Heterojunction bipolar transistors (HJBT) on the basis of GaN/SiC heterostructures have several advantages over group III nitride heterostructures grown on sapphire. For example 6H-SiC has less of a thermal and structural mismatch to GaN than sapphire has. Furthermore there is not the problem of optical recombination in a highly doped base region as there is for the group III nitrides on account of their direct bandgap. However, despite the well acknowledged potential of the GaN/SiC material system there are still unanswered questions relating to the technology used as well as to some fundamental properties of GaN/SiC heterostructures. Therefore we investigated epitaxial growth and physical properties of n-GaN/p-SiC heterojunctions with respect to their significance to n-GaN/p-SiC/n-SiC HJBT. We grew n-type GaN (n = 10(18) cm(-3)) on p-type (p = 2 x 10(18) cm(-3)) and n-type (II = 4 x 10(18) cm(-3)) 6H-SiC substrates in a horizontal hot wall reactor. This approach is very similar to the more common HVPE. Instead of synthesizing GaCl in situ from HCl and metallic Ga we used GaCl3 as the Ga precursor. All our experiments were carried out at low pressures around I mbar resulting in a good homogeneity. As it is common for the more usual HVPE we grew GaN without a buffer layer. From thermal admittance spectroscopy (TAS) as well as temperature dependent I-V characteristics we gained knowledge about deep level defects and the role of interface traps. The microstructure of the interface was investigated by transmission electron microscopy (TEM). Furthermore we present details about device processing by ion beam sputter etching with carbon dioxide (CO2) as the working gas.