|
Engineering Design Standards and Constraints for the Micro-electro Mechanical Systems (MEMS) Project (September 2006) |
|
Andrew R. Lingley, Student Member, IEEE, Brad Pierson, and Matthew Leone, Student Member, IEEE |
Abstract�this paper discusses specific design choices made during the MEMS project in order to satisfy various constraints and standards that are regularly overlooked.� The MEMS project entails manufacturing various MEMS devices and incorporating the design process into an Electrical Engineering undergraduate introductory fabrication course. �Constraints covered fall into eight categories: economic, environmental, sustainability, manufacturability, ethical, health and safety, social, and political.� These categories are defined and examples of active design solutions are given for each applicable constraint.
Index Terms�constraints, MEMS,
microfabrication,
|
T |
HE
MEMS project goals are to create working MEMS devices and to integrate the
results into an Electrical Engineering course at
�� Economic constraints are defined here as those which arise because of limited funds, and are often a result of a project budget.� Budgets show monetary resources for various aspects of a project, such as salaries and equipment.� Although there was no formal budget describing the amount and distribution of monies for the MEMS team, the acknowledged budget was to use the utmost frugality in every way possible.
�� The design took into account economic limitations on several occasions.� Typically, clean room microfabrication processing equipment is very expensive, and therefore purchasing any significant new lab equipment for the project was not feasible, nor was it practical.� As a consequence, all processing steps were designed to work with lab equipment already within the clean room.
�� Equipment in the MMF includes an acid wet bench, an oxidation furnace, two diffusion furnaces, a sputterer, a plasma etcher, a spin coater, an alignment and exposure system, a vaporizer, and a myriad of characterization equipment.� Equipment restrictions led to processing and minimum feature size limitations.� For example, no procedures were designed for chemical vapor deposition or ion implantation, as no means to perform these tasks were available.� Also, feature sizes were designed to be no smaller than 10μm, a value that was chosen due to the resolution of the photolithographic techniques available within the clean room.� Particularly, the mask alignment system, unless operated by an expert, cannot consistently produce alignment errors of less than 1-2 micrometers.� These sizes also push the limits of the MMF ultraviolet exposure technique and photoresist developer.
��
Another pressing economic concern surfaced as a result of the integration
of the project into an undergraduate course.�
Choices for masks, wafers, and metal contacts were limited to keep
student lab costs at a reasonable level.�
The choice was made to order relatively inexpensive masks manufactured
by the
��
Environmental constraints are ecological issues or problems caused by
physical surroundings.� For instance, the
clean room sinks are connected to
�� Also, the immediate environment for the MEMS project is a clean room, and therefore special procedures were designed to ensure the clean room remains as sanitary as possible.� Smoking is not allowing within one hour of entering the clean room, and shoes with large treads are not allowed. �Gloves, two layers of booties, lab coats, and hair nets are required to minimize contamination (hair nets over the face are required for individuals with beards).� The lab manual will also be posted on the internet, because standard paper releases a large number of particles into the air.
�� Sustainability here refers to both the longevity of the devices and the ability of the project as a whole to remain a useful tool for students to learn along with an undergraduate microfabrication course.� Assuming that inexperienced students will handle wafers for a significant portion of the lab, there is always the possibility that the wafers may be mishandled.� Therefore, devices of various sizes and strengths were designed to ensure that some would be operational even after bumps and drops.� Also, the lab manual was designed to be thorough, incorporating many variations in process steps to account for changes that may be made to the course in future years.
�� Manufacturability is a measure of a product�s ability to be reproduced in a straightforward and timely manner.� As part of the initial project goals, the lab manual will provide complete instructions to ensure excellent manufacturability.� In addition, the Electrical Engineering course this lab will supplement spans only one semester, so the manufacturing process was designed to take no more than fifteen weeks.� The course lab meets for two hours, once weekly, which placed considerable constraints on processing steps.� One such limit was on the number of masking steps needed to complete the task; within fifteen weeks, including various other processing, it was determined that no more than six masking steps could be completed in the allotted time.� Although more complicated devices could have been planned (e.g. gears, movable mirrors, and other actuators), only accelerometers, pressure sensors, and piezoresistors were chosen as acceptable devices to finish in one semester using only six masks.
�� As with any laboratory class, proper precautions were taken in order to ensure the health and safety of all involved parties.� Standard microfabrication techniques generally incorporate potentially dangerous equipment and various toxic chemicals.� An oxidation furnace, for example, operates at temperatures up to and exceeding 1000 �C.� The lab manual describes proper procedures for loading and unloading wafers from the furnace; this includes allowing sufficient time for the oven to cool, using oven mitts, and allowing even more time for the wafers to cool before handling.
�� Secondly, many hazardous chemicals, namely hydrofluoric acid, hydrochloric acid, solvents, and photoresist, are used during the manufacturing process.� Many other chemicals used are potentially dangerous, and therefore all documentation on uses, hazards, and emergency planning is carefully described within the lab manual. �For example, Hydrofluoric acid (HF) is extremely dangerous; it can be fatal in small doses even with immediate medial care, and quickly dissolves glass.� Therefore, it must be contained in polyethylene or Teflon containers.� Photoresist fumes can also be potentially harmful, so the lab manual states procedures to ensure proper ventilation of workspaces where photoresist will be used [1].
�� Thirdly, in one instance an etchant was discounted partially on the premises of health concerns.� Ethylene Diamene Pyrocatecol (EDP) would have been the preferred etchant for the anisotropic etching of silicon because of its fast etch rate, good selectivity, and its high <100> to <111> plane etch ratio.� Unfortunately, EDP is also extremely corrosive, highly carcinogenic [2], and can be inhaled, ingested, or absorbed through the skin.� Consequently, EPD was not used as a silicon etchant.
�� Lastly, lab attire rules were created with safety in mind.� Lab goggles must be worn during many process steps including photoresist application, wafer cleaning, and etching.� Students must also wear close-toed shoes as well as pants at all times in the lab.� Additionally, a rubber lab apron, a face visor, and a second pair of gloves must be worn when working on the acid wet bench.
�� No design specifications were decided upon to overcome any social constraints.
�� The term �ethics� encompasses the concepts of morality and responsibility.� Many very specific designs for accelerometers, pressure sensors, and piezoresistors exist in text books, journal articles, and technical papers.� Therefore, with this in mind, the MEMS project team made certain to in no way plagiarized or steal design specifics from any outside party.�� The designs created for this class, although based on commonly known ideas, are original designs.� The lab manual format also incorporated references for basic processing procedures.
�� Although governmental politics were not in any way relevant to the project, University politics were of some concern.� The MMF is not open to the public, and consequently there is an application process required to gain entry to the lab.� Therefore, all necessary paperwork is included in the lab manual which can be printed and turned into the proper authorities.
References
[1] T. Kaiser (2006) Available: http://www.coe.montana.edu/ee/tjkaiser/EE407/notes/EE407-02ChemicalSafety.pdf
[2] R. B. Darling (date not included) Available: http://www.ee.washington.edu/research/microtech/cam/PROCESSES/PDF%20FILES/WetEtching.pdf
A. Lingley is a Senior at
�� His previous work experience includes two
National Security Internships at the Pacific Northwest National Laboratory in
�� He is also the IEEE Student Chapter Chair
and Tau Beta Pi Engineering Honor Society treasurer.
B. Pierson is a Senior Electrical Engineering student at Montana
State University Bozeman.
�� Drawing from past experience, he is trying
to further his knowledge in the power engineering field concentrating
particularly on system transmission.�
Past work experience, including time spent at a USGS facility working
with data acquisition from satellites and time spent at a regional power
company near
M. Leone is in his senior year at
�� In the summer of 2006 he worked in the
Montana Microfabrication Facility under the supervision of MSU professor Todd
Kaiser
�� Matthew Leone is a member of the IEEE and
Tau Beta Pi engineering societies.
Manuscript received September 21, 2006. �This project was funded by Todd Kaiser, an
Assistant Professor of Electrical Engineering at
A. Lingley is with the Montana State University
Electrical Engineering Department,
M. Leone is also with the Montana State University
Electrical Engineering Department,
B. Pierson is also with the Montana State University
Electrical Engineering Department,