Transcript NRM 8000 in English 20121010 - quest

A New Nonvolatile Residue Monitor
for the Semiconductor Industry
By
David Blackford Ph.D.
Fluid Measurement Technologies
Questions about Nonvolatile
Residue
• What exactly is nonvolatile residue?
• Why do I need to measure it?
• What good is it if you can’t tell me what the
residue consists of?
• And so far I have managed perfectly well
without measuring NVR, so why should I
start now?
What exactly is nonvolatile residue?
Nonvolatile residue is dissolved inorganic
material.
• Primarily silica, both dissolved and colloidal.
• Also ions and trace metals.
Nonvolatile residue is not.
• Particles.
Why do I need to measure NVR?
Nonvolatile Residue is increasingly being
used as a measure of overall water quality. It
is frequently the first on-line measurement
to indicate a problem in a UPW system.
Nonvolatile residue can now be monitored at
ppt levels in almost real-time.
What good is NVR if you can’t tell me
what the residue consists of?
On-line TOC monitors and particle counters
cannot identify the specific TOC compound
or particle composition.
BUT, speed is everything. The true value
of on-line monitors comes from a fast
indication of a problem. Then, you can
collect and analyze a sample to identify the
source of the contamination.
And so far I have managed perfectly well
without measuring NVR, so why should I
start now?
As semiconductor line-widths become smaller,
semiconductor manufacturing becomes more
susceptible to contamination.
As the ITRS Roadmap committee looks to the
future, it is working to find parameters to measure
at the required detection limits. So the day is
coming when maybe you should consider
monitoring NVR.
How the NRM was “born”
• 20 years ago, Bob MacIntosh asked me if there
was a way to get some response to the presence
of colloidal silica in UPW.
• In 1992 the original NRM was introduced.
• The basic concept was not new. In 1964 two
Australian professors had proposed using a
nebulizer, a drying column and a particle
counter to measure nonvolatile residue in
solvents.
The original NVR nephelometer system
(Salkowski and Werle, 1964)
Schematic for the new Nonvolatile
Residue Monitor
The NRM technique
“Drip-counter” for measuring very
low flow rates
Improvements over the original NRM
•
•
•
•
•
•
•
•
Expanded measurement range, 1ppt to 60 ppb.
No more n-butyl alcohol.
10x faster response time.
1/3rd the weight and 2/3rd the volume.
Innovative water inlet flow monitor.
Able to measure water up to 80 degrees centigrade.
Universal power supply.
Pressure actuated water regulator (no more water
leaks).
KCl
challenges
NRM Model 8000
14000
Raw NRM count data
3 ppb
2 ppb
12000
1 ppb
10000
8000
6000
4000
2000
0
0
50
100
Time (minutes)
150
200
9.8 ppt challenge of KCl
Data averaged every 60 seconds
Raw NRM count data
1500
1400
1300
1200
1100
1000
900
0
500
1000
1500
2000
Time (seconds)
2500
3000
3500
400
4
300
3
200
2
NVR
100
1
SiO2
0
0
4pm 5pm 6pm 7pm 8pm 9pm 10pm 11pm 12pm 1am 2am 3am 4am 5am 6am
Time
Dissolved Silica (ppb)
Nonvolatile Residue (ppt)
US Semiconductor Facility
Simultaneous data collected for the NRM and a Hach
Silica Monitor
Low level (<10 ppb) 22nm colloidal silica challenge
20 nm pore size filter
2000
1600
400 mesh
diffusion screen
through filter
1400
1200
filter bypassed
Raw NRM count data
1800
1000
800
600
400
200
0
0
10
20
30
40
50
Time (minutes)
Filter removal efficiency 72%
60
70
2007 ITRS “Round Robin”
Data collected at Point of Entry (POE)
Parameters monitored were particles by SEM, TOC
An/Cat and TM, DO and DN and NVR.
Average NVR data for UPW collected over 3 days
Site A
245 ppt
Site B
243 ppt
Site C
Site D
199 ppt
156 ppt
Site E
102 ppt
Site F
96 ppt
Conclusions
The new Nonvolatile Residue Monitor offers
improved contamination measurement of
UPW:
• Expanded measurement range.
• Smaller, lighter, faster.
• Knocking at the door of ppq.