sfr3 sensors 111402 - university of california, san...
TRANSCRIPT
11/14/2002 SFR Workshop - Sensors
1
Sensors
SFR WorkshopNovember 14, 2002
Eray Aydil, Nathan Cheung, Bruce Dunn, Kameshwar Poolla, Costas Spanos
Berkeley, CA
11/14/2002 SFR Workshop - Sensors
2
Sensors• Broad goals of project
– Develop platform for fully autonomous sensor arrays capable of operating in harsh environments
– Develop novel sensors and transduction schemes to be used with the autonomous platform
• Applications– Process modelling– Equip design/process optimization
• Outline– Battery development (Dunn)– Encapsulation (Cheung)– Plasma sensors (Aydil)– Tomography-based sensors (Poolla)
– Process Control– Diagnostics
11/14/2002 SFR Workshop - Sensors
3
Development of Lithium Batteries for Powering Sensor Arrays
SFR WorkshopNovember 14, 2002
Bruce DunnUCLA
Student contributors: Tim Yeh and Daren Chow
2003 GOALS: a) Battery operation between room temperature and 150°Cb) Battery survivability to sensor soldering operations
11/14/2002 SFR Workshop - Sensors
4
Project Objectives
In order to provide on-board power of SMART wafers, a low profile, thermally stable, high energy density battery must be used.
Battery Requirements:Temperature capability: 150°CVacuum (10-2 torr).Operating voltage: > 2.5 VDischarge current: 2 mA.Discharge time: > 10 minutes.Low Profile: 500 µm or less.Area: Less than 3 cm x 3 cm.Rechargeable; 10 cycles.
State-of-the-art batteries do not meet theserequirements:- Temp. limitations (button cell, Panasonic)- Too thick (button cell)- Low capacity (thin film)
Panasonic primary battery;60°C temperature limit
11/14/2002 SFR Workshop - Sensors
5
UCB: Fabrication of wells on a silicon wafer
UCLA: Fabrication of lithium polymer battery; thickness < 500 µm
UCB: Plasma-assisted bonding
Further tests will be carried out at UCLA to evaluate battery performance after bonding.
Current Program: Improvements in Battery PackagingPlasma-Assisted Bonding Approach (with N Cheung)
11/14/2002 SFR Workshop - Sensors
6
Current Program: Higher Operating Temperatures Battery operation between room temperature and 150°C
New nanocomposite electrolyte incorporated into battery structure• successful operation at 125oC and 150oC; designed capacity demonstrated• battery cycled several times between RT and 125 or 150oC without difficulty
0
0.5
1
1.5
0 2 4 6 8
Cap
acity
(mAh
)
25°C 125°C 150°C
0
0.5
1
1.5
0 2 4 6 8
Cycles
Cap
acity
(mAh
)
25oC 125oC 150oC
Batteries discharge nicely at elevated temperature. No capacity loss when alternate between RT and 125oC or 150oC
2
2.5
3
3.5
4
0 300 600 900 1200 1500 1800 2100Time (S)
Volta
ge (V
)
2 mA discharge for 30 min. at 25, 125 and 150oC
125 C
25 C150 C
11/14/2002 SFR Workshop - Sensors
7
SST/Gel Electrolyte/LiCoO2
22.5
33.5
44.5
0 2000 4000 6000 8000 10000Time(s)
Volta
ge v
s. L
i/Li+
First generation design: Stainless steel anode
SST anode
SST current collectorPolymer electrolyte
LiCoO2 cathode
• Successful lithium deposition atSST/electrolyte interface
• First discharge works well, then cell degrades rapidly• Capacity decreases with cycling;
0.31mAh 0.04mAh in 5 cycles
Current Program: Higher Operating TemperaturesDevelopment of Lithium-Free Battery
Lithium-free approach avoids problems with exposing battery to temperatures above Tm of lithium (180oC): lithium electrode created after exposure
11/14/2002 SFR Workshop - Sensors
8
Current Program: Higher Operating Temperatures Development of Lithium-Free Battery
• In-situ deposition of lithium at carbon/polymer interface• Only small capacity loss in 7 cycles:
0.33mAh 0.28mAh• Good reversibility and cycling
SST current collectorPolymer electrolyte
SST current collector
LiCoO2 cathodeCarbon paper anode
SST/Carbon Paper/Gel Electrolyte/LiCoO2
22.5
33.5
44.5
0 2000 4000 6000 8000 10000
Time(s)
Volta
ge v
s. L
i/Li+
Results show that the anode host material is critical for in-situlithium formation
• Second generation design: Thin carbon anode
11/14/2002 SFR Workshop - Sensors
9
Summary and Research Plans
• Accomplished Goals for 9/30/2002New nanocomposite electrolyte integrated into batteryLithium-free battery demonstrated (in-situ lithium formation)
• The Next 6 MonthsStill thinner batteries for plasma-assisted bonding approach
Improve the electrochemical properties of lithium-free battery;demonstrate full-size battery operation and temperature testing
Battery is about the same height as the etched well (300~320μm)
11/14/2002 SFR Workshop - Sensors
10
Smart Wafer Integration byWafer Bonding and Layer Transfer
Yonah Cho , Zhongsheng Luo, Zhengxin Liu, Vorrada Loryuenyong,
and Nathan Cheung
Collaborators : Bruce Dunn (UCLA)
2002 GOAL: Prototyping a SMART wafer with optical metrologyand encapsulated power source
11/14/2002 SFR Workshop - Sensors
11
Motivation
Self-powered sensor wafer Optical metrology capabilityWafer-scale deposition/etching uniformity and end-point mappingApplicable to a large variety of materials (low-k materials and metal)Monitor for hostile processing environments ( Plasma, wet etching, CMP).
SiBattery Detector Photon emitter
Data transmission
Photon emitter Data collecting, processing, storage unit
MEMS SensorDielectric layer
11/14/2002 SFR Workshop - Sensors
12
Milestones2002•Demonstrate subsystem with encapsulation power
2003•Demonstrate integration of signal processing systems with MEOMS on sensor wafer
Progress
•Low-temperature encapsulation demonstrated
•Completion of optical source, switch, and power components
11/14/2002 SFR Workshop - Sensors
13
Justification for ApproachSimulated Effect of incident wavelength on the reflectance
Incident wavelength (A)
Thic
knes
s (µ
m)
θ=35°460nm
The reflectance fluctuation periodicity (2p) is not very sensitive to the incident wavelength: ∆p/d ≈0.4% at λ=460±10 nm, where d is the film thickness.
λ=[400nm,600nm]
λ=[400nm,600nm]
SiO2 onsmartwafer
11/14/2002 SFR Workshop - Sensors
14
Progress: Prototype Components
Solid-stateThin film battery
-
+
Base Si Wafer
Top ViewGaN based LED
V
Electrolyte: LiPONLi based electrodesVmax = 4.2 VoltsCapacity = 0.4mA-hr
CdS optical switch
R dark = up to 10 MΩR light = ~ 10 Ω
λmax = 462 nmI = 20mA @ V< 4V
11/14/2002 SFR Workshop - Sensors
15
Wafer cleaning:wet and dry (O2 plasma)
Lid Si Wafer
Base Si WaferBattery
Optical switch
LED
Bonding
Insertion of components
BatteryOptical switch
11/14/2002 SFR Workshop - Sensors
16
Battery
Optical switch
LED
System-Off
System-On
11/14/2002 SFR Workshop - Sensors
17
Filters for Molecular Spectroscopy
Filter 1
Filter 2
θ
λ2 ± ∆λ
Detector
Tλ
λ 2
Ref
lect
anc e
( R) &
Tra
nsm
it tan
c e ( T
)
400 600 900 1000
λ2 = 700 nm
500 800
0.0
0.0
0.5
1.0
0.5
1.0
Filter 1
RR
R
Multilayer TiO2/Si3N4/ SiO2 filters Incidence angles:
Filter 1: θ = 35 °Filter 2: θ = 0 °
Filter 2T
T
Wafer surface
λ1 = 460 nm
λ1 λ2Surface molecules Simulated R and T
λ (nm)
11/14/2002 SFR Workshop - Sensors
18
Summary
• Design and prototyping of sensor wafer with optical metrology
• Completion of optical source, switch, and power components
•Photo-detector and filter components to be added• Demonstrate thickness monitoring capability
Goals for 2002-2003
11/14/2002 SFR Workshop - Sensors
19
Ion Flux Uniformity in Plasma Reactors
SFR Workshop & ReviewBerkeley, CA
November 14, 2002
Tae Won Kim and Eray S. AydilUniversity of California Santa Barbara
Chemical Engineering Department
2002 GOAL: build and demonstrate 8” diameter on wafer ion flux probe array in industrial plasma etcher with external electronics
9/30/2002.
11/14/2002 SFR Workshop - Sensors
20
Motivation and Problem Statement• Uniformity of ion bombardment flux is critical to plasma
etching because it determines the uniformity of etching and etching profile evolution.
• Uniform plasma etching processes are developed by trial and error and uniformity requirement is in addition to several otherconstraints imposed on selectivity, anisotropy, etch rate, etc.
• Our goal has been to provide tools that allow fundamental understanding of the factors that affect the plasma and etching uniformity in realistic plasma etching chemistries.
11/14/2002 SFR Workshop - Sensors
21
Experimental Approach•We have been developing and using on-wafer ion flux probe arrays capable of mapping J+ (r,θ) on a wafer and using this probe array to study the factors that affect the plasma and etching uniformity in realistic plasma etching chemistries.
Kim et al., Rev. Sci. Instrum. 73, 3494 (2002)
11/14/2002 SFR Workshop - Sensors
22
Progress SummarySince Program Started
Finished experiments in Lam TCP reactor with 8” wafers and realistic etching chemistries (Cl2/SF6/Ar/He/O2/HBr).
Implemented modeling to understand some of the “wall effects” onetching uniformity in Cl2 etching of Si in TCP reactors.
Imaged low frequency periodic spatio/temporal variations in ion flux in inductively coupled SF6 plasmas
Since April
Designed electronics and built an ICP reactor to implement probearray on rf-biased electrodes.
Developed a theoretical framework and conducted simulations to understand and predict ion flux uniformity in plasma mixtures.
11/14/2002 SFR Workshop - Sensors
23
Nondimensionalized Governing EquationsElectron energy balance
Charged particle mass balance
Boundary conditions
0)exp()83.3()1(exp~ 21
2 =−+−−+∇− eT
effeffinelasticrfeTelasticT DaJBDaDa θ
θβξςθθθ
ereceT
iziz DaDa γθθ
θβ
θαθ −−=∇+∇− ++ )exp()~(~~ 2 E
−+−− −=∇−∇− θθγθθαθ rece
att DaDa)~(~~ 2 E
−+ −= γθθθ e
0=∇ Tθ
TPe θϑθαθ +++ =∇+∇− )~(~~ 2 E
0=−θ
11/14/2002 SFR Workshop - Sensors
24
Framework for understanding Ion Flux Distribution
Ion flux distribution
Skin depthPenetration depth ofplasma powerδ~1/ne
0.5(SF6 >Cl2 >Ar)
p
LBδ2
=
e- energy relaxation lengthdiffusion length w/o losingenergyλe~ 1/ σ (Ar > Cl2 >SF6)
MDRkTDa
e
Reelastic )1(
22
2
γ−><
=
><−=
ee
Rgeffinelastic kTD
HkRnDa
)1(5.1
02
γ
Diffusion of ionsD+ ~ 1/p
+
=DRnk
Da Rgiziz
20
MkT
DRPe eR ><
=+
Ion loss mechanismsurface loss: loss to the wallsvolume loss: i-i recombination
+
+ ><=
DRnkDa Rrec
rec
20
MkT
DRPe eR ><
=+
11/14/2002 SFR Workshop - Sensors
25
Relative magnitudes dimensionless #s determine the ion flux uniformity
15 10 5 00
5
10
z (c
m)
5 10 150
5
10
r (cm)
kTe (eV) n+ (1010 cm-3)
Ar
2.3
2.4
1.8
4.2
5 10 150
5
10
15 10 5 00
5
10
r (cm)
z (c
m)
1.8
3.6
0.5
5.5
Cl2
15 10 5 00
5
10
r (cm)
z (c
m)
5 10 150
5
10
1
6
0
2.6
SF6
B~13Daelastic~0.01Dainelastic~1
B~4Daelastic~102
Dainelastic~104
Darec ~ 102
B~5Daelastic~1Dainelastic~102
11/14/2002 SFR Workshop - Sensors
26
Ion Flux Uniformity in Ar, Cl2, SF6
1.00
.92
1.00
0.92
-15 -10 -5 0 5 10 150.0
0.2
0.4
0.6
0.8
1.0
Imax = 1.42 mA/cm2
Imax = 1.39 mA/cm2
Nor
mal
ized
ion
flux
r (cm)-15 -10 -5 0 5 10 150.0
0.2
0.4
0.6
0.8
1.0
Imax = 0.35 mA/cm2
Imax = 0.42 mA/cm2
Nor
mal
ized
ion
flux
r (cm)-15 -10 -5 0 5 10 150.0
0.2
0.4
0.6
0.8
1.0
Imax = 0.22 mA/cm2
Imax = 0.12 mA/cm2
Nor
mal
ized
ion
flux
r (cm)
1.00
0.82
Ar Cl2 SF6
cross sections
11/14/2002 SFR Workshop - Sensors
27
Can be extended to mixtures
∑=
kkme
emixe m
kTD,
, ν
Electron diffusivity
Elastic collision loss
kelastick
kmixelastic DayDa ,2
, ∑= effkinelastic
kk
effmixinelastic DayDa ,
2, ∑=
Inelastic collision loss
∑=k
kizkmixiz DayDa ,,
Production and consumption of ions
∑=k
kreckmixrec DayDa ,, ∑=k
kattkmixatt DayDa ,,
Positive ion loss by Bohm flux
∑><
=+
kkk
eRmix My
kTDRPe
11/14/2002 SFR Workshop - Sensors
28
Comparison between Experiments and SimulationIout
0 20 40 60 80 1000.0
0.5
1.0
1.5 Measured Predicted
Ion
flux
SF6 mole fraction0 20 40 60 80 100
0.0
0.5
1.0
1.5 Measured Predicted
Ion
flux
Cl2 mole fraction
Ar/SF6 Ar/Cl2
11/14/2002 SFR Workshop - Sensors
29
Comparison between Measured and calculated Uniformity
-15 -10 -5 0 5 10 15
0.6
0.8
1.0
Predicted Measured
Nor
mal
ized
ion
flux
r (cm)
-15 -10 -5 0 5 10 150.6
0.8
1.0
Predicted Measured
Nor
mal
ized
ion
flux
r (cm)
-15 -10 -5 0 5 10 150.6
0.8
1.0
Predicted Measured
Nor
mal
ized
ion
flux
r (cm)
-15 -10 -5 0 5 10 150.6
0.8
1.0
Predicted Measured
Nor
mal
ized
ion
flux
r (cm)
0 % Cl2 10 % Cl2
50 % Cl2 100 % Cl2
11/14/2002 SFR Workshop - Sensors
30
2002 and 2003 Goals• September 30th, 2001 Build and demonstrate
Langmuir probe based on-wafer ion flux probe array using external electronics.
• September 30th, 2002 Build and demonstrate 8”on-wafer ion flux probe array in industrial plasma etcher with external electronics.
Continue to work towards getting the on wafer probe array to work with rf bias.
Increasing time resolution of the probe array.
11/14/2002 SFR Workshop - Sensors
31
Electrical Impedance Tomographybased Metrology
SFR Workshop & ReviewNovember 14, 2002
Michiel Krüger, Kameshwar Poolla, Costas SpanosBerkeley, CA
2003 GOAL: to demonstrate the feasibility of an EIT based sensor to measure plasma induced potential at wafer surface,
by 9/30/2003.
11/14/2002 SFR Workshop - Sensors
32
Motivation• In-situ Plasma Sensing is very important:
– Diagnostics (drifts, detection of process instabilities)– Design (electrode configuration in plasma tools)– Control (to reduce process variability)– Model verification
• Current metrology has shortcomings– Expensive and complex– Invasive, sometimes destructive– No real-time data available, only time-integral
⇒Develop class of sensors based on EIT– Spatially resolved– Time resolved– Inexpensive
11/14/2002 SFR Workshop - Sensors
33
The Problem
What was the state of the wafer during processing?
processingequipmentwafers to
be processed finished wafer
11/14/2002 SFR Workshop - Sensors
34
stage B - EITstage A – special wafer design
stage C – modeling
Our Approach
conductivity profile
physical orchemical processwafer
state
( )θ,rX ( )θσ ,r
EIT based sensor
on wafer
measured conductivity profile
( )θσ ,ˆ rinverse physicalor chemical
modelestimatedwafer state
( )θ,ˆ rX
11/14/2002 SFR Workshop - Sensors
35
Electrical Impedance Tomography• Widely used in biomedical applications, geology and
non-destructive product testing• Basic idea:
– force known current through interior of object from electrodes at edge
– measure potentials at edge electrodes• Stage B – EIT:
– reconstruct conductivity σ profile from potential measurements V at edge measurements
– simplified explanation•• parameterize estimate • minimize using NLP
electrode
Iin
Iout
( )θσ( )( )θσϕ ˆ−V
( )σϕ=V
11/14/2002 SFR Workshop - Sensors
36
• Hot spot:• Inverse algorithm works well
( )( ) ( )
v
ee yyxx
Keyx σσ22
,−+−
−
=∆
actual conductivity profile estimated conductivity profile
Simulation Example: two hot spots
hot spot centers
11/14/2002 SFR Workshop - Sensors
37
wafer
electrodesensing area
EIT based sensor wafer• Simple electronics• Electrodes placed at edge• Minimal wiring • Spatial information• Minimal processing• Many applications:
⇒ Etch rate/uniformity⇒ Plasma induced potential at wafer surface– Temperature
11/14/2002 SFR Workshop - Sensors
38
Etch Rate: Stage A – Wafer Design• Conductance of doped Poly-Si function of thickness• Simple process flow• Easy to test in XeF2 etcher
Al electrode Oxidized
wafer
dopedPoly-Si
32 equally spaced Al electrodes
Doped poly-Si disk (r = 3.5 cm)
Pads to connect via edge-board connector to DAQ
11/14/2002 SFR Workshop - Sensors
39
loadingeffect
∆σs[
/Ω]
sensor wafer
time [s]
4 point probe
1cm1 inch, rotated 15° CCW
σs = 0.156 /Ω
exposed to XeF2
In-line Results: XeF2 etch• Absolute measurements not possible due to:
– unaccounted contact resistances– nonlinearities in DAQ– finite size of electrodes
11/14/2002 SFR Workshop - Sensors
40
nanospec nanospec
In-line Results: Si wet etch∆ thickness after Tetch =45s ∆ thickness after Tetch =75s
EIT EIT
wet etch roughens surface
⇓∆telec > ∆tphys
11/14/2002 SFR Workshop - Sensors
41
Plasma Induced Potential• Conductance of transistor channel
function of gate potential• Build resistive network of
transistors on wafer• Transistor gates exposed to plasma
• 4 nodes ⇒ 12 current patterns• network of transistors implemented
in Hspice
source drain
plasma
gate+ + + +
Vgs > Vt , σ > 0Vgs < Vt , σ ≈ 0
source drain
Iin
Iout •Hspice simulation to generate “measurements”•EIT algorithm capable to extract correct gate potentials !
11/14/2002 SFR Workshop - Sensors
42
2003 Goals• Study sensitivity: what is optimal sensor layout (electrode
positioning, nominal conductivity, etc.) such that sensitivity in center of wafer is enhanced, by 9/30/2003.
• Demonstrate the feasibility of an EIT based sensor to measure plasma induced potential at wafer surface, by 9/30/2003.