Strukturelle und elektrische Eigenschaften von epitaktischen, metallischen und halbleitenden Siliziden
Strukturelle und elektrische Eigenschaften von epitaktischen, metallischen und halbleitenden Siliziden
The transition metal silicide FeSi$_{2}$ is very interesting for basic research and also for possible device applications in silicon technology. FeSi$_{2}$ crystallizes in two phases, tetragonal, metallic $\alpha$-FeSi$_{2}$, which is stable above $\approx$ 950°C, and orthorhombic semiconducting $\b...
Saved in:
Personal Name(s): | Radermacher, K. (Corresponding author) |
---|---|
Contributing Institute: |
Publikationen vor 2000; PRE-2000; Retrocat |
Imprint: |
Jülich
Forschungszentrum Jülich GmbH Zentralbibliothek, Verlag
1993
|
Physical Description: |
171 p. |
Document Type: |
Report Book |
Research Program: |
Addenda |
Series Title: |
Berichte des Forschungszentrums Jülich
2792 |
Link: |
OpenAccess |
Publikationsportal JuSER |
The transition metal silicide FeSi$_{2}$ is very interesting for basic research and also for possible device applications in silicon technology. FeSi$_{2}$ crystallizes in two phases, tetragonal, metallic $\alpha$-FeSi$_{2}$, which is stable above $\approx$ 950°C, and orthorhombic semiconducting $\beta$-FeSi$_{2}$, which is stable at lower temperatures . Especially semiconducting $\beta$-FeSi$_{2}$ is of particular interest because early optical absorption experiments indicate that $\beta$-FeSi$_{2}$ should be a direct bandgap semiconductor with a gap of about 0.85 eV, an energy corresponding to a wavelength of 1.45 pm, which is close to the absorption minimum of silica optical fibres (1.55 $\mu$m). Thus it should be possible to produce opto-electronic devices in the infrared region integrated in the Si technology. The formation of homogeneous and continuous $\beta$-FeSi$_{2}$ layers with solid phase epitaxy is very difficult due to nucleation-controlled growth kinetics. For the first time it was demonstrated that the relatively new technique called Ion Beam Synthesis is one way to fabricate well defined epitaxial $\alpha$- and $\beta$-FeSi$_{2}$ layers buried below a single crystalline Si top layer. These layers are fabricated by Fe-implantation in heated substrates and subsequenttwo-step annealing. The first anneal for 10 s at 1150°C is necessary to obtain a continuous $\alpha$-FeSi$_{2}$ layer with sharp interfaces and without any remaining residual precipitates on either side of the layer. A second anneal for several hours below the transition temperature transforms the $\alpha$-FeSi$_{2}$ layer to a continuous $\beta$-FeSi$_{2}$ layer. Interestingly enough, the $\beta$-FeSi$_{2}$ layers show different epitaxial relationships to the (111)Si-matrix compared to $\beta$-FeSi$_{2}$ grown by solid phase epitaxy. Here the $\beta$-FeSi$_{2}$(010) plane is observed to be parallel or slightly off-oriented to S(111). Furthermore, the layers reveal a much lower defect absorption at room temperature below the conduction band as compared to samples fabricated by other techniques. The analysis of the data indicates a direct transition at 0.83 eV and for the first time an indirect transition at 0.78 eV. Thus the realization of efficient light-emitting devices will be unlikely with $\beta$-FeSi$_{2}$. But nevertheless the electrical properties of $\beta$-FeSi$_{2}$ layers show very interesting behaviour. Hall measurements indicated hole concentrations of 9 $\cdot$ 10$^{18}$ cm 3 and mobilities of $\approx$ 104 cm$^{2}$ / Vs at room temperature (for comparision the hole mobility in Si at such high carrier concentrations is $\approx$ 65 cm$^{2}$ / Vs). Moreover, by implanting additional Mn it is possible to fabricate semiconducting Mn$_{0.03}$Fe$_{0.97}$Si$_{2}$ which showed an increased hole concentration. These results appear promising for electrical device applications.In a further part the magnetoresistance of clean metallic CoSi$_{2}$ layers was measured. CoSi$_{2}$ is a favorite candidate among the metallic because of high-temperature stability, the very low specific electrical resistivity of = 14 $\mu \Omega$cm, and the epitaxial growth on Si (lattice mismatch only -1.2 %). The analysis of the magnetoresistance data was performed in terms of weak localization, Coulomb interaction and superconducting fluctuation. The CoSi$_{2}$ layers withthicknesses of 115 $\mathring{A}$ in (111)Si and 230 $\mathring{A}$ in (100)Si were also fabricated by Ion Beam Synthesis. The magnetic field dependence of the resitance could be interpretated in terms of two-dimensional weak localisation with strong spin-orbit interaction and an additional classical contribution proportional to B$^{2}$. Long phase coherence lengths of $_{\phi} \approx$ 0.75 $\mu$m in (111)Si and I$_{\phi}$ $\approx$ 2.3 $\mu$m in (100)Si at 4.2 K were determined by fitting the magnetoresitance data. The inferred inelastic scattering time is interpretated as a sum of a clean limit electron-electron process (dominant below 6 K) and an electron-phonon process dominant at higher temperatures. Further, a general orientation dependence of the electrical transport properties of these CoSi$_{2}$ layers is observed, such as anisotropy in the residual resistance, Hall coefficient and the prefactor for the classical B$^{2}$-dependence of the magnetoresistance.This could be related to muilpie band effects in CoSi$_{2}$. The measured relatively long coherent length in CoSi$_{2}$ indicate the potential of single-crystalline CoSi$_{2}$ for exploring quantum interference effects and the fabrication of novel microstructure devices in which quantum effects play an essential role. |