1

Michael A. Fury, Ph.D.Director of Market DevelopmentMFury@VantageTechCorp.com May 16, 2012

Slurry monitoring practices today

Introduction to Vantage SlurryScope™

Sampling statistics

Implications for fab operations

Summary

2

Full particle size distribution (PSD)◦ Offline lab tool (e.g. Horiba LA-950)◦ 0.01 µm to 3,000 µm ◦ Dilution to single particle passing through detector◦ Sample size 0.25 ml to 1.0 ml typical

Large particle count (LPC) detectors◦ Offline and inline variations◦ 0.5 µm and above◦ Dilution to single particle passing through detector◦ Sample size 0.25 ml to 1.0 ml typical

3

Particles < 0.3 µm◦ Many particles, by design◦ PSD is well-represented by small sample volumes

Particles > 1.0 µm◦ Few particles, by design◦ PSD is poorly represented by small sample volumes◦ Particle counts are poorly represented by small sample

volumes◦ Sampling count error increases as particle size

increases (particle count decreases)

4

5

Continuous, real-time measurement @ 15 ml/min

Detection range 1-12 µm in 0.2 µm increments

Undiluted CMP slurry, all types

LFCSlurryScope™15 ml/min 15 ml/min

Full flow less 15 ml/min Full flowFull flow

Flow diverter directs slip stream to SlurryScopeS

lurr

ySco

pe

Inpu

tS

lurryScope

Return

6

0.1

1

10

100

1000

10000

100000

1000000

1 2 3 4 5 6 7 8 9 10 11 12

Tota

l Par

ticle

s >1

mic

ron

Particle Size (microns)

Particle Size Distribution

470000

480000

490000

500000

510000

520000

530000

1 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46 49

Tota

l Par

ticle

s >1

mic

ron

Time (arbitrary units)

Total Particles vs. Time

Continuous measurement at a regulated flow of 15 ml/min◦ 0.25 ml in a 1 second measurement period◦ This simulates rapid sampling with other methods◦ Key difference: undiluted, so no risk of soft optical

agglomerates

Use a series of 1 second measurement intervals from SlurryScope to simulate many, many sample measurements by prior methods

Ceria slurry used in these experiments

7

8

1

10

100

1000

10000

100000

1000000

1.2

1.8

2.4 3

3.6

4.2

4.8

5.4 6

6.6

7.2

7.8

8.4 9

9.6

10.2

10.8

11.4 12

1

10

100

1000

10000

100000

1000000

1.2

1.8

2.4 3

3.6

4.2

4.8

5.4 6

6.6

7.2

7.8

8.4 9

9.6

10.2

10.8

11.4 12

1

10

100

1000

10000

100000

1000000

1.2

1.8

2.4 3

3.6

4.2

4.8

5.4 6

6.6

7.2

7.8

8.4 9

9.6

10.2

10.8

11.4 12

1

10

100

1000

10000

100000

1000000

1.2

1.8

2.4 3

3.6

4.2

4.8

5.4 6

6.6

7.2

7.8

8.4 9

9.6

10.2

10.8

11.4 12

Particle size (µm)

Tota

l par

ticle

s >(

x-ax

is) µ

m

9

Each line of data = 1 second sampling interval = 0.25 ml slurry

Particle size bin (µm)

Tota

l par

ticle

s co

unte

d in

siz

e bi

n

10

0.1

1

10

100

1000

10000

100000

1000000

1.2

1.4

1.6

1.8 2

2.2

2.4

2.6

2.8 3

3.2

3.4

3.6

3.8 4

4.2

4.4

4.6

4.8 5

5.2

5.4

5.6

5.8 6

6.2

6.4

6.6

6.8 7

7.2

7.4

7.6

7.8 8

8.2

8.4

8.6

8.8 9

9.2

9.4

9.6

9.8 10

10.2

10.4

10.6

10.8 11

11.2

11.4

11.6

11.8 12

Part

icle

s/m

L

Particle Size (microns)

50 simulated small-volume dilution measurements and 1 SlurryScope measurement (solid blue)

11

12

0%

1%

10%

100%

1000%

0.1 1 10 100 1000 10000 100000 1000000

Percen

t deviation

 (σ/

μ)

Average particle count (per mL)

Percent deviation vs. Particle Count

0.1%

0%

1%

10%

100%

1000%

0.1 1 10 100 1000 10000 100000 1000000

Percen

t deviation

 (σ/

μ)

Average particle count (per mL)

Percent deviation vs. Particle Count

13

Low countsLarge particlesCMP scratches

0.1%

14

Slurry Stock

Sub-fab Operations

Slurry Delivery System

Fab Operations

CMP Polisher

Wafer In

Pad Conditioner

Defect-free Wafer Out

CMP Pad

SlurryScope In-line

Measurement Domain

15

Slurry Stock

Sub-fab Operations

Slurry Delivery System

Fab Operations

CMP Polisher

Wafer In

Pad Conditioner

Scratched Wafer Out

CMP Pad

SlurryScope In-line

Measurement Domain

Root causes unrelated to slurry events

Root causes that may

correlate to slurry events

Periodic online slurry monitoring can assist post-mortem diagnosis of wafer scratching◦ Cannot prevent wafer scratching from occurring

Continuous online slurry monitoring can identify patterns and practices that contribute to LPC excursions◦ Identify and eliminate the root causes of LPC spikes◦ Reduce the incidence of slurry-induced wafer scratching◦ Apply six-sigma principles to prevent wafer scratching

16

17

0

500

1,000

1,500

2,000

2,500

12:1

4:51

12:1

5:31

12:1

6:11

12:1

6:51

12:1

7:31

12:1

8:12

12:1

8:52

12:1

9:32

12:2

0:12

12:2

0:52

12:2

1:32

12:2

2:12

12:2

2:52

12:2

3:32

12:2

4:12

12:2

4:52

12:2

5:32

12:2

6:13

12:2

6:53

12:2

7:33

12:2

8:13

12:2

8:53

12:2

9:33

12:3

0:13

12:3

0:53

12:3

1:33

12:3

2:13

12:3

2:53

12:3

3:33

12:3

4:13

12:3

4:54

12:3

5:34

12:3

6:14

12:3

6:54

12:3

7:34

12:3

8:14

12:3

8:54

12:3

9:34

12:4

0:14

12:4

0:54

12:4

1:34

12:4

2:14

12:4

2:55

12:4

3:35

12:4

4:15

12:4

4:55

12:4

5:35

12:4

6:15

12:4

6:55

12:4

7:35

12:4

8:15

Periodic LPC Spikes Due to Sub-fab Equipment Cycling

LPC Shifts Due to Filter Changes *

*ASMC May, 2012; A. Kim, Mega Fluid Systems & M. Parkin, Vantage Technology Corp.

Small volume slurry sampling delivers the largest sampling error at low particle counts, typically the largest particles which are most critical for wafer scratching

Continuous LPC monitoring with SlurryScope provides a more statistically meaningful representation of the largest particle counts

Continuous monitoring of undiluted slurry provides new information allowing LPC origins to be traced and eliminated, bringing CMP closer to six-sigma process defect control

18

ASMC May, 2012; A. Kim, Mega Fluid Systems & M. Parkin, Vantage Technology Corp.

Feb 22 2012 webcast: http://techcet.com/presentations/ ICPT 2011 Solid State Technology, July 2011 http://www.vantagetechcorp.com/

Upcoming:◦ Semicon West 2012: Malema booth, Levitronix booth◦ Semicon West 2012 CMPUG◦ Clarkson CAMP CMP August 2012◦ ICPT October 2012

19