Structure of C o A g nanoparticles embedded in a matrix

Abstract

Since CoAg nanoparticles are composed of two immiscible metals, their theoretical structure is predicted to be Co@Ag due to the weaker surface energy of Ag [1]. In the literature these particles are synthesized by chemistry techniques [2‐3], the structure obtained is Co(core)Ag(shell) and has been confirmed by the optical response in UV‐Vis spectroscopy [2]. The localized surface plasmon resonance (LSPR) of silver shell and the magnetic anisotropy of cobalt core open the access to magneto‐plasmonic studies. That is why we have decided to elaborate those particles in a different approach using physical method under ultra high vacuum (UHV). We have elaborated by low energy cluster beam deposition (LECBD [4]) technique Co 50 Ag 50 nanoparticles of different diameter (approximatively 1 to 10 nm) in UHV which are mass selected by a quadrupole ion deflector. These clusters are embedded in different matrices (metal oxides, amorphous carbon) to protect them from oxidization. The clusters stoichiometry has been checked by EDX measurements. The structure of Co 50 Ag 50 clusters in carbon matrix is polycrystalline fcc. On high resolution transmission electron microscopy (HRTEM) images we can distinguish two different phases between the top and the bottom of the particle (figure 1.a) by using the intensity as a discrimination factor. Then we have determined by Fourier transforms analysis that the silver is in fcc structure and we measured approximatively the same lattice parameter than the silver bulk. Nevertheless we could not identify expected bcc, hcpor fcc [5]crystallinecobalt structure. High angle annular dark field electron microscopy (HAADF) measurements provided us informations about the contrast between cobalt and silver (figure 1.b), which depends of their respective atomic number. Inhomogeneous Janus‐like structures have been identified, with cobalt assumed to be in dark and silver in bright (figure 1.b), this segregation is partially in agreement with theoretical predictions [1] which expect a core/shell segregated structure. In Al 2 O 3 matrix another phenomena occurs, an amorphous shell appears on each particle. As we can see on TEM image (figure 1.c and 1.d), we can distinguish a core in the particles composed of a dark part and a brighter one and an amorphous bright shell, assumed to be a reaction with the matrix. SQUID magnetic measurements confirmed us that the clusters are still ferromagnetic, an antiferromagnetic behaviour is expected for CoO, which means that the Co cannot be completely oxidized. As in carbon matrix, we can distinguish an inhomogeneous structure and an intensity contrast between the topand the bottom of the particle (figure 1.d). Then the silver is also in fcc structure and the lattice parameter obtained by Fourier analysis measurements matches with the bulk one. Moreover the same problem encountered in carbon matrix appears when we try to identify the Co crystalline structure.