Thermal Degradation of SiGe Interfaces Studied by X-Ray Reflectivity and Diffraction
- PDF / 391,155 Bytes
- 6 Pages / 420.48 x 639 pts Page_size
- 59 Downloads / 148 Views
THERMAL DEGRADATION OF SiGe INTERFACES STUDIED BY X-RAY REFLECTIVITY AND DIFFRACTION t J. M. HUDSON', A. R. POWELL*, D. K. BOWEN*, M. WORMINGTON , B. K. TANNER" R.A.KUBIAKt AND E.H.C.PARKER*. 'Dept. of Physics,University of Durham,Durham,DH13LE,U.K *Dept. of Engineering,University of Warwick, Coventry,CV4 7AL, U.K tBede Scientific Software Division,Coventry,CV4 7EZU.K. 'Dept. of Physics,University of Warwick, CoventryCV4 7ALU.K
ABSTRACT We demonstrate the use of x-ray diffraction to provide accurate compositional information, together with grazing incidence reflectivity to provide information on layer thicknesses and surface and interface roughnesses, on Si/Si,-xGex superlattice structures of less than 200nm total thickness. The quality of SiGe interfaces has been investigated in superlattices where x varies from 0.1 to 0.5. At low Ge compositions the interfaces are shown to be smooth to a few angstroms. However, as the Ge composition in the SiGe layer approaches 50%, severe roughness is observed at the SiGe to Si interfaces, although the Si to SiGe interfaces remain relatively smooth. Upon annealing for one hour at 850"C the Ge diffuses outwards from the SiGe layers and can be closely modelled by inclusion of a (2.4±0.3)nm linearly graded layer either side of the SiGe layer into a simulation program. The long range roughness at the SiGe to Si interface is lost upon annealing leaving only a short range roughness of similar size to the Si to SiGe interface roughness. Reflectivity measurements have been shown to distinguish between interface roughness and interdiffusion. for the annealed system. INTRODUCTION Double crystal diffraction (DCD) techniques offer a rapid nondestructive method of characterizing semiconductor materials. For crystalline epitaxial layers the technique can measure layer thicknesses to the Angstrom level as well as yielding information on layer compositions and strains [1]. However, for situations where the epitaxial layer is very thin ( 0crtii1 the reflectivity from
the top surface of an ideal sample varies as the inverse fourth power of the scattering vector. By multiplying the reflected intensity by a factor of 0' and considering the gradient, surface roughness can be determined. The X-ray reflectivity from the samples was simulated using the REFS simulation program discussed earlier. Included in the simulation was a 2.5nm oxide layer, present in order to model the effects of the thermal oxide growing upon the sample since growth [9]. The superlattice period was determined From the peak separation. Having obtained this period the effects of the modulating envelope are examined in order to pull out the individual layer thicknesses. A roughness at only one of the two types of interface alters the effect of the modulation envelope and a detectable change in the relative peak intensities occurs. For the high content Ge samples (ap1315 and ap1313), simulation of the preannealed data gives differing values for the interfacial roughness at the two different types of interface. It is not possible u
Data Loading...
