NEXAFS and XPS Studies of (CH 3 ) 2 S Adsorption on Rh(PVP) Nanoparticle ∗

We have studied the adsorption reaction of dimethyl sulﬁde (DMS) on the Rh-polyvinylpyrrolidone (Rh(PVP)) nanoparticles by using AFM, XPS and NEXAFS techniques. The AFM images show the morphology of the Rh(PVP) nanoparticles depends on the amount of them. The XPS results indicate that the dissociation reaction of DMS into atomic S does not depend upon the existence of the Rh(PVP) nanoparticles. The NEXAFS results show that there is a strong chemical bonding between Rh(PVP) nanoparticle and atomic S. [DOI: 10.1380/ejssnt.2010.233]


I. INTRODUCTION
The automobile exhaust gas is contributing to aerial pollution, and causes serious environmental problem such as acid rain. To reduce the environmental load, the platinum-group metals (Rh, Pd and Pt) are used as the purification catalyst for the toxic substances like as NO X , HC and CO in the exhaust gas. Furthermore, using the platinum-group nanoparticles is efficient to improve catalytic activity since they have large surface area. However, there is a problem of sulfur poisoning that is preventing the reaction of the toxic substances and the metal surface when the sulfur-containing molecule in the automobile fuel remains on the catalyst surface [1]. Therefore, we have paid attention to Rh that has a high resistance and healing ability against the sulfur poisoning [2].
In this study, the Rh nanoparticles are synthesized by the chemical reduction method that is possible to massproduce at a low price and practically used for fabricating the automobile exhaust catalyst. In this method, the Rh nanoparticles are formed by reducing the Rh ions in ethanol-water solvent and prevented from aggregating with each nanoparticle by polyvinylpyrrolidone (PVP) [3,4]. Many kinds of synthesis method of the metal nanoparticles using PVP as the surfactant have been studied, and it is reported about controlling the size of them [5][6][7]. By using PVP, it is thought that the PVP remains on the Rh nanoparticles surface and may affect * This paper was presented at 7th International Symposium on Atomic Level Characterizations for New Materials and Devices, The Westin Maui Resort & Spa, Hawaii, U.S.A., 6-11 December, 2009. † Corresponding author: niwa.hironori@a.mbox.nagoya-u.ac.jp the chemical state of the surface [8]. To clarify the effect to the sulfur poisoning, it is necessary to investigate the adsorption reaction of the sulfur-containing molecules on the Rh(PVP) nanoparticles. In our previous studies, we have revealed the adsorption reaction of dimethyl sulfide (DMS: (CH 3 ) 2 S) on the Rh(100) single crystal surface [2,9]. It is supposed that the more step-edge structures exist on the Rh surface, the higher reactivity against the decomposition of DMS into atomic S is caused. It is expected that the Rh(PVP) nanoparticle possesses higher reactivity than the Rh single crystal surface, because the surface of the nanoparticle has more step-edge structures than single crystal. The purpose of this study is to reveal the adsorption reaction of DMS on the Rh(PVP) nanoparticles by using atomic force microscopy (AFM), X-ray photoelectron spectroscopy (XPS) and near edge X-ray absorption fine structure (NEXAFS) techniques.

II. EXPERIMENTAL
We prepared the Rh(PVP) nanoparticles by chemical reduction method [10]. The PVP is purchased from KISHIDA CHEMICAL Co., Ltd. and this molecular weight is 10,000 (K-15; research grade). The Rh(PVP) nanoparticles colloidal solution drops and spreads out on a Ni substrate (10×10 mm 2 , 0.50 mm thickness) by spin-coating method at 4,000 rpm and heated up to 723 K in the air. In order to investigate the difference in the adsorption reaction depending on the amount of the Rh(PVP) nanoparticle, three kinds of samples were prepared; sample (a) was put double drops of 200 µl Rh(PVP) nanoparticles colloidal solution, sample (b) was done single drop, and sample (c) was done double drops The samples (a) and (b) were cleaned by Ar + sputtering (3 keV, 90 min) to remove the burned embers of PVP and impurities derived from the air on the surface of the Rh(PVP) nanoparticles. The cleanliness of the Rh(PVP) nanoparticle surface was confirmed by XPS measurement (Rh 3d, S 2p, O 1s, N 1s and C 1s). The sample (c) was sputtered for the same time that it took in samples (a) and (b). Subsequently, they were exposed to DMS under a constant pressure of 1.0×10 −8 Torr after cooling down to 90 K. The exposure was kept up for 20 sec (0.2 L).
The particle sizes were evaluated by AFM measurement for the height value of the particles. The AFM measurement was carried out by NanoScope III-a (Veeco Instruments) with tapping mode. The XPS measurement was carried out at 90 K and room temperature by the MgKα X-ray (1253.6 eV) and the hemispherical electron analyzer (PHOIBOS100-5ch, SPECS). The NEX-AFS measurement was carried out with He-path system at the soft X-ray double crystal monochromator beamline BL-3 on Hiroshima Synchrotron Radiation Center (HSRC) [11,12]. The fluorescence yield detection was employed with using an UHV-compatible gas-flow type proportional counter with P-10 gas (10% CH 4 in Ar).  [10]. The particle size of annealed at 723 K is bigger than that of as prepared sample. It is considered that each Rh(PVP) nanoparticle is aggregated by annealing. Compared the image of sample (a) with (b), the valley of the substrate is more buried in sample (a) than (b). Moreover, the particles of sample (a) more aggregate than that of sample (b). It is thought that Rh(PVP) nanoparticles of sample (a) easily aggregate by heating since the surface of sample (a) is coated by double drops of Rh(PVP) nanoparticles. The differences of the morphology between the surface of samples (a) and (b) may cause the difference in the DMS adsorption reaction. We reveal the difference of that reaction through the following discussion of XPS and NEX-AFS results. are smaller than that for sample (c). It is considered that the Rh(PVP) nanoparticles are contributing to DMS dissociation at 90 K. In addition, the peak intensity of DMS for sample (b) is smaller than that for sample (a). This indicates that sample (b) has stronger activity against dissociation reaction of DMS than sample (a) at 90 K. This XPS results approve AFM results and show that the aggregation of nanoparticles decreases its activity. After rising temperature up to RT, the peak intensity of atomic S for all samples increases in comparison with the spectra at 90 K. This increase shows that adsorbing DMS, even it on the sample (c) without Rh nanoparticle, dissociates into atomic S with a rise of temperature. It is thought that the dissociation reaction does not depend upon the existence of the Rh(PVP) nanoparticles at RT. This result about sample (c) means that the PVP possesses the ability to dissociate the DMS into atomic S at RT. Figure 3 shows S K-edge NEXAFS spectra for samples (a) and (c). These spectra were normalized with the edge-jump to compare the difference of the chemical states per S atom for samples (a) and (c). Moreover, the spectra for standard samples are inserted as spectrum in each chemical state. The spectra of samples (a) and (c) have three characteristic peaks corresponding to the chemical states of atomic S, DMS and SO 2− 4 . The peak intensity of atomic S for sample (a) is larger than that for sample (c). Meanwhile, the peak intensity of SO 2− 4 for sample (a) is smaller than that for sample (c). Atomic S adsorbed on the sample (a) dose not easily oxidized to SO 2− 4 compared to that on the sample (c). The difference between samples (a) and (c) in the degree of this oxidation reaction of atomic S indicates that samples (a) and (c) have a different strength in the chemical bonding with atomic S. Therefore, there is the strong chemical bonding between Rh (PVP) nanoparticle and atomic S.

IV. CONCLUSION
We have investigated the adsorption reaction of DMS on the various amounts of the Rh(PVP) nanoparticles. The size and chemical states of the particle are estimated by using AFM, XPS and NEXAFS. The XPS results indicate that the DMS dissociation into atomic S is varied with the degree of dispersion of the Rh(PVP) nanoparticles at 90 K and does not depend upon the existence of the Rh(PVP) nanoparticles at RT. The NEXAFS results show that there is a strong chemical bonding between Rh(PVP) nanoparticle and atomic S.