Understanding Whisker Phenomenon: Driving Force for Whisker Formation Chen Xu, Yun Zhang*, C. Fan and J. Abys * Part of this work was carried out at Lucent 1 Enthone Confidential Outline • Whisker Background • Whisker Program – Phase I, whisker test – Phase II, whisker growth rate and whisker growth mechanisms • Conclusions and Recommendations 2 Enthone Confidential FINISH ADVANTAGES CONCERNS/ISSUES Sn Most cost effective 232 °C Compatible with Whisker formation soldering alloy Simple process control Sn(2-5)Bi Compatible with Whisker formation 220 – 225 °C soldering alloy Immersion plating Melting point lower Alloy control & measurement Sn3.5Ag Compatible with Whisker formation 221 °C soldering alloy Very narrow process window Melting point lower Cost increase Environmental issue with Ag Sn(0.7-1.5)Cu Compatible with Whisker formation 227 - 270 °C soldering alloy Immersion plating Melting point lower Accelerated Sn(II) oxidation 3 Enthone Confidential Lead-Free Implementation Barrier Spontaneous Whisker Growth! 4 Enthone Confidential Tin Whiskers Filament Nodule 120 µm 15 µm Column Mound 9 µm 2 µm 5 Enthone Confidential Whisker Program Understanding the Whisker Growth Phenomenon • Previous Results - Phase I – Established a reliable whisker test • 50°C, dry heat • Office Condition • Thermal Cycling – Investigated the relative importance of several factors • Grain size & shape, carbon content • External mechanical bending • Temperature & humidity • Current Work - Phase II – Whisker growth rate – Whisker growth mechanisms 6 Enthone Confidential Phase II & III Test Matrix SnO or “Cap” Au, Pd & Ag Flash Sn/Sn Alloy Sn Bright, matte and Satin Bright Tin IMC SnNi35SnCu46 Ni Underlayer Ni MLS Ni Substrate Cu Parameters Tested Effect of “Cap” Type of Sn Ni Underlayer External Stress Reflow Group I Group II Temperature 90° Tensile or Group III Compressive Bend Reflow Whisker Test 7 Enthone Confidential Method of SEM Observation 2 mm Bent area, stressed Flat area, not stressed 1 inch 2.5 inches Area of examination: 0.5 mm x 27.4 mm 8 Four magnifications: 500X, 1K, 5K & 10K Enthone Confidential Maximum Length Measurement Bright Tin 6 months 3 months 600 µm 30 µm 800 700 600 500 400 300 200 Max. length, microns 100 0 9 0 4 8 12 Months Enthone Confidential Maximum Length Measurement Satin Bright Tin 6 months 3 months 5 µm 4 µm 20 15 Phase II 10 Phase I 5 Max. length, microns 10 0 0 3 6 10 12 Months Enthone Confidential Whisker Index Measure for Whisker Propensity WI = S n*d*L*f(L) n: number L: length d: diameter f: weight factor on length, L 11 Enthone Confidential Whisker Index Relationship between factor f and whisker length WI = S n*d*L*f(L) Length, mm Weight Factor, f(L) < 1 0 1 – 5 1 5 – 10 5 10 – 50 50 > 50 500 12 Enthone Confidential Categorization in Measurement Bright Tin 13 Enthone Confidential Categorization in Measurement Satin Bright Tin Total number: 5 14 Enthone Confidential Whisker Index |--------------Whisker Index-------------| FINISH 4 MONTHS 6 MONTHS 10 MONTHS LONGEST* Bright 279 13,000 63,400 600 Bright + Tensile 244 2,800 45,200 350 Bright + 3850 13,500 193,000 750 Compressive Satin Bright 3.2 10.5 10 5 Satin Bright + 30 22 40 5 Compressive Satin Bright 4.1 8.6 907 10 90/10 Satin Bright / 2.1 4.0 8.9 4 Reflow Satin Bright / Ni 0.2 0 0 3 * Longest whisker (µm) observed up to 10 months 15 Enthone Confidential Whisker Propensity of Different Finishes Effect of Grain Size & Carbon Content Bright Vs. Satin Bright 80000 C = 0.2% GS = 0.2 µm 60000 Bright 40000 20000 Whisker Index Satin Bright C = 0.04% 0 GS = 5 µm 2 6 10 Aging Time, Months 16 Enthone Confidential Whisker Propensity of Different Finishes Effect of 90° Compressive Mechanical Bending Bright Vs. Satin Bright 250000 Bright, bend 200000 150000 100000 Bright, no bend 50000 Whisker Index Satin Bright, w & w/o bend 0 2 6 10 Aging Time, Months 17 Enthone Confidential Effect of External Mechanical Stress Tensile Vs. Compressive Bright Tin 200000 Compressive 150000 100000 No Bend 50000 Whisker Index Tensile 0 2 6 10 Aging Time, Months 18 Enthone Confidential Whisker Propensity of Satin Bright Tin Effect of Reflow & Ni Underlayer 68000 58000 48000 38000 28000 18000 Whisker Index 8000 -2000 2 4 6 8 10 12 12 Aging Time, Months 10 B SB SB / Reflow SB / Ni 8 6 4 Whisker Index 2 0 2 6 10 Aging Time, Months SB SB / Reflow SB / Ni 19 Enthone Confidential Understanding the Mechanisms Driving Force Whisker = f (A, B, C, D, E, F, G……..) Grain Size Residual stress External Stress Organic Inclusion Hydrogen Inclusion Grain Shape and Orientation Temperature and Humidity 20 Enthone Confidential Driving Force vs Factors Affecting Whisker Growth • Driving Force: Thermodynamic DG=DH-TDS • Other Factors: Kinetic 21 Enthone Confidential Presentation Outline • What Is Known about the Mechanisms? • Our Experimental Results: Ö Local Structure by FIB, SEM Ö Residual Stress by XRD • Conclusion 22 Enthone Confidential Mechanisms of Whisker Growth Whisker SnOx Sn CuSnx Cu 23 Enthone Confidential Mechanism of Spontaneous Whisker Growth To explain the observed phenomena, two models were developed • Dislocation Model by Eshelby and Frank • Recrystallization Model by Ellis 24 Enthone Confidential Recrystallization and Grain Growth Internal Energy Reduction grain Driving Force: Residual stress (compressive and tensile) Grain size Grain shape, orientation and distribution Organic and hydrogen inclusion 25 Enthone Confidential Open Questions •Does whisker grow in the x,y-direction ? ? ? release Compressive stress Lateral growth release Tensile stress release Compressive stress Vertical growth increase Tensile stress 26 Enthone Confidential Focused Ion Beam (FIB) Ga+ Ion Beam Detector 5-200 nm 27 Enthone Confidential Cutting Through a Whisker Sn Cu6Sn5 Cu Whisker 28 Enthone Confidential Bright Sn/Cu, aged at RT for 18 months Whisker Sn Cu Whisker ~400 mm Grain Boundary 29 Enthone Confidential SB Sn/Cu, aged at RT for 18 months Whisker Sn Cu6Sn5 Cu 30 Enthone Confidential FIB x x Tensile stress hinders whisker growth 31 Enthone Confidential Compressive vs Tensile Stress Compressive Stress: There will be excessive material in a given volume Tensile Stress: There will be material deficiency in a given volume Tensile stress hinders whisker growth 32 Enthone Confidential Driving Force vs Factors Affecting Whisker Growth • Driving Force: Thermodynamic Compressive Stress : Whisker Tensile Stress : No Whisker • Other Factors: Kinetic Grain Structure, Aging Condition, etc. 33 Enthone Confidential Open Questions •What does tensile stress do ? ? ? •How to measure and control the stress in electroplated Sn? 34 Enthone Confidential Motivation (cont.) Why using XRD for the residual stress measurement? Detector X-ray Non-destructive and spatial resolved Monitoring the stress evolution Stress measurement on real part Measuring the stress of individual layer 35 Enthone Confidential Stress Measurement using XRD u+1 u e (y) = s sin2y - (s + s ) x E x E x y Tensile Stress Strain Compressive Stress sin2y 36 Enthone Confidential Shear Stress? 0.0002 z 0.0000 y y -0.0002 x -0.0004 -0.0006 -0.0008 Strain -0.0010 -0.0012 +y -0.0014 -0.0016 -y 0.0 0.1 0.2 0.3 0.4 0.5 0.6 37 2 sin y Enthone Confidential 3 Months at RT 38 Enthone Confidential Stress Evolution in Sn Film Plated Directly Over Cu Stress in MPa As Plated 4 Months Bright -3±1 -10±1 SB Sn -1±1 -7±1 39 Enthone Confidential Stop Cu-Diffusion into Sn Sn CuSn Ni x Cu 40 Enthone Confidential Ni Underlayer on Stress Samples aged at RT for 4 months Stress in MPa Sn/Cu Bright -10±1 SB Sn -7±1 41 Enthone Confidential Ni Underlayer on Stress Samples aged at RT for 4 months Stress in MPa Sn/Cu Sn/Ni/Cu Bright -10±1 9±1 SB Sn -7±1 7±0 42 Enthone Confidential Effect of Ni-Underlayer 100 100 Bright Sn/Ni/Cu SB Sn/Ni/Cu Bright Sn/Cu SB Sn/Cu 50 50 Whisker Index 0 0 0 1 2 3 4 5 6 0 1 2 3 4 5 6 Aging Time (Months) Aging Time (Months) 43 Enthone Confidential Intermetallic Formation Sn-Cu vs Sn-Ni Sn Sn CuSn x SnNix Ni Cu Cu 44 Enthone Confidential Minimizing Whisker Formation Using Thick Sn Coating (8~12 m) 45 Enthone Confidential Thick Sn Coating (8~10 m) No stress Compressive stress Sn CuSnx Cu 46 Enthone Confidential Thick Sn Coating (8~10 m) Compressive stress Sn CuSnx Cu 47 Enthone Confidential Whisker Reduction Using Reflow 48 Enthone Confidential Effect of Reflow Stress after aging at RT for 12 months Satin bright Sn (MPa) Sn as plated -8±3 49 Enthone Confidential Effect of Reflow Reflow: Right After Plating Stress Measurements: 12 Months Aging at RT Sn as plated -8±3 reflow 0±2 50 Enthone Confidential Effect of Reflow Sn CuSnx Cu 51 Enthone Confidential Conclusion • Tensile stress hinders whisker growth. • Ni underlayer generates a tensile stress in the Sn film. • Reflow reduces whisker growth propensity. • Stress can be measured using XRD. 52 Enthone Confidential Recommendations • SB Sn/Ni/Cu • SB Sn/Cu, Reflow • SB Sn(thick)/Cu • The evolution of the stress in Sn should be monitored and used in conjunction with whisker test for predicting the whisker growth propensity. 53 Enthone Confidential Stress Evolution with Time No Whisker Tensile Aging Time Whisker Compressive 54 Enthone Confidential.