An Inexpensive, Fast, and Versatile Microfabrication Technique

What is Mock CMOS?
A versatile fabrication process that allows users to quickly and cheaply construct micromachined structures using a one metal, one silicon dioxide film stack on a silicon wafer. This simplified process, (a) starts from pre-processed wafers and requires only one photolithography step, (b) provides a conductor material for actuating electrostatic and thermal devices, (c) avoids electrical shorting between metal microstructures or to the silicon substrate by using silicon dioxide as an insulator, and (d) allows quick prototyping of MEMS structures similar to those designed at Carnegie Mellon University (the CMOS-MEMS process).

The main advantage of this technique is that the design-to-device turnaround can be cut down to just a week (the time needed for photomask production) and can be easily incorporated for an introductory course on microelectromechanical systems.

How did this fabrication process arise?
For a university that does not have a full set of semiconductor processing equipment or sufficient funds for MEMS devices fabricated through foundry services (such as MUMPs), this process can be used to fabriate vertical and lateral electrostatic resonators along with thermal actuators. By using the three fundamental materials of metal, oxide, and silicon, a designer can create microelectromechanical devices with just two major processing steps starting form pre-deposited wafers.

Why is the process called 'Mock CMOS'?
The CMOS-MEMS process at CMU is a post-CMOS fabrication process in which the etching masks are provided by the interconnect metal layers in the standard CMOS process. Prototyping of sensor and actuator devices in the CMOS-MEMS process requires considerable cost, waiting time for chip fabrication, and possible further iterations until satisfactory device performance is attained. By removing the CMOS component and limiting the process to one metal and one oxide layer, a designer can focus on the mechanical aspects of a microstructure with the capability to layout multiple device variations of arbitrary size onto as large a wafer.

Has this process been used extensively?
The simplicity and quickness of the process made it a perfect match for the Introduction to MEMS course offered this past Fall 2001. Approximately 30 students were able to design a range of devices.

Report highlighting this fabrication process click here for PDF

What are the fabrication steps?


Here are the processing steps performed:
Photoresist Patterning
Dehydration: 120C for 10 minutes
Photoresist used: Shipley S1813
Spin speed: 1000 rpm
Approx. thickness: ~2.5-3 um
Softbake: 90 seconds at 115C on hot plate
Exposure time: 16 seconds
Development: 1:1 MIF-312:DI water
Hard Bake: 2 hours at 120C on oven

Aluminum Etch
Type of etch: Wet Isotropic Etch
Chemical used: Transene Type APre-mixed Aluminum Etch (l)
Machine used: None
Temperature of etch: 45-50 degrees Celsius (Digital hot plate setting of 70C)
Advertised etch rate: 100 Angstroms/minute
Etch time: 40 seconds

Titanium/Tungsten Etch
Type of etch: Wet Isotropic Etch
Chemical used: Hydrogen Peroxide (l)
Machine used: None
Temperature of etch: 35-40 degrees Celsius
Advertised etch rate: 100 Angstroms/minute (Digital hot plate setting of 70C)
Etch time: 1:20 minutes

Oxide Etch
Make chamber is clean!!!! Run Cleanall.prc
Type of etch: Dry Anisotropic Etch
Chemical used: Trichloromethane (CHF3) and Oxygen
Machine used: PlasmaTherm 790 RIE
Power: 100 Watts
Gas Flows: 22.5 sccm CHF3; 5.0 sccm O2
Pressure: 125 mTorr
Etch time: 50 minutes
Notes: Silicon surface should turn dark after oxide etch, no rainbows visible

Silicon Etch
Type of etch: Dry etch (isotropic)
Chemical used: Xenon Difluoride (XeF2)
Machine used: Xactix Xetch Xenon Difluoride Etcher
Temperature of etch: Room Temperature
Pressure: 3 Torr XeF2, 0 Torr N2
Approx. etch rate for 1cm x 1cm samples: 1 micron/minute
Etch time: 15 minutes (15 cycles, 60 seconds each)

Current updates:
1. Initial experiments and difficulties - June 20, 2001
2. Polymerization difficulties - August 21, 2001
3. Aluminum and Oxide etching attempt - August 31, 2001
4. Aluminum etching: Room Temperature - September 13, 2001
5. MEMS course structures: Full Process (Polymerization) - September 14, 2001
6. MEMS course structures: Polymerization removal - September 17, 2001
7. Ion mill sample: Metal etch - September 17, 2001
8. Ion mill sample: Full process and MEMS course structures- September 21, 2001
9. Ion mill sample and MEMS course structuers: More silicon etching - September 21, 2001
10. Refining the TiW wet etch - September 28, 2001