MBE growth of (1 1 0) refractory metals on a-plane sapphire

The initial stages of growth of epitaxial (1 1 0) niobium, tantalum, and molybdenum films on (1 1 2 0) sapphire by molecular beam epitaxy, MBE, are analyzed in situ by reflection high-energy electron diffraction (RHEED), and ex situ by high-resolution electron microscopy (HREM) and X-ray diffraction (XRD) techniques. The RHEED analysis shows that both Nb and Ta initially deposit (for approximately the first two monolayers in the case of Nb and the first four monolayers for Ta) with an hexagonal surface symmetry that is not consistent with the normal bcc structure of these metals, in agreement with the previous result of Oderno et al. 1] for Nb on sapphire. On further deposition, the films are observed to relax into the normal bcc (1 1 0) structure. Intensity oscillations of the RHEED specular beam are observed during both stages of growth. The RHEED oscillations are damped out after approximately 20 monolayers of Nb (30 monolayers of Ta) as a steady-state surface roughness is reached. HREM analysis reveals that the initial hexagonal structure has transformed completely to the normal bcc structure in the thicker films. The epitaxial relationships at each stage of growth are identified and a model is suggested for the development of the initial hexagonal structure based on Nb–O bonding at the interface. As the thickness of the bcc film increases, strain relief is observed by the formation of misfit dislocations. The growth mode of Mo is found to be different from that of Nb and Ta. The initial deposit grows in a three-dimensional mode with poor atomic order; the RHEED patterns are not sufficiently distinct to identify whether the hexagonal structure is formed. However, the surface becomes progressively smoother as growth proceeds and thicker films of Mo are comparable in crystalline quality, as measured by X-ray rocking curves, to Nb and Ta. The different growth mode of Mo is attributed to greater mismatch with the sapphire substrate and possibly different M–O ionic bonding at the substrate interface.