Characterization and Oxidation of Fe Nanoparticles Deposited onto Highly Oriented Pyrolytic Graphite, Using X-ray Photoelectron Spectroscopy D.-Q. Yang † and E. Sacher* Regroupement Que ´becois de Mate ´riaux de Pointe, and De ´partement de Ge ´nie Physique, E ´ cole Polytechnique, C.P. 6079, succursale Centre-Ville Montre ´al, Que ´bec H3C 3A7 Canada ReceiVed: NoVember 19, 2008; ReVised Manuscript ReceiVed: February 24, 2009 The characterization of Fe 0 nanoparticles (NPs), both before and during oxidation, has been of concern for the last two decades. We have studied the 2p and 3p XPS core levels of Fe NPs evaporated onto highly oriented pyrolytic graphite (HOPG) under ultrahigh vacuum. Both components of the 2p spectrum of Fe 0 are found to be highly asymmetric to higher binding energy; each is composed of a major photoemission peak, plus several smaller peaks attributable to a vacancy cascade, a process known to occur in Fe. In contrast, the two Fe 3p spectral components are too close to be separated with precision, and were treated as one single component; as with the 2p components, it, too, is asymmetric, due to the vacancy cascade. The onset of oxidation affects both spectra somewhat differently, causing the introduction, and subsequent increase, of components on the high binding energy side of the 2p 3/2 spectrum, superimposed on the vacancy cascade; this is not as obvious for the 3p spectrum because of a larger probe depth, to which the surface contributes less. These new components represent the FeO, γ-Fe 2 O 3 , and Fe 3 O 4 formed on oxidation; their oxidation kinetics indicate that the initially formed FeO is rate controlling. Introduction Fe-based nanoparticles (NPs), including oxides and alloys, have received substantial attention for several decades because (1) such NPs constitute the key materials behind the recent development of rewritable electronic media, (2) improvements in their production have led to increased efficiency and reduced component size in many electronic products, (3) they can be used in the diagnosis and treatment of medical diseases and as electronic sensors, 1,2 (4) they represent an important catalyst used in the formation and cleavage of C-C bonds and, (5) under the name nanoscale zeroValent iron, and often abbreviated as nZVI, they are an effective reagent for the treatment of toxic and hazardous chemicals. 3,4 X-ray photoelectron spectroscopy (XPS) has been widely used to characterize the surface compositions of these materials. 5-10 The complexities of the Fe 2p core level peaks in the various Fe oxides (and halides) have been extensively noted in these references, although efforts have been made to use them in estimating chemical compositions and surface electronic states. Whereas the complexities have been discussed using many-body effects, 11,12 there is, nonetheless, substantial disagreement among the various models 13-18 that have been proposed. Early XPS studies focused on bulk 13-15 and thin film samples of Fe 19-22 and some of its compounds, 5-10 but comparatively little work has been carried out on Fe NPs. Here, we have used XPS to characterize the chemical state of Fe NPs, evaporated onto highly oriented pyrolytic graphite (HOPG), with which it reacts minimally. Such minimal interac- tion avoids surface wetting, permitting the surface retraction of the deposited material, to form NPs. Our motivations have been to acquire more detailed information on the core level peak shape as a function of deposited NP size, to furnish a better understanding of the XPS spectra, as well as to obtain standard spectra to be subsequently used in the quantitative analysis Fe/ oxide core-shell NP structures. 23,24 Experimental Section XPS were carried out on a VG ESCALab 3 Mark II, in which the sample preparation chamber is separated from the instrument analysis chamber by a gate valve, avoiding air exposure on sample transfer. Grade ZYA HOPG was obtained from Ad- vanced Ceramics, Inc.; it was cleaved with adhesive tape just prior to each experiment and immediately inserted into the spectrometer. This technique assures that an almost undetectable trace of oxygen is found on the HOPG, at the step edges where free radicals are created by the cleavage process. Measurements on samples so prepared indicate the relative concentration of oxygen to be ∼0.1%. High-purity Fe (>99.9995%) was evaporated in the prepara- tion chamber of our spectrometer, using an e-beam evaporator, at a deposition rate of 0.2 nm/min and a pressure of 2 × 10 -8 Torr (the base pressure was <1 × 10 -9 Torr); the deposition rate was monitored by a quartz microbalance, as previously described 17,25 our ability to control the deposition time is less than two seconds ((0.003 nm). The deposition rate was determined by a rate monitor at steady state evaporation and is, in fact, the rate of deposition of an equivalent layer; because Fe does not wet the HOPG substrate, the Fe immediately retracts, forming larger NPs, visible by AFM. The XPS instrument uses nonmonochromated Mg KR radia- tion (1253.6 eV), at 300 W, with an instrument resolution of ∼0.7 eV. After Shirley background subtraction, the spectra were analyzed by the freely available XPSPeak 4.1 program, 26 useful for asymmetric peak analysis, with peak positions and param- eters previously established in our laboratory. Energy normal- ization was accomplished by placing the major C 1s peak at 284.6 eV. Relative concentrations were obtained from high † Present address: Surface Science Western, Room G-1, Western Science Centre, The University of Western Ontario, London, Ontario N6A 5B7, Canada. J. Phys. Chem. C 2009, 113, 6418–6425 6418 10.1021/jp810171e CCC: $40.75  2009 American Chemical Society Published on Web 03/30/2009