Fe-Silicides Formed with Molecular Beam Epitaxy and Studied with Conversion Electron Mössbauer Spectroscopy

Abstract

Fe-sulicides were formed by annealing MBE-deposited thin 57 Fe layers with thicknesses between 20 A and 60 A on (7x7) reconstructed Si(1 11) substrates. During the growth the substrate was held at room temperature. The silicide formation upon annealing in the temperature range of 200'C to 900'C was studied in-situ with RH-EED. Samples were studied with in-situ CEMS (Conversion Electron Mossbauer Spectroscopy) as well as with ex-situ CEMS after covering with Ag to prevent oxidation. RBS/Channeling was used to study the epitaxial quality and the structure of these layers.Using the 57 Fe nuclear probe, it is shown that silicide formation occurs at the interface already at room temperature. Metastable silicides with cubic structure are observed in the low temperature annealing range, and characterized by their Mossbauer parameters. At higher temperatures stable e-FeSi and r3-FeSi 2 are formed.

Iron silicide formation at different layers of (Fe/Si)3 multilayered structures determined by conversion electron Mössbauer spectroscopy

“…However, Fe defective non-stoichiometric Fe 1Àx Si, or strain at the interface, give rise to a distribution of locally noncubic site symmetry, which is reflected in CEMS spectra as a doublet of large linewidth. 11,29 In the present spectra, the fitted doublets show quadrupolar splittings ranging from QS ¼ 0.65(2) to 0.67(1) mm/s, isomer shifts from d ¼ 0.223(3) to 0.25(1) mm/s (with respect to a-Fe) and large linewidths LW ¼ 0.55(1) to 0.62(3) mm/s. This is consistent with the defective/strained paramagnetic c-Fe 1Àx Si phase.…”

“…According to previous works on the Fe-Si interfaces, 3,5,11,12,15,[27][28][29][30][31] these compounds would have a magnetic Fe 1Àx Si x alloy, and a paramagnetic silicide. The non-stoichiometric c-Fe 1Àx Si phase (0 x 0.5) has been proposed as that paramagnetic compound.…”

“…The non-stoichiometric c-Fe 1Àx Si phase (0 x 0.5) has been proposed as that paramagnetic compound. 11,27,29,30,[32][33][34][35] We focus our attention on the analysis of the paramagnetic component given by the doublets. The c-FeSi phase with the cubic CsCl structure should show a singlet.…”

Morphology of the asymmetric iron–silicon interfaces

“…This fitted density is similar to that of the stable stoichiometric -FeSi phase (q ¼ 5:19 g/cm 3 [29]), but it also lies between the value 4.73 g/cm 3 of the highly Fe-defective and epitaxially stabilized c-Fe 0.5 Si [30] and the close to stoichiometric c-Fe 0.88 Si [31], both with the CsCl structure. c-Fe 1Àx Si has been claimed to appear in the Fe/Si interface [18,19,12,15]. The fitted thickness has to be compared to a previous determination by GIXRR [14], where the Si-on-Fe interface showed also a thickness of 1.4 nm and the fitted optical constant turned out to be that of FeSi.…”

“…Different silicide phases like FeSi [16,17,14], c-Fe 1Àx Si [18,19,12], FeSi 2 [20,17,15] and Fe 3 Si [11,21,22] have been proposed to be formed at the Fe-on-Si interface. In contrast, in the Si-on-Fe interface some authors report only the presence of the paramagnetic c-Fe 1Àx Si (0 6 x 6 0:5) [11,19,12,23], while others state that a ferromagnetic Fe rich Fe(Si) solid solution is also present [18,15,24]. The contradictions may be caused by differences in the samples used; specifically, the substrate and the number and thickness of the Si and Fe layers.…”