Chapter 4 Investigation of Formation Mechanisms for PtRu Core-Shell Nanostructures during
4.3.1 Materials characterizations on Ru@Pt/XC72/CC
The XRD patterns are able to provide qualitative evidences for the PtRu displacement reaction since the Pt and Ru adopt distinctive lattices of fcc and hcp, and the alloying of a relatively larger Pt atom (1.35 Å ) into Ru (1.3 Å ) structure is expected to render a slight expansion in lattice parameter.
Fig. 4.2(a) exhibits the XRD patterns for the XC72/CC, Ru/XC72, Ru/XC72/CC, and group A of Ru@Pt/XC72/CC from H2PtCl6 solution of pH 1, pH 2.2, and pH 8, respectively. As shown, the XC72/CC exhibited an amorphous background with a notable diffraction peak at 43.62°. This peak was also present for remaining samples in Fig. 4.2(a) and it is a characteristic carbon signal as evidenced by many studies.[119-122] For the Ru/XC72, the hcp phase of Ru was confirmed with relevant planes properly indexed. In addition, the XRD pattern suggested a polycrystalline structure with relative intensity consistent with that of JCPDS 06-0663. The observed stronger signals for the Ru/XC72 over those of Ru/XC72/CC were attributed to the sample preparation difference where a larger amount of Ru/XC72 was used for XRD measurement. For samples of Ru@Pt/XC72/CC, the
60
only discernible diffraction peaks were (100) and (101) which were associated with the Ru lattice.
Since the (101) overlapped with the background diffraction from the XC72 and CC, we repeated the XRD measurements at a slower scan rate around the (100) peak for comparison. The high-resolution XRD pattern for 36°-41° is displayed in Fig. 4.2(b). As shown, the Ru/XC72/CC revealed a (100) peak at 38.74°. For samples of pH 1, pH 2.2, and pH 8, the (100) peak was located at 38.7°, 38.5°, and 38.82°, respectively. The minor variation between these diffraction peaks was likely caused by poor crystallinity of Ru after displacement reaction as well as background noises from the XC72 and CC that compromised signal quality considerably. We believed that the possibility for PtRu alloying was rather remote as the displacement reaction took place at 40˚C and any interdiffusion between Pt and Ru was unlikely. Hence, the formation of core-shell Ru@Pt was presumed because from the standpoint of displacement reaction, the Pt was deposited upon removal of Ru and this dislocation process was occurring on the Ru surface exclusively. According to Alayoglu et al., for Ru@Pt nanoparticles with an approximately 1-2 monolayer-thick Pt shell, both hcp Ru and fcc Pt diffraction peaks were observed, albeit with considerable noises and reduced crystallinity.[108] In our case, unfortunately, the Pt signal was not identified form Fig. 4.2(b).
Earlier, Manandhar and Kelber studied the spontaneous deposition of Pt on Ru(0001) single crystals by X-ray photoelectron spectroscopy and confirmed that the Pt was partially reduced as Pt(II) cations on the Ru surface.[24] Therefore, it is likely that the Pt might exist in an oxidized form instead of metallic one in our case after the displacement reaction.
61
Figure 4.2. The XRD patterns for the XC72/CC, Ru/XC72, Ru/XC72/CC, and Ru@Pt/XC72/CC from group A of pH 1, pH 2.2, and pH 8 in scan range of (a) 30◦–90◦ and (b) 36◦–42◦.
62
To validate our premise that the Pt was not present in metallic state after the displacement reaction, we carried out the hydrogen reduction treatment and Fig. 4.3(a) demonstrates the XRD patterns for Ru@Pt/XC72/CC from samples of group B. Similar to what we observed in Fig. 4.2(a), a broad diffraction peak was recorded around 44° which was attributed to the combined effects of XC72/CC and Ru (101). Fig. 4.3(b) provides the high-resolution XRD pattern for 36°-42°. For samples of pH 1, pH 2.2, and pH 8, the Ru (100) peak was located at 38.5°, 38.54°, and 38.62°, respectively. These peaks were orderly shifted to lower angles as compared to that of Ru/XC72/CC at 38.74°. This suggested a moderate alloying of Pt in Ru matrix after the hydrogen reduction treatment. In addition, the degree of alloying increased with baths at smaller pH value. Moreover, for samples of pH 2.2 and pH 8, there appeared a minor diffraction peak of Pt (111) at 39.48°
corresponding to a fcc Pt with a lattice parameter of 3.95 Å . This value was slightly larger than the bulk Pt of 3.92 Å which was unexpected because the underlying Ru core is presumed to exert finite constrains for the Pt lattice above, causing it to shrink its lattice parameter slightly. However, we rationalized that the signals from CC and XC72 might interfere with diffraction responses from the Pt so the peak location might lose its fidelity. Nevertheless, we concluded that the Pt was present initially in some oxidized forms but transformed to metallic one after the hydrogen reduction treatment.
63
30 40 50 60 70 80 90
(a)
Ru@Pt XC72/CC pH 2.2
Intens ity (a .u.)
2 (degree)
Ru@Pt XC72/CC pH 8
Ru@Pt XC72/CC pH 1
Group B
36 37 38 39 40 41
(b)
Pt (111) Pt (111)
Ru/XC72/CC Ru@Pt XC72/CC pH 1 Ru@Pt XC72/CCpH 2.2 Ru@Pt XC72/CC pH 8
In tensity ( a.u. )
2 (degree)
Group B
Figure 4.3. The XRD patterns for the Ru@Pt/XC72/CC from group B of pH 1, pH 2.2, and pH 8 in scan range of (a) 30◦–90◦ and (b) 36◦–42◦.
64
The TEM images for Ru/XC72 and Ru@Pt/XC72 (group A) of pH 1, pH 2.2, and pH 8 are exhibited in Fig. 4.4. Also displayed in the insets are their respective pictures in high magnification.
As shown in Fig. 4.4(a), the Ru nanoparticles were irregular but uniformly distributed in the XC72 support with an average size of 3.6 nm. The high-resolution image confirmed a polycrystalline structure for each individual particle. Interestingly, their sizes were slightly increased after the displacement reaction from the TEM images in Fig. 4.4(b-d). The average size for the Ru@Pt nanoparticles of pH 1, pH 2.2, and pH 8 was 4, 3.8, and 4.6 nm, respectively. However, their morphologies were similar to that of Fig. 4.4(a). This minute variation in sizes between each group suggested that the Ru nanoparticles were rather stable in various environments against corrosive dissolution. This behavior was not unexpected as Pourbaix diagram predicted similar behaviors. In addition, due to the coulombic balance during the displacement reaction, the loss of Ru introduces deposition of Pt that rendered the Ru@Pt at similar sizes Moreover, results from EDX, listed in Table 4.1, confirmed the presence of Pt after the displacement reaction. The bath of pH 1 revealed the largest amount of Pt, followed by pH 2.2 and pH 8, indicating that the acidic environment was favored for Pt deposition or adsorption. These results were consistent with the XRD patterns in Fig.
4.3(b) since at pH 1 and pH 2.2, some of the Pt were alloyed with Ru so a relatively larger amount of Pt was reasonably expected.
Table 4.1. EDX results on Ru@Pt/XC72 from group A of pH 1, pH 2.2, and pH 8.
Ru (at%) Pt (at%)
pH 1 88.92 11.08
pH 2.2 93.35 6.65
pH 8 96.62 3.38
65
Figure 4.4. The TEM images for (a) Ru/XC72 and Ru@Pt/XC72/CC from group A of (b) pH 1, (c) pH 2.2, and (d) pH 8. The insets are their respective mages in high resolution.
Since the EDX results provided qualitative evidences at best, more accurate reading for the Pt and Ru amount after the displacement reaction was obtained via ICP-MS. Table 4.2 presents the ICP-MS results for the Ru@Pt/XC72 from group A of pH 1, pH 2.2, and pH 8, as well as their corresponding H2PtCl6 solution after the displacement reaction. As listed, the Pt loading was 1.26, 1.17, and 0.62 μmole for samples of pH 1, pH 2.2, and pH 8, respectively. Apparently, a lower pH bath allowed more Pt deposition or adsorption on the remaining Ru nanoparticles, a behavior consistent with EDX results in Table 4.1. Likewise, the Ru loading was 2.02, 2.94, and 4.86 μmole
66
for samples of pH 1, pH 2.2, and pH 8, respectively. For the corresponding H2PtCl6 solution, the amount of residual Pt at various pH values agreed well with what we expected. In addition, the amount of Ru cations in the H2PtCl6 solution was proportional to the Ru loss in Ru@Pt nanoparticles. Obviously, considerable Ru dissolution in conjunction with Pt deposition was observed at a low pH bath. From Table 4.2, the composition for the Ru@Pt nanoparticles was Pt38Ru62, Pt28Ru72, and Pt11Ru89 for samples of pH 1, pH 2.2, and pH 8, respectively.
In order to remove possible effect of Ru corrosion encountered in the acidic electrolyte, we also obtained ICP-MS results from samples of Ru/XC72/CC immersed in aqueous solution of pH 1, pH 7, and pH 8 (reference group). Since the Pt cations were not present in the solution, the amount of Ru recorded in the solution was caused entirely by corrosive dissolution instead of displacement reaction. The Ru amount in pH 1, pH 7, and pH 8 bath was 0.31, 0.01, and 0.03 μmol, respectively.
It can be seen that the Ru suffered from moderate corrosion in pH 1 bath but became relatively stable in pH 2 and pH 8 bath. Nevertheless, the amount of corrosive dissolution of pH 1 bath was still insufficient to account for the Ru content reported in Table 4.2. According to the Pourbaix diagram, the Ru is expected to be in a metallic state at 0 V for bath of pH 1, pH 2, and pH 8. Hence, we concluded that the observed displacement reaction was not driven by the corrosive dissolution of Ru but initiated by the difference in the redox potentials between the Ru and Pt complexes at different pH baths. According to Brankovic et al.[83], the driving force (ΔU) for the displacement reaction is the potential difference between the [PtCl6]2- reduction and Ru oxidation as shown in
67
In addition, it is believed that the oxidative dissolution of Ru occurs at a potential more positive than the equilibrium potential of Pt/[PtCl6]2- so it is unlikely to trigger the displacement
reaction.[123]
Table 4.2. ICP-MS results on Ru@Pt/XC72/CC from group A of pH 1, pH 2.2, and pH 8, as well as their corresponding H2PtCl6 solution.
Pt (μmol) Ru (μmol)
Ru@Pt/XC72/CC pH 1 1.26 2.02
pH 2.2 1.17 2.94
pH 8 0.62 4.86
H2PtCl6 solution pH 1 20.01 1.88
pH 2.2 21.61 1.37
pH 8 22.57 0.05