1 Introduction

The exploitation of suitable electrocatalysts with high catalytic activity for oxygen evolution reaction (OER) is a challenging and important topic for hydrogen production from water splitting [1, 2]. Currently, precious metal-based materials, such as IrO2 [3], RuO2 [4] and Pd [5] are believed to be state-of-the-art electrocatalysts for OER, which can efficiently drive the OER with low overpotentials. Nevertheless, the industrial-scale applications are constrained by their scarcity and high cost. Hence, numerous efforts have been devoted to fabricate efficient and stable OER catalysts based on earth-abundant metal.

As an abundant and inexpensive metal, copper has a wide range of oxidation states (Cu0, Cu+, Cu2+ and Cu3+) which makes it suitable for many potential applications [6], including organic transformation [7], electrocatalysis [8], and photocatalysis [9]. Recently, a lot of attention was paid to homogeneous Cu-based catalysts for OER due to their attractive electrocatalytic characteristics [10, 11]. Very recently, heterogeneous Cu-based hydroxides and oxides were also demonstrated as active composites for OER in alkaline media [12,13,14,15,16]. Therefore, there is an increasing interest in the fabrication of Cu-based nanostructures as efficient OER catalysts [17,18,19,20,21,22,23,24,25]. For example, Pawar et al. reported a two-dimensional CuO nanosheet film electrode, and the excellent OER performance in alkaline solution was realized after a heat treatment [17]. Among the fabricated Cu based catalysts, Cu nanowires achieved remarkable progress in OER due to their high electrical conductivity from the unique structure [18, 19]. Meanwhile, three dimensional porous architectures are often created to further improve the catalytic efficiency, by offering tremendous specific surface area and availability for OER [20,21,22]. Based on the hydrogen bubble dynamic template method, a series of porous Cu-based nanostructures, including dendritic Cu oxide [20], Cu2O–Cu hybrid foams [21], CuO/C hollow shell@Cu dendrites [22], have been fabricated and their enhanced OER activities have been confirmed.

In the past decade, chemical/electrochemical dealloying has been emerging as a promising approach to fabricate free-standing porous materials for electrocatalysis [26,27,28,29]. Dealloying ribbon or film precursor often generates a free-standing porous structure extending in three dimensions, which is highly desirable for free transport of medium molecules. In addition, a free-standing catalyst can largely minimize the complexity of fabrication process and completely eliminate the resistance between the support and the active ingredient. Adopting this dealloying strategy, Ding et al. fabricated a series of metal electrocatalysts with free-standing porous structures and their excellent catalytic performances towards many important electrochemical reactions were demonstrated [30,31,32]. Recently, by carefully designing the alloy precursor and electrochemical dealloying potential, Chen et al. fabricated a nanoporous Co/Fe phosphides with outstanding activity towards both hydrogen evolution and OER [33].

Inspired by the reported free-standing materials as mentioned above, here, we report a free-standing porous Cu-based nanowires (p-CuNWs) fabricated by a combined method of chemical dealloying and electrochemical treatment. The as-fabricated p-CuNWs exhibit excellent activity and offer large current density for OER at a low overpotential in alkaline solution.

2 Experimental

2.1 Fabrication of p-CuNWs

The fabrication procedure of p-CuNWs is illustrated in Fig. 1a. The precursor Mg72Cu28 master alloy with nominal compositions were fabricated by arc melting the mixtures of pure Mg (99.99%) and Cu (99.99%) under a high-purity argon gas atmosphere. Subsequently, a melt spinning technique was introduced to re-melted the as-prepared precursor alloy, followed by injection onto a spinning copper roller to obtain the alloy ribbon. The resulting Mg72Cu28 ribbon is of thickness ~ 20 μm. The p-Cu was made by dealloying Cu28Mg72 ribbon in a 0.68 M HCl solution at 25 °C for 60 s. The p-Cu samples were rinsed in pure water (18.2 MΩ cm) to remove the residual chemical ion impurities. The p-CuNWs were then fabricated from the anode oxidation of p-Cu in 1.0 M KOH on a CHI 660E electrochemical workstation, by performing a linear sweep voltammetry (LSV) from − 0.6 to 0.8 V versus Ag/AgCl with a scan rate of 1 mV/s. The p-Cu was directly used as the working electrode. An Ag/AgCl (sat. KCl) electrode and a Pt wire were employed as the reference electrode and counter electrode, respectively.

Fig. 1
figure 1

a Schematic description of the fabrication process of p-CuNWs. b XRD pattern of p-CuNWs, p-Cu and Cu28Mg72 ribbon. c SEM image of p-Cu. d, e SEM images of p-CuNWs with different magnification

Full size image

2.2 Characterizations

X-ray diffraction (XRD) pattern analysis of as-prepared materials were recorded on an X’Pert Pro MPD (PANalytical, The Netherlands) X-ray diffractometer at room temperature using Cu Kα (λ = 0.154060 nm) as radiation source (40 kV and 40 mA). The structure and morphology analysis of the p-Cu and p-CuNWs were conducted using scanning electron microscopy (SEM, FEI, Inspect F50) at 10 kV accelerating voltage. X-ray photoelectron spectroscopy (XPS) was performed on a Thermo Fisher Scientific ESCALAB 250 photoelectron spectrometer (Al Kα radiation, λ = 1486.6 eV).

2.3 Electrochemical Measurement

All electrochemical measurements were performed at room temperature in 1.0 M KOH electrolyte on a CHI 660E electrochemical workstation using a standard three-electrode cell, consisting of as-prepared p-CuNWs or Mg72Cu28 as the working electrode, a Pt wire as the counter electrode, and an Ag/AgCl (sat. KCl) as the reference electrode. The OER activities of the samples were determined by LSV with a scan rate of 1 mV s−1. A 90%-iR correction was applied to all of the original data, except when specified otherwise. The potentials reported in the text were converted to the reversible hydrogen electrode (RHE) potential according to the equation E(RHE) = E(Ag/AgCl) + 0.059 × pH + 0.197 V. The OER overpotentials (η) were calculated using the equation η = E(RHE) − 1.23 V.

3 Results and Discussion

The XRD patterns (Fig. 1b) demonstrate that the Cu28Mg72 ribbons composed primarily of orthorhombic CuMg2 phase (JCPDS No. 65-1116). In addition, several weak diffraction peaks (2θ = 32.4°, 34.6°, 36.8° and 47.9°) are ascribed to metallic Mg (JCPDS No. 01-1141), indicating the bimetallic precursor is a mixture of CuMg2 and Mg. Owing to the disparity in chemical stability of Mg and Cu element, the Mg was selectively and rapidly dissolved from the bimetallic ribbons in a HCl solution, accompanying with vigorous hydrogen gas evolution on the surface of ribbons. After dealloying, the main diffraction peaks at 43.4° and 50.5° of the p-Cu correspond to (111) and (200) planes of Cu (JCPDS No. 65-9026), respectively. The extremely weak peaks at 36.5° and 61.5° indicate the existence of a small quantity of Cu2O (JCPDS No. 65-3288) resulting from oxidation of surface Cu in atmospheric air. The disappearance of CuMg2 and Mg peaks is indicative of the dissolution of most the Mg during 60 s dealloying process. For the as-fabricated p-CuNWs, the XRD pattern suggests a mixture of Cu hydroxide/oxide. The diffraction peaks at 16.8°, 24.0°, 34.4° and 39.7° can be indexed to the (020), (021), (002) and (130) planes of orthorhombic Cu(OH)2 (JCPDS No. 72-0140), respectively, and the dominant pattern indicates the high percentage of Cu(OH)2 in p-CuNWs sample. Interestingly, the diffraction peaks of CuMg2, completely suppressed by the strong peaks of Cu in the p-Cu sample, clearly show off in p-CuNWs, indicative of the existence of little residual CuMg2 alloy and the relatively poor crystallinity of Cu(OH)2 as compared to metallic Cu. The additional weak peaks mainly result from a small quantity of CuO (JCPDS No. 03-0867) and residual Cu in p-CuNWs.

SEM images of p-Cu and p-CuNWs are shown in Fig. 1c–e. The p-Cu sample exhibits a porous structure with interconnected pore channels and Cu ligaments with a size of 100 ~ 200 nm (Fig. 1c). The anode oxidation of p-Cu results in the electrochemical generation of dense hydroxide/oxide needle-like nanowires on the surface, as shown in Fig. 1d. The high magnification SEM image (Fig. 1e) shows the nanowires with a uniform diameter of ca. 80 nm. The as-fabricated p-CuNWs exhibits integrated advantages of porous structure and nanowires, proving numerous active sites for OER and improved availability of active species. This in situ direct generation strategy also ensures a decent mechanical adhesion and electric connection to electrocatalysis application.

XPS was used to probe the valence states p-CuNWs, as shown in Fig. 2. The survey spectra shows signals for Cu, O and Mg (Fig. 2a). The weak signals of Mg indicate the low concentration of Mg on the surface. The high resolution of O 1 s spectra (Fig. 2b) could be deconvoluted into two peaks located at about 529.4 and 531.6 eV, corresponding to O2– and OH– respectively [34, 35]. The high intensity of OH− signals indicates that p-CuNWs surface is mainly made up of hydroxide. The Cu 2p spectra further confirms this observation, as shown in Fig. 2c. The appearances of the satellite peaks at about 942.9 and 962.8 eV are the evidences of Cu2+ [35,36,37]. The main Cu 2p3/2 (934.7 eV) and Cu 2p1/2 (954.6 eV) peaks indicate the Cu2+ connects with OH− [37, 38], in good agreement with the XRD results.

Fig. 2
figure 2

a XPS survey for p-CuNWs, b O 1 s and c Cu 2p regions

Full size image

The electrocatalytic OER performance of p-CuNWs was determined in 1.0 M KOH (pH 13.6), as shown in Fig. 3a. While the bare Cu28Mg72 alloy has poor catalytic activity for OER, p-CuNWs shows significantly superior catalytic activity and only needs an overpotential as low as 310 mV to drive current density of 10 mA cm−2. To drive 50 and 100 mA cm−2, p-CuNWs only needs an overpotential of 377 and 409 mV, respectively. The observed activity is also superior to the behaviors of most reported Cu-based OER catalysts in alkaline media, including 2D CuO nanosheet film (\(\upeta_{10\,{\text{mA cm}}^{-2}}\) = 350 mV) [17], Cu(OH)2 nanowire arrays (\(\upeta_{10\,{\text{mA cm}}^{-2}}\) = 430 mV) [19], Cu2O-Cu hybrid foams (\(\upeta_{10\,{\text{mA cm}}^{-2}}\) = 350 mV) [21]. The enhanced performance of the p-CuNWs could be attributed to the porous nanowire structure from dealloying and electrochemical treatment. The Tafel slope (Fig. 3b) of p-CuNWs is 85 mV dec−1, indicating that the discharge of OH− (M + OH− → M-OH + e−, where “M” denotes the active site) is most likely to be the rate determining step in alkaline conditions [18, 39].

Fig. 3
figure 3

a LSV curves of p-CuNWs and Cu28Mg72 ribbon in 1.0 M KOH. b The corresponding Tafel plots of p-CuNWs and Cu28Mg72 ribbon. c OER polarization curves of the p-CuNWs before and after 1000 cycles in 1.0 M KOH (inset: chronoamperometry curves of the p-CuNWs in 1.0 M KOH). d SEM image of p-CuNWs after chronoamperometry for 10 h

Full size image

The high catalytic performance of p-CuNWs electrode could be attributed to the Cu hydroxide/oxide nanowires on the surface. It is accepted that OER on oxide surfaces consists of four consecutive electrochemical reactions, and the metal hydroxide/oxide could be oxidized to active intermediate during the reaction process [40, 41]. In our case, the formation of CuIII oxide from Cu hydroxide/oxide at anodic potential is considered as the active intermediate for catalyzing OER, as revealed by in situ Raman spectroscopy on Cu-based catalyst in Yeo group [42, 43]. Meanwhile, the small amount of residue Cu and CuMg2 inside the catalyst provides a high conductivity that facilitates the transport of electron during the catalysis. The free-standing porous structure of p-CuNWs offers a large amount of active surface area and the free transport of medium molecules. All of these features make the p-CuNWs a superior OER catalyst as compared to other Cu-based materials.

Long-term stability of p-CuNWs was also evaluated by continuous cyclic voltammetry (CV) scanning for 1000 cycles in 1.0 M KOH (Fig. 3c). A small decay of activity is observed, indicating the imperfect durability of this material. The durability is further measured by electrolysis at an overpotential of 310 mV over 10 h (Fig. 3c inset). The p-CuNWs electrode exhibits an obvious decay of activity in the first hour, and then the current density keeps at a value of ca. 10 mA cm−2. To explore the reason of the observed activity decay, the surface morphology of p-CuNWs after electrolysis was analyzed by SEM, as shown in Fig. 3d. An observable coarsening and agglomeration of the nanowires occurred during the electrolysis, which may result in the small decay of long-term catalytic performance. Further study could focus on the improvement of the durability of p-CuNWs, and related work on this issue is currently underway.

4 Conclusion

A free-standing Cu-based electrocatalyst was fabricated by chemical dealloying Mg72Cu28 alloy ribbon and following electrochemical anodization process. The dissolution of Mg from the alloy during chemical dealloying resulted in a formation of porous Cu, and the following electrochemical anodization process converted most of Cu into Cu hydroxide/oxide nanowires with a uniform diameter of ca. 80 nm. The free-standing structure, high surface area and high porosity makes the as-prepared p-CuNWs a robust catalyst for OER. The p-CuNWs exhibits enhanced OER activity in alkaline solution with an overpotential of only 377 mV to offer current density of 50 mA cm−2. This work provides a pathway for designing Cu-based 3D architectures electrocatalyst for OER.