View Article Online View Journal

PCCP Accepted Manuscript

This article can be cited before page numbers have been issued, to do this please use: G. Portale, L. Sciortino, C. Albonetti, F. Giannici, A. Martorana, W. Bras, F. Biscarini and A. Longo, Phys. Chem. Chem. Phys., 2014, DOI: 10.1039/C3CP54847C.

This is an Accepted Manuscript, which has been through the Royal Society of Chemistry peer review process and has been accepted for publication. Accepted Manuscripts are published online shortly after acceptance, before technical editing, formatting and proof reading. Using this free service, authors can make their results available to the community, in citable form, before we publish the edited article. We will replace this Accepted Manuscript with the edited and formatted Advance Article as soon as it is available. You can find more information about Accepted Manuscripts in the Information for Authors. Please note that technical editing may introduce minor changes to the text and/or graphics, which may alter content. The journal’s standard Terms & Conditions and the Ethical guidelines still apply. In no event shall the Royal Society of Chemistry be held responsible for any errors or omissions in this Accepted Manuscript or any consequences arising from the use of any information it contains.

www.rsc.org/pccp

Journal Name

Physical Chemistry Chemical Physics

Dynamic Article Links ► View Article Online

DOI: 10.1039/C3CP54847C

Cite this: DOI: 10.1039/c0xx00000x

ARTICLE TYPE

Published on 10 February 2014. Downloaded by Lomonosov Moscow State University on 10/02/2014 16:47:51.

www.rsc.org/xxxxxx

Influence of metal-support interaction on the surface structure of gold nanoclusters deposited on native SiOx/Si substrates Giuseppe Portale*,a, Luisa Sciortinob, Cristiano Albonettic, Francesco Giannicib, Antonino Martoranab, Wim Brasa, Fabio Biscarinid, and Alessandro Longo*,a,e 5

10

15

20

25

30

35

40

45

Received (in XXX, XXX) Xth XXXXXXXXX 20XX, Accepted Xth XXXXXXXXX 20XX DOI: 10.1039/b000000x The structure of small gold nanoclusters (around 2.5 nm) deposited on different silica-on-silicon (SiOx/Si) substrates is investigated using several characterization techniques (AFM, XRD, EXAFS and GISAXS). The grain morphology and the surface roughness of the deposited gold cluster layers are determined by AFM. The in-plane GISAXS intensity is modelled in order to obtain information about the cluster size and the characteristic length scale of the surface roughness. The surface morphology of the deposited clusters depends on whether the native defect-rich (n-SiOx/Si) or the defect-poor substrate obtained by thermal treatment (t-SiO2/Si) is used. Gold clusters show a stronger tendency to aggregate when deposited on n-SiOx/Si, resulting in films characterized by a larger grain dimension (around 20 nm) and by a higher surface roughness (up to 5 nm). The more noticeable cluster aggregation on n-SiOx/Si substrates is explained in terms of metal-support interaction mediated by the defects located on the surface of the native silica substrate. Evidence of metal-support interaction is provided by EXAFS, demonstrating the existence of an Au-O distance for clusters deposited on n-SiOx/Si that is not found on t-SiO2/Si.

Introduction Nanocomposite materials containing metal clusters deposited on a dielectric matrix have wide technological applications in different fields, such as catalysis, photonics, magnetics and electronics.1-5 New synthetic strategies and preparation methods are continually developed, in order to improve device performances and to obtain novel classes of materials with unprecedented properties connected to the reduced dimensionality of the metal clusters. The structure and electronic properties of the metal clusters can be strongly influenced at the nanoscale level by the support on which the particles are deposited. Gold clusters deposited on different substrates have been extensively studied to understand the support influence and to elucidate the structural and electronic changes induced by the particle-support interaction. The main aspects of such a metal-support interaction are matter of debate, and they still require extensive theoretical and experimental investigations to be clarified. Nevertheless, this effect has been proposed to explain the behavior of many chemical systems, not only limited to heterogeneous catalysis.6-14 Various techniques, like EXAFS, TEM, XPS and GISAXS, have been recently used to understand the nature of the metal-support interaction.15-18 The experimental approach has been often paralleled with computational methods based on first principle calculations in which the metal cluster growth mechanism is divided into a series of elementary steps, eventually obtaining a complete microscopic characterization of the process.19-22 Guczi This journal is © The Royal Society of Chemistry [year]

50

55

60

65

70

and co-workers have experimentally studied the catalytic activity of Au clusters deposited on n-SiOx/Si substrates, which are easy to prepare. Considering the non-reducible nature of the support and neglecting the Au-SiO2/Si interaction, the authors demonstrated the correlation between catalytic activity and valence band density of states of nanometric gold particles.23 On the other hand, Pacchioni et al. showed via DFT calculations that the metal-support interaction cannot be ignored and plays an important role in the stabilization of highly dispersed metal nanoparticles.24-26 In particular, the chemical and physical properties of an ideal porous SiO2 film can be modified by surface doping, hereby producing a sharp increase in oxygen vacancies and inducing charge transfer to supported Au atoms.27, 28 This effect enables a strong Au-silica interaction mediated by a polaronic distortion of the oxide lattice thus opening new perspectives for the synthesis of improved silica-based supports for metal clusters.27, 28 Recently, new metal/SiO2 catalysts have been synthesized, exhibiting strong metal-support interaction. For example, new supported Ni/SiO2, Pt/CeO2 and Pt/CeOx/TiO2 catalyst with enhanced metal support interaction have been prepared to sustainably produce hydrogen.29, 30 To date, a detailed investigation of the structural mechanisms underlying the metalsupport interaction is missing, and its resolution is fundamental in order to understand the main steps in the synthesis of materials with improved performances. The aim of this work is to gain insight on the metal-support interaction in Au-silica systems by investigating the influence of [journal], [year], [vol], 00–00 | 1

Physical Chemistry Chemical Physics Accepted Manuscript

Page 1 of 8

Physical Chemistry Chemical Physics

Page 2 of 8 View Article Online

Published on 10 February 2014. Downloaded by Lomonosov Moscow State University on 10/02/2014 16:47:51.

5

10

15

20

the support surface on the structure and morphology of the deposited metal particles. n-SiOx/Si has been chosen as model system to simulate a doped silica matrix due to the different type of defects on the surface that grow spontaneously in an amorphous arrangement.31 Metal vapour deposition has been used to deposit the gold clusters at an early growth stage, in order to maximize the effect of the surface defects on particle aggregation. The gold clusters were deposited on three different native nSiOx/Si substrates. For comparison, gold clusters were also deposited on a thermally annealed t-SiO2/Si substrate which has a definitively lower concentration of surface defects. Chemical states, crystal structure, and surface morphology of the nanocomposite systems were characterized by using grazing incidence small-angle X-ray scattering (GISAXS), atomic force microscopy (AFM), X-ray diffraction (XRD), and extended Xray absorption spectroscopy (EXAFS) to investigate the structure of the deposited gold metal clusters and their aggregation behaviour at different length scale.

Experimental methods

55

Extended X-ray absorption fine structure (EXAFS)

60

65

70

Sample Preparation

25

30

Four different Si substrates have been used: one obtained by thermal treatment in air at 700 °C for 12 hours, having a defectfree oxide surface layer ~580 nm thick, and three bearing a native oxide surface layer. The latter forms with a thickness of about 5– 7 Å on the silicon surface after a few minutes of air exposure and can grow in a few days as far as 20 Å, depending on doping concentration, Si crystallographic orientation and environmental conditions.19 The oxide surface roughness σ, measured by AFM, was the same for all substrates, equal to 2.3±0.3 Å.

75

80

Table 1. Samples and substrates description.

35

40

45

SAMPLE

Support

Doping

A B C D

thermal native Native Native

Boron Boron Boron Phosphorus

Au load Doping level (atoms/cm3) (atoms/cm2) 4.24·1015 6.5·1016 1.62·1016 6.5·1016 2.35·1016 6.5·1016 16 7.14·10 6.5·1016

Samples are named with capital letters from A to D according to Table 1. Gold nanoclusters have been deposited by solvated metal vapour deposition.32, 33 This synthetic route consists of Au foil (99.99%, Sigma-Aldrich) evaporation by resistive heating in a Mo crucible under vacuum. The substrates were placed 10 cm above the crucible to ensure a homogenous deposition of Au nanoclusters. The deposition flux was equal to 3.6·1014 atoms/(s·cm2) and a deposition time of 180 s was used for all the samples. For each substrate, two specimens were prepared; one was used to measure the amount of deposited Au by using inductively coupled plasma mass spectrometry (ICP-MS). The metal load per surface unit was the same for all samples 6.5·1016 atoms/cm2. A deposited thickness of about 55 nm was estimated from X-ray reflectivity curves for all samples (see supporting information section). X-ray diffraction (XRD)

50

The X-ray diffraction measurements have been carried out using 2 | Journal Name, [year], [vol], 00–00

a Bruker D5000 vertical goniometer equipped with CuKα radiation and a graphite monochromator. A proportional counter and a 0.05° step size in 2θ have been used to collect the XRD curves. The assignment of the crystalline phases has been based on the JCPDS database.

85

90

95

X-ray absorption spectra at the Au L3 edge (11.9 keV) were recorded at the GILDA-BM08 beamline at the European Synchrotron Radiation Facility (ESRF), using a Si (311) doublecrystal monochromator and Pd-coated mirrors for higher-order harmonics rejection. Absorption spectra have been collected in fluorescence mode with a 13-elements Ge detector. All measurements were performed at liquid nitrogen temperature (LN) to limit the thermal disorder effects which damp the EXAFS oscillations, thus allowing the collection of high quality data well above the instrumental noise in a wide range of k. Standard Au foil and AuCl3 have been measured as reference samples. Downstream to all the samples, the transmitted beam has been used as the incident beam to measure the X-ray absorption spectra of a standard Au foil, so to obtain a very accurate edge energy calibration for all the EXAFS data sets. Data extraction and EXAFS analysis has been carried out using the GNXAS package.34 Further details about EXAFS analysis are reported in the supporting information section. Grazing Incidence Small Angle X-ray Scattering (GISAXS) GISAXS experiments have been carried out at the beam line BM26B-DUBBLE at the ESRF35, 36 using a newly developed GISAXS set-up. A sample-to-detector distance of 1.03 m with an X-ray wavelength of λ = 0.124 nm and a beam size of 300×100 µm (horizontal×vertical) have been used to perform the measurements. 2D-GISAXS images have been recorded using a Photonic Science CCD camera with pixel size of 44×44 µm. An incident angle αi = 0.6°, well above the nominal gold critical angle αc (0.443° for a wavelength of 0.124 nm), has been used. Transverse q∥ scans have been obtained from the 2D-GISAXS images using the 2D software package Scatter37 at fixed q┴ = (2π/λ)[sin(αf≈αc)+sin(αi)], i.e. at the height of the Yoneda peak.38 αi, αf and αc are the incident, exit and critical angles, respectively. q∥ intensities have been analyzed in order to study the lateral structural dimensions and the in-plane ordering between deposited Au nanoclusters. For incident angles αi >> αc, the scattered intensity in the frame of the Distorted Wave Born Approximation simplifies, and for a constant q┴ cut, the Fresnel transmission functions for the incident and reflected beams reduce just to scaling factors.39-41 Atomic Force Microscopy (AFM)

100

105

The surface morphology of the deposited gold clusters has been characterized by AFM in Intermittent Contact Mode. Images have been recorded with a standalone atomic force microscope (SMENA NT-MDT, Moscow) operated under ambient conditions (ambient temperature with a relative humidity of 55%). Silicon cantilevers (NT-MDT NSG10) with typical resonant frequency of 255 kHz and tip curvature of 10 nm have been used. All topographic images have been corrected by second-order polynomial fitting to remove the background trend and step line correction to correct misaligned lines. Image analysis has been This journal is © The Royal Society of Chemistry [year]

Physical Chemistry Chemical Physics Accepted Manuscript

DOI: 10.1039/C3CP54847C

Page 3 of 8

Physical Chemistry Chemical Physics View Article Online

Published on 10 February 2014. Downloaded by Lomonosov Moscow State University on 10/02/2014 16:47:51.

5

10

15

20

25

30

carried out by means of the IMAGE-SXM 1.75 and Gwyddion software packages. The surface roughness, σ, of the deposited gold cluster layers was measured through the root mean-square (RMS) method, where σ represents the standard deviation of the average surface height. The grain morphology was analyzed quantitatively using the height-height correlation function H(r) and the autocorrelation function g(r). The autocorrelation function g(r) was used to measure the average grain size through the correlation length ξ that defines the lateral distance within which the heights of any two points are correlated. ξ was calculated by fitting the initial portion of g(r) versus distance r curves with a Lorentzian function.42 AFM images show an uniform gold covering of the substrate with no evidence of neither isolated small nodules nor large structures. Consequently the correlation length ξ closely reflects the average grain size and will be referred as grain size in the following. This procedure was applied to three different g(r) measured from topographic images recorded from different 5×5 µm surface regions for each sample, and an average value of ξ was determined. The surface morphology was also characterised using the H(r). H(r) was calculated according to the equation 2 H (r ) =  h( x) − h( x + r )  where h is the height distribution function over a scanned surface and r is the lateral distance along the fast scanning direction. The angle brackets denote an average over the entire area scanned in the AFM images. For r

Si substrates.

The structure of small gold nanoclusters (around 2.5 nm) deposited on different silica-on-silicon (SiOx/Si) substrates is investigated using several c...
1MB Sizes 2 Downloads 0 Views