The 2,2,6,6-Tetramethyl-2-Sila-3,5-Heptanedione Route to the Chemical Vapor Deposition of Copper for Gigascale Interconnect Applications

Abstract

A new class of copper(II) precursors containing silylated β-diketonate ligands has been developed for the chemical vapor deposition (CVD) growth of copper for applications in ultralarge scale integration interconnect schemes, including conformal seed layer for gigascale Cu integration and ultrathin Cu lines with enhanced conductivity characteristics. Cu(tmshd)2 (tmshdH = 2,2,6,6-tetramethyl-2-sila-3,5-heptanedione) has been studied as a representative compound and is appreciably more volatile than nonsilylated compounds such as Cu(tmhd)2 or Cu(tmod)2 (tmhdH = 2,2,6,6-tetramethyl-3,5-heptanedione; tmodH = 2,2,7-trimethyl- 3,5- octanedione). The CVD process employs Cu(tmshd)2 as the metalorganic precursor and hydrogen as the reducing and carrier gas. These films were deposited using a custom made, cold wall, stainless steel CVD. Copper films were produced at a substrate temperature of 250 – 320 °C, hydrogen flow rates of 20 - 100 sccm, deposition pressure of 0.2 - 1 Torr, and a source temperature of 120 – 135 °C. The films were analyzed by X-ray photoelectron spectroscopy, cross section scanning electron microscopy, transmission electron microscopy, four-point resistivity probe, Rutherford backscattering spectrometry and Auger electron spectroscopy.

Investigations into the preparation of sila-β-diketones via 2-trimethylsilyl-1,3-dithianes: structural characterization of a second polymorph of bis(2,2,6,6-tetramethyl-2-sila-3,5-heptanedionato)copper(II)

“…1) in order to prepare a wide variety of sila-bdiketones [8][9][10][11]. In general, this method employs the condensation of the lithium salt of an acetyltrialkylsilane with an activated carbonyl compound, such as an acyl chloride [8][9][10][11].…”

“…1) in order to prepare a wide variety of sila-bdiketones [8][9][10][11]. In general, this method employs the condensation of the lithium salt of an acetyltrialkylsilane with an activated carbonyl compound, such as an acyl chloride [8][9][10][11]. However, attempts to prepare a disilylated b-diketone using this protocol proved unsuccessful, which prompted us to explore a synthetic strategy involving 1,3-dithiane chemistry [12,13], based upon disconnection between the c-carbon atom and the b-carbonyl carbon (Fig.…”

“…While the C(1)-C(8) bond length (1.563(7) A A) in 3 is slightly shorter than the corresponding bond length in the two crystallographically independent molecules of 2; 2 0 -diphenyl-2; 2 0 -bi-1,3-dithiane (1.592(4) and 1.593 (4) A A) [38], the Si(1)-C(1) and Si(2)-C(8) bonds (1.969(5) and 1.974 (6) A A) are significantly longer than similar ones in related silylated mono-1,3-dithianes (1.938(4) and 1.925 (3) A A) [39,40]. There is also a marked flattening of the tetrahedra around the C(1) and C (8) atoms. Thus, the combination of steric congestion, due to the presence of the two SiMe 3 substituents, and stereoelectronic effects from the sulfur atom lone pairs leads to the structural distortions observed for 3.…”

“…Recently, we have reported on efforts to increase the volatility of homoleptic Cu(II) b-diketonate complexes by the incorporation of trialkylsilyl substituents in the supporting ancillary ligands [8][9][10][11]. In particular, the complex Cu(tmshd) 2 (where tmshd is the anion of 2,2,6,6-tetramethyl-2-silaheptane-3,5-dione) sublimes at a significantly lower temperature than its carbon analogue, Cu(tmhd) 2 (where tmhd is the anion of 2,2,6,6-tetramethylheptane-3,5-dione) [9] and is a viable, nonfluorinated candidate for MOCVD [8].…”

“…Recently, we have reported on efforts to increase the volatility of homoleptic Cu(II) b-diketonate complexes by the incorporation of trialkylsilyl substituents in the supporting ancillary ligands [8][9][10][11]. In particular, the complex Cu(tmshd) 2 (where tmshd is the anion of 2,2,6,6-tetramethyl-2-silaheptane-3,5-dione) sublimes at a significantly lower temperature than its carbon analogue, Cu(tmhd) 2 (where tmhd is the anion of 2,2,6,6-tetramethylheptane-3,5-dione) [9] and is a viable, nonfluorinated candidate for MOCVD [8]. Further, the silylated Cu(II) compounds generally show greater stability than the corresponding non-silylated complexes [9][10][11]; the larger differences between the sublimation and decomposition temperatures should allow for the ability to attenuate the properties of the complex to fit a specific MOCVD application.…”