A transmission electron microscopy study of low-temperature reaction at the Co-Si interface

Abstract

An efficient preparation method, which provides wedge-shaped cross-section transmission electron microscopy samples, has been developed. It was then used to investigate the structure of as-deposited cobalt multilayers on silicon substrates by rf plasma sputtering. It was found that an extended reaction takes place between Co and Si probably during the deposition. The cobalt atoms react with the silicon substrate to form an amorphous silicide layer. When the deposited layer is <3 nm thick, it entirely reacts with the substrate and can form an amorphous silicide as large as 5 nm. Above 4–5 nm thickness, growth of Co crystallites comes in competition with the formation of the amorphous silicide and limits it to 2 nm. The composition of this amorphous silicide is estimated to be Co2Si. In Co/C multilayers, the reactivity between the two materials is negligible, and the coalescence thickness of cobalt is 2–3 nm. At 2 nm, the cobalt layers are noncontinuous and very rough, whereas at 3 nm the critical thickness for crystalline nuclei coalescence has already been reached. The cobalt layers are then polycrystalline and have a reasonable roughness.

Linear growth kinetics of nanometric silicides in Co/amorphous-Si and Co/CoSi/amorphous-Si thin films

“…Our observations refine the earlier picture because we observed simultaneous growth of the Co 2 Si layer and consumption of the Co and CoSi layers. Moreover, in all previous investigations [10][11][12][16][17][18][19][20][21] the kinetic exponents for the growth of CoSi and Co 2 Si phases were found to be 0.5. In contrast, we observed a very definite deviation from the parabolic growth law.…”

“…Thin reaction layers can be observed at the Co/a-Si interfaces already in the as-deposited sample. According to the literature, [10][11][12] these are amorphous CoSi layers formed during the deposition. Bright-field TEM images, taken on a CoSi intermetallic layer deposited at 423 K onto NaCl, contained dark spots, indicating that the layer was partially amorphous.…”

“…It is known from the literature 11,12 that amorphous CoSi layers can be formed at the interfaces already during the deposition. Furthermore, it was also reported in Ref.…”

“…9 In order to avoid the difficulties which may originate from the nucleation of the new phase, we artificially built a thin layer of the CoSi phase among the initial constituents. Although there are indications in the literature [10][11][12] that such a phase ͑amorphous or crystalline͒ can already be formed during the deposition at the Si/ metal contacts, our main aim was to investigate samples having such a complete intermediate layer at the interface. Besides specimens produced in this trilayered geometry, biand multilayered specimens with usual preparation technique has also been studied for comparison.…”

Features of the magnetic properties of Co/Si/Co thin-film systems

“…The most inter esting experimental investigations of the structural and magnetic properties have been reported for Fe/Si [7⎯11] and Co/Si [10][11][12][13][14][15][16][17][18][19] thin film systems. Analysis of the available data showed that the antiferromagnetic exchange between iron layers in Fe/Si/Fe systems was observed for silicon layer thicknesses within 1.4-1.7 and 1.4-2.2 nm (see, e.g., [20,21]).…”