Rare-earth ion doped semiconductors are usually investigated as potential candidate materials for promising applications in optoelectronic devices. Many reports concern the structural and optical investigations of materials doped with optically active Ce3+ ions (e.g. YAG:Ce, Lu2SiO5:Ce, Ce-doped fluoride crystals) for applications as phosphors, scintillators and tunable lasers operating in the UV-blue region [1-4].
The Ce-doping of SiC is an interesting issue, because there is a lack of data in the literature, concerning this topic. There exists only one communication by Itoh et al. [5] concerning the PVT growth of bulk SiC crystals. For 4H-SiC crystals, which were grown using CeO2 or CeSi2 as the sources of the Ce impurity, the authors report the non-uniform concentrations of Ce atoms at levels less than 1017 cm-3 (which was a detection limit in their secondary ion mass spectroscopy (SIMS) measurements) up to 1020 cm-3. This constituted a motivation for us to search for a confirmation of these results, in particular, if such a high cerium doping levels are possible in SiC crystal grown by PVT method. Our goal was: (1) to confirm that the cerium vapor was present in the growth atmosphere, (2) to study the effects of possible cerium incorporation into the crystalline lattice of SiC, and (3) to investigate possible effects induced by the presence of the oxide impurity in the growth atmosphere.
In this study we present a detailed investigations on structural, optical and electrical properties of SiC bulk crystals, which were grown on the Si- or C-faces of the crystal seeds and in the presence of a different Ce impurity content (from 0.1wt% up to 5.0wt%) in the SiC source materials. As the source of the cerium we used a commercial cerium dioxide (CeO2) powder or granules of cerium silicide (CeSi2).
The SiC crystals were studied by using a variety of experimental techniques, such as: X-ray diffraction (XRD), energy dispersive X-ray spectroscopy (EDX), X-ray photoemission spectroscopy (XPS), scanning electron microscopy (SEM), optical absorption (OA), photoluminescence measurements (PL), secondary ion mass spectroscopy (SIMS) and the contactless method at microwave frequencies at 300 K (to characterize the electrical properties). The results obtained by EDX and XPS point out the condensation of cerium vapor species at the crystallization fronts of the SiC crystals upon cooling-crystallization process, after the effective crystal growth stage. When CeO2 was used as the source of Ce impurity, the cerium oxides Ce2O3 and CeO2 were found to coexist on the SiC crystallization fronts, whereas in the case of the CeSi2 source, there was found a mixture of CeO2 and some unidentified (at this stage) Ce-species on the crystallization fronts. The presence of the cerium oxides and Ce-species on the SiC crystallization fronts proves of the gradual dosage of cerium from the SiC source materials and also, confirms the continuous presence of the cerium vapor over crystallization fronts during the crystal growth processes. A possible concentration of Ce atoms in bulk SiC crystals is at the level less than 1017 cm-3 (which is a detection limit in our SIMS measurements), while at the SiC crystallization fronts it reaches a value of 0.1 at.% (estimated from the XPS results).
For SiC crystals grown in the presence of the cerium impurity, a very interesting results of the investigation of electrical properties have been obtained. The electrical properties of bulk SiC crystals turned out to be a completely different in two cases, when CeO2 or CeSi2 were used as the source of Ce impurity. For both 4H- and 6H-SiC crystals, which were grown in the presence of CeSi2, there were found the low values of resistivities at 300 K, within the range of ~0.02-0.57 Ωcm. From SIMS measurements it was found that the nitrogen donor impurity was a dominating impurity in all these crystals. For 4H-SiC crystals, which were grown in the presence of CeSi2 on the C-face of the crystal seed, there was observed a very low concentration of boron atoms ([B] ~1012 cm-3), in comparison with the SIMS results obtained for the reference SiC crystal ([B] ~1016 cm-3), which was grown without the cerium additive, on the C-face of the crystal seed, and with the same growth parameters. For 6H-SiC crystals, which were grown in the presence of CeSi2 on the Si-face of the crystal seed, the concentrations of boron atoms were at the same level as in the case of the undoped reference 6H-SiC.
For 6H-SiC crystals, which were grown in the presence of CeO2 on the Si-face of the crystal seed, there were found the high values of resistivities at 300 K. For some areas of different SiC wafers cut from these crystals, the high resistivity values up to ~1.70 x 104 Ωcm were observed. From SIMS measurements, the oxygen impurity was found to be a dominating impurity in the SiC wafers, besides the wafers, which were cut near the crystal seeds and thus represented the early stage of the crystal growth (they were the low-resistivity wafers with a dominating nitrogen impurity). But, regarding to a weak incorporation of Ce into the SiC lattice, it can be concluded that these high resistivities may be the result of a dominant oxygen impurity and the charge-carrier compensation effects between common residual donors (such as N or O) and acceptors (such as Al or B) in SiC, rather than from the impact of the Ce impurity. Moreover, some intrinsic defects and/or complex defects, e.g. B-O complexes, can provide deep levels in the band gap of SiC and, as compensating defects, they can take part in electrical compensation. The formation of electronically active Bs-O2i defect (consisted of a substitutional boron atom and an interstitial oxygen dimer), acting as a electron-hole recombination center, was reported in the literature e.g. for compensated p- and n-type samples of Czochralski-grown silicon wafers [6].
Acknowledgements: This work was supported by the Polish Research Center under Grant No. 5043/B/T02/2011/40.
References:
[1] Y.C. Kang et al., Mater. Res. Bull. 35 (2000) 789.
[2] G. Ren et al., Nucl. Instr. Meth. Phys. Res. A 531 (2004) 560.
[3] F. Okada et al., J. Appl. Phys. 75 (1994) 49.
[4] C.D. Marshall et al., J. Opt. Soc. Am. B 11 (1994) 2054.
[5] A. Itoh et al., Appl. Phys. Lett. 65 (1994) 1400.
[6] J. Geilker et al., J. Appl. Phys. 109 (2011) 053718_1.
|