Introduction

Direct methanol fuel cells (DMFC) are considered as one of the most promising options to solve the future energy problem, because of their high energy conversion efficiency, low pollutant emission, low operating temperature, and simplicity of handling and processing liquid fuel [1–6]. Electrocatalysts with higher activity on the methanol oxidation at room temperature are critically needed to enhance their performance for commercial applications. It is well known that platinum is the only single-component catalyst that shows a significant activity for methanol oxidation. Pure platinum, however, is readily poisoned by CO, the by-product in methanol electrooxidation. Considerable efforts have been devoted to design and synthesize Pt-based alloy catalysts with higher poison tolerance and greater activity on the methanol oxidation. Improved activity has been observed by alloying platinum with one or two other elements such as Ru [7–12], Au [13], Sn [10, 14], Ni [15, 16], Os [17], Bi [14], Pb [14], In [14], Sb [14], Mn [14], Cu [18–21], and W [22] or oxides, RuO2 [23, 24], WO x [25], CuO and NiO [26], IrO2 [27], and so on.

However, catalysts need to be supported on a solid substrate, such as carbon. Recently, carbon nanotubes, because of their large surface area, excellent conductivity, and high level of chemical stability, have been widely studied as supports for catalysts in DMFCs [8, 11, 28]. Some research groups have also suggested that carbon nanotube support produces better electrocatalytic activity on the methanol oxidation than other types of carbon support. However, in electrocatalysis, the use of support materials can create a potential problem with charge transport due to grain boundaries between the supporting materials. Therefore, the formation of a continuous network with minimal grain boundaries might be beneficial to the increase of electronic transportation efficiency. A nanoporous network of Pt metal was suggested as a solution to settle these problems [29]. A bimodal pore size (d) distribution with large (d > hundreds of nanometers) and small pores (d < tens of nanometers) for reactant accessibility and high surface area is useful for electrocatalytic nanoporous materials, respectively [30].

Porous metallic materials have received increasing attention in view of their potential applications in catalysis, chemical sensors, selective filtration, and electrochemistry [31–34]. Porous nanostructured metal can be prepared by selectively etching a single-phase binary alloy that consists of two kinds of metals with different chemical activity in appropriate corrosion conditions [35, 36].

Here, we report a simple and rapid route for the preparation of nanostructured Cu/Pt catalyst using a method based on replacement of Cu by Pt. For the copper deposition, extremely high cathodic potential was applied to gold film cathode inserted into an electrolyte with high acidity for deposition of rough Cu film. Cu/Pt catalyst was also obtained by partial replacement of metallic copper by platinum via immersion of Cu-film-coated gold film electrode in a K2PtCl6 aqueous solution. The electrocatalytic activity of the resulting modified electrode toward methanol oxidation was evaluated by cyclic voltammetry. We found that stability of the Cu/Pt catalyst depends on deposition potential of Cu film. This new approach can be considered as an effective method to prepare other bimetallic systems based on replacement of the first metal layer by more noble metal.

Experimental

Chemicals and reagents

The working electrode used throughout this work was constructed with small part of recordable disk (Memorex Pro Gold) as gold source to construct gold film electrode. CuSO4, K2PtCl6, H2SO4, HClO4, and other chemicals were of analytical grade of Fluka brand.

Measurements

Electrochemical measurements were performed using a potentiostat and galvanostat (model Autolab, PGSTAT30, Eco Chemie, Netherlands) with GPES software. All measurements were carried out in a three-electrode cell. A recordable disk was cut into smaller pieces and soaked in concentrated nitric acid to remove the protective film from its surface. Then, it was sonicated in absolute ethanol and after that water for 1 min. The obtained gold substrate was secured to the bottom of a Teflon cell with an O-ring to produce as working electrode. The apparent electrode area was 0.24 cm2. GFE was electrochemically cleaned by cycling the potential between −0.3 and 1.5 V vs. Ag|AgCl|KCl3M in 0.5 M H2SO4 at a scan rate of potential 0.1 V s−1. The cyclic voltammetry characteristics for a clean Au electrode were obtained after 3–5 cycles. A Pt wire and an Ag|AgCl|KCl3M were served as the counter electrode and reference electrode, respectively.

The morphology and elemental analysis of the fabricated Cu and Cu/Pt species were studied by Philips XL30 scanning electron microscope (SEM) fitted with an energy dispersive X-ray spectrometer (EDX).

Fabrication of nanostructured Cu/Pt catalyst on gold film electrode

A high cathodic potential (as high as −1.5 V) was applied to the cell in the electrolyte of 0.4 M CuSO4 + 1.5 M H2SO4 for 500 ms for the deposition of copper. For investigation of the effect of deposition potential on the catalytic ability of the modified electrode, various deposition potentials were employed. Deposition was performed in a stationary electrolyte solution (without stirring or N2 bubbling) at the room temperature. Cu/Pt catalyst was fabricated by simply immersing the Cu-coated gold film electrode (GFE) into an aqueous solution containing 5.0 mM K2PtCl6 and 0.1 M HClO4 for 15 min.

Results and discussion

For the copper deposition, we adopted extremely high cathodic potential in an electrolyte with high acidity. This simple but very effective technique involves electrochemical deposition of Cu, accompanying with hydrogen evolution. In this process, hydrogen bubbles serve as a dynamic template for metal deposition, resulting in a rough Cu film. The use of a highly acidic deposition bath ensures vigorous H2 evolution on the surface of newly generated copper particles during deposition, possibly creating a local turbulence of electrolyte around the copper deposits and thus disrupting continuous growth of it consequently a rough film of nanosize interconnected particles was formed.

Figure 1a depicts structures of copper film created by electrodeposition in an electrolyte of 0.4 M CuSO4 and 1.5 M H2SO4 for 500 ms. A Cu/Pt catalyst can be generated by partial galvanic replacement of the copper film by Pt. The replacement of Cu occurs by an irreversible and spontaneous redox process. Cu is oxidatively dissolved by a complex (K2PtCl6) of more noble Pt metal, which is simultaneously reduced and deposited on the Cu layer. According to this redox reaction, the amount of Pt deposited by replacement of Cu is limited to half of the Cu metallic layer, because Cu oxidation can supply two electrons per atom, while four electrons are necessary for the reduction of PtIV+ to metallic platinum (Fig. 1b).

Fig. 1
figure 1

SEM images of a Cu film fabricated by electrodeposition in an electrolyte of 0.4 M CuSO4 and 1.5 M H2SO4 at an applied cell potential of −1.5 V vs. Ag|AgCl|KCl3M for 0.5 s and b Cu/Pt-modified GFE perpetrated by immersing the Cu-film-modified GFE in an aqueous solution containing 5.0 mM K2PtCl4 and 0.1 M HClO4 for 15 min

Full size image

During Cu film deposition, hydrogen evolution also happens strongly. So, it is impossible to make estimates about the amount of Cu deposited from the charge passed during deposition process. To estimate the amount of Cu deposited on the GFE surface (ΓCu) and consequently loaded Pt, stripping voltammetry was used (Fig. S1A). The stripping voltammogram recorded from −0.3 to 0.9 V in 0.5 M H2SO4 shows good agreement with cyclic voltammogram of CuSO4 at the surface of gold electrode in a H2SO4 solution, which has been reported in the literature [37]. ΓCu estimated by stripping voltammetry in 0.5 M H2SO4 is about 1.25 × 10−6 mol cm−2 (0.08 mg cm−2). Assuming that all of the Cu atoms are galvanically replaced by Pt, resulted in 0.12 mg Pt cm−2 as maximum possible loaded Pt. To find the time that replacement process reaches to equilibrium, the replacement process was monitored by recording open circuit potential (OCP) (Fig. S1B). The Cu-film-coated GFE was soaked in 0.1 M HClO4, and potential in open circuit voltage was recorded. When cell potential reached to constant value, the certain amount of K2PtCl6 was injected to concentration of K2PtCl6 in solution which reached 5.0 mM. Potential was controlled by Cu2+/Cu0 redox couple and enhanced with reaction progress and increase the concentration of prepared Cu2+ in solution. It seems the equilibrium was attained within about 15 min.

Energy dispersive X-ray (EDX) spectroscopy was used to characterize the surface composition of the obtained Cu and Cu/Pt species structures before and after replacement by Pt. Elemental analysis data derived from EDX spectroscopy before and after replacement of Cu film by Pt are listed in Table 1. Data from EDX show the amount of loaded Pt is only about 12 % Wt of total weight. The amounts of remained Cu and loaded Pt calculated to be 0.05 and 0.042 mg cm−2, respectively, which show that about one-third of the deposited Cu galvanically was replaced by Pt. Obtained results also confirm that the galvanic replacement method is a surface-limited process.

Table 1 Elemental analysis data given from EDX spectroscopy before and after replacement of Cu-film by Pt
Full size table

For this type of potentiostatic electrodeposition, the applied potential on the working electrode (GFE) plays an important role both thermodynamically and kinetically [38, 39]. So, the copper deposition potential can affect copper film morphology and consequently its effective surface area, electrochemical, and catalytic properties. The effect of Cu film deposition potential on electrochemical behavior of the catalyst was investigated. Figure 2 shows cyclic voltammograms of prepared modified electrodes under different Cu deposition potentials in methanol-free solution. Although well-accepted cyclic voltammogram of pure Pt in H2SO4 usually exhibits broad or invisible oxidation peak in positive and sharp reduction peak in negative direction [40], cyclic voltammograms of Cu/Pt-modified GFEs display a pair of reversible peaks around 0.45 V in the methanol-free solution. The reversible behavior may occur since remained Cu oxidation signal overlaps with Pt oxidation peak, and additionally, PtO electrochemical reduction signal is suppressed as the result of the presence of Cu atoms on the surface of modified electrode. Same reversible behavior has been reported for some bimetallic alloys, such as Pt/Ru [7, 10] and Pt/Sn [10].

Fig. 2
figure 2

Cyclic voltammograms of the Cu/Pt-modified GFE with various Cu layer deposition potentials a −3.5 V, b −2.5 V, and c −1.5 V vs. Ag|AgCl|KCl3M, in 0.5 M H2SO4 solution at scan rate 10 mV s−1

Full size image

The real surface area of the Cu/Pt catalysts can be estimated from the integrated charge in the hydrogen adsorption/desorption region of the cyclic voltammogram. The real surface area (A r) can be obtained from the following equation:

$$A_{\text{r}} = \, Q_{\text{H}} /0. 2 1 {\text{ mC cm}}^{ - 2},$$

where 0.21 mC cm−2 stands for the charge/cm2 of Pt with a monolayer adsorption of hydrogen. Similarly, the roughness factor (R f) can be estimated by dividing A r to geometric area, A g of the electrode. Table 2 summarizes surface parameters for modified electrodes prepared under different Cu deposition potentials. Results show that the Cu deposition potential does not have a significant effect on the active surface area and consequently on the roughness factor. However, it seems in more negative potential the loading and thickness of Cu film are increased as it reflected in a sharp oxidation signal around 0.45 (Fig. 2, curve a). But as the galvanic replacement is a surface-limited method, the amount of loaded Pt and consequently active surface area do not show significant changes.

Table 2 Surface parameters for modified electrodes prepared under different Cu deposition potentials
Full size table

The electrochemically active surface area (ECSA), measured on nanostructured Cu/Pt film using calculated Pt loading and hydrogen adsorption voltammograms, was 1190 cm2/mgPt, which is comparable with those of other nanostructured bimetallic catalysts reported by other research groups (see Table 3). It may be attributed to the unique structure of rough Cu layer and ultrafine Pt film thickness, which allow most of the Pt surface to be accessible.

Table 3 Comparison of the efficiency of some modified electrodes used for electrocatalysis of methanol oxidation reaction
Full size table

In the case of bimetal catalysts for methanol oxidation, the activity of the catalyst can be described according to the bifunctional mechanism. The Pt and second atoms, here Cu, independently play their role in catalysis of methanol oxidation. This suggests that it would not be critical where bimetal catalyst is present in the form of alloy [41], bimetallic compound [18], layer-by-layer electrodeposited [19] or just absorbed [42]. In this regard, the geometric distribution and composition of bimetal catalyst play crucial role [43].

Figure 3 shows the cyclic voltammogram of 1.0 M methanol at the surface of Cu/Pt-modified GFE in 0.5 M H2SO4 solution. As can be seen in Table 3, this modified electrode shows good electrocatalytic activity in both terms of current density and mass activity toward methanol oxidation reaction comparable with those of other nanostructured bimetallic catalysts.

Fig. 3
figure 3

Cyclic voltammogram of the Cu/Pt-modified GFE in 0.5 M H2SO4 solution containing 0.1 M methanol at scan rate 50 mV s−1. The Pt loading was 0.042 mg cm−2

Full size image

The effect of Cu deposition potential also was studied on the stability of the catalyst for the oxidation of methanol. It is known that the signal of methanol oxidation on polycrystalline platinum electrode decreases as the number of cycles increases, because methanol is irreversibly adsorbed and dissociated to form linear and multibonded carbon monoxide [44]. The dependence of the electrochemical response with the number of cycles for prepared modified electrodes under different Cu deposition potentials is shown in Fig. 4. Response of all modified electrodes dropped in second cycle. However, the stability of Cu/Pt-modified electrodes was improved with decreasing of Cu deposition potential and reached complete stability after second cycle for modified electrode prepared via Cu deposition potential at −1.5 V vs. Ag|AgCl|KCl3M. Although the modified electrode prepared via Cu deposition potential at −3.5 V showed higher methanol oxidation current in first cycle, the methanol oxidation current in 10th cycle was lower than of the modified electrode prepared via Cu deposition potential at −1.5 V.

Fig. 4
figure 4

10 consecutive cyclic voltammograms of the Cu/Pt-modified GFE with various Cu layer deposition potentials a −3.5 V, b −2.5 V, and c −1.5 V vs. Ag|AgCl|KCl3M, in 0.5 M H2SO4 solution containing 0.5 M methanol at scan rate 50 mV s−1

Full size image

The stability of Cu/Pt-modified GFE for methanol oxidation was also investigated with chronoamperometric study in 0.5 M methanol + 0.5 M H2SO4 (Fig. S2). These results demonstrated that Cu/Pt-modified GFE prepared via Cu deposition potential at −1.5 V has lower initial current but shows the slowly decay and current reached to 20 % of the initial value even though the oxidation reaction achieved the steady state.

Figure 5 shows the potential range dependence of a Cu/Pt-modified GFE in 0.5 M H2SO4 solution containing 0.5 M CH3OH. The reverse peak corresponding to re-initiating oxidation of methanol becomes smaller and its potential shifts negatively with gradually increasing positive potential limiting. The formation of PtO was decreased with decreasing switching potential, consequently the electrode surface maintained relatively clean.

Fig. 5
figure 5

Cyclic voltammograms of the Cu/Pt-modified GFE in 0.5 M H2SO4 solution containing 0.5 M methanol with different positive potential limits a 800, b 850, c 900, d 950, and e 1000 mV vs. Ag|AgCl|KCl3M at scan rate 50 mV s−1

Full size image

Figure 6 shows the effect of methanol concentration on the anodic current of methanol oxidation at the surface of this modified electrode. The anodic peak was increased with increasing methanol concentrations from 0.10 to 1.75 M and then decline. The slope of changes was decreased with increasing of methanol concentration. This phenomenon is ascribed to the saturation of active sites at the electrode surface and catalyst poisonous in high concentrations.

Fig. 6
figure 6

a Cyclic voltammograms of the Cu/Pt-modified GFE in 0.5 M H2SO4 solution containing various concentrations of methanol a 0.05, b 0.2, c 0.3, d 0.6, e 0.7, f 1, g 1.25, h 1.5, i 1.75, and j 2 M at scan rate 50 mV s−1. b Plot of the anodic peak current as a function of methanol concentration

Full size image

Three elementary reactions were recommended for the electrocatalysis of the methanol oxidation reaction (MOR) on a Pt electrode under acidic conditions [45].

$${\text{Dehydrogenation reaction: Pt-}{\rm CH}_{ 3} {\text{OH}}\;\rightarrow {\text{Pt-}}{\rm CO} + 4 {\rm H}^{ + } + 4 {\rm e}}^{ - }$$
(1)
$${\text{Water dissociation reaction: Pt-}}{\rm H}_{ 2} {\text{O}}\rightarrow \;{\text{Pt-}}{\rm OH} + {\rm H}^{ + } + {\rm e}^{ - }$$
(2)
$${\text{CO oxidation reaction:}}\;\,{\text{Pt-}}{\rm CO} + {\text{Pt-}}{\rm OH}\; \rightarrow \;{\text{CO}}_{ 2} + {\rm H}^{ + } + {\rm e}^{ - }$$
(3)

Methanol molecules are adsorbed on the Pt sites and further undergo a stepwise dehydrogenation. Dehydrogenation reaction is demonstrated to be not a rate determining step along the MOR pathway within the potential range from 0.0 to 1.0 V vs. MOR applied on the bulk Pt catalysts and with low methanol concentration of 5.0 mM even at room temperature [45], which is consistent with the theoretical calculations [46]. Following the dehydrogenation reaction, the active sites of Pt catalysts are poisoned by the strongly adsorbed CO (COads), as shown in Eq. (1), which eventually results in slow reaction kinetics. When a higher potential is applied (>0.60 V), the adsorbed hydroxyl groups (OHads) are formed by the water dissociation on the Pt catalysts, as shown in Eq. (2). Once OHads is formed on Pt sites, they will desorb CO from Pt involving the reaction between COads and OHads species via Langmuir–Hinshelwood mechanism as indicated in Eq. (3) [47].

The effect of the potential scan rates, from 0.005 to 1.0 Vs−1, was investigated on the voltammetric behavior of methanol oxidation at the Cu/Pt-modified GFE (Fig. 7a). The forward peak currents were found to depend linearly on the square root of the scan rate (Fig. 7b), suggesting methanol diffusion-controlled process as has been reported previously [48].

Fig. 7
figure 7

a Cyclic voltammograms of the Cu/Pt-modified GFE in 0.5 M H2SO4 solution containing 1.0 M methanol at various scan rates a 5, b 10, c 25, d 50, e 100, f 200, g 300, h 400, i 500, j 600, and k 800 mV s−1. b Plot of the anodic peak current as a function of ν 1/2

Full size image

Conclusion

In this article, we have developed a method to prepare a nanostructured Cu/Pt catalyst. After electrochemical deposition of copper film, bimetallic modified electrodes based on Cu and Pt were prepared by partial galvanic replacement of copper film by platinum. The Cu/Pt-modified GFE exhibits remarkable electrocatalytic activity for the MOR at much more positive potentials in acid media. This method is very simple and allows for easy preparation. This procedure can be readily extended to other bimetallic species using a more noble metal for the replacement of the first metal layer.