The Adsorption and Decomposition of Gallane Adducts on GaAs (100)
- PDF / 425,766 Bytes
- 6 Pages / 420.48 x 639 pts Page_size
- 85 Downloads / 189 Views
THE ADSORPTION AND DECOMPOSITION OF GALLANE ADDUCTS ON GaAs (100) J.S. FOORD*, A.T.S. WEE, N.K. SINGH, T.J. WHITAKER AND D. O'HARE University of Oxford, Physical and Inorganic Chemistry Laboratories, South Parks Road, Oxford OXl 3QZ, United Kingdom. * corresponding author.
ABSTRACT The adsorption and decomposition of GaH 3NMe 3 and GaH3PMe 3 on GaAs (100) has been investigated using XPS, TDS and molecular beam techniques. Both precursors react in a similar way with the adsorbed layer formed at low temperatures decomposing as the temperature is raised to produce metallic Ga, gaseous gallium hydride(s) and adsorbed hydrogen and the Lewis base, which also undergo desorption. Molecular beam measurements show that both compounds exhibit similar decomposition kinetics and decompose more readily than the aluminium analogues. The results indicate that the site-blocking effects, which prevent the low temperature decomposition of alane adducts, are less significant for the gallane compounds. Introduction
The growth of semiconductors by metal organic vapour phase epitaxy (MOVPE) and chemical beam epitaxy (CBE) now represent major techniques for the controlled deposition of III-V epilayers for semiconductor devices (1). This has led to great interest in the development of metal organic compounds which display clean, efficient decomposition routes to the required elements. MOVPE is carried out under reaction conditions where a complex range of surface and gas phase reactions between gp III and gp V precursors leads to formation of the desired compound semiconductor. In contrast, CBE takes place at low pressures; the reactions involved occur exclusively in the adsorbed phase at the growing surface and the rate-limiting step is thought to be the unimolecular reaction: GpIIL3 (ads) -+ GpIII (ads) + 3L' (g) [1] It has proved extremely difficult to identify satisfactory precursors for use in CBE. Two problems arise in particular. Firstly reaction 1 proceeds via a series of Ga-containing intermediates when Ga alkyl compounds are used and some of the intermediates are volatile (2). As a result desorption takes place and only a fraction of the incident Ga flux is converted to surface Ga. Problematically the branching ratio for desorption:cracking depends sensitively on the growth conditions (temperature, Gp V flux etc) and on the growth surface (Ga-In-As vs GaAs etc) (2) and this means it is very difficult to achieve high reproducibility in the CBE growth of complex phases when Ga is present as one of the components. The second problem concerns background C incorporation levels. High purity films can only be grown according to reaction 1 above if pathways exist whereby the volatilising ligands are eliminated efficiently from the metal centre and desorbed into the gas phase. Radical desorption and/or P-hydride elimination/desorption do take place quite efficiently to remove the ligand from the surface, when L is an alkyl group, but nevertheless other irreversible cracking processes tend to occur to deposit C on the surface (3). Conseq
Data Loading...
