MOVPE Mechanisms from studies of specially designed and labelled precursors

Studies of the reaction or decomposition products of precursors for metal organic vapour phase epitaxy (MOVPE) do not always give enough information to allow the unequivocal determination of decomposition or growth mechanisms. By studying deuterium labelled precursors in the presence or absence of their protio analogues, other precursors and/or He, H-2, D-2 Or by studying precursors carrying substituents that are designed to give different products if different mechanisms operate, it is possible to draw more definitive conclusions.

Using these studies coupled with semi-empirical molecular orbital calculations, it is shown that primary arsines decompose by reductive elimination of H-2 followed by beta-abstraction ((BuAsH2)-As-t) or reaction with the parent arsine to form RAsH ., which undergoes reductive elimination (PhAsH .), beta-abstraction ((BuAsH)-As-t .) or As-C bond cleavage ((BuAsH)-As-t .). Hex-5-enylarsine has been used to show that adduct formation is not important during growth of GaAs.

For group 16 dialkyls (R2E, E = S, Se or Te), the predominant decomposition mechanism is homolytic E-C bond cleavage. Subsequent reactions involve abstraction of H from the beta-position of the intact R2E to give alkane, two molecules of alkene E and H . (E = Te or Se). For E = Te, H . does not react significantly with (Pr2Te)-Te-i, but for (Bu2Se)-Se-t a short chain-reaction is initiated by H .. The importance of free radicals is confirmed by studies of (but-2-enyl)(2)Te, (hex-5-enyl)(2)E (E = S, Se, Te), (pent-5-enyl)(2)Te and (hex-5-enyl)SH, as well as of secondary and tertiary analogues. Reactions of the labelled and designed group 16 precursors with Me2M (M = Cd or Zn) are also discussed.