Semiconductor device manufacturing: process – Packaging or treatment of packaged semiconductor – Making plural separate devices
Reexamination Certificate
2000-05-16
2002-01-01
Niebling, John F. (Department: 2812)
Semiconductor device manufacturing: process
Packaging or treatment of packaged semiconductor
Making plural separate devices
C438S460000, C438S465000, C438S113000
Reexamination Certificate
active
06335224
ABSTRACT:
CROSS-REFERENCE TO RELATED APPLICATIONS
None applicable
BACKGROUND OF THE INVENTION
This invention relates generally to the field of microelectronics and more specifically to packaging of microelectromechanical systems (MEMS) and integrated microelectromechanical systems (IMEMS) devices.
Examples of MEMS and IMEMS devices include airbag accelerometers, microengines, optical switches, gyroscopic devices, sensors, and actuators. IMEMS devices can combine integrated circuits (IC's), such as CMOS or Bipolar circuits, with the MEMS devices on a single substrate, such as a multi-chip module (MCM). All of these devices use active elements (e.g. gears, hinges, levers, slides, and mirrors). These freestanding structures must be free to move, rotate, etc. Additionally, some types of microelectronics devices, such as microsensors, must be freely exposed to the environment during operation (e.g. for chemical, pressure, or temperature measurements).
For current commercially packaged MEMS and IMEMS components, the steps of packaging and testing can account for at least 70% of the cost. The current low-yield of MEMS packaging is a “show-stopper” for the eventual success of MEMS. Conventional electronic packaging methods, although expensive, are not presently adequate to carry these designs to useful applications with acceptable yields and reliability.
During conventional MEMS fabrication, silicon dioxide or silicate glass is a sacrificial material commonly used at the wafer scale to enable creation of complex three-dimensional structural shapes from polycrystalline silicon (e.g. polysilicon). The glass sacrificial layer surrounds and covers the multiple layers of polysilicon MEMS elements, preventing them from moving freely during fabrication. At this stage, the MEMS elements are referred to as being “unreleased”.
The next step is to “release” and make free the MEMS elements. Conventionally, this is done by dissolving or etching the glass sacrificial coating in liquid mixtures of hydrofluoric acid, hydrochloric acid, or combinations of the two acids. This wet etching step is typically done at the wafer scale in order to reduce processing costs. Alternatively, a dry release etch may be performed by exposing the wafer to a plasma containing reactive oxygen, chlorine, or fluorine ions. Herein, the word “wafer” can include silicon; gallium arsinide (GaAs); or quartz wafers or substrates (e.g. for MEMS structures).
After releasing the active elements, the MEMS devices can be probed to test their functionality. Unfortunately, probed “good” MEMS are then lost in significant quantity due to damage during subsequent packaging steps. They can be damaged because they are unprotected (e.g. released). Subsequent processing steps can include sawing or cutting (e.g. dicing) the wafer into individual chips or device dies (e.g. dicing); attaching the device to the package (e.g. die attach), wirebonding or other interconnection methods, such as flip-chip solder bumping, or direct metallization (e.g. interconnecting); pre-seal inspection; sealing of hermetic or dust protection lids; windowing; package sealing; plating; trim; marking; final test; shipping; storage; and installation. Potential risks to the delicate released MEMS elements include electrostatic effects, dust, moisture, contamination, handling stresses, and thermal effects. For example, ultrasonic bonding of wirebond joints can impart harmful vibrations to the fragile released MEMS elements.
One solution to this problem is to keep the original sacrificial glass coating intact for as long as possible. In one approach, the MEMS elements would be released after all of the high-risk packaging steps have been completed (including sawing of the wafer into chips). Another approach (which relates to the present invention) would be to release the MEMS elements at the wafer scale; apply any performance-enhancing coatings; re-apply a temporary, replacement protective coating prior to wafer sawing; and, finally, remove the protective coating after all of the high-risk packaging steps have been completed.
In order to reduce the costs of MEMS fabrication and packaging, it is desirable to perform as many fabrication steps at the wafer scale (e.g. before sawing the wafer into individual device dies). An example of a wafer scale process is deposition of performance-enhancing coatings on released MEMS elements (e.g. anti-stiction films and adhesion-inhibitors). Unfortunately, if these coatings are applied at the wafer scale (obviously, on released MEMS elements), then some of these performance-enhancing coatings may also be unintentionally deposited on the backside of the wafer. These unwanted coatings can interfere with the subsequent die attachment step. Also, the subsequent removal of these unwanted backside coatings can damage or contaminate the released MEMS elements.
Likewise, application of adhesion-promoting coatings (e.g. for die attachment) to the backside of the wafer may similarly damage the released MEMS elements by unintentional contamination or adsorption of harmful materials.
What is desired is a process that first releases the MEMS elements at the wafer scale; then applies all of the desired performance-enhancing coatings to the released MEMS elements; followed by re-application of a temporary protective coating; followed by cleanup of unwanted coatings unintentionally applied to the wafer's backside, followed by cutting the wafer into individual device dies.
Another desirable goal is to replace conventional wet etching processes with dry etching processes; to reduce costs, and because of an increasing emphasis on using environmentally friendly fabrication and cleaning processes. Especially for IMEMS devices, which can contain CMOS or Bipolar structures and other semiconductor materials, aggressive wet etchants used to release MEMS elements can damage the CMOS or Bipolar structures if they are not sufficiently protected. Standard photoresist protection used on CMOS or Bipolar chips may not provide sufficient protection from attack by acid etchants.
Wet etching processes can have other problems. Large hydrodynamic forces may be unintentionally applied to the fragile released MEMS elements, such as when agitating within a bath, and can fracture thin elements. Also, improper removal of any liquid film can create stiction problems resulting from capillary effects during the process of immersion in, removal from, and drying of liquid solutions. Using dry processing can eliminate these potential sources of damage.
What is also desired is using a dry etching process, preferably single-step, for removing protective coatings from MEMS devices. It is also desired that the dry etching process can be stopped before completely removing all of the protective coating, thereby leaving some desirable residual material that may reduce friction and may reduce tolerances between bearing surfaces, potentially reducing wobble in rotating gears and discs.
The list of desired objects described above could apply to microelectronics devices other than MEMS or IMEMS, such as microsensors. Microsensors also have a need for temporary protection of sensitive areas during packaging steps.
None of the approaches discussed above provides a low-cost, high-yield, high-capacity commercial wafer-scale, water-insoluble, protection film or coating that provides MEMS and IMEMS stabilization and protection during device packaging.
Use of the word “MEMS” is intended to also include “IMEMS” devices, unless specifically stated otherwise. Likewise, the word “plastic” is intended to include any type of flowable dielectric composition. The word “film” is used interchangeably with “coating”, unless otherwise stated. The phrases “released MEMS elements”, “released MEMS elements”, “sensitive area”, and “active MEMS elements” are used interchangeably to refer to the freely-movable structural elements, such as gears, pivots, hinges, sliders, etc.; and also to exposed active elements (e.g. flexible membranes) for microsensors (e.g. chemical, pressure, and temperature microsensors).
REFERE
Conley William R.
Peterson Kenneth A.
Lattin Christopher
Niebling John F.
Sandia Corporation
Watson Robert D.
LandOfFree
Protection of microelectronic devices during packaging does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Protection of microelectronic devices during packaging, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Protection of microelectronic devices during packaging will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2835176