In Situ Meas Ecs 213 May 2008

In-Situ Measurement of the Relative Thermal Contributions of Chemical Reactions and Ions During Plasma Etching

Plasma and CVD Processes 17

213th Meeting of the Electrochemical Society

M.R. Tesauro Qimonda Dresden GmbH & Co. OHG, Germany

G.A. Roche KLA-Tencor, SensArray Division

Qimonda · Tesauro, Roche · Plasma and CVD Processes 17, 213th Meeting of the Electrochemical Society · Page 2 Copyright © Qimonda AG 2008 · All rights reserved.

Overview

Background: Chemical + Physical RemovalBackground: Chemical + Physical Removal

Results and DiscussionResults and Discussion

ConclusionsConclusions

Experimental MethodExperimental Method







Chemical & Physical Removal

J.W. Coburn, H. Winters J. Appl. Phys 50(5) May 1979

• Synergistic Effects- Chemical � Species� Temperature

- Physical� Ion Current� Ion Energy

• Synergistic Effects- Chemical � Species� Temperature

- Physical� Ion Current� Ion Energy


Attributes of Substrate Temperature Measurement

• Heat in- Ion current (incl.

recombination)- Heat of condensation- Exothermic reactions- Process gas convection- Radiation

• Heat in- Ion current (incl.

recombination)- Heat of condensation- Exothermic reactions- Process gas convection- Radiation

• Heat Out- Convection to process gas - Loss through electrostatic

chuck- Radiation

• Heat Out- Convection to process gas - Loss through electrostatic

chuck- Radiation

Tsubstrate = Heat in - Heat OutTsubstrate = Heat in - Heat Out


Why we care about this topic

• Evidently temperature spatial correlation to CD var iation.

• Routine recipe variation does not change response.

• Many hardware components are potentially responsibl e

• Objective: Identify mechanism and cut troubleshooti ng time.

• Evidently temperature spatial correlation to CD var iation.

• Routine recipe variation does not change response.

• Many hardware components are potentially responsibl e

• Objective: Identify mechanism and cut troubleshooti ng time.

Sensor Wafer Temperature

Sensor Wafer TemperatureIn-line CDIn-line CD







Sensorwafer Description

VRF ~Vs·ns·ω2


Sensor Wafer Pre-Processing

Integral Sensor WaferCoated with 3um PFR IX420Hi-line Resist, 2mm EBR

Manual Solvent Removalof Resist on Wafer HalfLeft of Notch

Data Not UsedData Not Used

"Resist" Sensor Group

"Resist" Sensor Group

"No Resist" Sensor Group"No Resist"

Sensor Group


Tem

pera

ture

Time

Si Tavg, steady-state (sensors on bare Si half)

Resist Tavg, steady-state (sensors on resist coated half)

∆Tss avg = k1A( Φr – ΦSi)∆Tss avg = k1A( Φr – ΦSi)

Assumption: for resist etch highly selective to Si: (Φr – ΦSi) = Heat of Chemical Reaction

Assumption: for resist etch highly selective to Si: (Φr – ΦSi) = Heat of Chemical Reaction

Exothermic Chemical Etch Components: Measurement Theory


Experimental Run Summary

1. Coat temperature wafer – remove resist from left s ide

2. CCP Etch: 11 step N2 / H2 DOE

3. CCP Etch: 5 single-factor N2 / H2 test runs

4. CCP Etch: 5 step N2 / H2 screening test

5. CCP Etch: 5 step Ar / O2 screening test

6. CCP Etch: 2 step N2 / H2 & Ar / O2 high pressure test

7. ICP Etch: 2 step Cl2 test (5mT, 50mT)

8. ICP Etch: 2 step O2 test (5mT, 50mT) – 50mT to end point

Both sensorwafers were run through the following conditions


Sensorwafer data traces

N2H2,30mT

O2Ar,30mT

POR, 100W, 300W, FRC-11, FRC+5 POR, 100W, 300W, FRC-11, FRC+5

O2Ar

200W,200mT

N2H2

200W, 200mT

Temperature

POR100W

300W

FRC-11 FRC+5

Vrf

CCP reactor

P=200mT, 200Wb,

a)N2=200, H2=250,

b) Ar=100, O2=10







Exothermic Chemical Etch Components: Experimental Data

N2 / H2 Chemistry: CCP Etch Chamber

43.8

44.0

44.2

44.4

44.6

44.8

45.0

1 6 11 16 21 26 31 36 41 46 51 56 61 66

Time (s)

Tem

p (C

)

1. Exponential Data Fit

2. Extrapolate to Steady-State

3. Calculate ( Φr – ΦSi)

4. Estimate removal rate based on C-C bond strength

1. Exponential Data Fit

2. Extrapolate to Steady-State

3. Calculate ( Φr – ΦSi)

4. Estimate removal rate based on C-C bond strength


N2 / H2 Chemistry: CCP Etch Chamber

43.8

44.0

44.2

44.4

44.6

44.8

45.0

1 6 11 16 21 26 31 36 41 46 51 56 61 66

Time (s)

Tem

p (C

) Bare SiBare Si

ResistResist

• However, the Si half of sensorwafer was hotter!

• Assumption is that thick resist (~3 µm) acts as an insulator to ion bombardment

• However, the Si half of sensorwafer was hotter!

• Assumption is that thick resist (~3 µm) acts as an insulator to ion bombardment

Exothermic Chemical Etch Components: Experimental Data


Etch Rate / Vrf and Temperature Response to RF Power (CCP chamber)

CCP, Ar / O2

120013001400150016001700180019002000

50 150 250 350

13 MHz RF Power (W)

Vrf

(V)

70

80

90

100

110

120

Etc

h R

ate

(nm

/ m

in)

CCP Ar / O2

44.4

44.6

44.8

45

45.2

45.4

45.6

50 150 250 350

13 MHz RF Power (W)

Ste

p T

max

(C

)

Resist

Si

Both Etch rate and Vrf show strong response to RF B ias power.Temperature has a modest response to RF power. This data supports the notion that Etch Rate is dom inated by physical mechanism.

Both Etch rate and Vrf show strong response to RF B ias power.Temperature has a modest response to RF power. This data supports the notion that Etch Rate is dom inated by physical mechanism.


Sensorwafer Response: Resist ClearingMean temperature by half wafer zones …

Now that resist is thinned ( ≤ 550 nm), resist half runs hotter than bare Si.Temperatures of each half converge as resist clears .Now that resist is thinned ( ≤ 550 nm), resist half runs hotter than bare Si.Temperatures of each half converge as resist clears .

Time (s)

… ICP reactor, 50mT O2 Clearing Resist from sensorwaf er


ICP Etch Chamber signals for Resist etch end point O 2 Chemistry

Resist clearing


Etch-to-Clear Resist, 50mT O 2(All sensors, same scale throughout)

RF on 176sRF on 176s clearing

"Hot spot" disappears as resist clears"Hot spot" disappears as resist clears

RESSi


Case Study: Chemical / Physical Components and Isotropic Mask Trim Etch

Width CD Depends on

Isotropic Etch Component

Width CD Depends on

Isotropic Etch Component

Poor In-line CD Uniformity


+ + +

CD

Lower substrate temperature=reduced isotropic etch rate=wider CDLower substrate temperature=reduced isotropic etch rate=wider CD

NN


Case Study: Root Cause Identification



Temperature Vrf Temperature Vrf

Sensorwafer responseSensorwafer response

“Heat Out” problemnot “Heat In”

“Heat Out” problemnot “Heat In”


Correction Verification


Poor In-line CD Uniformity Temp 20C Cathode Temp 40C Cathode

range = 10 C range = 3.5 C

Temp 20C Cathode Temp 40C Cathode

range = 10 C range = 3.5 C

Sensorwafer responseSensorwafer response

CD non-uniformity problem localized to heating/cool ing apparatus of ESCCD non-uniformity problem localized to heating/cool ing apparatus of ESC







Conclusions / Summary

The Resist/Bare Sensorwafer Method Shows Promise

• Can detect small but consistent differences in temp erature

and Could benefit from:

• Thinner coating for improved sensitivity

• "Checkerboard" pattern to avoid chamber-related asy mmetries

• Examine isotropic removal vs. vertical removal

Use of Temperature & Volt Sensorwafers together is proving useful

• Quickly ascertain source of chamber asymmetry

The Resist/Bare Sensorwafer Method Shows Promise

• Can detect small but consistent differences in temp erature

and Could benefit from:

• Thinner coating for improved sensitivity

• "Checkerboard" pattern to avoid chamber-related asy mmetries

• Examine isotropic removal vs. vertical removal

Use of Temperature & Volt Sensorwafers together is proving useful

• Quickly ascertain source of chamber asymmetry


Acknowledgements

AcknowledgementsAcknowledgements

Professor Costas Spanos The University of California, Berkeley

Matthias VoigtLithography Track Expert

Qimonda Dresden GmbH & Co. OHG

Professor Costas Spanos The University of California, Berkeley

Matthias VoigtLithography Track Expert

Qimonda Dresden GmbH & Co. OHG

Thank you

The World's LeadingCreative Memory Company




50mT O2

Date post:	14-Jul-2015
Category:	Education
Upload:	gregaroche
View:	270 times
Download:	1 times

In Situ Meas Ecs 213 May 2008

Education