Pd-Dispersed CeO₂ Catalyst Prepared from Dealloying the Pd–Ce–Al Ternary Amorphous Alloy Used for Oxidation Reaction

Abstract

The relationship between preparation method and the catalytic activity as well as reusability for Pd–CeO₂ catalyst was investigated. Pd-dispersed CeO₂ prepared from dealloying in NaOH aqueous solution of a Pd–Ce–Al ternary amorphous alloy as a precursor. The dealloying proceeded by active dissolution of Al and Ce to form sodium aluminate ion, Na[Al(OH₄)](aq), and CeO₂ particles, respectively. At the same time, Pd acting as cathode resulted in directly to be metallic nailed-particles exposing partly from the matrices of the CeO₂ particles. Its catalytic activity and reusability as a function of its surface area of exposed-Pd as well as sample surface area were investigated, compared with those of Pd-supported CeO₂ which was prepared by conventional liquid phase reduction. Pd-dispersed CeO₂ was higher in exposed Pd surface area than Pd-supported CeO₂ despite its lower sample surface area. It is likely that the nailed Pd on the CeO₂ prevents themselves from aggregation during sample preparation and reusability study. As a result, in the oxidation reaction of methyl benzyl alcohol, Pd-dispersed CeO₂ showed higher catalytic activity and reusability than Pd-supported CeO₂. The oxidation reaction of methyl benzyl alcohol is a measure of the oxidative decomposition reaction of harmful substances. Pd-dispersed CeO₂ is expected to be widely used for the oxidative decomposition reaction of harmful substances.

1. Introduction

Catalyst is a substance that promotes a specific chemical reaction without changing itself. Catalysts allow to produce various products that enhance the efficiency of chemical reaction processes, reduce environmental burden, and enrich life. To develop better metal catalysts, it is important to increase the active area per mass by increasing the surface area of the catalyst itself. However, with a decrease in the particle size and an increase in the surface area, the handling of catalyst becomes more difficult. Therefore, it is necessary to deposit the metal catalyst on a oxide support (e.g., SiO₂ and TiO₂) with porous structure, thermal and chemical stability.^1–3) Bell¹⁾ dimensioned the importance of metal oxides with high surface area for the dispersion of nanoparticles. Haruta²⁾ reported that the oxide supports plays important roles not only in the dispersion of Au catalyst but also in the oxidation reaction activity. Monyoncho et al.³⁾ reported the promotional role of oxide supports (CeO₂, SnO₂, TiO₂) on oxidation over Pd nanoparticles (NPs).

Yao et al.⁴⁾ showed that CeO₂ has excellent oxygen storage-release capacity which is advantageous property of a catalyst. Breysse et al.⁵⁾ and Mendelovici et al.⁶⁾ revealed that using CeO₂ as a catalyst support led to the excellent catalytic activity in CO oxidation⁵⁾ and ethylene oxidation reactions,⁶⁾ respectively. Fink et al.⁷⁾ and Masui et al.⁸⁾ prepared fine CeO₂ particles with high surface area from precipitation⁷⁾ and hydrothermal⁸⁾ synthesis. Recently, Ishikawa⁹⁾ clarified that CeO₂ prepared from a crystalline alloy by the dealloying method was advantageous in high surface area. The dealloying method is a method in which specific components are selectively extracted from a mother alloy by chemical treatment.⁹⁾ In addition, a sample with high surface area can be obtained due to the porous structure, which is formed by removing specific components.⁹⁾

In our previous studies,^10–13) the CeO₂ porous particles prepared from dealloying of a Ce–Al amorphous alloy as the precursor alloy were found to be higher in surface area than a commercial CeO₂, controlling their size and morphology as a function of dealloying time and precursor crystallinity. It was clearly showed that the catalysts composed of Au–Pd,¹⁰⁾ Ru,¹¹⁾ Ni,¹²⁾ and Au¹³⁾ supported-CeO₂ from an amorphous alloy showed higher catalytic activity in hydrogen generation^10–12) and oxidation¹³⁾ reactions than use of CeO₂ prepared from a crystalline alloy.

In the present study, the catalysts composed of Pd and CeO₂ prepared from an amorphous alloy was investigated. The Nobel Prize was awarded to Heck, Negishi and Suzuki for Pd-catalyzed cross-couplings reaction.^14–16) Pd catalysts were used for the reduction of nitrogen oxides¹⁷⁾ and the oxidation of hydrocarbons¹⁸⁾ that are indispensable for the purification of exhaust gas. In this study, we investigated the effect of morphology of Pd–CeO₂ on the oxidation catalytic reaction was investigated by preparing: (A) Pd-supported porous CeO₂ (denoted as Pd/CeO₂) prepared from dealloying of a Ce–Al binary amorphous alloy; (B) Pd-dispersed CeO₂ (denoted as Pd_CeO₂) prepared from dealloying of a Pd–Ce–Al ternary amorphous alloy. The catalytic characteristics of the former and latter were compared with altogether to understand from view of morphology of Pd–CeO₂. The metal deposited-oxides catalyst sometimes shows low reusability because the aggregation of deposited metal particles during the reaction. The reusability of the former, Pd/CeO₂ and the latter, Pd_CeO₂ was investigated. The catalytic activity and reusability of Pd–CeO₂ in the oxidation reaction of methyl benzyl alcohol were investigated. This oxidation reaction can be used as a practical measure of oxidation reaction activity such as purification reaction of various harmful substances.

2. Experimental Procedure

2.1 Catalyst preparation

Nanoporous CeO₂ was prepared from a Ce–Al amorphous alloy using the previously described preparation method.^10–13) The preparation method is shown below. Inoue et al.¹⁹⁾ reviewed that Ce–Al binary amorphous alloy can be prepared in the range of 7–10% Ce. Ce–Al alloy (Ce:Al = 8:92) ingots were prepared by the arc-melt technique from pure Ce and Al lumps in a highly purified Ar atmosphere. Ce₈Al₉₂ amorphous alloys in the form of ribbons were prepared from Ce–Al ingots through liquid quenching using the single roller melt-spinning method. As in the same way, Ce–Al–Pd ternary alloy (Ce:Al:Pd = 8:92:0.4) ingots were prepared by the arc-melt followed by preparation of Ce₈Al₉₂ Pd_0.4 amorphous alloys. The atomic ratio (Ce:Al:Pd = 8:92:0.4) was determined so that the sample after NaOH treatment was composed of 3 mass% of Pd and 97 mass% of CeO₂.

The prepared Ce–Al binary amorphous alloy was treated with a 1 M NaOH aqueous solution (20 mL) at 343 K for 10 h to selectively extract Al forming sodium aluminate aqueous ion followed by washing with distilled water and drying. At the same time, nanoporous CeO₂ particles were formed. The nanoporous CeO₂ prepared from an amorphous alloy was denoted as a-CeO₂. As in the same way, Pd-dispersed nanoporous CeO₂ was prepared from the Ce–Al–Pd ternary alloy, being denoted as a-Pd_CeO₂. Where only Al was extracted preferentially and Pd particles were dispersed in nanoporous CeO₂ directly to be that amount of Pd loading was 3 mass%.

Pd/a-CeO₂ catalysts (Pd loading: 3 mass%) were synthesized by depositing Pd on the as-synthesized CeO₂ supports (a-CeO₂) by the liquid phase reduction method. CeO₂ (0.2 g) was mixed with 14.19 mL of an aqueous solution containing Pd(NH₃)₄Cl₂ (10 mmol/L, 5.81 mL). The mixture of CeO₂ and metal ions was stirred at 293 K for 1 h; then, the samples were reduced with ammonia borane as a reducing agent. The suspension was evaporated under vacuum, and the obtained powder was dried overnight.

The Pd–Ce–Al crystalline alloy (c-Pd–Ce–Al) was prepared by heating the Ce–Al–Pd ternary amorphous alloy at 573 K for 2 h. As in the same way of a-Pd_CeO₂, Pd-dispersed nanoporous CeO₂ was prepared from the crystalline alloy, being denoted as c-Pd_CeO₂.

2.2 Catalyst characterization

The X-ray diffraction (XRD; Rigaku, Ultima IV) measurements were performed to analyze the crystallinity of the samples. Field emission scanning electron microscopy (FE-SEM; JEOL, JSM-7001) was used to determine the surface morphology of the samples. A sputtering instrument was used to coat the sample surfaces with carbon. Energy dispersive X-ray spectrometry (EDX, EDAX Ltd. DX-4) measurements were performed to analyze the bulk atomic ratio. The surface areas of the samples were estimated by the Brunauer–Emmett–Teller (BET) method using nitrogen physisorption isotherms obtained at 77 K (MicrotracBEL Corp., BEL-SORP mini). The adsorbed amount of CO was measured by pulse method (MicrotracBEL Corp., BEL-METAL-1) in order to quantify the Pd surface area. Komai et al.²⁰⁾ reported a method of measuring the dispersity and surface area of metals supported on CeO₂ by CO pulse adsorption measurement. CO is adsorbed only on the Pd surface of the sample, and the surface area of Pd can be determined from the amount of adsorbed CO.²⁰⁾ The X-ray photoelectron spectroscopy (XPS) analyses were carried out with photoelectron spectrometer (ULVAC-PHI Inc., PHI5000) using monochromatic X-rays (Al Kα, 1486.6 eV) to evaluate the Pd oxidation state on the sample surface. A charge neutralizer (low energy electrons and Ar⁺ ions) was used to compensate for the charging effect. The XPS spectra were analyzed after the Shirley-type background subtraction. The binding energy of 284.6 eV for the C 1s (adventitious carbon) was used for a calibration.

2.3 Catalytic reactions

The oxidation of methyl benzyl alcohol was performed in a quartz reaction vessel at 373 K in the presence of atmospheric air. The relevant Pd–CeO₂ catalyst (10 mg, 3 mass% Pd) and K₂CO₃ (0.5 mmol) were placed in a quartz reaction vessel. Then, a mixture of methyl benzyl alcohol (0.5 mmol) and toluene (5 mL) was injected to initiate the reaction. During the reaction, the reaction mixture was magnetically stirred at 373 K in the presence of atmospheric oxygen. At appropriate intervals, a portion of the reaction mixture was withdrawn and analyzed using a gas chromatograph (Shimadzu, GC-2014) equipped with a flame ionization detector. The quantification of organic substrate and products was performed using biphenyl as an internal standard.

3. Results and Discussions

The Pd–Ce–Al alloy ribbon prepared by the single-roll method can withstand bending, while the heat-treated Pd–Ce–Al ribbon is brittle. It was suggested that the Ce–Al amorphous alloy ribbon was crystallized by the heat treatment. XRD measurements were performed to determine the crystallinity of each sample (Fig. 1). The XRD pattern of a-Pd–Ce–Al did not show a sharp peak attributed to the crystalline alloy and a broad peak at approximately 38° was observed. The XRD patterns of without sample and c-Pd–Ce–Al did not show the broad peak at approximately 38°. This result indicated that a-Pd–Ce–Al primarily has an amorphous structure. The XRD pattern of c-Pd–Ce–Al showed a sharp peak attributed to Ce₃Al₁₁ and Al, which indicated that c-Pd–Ce–Al was crystalline. Further, there was no peak attributed to Pd; therefore, it is assumed that Pd moieties were dispersed in Ce–Al alloys.

Fig. 1

XRD patterns of a-Pd–Ce–Al, c-Pd–Ce–Al, a-Ce–Al, and without sample.

Figure 2 shows the XRD pattern of the sample after the prepared alloy was treated with the NaOH solution. The peaks attributed to CeO₂ were observed in all samples. These results suggested that Al moieties were selectively extracted from the mother alloy, and CeO₂ was prepared by performing the NaOH treatment. In addition, a small peak attributed to PdO was observed, which indicated that Pd species in the sample existed as PdO separately from CeO₂. As described later, the Pd species existing metallic state primarily of a-Pd_CeO₂ was found to be partially oxidized state from the Pd 3d XPS spectra. The PdO peaks shown in the XRD patterns were correlated with the partial oxidized states from the XPS spectra.

Fig. 2

XRD patterns of a-Pd_CeO₂, c-Pd_CeO₂, and Pd/a-CeO₂.

FE-SEM images were used to examine the surface morphology of the prepared sample (Fig. 3). a-Pd–Ce–Al had a fairly smooth surface, and no pores or cracks were observed. After NaOH treatment, porous structures were observed on the surface of a-Pd_CeO₂, c-Pd_CeO₂, and Pd/a-CeO₂.

Fig. 3

FE-SEM images of a-Pd–Ce–Al, a-Pd_CeO₂, c-Pd_CeO₂, and Pd/a-CeO₂.

Table 1 shows the composition ratio, the BET specific surface area and Pd surface area of the prepared Pd–Ce–Al ribbon and porous Pd–CeO₂. The composition ratios of a-Pd–Ce–Al were as follows: approximately 8.8 at% Ce, 90.6 at% Al, and 0.6 at% Pd. After NaOH treatment, the low Al composition ratios of a-Pd_CeO₂ and c-Pd_CeO₂ indicated that Al was selectively extracted. By comparing the specific surface areas of a-Pd_CeO₂ and c-Pd_CeO₂, it was determined that a-Pd_CeO₂ had a higher surface area. The difference in the sample specific surface area of the samples was due to the difference in the elution behavior of Al owing to the precursor atomic arrangement. Because Al is extracted slower from an amorphous alloy, a fine structure can be obtained and compared with that of a crystalline alloy. Pd/a-CeO₂ has a higher surface area than that of a-Pd_CeO₂. This result was obtained because the addition of the Pd served as the cathode during dealloying and accelerated the preferential extraction of Al as anode. CO adsorption measurement was performed to measure the surface area of only Pd, which is the active site of the sample. CO adsorption measurement results showed that a–Pd_CeO₂ and c–Pd_CeO₂ were higher in Pd surface area than Pd/a–CeO₂. This result suggest that dispersibilities of Pd in a–Pd_CeO₂ and c–Pd_CeO₂ are higher than in Pd/a–CeO₂. The sample surface area of a–Pd_CeO₂ (93 m² g⁻¹_cat) was slightly less than twice that of c–Pd_CeO₂ (50 m² g⁻¹_cat), whereas the Pd surface area of these samples (a–Pd_CeO₂: 50 m² g⁻¹_Pd, c–Pd_CeO₂: 42 m² g⁻¹_Pd) showed no significant difference. This result appears to indicate that the surface area of the sample does not significantly affect the surface area of the exposed Pd, and that Pd in Pd_CeO₂ tends to come out to the surface of the catalyst.

Table 1 Bulk atomic ratios, surface area, and exposed Pd surface area of the samples.

In the Pd 3d XPS spectra of Pd–CeO₂, peaks at 335.2, 336.3, 340.2 and 341.4 eV were observed (Fig. 4). The peaks at 335.2 and 340.2 eV can be assigned to Pd⁰ and the peaks at 336.3 and 341.4 eV can be assigned to Pd²⁺. The Pd species in the Pd/a-CeO₂ existed as its metallic state (Pd⁰ content of 61%) and its oxidized state (Pd²⁺ content of 39%). On the other hand, the Pd species in a-Pd_CeO₂ exist as its metallic state (Pd⁰ content of 71%) and its oxidized state (Pd²⁺ content of 29%). a-Pd_CeO₂ was higher in Pd⁰ ratio, thus, the preparation method of Pd–CeO₂ affects the charge state of the Pd species.

Fig. 4

Pd 3d XPS spectra of a-Pd_CeO₂ and Pd/a-CeO₂.

The catalytic activities of the oxidation reaction of methyl benzyl alcohol were examined. Figure 5 shows the conversion of methyl benzyl alcohol on a-Pd_CeO₂, c-Pd_CeO₂, and Pd/a-CeO₂ after 6 h. By comparing the conversion after 6 h, it was determined that a-Pd_CeO₂ showed an excellent activity. From the results of the CO adsorption measurement, it is determined that a-Pd_CeO₂ had the highest catalytic activity because Pd had the most exposed Pd species on the surface. By preparing the a-Pd_CeO₂ catalyst from an alloy containing Pd in a Ce–Al alloy, Pd was successfully supported with a high dispersion, which improved the catalytic activity. Furthermore, the initial reaction rates per Pd surface area of a-Pd_CeO₂ and Pd/a-CeO₂ were 2.42 and 1.17 µmol·h⁻¹·m⁻²·g_Pd, respectively. The activity per Pd surface area of a-Pd_CeO₂ was higher than that of Pd/a-CeO₂. From the XPS results, Pd⁰ ratio in a-Pd_CeO₂ was higher than that in Pd/a-CeO₂, and it is assumed that Pd⁰ effectively led to the activity improvement.

Fig. 5

Conversion of methyl benzyl alcohol on Pd–CeO₂ catalysts at 6 h.

The recyclability of the prepared samples was examined. After the first reaction of a-Pd_CeO₂ and Pd/a-CeO₂, the reaction solution was added again, and the recycling test was performed in the same way as during the first reaction to compare the catalytic activities. Figure 6 shows the results of the recycling test. a-Pd_CeO₂ and Pd/a-CeO₂ showed low reusability below 50%. It is assumed that the low reusability attributed to the aggregation of Pd particles during the reaction at 373 K. Although 3 mass% Pd–CeO₂ catalysts were used in this study, it is necessary to prepare a sample with lower Pd loading amount and suppress aggregation of Pd particles to improve reusability. The recyclability of a-Pd_CeO₂ was higher than that of Pd/a-CeO₂. a-Pd_CeO₂ suppressed the aggregation because Pd appeared be nailed in CeO₂.

Fig. 6

Reusability of a-Pd_CeO₂ and Pd/a-CeO₂.

Yao et al.⁴⁾ showed that CeO₂ has excellent oxygen storage-release capacity which is advantageous property of a catalyst support. Trovarelli²¹⁾ dimensioned that utilizing CeO₂ as a noble metal support improved the durability of the noble metal catalyst, and its high oxygen storage-release capacity enhanced the catalytic activity of noble metal. Oxygen storage-release capacity is an important function that leads to the enhancement of the catalytic activity of the noble metal in the oxidation reaction. CeO₂ is converted into Ce₂O₃ by redox reaction accompanying rapid valence change (Ce⁴⁺ → Ce³⁺) of cerium ion, and the oxygen species released during its process promotes the oxidation reaction over the noble metal.²¹⁾ In this study, it is assumed that the role of CeO₂ in Pd-dispersed CeO₂ is not only to improve the dispersibility and durability of Pd, but also to enhance the catalytic activity of Pd by the oxygen species released on CeO₂. The oxidation reaction of methyl benzyl alcohol is a measure of the oxidative decomposition reaction of harmful substances. Pd-dispersed CeO₂ is expected to be widely used for the oxidative decomposition reaction of harmful substances.

4. Conclusion

The effect of morphology of Pd on the oxidation catalytic reaction of methyl benzyl alcohol was investigated by synthesizing: (A) Pd-supported porous CeO₂ prepared from dealloying of a Ce–Al binary amorphous alloy; (B) Pd-dispersed CeO₂ prepared from dealloying of a Pd–Ce–Al ternary amorphous alloy. For both samples (A) and (B), dealloying in NaOH aqueous solution proceeded by active dissolution of Al and Ce to form sodium aluminate ion, Na[Al(OH₄)](aq), and CeO₂ particles. In the former (A), Pd was supported on CeO₂ particles by conventional liquid phase reduction method. In the latter (B), at the same time of anodic dissolution of Al and Ce, Pd acting as cathode resulted in directly to be metallic nailed-particles exposing partly from the matrices of the CeO₂ particles. The catalytic activity and reusability as a function of its surface area of exposed-Pd as well as sample surface area were investigated. The results obtained were as follows: (1) Pd-dispersed CeO₂ was higher in exposed Pd surface area than Pd-supported CeO₂ despite its lower sample surface area; (2) It is likely that the nailed Pd on the CeO₂ in the latter (B) prevents themselves from aggregation during sample preparation and reusability study; (3) In the oxidation reaction of methyl benzyl alcohol, Pd-dispersed CeO₂ resulted in higher catalytic activity and reusability than Pd-supported CeO₂, consistent well with its high exposed Pd surface area. The oxidation reaction of methyl benzyl alcohol is a measure of the oxidative decomposition reaction of harmful substances. Pd-dispersed CeO₂ is expected to be widely used for the oxidative decomposition reaction of harmful substances.

Acknowledgments

This study was partially supported by a Grant from university of Hyogo. A.N. also acknowledged to Kansai Research Foundation for technology promotion and Japan Association for Chemical Innovation, Japan.

REFERENCES

• 1)   A.T.  Bell: Science 299 (2003) 1688–1691.
• 2)   M.  Haruta: Cat. Tech. 6 (2002) 102–115.
• 3)   E.A.  Monyoncho,  S.  Ntais,  N.  Brazeau,  J.J.  Wu,  C.L.  Sun and  E.A.  Baranova: Chem. Electro. Chem. 3 (2016) 218–227.
• 4)   H.  Yao and  F.  Yao: J. Catal. 86 (1984) 254–265.
• 5)   M.  Breysse,  M.  Guenin,  B.  Claudel,  H.  Latreille and  J.  Véron: J. Catal. 27 (1972) 275–280.
• 6)   L.  Mendelovici and  M.  Steinberg: J. Catal. 96 (1985) 285–287.
• 7)   C.  Fink,  A.  Hagemeyer,  Z.  Hogan,  A.  Volpe and  J.  Yoder: Curr. Catal. 5 (2016) 182–202.
• 8)   T.  Masui,  H.  Hirai,  R.  Hamada,  N.  Imanaka,  G.-y.  Adachi,  T.  Sakata and  H.  Mori: J. Mater. Chem. 13 (2003) 622–627.
• 9)   Y.  Ishikawa,  M.  Takeda,  S.  Tsukimoto,  K.S.  Nakayama and  N.  Asao: Adv. Mater. 28 (2016) 1467–1471.
• 10)   A.  Nozaki,  A.  Yamashita,  R.  Fujiwara,  C.  Ueda,  H.  Yamamoto and  M.  Morishita: Bull. Chem. Soc. Jpn. 91 (2018) 1710–1714.
• 11)   A.  Nozaki,  C.  Ueda,  R.  Fujiwara,  A.  Yamashita,  H.  Yamamoto and  M.  Morishita: Mater. Trans. 60 (2019) 845–848.
• 12)   A.  Nozaki,  R.  Fujiwara,  C.  Ueda,  A.  Yamashita,  H.  Yamamoto and  M.  Morishita: Mater. Trans. 60 (2019) 1964–1967.
• 13)   A.  Nozaki,  T.  Yasuoka,  Y.  Kuwahara,  T.  Ohmichi,  K.  Mori,  T.  Nagase,  H.Y.  Yasuda and  H.  Yamashita: Ind. Eng. Chem. Res. 57 (2018) 5599–5605.
• 14)   R.F.  Heck: Org. React. 27 (1982) 345–390.
• 15)   E.  Negishi: Acc. Chem. Res. 15 (1982) 340–348.
• 16)   A.  Suzuki and  N.  Miyaura: Chem. Rev. 95 (1995) 2457–2483.
• 17)   J.F.  Trillat,  J.  Massardier,  B.  Moraweck,  H.  Praliaud and  A.J.  Renouprez: Stud. Surf. Sci. Catal. 116 (1998) 103–112.
• 18)   W.R.  Schwartz,  D.  Ciuparu and  L.D.  Pfefferle: J. Phys. Chem. C 116 (2012) 8587–8593.
• 19)   A.  Inoue,  K.  Ohtera and  T.  Masumoto: Jpn. J. Appl. Phys. 27 (1988) L736–L739.
• 20)   S.  Komai,  Y.  Yazawa,  A.  Satsuma and  T.  Hattori: J. Jpn. Petrol. Inst. 48 (2005) 173–177.
• 21)   A.  Trovarelli: Catal. Rev. 38 (1996) 439–520.