X-ray absorption spectroscopy analysis

X-ray absorption spectroscopy (XAS) studies were carried out to investigate the atomic structures of PtRu nanoparticles with T_off of 100, 400 and 600 ms, respectively, and the samples were designated as TF100, TF400, and TF600, respectively. XAS spectra for the Pt LIII-edge were obtained. Fig. 3.9(a) demonstrates the Pt L_III-edge XANES and EXAFS spectra from PtRu nanoparticles with different Toff. The Pt LIII-edge involves the energy associated with the electronic orbital transitions including 2p_3/2 → 5d5/2 and 2p_3/2 → 5d_3/2, also known as the white line. In general, the white line intensity can be used to determine the oxidation state of Pt due to its relevance to the vacancy of the d-band orbitals such as d5/2 and d_3/2. Usually a larger white line intensity indicates more vacant d-band orbitals, and consequently a higher oxidation state or a lower metallic state, and vice versa. The inset of Fig. 3.9(a) magnifies the white lines for our samples. It is noted that with a longer Toff, the metallic state of Pt inferred by the white line intensity of Pt LIII-edge became higher.

The TF600 sample with the longest T_off was the closest to the metallic state among all our samples.

This agrees well with the Pt/Ru atomic ratios measured by ICP-MS (Table 3.9). The higher Pt/Ru atomic ratios in PtRu alloys indicated a lower alloying degree and hence a metallic state. The TF600 sample with the longest Toff revealed the highest Pt/Ru atomic ratios rendering the highest metallic state among our samples. It has been reported that the binding energy between Pt atoms with neighboring ones decreased in the PtRu alloy.³⁶ Because the Ru atoms incorporating in the Pt lattice substituted some Pt atoms while the overall crystalline structure remained as fcc structure, the binding energy for Pt was altered magnificently. This resulted in the observed white line increase for PtRu nanoparticles as compared to that of Pt foil.

44

Figure 3.9 (a) The Pt LIII-edge XANES and EXAFS spectra and (b) the Pt LIII-edge k-space spectra for PtRu nanoparticles with different values of Toff (100, 400, and 600 ms) and fixed values of Ton

(50 ms), Ja (50 mA/cm²), and coulombic charge (8.0 C/cm²), along with Pt foil serving as the reference in (b).

45

Fig. 3.9 (b) displays the Pt L_III-edge k-space spectra for three samples, along with Pt foil serving as the reference. The phases of oscillation, i.e. the spectrum shapes, for three samples from TF100 to TF600 with prolonged T_off, revealed more resemblance to that of Pt foil. This indicates that the number of Pt-Pt bond increased with longer Toff. On the other hand, the amplitudes of oscillation for all three samples were much smaller than that of Pt foil, suggesting that nanoparticles contained fewer atoms with low coordination numbers as opposed to that of Pt foil.

Fig. 3.10 (a) demonstrates the Ru K-edge XANES spectra from PtRu nanoparticles with different Toff. The Ru K-edge involves the ionization of electrons from the 1s orbital. The inflection points of absorption for the three samples all close to 22117 eV, slightly increase with the metallic state value of Ru metal. So the Ru K-edge XANES spectra alone couldn’t provide distinctive information about the oxidation state or metallic state of Ru in three samples under different pulse deposition conditions.

Fig. 3.10 (b) displays the Ru K-edge k-space spectra for our samples, along with Ru metal serving as the reference. The phases of oscillation for the samples TF100 and TF400 were rather similar to that of Ru metal. However, the phase of oscillation for sample TF600 obviously deviated from that of Ru metal. This indicates that with a longer Toff, Ru-Ru bond is broken due to Ru dissolution in displacement reaction, leading to a higher possibility for the remaining Ru atoms to be bonded with Pt atoms. Consequently, the sample TF600 with the longest Toff demonstrated a larger number of Ru—Pt bond than the other two samples. On the other hand, the amplitudes of oscillation for all three samples were notably smaller than that of Ru metal, suggesting that nanoparticles contained fewer atoms with low coordination numbers than bulk Ru metal.

46 nanoparticles with different values of Toff (100, 400 and 600 ms) and fixed values of Ton (50 ms), Ja

(50 mA/cm²), and coulombic charge (8.0 C/cm²), along with Ru metal serving as the reference in (b).

47

The k³ weighted χ(k) data in k-space for the Pt L_III-edge (Fig. 3.9(b)) and Ru K-edge spectra (Fig. 3.10(b)) were Fourier-transformed to r-space. Fig. 3.11(a) shows the Pt LIII-edge Fourier-transformed EXAFS spectra. There were two obvious peaks for our samples. We surmised that the peak with a shorter bond distance (R) was associated with Pt-Ru bond, and the peak with a longer R was Pt-Pt bond, respectively. This is because at the first shell coordination with the Pt atom at the center, the Pt-Ru bond was reported to have a shorter bond distance than that of Pt-Pt bond[97, 98]. With a longer T_off, the amplitude of the peak associated with Pt-Ru bond became smaller and the amplitude of the peak associated with Pt-Pt bond became larger. This suggests that Pt atoms tended to form Pt-Pt bond rather than Pt-Ru bond after displacement reaction.

Fig. 3.11(b) shows the Ru K-edge Fourier-transformed EXAFS spectra. There were two notable peaks for our samples. We realized that the peak with a shorter bond distance (R) was associated with Ru-Ru bond, and the peak with a longer R was Ru-Pt bond, respectively. At the first shell coordination with the Ru atom at the center, the Ru-Ru bond was demonstrated to have a shorter bond distance than that of Ru-Pt bond[97, 98]. With a longer Toff, the amplitude for the peak associated with Ru-Ru bond became smaller and the amplitude for the peak associated with Ru-Pt bond became larger. This suggests that the Ru atoms tended to form Ru-Pt bond rather than Ru-Ru bond after displacement reaction. This fact is consistent with the result from Pt L_III-edge Fourier-transformed EXAFS spectra (Fig. 3.11(a)).

The EXAFS spectra (Fig. 3.11) in r-space were fitted by the data analysis package (Athena and Artemis) to analyze the coordination numbers between the center atom and the neighboring atoms.

Fixed bond distances were employed for Pt-Ru bond (R_Pt-Ru = 2.65 A) and Pt-Pt bond (R_Pt-Pt = 2.70 A). The EXAFS fitting results are summarized in Table 3.9. The NPt-Ru is the coordination number between the Pt center atom and the neighboring Ru atom. The P_Pt is the possibility for the Pt center atom to encounter a heterogeneous Ru atom. The equations for the calculation of PPt and PRu are as follows[99].

48

Ru Pt Ru Ru Pt

Ru Pt

Pt N

P N N

P N

 

  ^{ }

Combining the spectra in k-space and r-space, and the fitting results, we conclude that in pulse current deposition, for Pt atoms at the center, with a longer T_off, i.e. with a longer displacement reaction time, the Pt atomic ratio would become higher, the number for Pt-Pt bond would be larger and the number for Pt-Ru bond becomes smaller, and hence the possibility of P Pt is lower. On the other hand, for Ru atoms at the center, with a longer Toff, i.e. with a longer displacement reaction time, the Ru atomic ratio would be lower, the number for Ru-Pt bond would become larger and the number for Ru-Ru bond would become smaller, and hence the possibility PRu is higher. In addition, during galvanic displacement reaction in pulse current electrodeposition, Pt ions react directly with Ru atoms, leading to effective deposition of Pt ions on the surface of nanoparticles and an increased numbers of Pt-Pt bond and Ru-Pt bond. While Ru atoms are oxidized to ions by releasing electrons and subsequently dissolve in electrolytes, leading to a decreased numbers of Ru-Ru bond and Pt-Ru bond. All these results strongly confirm that the displacement reaction between Ru atoms and Pt ions has occurred when the current was resting in pulse current electrodeposition.

49 PtRu nanoparticles with different values Toff (100, 400 and 600 ms) and fixed values of Ton (50 ms), Ja (50 mA/cm²) and coulombic charge (8.0 C/cm²).

50

Table 3.9 EXAFS fitting parameters at Pt L_III-edge and at the Ru K-edge for PtRu nanoparticles with different values T_off (100, 400 and 600 ms) and fixed values of T_on (50 ms), J_a (50 mA/cm²) and coulombic charge (8.0 C/cm²).

N_Pt-Ru N_Pt-Pt ΣNPt N_Ru-Pt N_Ru-Ru ΣNRu P_Pt P_Ru Pt

(at%)

Ru (at%) TF100 4.16 5.91 10.07 4.69 3.54 8.23 0.41 0.57 51.9 48.1

TF400 2.25 5.9 8.15 6.2 2.98 9.18 0.28 0.68 65.5 34.5

TF600 1.0 8.98 9.98 9.01 1.52 10.53 0.1 0.85 83.4 16.6

Furthermore, by comparing the values of ΣNPt and ΣNRu (Table 3.9), we can analyze the atomic distribution of bimetallic PtRu nanoparticles, and then determine which element was located at the core and which element was located in the shell. Ru atoms incorporating in the Pt lattice substituted some Pt atoms while the overall crystalline structure of PtRu nanoparticles remained the Pt-fcc structure. So the element with a larger total coordination number (ΣN) resided at the core, and the element with a smaller ΣN resided in the shell. Guided by this principle, we obtained the schematic diagrams for the cross sections of PtRu nanoparticles prepared by pulse current electrodeposition under different conditions (Fig. 3.12). For the sample TF100 with the short T_off and the high duty cycle, its ΣNPt was larger than ΣNRu, suggesting that the nanostructure was composed of the Ru-rich shell and the Pt-rich core. While for the sample TF600 with the long T_off and the low duty cycle, its ΣNPt was smaller than ΣNRu, suggesting that the nanostructure was composed of the Pt-rich shell and the Ru-rich core[99]. It is demonstrated that not only the elemental composition but also the atomic distribution of bimetallic PtRu nanoparticles can be effectively controlled and adjusted by the pulse current electrodeposition technique.

51

Figure 3.12 Schematic diagrams for cross sections of PtRu nanoparticles with different values Toff

(100, 400 and 600 ms) and fixed values of Ton (50 ms), Ja (50 mA/cm²) and coulombic charge (8.0 C/cm²).