the first and only in-situ implanter cleaning process. · 2019. 6. 6. · >15% reduction in...

Copyright 2008 ATMI, Inc. All Rights Reserved.

ATMI AutoClean®

Predictable Process. Extended Source Life.

The First and Only In-SituImplanter Cleaning Process.

2Copyright 2008 ATMI, Inc. All Rights Reserved.

Agenda

Issues Impacting Implant Efficiency

ATMI AutoClean Demonstrated Benefits

System Features


Problems Caused by Dirty Implanters

Short and unpredictable source life

Beam instability due to glitching or arcing

Extended set-up time from beam instability

Undesired personnel exposure to toxic waste

Lost productivity due to decreased tool availability


Problems Caused by Dirty Implanters







Short and Unpredictable Source Life

1/25/2007 2/5/2007 2/22/2007 3/6/2007 3/9/2007 3/31/2007 4/10/2007 4/16/2007 4/29/2007 5/10/20070

100

200

300

400

500

600

Sou

rce

Life

(h

rs.)

Short

Target

Data Courtesy of Atmel

312 317296

343

230

134

313

249

277

74

Unpredictable

150mm Applied MaterialsXR80 with IHC source


Problems Caused By Dirty Implanters







Beam Instability due to Glitching or Arcing

Data source: Implanter Source Life and Stability Improvement using In-Situ Chemical CleaningTerry Romig, Michael Mitchell, Texas Instruments Inc

David Eldridge, Jim Mayer, ATMI

http://ismi.sematech.org/ismisymposium/abstracts/08051924424.pdf

Glitch occurrence increases as the source ages

0

100

200

300

400

500

600

700

4/25

4/26

4/27

4/28

4/29

4/30 5/1

5/2

5/3

5/4

5/5

5/6

5/7

5/8

5/9

5/10

5/11

5/12

5/13

5/14

5/15

5/16

5/17

5/18

5/19

5/20

5/21

5/22

5/23

5/24

5/25

5/26

5/27

5/28

5/29

5/30

5/31

6/1

6/2

6/3

6/4

6/5

6/6

6/7

6/8

6/9

6/10

6/11

6/12

6/13

6/14

6/15

6/16

6/17

6/18

6/19

6/20

6/21

6/22

6/23

6/24

6/25

Glitch CountSource Hours


Extended Recipe Set-Up Time due to BeamInstability

Beam non-uniformity requiring multiple recipe set-ups &potential / frequent manual intervention

300mm Varian VIISta HC with IHC source


Undesired Personnel Exposure to Toxic Waste

Operator exposed 2-3 times per month, per implanter


ATMI AutoClean®

Complete System Solution to Implanter In-Situ Cleaning

CleaningReagent

ProcessRecipes

& BKM

SolidSourceDelivery

CustomerSpecific

ApplicationDevelopment

PredictableProcess.

ExtendedSource Life.


ATMI AutoClean DemonstratedProcess Benefits

>40% increase in source life

>15% reduction in recipe set-up time

>80% reduction in glitching

>40% reduction in PM frequency

Dramatic reduction in toxic waste and cleaning time

7% improvement in implant productivity

Up to $100K annual cost savings per implanterImplanter &/or process dependent


>40% Increase in Source LifeField DataMultiple Tool Types, Source Types, Device Types, Wafer Sizes

Sou

rce

1 IH

C

Sou

rce

2 IH

C

Sou

rce

3 IH

C

Sou

rce

4 IH

C

Sou

rce

5 IH

C

Sou

rce

6 IH

C

Sou

rce

7 B

ern

as

Sou

rce

8 B

ern

as

Sou

rce

9 IH

C

Sou

rce

10

IHC

Sou

rce

11

IHC

Sou

rce

12

IHC

Sou

rce

13

IHC

Sou

rce

14

Sou

rce

15

Sou

rce

16

Sou

rce

17

Sou

rce

18

Sou

rce

19

Sou

rce

20

Sou

rce

21

Sou

rce

22

Sou

rce

23

IHC

Sou

rce

24

IHC

Sou

rce

25

IHC

Sou

rce

26

IHC

Sou

rce

27

IHC

Sou

rce

28

Bern

as

Sou

rce

29

Bern

as

Sou

rce

30

Bern

as

Sou

rce

31

Bern

as

Sou

rce

Ho

urs

AutoCleanBaseline

240

240

240

240

240

240

190

190 200

165

165

165

270

120

120

120

120

120

120

120

120

120

162

162

162

162

162

230

134

313

249

336

336

336

336

336

440

248 26

0

440

250

310 32

3

636

281

225

174 19

8

179

336

221

168

336

319

404 42

3

410

445

659

593

683

589


336 336 336

240

336 336

456

336 336 336

0

100

200

300

400

500

Source1

Source2

Source3

Source4*

Source5

Source6

Source7

Source8

Source9

Source10

Sour

ce

Ho

urs

AutoClean Source Life 3-week PM experiment

2-week calendar PMtarget

*Source #4experienced filamentpower supply failure,not source related

Opportunity to Reduce PMs By 50%

Baseline Source Life

Major Logic Manufacturing Customer 300mm mixed logic productionVarian VIISta HC with IHC source


Without AutoCleanWith AutoClean

Sour

ce

Life

(hr

s.)

0

100

200

300

400

500

600

700

800

1/25/2007 2/5/2007 2/22/2007 3/6/2007 3/9/2007 3/31/20074/10/20074/16/20074/29/20075/10/20076/15/20077/14/20078/15/2007

312 317296

343

230

134

313

249

74

277

Target

659

518

683

>2X Increase in Source Life

Data Courtesy of Atmel 150mm Applied MaterialsXR80 with IHC source

~100%


>30% Improvement in Beam Uniformity

0.00%

0.20%

0.40%

0.60%

0.80%

1.00%

1.20%

1.40%

Average beam uniformitywithout AutoClean

Average beamuniformity with

AutoClean

Un

iform

ity

1.234%

0.841%

32% improvement in beam uniformity on the wafer

Beam uniformity for all recipes

>30%reduction

Major Logic Manufacturing Customer300mm mixed logic production Varian

VIISta HC with IHC source


>15% Reduction in Recipe Set-Up Time



0.350.3

100.3150.3200.3250.3300.3350.3400.3450.3500.3

AS15K

AS2K9

AS35K

AS50K

B2K2

B3K32

B4K7

B6K

BF17

KF1

0KF2

0K P3K P8K

Tim

e to

se

t up

Re

cip

e (

sec

s)

Standard Std. AvgAutoClean AutoClean avg

17%

Set-Up Time for Multiple Recipes


>80% Reduction in Glitching

Week 1 Week 2 Week 3 Week 4 Week 5 Week 6 Week 7 Week 8

605

992

2933

981

128

679

286

834

467

29 94 142 86 13135

184

AutoClean

Baseline

23% 97% 97% 86% 32% 81% 88% 78%

Glit

ch

Co

un

t




>80% Reduction in Glitching7720

1485

No AutoClean AutoClean

81%Reduction

Glit

ch

Co

un

t




Glitch Count

Source Life (hrs)

0

100

200

300

400

500

600

700

4/25

4/26

4/27

4/28

4/29

4/30 5/1

5/2

5/3

5/4

5/5

5/6

5/7

5/8

5/9

5/10

5/11

5/12

5/13

5/14

5/15

5/16

5/17

5/18

5/19

5/20

5/21

5/22

5/23

5/24

5/25

5/26

5/27

5/28

5/29

5/30

5/31 6/1

6/2

6/3

6/4

6/5

6/6

6/7

6/8

6/9

6/10

6/11

6/12

6/13

6/14

6/15

6/16

6/17

6/18

6/19

6/20

6/21

6/22

6/23

6/24

6/25

With AutoClean GlitchingDecreases As Source Ages

0

100

200

300

400

500

4/25

4/26

4/27

4/28

4/29

4/30 5/1

5/2

5/3

5/4

5/5

5/6

5/7

5/8

5/9

5/10

5/11

5/12

5/13

5/14

5/15

5/16

5/17

5/18

5/19

5/20

5/21

5/22

5/23

5/24

5/25

5/26

5/27

5/28

5/29

5/30

5/31 6/1

6/2

6/3

6/4

6/5

6/6

6/7

6/8

6/9

6/10

6/11

6/12

6/13

6/14

6/15

6/16

6/17

6/18

6/19

6/20

6/21

6/22

6/23

6/24

6/25

600

700 Glitch Count

Source Life (hrs)No AutoClean

AutoClean

6 SourceChanges

4 SourceChanges

Data source: Implanter Source Life and Stability Improvement using In-Situ Chemical CleaningTerry Romig, Michael Mitchell, Texas Instruments Inc, David Eldridge, Jim Mayer, ATMI

http://ismi.sematech.org/ismisymposium/abstracts/08051924424.pdf


>40% Reduction in PM FrequencyMaintenance frequency is determined by source life

Without AutoClean

With AutoClean

Unscheduledmaintenance nowbecomes scheduledmaintenance.

More predictablefab operations.

Less personnelexposure.

Increased implanteravailability.

Data Courtesy of Atmel 150mm Applied Materials XR80 with IHC source

S M T W Th F Sa

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

FebruaryS M T W Th F Sa

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

MarchS M T W Th F Sa

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

April

S M T W Th F Sa

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

JuneS M T W Th F Sa

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

JulyS M T W Th F Sa

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

August


Predictable Process. Extended Source Life.Source Stays Cleaner… Longer

Photos courtesy of AtmelMid-current source chamber at source PM

In many cases only requiring a DI-H20 wipe

270 hours without AutoClean 636 hours with AutoClean


LongerSource

Life

GreaterImplant

Productivity

FewerPMs

FasterRecipe

Changes

ReducedGlitching

FewerManualAssists

What’s a 13th month of production worth?

7% Improvement in Implant Productivity

“Throughput modeling estimates approximately2,000 wafers per tool in increased wafer output.”

Presentation by Major 300mm Logic Customer at ISMI, Oct 24, 2007








Up to $100K annual cost savings per implanter

Implanter &/or process dependent


$0

$2,000

$4,000

$6,000

$8,000

$10,000

$12,000

$14,000

High Case Benefit Low Case Benefit Cost

$61,728 - $94,776 Savings Annually per Implanter

$5144monthly

$7,898monthly

Low Case Improvements Target:+40% life time source, -15% source PM time, -15%species change time, +40% glitch reduction.Implanter Assumptions:20 min of cleaning/day each implanter, BaseSource life=240h, uptime 80%, Ion Source cost=$1,500, Source PM: 4h

High Case Improvements Target:+60% life time source, -20% source PM time, -20%species change time, +60% glitch reduction.Implanter Assumptions:20 min of cleaning/day each implanter, BaseSource life=240h, uptime 80%, Ion Source cost=$1,500, Source PM: 4h

ReducedGlitching

$6,083

Quicker SpeciesChange

$162

Fewer PM’s

$2,295

PM Labor

$911

SourceConsumables

$2,869

ReducedGlitching

$4,056

Quicker SpeciesChange

$122

Fewer PM’s

$2,034

PM Labor

$812

SourceConsumables

$2,543

AutoClean Cost

$2,800

AutoCleanProcess

$1,622


ATMI AutoClean

CleaningReagent

ProcessRecipes

& BKM

SolidSourceDelivery

CustomerSpecific


Complete system solution to implanter in-situ cleaning


AutoClean Process using XeF2Engineered cleaning reagent first introduced by ATMI

Increased effectivenessStrong fluorinating agent volatilefluorides of As, P, B, Si, WNo need for plasma activation (NF3requires plasma activation)

Ease of UseCan use a simple, pressure-based,delivery system (SDS® gas stick)

Enhanced safetyNo need for dangerous highpressure gasAlways sub-atmospheric forenhanced safetyHigh density fluorine source; likelythe safest source of fluorine

Physical PropertiesWhite/Colorless Crystalline PowderMW=169.29, Density=4.32kg/l, Melting Point_135° C, Vapor Pressure 0.5kPa (3.8 torr) @25° C

Xe+ Beam

To Pump

SourceVacuumChamber

IonSource

XeFF

XeFF

Reactant flows just like dopant


XeF2 Solid Delivery SystemTailored delivery system developed by ATMI

Vapor is delivered at room temperature. No heating means no

condensation inside the implanter.

Ensures that vapor delivery from the solid material is consistent and reliable (no lumps!)

XeF2 Flow Comparison

0

10

20

30

40

50

60

0% 20% 40% 60% 80% 100%% Used

Flo

w

(sc

cm

)

AutoClean Cylinder

Conventional Cylinder

•75g fill weight

•Smaller than SDS ® cylinder

•Installs on any spare SDS slot

•_” MVCR connector


Vapor is delivered at room temperature.No heating means no condensationinside the implanter.

Ensures that vapor delivery from the solidmaterial is consistent and reliable (nolumps!)

XeF2 powder is distributed within thematrix of a porous aluminum material,ensuring good thermal contact andmaximum solid surface area forvaporization.

Al Foam

HighConductanceValve

•75g fill weight

•Smaller than SDS ® cylinder

•Installs on any spare SDS slot

•_” MVCR connector

XeF2 Solid Delivery SystemTailored delivery system developed by ATMI


AutoClean

CleaningReagent

ProcessRecipes

& BKM

SolidSourceDelivery

CustomerSpecific




Process BKM’s and RecipesImplementation Strategy

Multiple approaches to address variety of customer issues

Running multiplespecies on same

implanter requiringlengthy set-up or

prep.

Implement AutoCleanto replace Argonrecipe between

dopant changes.

Short source life, highglitch-rate, or highconsumable costs.

Implement AutoCleanrecipe to

automatically run for10-15 minutes every 8

hours.

Unstable sourceperformance due toprocess by-productaccumulation or…?

Implement AutoClean“recovery” recipe torecover from source

instability.


Process BKM’s and Recipes(Customer Examples)

Implement AutoClean recipe to automatically run for10-15 minutes every 12 hours.

Implement AutoClean “recovery” recipe to recoverfrom source instability.

Replace Argon recipe with AutoClean.


AutoCleanTM

CleaningReagent

ProcessRecipes

& BKM

SolidSourceDelivery

CustomerSpecific




Customer Specific Considerations

Situational Considerations

• Tool types in installed base

• Implant process types

• Manufacturing protocols

• Production Mix

Situational Response

• Flow of XeF2

• Duration of clean process

• Plasma intensity & source parameters

• Frequency of Clean


Customer Specific ApplicationsMajor 300mm Logic CustomerNeeds

Improved PM cycles on IHC source (less frequent)Improved Beam instability recoveryHigh consumable costs – throwing away costly 300mm parts because there is no way toclean up

Situation300mm mixed logic productionVarian VIISta HC with IHC source2 week PM schedule interrupted by early source failure (240 hours average)

SolutionAutoClean 15 mins every 12 hours

Results / Proof>336 hour source life on 6 sources40% improvementPM schedule changed from 2 weeks to 3 weeksBeam instability after 100 hours recovered and source run to 2 weeksCustomer noticed cleaner manipulator and bushingGlitching (beam interruption) reduced 80%


Customer Specific ApplicationsMajor 300mm Memory Customer

NeedsImproved PM cycles on IHC & Bernas sourcesImproved Beam instability recovery

Situation300mm memory productionVarian VIISta HC with IHC and Bernas sourcesAverage source life 240 hours

SolutionAutoclean10-15 minutes every 8-12 hours

Results / Proof1st source (IHC) ran 450 hrs (average life without AutoClean 200 hrs)AutoClean then moved to VIISta with Bernas source

1st source (Bernas) was glitching after 200 hrs.Source life extended to 248 hrs with 30 mins AutoClean conditioning recipe.

2nd source (Bernas) ran to 260 hours, well above average life (190 hrs).


Customer Specific ApplicationsAtmel

NeedsImproved PM cycles on IHC source (less frequent)Improved Beam instability recovery

Situation150mm simulated productionApplied XR80 HC with IHC sourceAverage source life was 165 hours

SolutionAutoClean 15 minutes every 8 hours

Results / ProofSource lifetimes with AutoClean were 250, 310 and 323 hours78% average improvement


AutoClean

CleaningReagent

ProcessRecipes

& BKM

SolidSourceDelivery

CustomerSpecific



PredictableProcess.

Extended SourceLife.

Copyright 2008 ATMI, Inc. All Rights Reserved.

AutoClean Process Development

1) Dynamic clean – ion source focused (currently released for customers)

2) Static clean – scheduled with PM; focus on source housing, extraction assembly, etc. (currently in field alpha testing, not released)

3) Ex-situ – stand alone process chamber for cleaning parts, validating capabilities and materials compatibilities

(currently in R&D use, exploring customer interest; not released)


Before AutoClean After AutoClean

Predictable Process. Extended Source Life.Cleaning of Bernas Source Alumina Insulators with XeF2


Predictable Process. Extended Source Life.Cleaning Bernas Source Repeller Plate With AutoClean

Used, before AutoClean After AutoCleanNew / Unused


AMAT Extraction Manipulator Bellows

Pre-Clean XeF2 Clean1 Cycle

XeF2 Clean4 Cycles


Viton O-ring

Before After

VespelSample

Teflon Sample

Before After

XeF2 Material Compatibility StudiesViton, Vespel, Teflon

Conclusions:Viton, Vespel and Teflon are inert withXeF2 – no weight loss, visual surfacedegradation or RGA / IR reactionproducts were observed.


Customer FeedbackAtmel

“Our initial testing demonstrated the outstanding cleaning properties of ATMI’s AutoCleansystem.”

“Regular periodic use of AutoClean for in-situ cleaning of the ion source resulted in a 40% -50% increase in source life.”

“It also improved fab predictability and productivity by eliminating ion source failures dueto beam instability and high suppression leakage current.”

“In addition, based on our initial data, our regular preventative maintenance activities willbe easier and require less tool downtime as a result of using AutoClean.”

Jim DunnImplant Equipment Engineering Section ManagerAtmelColorado Springs, CO


Customer FeedbackMajor 300mm Logic Customer

“AutoClean operation resulted in consistent source performance with zero unscheduleddowntime caused by early source failure”

In side-by-side testing, the AutoClean system experienced 25-30% fewer source changesthan the standard system.”

“Data shows clearly that the AutoClean process significantly reduced the glitch rate as thesource ages, producing glitch rates at the end of source life that are comparable to newsource performance.”

“Enhanced Safety with reduction in personnel exposure to hazardous materials & processbyproducts = parts were cleaner and required less scrubbing during PMs.”

“Throughput modeling estimates approximately 2000 wafers per tool in increased waferoutput.”


Implant Platform Experience

AMAT, XR80, Quantums X, 1-3

Axcelis, GSD E2, Ultra, Optima MDHi Energy tools – moving towards evaluations

Japan - SEN, Nissin – initiating evaluations

VSEA, VIISta’s - 80, HCP, HCS, 810, 900, etcVSEA, E500/220, VIISion 80/200

Note: ion source types, both IHC and Bernas.


AutoClean®

30% increase in Beam uniformity

80% reduction in glitching

40% reduction in PM frequency

Increased implanter availability

More predictable fab operations

Less personnel exposure

Unscheduled maintenance nowbecomes scheduledmaintenance

Without AutoClean

With AutoClean

Data Courtesy of Atmel 150mm Applied Materials XR80 with IHC source

S M T W Th F Sa

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

FebruaryS M T W Th F Sa

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

MarchS M T W Th F Sa

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

April

S M T W Th F Sa

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

JuneS M T W Th F Sa

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

JulyS M T W Th F Sa

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

August

With AutoCleanWithout AutoCleanPredictable Implant Process.Extended Source Life.

Winner of 2008 R&D 100 Award; a distinction known to industry, government, and academia as proof that the goods or service is among 100 of the most innovative ideas of the year, focusing on products and processes that can change people's lives for the better, improve the standard of living for large numbers of people, save lives, promotegood health, and/or clean up the environment.

Copyright 2008 ATMI, Inc. All Rights Reserved.Copyright 2008 ATMI, Inc. All Rights Reserved.Copyright 2008 ATMI, Inc. All Rights Reserved.

ATMI, the ATMI logo, AutoClean, The Source of Process Efficiency, are trademarks or registered trademarks of Advanced Technology Materials, Inc.in the United States, other countries or both.

Other company, product, or service names may be trademarks or service marks of others.

the first and only in-situ implanter cleaning process. · 2019. 6. 6. · >15% reduction in...

Documents