Optical waveguides – Planar optical waveguide – Thin film optical waveguide
Reexamination Certificate
1999-04-15
2003-05-13
Ullah, Akm E. (Department: 2874)
Optical waveguides
Planar optical waveguide
Thin film optical waveguide
Reexamination Certificate
active
06563998
ABSTRACT:
FIELD OF THE INVENTION
This invention relates to polished polyimide substrates and polymer laminate structures formed on those substrates, and more particularly, to polymer devices for optical applications.
BACKGROUND OF THE INVENTION
Optical waveguide devices are typically made on silicon substrates. It is desirable that materials used for optical waveguide devices exhibit certain optical, thermal and mechanical characteristics, besides low optical loss. Common silicon micromachining technologies include anisotropic chemical etching and reactive ion etching (RIE). Passive optical waveguides exhibiting acceptable losses between 0.1 and 10 dB/cm have been demonstrated in a number of materials, most notably optical grade glasses (silica) and PMMA and polystyrene polymers. The highest quality silica waveguides with very low losses of 0.1 dB/cm have been deposited on silicon wafers by the flame hydrolysis technique which yields good control over the index and thickness of the film but requires heating the porous glass layer to 1250° C. for consolidation. This high temperature perturbs the crystallographic micro-structure of silicon which affects its anisotropic chemical micromachining. Furthermore, the flame hydrolysis technique requires specialized and expensive equipment and involves the use of silane which is a toxic gas.
The fabrication of channel silica ridge waveguides requires deep RIE of several microns. Also vertical deep sidewalls and high aspect ratios, which are desirable in a micro-mechanical structure, such as accelerometer, can be achieved in silicon with deep RIE. However, RIE is an expensive process and requires use of high vacuum equipment which is prone to frequent failure. Furthermore, another problem with deep RIE is the erosion of the masking layer due to poor selectivity, which limits the etch depth in the silica film to the thickness of the masking layer, which is usually on the order of one micron. The selectivity of RIE can be improved with the proper selection and careful control of process parameters such as pressure and voltage. However, maintaining careful control over process parameters and finding a suitable masking material for a certain film can be limiting factors in the use of RIE.
It is desired in certain applications to incline the end faces of cantilevered film waveguides relative to the axis of the waveguide, especially at air gaps between cantilevered and fixed waveguides. This cannot be readily achieved with RIE because the electric field lines in a plasma, which define the trajectory of the energetic ions doing the etching, terminate perpendicularly to the wafer surface. Thus, the desired oblique walls at the end faces cannot be obtained with silicon micromachining technology.
Silicon micromachined cantilevers carrying film waveguides have made use of films such as silicon dioxide (silica) and nitride. However, there are problems associated with fabricating micro-structures from the bulk of silicon substrates, such as the undercutting of convex corners which alters the shape of micro-structures, e.g. the inertial mass at the end of a cantilever. This prevents the reproducible fabrication of microstructures with 90° corners such as accelerometers. This problem can be partially corrected with the use of proper corner compensation in the mask layout, however this requires significant experimentation by trial and error to determine the correct compensation for each mask design. Another problem with using silica films for waveguides in micro-mechanical applications which is not encountered in micro-electronic processing is that thick films (up to 15 &mgr;m) are needed. The problem with such films is that they tend to crack and peel off due to the large residual stresses built-in during the deposition. Furthermore, the deposition of silica films is not compatible with silicon micromachining because it requires heating the wafer to a very high temperature which can affect the crystallinity of silicon on which anisotropic etching depends. Another drawback of high silica films is the necessity of deep RIE to form ridge waveguides, which is an expensive process and which is limited to etching thin films (below 1 &mgr;m) due to mask erosion.
Certain polymers have been used as waveguide materials. Low loss polymer waveguides have been most commonly achieved in poly-methyl-methacrylate (PMMA) or polystyrene. However, a limitation of PMMA is its poor thermal and environmental stability. For example, polyimides are affected by bases such as KOH or NaOH, which are used in anisotropic chemical silicon micromachining.
The use of polyimides on silicon presents problems in regards to wet and dry etching and to the mismatch in the coefficient of thermal expansion, so that polyimide films on silicon wafers tend to have limited utility in fabricating micromachined structures for optical waveguiding applications. For optical applications it is desired to cure polyimide films at temperatures not exceeding 250° C. in order to reduce optical losses.
The residual side wall angle of a wet etched air gap or slit is unpredictable due to the swelling when a developed film dries at elevated temperatures. This is aggravated in the case of a multilayered film wherein solvent attack at the interfaces between the layers results in uneven surfaces at the end faces of the film.
When a silicon wafer carrying a polymer film is cut or cleaved, the polymer film waveguide tends to lift off the cut edge of the wafer. The width of the lifted-off regions can extend up to 300 &mgr;m inward from the edge. This necessitates removing the entire lifted region of the film, such as for example by ablating with a laser to improve coupling of light in and out of the waveguide. However, this is problematic because it creates a relatively long step that the light must traverse between edge of the wafer and edge of the film. If this step is at the input edge of the waveguide where light is focused as a cone or wedge then a substantial portion of the light can be blocked off. Whereas if the step is at the output edge then it interferes with the collection of the light by a lens for feeding into a pick up fiber. This step is particularly problematic over silicon wafers. It was necessary to control the end face of a polymer channel waveguide within 5 &mgr;m from the cleaved silicon substrate edge in order to achieve acceptable coupling of the light (J. C. Chon and P. B. Comita, “Laser ablation of nonlinear-optical polymers to define low-loss optical channel waveguides”, Opt. Lett. 19, 1840, 1994). The cleavage of the silicon wafer must be done very carefully so that the least amount of film is peeled off at the cleaved edges.
To couple light in and out of waveguides single mode optical fibers are typically attached to single mode channel waveguides. This requires alignment of the axes of the fiber and waveguide with submicron accuracy. For example, V-grooves can be etched in silicon substrates and the alignment between the fiber and waveguide is adjusted while actively monitoring the coupling efficiency. At the point of maximum efficiency, the fiber is attached to the substrate. It would be desirable to couple light efficiently between single mode fiber and waveguide passively without monitoring the light intensity during the attachment.
It would therefore be desirable to provide a flexible polyimide substrate and a polymer laminate wherein the materials used for the different layers are highly compatible in terms of thermal, mechanical, chemical and machining properties.
It would also be desirable to cost-effectively fabricate, for example, by laser processing in a polymer or a polymer laminate a micro-structure, for example, a micro-mechanical cantilevered waveguide.
It would also be desirable to couple light efficiently and passively between a single mode fiber and a single mode waveguide.
It would also be desirable to fabricate an opto-mechanical device, such as an accelerometer incorporating a micromachined cantilevered waveguide.
SUMMARY OF THE INVENTION
This invention is directed to polished
Farah John
Sudarshanam Venkatapuram S.
LandOfFree
Polished polymide substrate does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Polished polymide substrate, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Polished polymide substrate will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3072368