692 Research Article Received: 30 July 2009 Revised: 11 December 2009 Accepted: 11 January 2010 Published online in Wiley Interscience: 25 February 2010 (www.interscience.wiley.com) DOI 10.1002/sia.3217 Chemical composition of superconducting SmFeAsO doped with fluorine † S. Kaciulis, a∗ A. Mezzi, a C. Ferdeghini, b A. Martinelli b and M. Tropeano b Bulk samples of SmFeAsO with and without F doping were prepared by thermal synthesis of the powders. Some samples of obtained oxy-pnictides were further pressed and sintered at 1300 ◦ C. The crystalline structure of the samples was investigated by X-ray powder diffraction. The samples were also characterized by resistivity measurements and SEM-Energy Dispersive Spectroscopy (EDS) analysis. The chemical composition (atomic concentration and chemical states of constituent elements) of the samples’ surface was analyzed by means of XPS before and after the sample cleaning by Ar ion sputtering. For the analysis of the bulk composition, the samples were scraped by using superfine diamond file and immediately transferred into XPS apparatus. In this way, the in-depth distribution of fluorine and the contribution of surface oxides were revealed. Copyright c  2010 John Wiley & Sons, Ltd. Keywords: oxy-pnictide; SmFeAsO; XPS; XRD; SEM Introduction Compounds referred to as Fe-based oxy-pnictides recently have attracted much attention after the discovery of superconductivity in F-substituted LaFeAsO. [1] Since this discovery, notable efforts of numerous research groups have been devoted to increase the critical temperature by substituting La with other Rare Earth (RE) atoms and by tuning the doping properly. [2] The doping can be varied in different ways: (i) by heterovalent substitution on the RE site, in this case, the record critical temperature of 56 K is obtained in the donor-doped Gd–Th system [3] ; (ii) by varying the oxygen stoichiometry to obtain acceptor-doped samples [4] and (iii) by F substitution in the O site resulting in donor-doped system. [1] Doping with fluorine is the most common way to dope efficiently the oxy-pnictides. [5] However, the optimization of the synthesis procedures and a deeper characterization is still required in order to obtain controlled and repeatable samples and to determine the precise chemical composition of this oxy-pnictide, in particular, with regard to the effective F concentration. Indeed, the effective F concentration seems to depend on the synthesis procedure that can result in uncontrolled loss of fluorine. To our knowledge, no systematic investigations on surface chemical composition of SmFeAsO have been reported until now. Therefore, in this article we report on XPS, SEM and XRD studies of bulk SmFeAsO doped with fluorine. Experimental Bulk samples of SmFeAs(O 1−x F x ) with x = 0, 0.07 and 0.15 were prepared in two steps: (i) synthesis of SmAs starting from pure elements in evacuated glass flask at a maximum temperature of 550 ◦ C and (ii) synthesis of the oxy-pnictide in a welded tantalum crucible closed in a sealed quartz ampoule, by reacting SmAs with stoichiometric amounts of Fe, Fe 2 O 3 and FeF 2 at 1200 ◦ C for 24 h in the form of a cylindrical pellet (diameter ∼1 cm; mass ∼3–4 g). After this reaction, the samples were furnace-cooled. The use of closed tantalum crucible reduces F losses, since it avoids the partial reaction of fluorine with the quartz vessel. It guarantees that the doping content strictly scales with the nominal one, which is intended as an upper limit to the real content. Electrical resistivity measurements were recorded by standard four probe technique. Phase identification was performed by X-ray powder diffraction (XRPD) (Philips PW1830; Bragg-Brentano geometry; CuK α ; range 20 ◦ –110 ◦ 2θ ; step 0.025 ◦ 2θ ; sampling time 12 s) and the microstructure of the samples was investigated both on the fractured samples and after metallographic preparation by SEM (Leica Cambridge S360) equipped with an Energy Dispersive Spectroscopy (EDS) microprobe (Oxford Link Pentafet). X-ray photoelectron spectra were collected by using an Escalab Mk II (VG Scientific) equipped with 5-channeltron detection system. Photoelectrons were excited by using a standard Al K α excitation source. The spectra were registered at 20 eV pass energy of the analyzer at a base pressure of about 10 −10 mbar that was increased up to 1 × 10 −7 mbar during Ar + ion sputtering at the beam energy of 2.0 keV. The binding energy (BE) scale was calibrated by positioning the C 1s peak at 285.0 eV. XPS analysis was carried out on the sample area of diameter ≈3 mm. In order to remove the surface layer and analyze the bulk composition, after the first measurement the samples were scraped by using superfine diamond file and immediately retransferred into XPS apparatus. ∗ Correspondence to: S. Kaciulis, Institute for the Study of Nanostructured Materials, ISMN-CNR, PO Box 10, 00015 Monterotondo Stazione, Roma, Italy. E-mail: saulius.kaciulis@ismn.cnr.it † Paper published as part of the ECASIA 2009 special issue. a Institute for the Study of Nanostructured Materials, ISMN-CNR, PO Box 10, 00015 Monterotondo Stazione, Roma, Italy b Artificial and Inovative Materials Laboratory, CNR-INFM-LAMIA, Corso Perrone 24, 16152 Genova, Italy Surf. Interface Anal. 2010, 42, 692–695 Copyright c  2010 John Wiley & Sons, Ltd.