A comparison of optical particle counting and aerosol-based … · 2019. 2. 13. · Van Schooneveld, et al., Ultrapure Water Micro 2017 –Portland, Oregon Slide 1 A comparison of

Slide 1Van Schooneveld, et al., Ultrapure Water Micro 2017 – Portland, Oregon

A comparison of optical particle counting and

aerosol-based particle counting technologies for

UPW monitoring and diagnostics.

Gary Van Schooneveld1, Siqin He2, Pat Keady3

1CT Associates, Inc., 7121 Shady Oak Road, Eden Prairie, MN 55344

2Kanomax FMT, 4104 Hoffman Road, White Bear Lake, MN 55110

3Aerosol Devices Inc., 2614 S. Timberline Road, Fort Collins, CO 80525

CTAssociatesinc.com


Introduction

• Liquid optical particle counting (OPC) utilizing light-scattering has been and

continues to be a key technology used by the semiconductor industry to count

particles in ultrapure water (UPW).

• Demand for detecting particles smaller than the current capabilities of these

instruments (<< 20 nm) has resulted in development and introduction of

instruments that utilize alternate detection techniques such as acoustic

emission1, laser induced breakdown detection2 and nebulization/aerosol

particle counting3.

• Each of these techniques deploy technologies that detect particles, and

potentially other forms of contamination, that differ from the OPC.

• Correlating the data generated by the new technologies to the historical data

base available from OPC’s is important for these new technologies to gain

acceptance and to understand what new and potentially insightful information

may be available using these techniques.

1. Madanshetty, Sameer, Particle profiling of UPW and suspensions, Ultrapure Water Micro 2016, Austin, TX.

2. Boj, Sylvain, et al, “Has the LIBD technique have potential for online Nano-particle detection in UPW?, Ultrapure Water Micro 2015, Portland, Oregon.

3. Blackford, David, et al, Introducing a non-optical 10 nm particle counter for ultrapure water, Ultrapure Water Micro 2014, Phoenix, AZ.


Outline

• Instrumentation overview and capabilities

• Test plan

• Water system description and sampling locations

• Online instrumentation results

• Off-line analysis (SEM/EDS) results

• Key observations and recommendations


Instrumentation

Lighthouse Worldwide

Solutions NC30+ OPC

Kanomax FMT

NanoParticle Nebulizer Model 9110

Aerosol Devices Model SSS110

Sequential Spot Sampler

Kanomax FMT

Liquid Nanoparticle Sizer (LNS)


Theory of Operation and Capabilities -

Optical Particle Counting (OPC)

Courtesy of Noria Communications

Lighthouse Worldwide

Solutions NC30+

Sizing Channels: 30, 50, 80, 100 nm

Source: Van Schooneveld, et al., UPW Micro 2013

OPC response to mono-dispersed PSL


Differential Mobility Analyzer (DMA)

Condensation ParticleCounter (CPC)Precision Nebulizer


Liquid Nanoparticle Sizing (LNS)

CDA

PrecisionNebulizer

Scanning Mobility Particle Sizer (SMPS)ParticleAerosol

ChargeConditioner DMA CPC

monodispersed

aerosolIncoming sample



Liquid Nanoparticle Sizing (LNS)

20-125 nm PSL 8E4/mL > 30 nm

Particle Diameter (nm)

20 30 40 50 60 70 80 90 100 150 200 300

Cum

ula

tive

Co

nce

ntr

atio

n (

#/m

L)

100

101

102

103

104

105

106

LNS, diluted

NC30+

M50e

Excellent correlation observed between

LNS and OPC data using PSL.

Source: Van Schooneveld, et al., UPW Micro 2013



Focused Aerosol Deposition (FAD)

Kanomax FMT NanoParticleNebulizer Model 9110

Aerosol Devices Model SSS110Sequential Spot Sampler

→

CDA

UltrafineNebulizer

ParticleAerosol

Sequential Aerosol Spot Sampler

UPW


Test Plan

• Measure selected points on CTA’s UPW system after

major cleaning, new component install and installation of

new final resin.

• Measure selected points for 12-24 hours before moving to

new location.*

• Eliminate first two hours of data from data analysis.

• Analyze focused aerosol deposition samples via

SEM/EDS.

* In subsequent testing, instrumentation collected data for several days at each sample point.


UPW System Sample Points

1

2

3

4

5

Operating conditions:

• 20 – 24 LPM

• 4 Bar

Ultra-HighPurityTank

185 nm UV Light

IX MixedBed (50 liter)

NRMMonitor

TOCMonitor

F4

P9 P10

P11

P12

T3

R3 PT2

ParticleCounter

High PurityWater System

20" - 20 nm cartridge filter

UF

20" - 20 nm cartridge filter


Relevant Dates and Instruments

• January 26-27, 2017

– Sanitized system with Minncare® cold sanitizer.

– Installed UF Module

– Installed 2-20” filters

– Replace final polish resin.

• February 17-19, 2017

– CTA LNS#1

– Lighthouse NC30+

– Tandum KFMT Nanoparticle Nebulizer with AD Spot Sampler.

• April 5 – May 4, 2017

– CTA LNS#1

– Tandum KFMT Nanoparticle Nebulizer with AD Spot Sampler.


Test Results


Lighthouse NC30+

Date

2/18/2017 2/19/2017 2/20/2017

Cum

ula

tive

Co

nce

ntr

atio

n (

#/m

L)

0.1

1

10

100

1000

10000

> 30 nm

> 50 nm

> 80 nm

> 100 nm

Out of final filter Out of UF Out of IX Out of TOC

LampOut of pump

Liquid Nanoparticle Sizing System

Date

2/18/2017 2/19/2017 2/20/2017

Cum

ula

tive C

oncentr

ation (

> 5

nm

)

1e+6

1e+7

1e+8

1e+9

1e+10

Out of final filter Out of UF Out of IX

Out of TOCLamp

Out of pump

(lost air supply)

Test Results – LNS vs OPC

Observations:

1. Highest LNS and OPC concentrations are out of the TOC lamp.

2. Lowest OPC concentration is out of the UF/MF.

3. Lowest LNS concentration is out of the ion exchange resin


Liquid Nanoparticle Sizing System

Date

2/18/2017 2/19/2017 2/20/2017

Cum

ula

tive

Co

nce

ntr

atio

n (>

5 n

m)

1e+6

1e+7

1e+8

1e+9

1e+10

Out of final filter Out of UF Out of IX

Out of TOC

Lamp

Out of pump

(lost air supply)

LNS Particle Size Distributions


5 6 7 8 9 10 12 15 20 25 30 40 50

Cum

ula

tive C

oncentr

ation (

#/m

L)

1e+5

1e+6

1e+7

1e+8

1e+9

Out of pump (-2.45)Out of TOC Lamp (-2.91)Out of IX resin (-2.44)Out of UF (-2.78)Out of final filter (-2.61)

Test Results – LNS Particle Size Distribution

Observations:

1. Increase in small particle size distribution out of the TOC reduction lamp. Could be

small particle formation/generation in the TOC lamp and/or increased level of

dissolved species (organic or silica).

2. Sharp drop in concentration after UF at 10 nm.

3. Lowest LNS concentration is after IX, not UF or MF


Lighthouse NC30+

Date

Sat 18 Sun 19 Mon 20

Cum

ula

tive

Co

nce

ntr

atio

n (

#/m

L)

0.1

1

10

100

1000

10000

> 30 nm> 50 nm> 80 nm> 100 nm


LampOut of pump

Ion Exchange Particle Retention (NC30+)

Channel (nm)

20 40 60 80 100 120

Pa

rtic

le R

ete

ntio

n (

%)

99.5

99.6

99.7

99.8

99.9

100.0

Lighthouse NC30+

Channel (nm)

20 30 40 50 60 80 100 150

Cu

mu

lative

Co

nce

ntr

atio

n (

#/m

L)

0.01

0.1

1

10

100

1000

10000

Out of pump (-1.79)Out of TOC lamp (-1.78)Out of IX resin (-2.96)Out of UF (-2.46)Out of final filter (-2.08)

Test Results – OPC Particle Size Distribution

Observations:

1. The ion exchange resin exhibited significant particle retention, >

99% of particles detected by OPC ≥ 30 nm.

2. Unusually flat slope (log-log < -2) out of pump and TOC lamp.

Possibly consequence of the sanitizing.

3. Flat slope out of UF/MF is also unusual. Would expect -3.5 to 4.

Possibly shedding from the UF and MF’s.


Particle Size Distributions


5 10 20 30 40 50 60 80 100 120

Cum

ula

tive

Co

nce

ntr

atio

n (

#/m

L)

1e-1

1e+0

1e+1

1e+2

1e+3

1e+4

1e+5

1e+6

1e+7

1e+8

1e+9

Out of pump (LNS)Out of TOC lamp (LNS)Out of IX resin (LNS)Out of UF (LNS)Out of final filter (LNS)Out of pump (NC30+)Out of TOC lamp (NC30+)Out of IX resin (NC30+)Out of UF (NC30+)Out of final filter (NC30+)

Test Results – LNS vs OPC

Observations:

1. Sizing order different between two

techniques out IX resin. Measuring

different particles?

2. Large drop (3 orders) in concentration

after IX observed with OPC but not

observed by LNS. Why??

3. Concentration measured by LNS are 3 to 6

orders of magnitude higher than OPC.

Why??

Particle Size Distributions


30 40 50

Cum

ula

tive

Co

nce

ntr

atio

n (

#/m

L)

1e+5

1e+6

1e+7

Out of pump (LNS)Out of TOC lamp (LNS)Out of IX resin (LNS)Out of UF (LNS)Out of final filter (LNS)


PSD change with run-time (6 weeks)

LNS Particle Size Distributions(February 17 - 19, 2017)


5 6 7 8 9 10 12 15 20 25 30 40 50

Cu

mu

lative

Co

nce

ntr

atio

n (

#/m

L)

1e+5

1e+6

1e+7

1e+8

1e+9

Out of pump (-2.45)Out of TOC Lamp (-2.91)Out of IX resin (-2.44)Out of UF (-2.78)Out of final filter (-2.61)

LNS Particle Size Distributions(April 5, 2017 to May 4, 2017)


5 6 7 8 9 10 12 15 20 25 30 40 50

Cum

ula

tive C

oncentr

ation (

#/m

L)

1e+5

1e+6

1e+7

1e+8

1e+9

Out of PumpOut of TOC lampOut of IX resinOut of UFOut of final filter

Observations:

1. Little change in PSD prior to IX (after pump and TOC lamp).

2. Cleanup observed in IX, UF and final filter. Small particle (< 20 nm)

concentration are similar post IX.

3. Trend in large particles (> 20 nm) are consistent with prior OPC data.

Clean up of UF and

final filter


Focused Aerosol Deposition

SEM/EDS Analysis


Spot Sampler SEM/EDS Analysis

Collection Location: Out of IX resin

Collection Time: 41 hrs

Collection Date: 4/7/2017

1mm




10 - 50nm silica

1000nm organic particles


Elements Detected:

• Carbon

• Oxygen

• Silicon

• Sulphur

• Zinc

• Magnesium





Primarily organic

with silica inter-dispersed

Elements Detected:

• Carbon

• Oxygen

• Sulphur

• Sodium

• Magnesium

• Silicon


Estimated Spot Size – OPC vs LNS

Lighthouse NC30+

Date


Cum

ula

tive

Co

nce

ntr

atio

n (

#/m

L)

0.1

1

10

100

1000

10000

> 30 nm> 50 nm> 80 nm> 100 nm


LampOut of pump



5 6 7 8 9 10 12 15 20 25 30 40

Cu

mu

lative

Co

nce

ntr

atio

n (

#/m

L)

1e+5

1e+6

1e+7

1e+8

1e+9

Out of IX resin

Approach:• Integrate differential size data.

• Only consider particles > 7 nm.

• Extrapolate OPC PSD to 7 nm using power

law: 𝑓 𝑥 = 𝐾 ∗ (1/dn) where n=3.

• Assume that the OPC measurements from

February are still valid.

• Assume 1% packing efficiency.

• Assume 2D packing only.


Lighthouse NC30+

Date


Cum

ula

tive

Co

nce

ntr

atio

n (

#/m

L)

0.1

1

10

100

1000

10000

> 30 nm> 50 nm> 80 nm> 100 nm


LampOut of pump



5 6 7 8 9 10 12 15 20 25 30 40

Cu

mu

lative

Co

nce

ntr

atio

n (

#/m

L)

1e+5

1e+6

1e+7

1e+8

1e+9

Out of IX resin

Projected spot diameter based

on integrated OPC data – 0.004 mm

Estimated Spot Size – OPC vs LNSApproach:

• Integrate differential size data.









Lighthouse NC30+

Date


Cum

ula

tive

Co

nce

ntr

atio

n (

#/m

L)

0.1

1

10

100

1000

10000

> 30 nm> 50 nm> 80 nm> 100 nm


LampOut of pump



5 6 7 8 9 10 12 15 20 25 30 40

Cu

mu

lative

Co

nce

ntr

atio

n (

#/m

L)

1e+5

1e+6

1e+7

1e+8

1e+9

Out of IX resin


on LNS data – 0.63 mm



Estimated Spot Size – OPC vs LNSApproach:

• Integrate differential size data.









Lighthouse NC30+

Date


Cum

ula

tive

Co

nce

ntr

atio

n (

#/m

L)

0.1

1

10

100

1000

10000

> 30 nm> 50 nm> 80 nm> 100 nm


LampOut of pump



5 6 7 8 9 10 12 15 20 25 30 40

Cu

mu

lative

Co

nce

ntr

atio

n (

#/m

L)

1e+5

1e+6

1e+7

1e+8

1e+9

Out of IX resin


on LNS data – 0.63 mm



Estimated Spot Size – OPC vs LNS

Focused Aerosol Deposition SEM results

seem to correlate better with LNS data than

OPC data in this particular study.

Additional work will be required to validate

this observation in other locations within the

UPW system.


Key Observations

• While particle concentrations varying significantly between the LNS and OPC,

trends were consistent except out of the IX resin.

• Focused Aerosol Deposition with SEM/EDS appear to support the higher

concentrations measured by the LNS.

• The vast majority of the material deposited by the AD Spot Sampler were

organics and silica. Some organics materials are more difficult to detect by

light scattering due to their refractive index being close to water. Same is true

for silica. This may be one of the reasons for the differences observed

between the instruments.

• The TOC reduction lamp was a significant contributor to sub-20 nm

contamination. It is uncertain if these are particles, dissolved material or both.

Further analysis using Focused Aerosol Deposition may help to determine the

nature of these particles.

• The mixed-bed IX resin used in this test has significant particle removal

capability, especially greater than 30 nm, except for silica.

CTAssociatesinc.com


Recommendations

• Continue developing our understanding of the differences

between aerosol-based and light-scattering particle

measurement techniques:

– Conduct parallel testing with aerosol-based and light scattering

instruments.

– Include alternate technology instruments as they become available.

– Utilize new tools such as Focused Aerosol Deposition to collect

and analyze the nanoparticles found in modern UPW systems.

– Work to establish a quantitative correlation between FAD, aerosol-

based and OPC results.

– Continue controlled testing to measure the response of the

instruments to nanoparticles of varied compositions.


Acknowledgements

• University of Minnesota’s Materials Charactization Lab

for the use of the Hitachi SU 8230 FE/SEM and Thermo

Noran System 7 Spectral Imaging (EDS) instruments.

CTAssociatesinc.com

Documents

A comparison of optical particle counting and aerosol-based … · 2019. 2. 13. · Van Schooneveld, et al., Ultrapure Water Micro 2017 –Portland, Oregon Slide 1 A comparison of