Characterization of Organic Contaminants Outgassed Form

Characterization of Organic Contaminants Outgassed from Materials Used in Semiconductor Fabs/Processing

Peng Sun and Caroline Ayre California Materials Technology Intel Corporation

3/25/03 P. Sun/C. Ayre, CA MATTEC, Intel 1 Outline

• Introduction • Analytical Methods for Materials’ Outgassing Measurements • Characterization of Outgassed Organic Contaminants Using Thermal Desorption-Gas Chromatography- Mass Spectrometry (TD-GC-MS) • Ammonia/Amines’ Outgassing • Summary

3/25/03 P. Sun/C. Ayre, CA MATTEC, Intel 2 Introduction: Airborne Molecular Contamination (AMC) Airborne contaminants that: – Are smaller than particles, << 0.01 um or 100 Å, typically: 2-50 Å. – Are in the form of molecular vapor. – Pass through HEPA/ULPA filters.

Molecules Special filters needed (vapor) Airborne contamination Std HEPA/ULPA Particles filters sufficient

• Note: Most cleanrooms are clean for particles, but contaminated with AMC. HEPA/UPLA filtration systems can be AMC sources.

3/25/03 P. Sun/C. Ayre, CA MATTEC, Intel 3 Introduction: Airborne Molecular Contamination (AMC)

• AMC Classification (SEMI F21-95): – Acids: HCl, HF, HNO3, H2SO4 – Bases: NH3, amines, NMP – Condensables: organics with boiling point (bp) > 150 oC – Dopants: P, B, As compounds • Material outgassing is one of the major sources of AMC.

• Notes: 1. Some compounds belong to more than one class. 2. Some compounds are not covered by this classification. 3. NMP: N-methyl-pyrrolidinone, photoresist stripper or polyimide solvent.

3/25/03 P. Sun/C. Ayre, CA MATTEC, Intel 4 Introduction: Molecular Condensables (MC) • Known Problematic Organic Contaminants: – Flame retardants, e.g. organophosphorus – Siloxanes – Polymer/plastic additives, e.g. plasticizers, antioxidants – Amines/amides (such as NMP) – Decomposition compounds

• Effects of Organic Contamination: – Surface properties (e.g. hydrophobicity) – Lower breakdown voltage – Unintentional doping – Haze degradation – Oxide growth rate and quality – Post-CVD defects – Delamination – Optic and mask hazing 3/25/03 P. Sun/C. Ayre, CA MATTEC, Intel 5 Introduction: Effects of Organic Contamination - Unintentional Doping Due to Outgassing Phosphorus contamination vs. exposure time of Sheet resistance map of a contaminated wafer witness wafer to cleanroom contaminated by Fyrol

• Unintentional doping on Si device wafers during furnace operation was observed. • Witness wafer test showed that phosphorus contamination on witness wafers was roughly linear with exposure time. • TD-GCMS analysis of various components of HEPA filters identified that the contaminant as Fyro PCF flame retardant outgassed from the HEPA polyurethane potting material.

Reference: J.A. Lebens, et al., “Unintentional doping of wafers due to organophosphates in the cleanroom ambient”, J. Electrochem. Soc., 143(9), 2906-2909 (1996). 3/25/03 P. Sun/C. Ayre, CA MATTEC, Intel 6 Introduction: Effects of Organic Contamination - Breakdown Voltage Reduction

Wafers exposed to cleanroom w/o chemical filters

Wafers exposed to cleanroom with chemical filters

Device structure used to measure breakdown strength of inter-polysilicon SiO2

• Notes: 1. Organic contamination on the first SiO2 surface causes degradation in the polysilicon layer, resulting in breakdown field strength reduction. 2. Storage time in cassettes prior to the first polysilicon deposition can affect breakdown field strength. Reference: M. Tamaoki, et al., “The effect of airborne contaminants in the cleanroom for ULSI manufacturing process”, 1995 IEEE/SEMI Advance Semiconductor Manufacturing Conference, 322-326 (1995). 3/25/03 P. Sun/C. Ayre, CA MATTEC, Intel 7 Introduction: Sources of Organic Outgassing Contamination – sealants – adhesives – paints – lubricants – plastics – HEPA/ULPA filtration systems – floor tiles – wafer carriers – cleanroom consumables: gloves, garments, tapes, cleaners… – process chemicals – people …

3/25/03 P. Sun/C. Ayre, CA MATTEC, Intel 8 Introduction: ITRS Roadmap for Wafer Surface Organics

Wafer surface organic contamination: Carbon atoms/cm2:

2001 2002 2003 2004 2005 2006 2007 Technology (nm): 130 90 65

‘99 edition: 6.6x1013 5.3x1013 4.9x1013 4.5x1013 4.1x1013 2.1x1013 (2008)

‘02 edition: 2.6x1014 2.1x1013 1.8x1013 1.5x1013 1.3x1013 1.1x1013 1.0x1013

• Trend: Tighter wafer surface organic contamination requirements due to increasing sensitivity to organic condensables’ contamination.

3/25/03 P. Sun/C. Ayre, CA MATTEC, Intel 9 Outline

3/25/03 P. Sun/C. Ayre, CA MATTEC, Intel 10 Analytical Methods for Outgassing Measurements • Most widely used methods in semiconductor industry: – Direct Outgassing • Weight loss method – for total outgassing assessment • TD-GC-MS method – for outgassing compound identification and quantification

– Indirect Outgassing Measurement Using Witness Wafers • TD-GC-MS analysis of witness wafers – full wafer analysis • Time-of-flight Secondary Ion Mass Spectrometry (TOF-SIMS) analysis of witness wafers – spot analysis

3/25/03 P. Sun/C. Ayre, CA MATTEC, Intel 11 Analytical Methods for Outgassing Measurements • Industry Standards: – ASTM F1227-89, “Standard Test Method for Total Mass Loss of Materials and Condensation of Outgassed Volatiles on Microelectronics-Related Substrates”

– IEST WG-CC031 Recommended Practice: “Method for Characterizing Outgassed Organic Compounds from Cleanroom Materials and Components” (to be published in 2003)

– SEMI Environmental Contamination Control (ECC) WG: SEMI E108-0301: “Test Method for The Assessment of Outgassing Organic Contamination from Minienvironments Using Gas Chromatography/Mass Spectroscopy”

– ASTM F1982-99: “Standard Test Methods for Analyzing Organic Contaminants on Silicon Wafer Surfaces by Thermal Desorption Gas Chromatography”

3/25/03 P. Sun/C. Ayre, CA MATTEC, Intel 12 Outline

3/25/03 P. Sun/C. Ayre, CA MATTEC, Intel 13 Characterization of Outgassed Organic Contaminants Using TD-GC-MS

• Schematic Diagram of TD-GC-MS Technique:

3/25/03 P. Sun/C. Ayre, CA MATTEC, Intel 14 Detection of Outgassed Polycyclodimethyl- siloxanes from a Cleanroom Sealant

• TD-GC-MS Analysis Results:

Total outgassing: 680 ug/g Direct outgassing of Siloxanes: 290 ug/g a sealant

Surface analysis of a witness wafer exposed to Total organics: 7.6 ng/cm2 the sealant: Siloxanes: 6 ng/cm2 Siloxane: 2.5E14 Carbon atoms/cm2, 10x > ITRS requirement (1.8E13)

Retention Time (min) ug = microgram, ng = nanogram

3/25/03 P. Sun/C. Ayre, CA MATTEC, Intel 15 Detection of Outgassing Poly- Cyclodimethylsiloxanes • Identification of Dodecamethyl-cyclohexasiloxane: Abundance Scan 1815 (10.891 min): 10300202.D 341 Unknown 9000 8000 73 spectrum 7000 6000 5000 429 4000 3000 2000 147 1000 45 91 117 179 207 225 251 281 311 367 397 0 40 60 80 100 120 140 160 180 200 220 240 260 280 300 320 340 360 380 400 420 440 m/z--> Abundance Dodecamethyl-cyclohexasiloxane 73 #27931: Cyclohexasiloxane, dodecamethyl- 9000 standard spectrum 8000 341 7000 6000 5000 429 4000 3000 2000 147 1000 45 117 207 267 311 397 0 40 60 80 100 120 140 160 180 200 220 240 260 280 300 320 340 360 380 400 420 440 m/z-->

3/25/03 P. Sun/C. Ayre, CA MATTEC, Intel 16 Detection of Triethyl Phosphate (TEP) Outgassed from a Cleanroom Sealant A bundance TIC: 12060206.D 9500000 9000000 Direct outgassing of a 8500000 8000000 7500000 cleanroom sealant 7000000 Triethyl phosphate: 6500000 6000000 5500000 49 ug/g 5000000 4500000 4000000 3500000 3000000 2500000 2000000 1500000 1000000 500000 5.00 10.00 15.00 20.00 25.00 30.00 35.00 Ti me--> A bundance TIC: 12160202.D 1.4e+07 1.3e+07 Surface analysis of 1.2e+07 1.1e+07 Triethyl phosphate: witness wafer exposed 1e+07 9000000 5.8 ng/cm2 8000000 to the same sealant 7000000 6000000 5000000 4000000 ~2E13 P atoms/cm2, 3000000 2000000 >> critical level: 1E12 1000000 5.00 10.00 15.00 20.00 25.00 30.00 35.00 Time--> Retention Time (min)

3/25/03 P. Sun/C. Ayre, CA MATTEC, Intel 17 Detection of Triethyl Phosphate (TEP) Outgassed from a Cleanroom Sealant • Identification of Triethyl phosphate (TEP):

Abundance Scan 1929 (11.563 min): 12160202.D 155 8000 99 6000 127 Unknown spectrum 4000 81 2000 111 139 33 45 65 167 182 207 235 263 291 0 20 40 60 80 100 120 140 160 180 200 220 240 260 280 Abundancem/z--> 99 #43455: Triethyl phosphate 8000 155 Triethyl phosphate 6000 81 127 4000 29 standard spectrum 2000 45 111 15 65 139 167 182 0 20 40 60 80 100 120 140 160 180 200 220 240 260 280 m/z-->

3/25/03 P. Sun/C. Ayre, CA MATTEC, Intel 18 Characterization of Outgassed Organic Contaminants Using TD-GC-MS • Polymer materials chosen outgas organic contaminants commonly seen in semiconductor Fab cleanrooms. Organic contaminants monitored included: • polycyclodimethylsilocxanes • Dibutyl phthalate (DBP) • Butylated hydroxytoluene (BHT) • 2-Ethyl-1-hexanol • Methylstyrene • Naphthalene

• Outgassing behaviors of selected contaminants were studied by varying the following parameters: • outgassing time • outgassing temperature • surface area •weight

3/25/03 P. Sun/C. Ayre, CA MATTEC, Intel 19 Outgassing Concentration vs. Outgassing Time: Siloxanes

Outgassing Time Dependance (at 50oC) 1.20

1.00 y = 0.0358x + 0.2154 y = 0.0394x + 0.0609 2 R2 = 0.9816 R = 0.9872 w, ng/g)

0.80

0.60

0.40

Octamethyl-cyclotetrasiloxane 0.20 y = 0.0236x + 0.0744 Decamethyl-cyclopentasiloxane Outgassing Concentration (ppm R2 = 0.9993 Dodecamethyl-cyclohexasiloxane 0.00 0 5 10 15 20 25 30 35 Outgassing Time (min) • Linear relationship between outgassing concentration and outgassing time within a time range. Outgassing rate can be estimated from the slope. • Same findings for Dibutyl phthalate (DBP) and Butylated hydroxytoluene (BHT)(not shown). • Linear relationship was also observed at 75 oC with higher slope values, indicating higher outgassing rate at 75 oC.

3/25/03 P. Sun/C. Ayre, CA MATTEC, Intel 20 Outgassing Concentration vs. Temperature: Siloxanes and DBP

Outgassing vs. Temperature

2.50

2.00 y = -3.9015x + 11.825 R2 = 0.9726 1.50 y = -1.782x + 5.4216 y = -1.724x + 5.301 R2 = 0.9977 2 1.00 R = 0.9945 y = -1.5764x + 4.6101 R2 = 0.9841 0.50 Log (Outgassing Conc.) 0.00 2.50 2.60 2.70 2.80 2.90 3.00 3.10 3.20

-0.50 Octamethyl-cyclotetrasiloxane Decamethyl-cyclopentasiloxane Dodecamethyl-cyclohexasiloxane -1.00 Dibutyl Phthalate (DBP) 1/T x 103 (1/K x 103) • Linear relationship between Log (Outgassing Conc.) and inverse of outgassing temperature (1/T). • In agreement with findings reported by K. Takeda, et al., IEST 2000 ESTECH Proceedings, pp. 71-77 (2000). • Same result was observed for Butylated hydroxytoluene (BHT) (not shown).

3/25/03 P. Sun/C. Ayre, CA MATTEC, Intel 21 Outgassing Quantity vs. Surface Area: Siloxanes

Siloxane Outgassing Dependence on Sample Surface Area (outgassing condition: 50oC, 15 min) 700 Outgassing ∝ Surface Area 600

500

400

300

200 Total Outgassing (ng) 100

0 Octamethyl- Decamethyl- Tetradecamethyl- cyclotetrasiloxane cyclopentasiloxane cycloheptasiloxane 1x surface area, 1x sample weight Outgassing Compound 2x surface area, 1x sample weight

• Outgassing quantity ∝ surface area when sample weight is constant. 1x surface area 2x surface area 1x weight 1x weight

3/25/03 P. Sun/C. Ayre, CA MATTEC, Intel 22 Outgassing Quantity vs. Sample Surface Area

Outgassing Dependence on Sample Surface Area (outgassing condition: 50oC, 15 min) 600 Outgassing ∝ Surface Area 500

400

300

200 Total Outgassing (ng) 100

0 2-ethyl-1-hexanol Methylstyrene Naphthalene

Outgassing Compound 1x surface area, 1x sample weight 2x surface area, 1x sample weight

• Outgassing quantity ∝ surface area when sample weight is constant. 1x surface area 2x surface area 1x weight 1x weight

3/25/03 P. Sun/C. Ayre, CA MATTEC, Intel 23 Outgassing Quantity vs. Sample Weight

Outgassing Dependence on Sample Weight (50oC, 15min) 600 Outgassing is independent 500 of sample weight 400

300

200

Total Outgassing (ng) 100

0 Octamethyl- Decamethyl- Tetradecamethyl- cyclotetrasiloxane cyclopentasiloxane cycloheptasiloxane Outgassing Compound 1x sample weight, 1x surface area 2x sample weight, 1x surface area

• Outgassing is independent of sample weight when surface area is constant. • Same results for methylstyrene, 2-ethyl-1-hexanol and naphthalene (not shown).

3/25/03 P. Sun/C. Ayre, CA MATTEC, Intel 24 Conclusions: Characterization of Outgassed Organic Contaminants Using TD-GC-MS

• Linear relationship between outgassing concentration and outgassing time in a certain time range. – where the slope equals the outgassing rate. • Linear relationship between logarithm of outgassing concentration and 1/T; where ‘T’ is the absolute outgassing temperature. • Outgassing amount ∝ sample surface area. • Outgassing is independent of sample weight when surface area remains constant. – Suggests that outgassing is indeed a surface phenomenon.

Note: These findings are based on the limited number of contaminants tested. – Materials’ outgassing can be far more complicated due to large number of variables in the physical and chemical properties of different materials used in cleanrooms.

3/25/03 P. Sun/C. Ayre, CA MATTEC, Intel 25 Outline

3/25/03 P. Sun/C. Ayre, CA MATTEC, Intel 26 Ammonia/Amines’ Outgassing • T-topping defect caused by NH3/amines’ Outgassing:

T-topping SEM image

• Airborne bases outgassed from cleanroom materials can neutralize the photo-generated acid during the time delay between the exposure and post-exposure bake, generating insoluble products, which cannot be dissolved by developer solvent. A lip forms at the top of the developed resist profile, known as “T-topping”.

3/25/03 P. Sun/C. Ayre, CA MATTEC, Intel 27 Ammonia/Amines’ Outgassing • Gas Diffusion-Conductivity Detection-based NH3/Amines’ Outgassing Method:

Capable of detecting 1 ng/g (ppb) of NH3/amines’ outgassed

ppb = parts per billion

3/25/03 P. Sun/C. Ayre, CA MATTEC, Intel 28 NH3/Amines Outgassing of Two Cleanroom Sealants

Sealant A caused significant T- topping defects in E0 delay photoresist test due to its high NH3/amine outgassing.

3/25/03 P. Sun/C. Ayre, CA MATTEC, Intel 29 Time Dependence of NH3/Amines’ Outgassing

NH3/Amine Outgassing vs. Cure Time

g 800 n i 700 s

s Time dependent NH3/amine

a 600 g t

u 500

g) outgassing of a sealant: room 400 e O ng/ ( n

i 300 temperature

m 200 A

3/ 100 H

N 0 0 1020304050607080 Cure Time (hour)

Wafer Surface NH4+ vs. Cure Time (1 hour exposure at room temperature) 3000 Surface NH4+ of witness 2500 + wafers exposed to same ) 4 2 H

m 2000

c sealant: room temperature / e N ac 4+ 1500 f r H u N 1000 S 10 er E f

( control wafer

a 500 W 0 0 1020304050607080 Cure Time (hour)

3/25/03 P. Sun/C. Ayre, CA MATTEC, Intel 30 NH3/Amines’ Outgassing vs. Wafer Surface NH4+

Wafer Exposure Test (1 hour exposure at room temperature) 3000 Linear relationship between witness wafer surface NH4+ 2500 and sealant NH3/amines’

2000 outgassing concentration

1500

1000 (E10 NH4+/cm2) Wafer Surface NH4+ 500

0 0 100 200 300 400 500 600 700 800 NH3 Outgassing (ng/g)

3/25/03 P. Sun/C. Ayre, CA MATTEC, Intel 31 Summary

• Many AMC contaminants found in semiconductor Fabs come from outgassing of cleanroom materials. Outgassed contaminants can adversely affect many processes, resulting in yield loss, shortened tool life and reduced long-term device reliability.

• The large variety of cleanroom materials and numerous outgassing contaminants combined with the complexity of process steps makes understanding detrimental levels of specific contaminants in particular processes very challenging.

• Screening materials for both condensable and NH3/amines’ outgassing prior to bringing them into Fabs can be used as a first- line-of-defense against molecular contaminants such as organophosphorus, siloxanes, plasticizers and ammonia/amines.

3/25/03 P. Sun/C. Ayre, CA MATTEC, Intel 32 Summary

• New microelectronic fabrication processes with decreasing device geometries are increasingly susceptible to AMC.

• As processes and chemistries change, requirements for monitoring, control and analysis of materials’ outgassing will evolve as well.

• Cooperative efforts among manufacturers of integrated circuit, materials and analytical tools are needed to better understand the impact of molecular contaminants and to properly define specifications applicable to new ULSI technologies.

3/25/03 P. Sun/C. Ayre, CA MATTEC, Intel 33 Acknowledgements

• Yaacov Maoz of Intel Fab 18 for providing the “T- topping” defect SEM image.

• Zari Pourmotamed of Intel California Materials Technology for collecting time-dependent NH3/amines’ outgassing data.

• Joseph O’Sullivan of Intel Facilities Technology for providing valuable technical inputs.

3/25/03 P. Sun/C. Ayre, CA MATTEC, Intel 34