Title: 150mm Wafer Transfer
1150mm Wafer Transfer
- Keith Warner, Andy Loomis
- April 7, 2000
2Transfer Process Goals
Adhesive-based wafer-to-wafer bonding 2 5 µm
bondline thickness, thermal limits on
post-transfer processing, /- 2 µm alignment
accuracy with existing tools Complete transfer
of NU processes, also useful for other device
types Oxide fusion bonding (Adopt existing
process from CCD work) lt 1 µm bond layer,
post-transfer processing 400C (Tungsten CVD
plugs, hot Al), /- 1 µm alignment with tool to
be purchased Determine technology limits on this
program Sub-micron via placement with oxide
bonding Build/specify tools using stepper
technology techniques
3Wafer Bonding - Voids
- Voids are formed after the SOI handle wafer is
removed - Particulate - Cleaning, clean area, inspection,
filtration - Pinholes and oxide quality - Starting material
- Adhesion Cleaning
- Bubbles Vacuum outgassing of epoxy
4Wafer Bonding - Voids
(After wafer thinning decorative etch)
12/16/99
3/2/00
Void free wafer transfer achieved
5Wafer Bonding - Alignment
- Alignment method depends on type of tool
- Early work used Research Devices flip-chip bonder
- Limitations of optical system restricted accuracy
to /- 5 µm
6Wafer Bonding - Alignment
- Karl Süss mask aligner with infrared optics
- Better than /- 2 µm accuracy, repeatable
7Wafer Bonding - Alignment
(Same wafer without voids)
Accurate alignment achieved
8Wafer Bonding Bondline Uniformity
(2-5 µm required across wafer)
- Wafer flatness Double side polished ultra-flat
wafers - Adhesive curing cycle Experiments started
(Research Devices tool is not designed for
wafer-to-wafer bonding available pressure is
inadequate) - Spacers Latex microspheres
- Particulate Cleaning, inspection, filtration
Still working on optimization
9Wafer Bonding Apparatus and Operation
10Status
- Void-free transfer successful
- Alignment accuracy adequate for devices currently
available (North2) - Method identified to meet bondline requirements
experiments are underway - Full-time effort has started