There have been numerous comments on the MEMS line about the need for low cost
and/or special masks for MEMS applications. We, too, have been hampered in our
ability to obtain gray-scale masks, masks with circular patterns, masks with
extremely fine detail, masks defined by bit-maps, and masks with very large
amounts of irregular data.

In addition, there is a need to be able to pattern deep submicron features
(order 0.1 um) directly on wafers for experimentation, process development and
small scale production.

By leveraging our current work, we have responded to these needs by beginning
the development of a new, MEMS-enabled, mask- and wafer-exposure system. Each
desk-top system would be placed in a clean room and connect to a
customer-supplied, high end PC with a high performance and capacity memory
system. Thanks to several breakthroughs we have made, we foresee the sales of
these new systems (not including the PC) for only US $25,000 to $50,000 -- less
than 1/100 the cost of conventional mask-exposure machines, which are, however,
faster, more automated, more accurate and more flexible than our planned
first-generation machine.

We expect that our new system will be able to FULLY pattern production-scale
masks and wafers, e.g., up to 125 mm wafers and mask plates, not just create
miniature patterns only a fraction of a mm square. The target, minimum address
unit is 25 nm (1 microinch) over this entire area. Using a very thin resist,
the target, minimum dot size is 100 nm (4 microinches). The target, absolute,
dot-placement accuracy over this entire area is 100 nm (4 microinches). Our
proprietary servo systems and fabrication techniques compensate for thermal
expansion / contraction and provide precise registration and reregistration
during writing. The estimated time required to write such a large, fine pattern
is several hours at a rate of 1 sq. mm/S.

(I cannot go into how we plan to do this, but we are using relatively
conservative technology that I believe we can bring to market fairly quickly.)

The system can be programmed to operate in both polar and Cartesian
coordinates. We would provide a public file format to control it and be able to
support existing, public, mask data formats.

We are looking for:

 1. additional funding for this development
 2. collaboration to accelerate this development
 3. customer input as to their requirements for, and interestin purchasing,
    such low-cost, high precision, mask- and wafer- exposure systems
 4. customer input as to their problems with conventional mask makers and their
    interest in having specialty masks made by us using our systems

Please contact me if you have any comments or interest in this work.

For those of you who don't know us, we specialize in Electronic Imaging
Devices. Extensive information about us is available on our Web page, whose
address is given below.

Sincerely,

Steven G. Morton, President
Oxford Computer, Inc.
39 Old Good Hill Road
Oxford, CT 06478 USA
e-mail: [email protected]
World Wide Web: http://www.oxfordcomputer.com
tel. 203-881-0891, fax 203-888-1146