250 likes | 467 Views
Accelerated Testing of Flip Chip Underfill - Sensitivity to Aging/Moisture. Nikhil Vichare Graduate Research Associate Department of Systems Science and Industrial Engineering State University of New York at Binghamton Binghamton NY 13902. Experiment.
E N D
Accelerated Testing of Flip Chip Underfill - Sensitivity to Aging/Moisture Nikhil Vichare Graduate Research Associate Department of Systems Science and Industrial Engineering State University of New York at Binghamton Binghamton NY 13902
Experiment Performance in cycling is relative measure of degradation in preconditioning • Preconditioning Time, Temperature, Humidity • 72 hours - 6 months • Room temperature - 130oC • Dry N2 - 85%R.H. • No Preconditioning Subsequent Cycling in Liquid to Liquid Thermal Shock (LLTS) • 5 minutes hold at -55oC and 125oC • 10 seconds transfer time Measure corner fillet cracking and delamination around bumps v/s number of LLTS cycles
Experiment • 1056 die 4224 corner fillets • 3 underfills • Dexter FP4549 • Namics U8437-3 • Honeywell JM8802 • 2 fluxes • Kester TSF-6522 • Indium NC-SMQ 75 PB08 chips on 62 mil board with Taiyo PSR4000 mask
Effects of Moisture • It is well “established” that moisture exposure may weaken the underfill. • Pressure Cooker Test (PCT) is a common accelerated test of sensitivity to humidity • Also called unbiased HAST • 85% R.H.,130°C, 2 atm.pressure (JESD22-A118) • Is PCT meaningful?
30 Crack No Crack No Preconditioning Direct LLTS 25 20 144 Hours PCT + LLTS Number of Fillets 72 Hours PCT + LLTS 15 10 5 0 250 500 250 500 250 500 1000 1500 2000 2500 1000 1500 2000 2500 1000 1500 2000 2500 LLTS Cycles Fillet Cracking in Cycling after PCT • Clear effect on subsequent cycling 14-16 mil Thick Namics U8437-3 Fillets (Kester TSF-6522 Flux)
PCT? • Characteristics of damage in cycling after PCT different from • that after other humidity/temperature exposure PCT 72 Hours + LLTS 85°C/85%R.H. 3 Weeks + LLTS • Also discoloration of underfill and solder mask (see poster)
Time/Temperature and/or Moisture has an Effect ! • Faster Cracking of Namics (TSF-6522) after 85°C/30% R.H. 14-16 mil Thick Namics U8437-3 Fillets (Kester TSF-6522)
Crack No Crack 85°C/30%R.H. 30 85°C/85%R.H. 20 Number of Fillets 10 0 250 500 250 500 1000 1500 2000 2500 1000 1500 2000 2500 LLTS Cycles More Moisture is Worse 2 Weeks of Preconditioning of 14-16 mil Thick Namics U8437-3 Fillets (Kester TSF-6522)
Also Faster Delamination from Solder Bumps after Moisture Exposure • After 3 Weeks of 85°C/85% R.H. 250 1000 2000 • Without Preconditioning 1000 2000 3000 Namics U8437-3 (Kester TSF-6522)
Acceleration? • Similar cracking in subsequent cycling for some encapsulant/ flux/ fillet thickness combinations • Example • 3 months aging in ambient 85°C/30%R.H. for 3 weeks • 6 months aging in ambient 85°C/85%R.H. for 3 weeks
Acceleration? • 3 months ambient 85°C/30%R.H. for 3 weeks • Namics(Kester) 14-16 mil fillets
Acceleration? • 3 months ambient 85°C/30%R.H. for 3 weeks • Same material thinner fillets (10-12 mils)
Acceleration? • 3 months ambient 85°C/30%R.H. for 3 weeks • Dexter(Indium) 10-12 mil fillets
Crack No Crack 30 6 Months Ambient 85°C/85%R.H. 3 weeks 20 Number of Fillets 10 0 500 500 1000 1500 2000 2500 1000 1500 2000 2500 LLTS Cycles Acceleration? • 6 months aging in ambient 85oC/85%R.H. for 3 weeks Namics U8437-3 (Kester TSF-6522) 14-16 mils
Acceleration? • 6 months ambient 85°C/85%R.H. for 3 weeks • and 3 months ambient 85°C/30%R.H. for 3 weeks (totally different cracking in subsequent cycling) for other encapsulant/ flux/ fillet thickness combinations
30 No Crack Crack 20 3 Months Ambient Number of Fillets 85°C-30% R.H. 3 week 10 0 500 500 1000 1500 2000 2500 1000 1500 2000 2500 LLTS Cycles Acceleration? • 3 months ambient 85°C/30%R.H. for 3 weeks • Honeywell (Kester) 12-14 mil fillets Honeywell (Kester) 12-14 mils
NoCrack 30 Crack 85°C-85% R.H. 3 week 20 6 Months Ambient Number of Fillets 10 0 500 500 1000 1500 2000 2500 1000 1500 2000 2500 Acceleration? • 6 months ambient 85°C/85%R.H. for 3 weeks • Dexter (Indium) 10-12 mil Thick Fillets LLTS Cycles Dexter FP4549 (Indium NC-SMQ 75) 10-12 mils
Moisture isn't Everything! • Underfill also weakens with time (even in dry storage) • That’s why we didn't have a single acceleration
Effect of Time As Well • Aging in Nitrogen Cabinet 10-12 mil Thick Namics U8437-3 (Kester TSF-6522)
Effect of Time As Well • Faster Delamination from Solder Bumps • After 3 Months Aging in Nitrogen Cabinet 1000 2000 3000 • Without Preconditioning 1000 2000 3000 Namics U8437-3 (Indium NC-SMQ 75)
Preconditioning Sometimes Decreased Cracking • 85°C/30%R.H. reduces subsequent cracking of Honeywell (Kester) • Effect of time/temperature or moisture? No Preconditioning 10-12 mil Thick Honeywell JM88202 (Kester TSF-6522)
Crack No Crack 30 85°C/30%R.H. 2 Weeks 85°C/85%R.H. 85°C/85%R.H. 20 3 Weeks 2 Weeks 85°C/30%R.H. Number of Fillets 3 Weeks 10 0 250 500 250 500 250 500 250 500 1000 1500 2000 2500 1000 1500 2000 2500 1000 1500 2000 2500 1000 1500 2000 2500 LLTS Cycles More Moisture Increased Cracking • At same temperature / time more moisture (85°C/85%R.H.) is worse in this case 10-12 mil Thick Honeywell JM88202 (Kester TSF-6522)
Moisture Not Always Bad • A little moisture seems to help (same time /temperature) • Dexter (Indium) 8-10 mil fillets
SUMMARY • Moisture and time/temperature each affect underfill properties • Time/temperature may not always be bad (depends on encapsulant) • Sometimes a little moisture reduces subsequent cracking • Relative contribution of time/temperature Vs moisture depends on encapsulant/ flux combination • Design/choice of accelerated test must account for this