The electrical degradation of 4H–SiC PiN diodes has recently attracted much interest and is a critical material problem for high power applications. The degradation is caused by stacking faults observed as an increased forward voltage drop after forward injection operation. In this article we have combined electrical, optical, and structural techniques to study the formation and growth of the stacking faults causing degradation. We will show three different sources causing two different types of stacking fault properties.

4H-SiC commercial wafers and sublimation grown epitaxial layers with a thickness of 100 μm have been studied concerning crystalline structure. The substrates and the epitaxial layers have been separately investigated by high-resolution x-ray diffraction and synchrotron white beam x-ray topography. The results show that the structural quality was improved in the epitaxial layers in the [112̄0] and [1̄100] directions, concerning domain distribution, lattice plane misorientation, mosaicity, and strain, compared with the substrates. Misoriented domains have merged together to form larger domains while the tilt between the domains was reduced, which resulted in nonsplitting in diffraction curves. If the misorientation in the substrate is large, we can only see a slight decrease in the misorientation in the epilayer. At some positions on the substrates block structures (mosaicity) were observed. ω-rocking curves showed smaller full width at half maximum values and more uniform and narrow peaks, while the curvature was almost the same in grown epilayers compared with the corresponding substrates. We show that threading edge dislocations along the c axis in silicon carbide grown crystals transform to deflected dislocations in the epilayer. A formation mechanism for deflected dislocations and supporting facts are presented. We further show that these deflected dislocations are one possible source for the creation of stacking faults that recently has been reported to cause degradation in processed SiC bipolar diodes.

Triangular structural defects are occasionally generated during the long-term operation of 4H-SiC pin diodes and degrade the forward characteristics of the diode. We have used synchrotron white beam x-ray topography, scanning electron microscopy, in situ cathodo luminescence, and transmission electron microscopy to characterize the structure and formation of these defects. It is shown that the defects are stacking faults on the (0001) basal planes, bound by partial dislocations with Burgers vectors 1/3〈101̄0〉 and 1/3〈011̄0〉. These partials are suggested to form by the dissociation of existing dislocations.

Articles you may be interested inA method to determine fault vectors in 4H-SiC from stacking sequences observed on high resolution transmission electron microscopy images J. Appl. Phys. 116, 104905 (2014); 10.1063/1.4895136 X-ray microbeam three-dimensional topography for dislocation strain-field analysis of 4H-SiC

High quality double-position-boundaries free 3C-SiC epilayers have been successfully grown on on-axis (0001) 4H-SiC by chemical vapor deposition at optimized conditions as observed with optical microscopy and X-ray diffraction. The effect of the growth parameters, including temperature, C/Si ratio, ramp-up condition, Si/H-2 ratio, N-2 addition and pressure, on the quality of the grown layers is investigated. Different techniques, including microscopic and spectroscopic techniques, are used to characterize the epilayers. High resolution X-ray diffraction shows 2 theta-omega curve with full width at half maximum of only 16 arcsec for the (111) reflection detected from a 35 mu m thick 3C-SiC layer, showing the good structural quality of the layer. Reciprocal space maps confirm the absence of double-position-boundaries in a large depth of the layers. Low temperature photoluminescence measurement shows clear near-bandgap emission with sharp and single peaks, which further verifies the high quality of the epilayers

Changes in the environment of europium ions in solution bring about changes in the structure of the absorption spectrum. For example, an aqueous solution of europium chloride upon having potassium nitrate added to it gives rise to the structures in the spectra of europium nitrate, in addition to the spectrum of europium chloride. Structures in the spectra of europium salts in alcohol differ from those in water. In a mixed solvent of water and alcohol both spectra are present with relative intensities which vary according to the relative amounts of the components. Europium chloride in the mixed solvent has a spectrum which cannot be distinguished from the exact superposition of the spectra of europium chloride in pure water and in pure alcohol. However, the spectrum of europium nitrate in the mixed solvent has its lines shifted from the positions they occupy in the spectra of the nitrate in the pure solvents. The shifts seem to vary continuously with the proportions of water to alcohol. Nevertheless, the lines remain sharp as successive amounts of water are added, with no suggestion that the lines are a blur, between the spectra of the limiting concentrations. The implications of these results are briefly discussed in their bearing upon the behavior of electrolytes in mixed solvents.

No abstract

No abstract