Porous SiC materials

Porous SiC material, to be used for various device applications, requires additional processing steps which are common for the fabrication of electronic devices. These steps include cleaning, etching, doping, oxidation, metallization, film deposition, annealing, etc. Such treatments of the porous layer can modify the properties of the porous layer and, moreover, the kinetics of the processes in porous material might be different from similar processes in nonporous substrates of the same material. 

The microstructure of an SiC-filter made from silicon infiltrated polymer foam is shown in Fig. 19. Cell size, cell geometry, and cell anisotropy is controllable during processing [283]. The structural variability of this material reaches from tube-like anisotropic to isotropic pore nets with several pore and bridge averages. This porous SiC material cannot only be applied to filters and membrane supports it can also be used as catalyst support or heat exchanger [284]. 

Work is ongoing to reduce defects in SiC material. One of the more interesting concepts is the reduction of defects through epitaxial growth on porous SiC substrates [64]. This approach has clearly demonstrated a reduction in intrinsic defects, as evidenced by photoluminescence measurements. It is too early to tell whether this technique can provide a path forward for the bipolar devices but it will clearly find its applicability in several areas where SiC will have a market. 

The thermal shock behaviour of (a) monolithic alumina, and (b) porous SiC at different sintering temperatures (reprinted from Figure 11.6 on p 212 of Ceramics Mechanical Properties, Failure Behaviour, Materials Selection by Munz and Fett, 1999, published with permission from Springer-Verlag GmbH. 

S.E. Saddow, M. Mynbaeva and M.F. MacMillan, Porous SiC technology, in Silicon Carbide Materials, Processing and Devices, Zhe Chuan Feng and Jian H. Zhao (Eds), Taylor and Francis, New York, 2004, Chapter 8, pp. 321-385. 

In order to produce SiC material of the level of quality required for device applications, chemical vapor deposition (CVD) is currently used as the primary growth technique for SiC epitaxy [2], Due to the continuous improvements in commercial substrate quality, the presence of micropipes in SiC epilayers is not the device yield limiting issue as it was a decade ago. However, the epitaxially grown SiC films still suffer from other extended defects such as basal plane and threading edge dislocations as well as point defects. The vision of growing SiC on porous SiC was to reduce the concentration of these defects and thus improve the epitaxial layer quality for device applications. 

The porous SiC is fabricated from commercial SiC substrate (4H or 6H) by electrochemical etching. An electrolyte is placed in contact with the SiC substrate. A bias is introduced across the electrolyte and the semiconductor materials causing a current to flow between the electrolyte and the semiconductor material. The SiC partially decomposes in this electrolyte and forms high density of pores with nano-scale diameter. This decomposition initiates from the carbon-face of SiC substrate because the carbon-face is less chemically inert compared with the silicon-face. These as-etched pores have a depth of approximately 200 pm but do not reach the silicon-face of SiC. To fabricate porous silicon-face SiC (silicon-face is used as the growth plane for GaN), SiC with thickness of tens of micrometers is polished away from the silicon-face to expose the surface pores. Two surface preparation procedures, hydrogen polishing and chemical mechanical polishing, have been applied to the as-polished silicon-face porous SiC to improve its surface perfection. 

Note that when designating the fabricated porous structures in this way, we do not follow classification recommended for structural characterization of porous inorganic materials by the International Union of Pure and Applied Chemistry (IUPAC) [6], but simply proceed from the actual size of the pores. Classification of porous structures fabricated in SiC cannot yet be considered as established [7,8]. 

Briefly, three points of porous SiC-based catalytic support properties can be emphasized (i) SiC shows very good mechanical properties which gives resistance to erosion and attrition, in addition to a high thermal stability (ii) SiC has a higher thermal conductivity compared with the more conventional supports which could prevent the metal sintering (iii) SiC is particularly inactive with respect to chemical reagents such as acids or bases. Therefore, the active phase can be easily reprocessed after simple acidic or basic treatments. Among refractory materials, the thermal conductivity of silicon carbide, SiC (500 W m-1K-1 for crystalline state, at room temperature) is close to that of metals such as Ag or Cu (400-500 Wm K-1). 

Much work has been done with polymers to create biocompatible surfaces that resist protein fouling, i.e., non specific adsorption. A membrane composed of modified polyethersulfone (PES) (Omega Membrane) resisted albumin (ALB) absorption 24 times better than unmodified PES, and three times better than regenerated cellulose, a material known to resist biofouling. We compared porous SiC, both n-type and p-type, to the Omega Membrane [14] and found a very similar resistance to ALB absorption (Figure 12.3). 

Figure 12.5(a) and (b) shows the results of the diffusion of the six proteins through our n-type and p-type porous SiC membranes, respectively. Membranes of both n-type and p-type SiC exclude proteins in the same size range. The n-type material allowed much more protein to diffuse through, by a factor of as much as four times (for myoglobin). We have not yet investigated the effect of membrane thickness. Each membrane... 

Porous GaN can be produced without using electrical excitation as described in Chapter 4. As for porous SiC, there is potential for using the porous GaN as a substrate for homo-epitaxial growth of GaN. Other applications with strong potential include chemical sensors, with the large surface area of the porous material providing enhanced sensitivity of such devices. 

Chapters 5-8 discuss uses of porous SiC, and porous intermediate layers of other materials, as substrates for GaN epitaxy. The lattice mismatch between GaN and SiC leads to dislocations in GaN films, and use of a porous template offers a mechanism for reducing the dislocation density. 

Important electrical properties of porous SiC are discussed in Chapter 9. Not surprisingly, porous SiC has a higher specific resistivity than its host crystals. More significantly, pores appear to trap charge carriers, which renders p-SiC semi-insulating. Such effects strongly influence the use of porous material as electrical sensors. 

Thin walled structures made of C/SiSiC and SiC materials based on biocarbon and carbon/carbon preforms showed excellent long term stability in porous burner systems and are in development for... 

For this article, the term melt infiltrated ceramic matrix composite (MI-CMC) will refer only to continuous fiber composites whose matrices are formed by molten silicon (or silicon alloy) infiltration into a porous SiC- and/or C-containing preform. GE holds nmnerous patents on the composition and fabrication of these materials, only a few of which are listed in reference 10. Such composites can be made from a variety of constituents and processes. A detailed description of the material variations and processes is also given in reference 1, so only an abbreviated description will be given here. 

The ceramic materials investigated in this paper are hot-pressed silicon nitride (HPSN), reaction-bonded silicon nitride (RBSN), zirconia-toughened alumina (AI2O3), and porous SiC. The available material properties for these ceramics are given in Table 4.3. The HPSN and zirconia-toughened alumina ceramics were in the shape of flexural strength test bars cut from billets. The RBSN ceramic was molded into bars of the shape of flexural test specimens. The porous SiC ceramic was provided in the shape of flexural test specimens. 

In recent years there has been tremendous interest in porous ceramics because of their applications as filters, membranes, catalytic substrates, thermal insulation, gas-burner media and refractory materials. These are due to their superior properties, such as low bulk density, high permeability, high temperature stability, erosion/corrosion resistance and excellent catalytic activity. One branch of this field is porous SiC ceramics, owing to their low thermal expansion coefficient, high thermal conductivity and excellent mechanical properties. However, it is difficult to sinter SiC ceramics at moderate temperatures due to their covalent nature. In order to realize the low temperature fabrication of porous SiC ceramics, secondary phases may be added to bond SiC. Oxidation bonded porous SiC ceramics have been found to exhibit good thermal shock resistance owing to the microstructure with connected open pores. 

In this study, three kinds of p-SiC powders with different particle sizes were used as the starting materials, and the effects of p-SiC particle size and sintering temperature on microstructure of porous SiC ceramics based on in-situ grain growth were investigated. 

---