IR and NMR Study of the Chemisorption of Ammonia on Trimethylaluminum-Modified Silica

Abstract

The saturating reaction of ammonia was studied on trimethylaluminum (TMA)-modified porous silica. This reaction step completes a reaction cycle of TMA and ammonia in aluminum nitride growth by atomic layer chemical vapor deposition (ALCVD), a technique based on well-separated saturating gas-solid reactions. The reaction was studied from 423 to 823 K. In addition, the separate reactions of TMA at 423 K and ammonia at 823 K were studied on silica dehydroxylated at 1023 K. The reaction products on the surface were identified by IR and 29 Si, 13 C, and 1 H NMR spectroscopy, and they were quantified by element determinations and 1 H NMR. In the reaction of TMA on silica, methyl groups were attached to the surface indirectly through aluminum and through direct bonding to silicon. In the subsequent ammonia reaction, ligand exchange of ammonia with the methyl groups occurred at all reaction temperatures, resulting in primary amino groups and the release of methane. Also, secondary amino groups were found on the surface, and quantitative determinations indicated the presence of tertiary amino groups, especially at high reaction temperatures. In addition, especially at low reaction temperatures, ammonia chemisorbed associatively on the TMA-modified silica. All of the methyl groups bonded to aluminum were removed with ammonia at 573-623 K, and about 80% of the methyl groups bonded to silicon were removed at 823 K; amino groups bonded to both aluminum and silicon were left behind. The higher the reaction temperature, the smaller was the average number of hydrogen atoms (x) in the amino groups (NH x ).

An investigation of catalyst/cocatalyst/support interactions in silica-supported olefin polymerization catalysts based on Cp*TiMe3*

“…Coordination of 6 to a nearby siloxane oxygen is likely to occur in order to maintain the preferred tetracoordination at Al. This hypothesis is nevertheless consistent with the ability of AlMe 3 /silica to adsorb Lewis bases such as Et 2 O [47] and NH 3 [43,51]. Adsorption may occur by displacement of the coordinated siloxane oxygen atom, instead of being interpreted as evidence for persistent threecoordinate Al sites [47].…”

“…This assertion was supported by elemental analysis of AlMe 3 adsorbed on silica gel [44]. However, recent studies involving AlMe 3 -saturated MCM-41 [45] and silica gel [43] reported that a substantial amount of (BSiO)AlMe 2 persists in those materials. Finally, a recent solid state NMR study showed that AlMe 3 caused little methylation of a silica surface [29].…”

“…The peak at )6 ppm which is clearly seen only after hydrolysis is therefore assigned to O 3 SiMe [43,44], while the peak at )1 ppm is assigned to O 2 SiMe 2 . The presence of small amounts of each of these sites was confirmed by the appearance of two signals in the 29 Si CP/MAS spectrum, Figure 7.…”

“…Most investigations of the AlMe 3 -support interaction have involved fumed silicas, notably Aerosils and Cab-O-Sils, however, studies of AlMe 3 adsorption on porous glass [40], silica gels [41][42][43][44] and MCM-41 [45,46] have also appeared. Regardless of the origin of the silica, chemisorption was suggested to involve formation of isolated three-coordinate methylaluminum and dimethylaluminum sites, associated with O or Si atoms of the surface, Scheme 2.…”

“…)8 ppm, Figure 6(a). They have been assigned to Si-CH 3 and Al-CH 3 sites, with their intensities used to estimate the relative amounts of these sites [43,45], implicitly assuming that the cross-polarization efficiencies of methyl carbons bound to either Si or Al are similar (a reasonable assumption, although CP/MAS spectra are not in general quantitative). The greater line width of the Al-CH 3 signal is consistent with the effect of quadrupolar line broadening due to the proximity of 27 Al, as well as unresolved J-coupling between 13 C and 27 Al.…”