Synchrotron Radiation X-ray Photoelectron Spectroscopy as a Tool To

---

Article pubs.acs.org/JPCC

Synchrotron Radiation X‑ray Photoelectron Spectroscopy as a Tool To Resolve the Dimensions of Spherical Core/Shell Nanoparticles Won Hui Doh,*,† Vasiliki Papaefthimiou,† Thierry Dintzer,† Véronique Dupuis,‡ and Spyridon Zafeiratos*,† †

Institut de Chimie et Procédés pour l’Energie, l’Environnement et la Santé (ICPEES), ECPM, UMR 7515 du CNRS, University of Strasbourg, 25 rue Becquerel Cedex 2, F-67087 Strasbourg, France ‡ Institut Lumière Matière, UMR5306 Université Lyon 1-CNRS, F-69622 Villeurbanne cedex, France S Supporting Information *

ABSTRACT: In this work we demonstrate the potential of synchrotron X-ray photoelectron spectroscopy (XPS) to provide quantitative information on the intrinsic dimensions of core−shell nanoparticles. The methodology is based on the simulation of depth profiling curves, using simplified quantitative models earlier proposed in the literature. Three model systems consisting of X@ Fe2O3 (with X = Au, Pt, and Rh) metal−iron oxide core−shell nanoparticles, formed via oxidation of size-selected 5 nm bimetallic FeX nanoparticles inside the spectrometer, were measured in situ by near ambient pressure XPS. We show that when the shell layer is composed of a unique component, the experimental depth profiling curve can be simulated by the quantitative calculations and reveal the core and the shell thickness of the nanoparticles. On the contrary, a significant offset between the experimental and the theoretical depth profiling curves implies intermixing between the core and the shell layers. In this case the theoretical model has been modified to represent the more complex particle morphology. Transmission electron microscopy results are in good agreement with the XPS findings, confirming the validity of the model to predict the nanoparticle dimensions.

1. INTRODUCTION

is combined with spectroscopic methods, like electron energy loss spectroscopy. Alternatively, X-ray photoelectron spectroscopy (XPS) is well established as a powerful technique to probe the surface elemental composition and the oxidation state (up to 10 nm)14 of NPs. A unique advantage of the XPS technique is the possibility to perform depth-dependent measurements and to construct a concentration depth profiling without destruction of the sample.15,16 This is usually done either by collecting photoelectrons with different escape depths or at different detection angles (angle resolved XPS or ARXPS). It is imperative to note that ARXPS can be applied for planar samples, but not for csNPs. In the latter case, the XPS signal is not affected by the detection angle, since the projection area does not change at the different measurement angles,17,18 for example, the thickness of the shell “seen” from the electron analyzer is the same for all angles. Therefore, depth profiling on csNPs requires collection of spectra emitted from the same elements but with different escape depths, using laboratory11,12 or synchrotron19,20 based X-ray sources.

Nanoparticles are used increasingly in innovative products and this has stimulated a growing interest in order to understand and control their fundamental characteristics. In particular, core−shell nanoparticles (abbreviated as csNPs) have been extensively studied owing to their promising technical applications in photonics, electronics, catalysis, and so on.1−7 Several physical and chemical methods have been described for the fabrication of csNPs, such as evolution of a surface oxide layer over metallic NPs,1,2 shell coating by polymers,3,4 reduction of metal cations in aqueous solution,5,6 phase segregation of bimetallic alloys,7−9 and by wet chemical routes with or without organic capping (also called passivation) layer.10−12 To achieve the desired properties in the applications, the control of the dimensions and the oxidation state are crucial, and adequate analytical tools are essential in order to trace their properties. Transmission electron microscopy (TEM) is usually employed to measure the size, the structure, and often the shape of the NPs.13 However, in the case of csNPs, the core and shell materials should have sufficiently high contrast and relatively sharp interface in order to be clearly distinguished in a TEM micrograph. Furthermore, elemental and chemical analyses are not straightforward even if the TEM instrument © 2014 American Chemical Society

Received: February 21, 2014 Revised: October 23, 2014 Published: October 27, 2014 26621

dx.doi.org/10.1021/jp508895u | J. Phys. Chem. C 2014, 118, 26621−26628

The Journal of Physical Chemistry C

Article

their validity and their accuracy in describing the dimensions of csNPs. In this work we apply two of the simplified quantification methods28,34−36 to describe model samples containing csNPs with well-defined structure, narrow size distribution and no surface contamination. The samples consist of 5 nm iron-based alloyed nanoparticles (FeAu, FePt and FeRh) supported on a planar carbon substrate. In order to form noble metal-core/iron oxide-shell structures the alloyed NPs were exposed into 0.2 mbar O2 and measured in situ in order to avoid surface contamination, in a near ambient pressure photoelectron spectrometer (NAP-XPS). We show that fitting of the experimental depth profiling curves by using simplified theoretical equations allows the estimation of both the core radius and shell thickness of the csNPs with quite good accuracy as confirmed by TEM measurements. Finally, we discuss the potential of the simplified model to distinguish between csNPs with sharp or mixed interfaces.

One of the early attempts to distinguish between core/shell and homogeneous structures using XPS was described by J. M. White and co-workers for ZnSe@CdSe and CdSe@ZnSe NPs.10 They used a MgKα XPS source to collect the Zn-2p photoelectron peak (low kinetic energy; KE) and the Zn LMM Auger peak (high KE) from core/shell and randomly distributed particles. Since the escape depth of the detected electrons emitted from the same atomic species increases with energy, the contribution of surface components is less influential as the electron kinetic energy increases (i.e., for the Zn LMM peak). Therefore, different NPs structures have a distinct effect to the detected XPS signal. The integrated spectral intensity (or for brevity just intensity), or the shape of the photoelectron peaks, has been used in order to determine the thickness of a supported overlayer, for both flat21−24 and spherical geometry.17,18,25,26 For flat, thin overlayers with sharp interface, the estimation of layer(s) thickness(es) by XPS is quite accurate and relatively simple to calculate.24,27 On the other hand, in the case of core/ shell spherical nanosized particles, a shape-dependent parameter, the so-called geometric correction factor, has been introduced to take into account the spherical symmetry around the particle core. The accuracy of the methods is improved when the photoelectrons arising from the core and the shell have similar kinetic energies.14,28 Sarma and co-workers proposed a detailed quantitative analysis model to determine the dimensions of the csNPs from laboratory29,30 and variableenergy synchrotron31 based XPS measurements. In their method, the theoretical XPS intensities were obtained by integrating the infinitesimal intensity contribution from a volume element dv of the csNPs (eq 1) by using the spherical polar coordinates. The infinitesimal XPS intensity, dI, is expressed as follows; ⎛ x⎞ dI = nσI0 exp⎜ − ⎟dv ⎝ λ⎠

2. EXPERIMENTAL SECTION a. Sample Preparation. The iron-based bimetallic core− shell structured nanoparticles were prepared as described elsewhere.8,37,38 Briefly, chemically disordered fcc FeAu, FePt, and FeRh NPs in equimolar composition were produced separately by the Low Energy Cluster Beam Deposition (LECBD) technique. The mass-selected LECBD method consists of two steps; the bimetallic NPs are produced by a laser vaporization-gas condensation source and are consequently subjected to size selection using a quadrupolar electrostatic deviator acting as a mass-filter. Subsequently the NPs were soft-landed under ultrahigh vacuum (UHV) conditions on a 10 nm thick amorphous carbon (aC) film predeposited either on a SiO2/Si substrate (for photoemission measurements) or on an aC coated commercial copper TEM grid (for TEM measurements). The NPs size distribution follows a Gaussian shape with a median diameter Dm = 5 nm and size distribution