Adrian Kiermasz

Plasma-Enhanced Atomic Layer Deposition Sealing Property on Extreme Low-k Film with k = 2.0 Quantified by Mass Metrology

We have investigated plasma-enhanced atomic layer deposition (PEALD) SiN pore-sealing film formation and diffusion behavior on highly porous SiOCH films. Mass measurement revealed the diffusion of the precursor used in PEALD into pores of SiOCH films, which was enhanced for higher-porosity SiOCH films. The diffusion of the precursor into the pores was reduced by applying UV-assisted restoration treatment before the pore-sealing process, which helped the formation of hermetic pore-sealing films. The results indicated that a 1-nm-thick SiN film was sufficient to seal the surface of the restored SiOCH film with k = 2.0. It was found that the decrease in k due to the pore-sealing deposition was as small as 0.02. The results indicated that the sequential application of UV-assisted restoration and PEALD-SiN pore sealing is a promising method of introducing the use of highly porous SiOCH films with k = 2.0 into interconnect integration.

Characterization of Oxidation Induced Substrate Loss

Post implant resist strip for sub-65 nm highdose implant (HDI) poses challenges with regard to Si substrate loss due to oxidation. In order to ensure the desired device characteristics, this loss must be kept to a minimum, typically less than 4A loss per strip/wet clean cycle. Conventional metrology techniques such as Ellipsometry, XPS and EOT (effective oxide thickness measurements using capacitance) have difficulty in measuring substrate loss as they make a measurement of native oxide thickness rather than substrate loss. Additionally quantitative measurements such as spectrophotometry and ICPMS measure dissolved silicon and are limited to wet clean process steps only. This paper reports on wafer mass loss as a direct linear measurement technique to quantify substrate loss. Mass loss measurement displays the ability to differentiate with a high degree of resolution mass loss in the sequential ash / clean / anneal process.