Further Evaluation of Two Plasma Technologies for PFC Emissions Reduction

SSA 2000 Annual Conference, Arlington, Virginia Further Evaluation of Two Plasma Technologies for PFC Emissions Reduction Walter Worth Program Manager, ESH April 27, 2000

Background Perfluorocompounds (PFCs) are potent greenhouse gases In April of 1999, World Semiconductor Council (WSC) established international PFC emissions reduction goal -- 10% absolute reduction by 2010, over 1995 baseline To meet aggressive goal, fabs will need a 95% PFC utilization or destruction for PECVD chamber cleans and, selectively, some etch tools Currently, plasma technologies appear to offer the most cost-effective way of achieving this goal p:\955\a-pfc_us\ssa200.ppt 2

Growth of Emissions with No Action (linear extrapolation of 95-98 data for U.S.EPA MOU Partners) 3.0 Emissions (MMTCE) 2.0 1.0 0 10% Goal '95 '96 '97 '98 '99 '00 '01 '02 '03' 04' 05' '06 '07 '08 '09 '00 Year p:\955\a-pfc_us\ssa200.ppt 3

PFC Emissions Reduction Options Alternative Chemicals Destruction/ Decomposition Global Warming Gas Reduction Process Optimization Recovery & Recycle Combustion Plasma Thermal/ Chemical p:\955\a-pfc_us\ssa200.ppt 4

Introduction At the SSA 99 Conference, we reported the results from preliminary evaluations of: AMAT s µclean microwave plasma technology for PECVD chamber cleaning Litmas Rf plasma abatement device for etch tool emissions Both were first-time evaluations of these technologies in a manufacturing environment The results were very encouraging, but the tests were short and long-term reliability was still a question The results presented today are from evaluations of commercial, improved versions of these two devices p:\955\a-pfc_us\ssa200.ppt 5

Different Technology Options NF 3 /Ar Remote Plasma To House Scrubber Alternative Chemistries (e.g., dilute NF 3 ) Turbo Pump Pre-Pump Abatement CVD/Etch TOOL Dry Pump Post-Pump Abatement p:\955\a-pfc_us\ssa200.ppt 6

Evaluation of Applied Materials Remote Clean Technology for DxZ Chambers

Applied Materials Remote Clean Technology Utilizes low-field toroidal plasma to generate atomic F upstream of CVD chamber Input gases: NF 3 and Argon (1:2 ratio) Compact, lid-mountable unit No additional rack components Has foreline clean endpoint detector Optical emission spectrometer monitoring SiF 4 p:\955\a-pfc_us\ssa200.ppt 8

Remote Clean Unit on DxZ Lid p:\955\a-pfc_us\ssa200.ppt 9

Demonstration Goals Conduct two 1000 wafer marathons with depositions in DxZ TM chamber PETEOS and doped PETEOS films Measure film parameters and particles Utilize endpoint detector to determine clean time Conduct Design of Experiments (DOE) Optimize clean time, NF 3 and Ar flows Characterize emissions Verify > 99% NF 3 utilization Quantify F 2 emissions p:\955\a-pfc_us\ssa200.ppt 10

Summary of Marathon Results Both 1000 marathons completed successfully NF 3 = 700 sccm, Ar = 1400 sccm Clean time for 1 µm PETEOS film = 55 seconds did not vary throughout marathons All film parameters within specification limits No particle issues p:\955\a-pfc_us\ssa200.ppt 11

10KÅ Film Thickness/Uniformity PETEOS 10,200 2.5 Thickness (A) 10,100 10,000 9,900 2 1.5 1 0.5 Non-Uniformity (%) 9,800 0 0 200 400 600 800 1000 Wafer # Thickness (A) % Non-uniformity p:\955\a-pfc_us\ssa200.ppt 12

Particles - PETEOS 50 40 Particle Adders 30 20 10 0 0 200 400 600 800 1000 Wafer # P articles p:\955\a-pfc_us\ssa200.ppt 13

Film Stress/Refractive Index PETEOS Stress R efrac tive Index -80 1.465-90 Film stress -100-110 -120-130 Refractive Index 1.455-140 0 200 400 600 800 1000 Wafer # Stress - 4K Stress - 8K 1.445 0 200 400 600 800 1000 wafer # R efrac tive Index p:\955\a-pfc_us\ssa200.ppt 14

DOE Results for 1µm Oxide Film NF 3 NF 3 Clean Used Flow Rate Time (scc) (sccm) (sec.) Min. NF 3 Usage 453 400 68 Baseline Clean 555 605* 55 Min. Clean Time 615 997 37 *used 700 sccm and 55 sec. for marathon p:\955\a-pfc_us\ssa200.ppt 15

Baseline Conditions for Remote Clean Process NF 3 Consumption = 610 scc/wafer Time (sec) 52.3 40 700 1400 4 400 1000 800 2000 2 8 NF (sccm) Argon (sccm) Pressure (Torr) 3 p:\955\a-pfc_us\ssa200.ppt 16

Sampling Apparatus To Acid Exhaust FTIR PI SAMPLE PUMP CALIBRATION SYSTEM 750 torr PI SAMPLE PUMP TO CORROSIVE EXHAUST N2 PURGE PROCESS PUMP MASS SPECTROMETER Tool Exhaust GASES p:\955\a-pfc_us\ssa200.ppt 17

Emissions from Standard TEOS Process (per wafer pass) Film 2.5K Å 6.5K Å 10K Å NF 3 IN (scc) 316 359 495 476 496 688 638 628 NF 3 (scc) 4 3 5 5 8 6 6 5 SiF 4 (scc) 51 43 78 78 74 123 119 115 F 2 (scc) 386 408 551 556 530 771 717 690 HF (scc) 16 17 16 14 19 25 27 22 Utilization (% ) 99 99 99 99 98 99 99 99 F-B alance 1.06 0.94 0.97 1.02 0.94 1.01 1.02 1.00 p:\955\a-pfc_us\ssa200.ppt 18

Comparison of PFC Emissions 20.00 Global Warming Potential lbs (C eq)/wp 15.00 10.00 5.00 0.00 In situ C 2 F 6 RF Clean Remote Plasma NF 3 Clean p:\955\a-pfc_us\ssa200.ppt 19

Comparison of HAP Emissions 0.004 Hazardous Air Pollutants lbs (HF eq)/wp 0.003 0.002 0.001 0.000 In situ C 2 F 6 RF Clean Remote Plasma NF 3 Clean p:\955\a-pfc_us\ssa200.ppt 20

Simplified Cost of Ownership Throughput (continuous) 55 wafers/hour Availability 90% Utilization 70% Gross throughput 291,060 wafers/year Retrofit capital costs $47,290/year Unit retrofit capital $0.16 Installation cost $0.00 Clean gas cost $0.52 Consumables cost $0.15 Total $0.83/wafer Assumptions: TEOS oxide film deposition = 1µm A chamber clean after every wafer 4-chamber Centura PECVD tool Retrofit capital cost/chamber = $59,900 Installation labor cost/chamber = $3,360 Capital & labor amortized over 5 years NF 3 and argon are available at tool Cost of NF 3 = $110/lb Maintenance kit = $11,000/yr/chamber p:\955\a-pfc_us\ssa200.ppt 21

Summary Remote Clean is viable alternative to in-situ C 2 F 6 clean Virtually eliminates PFC emissions No negative effects on deposition process Only negatives are: Cost of NF 3 is high Tool retrofit cost is significant By-product is large amounts of F 2 p:\955\a-pfc_us\ssa200.ppt 22

Reliability Study of Litmas Plasma Abatement Device

Test Program DRE/By-Product Evaluation DOE and 1050 wafer marathon on dielectric etch tool Blanket I-line photoresist wafers High flow tests for CF 4, NF 3, SF 6 and C 4 F 8 Analysis by extractive FTIR (10 cm) and QMS Long-Term Operation Evaluation Exposure of pump internals to HF and H 2 O Effect of repeated cycling on device (e.g., plasma tube) Service requirements Process Impact Evaluation Foreline pressure increase Process drift, particle generation Evidence of back-streaming p:\955\a-pfc_us\ssa200.ppt 24

Litmas Plasma Abatement Device Specifications: 1200 watts of RF power (1800 W supply) 1.7-2.2 MHz variable frequency Solid-state match (no moving parts) 25-100 ms typical plasma strike time 1.92 i.d. x 9 Al 2 O 3 tube 4 lpm cooling water 16 status and power LED s plasma turned on & off by tool variable shutoff delay p:\955\a-pfc_us\ssa200.ppt 25

Turbomolecular Pump Auxiliary Gas Isolation Valve Metering Valve Medium Density Oxide Etcher Tool Effluent O 2 In Isolation Valve To Scrubbed Exhaust Schematic of the Device Installation Water Infusion System Cooling Water Ports Power Supply and Matching Circuitry To FTIR 100 300 mtorr Signal From Tool Final Valve N 2 Purge Gate Valve Dry Pump p:\955\a-pfc_us\ssa200.ppt 26 MF MFC C N 2 Purge 45 slm

Results from the DOE CF 4 CHF 3 Argon CF 4 Emission Matrix Flow Flow Flow DRE Reduction Point (sccm) (sccm) (sccm) (%) (% Kg CE) 1 4 40 50 96.5 97.7 2 4 40 70 92.5 98.6 3 4 60 50 92.3 96.8 4 4 60 70 94.5 98.6 5 6 40 50 96.9 98.7 6 6 40 70 97.5 98.9 7 6 60 50 92.8 97.9 8 6 60 70 93.2 97.9 center 5 50 60 95.9 98.8 p:\955\a-pfc_us\ssa200.ppt 27

FTIR Spectra for Centerpoint Etch Recipe Abatement OFF 5 sccm CF 4 50 sccm CHF 3 60 sccm Ar 166 mtorr pressure Absorbance.3.25.2.15.1.05 HF 134 ppm CF 4 CHF 3 0 4500 4000 3500 3000 2500 2000 1500 1000.25 HF Abatement ON 1200 W power 150 sccm H 2 O 300 mtorr pressure Absorbance.2.15.1.05 CO 2CO H 2 O 4 ppm CF 4 0 p:\955\a-pfc_us\ssa200.ppt 28 4500 4000 3500 3000 2500 2000 1500 1000 Wavenumber (cm -1 )

Effect of Rf Power on CF 4 DRE CF 4 DRE vs. Applied Abatement Power CF 4 DRE (%) 100 90 80 70 60 50 40 30 20 10 0 400 500 600 700 800 900 1000 1100 1200 Pow er (Watts) 5ccm CF 4 50 sccm CHF 3 60 sccm Ar 75 Watt increments 1200W p:\955\a-pfc_us\ssa200.ppt 29

Effect of Water Concentration on CF 4 DRE CF4 DRE (%) 100 80 60 40 20 0-20 -40-60 CF 4 DRE (%) 0 1000 2000 3000 4000 W a te r Conce ntra tion (ppm ) Optimum Water Concentration ~3300 ppm p:\955\a-pfc_us\ssa200.ppt 30

Destruction of CF 4 at High Flows* Foreline CF 4 Flow Ar Flow CF 4 DRE P re ssure sccm sccm % m Torr 25 75 98.6 238 50 100 96.8 280 62 100 95.4 316 75 100 93.5 336 87 100 91.8 352 100 100 85.7 363 * with 150 sccm of H 2 O injection p:\955\a-pfc_us\ssa200.ppt 31

Destruction of Other Gases at High Flows Foreline Gas Flow Ar Flow DRE Pressure Gas sccm sccm % mtorr a NF 3 100 100 95.5 336 b SF 6 50 50 81.8 230 c C 4 F 8 20 100 >99 319 a with 100 sccm O 2 addition b with 150 sccm O 2 addition c with 130 sccm of H 2 O addition p:\955\a-pfc_us\ssa200.ppt 32

C 4 F 8 /CF 4 By-products Input: 8 sccm C 4 F 8, 12 sccm CF 4, 140 sccm H 2 O Abatement OFF Abatement ON Absorbance.4.3.2.1 0 ppm C 4 F 8 49 C 3 F 8 17 C 2 F 6 40 CF 4 186 CO 770 SiF 4 10 HF CO CF 4 C 4 F 8 Absorbance.2.15.1.05 0 HF ppm ppm CF 4 9.7 H 2 O 529 CO 1702 HF 2408 CO 2 97 SiF 4 12 COF 2 13 CO 2 CO CF 4 4500 4000 3500 3000 2500 2000 1500 1000 Wavenumber (cm -1 ) 4500 4000 3500 3000 2500 2000 1500 1000 Wavenumber (cm -1 ) p:\955\a-pfc_us\ssa200.ppt 33

SF 6 By-Products (no etching in tool, abatement ON, H 2 O addition) 12.5 sccm SF 6 140 sccm H 2 O 50 sccm SF 6 140 sccm H 2 O Absorbance.15.1.05 HF H 2 O ppm SF 6 0 SO 2 293 HF 1426 H 2 O 1386 H 2 O Absorbance.25.2.15.1.05 HF ppm SF 6 50 SO 2 428 HF 2401 H 2 O 177 SO 2 F 2 SO 2 SF 6 SOF 2 0 0 4500 4000 3500 3000 2500 2000 1500 1000 4500 4000 3500 3000 2500 2000 1500 1000 Wavenumber (cm -1 ) Wavenumber (cm -1 ) p:\955\a-pfc_us\ssa200.ppt 34

High Resolution OES Spectrum of CF 4 /Ar Plasma During H 2 0 Backstreaming Tests 70000 High Resolution Charge-Coupled Device (CCD) OES Spectrum 60000 50000 No peaks detected for H and OH (H at 286.1,656.2 nm; OH at 282.9, 308, 356.5 nm) Signal Counts 40000 30000 20000 10000 0 200 300 400 500 600 700 800 900 Wavelength (nm) p:\955\a-pfc_us\ssa200.ppt 35

Pump Inspection Results Etch tool processed 11,500 wafers over 8 months Plasma in Litmas device was on for 357 hours Pump was disassembled and visually inspected at end of test. Found: no visible corrosion in inlet no particles or films in foreline no pump inlet o-ring degradation no particles or films in silencer no visible contamination of bearing grease p:\955\a-pfc_us\ssa200.ppt 36

Close-up of Dry Pump Claw Indicating No Teflon Degradation p:\955\a-pfc_us\ssa200.ppt 37

Close-up of Lobes & Stator Surfaces Showing No Abraison or Film Buildup p:\955\a-pfc_us\ssa200.ppt 38

Summary 8-month performance of device was excellent: Totally integrated with etch tool; plasma on when gases flow Required no service over the whole period Refilling of water reservoir was only requirement Increased foreline pressure and H 2 O addition had no effect on etch tool performance: No increase in particles No backstreaming was observed Emission reduction for CF 4 /CHF 3 /Ar etch recipe was ~96.8-98.9%; for CF 4 /C 4 F 8 /Ar ~96.6-97.8 At CF 4 flows > 50 sccm, DRE drops off significantly There was no visible impact on pump hardware p:\955\a-pfc_us\ssa200.ppt 39

Acknowledgements The following persons participated in the evaluations and the preparation of the reports. Their extraordinary efforts are greatly appreciated: Applied Materials Remote Plasma Device Laura Mendicino, Paul Thomas Brown & Stan Filipiak (Motorola) Alan Atherton, Martin Seamons, BipinThakur, Heath DeShong, Toni Vaughan, Thomas Nowak & D.Silvetti (Applied Materials) Andrew Johnson & Richard Pierce (Air Products) Litmas Abatement Device Victor Vartanian, Laurie Beu, T.Stephens, J.Rivers & B. Perez (Motorola) Eric Tonis, Mark Kiehlbauch & David Graves (U of CA at Berkley) p:\955\a-pfc_us\ssa200.ppt 40