Browse Prior Art Database

Method for parallel fabrication of large area waveguide structures

IP.com Disclosure Number: IPCOM000016414D
Original Publication Date: 2002-Nov-11
Included in the Prior Art Database: 2003-Jun-21
Document File: 8 page(s) / 82K

Publishing Venue

IBM

Abstract

*Main Idea

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 28% of the total text.

Page 1 of 8

Method for parallel fabrication of large area waveguide structures

*Main Idea

Transmission of very fast signals (10Gbit/s and more) over large
distances (1 m) is becoming increasingly difficult for
conventional copper based technologies. The use of optical signal
transmission is generally considered as a promising solution to
this problem.

There has been a wealth of approaches for optical interconnects,
depending on the speed, length and type of application. This
invention addresses the approach of using waveguides that are
integrated directly with a large backplane-type printed circuit
board. Such boards can be up to 130 cm x 65 cm large. To
practically integrate waveguide structures into or onto such
large boards, there have been again several approaches, but all
of these approaches are prototype/feasibility demonstrations or
solutions for a restricted (low-volume) application space.

As far as known, none of the current waveguide fabrication
methods is able to provide a mass-producible way of defining
large structures with high-quality waveguides. In general, there
is a trade-off between overall size and resolution for
technologies such as etching and lithography. While the currently
known approaches fail to satisfy all requirements simultaneously,
this will be a "conditio sine qua non" for waveguide based
intra-shelf interconnects. The need for optical interconnects
within high-end systems mainly comes from length limitations of
current copper technology. Optics will therefore be needed for
larger in-box distances first (i.e. 1m card-to-card over a
backplane and not 30 cm on-board). To provide the tens of
Terabit/second aggregate throughput that future high-end systems
will need, individual line speed and density will also have to be
pushed to their limits. For the waveguide technology, this will
result in a need for small waveguide structures (50 micrometer
and smaller) in order to provide the high density and to avoid
modal dispersion problems of large (> 100 um) waveguide cores at
very high channel-speed. As mentioned above, none of the
currently known approaches provides large sample size,
fine-resolution and mass-producibility simultaneously.

A potential solution to use established mass-production
technologies to fabricate large samples of fine-resolution
waveguides is as follows:

An additional process step is added, or more specifically an
"auxiliary lens layer". Instead of trying to directly write very
large samples with fine-resolution structures, a relatively

1

Page 2 of 8

coarse auxiliary photoresist structure is defined first by UV
exposure and etching. This step can be done by established
technologies from large board fabrication. In one potential
impelmentation these photoresist structures are then heated, and
upon melting, surface tension effects will cause them to form
cylindrical lens structures. These cylindrical lenses are then
used to do the actual exposure of the waveguide polymer. By
exposing through the previously formed lenses, much finer
structure...