Browse Prior Art Database

Wafer Surface Analysis and Sorting Disclosure Number: IPCOM000124038D
Original Publication Date: 2005-Apr-06
Included in the Prior Art Database: 2005-Apr-06
Document File: 3 page(s) / 40K

Publishing Venue



Our process sends silicon wafers thru a standard cleaning process and then through a Top anti-reflective coating process. Once the wafer are clean, and coated with 440 angstroms of NFC-540, we measure the wafers on a Tencor F5 thin films measurement tool. We analyze the data and compare the mean film thickness and standard deviation of the NFC-540 to where the tools would normally run. Instead of using the wafer to characterize the film, we?ve used the film to characterize the wafer?s surface. We gather the tools recent NFC-540 coating data, average it, and then post Upper and Lower control limits accordingly. We've determined if a wafer falls within a range of ~30 angstroms for mean and less than ~80% of the tools normal standard deviation, it does not possess any of the 3 major surface abnormalities, Native Oxide, Residual HMDS/High Surface tension, or Scratches/FM.

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 52% of the total text.

Page 1 of 3

Wafer Surface Analysis and Sorting

Wafer Surface Analysis and Sorting

By Jason Duncan

IBM, 1000 River Street, Essex Jct., VT 05452

Co-Inventors : Bruce Bosley and William Forant

This paper involves the analysis and sorting of bare silicon wafers using Top anti-reflective coatings. It will describe how to use Top anti-reflective coatings to characterize surface abnormalities that affect the Coatability of DUV Resists. Silicon wafers have 3 major surface problems that can cause inaccurate thin film measurements.

Problem #1: Native Oxide . Native Oxide growth can cause film thickness to appear thicker than they actually are. Oxide will grow on silicon in manufacturing environments, when exposed to the air. Cleaning operations, or contact with TMAH (DUV Developer) can accelerate this growth and render the silicon useless.

Standard Solution: Typically, manufacturers will sort for Native Oxide after a clean operation, using a Rudolph Ellipsometer. The tool will measure the wafers and segregate any wafer with higher than normal Oxide growth. Typically less than 15 angstroms Native Oxide is acceptable.

Drawbacks: Measurement tools are highly utilized for measuring films on Product. Sorting silicon typically runs at approximately 25 wafers per hour. Since high output fabricators will measure resist film thickness on 100-200 wafers per day, it's not feasible to measure wafers for Native Oxide in this fashion.

Problem #2: Adhesion Promoters. Some Materials require the use of an adhesion promoter, Hexamethyldisilazane (HMDS). Adhesion promoters do not add thickness to the wafer, but over time, the cumulative affects can change the surface characteristics enough that uniform coatings can not be attained. In extreme cases, tools are adjusted using the inaccurate data. In reality, the final film thickness is driven out of spec. By rule of thumb, any material that does not require an adhesion promoter will not coat well over wafers with residual adhesion promoter on the surface.

Standard Solution: Typically, Manufacturers will use a Goiniometer Surface Tension measurement tool and the "water droplet method" to determine if the surface of a wafer has been in contact with HMDS vapor. This is done by placing a drop of water on the surface of the wafer, in ~ 5 locations, and measuring the angle of contact. See Below.

The contact angle is measured in degrees. Favorable contact angles for coating DUV materials are between 10 and 40 degrees. Contact angles above 40 degrees are considered coating "challenges" and require high volumes of resist for uniform coatings.

Drawbacks: Once wafers have been handled and measured in this fashion, they need to be cleaned and sorted


[This page contains 1 picture or other non-text object]

Page 2 of 3

again. This...