Continuous atmospheric atomic layer deposition (ALD) is based on the separation of ALD reactants in space rather than in time. In this study, operating conditions for continuous atmospheric ALD were evaluated using a multiple slit gas source head. A test apparatus was constructed consisting of a gas source head that sits above a substrate that is driven by a programmable stepper motor. The gas source head was composed of a series of rectangular channels that spatially separate the ALD reactants. With Al2O3 ALD as the model system, the trimethylaluminum (TMA) reactant channel was positioned in the center and the H2O reactant channels were located on both sides of the TMA channel. This design allowed for deposition of two Al2O3 ALD cycles during one complete back-and-forth translation of the substrate. The gap spacing between the gas source head and substrate was fixed and controlled using micrometers. A series of experiments was conducted using He as a tag gas to determine operating conditions that prevent reactant cross diffusion. There was a wider range of suitable exhaust pumping speeds at a 30 μm gap spacing than at a 100 μm gap spacing. However, mechanical tolerances were not sufficient to allow back-and-forth translation of the substrate at a gap spacing of 30 μm. For Al2O3 ALD at a 100 μm gap spacing, the best Al2O3 film uniformity was achieved when there was a slight vacuum of ∼−5 Torr beneath the gas source head relative to ambient. A larger vacuum beneath the gas source head occurred if the pumping speed of the exhaust channel was higher relative to the N2 supplied to the gas source head. These conditions produced a smaller footprint of Al2O3 deposition resulting from an influx of N2 from the higher pressure ambient. The gas source head had a higher pressure than ambient if the exhaust pumping speed was too low relative to the N2 supply. Under these conditions, reactant gas leaked out the sides and formed Al2O3 powder around the perimeter of the gas source head resulting from the chemical vapor deposition reaction of TMA with ambient moisture. A response surface model was generated to predict the dependence of the pressure beneath the gas source head relative to ambient on the various operating conditions.
Metal insulator semiconductor structures were fabricated from n-Si(100) and n-Ge(100) wafers passivated with thin (4.5–5 nm) films of N-rich BCxNy (0.09≤x≤0.15, 0.38≤y≤0.52) and with atomic layer deposition HfO2 (10 nm) as the gate dielectric. C-V and I-V characteristics of devices with BCxNy films grown at 275–400 °C by chemical vapor deposition showed that lower deposition temperatures resulted in improved electrical characteristics, including decreased hysteresis, lower VFB shift, lower leakage current, and less C-V stretch out. The electrical improvement is attributed to decreased bulk and interfacial defects in lower temperature deposited BCxNy films, which also had a higher optical bandgap [Eg=3.55 eV at 275 °C on Ge(100)], lower subbandgap absorption, lower index of refraction [n(633 nm)=1.84 at 275 °C on Ge(100)], reduced O uptake during ambient exposure, and increased percentage of B. Even for the lowest growth temperature studied (275 °C), BCxNy-passivated Ge(100) devices had considerable hysteresis (1.05 V), and electrical characteristics worsened after a postmetallization anneal. BCxNy-passivated Si(100) devices outperformed similar Ge(100) devices likely due to the higher interface state densities at the BCxNy–Ge(100) interface associated with the higher relative inertness of Ge(100) to thermal nitridation. C-rich BC0.61N0.08 films were also investigated but large amounts of hysteresis and fixed negative charge motivated the abandonment of these films.
The interfacial adhesion energy was studied using the four-point bend method for boron carbo-nitride (BC x N y ) deposited on dielectric and copper films. Twenty-five nanometer BC x N y films were deposited by chemical vapor deposition at 360°C and 1 Torr using dimethylamine borane with no coreactant, NH 3 , or C 2 H 4 , producing different composition films, BC 0.37 N 0.15 , BC 0.11 N 0.49 , BC 0.92 N 0.07 , with dielectric constants of 4.1, 4.2, and 3.8, respectively. BC x N y films were deposited on dense and porous dielectrics, and copper. BC x N y films adhered strongly to the dielectric films and the composite beams snapped before debonding, revealing that the critical debond energy G c exceeded 10 J/m 2 . The adhesion of BC x N y to oxidized copper increased with carbon content in the film, with the BC 0.92 N 0.07 film beams snapping, and is possibly related to covalent bonding between surface oxygen and carbon in the film. a) Present address: Texas Instruments, 13560 North Central Expressway, Dallas, TX 75243.
Porous ultralow dielectric film pores were sealed by 1 -5 nm films of boron carbonitride, itself a dielectric. The BC 0.9 N 0.07 films were deposited by chemical vapor deposition at 335°C on etched and ashed blanket films and patterned films of porous methyl silsesquioxane ͑PMSQ͒. The penetration of Ta through the boron carbonitride film and into the porous low substrate following exposure to TaF 5 at 230°C, or TaSi x growth from TaF 5 and Si 2 H 6 at 230°C, was used to test pore sealing. Tantalum profiles were evaluated using x-ray photoelectron spectroscopy depth profiling, back side secondary ion mass spectroscopy, and energy dispersive x-ray spectroscopy. A 3.9 nm boron carbonitride film sealed PMSQ, which has an average pore diameter of 1.9 nm before etching. The diffuse nature of the BC 0.9 N 0.07 -PMSQ interface in electron energy loss profile maps suggests that some of the BC 0.9 N 0.07 penetrates into the PMSQ until the pore openings are pinched off as the sealing film deposits.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
customersupport@researchsolutions.com
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.
Copyright © 2025 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.