MOVPE growth of InP by reaction of usual indium alkyls with phosphine displays different problems [ 11 : incomplete pyrolysis of phosphine, non volatile polymer deposit, hazardous manipulations of phosphine and trimethylindium, contamination of the epitaxial layers ...In view to substitute trimethylindium by a new organoindium precursor, trimethylindium-trimethylamine adduct was chosen. This is an addition compound between a Lewis acid, trimethylindium and a Lewis base, trimethylamine. Complex formation stabilizes trimethylindium and so improves security of storage and handling of this material ; adducts are generally non pyrophoric and relatively stable to air and moisture compounds. First we synthesized trimethylindium-trimethylamine adduct by simple original way through an organomagnesium compound [2]. The liquid trimethylindium-diethyloxide intermediate compound avoided the manipulation of base free bimethylindium.The solid adduct was first purified by vacuum sublimation ; its purity was then improved through zone refining. It was characterized by vibration spectroscopies (infrared and Raman) and proton nuclear magnetic resonance. Its decomposition temperature determined by Raman spectroscopy was found to be about 138 "C.Our vapor pressure measurements on trimethylindium, trimethylindiumtrimethylamine and DADI (3-dimethylaminopropyldimethylindium) and the comparison with some other indium precursors described in the literature shows that trimethylindiumtrimethylamine pressure is the closer to that of trimethylindium. We measured 1.4 1 hPa, 0.97 hPa and 0.29 hPa at 20 OC for trimethylindium, trimethylindium-trimethylamine and DADI respectively. The phosphine trialkyls adduct with trimethylindium have also lower vapor pressure than trimethylindium-trimethylamine [3].Calorimetric experiments and vapor pressure studies of trimethylindiumtrimethylamine adduct exhibits a melting point of 71.6 OC and 73.9 "C respectively, which is much higher than those given in the literature [4] (66.2-66.4 "C). This can be explained by the good purity of our product which has been purified two times ; moreover the quality of the source materials synthesized may have improved since 1956. This is confirmed by the pressure data we have found lower than those given by Coates [4] ; our pressure measuring procedure may eliminate the effect of volatile contaminants.The comparison of the behavior of trimethylindium and trimethylindiumtrimethylamine compound with phosphine by growing indium phosphide epitaxial layers is carried out.Trimethylindium-trimeth ylamine adduct has already been tested for epitaxial growth of indium phosphide [5,6] showing good quality of the layers obtained, 600 O C being the best temperature growth and 90 the best V/III ratio.
11