Presentation on theme: "KBG Ken Gilleo - ET-Trends David Blumel Alpha Metals with Reworkable Wafer-Level Underfill"— Presentation transcript:
KBG Ken Gilleo - ET-Trends David Blumel Alpha Metals with Reworkable Wafer-Level Underfill
KBG Outline The Packaging Revolution Flip Chip vs. CSP Why Underfill? Classes of Underfill Final Generation FC; a CSP Conclusions
smallersmaller faster cheaper
KBG Is the KEY to achieving:
Smaller The whose time already came! Faster = more leads Cant be solved by packaging evolution
KBG n 2 > 4n-4
KBG Flip Chip Strip Spider Chip-on-Board BGA Lead Frame TCP (TAB) (re-Engineered) 1960's Hot for the 21st Century SMT Flip Chip (C4) micro-SMT TBGA SMT Feed-thru Beam Lead Chip FC-BGA CSP/FC-BGA The Entire Packaging History Flip Chip BGA
KBG Perimeter Leads Area Array Size: chip scale Packaging: minimal Packaging: post- and concurrent Paths/lead length: shorter SMT 3 COB & TAB 2 DCA IC 1 ULTIMATE
KBG Is Flip Chip a True PACKAGE? Is Flip Chip a True PACKAGE?
KBG the Removability Translation: IC to PCB Standardization Environmental Protection
KBG IBM The Original Flip Chip was a CSP
KBG FLIP CHIP 360 o REVOLUTION s CSP High lead, ceramic substrate New bumps, organic substrate + Underfill CSP again
KBG metal vapor Bumping Methods Attach discrete spheres; Au, Cu, Sn/Pb Print joining mat's; Sn/Pb or Conductive Adhesive Vacuum deposit metal: old, still alive Electrolytic plating; Au, Cu, Sn/Pb, Ni (cost issue?) Electroless plating; Au, Cu, Ni (NEWER) Fluid jet molten metal; Sn/Pb (VERY NEW) Stud bump with; Sn/Pb, CU or Au (single chip) Material transfer; Sn/Pb or Cond. Adhes.; paste or film
KBG Switch to organic substrate –Causes large thermal mismatch –Low reliability in thermocycle Mismatch must be addressed –low CTE organic substrate –columns instead of bumps –non-fatiguing joints??? –mechanical coupling: chip-to-substrate
KBG H e a t i n gC o o l i n g Thermal Mismatch Kills Reliability CHIP Sn/Pb
KBG UNDERFILL Mechanism Y p o s i t i o n constrained Y p o s i t i o n
KBG Underfill: What You A real aggravation Added equipment Added floor space Added cost Reduced yield
KBG Underfill: What You Self-Dispensing Self-Fluxing No added equipment No added time required Cost-effective Reworkable
KBG Underfill Events Underfill Effect discovered: 1960s Slow flow, slow cure the norm: early 1990s Fast flow (>2.5cm/min.), 30 min. cure: 1995 Pre-dispense flux-fill R&D: mid-1990s Snap flow (>3 cm/min) /Snap cure (5 min.): 1997 Convert FC to SMT: Wafer-level: coming in
KBG Types of Underfill SUBSTRATE Chip/Wafer APPLIED to LiquidAvailable Solid PHASE Pre- Dispensed Post- Dispensed Chip & Substrate Concurrently Liquid Solid Liquid Solid Available NA ? R & DNA
KBG Capillary Type (post dispensed) Flow rate is close to max. Cure time is close to min. Still adds –equipment –space –time –cost Result: FC = SMT
Pre-Dispensed Liquid Process control is critical Requires dispenser/printer Solder reflow oven provides cure Enables FC = SMT Result: next generation underfill
KBG Not Assembled Pre-Assembled Pre-Dispensed Post-Dispensed Underfill Flux/Underfill
KBG Pre-Dispense Solid on Substrate Film-on-PCB –Special, expensive equipment –Not an SMT process –Doesnt address underfill problems An old concept?
KBG Anisotropic Conductive Adhesive ACA film has a built-in underfill and is the 1st example of pre-dispensed solid underfill.
KBG Wafer-level applied Self-fluxing Dry solid Integral to Flip Chip True SMT process Transparent to assembler Can be reworkable Pre-Dispense Solid on Chip
KBG Liquid polymer-based composition is coated onto Flip Chips at wafer-level and then converted to a SOLID that: (1) Permits a bumped wafer to be diced into Flip Chips. (2) Provides flux for assembly. (3) Liquefies to a thermoplastic underfill during reflow. (4) Polymerizes and wets substrate during reflow step. (5) Remains reworkable after reflow stage cure.
KBG Ramifications Ramifications : FC becomes a std. SMT process. FC becomes CSP if reworkable. Underfill becomes a semiconductor process. The ready-to-bond FC becomes the most cost- effective minimal package. Success can make this package the dominant micropackage.
KBG Assembly Process Pick & Place FC from any format Reflow flux melts/activates underfill liquefies/wets solder melts/forms joint underfill solidifies Test Rework if required
KBG TIME in Solder Reflow Oven TEMPTEMP Melts; flux activates, begins to bond to substrate Solder joints form, underfill properties generated Flux has deactivated, material is now an underfill
KBG Issues & Challenges Materials; single or multiple? Shelf life, what is required? What wafer Coating process? Dicing with polymer in place? Assembly –voiding, filleting, adhesion –process sensitivity
KBG Solid Flux FLIP CHIP INTEGRTATED/FLUXFILL Type 1 - Single material converts from flux to underfill during reflow
KBG FLIP CHIP INTEGTRATED FLUX/UNDERFILL Solid Flux Solid Thermoplastic Underfill Type 2 - Two separate materials Many variations
KBG Status Technology 2-LAYER1-LAYER MATERIALS WAFER COATING DICING FC ASSEMBLY RELIABILITY completebeing optimized selection stagebeing optimized Feasibility confirmed to be determined confirmed to be determined to be determined to be determined
KBG Transparent 12 mm x 12 mm Flip Chip Bonded to Copper with single-layer Flux/Underfill by running through an IR reflow oven at 220 o C Delco is not a sponsor or participant Copper sheet Flux/underfill after heating Purchased quartz FC with Sn/Pb bumps Phase 1 Test Platform
KBG Conclusions Todays underfills impede FC FC = SMT: required for max. success Underfill can be a semicon process FC will become a CSP again Result: best micropackage solution
KBG The Ultimate Micro Package Everything should be made as simple as possible but not simpler. Albert Einstein Just add heat; some assembly required.