Wireless Metrology in Semiconductor Manufacturing

1
Wireless Metrology in Semiconductor Manufacturing

• Costas J. Spanos
• FLCC Seminar
• 5/2/2005

2
Outline

• Historical perspective
• Hardware and software applications
• Breakthroughs that have yet to be realized
• Distributed control and diagnostics
• Hardware requirements

3
Past And Present

• Ideas and even patents circulated in the industry
  since the early 90s.
• No known implementation since UCB work started in
  late 1997
• Commercialized in 2000 by OnWafer Technologies
• Founders from UCB and LAM Research
• BakeTemp for Post Exposure Bake
• PlasmaTemp
• An expanding software applications suite

4
Smart Sensor Wafers - 1997

• In-situ sensor array, with integrated power and
  telemetry
• Applications
• process control, calibration,
• diagnostics monitoring,
• process design

5
Sacrificial, On-wafer Sensors
Temperature Compensated Design of Etch Rate
Sacrificial Sensor Each sensor has a buried
temperature reference
6
Wired Proof of Concept Implementation
100mm
7
Results - bakeplate

• Wafer placed on bakeplate, Rs measured for both
  sensors

8
Results - XF2 Etch

• Temperature-induced offset cancelled

13Å average
9
Wireless Prototype 1999Off-the Shelf
Components, Ni on Al, Solder Paste Mount
4 Sensors 4oC Accuracy 2oC Precision 1oC
Resolution Primary Batteries No Memory Fixed
Behavior 15grams/ 5mm
100mm
10
Thermal Stress Tests on Calibrated Plate
11
Autonomous Passage Through Wafer Track
Wafer arrives at chill plate
Non-uniform bake
Non-uniform cooling
Wafer leaves bake plate
12
RF isolation of wafer isnt it ugly?

• Standard temperature wafer covered with layer of
  epoxy
• Epoxy is transparent to infrared LED can be used
  for data transmission

13
Test results in plasma RF 50W, 0.76 Torr, O2
sensor 1
sensor 2
Words first Wireless Plasma Monitoring Late
1999!
sensor 3
sensor 4
14
Test results in plasma RF 100W, 0.76 Torr, O2
sensor 1
sensor 2
Words first Wireless High Power Plasma
Monitoring!
sensor 3
sensor 4
15
A Commercial Smart Dummy? (Sept 2000)
Courtesy of OnWafer Technologies, Inc.
16
Outline

• Historical perspective
• Hardware and software applications
• Breakthroughs that have yet to be realized
• Distributed control and diagnostics that go
  beyond lithography and etch
• Hardware requirements

17
Today The BakeTemp Sensor Wafer
polyimid
SiO2
1999 4 Sensors 4oC Accuracy 2oC Precsion 1oC
Resolution Primary Batteries No Memory Fixed
Behavior 15grams/ 5mm
2005 64 Sensors 0.05oC Accuracy 0.02oC
Precision 0.001oC Resolution Secondary
Batteries Memory Programmable/Adaptive
Behavior 1.5grams/ 3mm
module
Courtesy OnWafer Technologies
18
Much more than you ever wanted to know about Post
Exposure Bake
Overshoot
Steady
Heating
Cooling
200mm ArF 90nm 130oC 60sec
Chill
Courtesy OnWafer Technologies
19
Today The PlasmaTemp Sensor Wafer
Demonstrated in up to 7,000W Chambers Reusable Cle
an and safe enough to be adopted by all the top
tier fabs
20
On-Wafer Plasma Monitoring200mm Poly Etching
Reduced He
Routine He
main etch
pre-etch
de-chuck
over etch
21
Cool chuck - 200mm Poly Etching
main etch
Temperature fluctuations during main etch
22
Can see rotating magnetic field !
phase delay in temp fluctuation Can calculate
B-field period Can see rotation is clockwise
23
What are the Killer Apps?

• Temperature Monitoring has intuitive diagnostic
  value for some users.
• To expand use from Pilot to Production, we must
  address different, routine needs.
• Calibration
• Acceptance
• SPC/SQC

24
Post Exposure Bake Track Equipment Complexity is
Increasing
10 Years of Product Evolution
Multi-Zone Control
Single Zone Control
PEB Evolved from a Single Zone to Multi-Zone
Control System Why?
25
PEB Hotplate Thermal Profile Optimization System
Plate Type Specific Thermal Profile Modeling
Engine
AutoCal
Input
Offset Values Optimized for Both Within-Plate and
Plate-to-Plate Thermal Profile Uniformities
Offset Generator Engine
Output
Input
Baseline Thermal Profile Condition
BakeTemp OnView
OnWafer Technologies
26
PEB Temp Control
After
Before
Target 120oC
2.700oC
0.175oC
16 plates, 120 ºC Target
OnWafer Technologies
27
Spatial PEB/CD Distribution Correlation
28
PEB Hotplate Critical Dimension Optimization
System
AutoCal
AutoCD
Plate Type Specific Thermal Profile Modeling
Engine
Resist Litho Cell Specific CD Modeling Engine
Input
Input
Offset Values Optimized for Both Within-Plate and
Plate-to-Plate Critical Dimensions Uniformities
Offset Generator Engine
Output
Input
Input
Baseline Thermal Profile Condition
Baseline CD Profiles per Plate
Customer Provided
BakeTemp OnView
OnWafer Technologies
29
CDU Improvement
OnWafer Technologies
30
What we learned about Lithography

• Weight / Form Factor
• module electronics, height, wafer flatness
• Precision
• Speed
• Equipment Compatibility
• Contamination
• COST OF OWNERSHIP
• Lifetime
• Ease of Use

31
Killer Lithography Applications

• Monitor/Diagnose
• Multi-zone plates, complex exhaust systems,
  precise transport timing and placement.

Plate C is still stabilizing
Plate response errors.
32
What we learned from Plasma

• RF/heat tolerance
• Arcing resistance
• Shape/size
• Contamination
• Lifetime and reliability

33
Killer Plasma Apps

• Temperature monitoring
• Automatic Calibration/Diagnosis???
• Equipment Compatibility
• Power Range
• RF Tolerance
• Chemistry
• Contamination
• Arcing Tolerance

34
Outline

• Historical perspective
• Hardware and software applications
• Breakthroughs that have yet to be realized
• Distributed control and diagnostics
• Hardware requirements

35
The limitations

• Shape, Size and Weight
• Surface mount electronics min 2mm
• Needs thin film wafer interconnect to achieve
  10mm flatness
• Integrated (in Si) electronics still prohibitive
  for a consumable product
• Lifetime
• Battery
• Films (Plasma, CMP)
• Performance Envelope
• Battery up to 200oC
• Electronics up to 300oC

36
Shape Size and Weight Zero Footprint Wafer
Metrology wafer to monitor and map optical
reflectance and interference of surface layers
Prototyping a zero-footprint optical metrology
wafer for real-time monitoring of dielectric
film deposition/etching, resist
curing/development and metal etch end-pointing
Professor Nathan Cheung and Students
37
Lifetime

• Batteries
• Button cell secondary commercial solutions not
  available above 80oC.
• Primary solutions are a bit bulky but can go up
  to 140oC.
• Lithium-based thin film technologies still
  immature, not focused on high temp applications.
• Temporary solution screen commercial batteries,
  make them field replaceable (160oC ceiling, 10s
  of hours of operation)
• Long term solution wait for high temperature
  cathode thin film batteries (200oC ceiling, 100s
  of hours)
• Films
• Inherent Limitation in Etch just make it thick
  and provide visible thread wear marks.
• Might be replaceable in CMP

38
Performance Envelope

• High Temperature Electronics
• Off the shelf electronics (surface mount or
  hybrid) 180oC.
• Off the shelf electronics with custom clocking
  240oC.
• High Temperature Batteries will probably not
  happen above 200oC
• Parasitic power sources / caps

39
Expanding

• Aerial Image
• Chemical Mechanical Planarization
• Wet processing
• Ion Beam
• PVD
• CVD
• RTP???
• Automatic Deployment

40
Aerial Image Sensor Concept

• Data acquisition process

High spatial frequency aerial image
Aperture mask transmission
Low spatial frequency detector signal
41
Choosing Aperture Width Thickness (cont.)

• Near-Field Simulation

x
Aperture groups phase moving at max position
Aperture groups phase moving at min position
42
Outline

• Historical perspective
• Hardware and software applications
• Breakthroughs that have yet to be realized
• Distributed control and diagnostics
• Hardware requirements

43
Pattern Transfer Control
Aerial Image
Temperature, Plasma Voltage, Ion Current FF/FB
Control, chuck diagnostics
Exposure
Temperature Feed-forward control
Etch
PEB
Poly Etch System
Photoresist Removal
Etch
Etch
Thin Film FB/FF Control
Develop
Scatterometry Feedback Control
Scatterometry FB/FF Control
HMDS
Scatterometry Profile Inversion Feedback Control
Electrical Testing
44
Outline

• Historical perspective
• Hardware and software applications
• Breakthroughs that have yet to be realized
• Distributed control and diagnostics
• Hardware requirements

45
Automatic Deployment

• Wireless Wafers must resemble real wafers as
  much as possible.
• In a 300mm factory, they must reside in FOUPs and
  move around via overhead transport.
• Deployment, data collection, analysis and
  resulting actions must be automated.

46
Other Breakthroughs Needed

• Batteries!
• Or, maybe, parasitic power sources
• full-wafer gate cap is 2 orders of magnitude too
  small
• RF pickup coils in plasma have almost unlimited
  power
• High temperature electronics!
• Silicon Carbide Devices about 250oC
• Or, maybe thermal isolation for limited time

47
Conclusion

• Wireless Semiconductors Metrology has gone a long
  way since we started in 1998
• Three vendors
• Officially Adopted by Fabs and Tool Makers
• Best Known Method (BKM) in several applications
• Target processes remain in Lithography and Etch,
  but others are not too far behind
• Next generation of zero footprint metrology is
  likely to expand the application base even more