Thermal Stability and CO Oxidation Property of Non-Equilibrium Pd–Ru Alloy Catalyst

Takeru Fukushima; Ryota Tsukuda; Satoshi Ohhashi; Nobuhisa Fujita; Satoshi Kameoka

doi:10.2320/matertrans.MT-MH2022010

Abstract

The thermal stability and CO oxidation activity of a non-equilibrium Pd–Ru alloy obtained by leaching Al–Pd–Ru alloy (3/2 approximant, P₄₀ phase: Al₇₂Pd_16.4Ru_11.6 (at%)) with 20 mass% NaOH aqueous solution were investigated. When the Pd–Ru alloy was annealed under a He or H₂ atmosphere, phase separation of the Pd–Ru alloy proceeded under H₂ at 500°C, whereas the non-equilibrium state was relatively stable under He at temperatures as high as 500°C. In addition, the Pd–Ru alloy sample annealed under He (at 300 or 500°C) showed substantially greater CO oxidation activity than those annealed under H₂ (at 300 or 500°C). The results suggested that there is a more suitable microstructure of Pd–Ru alloy in the nanocrystals for CO oxidation than the solid solution state or the phase-separated state.

1. Introduction

The Raney method has long been known as an activation treatment method for bulk alloys.¹^,²⁾ It is carried out by leaching (i.e., selective dissolution) of Al from Al-based intermetallic compounds (Al–TM: TM = Ni, Cu, Co, etc.) containing a catalytically active transition metal TM in an alkaline aqueous solution, resulting in active materials with a high specific surface area (∼100 m² g⁻¹). In addition, bulk catalyst materials with unique shapes and properties can be produced in a single step by this method. When the Raney method is used for Al-based ternary intermetallic compounds (Al–TM₁–TM₂, where TM₁ and TM₂ are immiscible), non-equilibrium TM₁–TM₂ alloys such as supersaturated solid solutions can be readily obtained in many systems (e.g., Cu–Co,³⁾ Ag–Co,⁴⁾ Cu–Ru,⁵⁾ and Pd–Ru⁶⁾) even at room temperature.

Kitagawa and coworkers synthesized Pd–Ru alloy nanoparticles via a wet-chemical synthesis based on a modified polyol method⁷^–¹⁰⁾ even though Pd and Ru are immiscible at the atomic level in the bulk state. Interestingly, compared with Pd, Ru and Rh monometallic nanoparticles, these Pd–Ru alloy nanoparticles were found to exhibit high catalytic activity towards CO oxidation⁷^,⁸⁾ and NO_x reduction.⁹⁾ Although Pd–Ru alloy nanoparticles are expected to exhibit excellent catalytic properties, their preparation methods involve multistep chemical processes.⁸^,¹⁰⁾ Other potential alternative facile methods for preparing non-equilibrium alloys with high catalytic performance are needed. Recently, Mashimo et al. reported a one-step physical method that uses pulsed plasma in liquid (PPL) for the synthesis of Pd–Ru solid-solution nanoparticles.¹¹⁾ However, the thermal stabilities of these Pd–Ru alloy nanoparticles have not been fully investigated.

We have studied quasicrystals (QCs) and their approximants (APs) in Al–Pd–TM (TM = transition metal) system from the viewpoint of their unique structures (e.g., icosahedral clusters)¹²^–¹⁴⁾ and novel catalysis materials.¹⁵⁾ Recently, we found that Al–Pd–Ru alloys (icosahedral QC (IQC: Al₇₁Pd₁₉Ru₁₀), 3/2 approximant (3/2AP: Al₇₂Pd_16.4Ru_11.6), 1/1 approximant (1/1AP: Al_70.4Pd_14.7Ru_14.9)) become very useful precursors as catalysis materials.⁶^,¹⁶⁾ Non-equilibrium Pd–Ru alloys obtained by leaching of Al–Pd–Ru QC or APs alloys with 20 mass% NaOH aqueous solution (so called Raney method) exhibited greater CO oxidation activity at lower temperatures than Pd and Ru monometal catalysts,⁶^,¹⁶⁾ consistent with previously reported results.⁷⁾ However, the relationship between the thermal stabilities and catalytic properties of those Pd–Ru alloys has not yet been clarified. In the present study, we examined the thermal stability and CO oxidation property of a non-equilibrium Pd–Ru alloy obtained by leaching icosahedral approximant crystals of Al₇₂Pd_16.4Ru_11.6 (3/2 approximant; P₄₀ phase) of an Al–Pd–Ru alloy with 20 mass% NaOH aqueous solution.

2. Experimental Procedure

Al_72.0Pd_16.4Ru_11.6 (P₄₀ phase; 3/2 approximant),¹³^,¹⁷⁾ Al₁₃Ru₄ (decagonal approximant: DAP),¹⁸⁾ and Al₃Pd (decagonal quasicrystal: DQC)¹⁹⁾ were prepared under an Ar atmosphere using 99.99% Al, 99.99% Pd and 99.95% Ru in an arc-melting furnace. The as-cast Al_72.0Pd_16.4Ru_11.6 alloy and Al₁₃Ru₄ were annealed at 1000°C for 120 hours under an Ar atmosphere.⁶⁾ The as-cast Al₃Pd alloy was quenched from the liquid state via single-roller melt spinning, and a metastable DQC was obtained.¹⁹⁾

The precursors were crushed with a pestle and mortar and sieved to obtain particles in the size range 20–75 µm. The sample powders were leached in a 20 mass% NaOH aqueous solution, kept in the solution for 24 h and then collected by filtration and thoroughly washed with distilled water until no alkali was detected in the filtrate. The samples were dried at 40°C for 24 h. The composition (at%: atomic percent) of the samples after leaching treatment were determined by inductively coupled plasma analysis. To investigate the thermal stability of the samples, the leached sample with NaOH (aq) was annealed under He at 150, 300 and 500°C or in H₂ at 300 and 500°C for 3 h (a total flow rate of 30 mL min⁻¹: 0.1 MPa). The samples prepared from Al₃Pd, Al₁₃Ru₄ and Al_72.0Pd_16.4Ru_11.6 are designated as Pd(150), Ru(150) and PdRu(150, 300, 500, H300 and H500) (Table 1).

Table 1 List of each sample’s annealing conditions, surface areas, reaction rates.

Structural analysis was conducted by powder X-ray diffraction (XRD) (Rigaku, Ultima IV) using Cu-Kα radiation. The bulk chemical composition of samples was analysed by inductively coupled plasma atomic emission spectroscopy (ICP-AES) (SPECTRO, ARCOS EOP) while surface analysis was conducted using X-ray photoelectron spectroscopy (XPS; PHI 5600, ULVAC-PHI). The specific surface area of particles was determined using the Brunauer–Emmett–Taylor (BET) method with N₂ and/or Kr adsorption (BEL Japan, BSORP Max).

We investigated the catalytic activity of the leached samples towards CO oxidation using a fixed-bed flow reactor. Each sample powder was loaded into a quartz glass reactor with quartz wool. The loading amount of the leached Al–Pd–Ru alloy was 150 mg, and that of the leached Al₃Pd and Al₁₃Ru₄ catalysts was 50 mg. CO/O₂/He mixed gas (0.5% CO, 0.25% O₂, 99.25% He) was passed over the catalysts at a total flow rate of 30 mL min⁻¹ and a total pressure of 0.1 MPa. The catalysts were heated to a temperature at which CO was completely oxidized. The effluent gas was sampled at each temperature during the cooling process and analysed by gas chromatography (Shimadzu GC-8A). The catalytic activities were evaluated on the basis of the CO conversion rate calculated according to eq. (1):

\begin{equation} \text{Conv.} = \frac{C_{\text{blank}} - C_{T}}{C_{\text{blank}}} \times 100\ (\%) \end{equation}

(1)

where C_blank is the concentration of CO in the absence of a catalyst and C_T is the concentration after the CO was passed through the catalyst at temperature T.

3. Results and Discussions

The powder XRD patterns for the PdRu((150), (300), (500), (H300) and (H500)), Pd(150) and Ru(150) samples are shown in Fig. 1. The diffraction peaks for each original precursor alloy (Al₇₂Pd_16.4Ru_11.6, Al₃Pd and Al₁₃Ru₄) disappeared, and broad peaks corresponding to face-centred cubic (fcc) and/or hexagonal close-packed (hcp) phases were observed (data not shown) after the leaching treatment.⁶⁾ The broad peaks of PdRu(150) were indexed as the 111_fcc, 200_fcc and 220_fcc diffraction planes of an fcc phase and were shifted towards higher angles in comparison with the corresponding peaks of Pd(150). The fcc structure of the PdRu nanocrystals could be well preserved at temperatures as high as 500°C under He, which was confirmed by the broad peaks of the fcc phase. The broadness of the peaks indicates that the crystallite sizes of these samples were very fine. The XRD pattern of the sample annealed in H₂ at 500°C is dominated by 100_Ru, 111_Pd, 101_Ru, 200_Pd, 102_Ru and 220_Pd peaks, indicating a mix of fcc and hcp structures. The non-equilibrium Pd–Ru alloy decomposed into Pd and Ru phases after H₂ treatment at 500°C. As shown in Table 1, the surface areas of Pd–Ru alloy samples decreased with increasing annealing temperature (i.e., from 15.7 to 2.6 m² g⁻¹).

Fig. 1

XRD patterns for the leached alloy samples with each annealing (2 theta = 30°–80° region) and close-up of the 2 theta 35°–50° region. Miller indices of some of the peaks are shown in the figure for clarity.

Figure 2(A) shows the CO conversion rate as a function of the reaction temperature for the oxidation of CO over Pd–Ru alloy catalysts at different annealing temperatures. For comparison of the initial catalytic activities of each PdRu sample, the results for the Pd(150) and Ru(150) samples obtained from Al₃Pd and Al₁₃Ru₄ precursors, respectively, are also shown. The order of the catalytic activities was PdRu(300) > PdRu(500) > PdRu(150) > Ru(150) > PdRu(H300) > PdRu(H500) > Pd(150). The PdRu(300) sample exhibited the highest apparent catalytic activity. As shown in previous reports, the catalytic activities of the Pd–Ru alloy are substantially greater than those of the Pd or Ru catalysts.⁶^–⁸⁾ However, because the specific surface area of each catalyst is different (Table 1), it is more appropriate to compare the reaction rate per unit surface area of catalyst (mol min⁻¹ m⁻²-catal). For comparison among the bulk-type (i.e., unsupported) catalysts, the reaction rates in the units (mol min⁻¹ m⁻²-catal) are very useful in evaluating the catalytic activity because the outermost surface can interact with reactants.

Fig. 2

(A) Temperature dependence of CO oxidation and (B) Arrhenius plots for the CO oxidation reaction rate over each sample.

Figure 2(B) shows Arrhenius plots for the CO oxidation over the different catalysts, where the conversion of CO in Fig. 2(A) was converted to the areal reaction rate of CO (mol min⁻¹ m⁻²-catal). As shown in Fig. 2(B), the catalytic activities of the Pd–Ru alloy increased substantially with increasing annealing temperature (i.e., PdRu(150) < PdRu(300) ≤ PdRu(500)). Catalytic activity in terms of the areal reaction rate (mol min⁻¹ m⁻²-catal) of PdRu(300) is almost the same as that of PdRu(500). For example, the areal reaction rate of PdRu(500) at 75°C is one order magnitude higher than that of PdRu(150) (Table 1). From these Arrhenius plots, the apparent activation energies E_a for CO₂ formation over Pd(150), PdRu(300) and PdRu(500) were estimated to be 93.2, 80.8 and 83.3 kJ mol⁻¹, respectively. Thus, the apparent activation energies for the Pd–Ru alloys are lower than those for Pd(150) and Ru(150). Interestingly, the present results indicate that the catalytic activity of CO oxidation on Pd–Ru alloy does not depend on the surface area of the catalyst. These results indicate that the microstructures of Pd–Ru alloy in the nanocrystals are changed by annealing and that these variations lead to an improvement of the catalytic activity towards CO oxidation.

In a previous paper, the leaching rates of Al from Al_72.0Pd_16.4Ru_11.6, Al₁₃Ru₄ and Al₃Pd were reported to be 91.0, 91.3 and 91.1, respectively.⁶⁾ Even though the amount of residual Al species in each sample is similar, the catalytic performance of the Pd–Ru sample is substantially better than that of the Ru and Pd samples (Fig. 2). Therefore, we speculate that the contribution of the residual Al species to the catalytic activity is very small compared with the contribution of the alloying effect of Pd–Ru.

Figure 3 shows the Pd 3d and Ru 3d XPS spectra obtained from each sample. The Pd 3d binding energies shifted to higher values (ΔE ≈ +0.250 eV), and the Ru 3d binding energies shifted to lower values (ΔE ≈ −0.375 eV) when Pd and Ru were alloyed. In particular, the chemical shifts of Pd 3d and Ru 3d in PdRu(300) and PdRu(500) showed almost the same behaviours. These XPS results show the same tendency as those for Pd–Ru nanoparticles reported previously⁷⁾ and suggest that Pd and Ru developed a more positive and a more negative charge, respectively; that is, some of the electrons in Pd atoms apparently transferred to Ru atoms. These results indicate that the electronic states of Pd and Ru are also changed by annealing and that these also affect to the catalytic activity towards CO oxidation.

Fig. 3

Pd 3d and Ru 3d XPS spectra of each treated sample.

The CO oxidation reaction on transition-metal surfaces is known to occur between dissociatively adsorbed O and adsorbed CO (eq. (2) and eq. (3)) in a so-called Langmuir–Hinshelwood mechanism (eq. (4)). Notably, a gas-phase molecule and an adsorbed molecule are represented by symbols (g) and (a), respectively. Either O₂ dissociative adsorption or the CO oxidation reaction has been reported to be rate-determining for the CO oxidation reaction on a Pd catalyst.²⁰⁾ In general, transition metals (e.g., Group 8–11) towards the left side of the periodic table exhibit greater adsorptive power than transition metals towards the right side of the periodic table; they also exhibit greater molecular dissociation ability.²¹^,²²⁾ Thus, we expected Ru to exhibit greater adsorption power and O₂ dissociation ability than Pd. We speculated that O₂ dissociatively adsorbed onto Ru to compensate for the rate-determining O₂ dissociation on Pd, which enhanced the CO oxidation activity of the Pd–Ru alloy.

\begin{equation} \text{CO(g)} \to \text{CO(a)} \end{equation}

(2)

\begin{equation} \text{O$_{2}$(g)} \to \text{[O$_{2}$(a)]} \to \text{2O(a)} \end{equation}

(3)

\begin{equation} \text{CO(a)} + \text{O(a)} \to \text{CO$_{2}$(a)} \to \text{CO$_{2}$(g)} \end{equation}

(4)

In the conventional development of alloy catalysts, among other methods, the combination of metal elements, design of alloy compositions, formation of nanoparticles and the use of a non-equilibrium phase have been studied to maximize the ligand effect and the ensemble effect. The findings obtained in the present study have revealed a novel approach to the preparation of metallic catalysts. The results suggest that not only the alloy composition, nanoparticle synthesis method, and nanoparticle specific surface area but also the microstructure obtained using a non-equilibrium alloy will be an important method of tuning catalytic properties.

4. Conclusion

The thermal stability of the non-equilibrium Pd–Ru alloy obtained by leaching Al from Al–Pd–Ru alloy was investigated under various heat treatments (in a He or H₂ atmosphere). The Pd–Ru alloy was relatively stable under He (<500°C), whereas phase separation progressed under H₂ (e.g., 500°C). The effects of various heat treatments (under He or H₂) on the CO oxidation activity were investigated. The Pd–Ru alloy showed substantially better catalytic properties than Pd and Ru obtained from Al₃Pd and Al₁₃Ru₄ precoursor. A heat treatment under He enhanced the activity, whereas a heat treatment under H₂ reduced the activity because of phase separation. The results showed that there exists a more suitable microstructure of Pd–Ru alloy in the nanocrystals for CO oxidation than the solid solution state or the phase-separated state.

Acknowledgments

This work was supported in part by JSPS KAKENHI Grant Number (19H05819, 20H05260, 22H04581), a Grant-in-Aid for Scientific Research ((B) 22H01804) and the “Nanotechnology Platform” (A-21-TU-0034) from the Ministry of Education, Culture, Sports, Science and Technology (MEXT) of Japan, JST CREST (JPMJCR2203) and by the Dynamic Alliance for Open Innovation Bridging Human, Environment and Materials.

REFERENCES

Corresponding author

Version information

Register with J-STAGE for free!