Xuejiao Yan‡
a,
Haiyan Xiong‡b,
Qingguo Baia,
Jan Frenzelc,
Conghui Sia,
Xiaoting Chena,
Gunther Eggelerc and
Zhonghua Zhang*a
aKey Laboratory for Liquid-Solid Structural Evolution and Processing of Materials (Ministry of Education), School of Materials Science and Engineering, Shandong University, Jingshi Road 17923, Jinan, 250061, P.R. China. E-mail: zh_zhang@sdu.edu.cn; Fax: +86-531-88396978; Tel: +86-531-88396978
bCenter for Advanced Energy Materials & Technology Research (AEMT), and School of Chemistry and Chemical Engineering, Shandong University, Jinan 250100, China
cInstitut für Werkstoffe, Ruhr Universität Bochum, Bochum 44780, Germany
First published on 10th February 2015
Atomic layer-by-layer construction of Pd on nanoporous gold (NPG) has been investigated through the combination of underpotential deposition (UPD) with displacement reaction. It has been found that the UPD of Cu on NPG is sensitive to the applied potential and the deposition time. The optimum deposition potential and time were determined through potential- and time-sensitive stripping experiments. The NPG-Pd electrode shows a different voltammetric behavior in comparison to the bare NPG electrode, and the deposition potential was determined through the integrated charge control for the monolayer UPD of Cu on the NPG-Pd electrode. Five layers of Pd were constructed on NPG through the layer-by-layer deposition. In addition, the microstructure of the NPG-Pdx (x = 1, 2, 3, 4 and 5) films was probed by scanning electron microscopy (SEM), transmission electron microscopy (TEM) and scanning transmission electron microscopy (STEM). The microstructural observation demonstrates that the atomic layers of Pd form on the ligament surface of NPG through epitaxial growth, and have no effect on the nanoporous structure of NPG. In addition, the hydrogen storage properties of the NPG-Pdx electrodes have also been addressed.
Palladium nanomaterials, well-known for their remarkable capacity in hydrogen absorption and less expensive than platinum,6 are widely used in catalysis, optical sensors7 and hydrogen sensors.8 In order to maximize their performance in all of these applications, the size and shape of Pd nanomaterials are critical parameters.9 The atomic layer-by-layer construction of Pd plays a crucial role in its applications for different fields. For deposition of nanostructured Pd, the substrate with high chemical stability, oxidation resistance and good biocompatibility need be carefully chosen. Recently, nanostructured Au attracted a great deal of attention in applications such as catalysis, drug delivery, biological labeling, etc.9,10 Among kinds of Au nanostructures, unsupported nanoporous gold (NPG) demonstrates remarkable catalytic activity for CO oxidation at low temperatures11,12 and for direct catalytic oxidation of methanol.13 NPG film made by dealloying of Au/Ag alloys has a unique nanoporous structure with good electrical conductivity,14 and is a good substrate for Pd deposition.
Underpotential deposition (UPD) is electrodeposition of a single monolayer or submonolayer amount of foreign metals and non-metals on the substrate at a potential positive than the thermodynamic potential.6,15,16 Up to date, UPD has been widely used to prepare catalyst monolayers with precise coverage. When the interaction energy between M–S (metal M and metal S) is larger than that between M–M, UPD will take place with M deposition on the S substrate.17 In addition, the structure of the substrate surface plays an important role in the UPD process, and determines specific features of the growing deposition layer. Numerous UPD systems have been extensively investigated on polycrystalline or single crystalline noble metal substrates (Pt and Au), such as Cu2+, Pb2+, Bi3+, Ag+ etc.6,18–30 For example, Shao et al.28 designed the dealloyed PdNi/C core–shell catalysts involving Pt displacement of a UPD Cu monolayer.
Because Cu UPD can occur on both Au and Pd surfaces, this kind of operation can be repeated to form ultrathin Pd films/layers in a precise manner from one to several atomic layers. In this work, we demonstrate how the UPD technique can be applied to fabricate ultrathin Pd film on NPG surface in an atomic layer-by-layer mode through the combination of UPD with displacement reaction. The NPG films were firstly prepared by the dealloying of commercial Ag–Au leaves. Thus the UPD potential of Cu on NPG was determined through time- and potential-sensitive stripping experiments. We have also found that the deposited Pd layer has a visible influence on the subsequent UPD of Cu, and further investigated the hydrogen storage of the Pd coated NPG electrodes.
NPG films were prepared through dealloying commercial white gold leaves (100 nm thick, 12 carat, Au50Ag50 in wt%, Noris-Blattgold GmbH, Germany) by floating them in a concentrated HNO3 (65 wt%) solution at room temperature for 30 min.17 The as-prepared NPG films were carefully rinsed with ultrapurified water several times. Afterwards, the NPG films were loaded onto the surface of a glass carbon (GC) electrode, and then were dried in vacuum. All electrochemical experiments were performed in a three-electrode cell with a CHI 760E Potentiostat at room temperature. The GC electrode loaded with NPG was used as the working electrode. A Pt plate was served as the counter electrode, and an Ag/AgCl electrode (KCl-saturated) was used as the reference electrode.
The UPD of Cu on NPG was carried out in a 0.5 M H2SO4 solution containing 1 mM CuSO4. Afterwards, the electrode was immediately immersed into a 250 mM HCl solution containing 0.5 mM PdCl2, holding for 10 min for displacement of Cu layer by Pd. Thus a Pd monolayer decorated NPG electrode was obtained and designated as NPG-Pd1. Because UPD of Cu can occur on both Au and Pd surfaces, NPG-Pdx with different atomic Pd layers (namely, NPG-Pd2, NPG-Pd3, NPG-Pd4 and NPG-Pd5 for 2, 3, 4, and 5 Pd layers respectively) can be obtained by changing the deposition potential of corresponding system and repeating the above UPD and displacement steps. The electrochemical behavior of the as-prepared NPG-Pdx electrodes was characterized by means of cyclic voltammetry (CV) in the 0.5 M H2SO4 solution and 0.1 M KOH solution.
Scanning electron microscope (SEM, Quanta FEG 250) was used to characterize the morphology of the NPG-Pdx electrodes. The microstructure of NPG-Pdx was also characterized using transmission electron microscopy (TEM, FEI Tecnai G2) and selected-area electron diffraction (SAED). The scanning transmission electron microscopy (STEM) images were also obtained under high angle annular dark field (HAADF) mode. In addition, the chemical compositions of the as-prepared NPG-Pdx electrodes were determined by energy dispersive X-ray spectroscope (EDX) in the SEM. Nanobeam-EDX (NB-EDX) analysis was also performed on the NPG-Pdx electrodes by the FEI Tecnai G2 microscope under the HAADF mode.
Fig. 1 (a) Macrographs of the NPG film before (left) and after (right) dealloying. (b) CV of the NPG electrode in the 0.5 M H2SO4 solution at the scan rate of 50 mV s−1. |
Fig. 2 CV of the NPG electrode in the 0.5 M H2SO4 + 1 mM CuSO4 solution at the scan rate of 2 mV s−1. |
Improper potential will result in deposition of either submonolayer or bulk Cu. In order to achieve maximum Cu monolayer coverage without bulk Cu deposition, two parameters including deposition potential and deposition time should be optimized. For optimizing the UPD potential, the NPG electrode was operated at different potentials in the 0.5 M H2SO4 + 1 mM CuSO4 solution in an amperometric i-t mode, all for a constant period of 240 s. Soon afterwards, the anodic stripping voltammograms were recorded by sweeping the electrode potential from the deposition potential to the final potential of 0.7 V vs. Ag/AgCl, which was high enough to completely remove the deposited Cu. Fig. 3a shows the stripping curves of the NPG electrode with Cu deposited at different potentials. Peak A′/B′ and C′ correspond to the stripping of UPD Cu and bulk-deposited Cu respectively. The stripping peak of Cu evolves quickly with the decrease of the deposition potential, which could be clearly observed from the enlarged plot of Fig. 3a (Fig. 3b). It is clear that the bulk deposition of Cu takes place below −0.01 V vs. Ag/AgCl. The present results are consistent with the report by Khosravi and Amini.37 Then we got the stripping charges of Cu by integrating the area under the stripping voltammogram. Fig. 3c shows the change of stripping charges against the deposition potential. The charge values increase slowly with decreasing deposition potential from 0.25 to 0 V vs. Ag/AgCl. However, the charge values increase sharply when the deposition potential further decreasing to a more negative value. And the charge for the deposition potential of −0.02 V vs. Ag/AgCl is even two orders of magnitude larger than that for the potential above 0 V vs. Ag/AgCl. Moreover, it can be seen that the deposition of Cu is quite sensitive to the change of potential in the UPD region. Thus we can roughly determine the monolayer UPD potential of Cu around −0.005 V vs. Ag/AgCl, as highlighted by an arrow in Fig. 3c. Fig. 3d shows the schematic illustration of Cu deposition on the NPG surface. Initially, Cu atoms deposit on the NPG surface forming a uniform monolayer through the UPD process. When the potential goes more negatively, the bulk deposition of Cu occurs on the electrode surface.
In order to find the more accurate deposition potential and deposition time, we conducted a series of experiments at −0.007 and −0.009 V vs. Ag/AgCl all for 60, 120, 180, 240, 300 and 360 s. Fig. 4 shows the stripping curves of the NPG electrode with the deposited Cu. The intensity of the stripping peaks (B′ and C′) increases obviously as time growing when the applied deposition potential is −0.009 V vs. Ag/AgCl (Fig. 4a). And the stripping peaks increase relatively slow when the deposition time is less than 240 s. Once the deposition time is longer than 240 s (for example, 300 s), the stripping peaks increase sharply, indicating the occurrence of bulk deposition of Cu. This suggests that 240 s is long enough to get monolayer Cu UPD on NPG when the applied potential is negative enough. Longer time will result in bulk Cu deposition. When the deposition potential becomes a little more positive (for instance −0.007 V vs. Ag/AgCl, Fig. 4b), however, the stripping peak intensity increases gradually and peak C′ does not emerge even though the deposition time reaches 360 s. This indicates that −0.007 V vs. Ag/AgCl is too positive to obtain monolayer UPD of Cu on the NPG surface. According to the above results, the optimum deposition potential and time were determined to be −0.009 V vs. Ag/AgCl and 240 s for the monolayer UPD of Cu on NPG, respectively.
Fig. 4 Stripping voltammograms of Cu on NPG deposited at the potential of (a) −0.009 and (b) −0.007 V vs. Ag/AgCl for different times in the 0.5 M H2SO4 + 1 mM CuSO4 solution. |
The next thing was to determine the deposition potential of monolayer Cu UPD on the surface of NPG-Pd1. First the voltammetric behavior of Cu on NPG-Pd1 electrode was probed in the 0.5 M H2SO4 + 1 mM CuSO4 solution (Fig. 5a). The CV profile of the NPG-Pd1 electrode is different from that of the NPG electrode, with the absence of peaks A and A′. For the NPG-Pd1 electrode, it is difficult to unambiguously discern the UPD and bulk deposition of Cu from the CV curve. Thus we try to determine the deposition potential for the monolayer UPD of Cu on the NPG-Pd1 electrode through the integrated charge control. That is, when the charge corresponding to the stripping peak (B′ and C′) is equal to that of monolayer Cu UPD on NPG, the applied potential is believed to be the UPD potential for the monolayer Cu on NPG-Pd1.
The NPG-Pd1 electrode was operated at different potentials in the 0.5 M H2SO4 + 1 mM CuSO4 solution in the amperometric i-t mode at a constant period of 240 s. The anodic stripping voltammogram (Fig. 5b) was recorded by sweeping from the deposition potential to 0.22 V vs. Ag/AgCl. Afterwards, the charge could be obtained through integrating the area under the stripping peaks (B′ and C′). The integrated charge (382.3 μC) at the potential of −0.005 V vs. Ag/AgCl is approximately equal to that (386.9 μC) of monolayer Cu UPD on NPG. Thus the optimum deposition potential and deposition time were determined to be −0.005 V vs. Ag/AgCl and 240 s for the monolayer Cu UPD on the NPG-Pd1 electrode. Under this condition we could obtain monolayer Cu coverage on the NPG-Pd1 substrate. Subsequently, the as-prepared NPG-Pd1-Cu electrode was immersed into the Pd2+ solution for galvanic replacement to prepare two-layer Pd decorating NPG electrode (NPG-Pd2). Similarly, we could determine the deposition potential of monolayer Cu on the NPG-Pd2, NPG-Pd3 and NPG-Pd4 electrode (Table 1). The corresponding stripping curves are presented in Fig. 5c–e, and analogous scenarios could be observed for all the stripping curves. Moreover, the deposition potential of the NPG-Pdx electrodes slightly increases with increasing Pd layers (Table 1).
Substrate | NPG | NPG-Pd1 | NPG-Pd2 | NPG-Pd3 | NPG-Pd4 | NPG-Pd5 |
---|---|---|---|---|---|---|
UPD potential (V vs. Ag/AgCl) | −0.009 | −0.005 | −0.005 | −0.003 | −0.003 | |
Charge for H adsorption/desorption (μC) | 229.9 | 260.1 | 331.0 | 398.0 | 455.1 |
Fig. 6 CVs of the NPG, NPG-Pd1 and NPG-Pd2 electrodes in the 0.1 M KOH solution at the scan rate of 50 mV s−1. |
Some NPG-Pdx electrodes were selected for SEM, TEM and STEM characterization. Fig. 7 shows the typical SEM image of the NPG-Pd2 electrode. The electrode displays a typical open, bicontinuous ligament-channel structure, which is quite similar to that of the bare NPG electrode. The corresponding EDX results verify the presence of minor Pd in the NPG-Pd2 electrode, and a typical EDX spectrum is presented as Fig. 7b. According to the CV (Fig. 6) and EDX (Fig. 7b) results, the atomic Pd layers were successfully constructed on the NPG electrode. In addition, we have inspected all the NPG-Pdx electrodes by SEM, and similar nanoporous structures were observed in spite of different Pd layers. As shown in Fig. S1,† the morphology of NPG-Pd5 still maintains the bicontinuous nanoporous structure with smooth surface.
Fig. 8 and S2† show the TEM results of the NPG-Pd2 and NPG-Pd4 electrodes respectively. The three-dimensional nanoporous structure could be clearly observed with the Au surface being covered by a uniform Pd deposit. The average length scale of ligaments/channels is around 30 nm. The related SAED patterns of the NPG-Pd2 (inset of Fig. 8) and NPG-Pd4 (inset of Fig. S2†) confirm the single-crystalline nature of Pd deposited NPG film in the selected area (∼200 nm in diameter). The zone axis of NPG-Pd2 is close to the [001] direction of face centered cubic (fcc) Au (inset of Fig. 8). And the zone axis of NPG-Pd4 is the [110] direction of fcc Au (inset of Fig. S2†). The electron diffraction results demonstrate that at least a significant portion of Pd layers keeps the same crystallographic orientation as the NPG substrate during the displacement process. Additionally, a typical STEM image of NPG-Pd4 is shown in Fig. S3a.† The ligaments also show relatively smooth surfaces. The NB-EDX analysis (Fig. S3b and c†) indicates that both the center and border of the ligament are covered by Pd atoms. On the basis of the above results, it is reasonable to assume that the atomic Pd layers decorate on the ligament surface through the epitaxial growth. In comparison, as reported in the literature,38–40 it is difficult to handle the atomic deposition of Pd on nanostructured gold substrate through regular electrodeposition. During regular electrodeposition, Pd tends to form nanoparticles. For example, Ke et al.40 have reported the different morphology of NPG before and after Pd electrodeposition. After electrodeposition, the NPG framework was packed by the dense and uniform Pd nanoparticles.
In addition, we have done comparison experiments to deposit Pt on NPG using similar UPD and displacement reaction. Even for NPG-Pt1, Pt atoms tend to form an island-like or particle-like structure on the NPG surface (Fig. S4†), which is quite different from the scenario of Pd deposition. Previous reports have shown that the deposition of Pt on Au surface is definitely one monolayer high, but they are partially interconnected nano-clusters.16,41 However, the morphology of Pt after galvanic replacement in our work (Fig. S4 in ESI†) is similar to that reported in the literature.42,43 Their results suggest that Pt nanoparticles grew over the substrate surface. The Pt ions appear to be reduced to Pt atoms which diffuse over the substrate surface due to very high Pt surface energy,43,44 creating very small Pt nanoparticles on the substrate surface. In the case of Pt deposition on NPG, similar morphology of Pt has also been reported in the previous report.17 One possible reason lies in the displacement reaction.42,45 For Pd, the reaction is given as, Cu + Pd2+ = Pd + Cu2+, and the atomic ratio of Pd/Cu is 1:1. For Pt, however, the reaction is given as, 2Cu + Pt2+ = Pt + 2Cu+, where the Cu atom is oxidized to Cu+.45 And the atomic ratio of Pt/Cu is 1:2. Even if the UPD-Cu is monolayer, the Pt layer is less than one atomic layer after displacement reaction and Pt atoms may form nanoclusters or nanoparticles. Of course, the underlying mechanism for this difference of Pd and Pt deposition should be probed in the following work (for example, theoretical calculation like DFT).
Fig. 9 CVs of the NPG and NPG-Pdx electrodes in the 0.5 M H2SO4 solution at the scan rate of 50 mV s−1. |
Footnotes |
† Electronic supplementary information (ESI) available. See DOI: 10.1039/c4ra17014h |
‡ The authors contribute equally to this work. |
This journal is © The Royal Society of Chemistry 2015 |