(Invited) Sublimation Growth of 3C-SiC: Status and Prospects

Abstract

Cubic silicon carbide (3C-SiC) has been considered suitable for a number of device applications like MOSFETs, MEMS, solar cells, water splitting devices, and etc. However, the growth technologies for 3C-SiC are lagging behind its hexagonal counterparts (6H- and 4H-SiC). A primary reason is the lack of high quality 3C-SiC seeds which could be explored for the growth of bulk crystals or homoepitaxial layers. Although, some thermodynamic aspects on 3C-SiC stability should be accounted as well. During the initial developments of hexagonal SiC crystal growth techniques the hexagonal SiC seeds were grown by Acheson or Lely processes. Unfortunately, the 3C-SiC is rarely obtained using these processes. Therefore, it has to be heteroepitaxially grown on silicon or SiC substrates. In both cases there are still challenges in controlling the formation of structural defects at the 3C-SiC/substrate interface. A reproducible growth of 3C-SiC with a crystal size and crystalline quality similar to the hexagonal equivalents has not been demonstrated. Commonly, growth studies of 3C-SiC on SiC substrates by sublimation techniques were done using nominally on-axis hexagonal SiC surfaces. However, this approach has not succeeded in growth of crystals with reasonable structural quality. The main reason is the mismatch of symmetries between the SiC (0001) and 3C-SiC (111) which induces rotational twinning and formation of structural defects – double positioning boundaries (DPBs). Such defects could be avoided only if one 3C-SiC domain or several with the same orientation nucleate on the surface. In contrast to on-axis substrates, the off-orientated substrates have been a typical choice for the growth of homoepitaxial hexagonal SiC layers or bulk crystals. The off-oriented surfaces contain high density of steps which ensures reproduction of the substrate polytype. This has been considered by most of the researchers as an obstacle towards stable 3C-SiC growth. However, under certain conditions, off-oriented SiC substrates can be used to grow 3C-SiC with well controlled initial nucleation. In this paper we will present our recent results on the growth of thick (~1 mm) 3C-SiC layers on off-oriented SiC substrates using sublimation epitaxy. The design of graphite crucible was optimized using numerical simulations to obtain high polytype stability and reproducible growth of 3C-SiC at temperatures below 2000oC in vacuum (10-5 mbar). Based on our knowledge of the initial 3C-SiC nucleation on nominally on-axis and low off-axis surfaces we developed and optimized 3C-SiC growth approach using 4H-SiC (0001) surfaces with the off-orientation of 4 degrees. We will also present results on the growth of 3C-SiC on 6H- and 15R-SiC. The 3C-SiC layer grown using this approach pass through three interconnected growth stages which include formation of a large terrace/facet with an on-axis area at the edge of the sample, preferential nucleation of 3C-SiC domains on the large terrace, and their lateral enlargement along the step-flow direction. In order to obtain a single domain layers we have developed a two-step growth process and applied geometrical control using spacers, separating the source and the substrate, with different geometries. The HRXRD and LTPL analysis of 3C-SiC layers indicate high crystalline quality. The average value of the full width at half maximum (FWHM) of ω rocking curves is about 40 arcseconds. The LTPL spectra demonstrate well resolved near-band-edge features and lines associated with multiple bound-exciton complexes with up to four electron−hole pairs. The layers are usually unintentionally doped with nitrogen and exhibit resistivity of about 10-50 Ωcm. Furthermore, such layers were used as substrates for sublimation growth of homopeitaxial layers with intentional doping of B and V. The dopants were introduced by co-doping from the source material. Single domain 3C-SiC layers were used as seeds to explore the growth of bulk (thickness 4-5 mm) material by sublimation. The absence of DPBs allows us to analyze the development of a bulk crystal and the effect of other defects like stacking faults, secondary inclusions and threading defects on the crystalline quality of the bulk material. We will present our recent results and discuss further prospects for the growth of bulk 3C-SiC crystals.