STUDY OF THE LOCATION OF IMPLANTED FLUORINE-ATOMS IN SILICON AND GERMANIUM THROUGH THEIR NUCLEAR-QUADRUPOLE INTERACTIONS

Time Differential Perturbed Angular Distribution (TDPAD) measurements of the nuclear quadrupole hyperfine parameters for F-19 implanted int o amorphous, polycrystalline and crystalline silicon and germanium are reported and reviewed. Two signals are observed in the crystalline ma terials (approximate to 35 and 23 MHz in silicon, approximate to 33 an d 27 MHz in germanium) while only one is detected in the amorphous and polycrystalline samples (approximate to 22 MHz in silicon, approximat e to 27 in germanium). Impurity sites in these materials were modeled using a Hartree-Fock cluster procedure. The Intrabond, Antibond, and S ubstitutional sites in the bulk were studied in both silicon and germa nium. The ATOP and Intrabond Surface sites were also studied in silico n and the results extended to germanium. Lattice relaxation effects we re incorporated by employing a geometry optimization method to obtain minimum energy configurations for the clusters modelling each site. Th e electronic wave functions were obtained for each optimized cluster b y applying Unresctricted Hartree-Fock theory, and these wave functions were used to calculate the nuclear quadrupole hyperfine parameters at the sire of the fluorine nucleus. Comparison of the theoretical hyper fine parameters to the experimental values indicates that F-19 locate d in the Intrabond and Intrabond surface sites could readily explain t he higher frequency signal that has been observed. F-19 in the Antibo nd and the surface ATOP sites yield hyperfine parameters consistent wi th the low frequency signal observed in the crystalline materials and the single signal observed in the amorphous (or polycrystalline) mater ials. Examination of these two sites, in view of other available exper imental evidence including the temperature dependence of the TDPAD sig nals, leads to the conclusion that the lower frequency signal is due t o F-19 implants which have come to rest at the site of dangling bonds in the bulk, These dangling bonds are created as a result of damage g enerated in the individual collision cascades during the implantation process.