Reactor for wafer backside polymer removal using plasma products in a lower process zone and purge gases in an upper process zone

Abstract

A reactor is provided for removing polymer from a backside of a workpiece. The reactor includes a vacuum chamber having a ceiling, a floor and a cylindrical side wall. The reactor further includes workpiece support apparatus within the chamber configured for a workpiece to be placed thereon with its front side facing the ceiling. The support apparatus is configured to leave at least an annular periphery of the backside of the workpiece exposed. A confinement member defines a narrow gap with an outer edge of the workpiece, the narrow gap being on the order of about 1% of the workpiece diameter, the narrow gap corresponding to a boundary dividing the chamber between an upper process zone and a lower process zone, the reactor further comprising a vacuum pump coupled to the lower process zone. An external plasma-generating chamber is coupled to the chamber, the external plasma-generating chamber configured to introduce a plasma by-product into the lower process zone, and a supply of a polymer etch precursor gas to the external plasma-generating chamber. The ceiling includes a gas distribution plate facing the upper process zone, the reactor further comprising a purge gas supply coupled to the gas distribution plate.

Description

CROSS-REFERENCE TO RELATED APPLICATIONS[0001]This application claims the benefit of U.S. Provisional Application Ser. No. 60 / 898,645, filed Jan. 30, 2007.BACKGROUND[0002]Plasma processing of a workpiece or semiconductor wafer, particularly dielectric etch plasma processing, typically employs carbon-containing process gases (e.g., fluorocarbon or fluoro-hydrocarbon gases) that enhance the etch selectivity of dielectric materials, such as silicon dioxide, relative to other materials such as silicon. These processes are used to treat the front (top) side of the wafer on which the microelectronic thin film structures are formed. The opposite (back) side of the wafer is typically unpatterned. One problem is that the carbon-containing process gases tend to form polymer precursors in the plasma, which can leave a polymer residue on the front side of the wafer and on the exposed portion of the backside of the wafer, and even some distance under the unexposed portion of the wafer backside. S...

AI Technical Summary

Problems solved by technology

One problem is that the carbon-containing process gases tend to form polymer precursors in the plasma, which can leave a polymer residue on the front side of the wafer and on the exposed portion of the backside of the wafer, and even some distance under the unexposed portion of the wafer backside.

However, the wafer edge is beveled, and the curved surface on the backside of the wafer edge is also exposed and therefore susceptible to polymer deposition during plasma processing.

Attempting to remove the back-side polymer film by heating and exposing the wafer to an oxygen-rich plasma (during a post-etch polymer removal step) will damage the ultra low-K film by removing carbon from it.

It is generally difficult if not impossible to avoid damaging the ultra low-K (ULK) film beyond the permissible 3 nm depth while exposing the wafer to an oxygen-rich plasma of a sufficient density and for a sufficient time to remove 700 nm of polymer from the backside of the wafer edge or bevel.

The required polymer-to-ULK film etch selectivity (over 200:1) for such a polymer removal process in general cannot be maintained reliably in conventional processes.

Such a collar tends to shield the wafer edge, but cannot be sufficiently close to the wafer edge to prevent polymer deposition on the backside of the wafer edge.

Therefore, it has seemed inevitable that unwanted polymer is deposited on the wafer, including the backside edge of the wafer.

However, this approach is not practical if the thin film structure on the wafer includes an ultra-low K film that is exposed on a sidewall of an etched opening.

Such a rich oxygen mixture in the etch plasma would cause unacceptable damage to the ultra-low K film.

Images