An approach to form planar structures based on ferromagnetic Fe1 − xSix films is presented. Epitaxial Fe1 − xSix iron-silicon alloy films with different silicon content (x = 0-0.4) were grown on Si(111) substrates. Structural in situ and ex situ characterization of the films obtained was made by X-ray diffraction, reflective high-energy electron diffraction, Rutherford backscattering spectrometry and transmission electron microscopy, which confirmed single crystallinity and interface abruptness for all films. Etching rates in the wet etchant (HF: HNO3: H2O = 1: 2: 400) for the films with various chemical composition were obtained. A nonmonotonic dependence of the etching rate on silicon content with a maximum for the composition Fe0.92Si0.08 was discovered. Moreover, the etching process is vertical and selective in the etching solution, i.e., the etching process takes place only in silicide film and does not affect substrate. As an example, a four-terminal planar structure was made of Fe0.75Si0.25/Si(111) structure using the etching rate obtained for this silicon content. Magneto-optical Kerr effect (MOKE) microscopy and transport properties characterization indicated successful etching process.

No abstract

This work investigates the Si/Fe flux ratio (2 and 0.34) influence on growth of β-FeSi 2 polycrystalline thin films on Si(100) substrate at 630 ˚C. Lattice deformation for the films obtained are confirmed by X-ray diffraction analysis (XRD). The volume unit cell deviation from that of β-FeSi 2 single crystal are 1.99 % and 1.1 % for Si/Fe = 2 and Si/Fe = 0.34, respectively. Absorption measurements show that the indirect transition (~ 0.813 eV) of the Si/Fe = 2 sample changes to the direct transition with a bandgap value of ~0.806 eV for the sample prepared at Si/Fe = 0.34. Along with this direct transition, the absorption spectra exhibit an additional feature with an excitation energy of ~0.56 eV. Surface magneto-optic Kerr effect (SMOKE) measurements detect ferromagnetic behavior of the β-FeSi 2 polycrystalline films grown at Si/Fe = 0.34 at T=10 K, but no ferromagnetism was observed in the samples grown at Si/Fe = 2. Theoretical calculations refute that the cell deformation can cause the emergence of magnetization and argue that the origin of the ferromagnetism, as well as the lower impurity direct transition, is β-FeSi 2 stoichiometry deviations. Raman spectroscopy measurements evidence that the film obtained at Si/Fe flux ratio equal to 0.34 has the better crystallinity than the Si/Fe = 2 sample.

1225Studying thin magnetic metal films on semicon ductor substrates is now of particular interest due to new physical properties of these structures and the prospects of their application in spintronics [1][2][3]. The film thicknesses in those works were varied from several fractions of a nanometer to tens of nanometers. The technological growth parameters are usually cho sen using preliminary calibrations, which can substan tially change during growth. Therefore, it is necessary to control the layer thickness in such a structure during growth.The experimental techniques most widely used to perform in situ film thickness control during growth are as follows: thickness measurement from the differ ence between beats in a quartz resonator, the high energy electron diffraction, and ellipsometry. To obtain reliable film thicknesses using a quartz resona tor, it should be located as close as possible to a synthe sis region and should be thermally stabilized, which requires an additional complication of the experimen tal technique. Moreover, this method is inconvenient during the formation of multicomponent systems. The diffraction of fast electrons can be used to calculate the coating thickness from specular reflection oscillations during layer by layer growth of epitaxial films [4] and cannot be used for thickness measurements during the formation of polycrystalline and amorphous struc tures.Among these methods, ellipsometry is a sensitive method that can almost continuously obtain informa tion on the processes occurring on the surface of a growing sample without affecting its structure. Apart from thickness, an ellipsometric experiment can also give information on the optical properties, in particu lar, the depth profiles of optical constants in a nonuni form layer [5][6][7][8][9]. Modern fast ellipsometers can main tain an almost continuous ellipsometric data flow measured during growth, which opens up fresh oppor tunities for solving the inverse problem of ellipsometry in a nondestructive manner. This problem can rather easily be solved when the thickness dependence of the ellipsometric parameters is known [8]. In a real exper iment, however, the ellipsometric parameters are mea sured during growth as functions of time and can only roughly be estimated if the growth rate is calibrated.Shvets [9] considered the problem of restoring optical constant profiles from a "growth curve," namely, the ellipsometric parameters measured during the growth of a nonuniform layer. He proposed an algorithm to determine these profiles, which achieved good results provided the gradient of the optical con stants is small, G = λ|dN/dz| Ӷ 1, where N = n -ik is the complex refractive index and λ is the wavelength. This algorithm is based on the relative derivatives of the ellipsometric parameters determined experimen tally from the growth curve at a high density of mea sured points. For the reverse inequality (G ӷ 1), gen Abstract-An algorithm is developed to perform rapid control of the thickness and optical constants of a film structure...