Beth Pruitt

1
Course DevelopmentME342 MEMS Laboratory

• Beth Pruitt
• Assistant Professor
• Dept. of Mechanical Engineering
• Stanford University
• http//me342.stanford.edu

2
Course Goal Multidisciplinary learning and
entrepreneurship

• Micro/nanotechnology
• Scaling laws
• Transduction mechanisms
• Design/manufacturing
• Processes and tolerances
• Material selection and limitations
• Innovation
• Biomedical device engineering
• Biocompatibility
• Safety/Ethics
• Multidisciplinary language

3
Course Structure project based course

• Two quarter sequence
• Spring
• Predesigned masks, device and process
• Lab teams assigned for diversity of majors and
  backgrounds
• Qualify on equipment in Stanford Nanofab
• Summer
• Defined projects with partners (design starts
  early May)
• Complete design, fabricate, and test cycle
• Partners
• Internal research collaboration needs (e.g.
  Cardiology, Material Science, Cell Physiology)
• Industry defined challenges (e.g. Intel,
  Honeywell)

4
AIM Course Development Funding

• 10,000 grant to help start this course
• Winter quarter TA support to debug the process
  and prepare course materials
• Prototyping supplies (wafers, masks, etc.)
• Thank you!
• I gratefully acknowledged assistance this quarter
  that also came from
• Nu Ions donation of ion implant service for
  course
• Center for Integrated Systems new user grants to
  fund team clean room charges
• Goal is self-sustaining course model

5
Day 1

• About 70 students attended the first class
• 20 students were admitted based on questionnaires
  of background and interests
• 4 teams of 5 (max. capacity this year) formed
  with at least 1 EE, 1 Med/Phys/Chem/MSE, and 2-3
  ME students (will cross-list in EE, not
  advertised this time)
• 1 team of 5 overqualified applicants accepted
  to audit A and participate fully in B
• Very tough to turn students away, an exciting
  amount of interest in microfabricated solutions
  for new areas of research exists at Stanford

6
Week 1

• Safety training sessions for all new students to
  obtain clean room access
• Safety tours of SNF (Stanford NanoFab Facility)
• Written safety test
• Cleanliness training
• Instill sense of MEMS/clean room community

7
Week 2-6 Processing

• Fabrication in earnest under wing of senior MEMS
  research students for 4 weeks
• Incredible SNF staff support to ensure thorough
  qualification of students as users
• 2 weeks and 2 masks as independent users (with
  support net of teaching team)
• Analysis/simulation in parallel with fabrication
• Week 7-9 Measurements
• Package, test, signal condition and calibrate
• Compare theory and experiment

8
ME342A MEMS LaboratoryQ1 Project Fabrication
and Testing of Piezoresistive Cantilevers for
nN-mN Force Measurement

• Beth Pruitt
• Dept. of Mechanical Engineering
• Stanford University

9
Background for Project

• Sensors designed as part of a MEMS based system
  for force-displacement measurements of electrical
  microcontacts
• Sensors originally incorporated gold contact pad
  at tip to study thin gold films as
  MEMS/micro-electrical contacts

10
MicroContact example under studyFormfactor
MicroSpringTM Interconnects

• 1st and 2nd level interconnect
• pressure connection from the die to the printed
  circuit board, e.g. 2-sided memory module

with permission
11
Trends and opportunities Separable Contacts for
Packaging, Testing, Switching

• Shrinking interconnect pitch and size
• Smaller probes for test
• Smaller off-chip interconnects
• Thinner wafers and organic dielectrics
• Low force probing
• Thinner metal stackups
• To support continued miniaturization need low
  force, small size, and low contact resistance

12
Design of Contact Characterization Sensors

• Measurement over 6! orders of magnitude (2
  designs)
• Fabrication of thin film metals in-situ with
  standard processing (evaporated, sputtered,
  plated)
• 4-wire contact resistance measurement
• Measure force and contact resistance
  simultaneously

Gold Pad
measurement leads
Piezoresistor
13
Complete Experimental SetupForce-Displacement
Contact Measurements
Piezoactuator and controller
Laptop with Labview
GPIB card
Voltage Measurements (7 Channels)
DAQ card
14
Design

• Cantilever Beam
• Equivalent spring constant, K (N/m)
• Goal maximize range and sensitivity
• Constraints
• 100 micron travel in 5nm steps (actuator
  selection)

z
PKz
x
Piezoresistor linearity with strain (Matsuda
Kanda)
Linear elastic beam equations (Young)
15
Design Space
40µm thick cantilever Pmax _at_ 100 µm 10mN
L max(m)
Kmin (N/m)
1E01 1E00 1E-01 1E-02 1E-03 1E-04
A require L gt w B piezo ? limited C linear
elastic ? limited D cantilever design 1 800µm
x 3mm x 40µm K 85 N/m
B
Kmin (N/m)
C
D
L max(m)
A
16
Design Space
25µm thick cantilever K 1.3 N/m Pmax 0.6mN
L max(m)
Kmin (N/m)
A require L gt w B piezo ? limited C linear
elastic ? limited E cantilever design 2
400µm x 6mm x 25µm K1.3 N/m
B
Kmin (N/m)
E
L max(m)
A
17
Comparison to AFM cantilever
W
L
L 180 ?m W 35 ?m t 2 ?m K 1.3 N/m
L 6 mm W 400 ?m t 25 ?m K 1.3 N/m
3.6mm
1.6mm
Custom Cantilevers K from 1.3 to 85 N/m 100?m
displacement range
Park Scientific dlevers K from 1.3 to 16
N/m Small displacement range
18
Cantilever Fabrication (omit gold pads!)
aluminum
doped conductor, B
silicon
SiO2
doped piezoresistor, B
aluminum
piezoresistor
silicon
conductor
7 mask process 25 micron SOI, 300micron handle
19
Processing alignment
Pattern resist and light Si etch (3000 angstroms)
to define alignment patterns
20
Processing protective oxide
Strip resist Grow protective screeening oxide
250 angstroms
21
Processing piezoresistors
Pattern resist 50 keV boron implant for
piezoresistors, e.g. dose 1e15 ions/cm2
22
Processing conductors
Pattern resist 50 keV boron implant for
piezoresistors, dose 1e16 ions/cm2
23
Processing oxide/anneal
Strip damaged oxide Wet Oxidation 900C, 2500A,
2 ?m depth, ?piezo 130 ?/? , ?conductors 45
?/ ?
24
Processing contacts
Open oxide Strip Resist Sputter 0.5 ?m
Aluminum Pattern and etch Al
25
Processing DRIE
Frontside Etch- 1.6 ?m resist, open oxide, etch
Si to buried oxide, 1.6 ?m resist frontside
protect
Backside Etch-, 10?m resist, open oxide, etch Si
to buried oxide, wet etch box
26
Cantilever Fabrication (shown w/ gold)
aluminum
doped conductor, B
gold
silicon
SiO2
doped piezoresistor, B
aluminum conductor piezo
gold
27
Cantilever SEM
28
ME342 Cantilevers-7 Masks, no Gold

• Mask Levels 1-3 completed by TAs
• Alignment Marks/Cantilever outline
• Conductive Interconnect Implants
• Piezoresistive Region Implants
• Team Processing Mask Levels 4-7
• Complete in Labs 2-6 plus some time outside of
  lab for levels 6 and 7
• Qualify individually on wetbenches, litho, DRIE
  during labs of ME342
• Note team stuck at mask 5 until all team members
  qualify on required equipment!

29
ME342 Processing

• Each team completes processing with same mask set
• Each team has 5-6 wafers to process
• 2 SOI wafers fully released by DRIE (300µm)
• 3 test wafers partially processed (Noise only)
• Sensor measurements, 2 die per person
• Packaging and Signal Conditioning
• Testing and Measurements (Sensitivity Noise)
• Analysis

30
Interconnect Levels wire bonding to dip package
0th level interconnect
1st level interconnect
2nd level interconnect
Silicon die
Package
Printed circuit board
31
Cantilever Calibration
Signal analyzer
Laser vibrometer
Vdisplacement
Vstrain

• Piezoresistor Bridge Voltage vs. Displacement
• Measure at resonant frequency of cantilever
• Typical sensitivity 1mV/µm
• Noise spectrum of piezoresistor
• lt 0.1µV/?Hz or 80pN/ ? Hz at 1Hz

32
Cantilever Calibration time frequency
?0
?1
?3
?2
?n 1st resonance K spring constant mc
concentrated mass md distributed mass
33
ME342A Analysis

• Simulate piezoresistor values (TSUPREM4)
• Each wafer receives different dose/anneal set,
  each student assigned a particular wafer to
  analyze
• Predict spring constant and gage factor
• Determine sensitivity and noise of cantilevers
• compare analysis by beam equations and noise
  characteristics to measurements
• Comparisons and Conclusions
• 15 min. talk 6/3, short report of results

34
ME342B Design Projects

• Project and team assignments early May
• Initial designs due end of May
• Mask designs must be submitted before start of
  summer quarter!
• Processing and testing completed in ME342B
• Seminars, team meetings and lots of lab time in
  summer quarter
• Project results Conference papers???
• e.g. MEMS05, ASME05, send 1 author per paper

35
Potential Projects for ME342B 2004

• Radial 100 strain gage for measuring deformation
  in animal model blood vessels, e.g. rat aorta
  (Taylor, ME/cardiology)
• Integrated touch sensitivity system for
  neurological examination (Goodman, molecular
  cell physiology)
• Out-of-plane actuated stage (Intel mirror
  steering)
• Active thermal isolation package (Honeywell chip
  scale atomic clock)
• Implanted piezoresistor design rule formulation
  (Pruitt)
• Optimization of miniature blood pressure sensor
  sensitivity by process and geometry (Feinstein,
  pediatric cardiology)
• Coupled beam microresonators for molecular assay
  (Melosh, MSE)

36
9 weeks to go and the whole Summer!

• A class full of enthusiasm
• The best teaching assistants anyone one could ask
  for
• A supportive clean room environment and
  technical staff
• A rich tradition of innovation in manufacturing
  and design
• Cool projects inspired by local industry and my
  Bio-X collaborators

37
Thank you AIM for your help and support!

• 2004-2005 MEMS projects wanted!
• Team of 3-4 multidisciplinary students May plus
  summer
• Innovative ideas, unique facilities, excellent
  coaching from faculty and industry
• Projects on the margin, something a company would
  like to try or know if it works but doesnt have
  manpower, expertise, or resources for it