Plasma monitoring under industrial conditions for semiconductor technologies · 2002. 8. 13. · August 13, 2002 Page 1 Plasma monitoring under industrial conditions for semiconductor

August 13, 2002Page 1

Plasma monitoring under industrialconditions for semiconductor

technologies

Michael Klick

ASI Advanced Semiconductor Instruments GmbH,Rudower Chaussee 30

D 12489 Berlin, [email protected]


Motivation

power dissipation for chemistry byelectrons (collision rate)energy and angle distribution of ionsdetermines etch profile

depending on both pressure p and gas (neutrals) temperature Tn.

nn =p

kTn

The important process parameter is the density of the neutrals nn

nn: density of the neutrals ⇒ crucial process parameterp: pressure ⇒ adjustable tool parameterTn: gas (neutrals) temperature ⇒ hardware parameter

(chamber... electrode temperature)k: Boltzmann constant

In a non-thermal, low pressure plasma – the neutral’s density is the core parameter!

Why should we measure plasma parameters?


Contents

! Introduction• Plasma monitoring for low pressure application, mainly

plasma etching• The IC manufactures application conditions and

requirements! Comparison of applicable monitoring methods

• OES• VI Probes• SEERS

! Application examples and conclusions• Plasma physical effects in production chambers• Fault detection and conditioning• Plasma and product parameter


The up-coming challenges

Increasedwafer size

CD’s below150 nm

Newproducts

Increased demand of process stability(smaller process window, process mix ...)

Fab-wide APC platform

Smart controlSmart sensors

Introduction


Process knowledge and economic benefit

Plasmaparameter

Processstability

Manufacturingissues

Radical densities…Electron densityElectron collisionrate…Potentials

Etch rateHomogeneityProfile…ArcingConditioning…

YieldMaintenance…Throughput…Up-time

Introduction


Plasma parameter and process state

Fastconditioning

after PM and dryclean, earlydetection ofproduct mix

issues

Earlychamber

faultdetection

as corrosionand arcing

Fastchambermatching

and processtransfer

Pre-process

faultsdetection ashard mask

issues

CriticalDimensions,

Yield

Test & Conditioning wafer usage

Up-time,Maintenancespare parts &

manpower

Introduction


Plasma parameter and process state

Tool Plasma Bulk Wafer Surface

MFC gas flow,Chamber lidtemperature

Pressure,Optical emission

endpoint

OpticalInterferometric

Endpoint

How tomeasure ?

Difficult tomeasure

Easy tomeasure

RF power inputinto plasma

Chambersurface

temperature

Wafer surfacetemperature

Ion currentdensity & energy,chem. reactions

Parameters ofelectron, ionsand neutrals

Electronexcitation,

chem. reactions

In-situ measurement techniques are needed

Many important process parameters and even tool parameters cannot be measured directly,or even not at all.Source: A. Steinbach (Infineon), 2nd AEC/APC Europe, Dresden, Germany, 2002.

Introduction


Which parameters should be measured

! RF power input into plasma?• VI probe delivers RF power into chamber including feed-

through, cooling system and e-chuck.! Electron parameters are the key parameters of the bulk

plasma (ionization, dissociation, fragmentation, excitation...).• Optical Emission Spectroscopy (OES) reflects electronic

excitation for different species.• Self Excited Electron Resonance Spectroscopy (SEERS)

provides global electron parameters directly.


Contents

! Introduction• Plasma monitoring for low pressure application, mainly plasma

etching• The IC manufactures application conditions and requirements

! Comparison of applicable monitoring methods• OES• VI Probe• SEERS



Applicable methods under industrial conditions

! RF voltage and current as well as phase angle at matchbox! Main application: RF Maintenance! Disadvantage

• Strictly speaking no plasma parameter, RF current 90%’chamber current’ and 10% ‘plasma current’. Measures mainlythe chamber’s impedance

• ‘New game’ in case of new chamber hardware (see above)• Provides only the fundamental or some harmonics

! Advantage• Easy to understand• Supports chamber development and analysis

VI probe

Comparison



! Relative intensities of excited species! Main and pure application: Endpoint detection! Disadvantage

• Interpretation needs high level of knowledge• Requires data compression due to large amount of data via

Principal Component Analysis (PCA) or ...! Advantage

• Easy access via sight window• Different species can be identified• Supports process development and analysis

Optical emission spectroscopy (OES)

Comparison



! Based on a broad-band RF current measurement! Approximately volume averaged density and collision rate of

electrons and the electronic bulk power! Main application: Process monitoring! Disadvantage

• Needs access to chamber via passive sensor on groundpotential

• Not available for all chamber types! Advantage

• Provides real plasma parameters (inside measurement)• Easy handling for process monitoring• Supports process and chamber development

Self Excited Electron Resonance Spectroscopy (SEERS)

Comparison


Robustness under industrial conditions

! Classical, optical approach• Optical sight window access• New intensity level after PM• Damping by polymers/by products• No absolute values and chamber matching

! SEERS and VI probe• Based on passive RF current sensor flat in chamber wall (SEERS) or at matchbox output• No impact of chamber wall polymer• Absolute values• Chamber matching

13 MHz13 MHz Optical emission(OES)

RFRFcurrent

Polymer

Polymer

Polymer

Polymer

Cha

mbe

r wal

lC

ham

ber w

all

Does an insulating layer disturb ?

Comparison


SEERS Theory

! SEERS Fundamentals:• Non-linearity between voltage and displacement current in space

charge sheath• The resonance capability of the plasma (damped series resonance

circuit)• Geometric resonance frequency depending on electrode gap• Treatment of full Fourier spectrum in model with free parameters• Parameter estimation provides electron density and collision rate as

core parameters! SEERS Assumptions

• Electron plasma frequency >> RF frequency >> ion plasma frequency• Plasma sheath assumed as one dimensional• Plasma bulk as cylindrical

SEERS - model-based measurement of plasma parameters

Comparison


SEERS Theory

! The current pitch ratio is determined dynamically.

Experimental setup of SEERS

Peak/dc biasvoltage

RFcurrent

Coaxialsensor and

cable

Dielectricwindow

Top power

Bottom power

Chamber

Example: TCP®

Fast ADC500 MHz, 2 GS/s

50 Ohms input

SEERSalgorithm

Process databank (ne, νννν, ...)

Comparison


SEERS Theory

SEERS equivalent circuit

matchbox andgenerator

feed-through andstray capacitance

~ s0 linear „part“ of RF sheath

~ s1 linear „part“ of wall sheath

~ u0 [ i dt] nonlinearity sheath RF electrode

urf~ ω l me inertia (imaginary) part thereof

~ u1 [ - i dt] nonlinearity sheath plasma-wall

~ n l me ohmic part of plasma bulk (ohmic and stochastic heating)

" The plasma can be regarded as a damped resonance circuit.

Comparison


SEERS Theory

! Shape of the FOURIER-transform provides the information• amplitude is not important!

! Needs non-iterative and robust algorithm due to nonlinearity, fluctuation ormodel violations

One signal - one parameter set

An easy explanation of SEERS

⇒ n⇒ν⇒ν⇒ν⇒ν

secondary resonance main peak

spectrum fit

Fourier transformTime domain data

without resonance(nonlinear sheath only)

- 1

0

1M

easu

red

sens

or cu

rren

t I P

[mA

]

p = 2 P a , U B = 5 2 0 V ( P = 7 5 W )p = 2 P a , U B = 5 2 0 V ( P = 7 5 W )

0 0 .5 1 1 .5 2N o r m a l iz e d t im e ττττ

- 2

- 1

0

1

2

p = 1 0 P a , U B = 5 7 0 V ( P = 1 2 5 W )p = 1 0 P a , U B = 5 7 0 V ( P = 1 2 5 W )

DFT

Comparison


SEERS Theory

! SEERS determines reciprocally volume averaged:• electron density:

• electron collision rate:

• (electronic) bulk power:

Reciprocally and spatially averaged plasma parameters

V≈ n-1 dV⌠⌡V

∼n1( )-1

∼PB

ν

n∼

≈ nν dVν

⌠

⌡Vn∼∼

V

[ I(k) ] 2∝ ∑

Due to 1/n averaging, ranges of lower density get a higher weight!

Comparison


SEERS sensors

Standard sensor types for etch tools and unaxis

DN16CF

unaxis (PlasmaTherm)

DN25KF

DN40KFcustomer modified

Comparison


SEERS sensors

Sensors for 300 mm etch tools

Source: V. Tegeder, AEC/APC-Symposium XII, Lake Tahoe, USA, 2000.

APPLIED MATERIALS®eMxP+™# peak voltage# rotating B-field# optical access for OES

LAM® 2300™# peak voltage# inductively coupled# PnP data interface

APPLIED MATERIALS®DPS™# peak voltage# inductively coupled# capacitive ≠ inductive coupled frequency

by courtesy of

Comparison


Contents

! Introduction• Plasma monitoring for low pressure application, mainly plasma

etching• The IC manufactures application conditions and requirements

! Comparison of applicable monitoring methods• OES• VI Probes• SEERS



! TCP® power effectsthe electron density,thus the bias voltageand therefore theplasma impedanceand the powerdissipation of thebottom power(capacitive).

Decreasing dc bias for increasing TCP® power

Source: S. Wurm, Lam Research Corp., 10th Ann. European Techn. Symp., Geneva, Switzerland, 1998. by courtesy of

LAM® TCP® 9600 (Al etch, Cl chemistry)

Application


Source: S. Wurm, Lam Research Corp., 10th Ann. European Techn. Symp., Geneva, Switzerland, 1998. by courtesy of

Bulk power and electron density for increasing TCP® power

LAM® TCP® 9600 (Al etch)

! TCP® power effects thedensity and collision rate ofelectrons and therefore theplasma impedance and thepower dissipation of thebottom power (capacitive).

! Mainly dependent oncollision rate, the bulkpower (top) decreases forincreasing TCP® power(>259 W). This is thereason for the plateau inthe electron density.

Stochastic heating withlarger sheath thickness

Application


SEERS Theory

Physical background of the electron collision rate

Stochastic heating,dominating for low pressure,< 10 Pa = 75 mTorr

Momentum transfer,dependent on gas mixture, pressure,and gas temperature directly

relative concentration of species i (partial pressure ratio)

Mean thermal velocity of electronsplasma length

cross section xthermal velocity

~ const.

pressure /gas temperature= density of neutrals

dc biasBohm criterion

No magnetic field (B = 0) and capacitive coupling

1.6νeff ≈l 2πme

σ( )νe Ubias p

kTnΣΣΣΣi

pip+

1νe

Application


! Pressure variation: collision rateshows nonlinear behavior

! Distinction of domination by ohmicheating / stochastic heatingpossible (= different modi ofpower conversion into plasma)

! Saturation / maximum at evenhigher pressures estimated

! Potential process instability,basic understanding isneeded for processdevelopment!

LAM® TCP® 9400 (poly-Si etch)

Collision rate depending on pressure

mean electron collision rate vs. pressure, chamber 1

2,0E+07

4,0E+07

6,0E+07

8,0E+07

1,0E+08

1,2E+08

1,4E+08

0 10 20 30 40

pressure [mTorr]

elec

tron

collis

ion

rate

[1/s

]

Source: C. Steuer (TU Dresden), 2nd Workshop of SEERS, Dresden, Germany, 2000. by courtesy of

Application


LAM® TCP® 9400 (poly-Si etch)

! Plasma parameter:electron density / collisionrate (ED/CR)

! Correlation between in-situand in-line measurementsmean etch rate, (blue) andquotient electron density /electron collision rate(purple) for pressurevariation.

Etch rate and plasma parameters depending on pressure

ER & f(CR, ED) vs. pressurechamber 2

160

170

180

190

200

0 10 20 30 40pressure [mTorr]

etch

rate

[nm

/min

]

10,00

20,00

30,00

40,00

50,00

60,00

f(CR

, ED

) [ar

b.u.

]

Mean_Etch Rate ED/CR


Application


reference chamber striking chamber

Electron density depending on top power and bottom power

Chamber Comparison, LAM® TCP® 9400/SE

! bad selectivity between poly-Si and gate oxide in striking chamber $ lower electron density for the same TCP® power

! RF power dissipation in plasma!! easy tool fault detection $ TCP® Matchbox

Mean Electron Density vs. VariationTop and Bottom Power, Chamber 1, t = 86 rfh

Mean Electron Density vs. VariationTop and Bottom Power, Chamber 2, t = 73 rfh

GC etch


Application


reference chamber striking chamber

Source: C. Steuer (TU Dresden), 2nd Workshop of SEERS, Dresden, Germany, 2000.

Collision rate depending on top power and bottom power

Chamber Comparison, LAM® TCP® 9400/SE

Quantitative difference in collision rate,local minimum along increasing bottom power$ separating ohmic + stochastic heating

Mean Electron Collision Rate vs. VariationTop and Bottom Power, Chamber 1, t = 86 rfh

Mean Electron Collision Rate vs. VariationTop and Bottom Power, Chamber 2, t = 73 rfh

by courtesy of

Application


Chamber lid temperature and electron collision rate

! First wafer effect(gas adsorption anddesorption atchamber wall)

! Gas compositiondrift in plasma bulk(„saw tooth“),heating of chamberkit and wafersurface cause driftof chemicalreactions there.

Electron collision rate and chamber lid temperature vs. process time

0

5

10

15

20

0 1000 2000 3000 4000 5000 6000

process time [s]

elec

tron

ollis

ion

rate

[10

7 /s]

lid te

mpe

ratu

re

[0C]

1/Tn

Gas temperature drift only duringhigh RF power step

Si etch in APPLIED MATERIALS® HART chamber, oxide wafers

by courtesy of

Application


Chamber lid temperature and electron collision rate

! Verifies again thetemperaturedependence of theelectron collisionrate (1/Tn).

! Temperaturecontrol of lid heaterleads to stableconditions.

Si etch in APPLIED MATERIALS® HART chamber, product wafers

Electron collision rate vs. wafer [mean]

2,5E+07

3,0E+07

3,5E+07

4,0E+07

4,5E+07

5,0E+07

5,5E+07

0 5 10 15 20 25 30 35

wafer

Elec

tron

col

lisio

n ra

te [s

-1]

without lid heaterwith lid heater

Application


Chamber: MxP™, B-field parallel to wafer

Electron collision rate versus magnetic field

∑⋅⋅+=k

kg

k

NB

g

e

eBStoch σP

pTk

PmTk

πνν 8

ν eff( ),,B n ω Re .ω e

2

.j ω1 .

.j ω ν.j ω

ω pe( )n2

( ).j ω ν 2 ω c( )B2

1

ω .sl

12

ω pe ( )n

! Simple ansatz using the dielectric tensor leads to an one-dimensional approximation.

s: sheath thickness,l: plasma length, without stochastic heating for B > 0

Application


Chamber: MxP™, B-field parallel to wafer

Electron collision rate versus magnetic field

Electron Collision Rate vs. B - Field and PowerParameter: Pressure 300 mTorr

Theory

s/l = 1/4 s/l = 1/16

300mTorr

Electron Collision Rate vs. B - Field and PowerParameter: Pressure 100 mTorr

Experiment

100mTorr

by courtesy of

Application


Comparison of power dissipation insidethe chamber, while nominal rf powerkept constant.

Etch rate BPSG vs. bulk power

R2 = 0,5187620

630

640

650

660

670

11 12 13 14 15 16 17 18

bulk power [mW / cm²]

etch

rate

[nm

/ m

in]

rf match 1rf match 2rf match 3rf match 4

Contact etch at APPLIED MATERIALS®MxP+™

one point -mean of one wafer

Source: A. Steinbach, et. al. SEMATECH AEC/APC Symposium XI, Vail, USA, 1999.

APPLIED MATERIALS® MxP+™ (CT etch)

RF matchbox comparison by power density measurement

! RF match box comparisonby bulk powermeasurement.

! Power coupling into thechamber differs about 30%as indicated by bulk power.

! Impact of chamberconditions.

! Oxide etch rate saturationat high power dissipation(rf match 4) possiblycaused by transportprocesses or surfacereactions.

by courtesy of

Application


Source: A. Steinbach et al. (SIEMENS Dresden), AEC/APC X, Vail, USA, 1998. by courtesy of

Chamber: MxP+™, B-field parallel to wafer

Conditioning after wet clean

! Wet clean! Stable chamber

conditions after about10 wafers.

Application


Source: A. Steinbach, et. al. SEMATECH AEC/APC Symposium XI, Vail, USA, 1999. by courtesy of

Real time detection for contact etch etch at MxP+™

Plasma in helium feed-through

Process instability was caused byparasitic plasma inside He feed-through.

Application


Arcing

! Arcing is basically a breakthrough in an insulating layer atthe chamber wall, wherever potential grown-up.Reasons for arcing are:

– inhomogeneous polymer build-up at chamber wall– incorrect grounding of parts of chamber

! Arcing leads to a reaction in collision rate:– increase $ large polymer molecules– decrease $ relative small metal ions

! SEERS enables the detection the process relevant chemical‘drag-mark’ of arcing.

Application

Introduction


Process gas and polymer

e-

ARCING

length of stay - some secondslength of stay - some seconds

G

eE

wafer

chamberwall

electronic oscillation

Generation and behavior of particles

Arcing

! in case of arcing - increase of collision rate at least for someseconds due to additional polymer particles from chamberwall

Application


•0

•20

•40

•60

•80

•100

•120

•140

•160

•0 •5 •10 •15 •20 •25 •30 •35 •40

•Ave

r. C

oll.

Rat

e [1

0•6 •s•

-1 •]

•Day

•0

•20

•40

•60

•80

•100

•120

•140

•160

•180

•200

•0 •20 •40 •60 •80 •100 •120 •140

•Etch time [s]•C

ollis

ion

Rat

e [1

0•6 • s•

1 •]

Source: V. Tegeder, AEC/APC-Symposium XII, Lake Tahoe, USA, 2000. by courtesy of

Fault detection at eMxP+™ (300 mm, F chemistry)

Arcing

Arcing Traces in Chamber $ Exchange of E-Chuck and Ion Shield

Application


0,00E+00

5,00E+07

1,00E+08

1,50E+08

2,00E+08

2,50E+08

3,00E+08

3,50E+08

4,00E+08

4,50E+08

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44

process time t[s]

colli

sion

rate

ν ννν e[s

-1]

arcing

fingerprint

Arcing traces at gas distribution

by courtesy ofSource: V. Tegeder, AEC/APC-Symposium XII, Lake Tahoe, USA, 2000.

Arcing at gas distribution, APPLIED MATERIALS® eMxP+™

Arcing

Arcing detected at new tool

RecipeStep 125mtorr / 215W /30G/ 50 sccm O2Step 225mtorr / 215W /0G/ 50 sccm O2

Application


20 40 60 80 100 120 140 160

process time [s]

468

10

20

406080

100

200pe

ak v

olta

ge [V

]

106

5

107

5

108

5

109

colli

sion

rate

[1/s]

100

500

1000

5000

10000

optic

al e

miss

ion

(EP)

Source: S. Wurm, Lam Research Technical Symposium, Geneva, Switzerland, 1998.

Wafer Fault Analysis: no resist, faultless dashed at LAM® TCP®

The impact of polymers onto the plasma

LAM® TCP® 9600SEFault: no resist, faultless dashed (reference wafer)

! The break-throughis not influenced(here separatestep).

! In case of the mainetch the collisionrate decreases byone order ofmagnitude due tono polymer on thewafer.

! Chemistry:Cl2/BCl3

by courtesy of

Application


Source: A. Steinbach, et. al. SemiPAC´99, San Antonio, USA, 1999.

Electron collision rate vs. etch time

3

5

7

9

11

13

15

0 50 100 150

time [s]

Col

lisio

n ra

te [1

0 7 s

-1]

Chamber A, quartz ring Chamber A, Si ring Chamber B, Si ring

Comparison at chamber A:- Quartz ring $$$$ isolating- Si ring $$$$ rf conducting

increase of effective cathode area $$$$decrease of power density

Conditioningwith resistwafers

Parameter Ratio Si ring /Quartz ring

Nitride etch rate 0.66Wafer temp. 0.71Inverse ring temp. 0.78El. collision rate 0.69Electron density 0.55Sqrt. Bulk Power 0.59Inverse ratio of thecathode areas 0.59

by courtesy of

Evaluation of shadow rings at MxP™

Application

Increase of virtual electrode size by Si shadow ring


Trench etch, wafer history - impact from Litho on Etch

Deep trench etch (Si etch using Cl and F chemistry)

M e a n e le c tro n c o llis io n ra te v s . wa fe r

2

6

1 0

1 4

1 8

0 5 1 0 1 5 2 0 2 5 3 0 3 5 4 0 4 5

wa fe r

mea

n el

ectr

on c

ollis

ion

rate

[107

/s] f irs t lo t s e c o n d lo t

! Wafer to wafer signature at second lot caused by alternating maskquality, due to pre– processes (Litho, CVD, ...).

! Drift during processing of both lots is caused by tool impacts.

Good etch result

Bad etch result

one point - one wafer

by courtesy ofSource: S. Bernhard, A. Steinbach, 3th AEC/APC Europe, Dresden, Germany, 2002.

Application


Monitoring of product mix impact on process stability

! Product mix impact of two different Si etch processes onprocess stability was monitored by

• Plasma parameter measurement using Self ExcitedElectron Resonance Spectroscopy (SEERS)

• Multichannel Optical Emission Spectroscopy (MPCA) atwavelengths from 200nm to 480nm during main etch step

! Analysis of measurements• Plasma parameters are calculated by plasma monitoring

system Hercules® internally in real time, no furthercalculation necessary

• Analysis of optical emission spectra offline by MultiwayPrincipal Component Analysis

Source: S. Bernhard, A. Steinbach, 3th AEC/APC Europe, Dresden, Germany, 2002.


Application


Impact on process stability by electron collision rate

Mean e lectron co llis ion rate v s . wafer

20

25

30

35

40

45

0 100 200 300 400

wafer

elec

tron

col

lisio

n ra

te [1

07/ s

]

P roduc t 1 P roduc t 2

conditioning wafers

Product 1 Product 2




Application


Off-line analysis of product mix impact on process stability by MPCA

Product 2Wafer product 2 afterWafer product 1Conditioning wafersproduct 2

Product 1Etch rate measurement on blankettest wafers destabilizes chamberconditionsConditioning wafers product 1




Application


by courtesy of

Si

poly-SiWSix

Si3N4

gate-oxide

Si

poly-Si

TEOS-oxide

gate-oxide

DRAM

LOGIC

Cross section of GC Stack

Gate contact etch (GC)

! Si3N4- hard maskWSix- metal layer etched with fluorinechemistryPoly-Silicon-Layer etched withChloride/Bromide chemistry

– Main-Etch etches main part of Poly-Si– Over-Etch etches only a fraction of Poly-

Si

! TEOS- oxide hard maskNo metal layer $ no fluorine chemistryPoly-Silicon-Layer etched withchlorine/bromide chemistry

Source: T. Dittkrist, A. Steinbach (Infineon), AEC/APC Europe, Dresden, Germany, 2001.

Application


by courtesy of

Si

poly-SiTiSix

GC stack

gate-oxidedrain source

gate

under diffusion

poly length

GC Stack cross section GC Stack SEM

Geometry definitions of GC Stack

Gate contact etch (GC)

! GC Stack cross section of logic product! Problem: under diffusion length and poly length to small

Source: T. Dittkrist, A. Steinbach (Infineon), AEC/APC Europe, Dresden, Germany, 2001.

Application


Source: T. Dittkrist, A. Steinbach (Infineon), AEC/APC Europe, Dresden, Germany, 2001. by courtesy of

e l. co llis io n ra te vs . w a fer (m ed ian ), P o ly -S i (o ver e tch )

0

50

100

150

200

250

2100 2200 2300 2400 2500 2600

w afe rn u m b er

colli

sion

rate

[106

s-1

]

20.0022.0024.0026.0028.0030.0032.0034.0036.0038.0040.00

unde

rdiff

usio

n

one point = one w afer

LOGIC CLEAN DRAM UNDER DIFFUSION

only conditioning plasma-clean

Product Mix observed by electron collision rate

Gate contact (GC)

! First wafer effect for electron collision rate and under diffusion lengthfor LOGIC processes after plasma clean and conditioning no firstwafer effect if only conditioning without plasma clean

– Processes are influenced by plasma clean.– Etching of chamber wall with fluorine is possible reason for bad

etch results after plasma clean.

Application


by courtesy of

Electron collision rate vs. poly length

Gate contact (GC)

! Correlation inside specification limits $ significant response for rather thanat serious process problems can be expected.

! From unit processing point of view the correlation is week, but from processintegration view it´s fine.

elec

tron

collis

ion

rate

at G

C et

ch

gate conductor width (electrically measured)

LSL USL

First part of resultsalready presented lastyear

Dedicated chamber

Electron collision rate vs. Gate conductor width

one point– one wafer

about 40.000wafer

Source: A. Steinbach, 3nd AECAPC, Dresden, Germany, 2002.

Application


Conclusions

! Plasma parameter measurement in IC manufacturingrequires robust methods and basic understanding of theplasma.

! The usage of plasma parameters is just at the beginningand requires more experience and knowledge in

• Handling and extraction of core parameters,• Automatic model building to predict product parameters.

! The best method depends on• Final goal (fault detection, process development…),• Chamber type and process,• Knowledge available.

! Electron parameters are most sensitive (OES, SEERS).

Plasma monitoring under industrial conditions for semiconductor technologiesMotivationContentsIntroductionThe up-coming challengesProcess knowledge and economic benefitPlasma parameter and process stateIn-situ measurement techniques are needed

Which parameters should be measured

Comparison of applicable monitoring methodsApplicable methods under industrial conditionsVI probeOptical emission spectroscopy (OES)Self Excited Electron Resonance Spectroscopy (SEERS)

Robustness under industrial conditionsSEERS TheorySEERS - model-based measurement of plasma parametersExperimental setup of SEERSSEERS equivalent circuitAn easy explanation of SEERSReciprocally and spatially averaged plasma parameters

SEERS sensorsStandard sensor types for etch tools and unaxis Sensors for 300 mm etch tools

Application examples and conclusionsLAM® TCP® 9600 (Al etch, Cl chemistry)LAM® TCP® 9600 (Al etch)SEERS TheoryLAM® TCP® 9400 (poly-Si etch)Collision rate depending on pressureEtch rate and plasma parameters depending on pressure

Chamber Comparison, LAM® TCP® 9400/SEElectron density depending on top power and bottom powerCollision rate depending on top power and bottom power

Chamber lid temperature and electron collision rateSi etch in APPLIED MATERIALS® HART chamber, oxide wafersSi etch in APPLIED MATERIALS® HART chamber, product wafers

Electron collision rate versus magnetic fieldChamber: MxP™, B-field parallel to waferChamber: MxP™, B-field parallel to wafer

APPLIED MATERIALS® MxP+™ (CT etch)Conditioning after wet cleanPlasma in helium feed-throughArcingIntroductionGeneration and behavior of particlesFault detection at eMxP+™ (300 mm, F chemistry)Arcing at gas distribution, APPLIED MATERIALS® eMxP+™

The impact of polymers onto the plasmaEvaluation of shadow rings at MxP™Deep trench etch (Si etch using Cl and F chemistry)Trench etch, wafer history - impact from Litho on EtchMonitoring of product mix impact on process stabilityImpact on process stability by electron collision rateOff-line analysis of product mix impact on process stability by MPCA

Gate contact etch (GC)Cross section of GC StackGeometry definitions of GC StackProduct Mix observed by electron collision rateElectron collision rate vs. poly length

Conclusions

Documents

Plasma monitoring under industrial conditions for semiconductor technologies · 2002. 8. 13. · August 13, 2002 Page 1 Plasma monitoring under industrial conditions for semiconductor