Strain and Interdiffusion Profiles in Epitaxied AU/NI (100) Multilayers Deduced From X-Ray Diffraction Experiments
- PDF / 319,999 Bytes
- 6 Pages / 414.72 x 648 pts Page_size
- 47 Downloads / 185 Views
379 Mat. Res. Soc. Symp. Proc. Vol. 356 01995 Materials Research Society
with Ni thicknesses ranging from 1 to 4 ML [3]. HREM cross section images have been quantitatively analysed to study the local distortion of the {002} planes near the Ni area. The profiles have been found to be asymmetrical and spread out over a distance larger than the nominal number of Ni ML. From these experimental strain profiles and by using elasticity theory, it has been deduced that a strong interdiffusion between gold and nickel has occured during the growth, induced by the high biaxial stress in the epilayer. In the present study, X-ray diffraction (XRD) experiments have been carried out on periodic multilayers. A structural refinement method has been used to determine the strain profile as well as the chemical composition near the Ni layers. EXPERIMENTAL
DETAILS
The samples were prepared by electron beam evaporation in an ultra-high vacuum (10-10 Torr range). Four multilayers were prepared on GaAs(100) substrates. The substrates were chemically etched with the standard procedure before annealing in vacuum at about 600 'C for some minutes. A seed layer of Fe (1.5 nm) was deposited before the growth of a 60 nm-thick Au(100) buffer layer. RHEED (reflection high energy electron diffraction) oscillations were recorded to calibrate both the Ni and Au growth rates. All the periodic multilayers were deposited at room temperature (RT). The designed thicknesses of the Au layers in the periodic multilayers were always 1.5 nm. The designed Ni thicknesses were respectively 1.5 ML, 2 ML, 4 ML and 5 ML. X-ray diffraction experiments were carried out on a conventional 2-circles diffractometer equipped with a graphite monochromator. We used the Copper KX radiation (X = 1.5418 A). XRD MODELING In this work, the entire X-ray diffraction profile was fitted with a structural refinement program which uses both the relative intensities and the line profiles to determine the average structure of the bilayers. We have basically followed the kinematical model proposed by Fullerton [5]. This model includes structural fluctuations which may occur in the superlattice and which are the main cause of the broadening of the superlattice peaks. We will recall the basic ideas that we use in our model and which are detailed in ref. 5 . The scattering factor F for a multilayer with N bilayer is N
F = •exp(iqxi)[FA, + exp(iqtAi)FB] (1) i=l
where xi is the distance between the substrate and the ith bilayer, FAi (FB i) is the scattering factor for layer i of material A (B), tAi is the thickness of layer Ai and q is the scattering
vector : q=4 t sin(0)
We used equation (7) in ref. 5 which gives the total averaged integrated intensity, I(q) = . This equation requires averaged quantities such as , ,
380
or . Because there is some deviation from an ideal layer-by-layer growth, and also because the number of atomic planes actually deposited in one layer may be a non-integer, lateral discrete layer thickness fluctuations have to be included when averaging FA and FB.
Data Loading...
