MOCVD Precursors
Copper β-Diketonate Precursors:
The substitution of a carbon atom by silicon provides an attractive, novel approach to modification of the thermal stability and volatility of metal-organic chemical vapor deposition precursors supported by β-diketonate ancillary ligands. The low temperature reaction of the lithium enolates of acetyltrialkylsilanes with acyl chlorides affords the sila-β-diketones, R'C(O)CH2C(O)SiR3 (R'=Me, Et, n-Pr, i-Pr, n-Bu, i-Bu, s-Bu, t-Bu; SiR3=SiMe3, SiEt3, SiMe2(t-Bu), SiMe2(t-hexyl), Si(i-Pr)3), in good yields. Multinuclear NMR studies suggest that the sila-β-diketones exist as the enolic tautomer with a vinylsilane isomeric structure. Homoleptic Cu(II) sila-β-diketonate complexes were prepared in a first pass study to evaluate how precursor performance is affected by modulation of the peripheral substituents in the ligands. Thermal analyses, (TGA, DSC) show that the silylated Cu(II) precursors (SiR3=SiMe3; R'=t-Bu or i-Bu) have greater volatility than the corresponding carbon analogues. Some of the new Cu(II) complexes exist as liquids or low melting solids, which are preferred states for industrial deposition processes. X-ray diffraction studies of selected copper complexes showed them to have typical, square planar geometry; calculations of molecular volumes suggest that packing in the solid-state is less efficient for the silicon-containing complexes than for the non-silylated analogues.
Silver β-Diketonate Precursors:
Currently, there is an ongoing interest in the search for electroluminescent materials. Applications range from lighting appliances such as fluorescent tubes to display apparatus such as cathode ray tubes or projection television. One of the most rapidly advancing and growing areas is the development of thin film electroluminescent displays. Inorganic phosphor materials currently are widely used in these applications while research on organic phosphors are undergoing. The most commonly employed phosphor, the ZnS:Mn with SrS:Ce combination generates red and green output by filtering the yellow emission from the ZnS:Mn couple while the SrS:Ce couple provides a bluish green output. Unfortunately, the SrS:Ce couple is only a weak blue emitter and as such, there is a need for bright blue emitters. Recent works show SrS:Cu as an alternative giving a good blue emission. Codoping of SrS:Cu with Ag gave a true blue color. However, current limitation towards this approach appears to be the lack of suitable volatile Ag precursors.
One of the techniques for preparation of thin films of these phosphors is metallo-organic vapor deposition. As such, precursor volatility and stability are important issues. Research on the preparation of stable yet volatile complexes is central to the development of new, more efficient MOCVD precursors. 1,3-β-diketones are one of the most common ligand types. Structural variations of this ligand may impart desired properties to the precursor complex. Typically, variations of ligand substituents can modulate volatility. Introduction of a second ligand may have complementary effects yet improves the stability of the complex. Combination of both strategies clearly can lead to highly desirable enhancements in activity. Majority of the available literature on silver β-diketonate complexes used in MOCVD have been concentrated on fluorinated β-diketones, e.g. hfac, fod. These ligands impart volatility, the most important property required for complexes with applications in MOCVD. Unfortunately, a major concern is fluoride contamination. The initial aim of this work is to present several alternatives to the use of fluorinated silver β-diketonates for MOCVD applications. Although there were some previous studies that indicated that silver complexes containing non-fluorinated β-diketonate ligands had limited volatility, we decided to explore a much wider range of structural alterations to the β-diketonate backbone, including unsymmetrically substituted ligands, as well as those in which silylated substituents were incorporated or in which an oxygen donor atom was replaced by sulfur or imino nitrogen. The replacement of one of the oxygen donor atoms of the β-diketonate was investigated, since such ancillary ligands might facilitate reduction of oxygen contamination, which accompanies the MOCVD processing of silver β-diketonates.
A series of Ag(I) complexes, [Ag(X)L], where X is the anion of 2,2,6,6-tetramethyl-3,5-heptanedione (tmhdH), 2,2,7-trimethyl-3,5-octanedione (tmodH), 2-sila-2,2,6,6-tetramethyl-3,5-heptanedione (tmshdH), 5-mercapto-2,2,6,6-tetramethyl-4-hepten-3-one (S-tmhdH), 5-mercapto-2,2,7-trimethyl-4-octen-3-one (S-tmodH), or 5-amino-2,2,6,6-tetramethyl-4-hepten-3-one (N-tmhdH) and L = triphenylphosphine (PPh3) or tri-n-butylphosphine (PBu3), were prepared by the treatment of silver nitrate with either the diketone derivative in the presence of a base or with the preformed sodium salt of the diketone derivative.
The thermal properties of the new complexes were investigated for potential application in chemical vapor deposition (CVD) processes. Results of thermogravimetric analysis showed that the vast majority of the silver complexes have little promise for CVD, since the ligands dissociate at elevated temperatures without volatilization. The first ever reported single crystal X-ray diffraction studies of silver complexes, with simple, ancillary, non-fluorinated β-diketonate supporting ligands, revealed [Ag(tmhd)(PPh3)] to be monomeric and [Ag(S-tmhd)(PPh3)]2 to be dimeric in the solid state, respectively.
Copper Dithiocarbamate Precursors:
A series of Cu(II) dialkyldithiocarbamate complexes, Cu(S2CNRR')2, with R=R'=n-Bu (1); i-Bu (2); c-Hex (3); CH2Ph (4); R=n-Bu, R'=Et (5); R=n-Pr, R'=c-PrCH2 (6); R=R'=n-Pr (7); i-Pr (8); allyl (9), were prepared. The thermal properties of the complexes were investigated to determine if their potential performance in chemical vapor deposition processes was affected by the nature of the peripheral substituents of the ancillary ligands. Modest gains in volatility were noted for 2 and 7 over the most often utilized complex with R=R'=Et, while 1 and 8 had thermal parameters and stability comparable to this standard. Unsymmetrical substitution, such as in 5, also improved volatility, with some loss of stability for this particular compound. X-ray diffraction studies of complexes 1-6 suggested that long range Cu···S interactions in the solid-state have little bearing on the thermal properties of this class of Cu(II) complexes.
