St Petersburg, Russia, May 27-29, 2002 | C.009p |

Lebedev Physical Institute, Moscow, 119991, Russia

X-ray reflectometry at small grazing angles is now used by many research teams for non-destructive characterization of surface layers and multilayer structures. It gives important information about surface and interface roughness, layer thickness and compositions, interdiffusion, and multilayer period. However, for many important cases the standard reflectometry technique with monochromatic radiation doesn't provide reliable results or can't be applicable.

The first problem is calibration of reflected radiation. If the sample size is 1 cm, that is typical for routine laboratory practice, a correct calibration of specularly reflected intensity is difficult or impossible, as the sample surface doesn't completely intercept the incident beam. Secondly, when a scattering diagram is measured from strained structures, it's impossible to distinguish correctly contributions of roughness and interface curvature in diagram broadening. Thirdly, for locally processed samples due to inhomogeneous illumination contributions of different parts of reflective structure can't be evaluated properly.

In our recent work we demonstrated that all above-mentioned problems could be solved if to meet the following requirements to the instrumentation and procession algorithm: a – to measure reflected or scattered radiation simultaneously in two or more narrow spectral bands; b – to use for calculations and inverse problem solution the ratio of intensities for chosen spectral bands; c – to provide identity of two-dimensional distributions of the X-ray flow density in normal cross-sections of the incident beam for both spectral lines (bands) at any work parameters of an X-ray source.

In this report we first present results of measurements and data procession received by the new technique for small and irregular form samples of SiGe/Si, AlGaAs/GaAs multilayer structures, and a diamond substrate locally processed by ion implantation.

The work is supported by RFFI No 02-02-17349, No 01-02-17432.

© Educational Centre at Ioffe Institute

(Search|About) Images: 1; Size: 3237; TeX size: 2595; Update: 27 May 2002; Converted: 28 May 2005 00:59:34; elapsed time: 0.027 sec.