Whole wafer MEMS release process

Semiconductor device manufacturing: process – Semiconductor substrate dicing – With attachment to temporary support or carrier

Reexamination Certificate

Rate now

  [ 0.00 ] – not rated yet Voters 0   Comments 0

Details

C464S051000, C464S062100

Reexamination Certificate

active

06586315

ABSTRACT:

FIELD OF THE INVENTION
This unit relates generally to apparatus and methods for manufacturing MEMS (micro-electromechanical systems) by forming a multiplicity of such devices on a silicon wafer. More specifically, the invention relates to such a manufacturing process which allows further processing and/or testing before each individual device is separated from the silicon wafer.
BACKGROUND OF THE INVENTION
Texas Instruments presently manufactures a two-axis analog micromirror MEMS device fabricated out of a single piece of material (such as silicon, for example) typically having a thickness of about 115 &mgr;m. The die layout consists of an oval micromirror, normally 3.8 mm×3.2 mm supported on a gimbal frame by two silicon torsional hinges. The gimbal frame is attached to the die frame by another orthogonal set of torsional hinges. The micromirror die (i.e. each individual device) is fabricated by etching the 115 &mgr;m thick silicon wafer in a specialized ICP (Inductively Coupled Plasma) plasma reactor.
MEMS devices are becoming more and more available and common. However, these devices are extremely small compared to regular machines, but still very large when compared to the individual circuits or components and features found on IC's and other electronic chips. Some MEMS devices such as the digital micromirror device arrays produced by Texas Instruments are made significantly smaller than most other types of MEMS devices, but are also very large compared to components on an IC or other chips and use existing geometry and patterning techniques common for the productions of semiconductor circuits. For example, small MEMS devices such as gimbal supported mirror
32
shown in
FIG. 2D
used for optical switching of transmitted data streams are presently on the order of 3.2×3.8 mm, whereas the mirrors on micromirror arrays used for display devices are typically between about 15-20 microns on a side. Thus, it is seen that MEMS devices are not comfortably compared with either full-size machines or devices (they are much smaller) or a true array of micro devices such as IC's, memory chips, and the like (they are much larger).
The present invention relates to individual mirror devices formed on a wafer using processing steps some of which have similarity to steps used in manufacturing IC's and other semiconductor devices.
SUMMARY OF THE INVENTION
The present invention provides a process for manufacturing a plurality of MEMS devices on a first layer of material, such as for example, a thin wafer of silicon typically having a thickness of about of 115 &mgr;m. The process comprises attaching the thin silicon wafer to a carrier or backing wafer and then defining features for each individual device of said plurality of devices with a first line width. The boundary or separation lines between the individual ones of the plurality of devices are defined with a second line width that has a thickness less than the thickness of the first line width used to define the device features.
After placing both the lines which define the features of the individual devices and the boundary or separation lines between individual devices, the wafer while attached to the backing wafer is etched such that the lines which define the features of the device are etched through the selected thickness. However, the etching is stopped before the thinner lines which define boundaries of the individual devices are etched through the thickness of the wafer. This is possible because of the phenomenon called microloading. Microloading is the differential etch rate between wide lines and narrow lines (wide lines etch faster) in a plasma reactor. Thus, it is seen that the individual devices are formed because of the fast etch rate of the wide lines, while at the same time all of the devices on the wafer remain attached together because of the slower etch rate of the thin separation line. The wafer with the devices still attached together is then separated from the backing layer. It should also be noted that the wafer with the devices could be silicon or another suitable material. Further, the wafer may also undergo other processes before the device is etched. For example, electronics, sensors or other mechanical features can be created by standard IC or MEMS fabrication before the process step of etching through the wafer is accomplished.
Therefore, according to embodiments of the present invention, the silicon wafer with all of the attached devices etched therein can then be further processed. For example, further processing may comprise testing of the torsional gimbals of the individual mirrors by moving the mirrors by either soft directed currents of air or spring pins. This is a much faster process than having to handle and test the gimbals on each separated mirror. In addition, it is also possible to better clean the attached mirror on the wafer after it has been released from its backing layer than it is to handle each individual device.


REFERENCES:
patent: 6383833 (2002-05-01), Silverbrook
patent: 6387778 (2002-05-01), Bonin et al.
patent: 6444138 (2002-09-01), Moon et al.

LandOfFree

Say what you really think

Search LandOfFree.com for the USA inventors and patents. Rate them and share your experience with other people.

Rating

Whole wafer MEMS release process does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Whole wafer MEMS release process, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Whole wafer MEMS release process will most certainly appreciate the feedback.

Rate now

     

Profile ID: LFUS-PAI-O-3064668

  Search
All data on this website is collected from public sources. Our data reflects the most accurate information available at the time of publication.