Beruflich Dokumente
Kultur Dokumente
KOH
CMP
Low-pH Mix
Lev 1
Slurry Feed Tank
pump
Cleanroom
Filtering & Lev 2
Packaging pump
13
75nm High Purity Colloidal Slurry
12.5
12
Spec. Limits
11.5
11
pH
10.5
10
Silco Process
9.5
Industry Standard
9
8.5 DAYS
180
135
1 50
165
120
105
75
60
90
15
30
45
0
Effect of Trace Metals on ILD Polish Performance Silco 75nm ILD Slurry Performance
Normalized Defects Down Force Normalized Defects Polish Rate
Slurry [Al] [Ca] [Cr] [Fe] [Ni] [Na] Slurry (psi)
Uniformity
(A/min)
& Microscratches & Microscratches
198 198
Silco <1 <0.1 <0.1 <1 <0.1 <10 Silco 7.0 6.0% 3800
164 164
588-658
Standard <100 <1.8 NA <6.5 <10 <50
268-316 588-658
Standard 7.0 7.4% 3700
233-277 268-316
Supplier X NA NA NA NA NA NA
451-533
NA 233-277
Typical <50 <5 <1 <20 <100 <100 Alternative 7.6 5.5 to 8.0% 3650
451-533
- Experiments run by a volume IC Fab
- All metals are specs and units in ppm - Experiments run by a volume IC Fab
- All slurries are based on colloidal silica particles - All slurries are based on colloidal silica particles
- Comparable removal rate and uniformity - All slurries have same solids content of silica
Slurry 1, Novellus w/ IC1000 Pads, Wafer #1 Silco EM-7530K, Novellus w/ IC1000 Pads, Wafer #1
Cumulative # of Particles / mL
Cumulative # of Particles / mL
Figure 2. Colloidal Silica Dispersion source and Figure 3. Colloidal Silica Dispersion source and Figure 4. EM-5530HP oxide silica slurry source and
Planargard® high retention filters LPC distribution. Planargard® 1 micron filters LPC distribution. Planargard® high retention filters LPC distribution.
The filtration with the tightest CS filter (Planargard® CS0.2) showed retention of ~99 % for the dispersion and 91
% for the oxide slurry, whereas the slurry filtration with Planargard® CS0.5 and CL0.3 resulted in ~78 % and 89 %
reductions, respectively, in the cumulative LPC ≥ 0.56 micron. The slurry source and filtrate samples collected at
different time points were analyzed for LPC using a Particle Sizing System Accusizer™ 780 APS system (employing a
top injection dilution chamber with single stage auto dilution with measurements started before on-line dilution of the
injected samples). Slurry samples were also tested with Horiba LA-930 mean particle size distribution (PSD) analyzer.
This study demonstrates that Entegris Planargard® CS0.5 and CL0.3 filters can be employed to achieve the
required retention of the cumulative LPC ≥ 0.56 micron in a single pass bulk filtration of colloidal silica dispersion.
Since Planargard® CMP1 filter has much lower initial pressure drop, it may be used on upstream side of the
Planargard® CS0.5 filters to achieve relatively longer filter lifetime for CS0.5 filter.
Another alternative would be to employ Planargard® CL0.3 filter for tighter retention, and use CMP1 filter
upstream. Since the tested silica dispersion and the oxide slurry have relatively smaller mean particle size than
majority of oxide abrasive slurries, using CMP1 filter in the above configuration is expected to provide good life-time for
CMP1 as well as CS0.5 (or CL0.3) filters. Based on the present study, Planargard® CS0.5 or CL0.3 filters may be
employed for the POU filtration of silica slurry.
Acknowledgments
The authors would like to thank Dr. Peter Burke, Clint Jones, and Dr. Ashwani Rawat for their contributions to this
study and numerous insightful discussions. Sincere thanks are due to slurry manufacturers for providing CMP slurry
samples for the slurry characterization evaluations, Levitronix GmbH for providing a magnetically levitated centrifugal
pump for this study, and Contaminations Control Solutions Team at Entegris, Inc., for their support of this project.
Planargard, Mykrolis, and Entegris are registered trademarks of Entegris, Inc., Levitronix is a registered trademark of Levitronix
GmbH, AccuSizer is a trademark of Particle Sizing Systems.