required for desorption of monolayer of oxygen on the Pd surface ^[32]. The Pd/INSAFs displays an outstandingly high ECSA of 13.82 m²·g⁻¹, which is much higher than that of the commercial Pd/C catalyst (3.34 m²·g⁻¹). This substantial difference indicates that Pd/INSAFs possess larger number of electrochemical active sites, which probably stems from the overall smaller size of Pd nanoparticle around INSAFs ^[33].

Fig. 2.Full-screen View

CV curves of insulin-GCE (a), commercial Pd/C-GCE (b) and Pd/INSAFs-GCE (c) in PBS buffer without (red) or with (blue) 0.4 mM H₂O₂.

Figure 3 shows typical CV curves recorded for the Pd/INSAFs in PBS solution containing 4 mM H₂O₂ at scan rates of 20−260 mV/s. The reduction current increases with the scan rate. For all the scan rates, the Pd/INSAFs exhibited excellent electrochemical performance. A plot of the cathodic peak currents versus the square root of the scan rate (ν^1/2) for Pd/INSAFs is also given in the inset of Fig. 3. The peak current is linearly proportional to ν^1/2, indicating that the diffusion limited electron transfer kinetics for the reduction of H₂O₂ at Pd/INSAFs ^[31]. Moreover, we notice that the reduction peaks shift slightly to more negative potentials at higher scan rates, suggesting that the electrochemical reduction process for H₂O₂ is irreversible ^[25].

Fig. 3.Full-screen View

CV curves of Pd/INSAFs for the catalystic reduction of 0.4mM of H₂O₂ in PBS (0.1M, pH: 7.0) at different scan rates between 20 to 260 mV·s⁻¹. The inset shows plot of cathodic peak currents vs the square root of the scan rate.

Typical CVs for Pd/INSAFs modified GCE in PBS buffer with H₂O₂ of various concentrations is shown Fig. 4. The Pd/INSAFs-GCE shows a wider linear range from ~0.2 μM to 0.8 mM for H₂O₂ ^[34,35]. When the concentration of H₂O₂ was less than 0.2 μM, the current peak gradually deviated from linear region. Thus the detection limit is estimated at 0.2 μM, which is comparable to some similar types of sensors ^[36]. This wide linear range and low detection limit are likely due to the 3D network structure of the Pd/INSAFs and the high electrocatalytic activity of Pd NPs dispersed on INSAFs ^[21].

Fig. 4.Full-screen View

The current responses of the Pd/INSAFs on different H₂O₂ concentrations (a) in PBS; and the cathodic peak current as function of the H₂O₂ concentration, the inset shows the cathodic peak current at low H₂O₂ concentration (b).

Stability of catalysts is an important factor affecting the durability of Pd-based electrocatalyst. Fig. 5(a) shows the current responses of the Pd/INSAFs after different days at room temperature. The initial current response was maintained in 15 days, suggesting a remarkable resistance of the Pd/INSAFs-GCE to environmental fouling. These advantages make Pd/INSAFs a promising material for electrochemical applications.

Fig. 5.Full-screen View

CV curves of Pd/INSAFs-GCE in PBS (0.1M, pH 7.0) on different days (a) and at different coverage of Pd/INSAFs (b).

We further investigated the effect of surface coverage of Pd/INSAFs on its electrochemical performance. Surface coverage was controlled by tuning the dipping duration of Au coated GCE in Na₂PdCl₄ solution. As expected, the reduction peak current enhanced with an increase of Pd/INSAFs coverage on GCE (Fig.5b). At low particle coverage (dipping duration: 40 minutes), the peak for the reduction of H₂O₂ at modified electrode was very broad. With an increase of dipping time, the current peak became much shaper. When the dipping duration was increased to 120 minutes, the peak intensity achieved at 63.77 μA, which was about 4 times as much as that at low coverage. This is likely due to the change of ensemble behavior of Pd/INSAFs on the electrode ^[37]. Meanwhile, the peak position was shift from 0.058 to −0.19 V, probably because the particle coverage increases the rate constant kinetically as reported by Kumar where they observed the positivity peak position shift when Pd/C was used to catalyze the H₂O₂ ^[38].

4 Conclusions

In conclusions, we have presented a simple and green approach for the synthesis of ultra-small Pd NPs supported by INSAFs. The key for the successful synthesis of Pd NPs is the use of INSAFs which poss special amino acid that can specific interaction with Pd NPs. The Pd-INSAFs are demonstrated to have very high electrocatalytic activity and stability toward the reduction of H₂O₂. We believe that this approach can be used for the development of metallic nanoparticles based ensemble electrode for a wide range electrocatalytic and electroanalytical applications.