Enhanced Polymer Passivation Layer for Wafer Level Chip Scale ...

More documents

Recommendations

Info

Abstract Wafer level chip scale package (WLCSP) have been used in many consumer products, and thus they are competitive in cost, size, yield, and technology. For advanced WLCSP, solder joint reliability is a major concern. Underfilling is a common solution to addressing WLCSP reliability concerns. Typical stress-relieving methods such as molding compounds and capillary underfills have proven successful in CSP protection, but their added cost to the assembly process is generally prohibited. Instead, successful low cost reliability solutions have generally been the adaptation of wafer level backend packaging processes such as modification of the redistribution layer materials, solder selection, or metal pad thickness. However, the increased performance is limited. In this research, a new approach is presented to reexamine the final passivation layer as more than a dielectric, but also a partial underfill. The new material, branded as "SolderBrace" as an alternative to underfill, is a photo-imageable molding compound with a low CTE. This layer of SolderBrace coating adds a mechanical buffer to the front side of the WLCSP and delivers improved reliability with conventional tools, short process times and lower costs. SolderBrace coated WLCSPs and standard non-coated WLCSPs, were designed and fabricated with known standard fabrication procedures. The processing of the SolderBrace coatings was achieved by two methods. The first is similar to that of standard polyimide processing: spin coat, bake, photo-image, solvent develop, and ball drop. The second application process involves printing the material on the already-balled wafers followed by solder cleaning and cure. These ii
SolderBrace coatings were low temperature cured and generated minimal wafer bow. The test WLCSPs were assembled to the circuit boards after the wafer singulation. The standard thermal cycling test was used for reliability testing. A finite element based approach was also used to gain a deeper understanding of the solder joint failure mechanism caused by the repeated thermal stress. According to the test results, the SolderBrace coated dies had much higher lifetime than the non-coated dies. SolderBrace technology may offer a unique method to package low cost high performance WLCSPs. The simulation results also give insight on the stress generation and can provide guidance to appropriate design adjustment. iii
Page 1: Enhanced Polymer Passivation Layer
Page 5 and 6: Table of Contents Abstract ........
Page 7 and 8: 4.4 Failure Analysis ..............
Page 9 and 10: List of Figures Figure 1.1 Trends i
Page 11 and 12: Figure 3.17 SAC305 Reflow profile .
Page 13 and 14: CHAPTER 1 INTRODUCTION In the era o
Page 15 and 16: This effect changes the package to
Page 17 and 18: 1.4 Solder Joint Fatigue Figure 1.2
Page 19 and 20: leading to the solder joint failure
Page 21 and 22: umped wafers. The coating is stenci
Page 23 and 24: 11 Bumped Wafer Print over ball Pre
Page 25 and 26: 2.1 Chip Scale Package Technology C
Page 27 and 28: Mountable with conventional assembl
Page 29 and 30: Figure 2.3 Cross section of a typic
Page 31 and 32: Table 2.1 Comparison between tradit
Page 33 and 34: In-Situ Bumped Wafers Placed Prefor
Page 35 and 36: of solder joint failure, and they o
Page 37 and 38: (a) (b) Figure 2.8(a). Metalized ph
Page 39 and 40: "underfilled" structure distributes
Page 41 and 42: have generally been the modificatio
Page 43 and 44: are “low temperature” wafer lev
Page 45 and 46: 2.3.4 Optimized SolderBrace Materia
Page 47 and 48: CHAPTER 3 WLCSP DIE FABRCIATION In
Page 49 and 50: Table 3.1 Test Die used for Reliabi
Page 51 and 52: Figure 3.4 Fabrication Process Flow
Page 53 and 54:
spin coating. A layer of light sens
Page 55 and 56:
layer, is investigated as a potenti
Page 57 and 58:
10. Final cleaning: After the passi
Page 59 and 60:
Cyclopentanone, a colorless liquid
Page 61 and 62:
used to check the basic spin and ph
Page 63 and 64:
Figure 3.7 SolderBrace film thickne
Page 65 and 66:
6. Post-UV exposure bake: This step
Page 67 and 68:
8. Pattern characterization: The su
Page 69 and 70:
offers the optimum flux release cha
Page 71 and 72:
fatigue resistance, lower cost SAC
Page 73 and 74:
Figure 3.12 Solder Ball Placement M
Page 75 and 76:
of solder balls to the holes on the
Page 77 and 78:
of thermal shock. If the temperatur
Page 79 and 80:
Figure 3.17 SAC305 Reflow profile -
Page 81 and 82:
foaming and high performance cleane
Page 83 and 84:
- Polish clean: used the optimized
Page 85 and 86:
Figure 3.22 SolderBrace printed waf
Page 87 and 88:
Voids Figure 3.25 SolderBrace Mater
Page 89 and 90:
cutting speed was set at a low valu
Page 91 and 92:
Figure 4.2 Process Flow of Board As
Page 93 and 94:
pitch, and bump diameter), and subs
Page 95 and 96:
4.2.4 X-Ray Inspection Figure 4.5 R
Page 97 and 98:
Figure 4.6 Air to air thermal cycli
Page 99 and 100:
Figure 4.7 Thermal Cycling Setup 87
Page 101 and 102:
4.4 Failure Analysis The main purpo
Page 103 and 104:
Figure 4.10 Pad cratering failure o
Page 105 and 106:
CHAPTER 5 FINITE ELEMENT ANALYSIS I
Page 107 and 108:
5.1.1 Modeling Approaches Due to th
Page 109 and 110:
are shown in Figure 5.2. However, t
Page 111 and 112:
5.1.2 Material Properties Many pack
Page 113 and 114:
strain rate sensitivity. The evolut
Page 115 and 116:
Morris et a1 [96] used a double pow
Page 117 and 118:
focus on the time-dependent effects
Page 119 and 120:
5.2 Modeling Procedure In this proj
Page 121 and 122:
Figure 5.6 The diagonal symmetry mo
Page 123 and 124:
5.2.3 Meshing Table 5.2 Anand const
Page 125 and 126:
direction, but that the surface is
Page 127 and 128:
PREP7 tref,398 ! set zero strain te
Page 129 and 130:
5.2.7 Thermal Fatigue Life Predicti
Page 131 and 132:
CHAPTER 6 CONCLUSIONS Wafer level c
Page 133 and 134:
References 1. Future Trends in Elec
Page 135 and 136:
23. Electronic Packaging and Interc
Page 137 and 138:
48. Ultra Low Stress and Low Temper
Page 139 and 140:
70. Factors for Successful Wafer-Le
Page 141 and 142:
92. Microstructural Dependendence o
Page 143:
111. Nonlinear Analysis of Full Mat
show all

Enhanced Polymer Passivation Layer for Wafer Level Chip Scale ...

Create successful ePaper yourself

Delete template?

Save as template?