Has Dry Cleaning Come of Age?
Maria Lester, Associate Editor -- Semiconductor International, 3/1/1999
For the past 10 years, the semiconductor industry has anticipated the replacement of wet chemistries with dry cleaning technologies. But dry cleans have not kept pace with advances in wet cleans with the shear cleaning strength of wet chemistries, in some instances, superior to dry. What then will be the role of dry cleans in future applications? According to Jerzy Ruzyllo, professor of Electrical Engineering at Penn State University (University Park, Pa.), dry cleaning will not directly replace wet chemistries in mainstream applications, but instead will supplement wet techniques at various points such as single wafer processing at 300 mm and in cluster tools. He suggested that simplified wet cleaning procedures performed ex-situ be followed by a dry surface conditioning sequence integrated with the next process step such as epitaxial deposition, deposition of high-k gate dielectric or contact metallization. So while dry cleaning has been effective in the removal of organic contamination, volatilization of most metallic contaminants, etching native oxides and altering the chemical condition of the wafer surface, ideally, it should be considered part of the overall surface preparation procedure.
The introduction of dry clean technology in the manufacturing arena has been slow. According to Ruzyllo, this is in part due to the lack of a single dominant driving force, as was the case with etch anisotropy, which spurred the introduction of dry etch techniques. Early experiments fell short in terms of performance and cost of ownership, adding further to delays into production. Today, dry chemistries have found a place. They can be used for in-situ conditioning of surfaces that are essentially free of particles and to remove some metallic contaminants, Ruzyllo said.
Conventional wet chemistries are very effective for most applications, though there are growing concerns regarding environmental safety and economics. 'Ultimately, at comparable wet-clean performance, cost will determine the extent of dry clean use in select IC manufacturing applications,' Ruzyllo said. In theory, larger wafers should favor dry chemistries because of increased demand on chemical, water and waste disposal. Cost effectiveness may also be seen if dry surface cleaning/conditioning modules are added to existing cluster tools. In addition to reduction of chemicals and water, a key advantage of dry surface processing technology is in its compatibility with process integration, Ruzyllo commented. These chemistries may also offer particular advantages in processing high aspect ratio structures.
However, the real test of dry cleaning technologies is in particle removal. If they fail to be greater than 90% effective for all particle sizes, wet methods such as megasonic SC1 will continue to be used to control particles, Ruzyllo said. Subsequently, particle-free wafers can then be loaded into the cluster tool using a dedicated module, and final surface conditioning can be carried out using dry chemistries. The goals of surface conditioning will depend on the requirements of subsequent processes. Different requirements apply to pre-oxidation surface treatments in which the presence of very clean, nearly stoichiometric ultra-thin oxide is not only acceptable, but may be desirable. The emphasis is on surface conditioning using gas-phase processing rather than its 'cleaning' functions.
| Gas-Phase Surface Cleaning Techniques | ||||
| Particles | Organics | Metals | Nat./chem. oxide | |
| Physical | Aerosol | -- | -- | Ar sputter |
| Thermal | Laser | Oxidation | HCl
anneal Metalorganics |
H2 anneal |
| Vapor | -- | -- | -- | HF/H2O vapor
HF/alcohol |
| Photochem. | -- | UV/O2 | UV/Cl2 UV/HCl |
-- |
| Plasma | -- | Remote O2 | Remote HCl | Remote H2 |
| (Source: Jerzy Ruzyllo, July 1998) | ||||
