Fraunhofer IZM Workshop November 25, 2002 Thin Semiconductor Devices
Effect of Wafer-Thinning Processes On Ultra-Thin Wafer and Die Strength Tony Schraub Ph.D.
A Cooperative Program between Three US Companies Strasbaugh Grinding & Polishing Tru-Si Technologies Dry Plasma Etching Saint-Gobain (Norton) Abrasive grind wheels Die Strength Testing
Thin Die Strength Dependency Stress-concentrating flaws due to sub-surface damage by grind forces on abrasive grits Damage removal method and amount Final thickness of each process step Edge condition (die and wafer edge chips) Handling stress during and between process steps Die size, shape and grind line orientation Brittle fracture test method This program addresses only items in yellow
Variables tested in this Program Amounts Removed: (all wafers final thickness 100 um) Coarse Grind ~625um Fine Grind 20 um Grind + No polish or Etch Grind +Polish 5 um Grind+ Polish 10 um Grind + Etch 5 um Grind + Etch 10 um Grind + Etch 25 um 350 mesh 350 mesh # 2000 YES none none YES YES None # 8000 YES YES YES YES YES YES
Previously Presented Die Strength Vs. Silicon Removal W. Kroninger, Infineon Technologies Grind and 10 µm Dry Etch Grind and 10 µm Wet Etch Rel. Die Strength in N Grind Only DBG Grind and 3 µm Polish Silicon Removal Method Source: IZM Conference 2001
Test Result Summary Die Strength 2000 Grit Grind 8000 Grit Grind Grind & Polish Grind & ADP Etch
How Was Test Done? Bare 200mm (1,0,0) Wafers Thinned to 100µm 1. Grind Only 2. Grind and Polish 3. Grind and Dry Plasma Etch (ADP) Dice Wafers into 32 Die 25mm X 25mm Measure all 32 Individual Die Fracture Loads 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
Wafer back grinder Strasbaugh 7AF (ntellect) Key Strasbaugh Technologies: In-Situ Thickness Measurement Linear Traversing Grind Spindles Innovative Abrasive grind wheels Force Adaptive Grinding
Saint Gobain FAVSTM grinding wheels Key Norton Technologies: Range of specifications engineered to suit applications In depth R&D and application support to develop best processes
Tru-Si Atmospheric Downstream Plasma (ADP) Key Tru-Si Technologies: Atmospheric Dry Chemical Etch Thinning & Damage Removal Thru-Silicon Connections NoTouch Handling
Die Fracture Load Method Biaxial flexture test on 25mm by 25mm die Pin-on-ring fixture on MTS test machine 12mm steel ball onto 16mm Diameter ring Ground side of die put in tension Tape (compressive side) holds fragments Load rate at 0.5mm/minute
MTS Material Test Tool 0.625 (15.9 mm) MOVEABLE CROSSHEAD M T S F R A M E RING FIXTURE PIN FIXTURE WAFER M T S F R A M E LOAD CELL MTS FRAME
Strengthen Wafer/Die by Removing Grinding Damage Sub surface damage Surface damage
Remove Subsurface Damage to Increase Strength After Grinding After ADP Etching
Examples of Die Fractures Fractured at High Load Fractured at Low Load Fractured at Medium Load
Weibull Curve Test Results Weibull Comparison of Grind, Polish, & ADP Etch Probability LN(LN(1-p)) 2 1 0-1 -2-3 -4-5 1 2 3 4 5 10 20 30 3540 Fracture Load (lbs) -1 0 1 2 3 4 5 LN (Fracture Load [lbs]) 2000 Grind 8000 Grind 5um Polish 10um Polish 5um ADP 10um ADP
Mean Fracture Loads (Pounds) 16.0 14.0 12.0 10.0 8.0 6.0 4.0 2.0 40 % improved die fracture load using #8000 without any treatment 0.8 1.1 0.0 2000 Grind 8000 Grind 5um Polish 10um Polish 5um ADP 10um ADP
Raw Test Results Die Strength 40.00 35.00 Die Fracture Load in Pounds 30.00 25.00 20.00 15.00 10.00 8000 Grind 5um Polish 5um ADP 5.00 0.00 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 Die Strength Sorted from Lowest to Highest
Comparisons of Min, Mean, and Max Min, Mean, & Max Comparisons 40.0 37.9 35.0 30.0 25.0 20.4 Die Fracture Load in Pounds 20.0 15.0 10.0 5.0 0.0 8000 Grind 1.1 1.8 0.6 1.8 5um Polish 8.6 2.4 5um ADP 14.1 Min Mean Max
Strength Comparison with Different Amounts of Etch Strength Equivalent after 5um ADP Etch Probability LN(LN(1-p)) 2 1 0-1 -2-3 -4 1 2 3 4 5 10 20 30 35 40 Fracture Load (lbs) 5um ADP 10um ADP 25um ADP -5 0 1 2 3 4 5 LN (Fracture Load [lbs])
Fracture Load vs.thickness No correlation of die strength with die thickness within the thickness range observed within the wafer 5 micron polished 10 micron ADP dry etch 25 40 Fracture Load (lbs) 20 15 10 5 Fracture Load (lbs) 35 30 25 20 15 10 5 0 98 100 102 104 106 Thickness (µm) 0 10 0 10 1 10 2 10 3 10 4 10 5 Thickness (µm)
#2000 Grit vs #8000 Grit with ADP: Mean Fracture Load (Lbs) 16.0 14.0 25 % improvement 14.1 10 % improvement 12.8 14.0 12.0 11.2 10.0 8.0 6.0 Etching 5 um after grinding with the #8000 gives better strength than etching 10 um after grinding with the #2000 4.0 2.0 0.0 2000 Grit 5um ADP Etch 8000 Grit 5um ADP Etch 2000 Grit 10um ADP Etch 8000 Grit 10um ADP Etch
Conclusions 8000 Grit Grind better than 2000 Grit Grind. The strength of the dies ground with the #8000 is better than the ones ground with #2000 Polishing or ADP Etching after grinding significantly strengthen die ADP Etching after #8000 fine grind produced the Strongest Die
Future Direction for Ultra thin Examine Wafer Level Strength For example: Effect of rounded edges from ADP Etching on wafer strength
Contact Information Strasbaugh www.strasbaugh.com Tru-Si www.trusi.com Saint-Gobain Abrasives www.saint-gobain.com Die Strength 2000 Grit Grind 8000 Grit Grind Grind & Polish Grind & ADP Etch
Additional Information Acknowledgements Test data summary Other suggestions for future work Equipment descriptions 8000 Grind wheel information
Test Data Summary Test Case Test Case Description Die Thickness (µm) Die Fracture Load (lbs) Fine Wheel Polish microns removed ADP Etch microns removed Min Mean Max Range Stdev Min Mean Max Range Stdev 1A 8000 101.2 102.1 102.6 1.4 0.3 0.6 1.1 1.8 1.2 0.3 4A 8000 10 99.0 102.1 103.9 4.9 1.4 2.3 8.1 23.5 21.2 4.2 3A 8000 5 99.8 103.0 105.3 5.5 1.7 1.8 8.6 20.4 18.6 3.8 5A 8000 0 5 101.0 102.7 105.1 4.1 1.0 2.4 14.1 37.9 35.5 10.9 7A 8000 25 98.0 101.2 105.2 7.2 1.6 2.2 15.0 30.2 28.0 8.4 8B 8000 10 101.0 102.7 104.1 3.1 0.9 2.9 14.0 34.2 31.4 10.1 11A 2000 5 100.0 101.1 102.4 2.4 0.7 3.6 11.2 24.7 21.1 6.3 12A 2000 10 100.6 102.3 103.7 3.1 0.9 2.5 12.8 43.9 41.4 9.8 12B 2000 10 101.2 102.5 103.8 2.6 0.7 1.5 10.8 26.9 25.5 8.2 13B 2000 96.9 98.7 100.2 3.3 0.9 0.5 0.8 1.2 0.7 0.2
Large increases in die strength with enhanced surface treatments Percentage Increase in Strength Over 2000 Grit Grind 3500% 3034% 3000% 3034% 2500% 2000% 1582% 1650% 1500% 1582% 964% 1000% 49% 31% 500% 0% 49% Max Mean 8000 Grind 5um Polish 5um ADP Min
Other Suggestions for Future Work Ultra-thin-wafer Dicing method influence Data for thinner wafers below 100 µm thickness Optimize grind recipes for new #8000 wheel Optimize each process removal amounts Smaller die; Rectangular die Whole wafer strength evaluations, including edge effects Polish after etching and/or etching after polishing Study great increase in range of die strength after polish or etch 2.5 µm polish removal data (to complement 5 and 10 µm)
Acknowledgements This work is a cooperative effort of these companies & individuals: Tony Schraub, Ph.D Consultant on thin wafers Sal Kassir Grinding R&D Engineer, Strasbaugh Bryan Vogt Polishing Process Engineer, Strasbaugh Sam Kao, Ph.D R&D Applications, Tru-Si Technologies Frank Kretz,Manager--Atmospheric Downstream Plasma Equipment, Tru-Si Technologies Jeri Ikeda, Ph.D R&D, Saint-Gobain Abrasives Rama Vedantham Saint-Gobain Abrasives Thomas Puthanangady Saint-Gobain Abrasives
Tru-Si Enables 50 Micron and Below Etching Wafer Etching Equipment Dry Chemical Etch Atmospheric Operation Applications: Grind Damage Removal Rough Wafer Backsides Damage Free Dicing Selective Film Etching Thin Wafer Handling Equipment NoTouch end Effectors and Wands for thin bowed, bumped wafers
#8000 Fine Grinding Wheel Designed and manufactured by Saint-Gobain Abrasives Supplied by Strasbaugh as part number #301093 Ultra fine grind wheel with 1-2 micron diamond grits Produces stronger wafers with less bow than <2000 wheels Improved surface finish to 10-30 Angstroms