0
Research Papers

# Underfill Filler Settling Effect on the Die Backside Interfacial Stresses of Flip Chip Packages

[+] Author and Article Information
Cheng-fu Chen

Department of Mechanical Engineering, University of Alaska Fairbanks, P.O. Box 755905, Fairbanks, AK 99775-5905ffcc@uaf.edu

Pramod C. Karulkar

Office of Electronic Miniaturization, University of Alaska Fairbanks, P.O. Box 755905, Fairbanks, AK 99775-5905

J. Electron. Packag 130(3), 031005 (Jul 30, 2008) (10 pages) doi:10.1115/1.2957324 History: Received July 25, 2007; Revised January 07, 2008; Published July 30, 2008

## Abstract

Underfill is usually modeled as an isotropic medium containing uniformly distributed filler particles. However, filler particles tend to settle (or segregate) and thus alter the mechanical response of the flip chip die attachment package. The integrity of such flip chip attachment is different from that with an ideal, isotropic underfill with very uniform distribution of filler. We analyzed the thermomechanical implications of filler settling to the stresses along the die/underfill interface by considering different profiles for the local concentration of filler and calculating their effective material properties by employing the Mori–Tanaka method. As the worst-case scenario, direct silicon die attach with solder bumps was assumed to analyze the interfacial stresses, which were predicted in trend by a simplified multilayered stack model and calculated in detail by finite element simulation. The filler settling has a localized but strong influence on the interfacial peeling stress near the edge of the die. The extent of this influence is determined by the profile of filler settling: (1) if the filler is assumed to settle in the form of a bilayer, then the peeling stress near the die’s edge increases and it is directly proportional to the average volume fraction of the filler; (2) if the filler is assumed to settle gradually, then the magnitude of the peeling stress near the edge of the die becomes smaller as the local filler volume fraction near the die interface increases. The filler settling has no significant effect on the other components of the interfacial stresses. The edge fillet of underfill in pure resin can locally reverse the direction of the interfacial peeling stress and increase the interfacial shearing stress near the die’s edge.

<>
Copyright © 2008 by American Society of Mechanical Engineers
Your Session has timed out. Please sign back in to continue.

## Figures

Figure 2

Filler distribution is classified into three categories: (a) fillers are uniformly distributed to form a single, macroscopically homogeneous underfill layer; (b) a bilayered distribution, one with pure resin and the other laden with filler (shown is with filler packed in a face-centered cubic packing style, known for the best packing efficiency (74.05%) for equisized spheres); and (c) gradual settling. Gravity is downward.

Figure 3

Six y-dependent profiles for filler settling (using equisized spheres). Case A: Filler in uniform distribution. An isotropic underfill model. ϕ(fillervolumefraction)=35%. Case B1: The most packed filler settling. A bilayered underfill model. ϕ=35%. Case B2: The most packed filler settling. A bilayered underfill model. ϕ=60%. Case C1: Gradual filler settling. An orthotropic underfill model. ϕ=35%. Case C2: Gradual filler settling. An orthotropic underfill model. ϕ=35%. Case C3: Gradual filler settling. An orthotropic underfill model. ϕ=35%. In group B’s, filler spheres in the bottom layer are packed in the face-centered cubic style, which is known for the best efficient packing of equisized spheres with a packing efficiency of 74.05%. In group C’s, underfill is divided into ten horizontal strips (for the simulation purpose) and each strip has a constant localized volume fraction, which is determined by the mid-y value of each strip from the associated distribution curve shown.

Figure 4

Approximation of the Mori–Tanaka method for the effective material properties of underfill. (a) The effective elastic modulus. (b) The effective CTE.

Figure 5

(a) Schematic of a half-symmetric flip chip package with underfill. (b) For theoretical estimation of the interfacial stresses, the flip-chip model is simplified by a multilayered stack with dissimilar materials by neglecting the solders, the overhung part of the substrate, and the underfill edge fillet. Dimensions in mm.

Figure 6

Figure 7

Comparison of the FE and analytical results of the interfacial stresses at the die/underfill interface for (a) no filler settling (left column), (b) bilayered filler settling (middle column), and (c) gradual filler settling (right column). (FE results are with the diamond-ridden lines. Analytical results are in solid lines.)

Figure 8

Distribution of the peeling stress (S22) along the die-underfill interface. Underfill fillet is not included in the model. The curves end at x=3.475mm where the die edge is.

Figure 9

Distribution of the peeling stress (S22) along the die-underfill interface by using the bilayered underfill model. Underfill fillet is not included in the model. The curves end at x=3.475mm where the die edge is.

Figure 1

Sign convention (shown positive) for planar stresses along the interface of two bonded layered materials: the interfacial peeling stress (S22), the interfacial shearing stress (S12), and the axial normal stress near the interface (S11).

Figure 13

The edge fillet of underfill changes the direction of the interfacial peeling stress and reduces its magnitude near the edge of the die, and also locally increases the interfacial shearing stress. (The comparison is based on the uniform filler distribution scenario with a volume fraction of 35%.)

Figure 10

Distribution of the peeling stress (S22) along the die-underfill interface by using the underfill model of gradual filler settling. Underfill fillet is not included in the model. The curves end at x=3.475mm where the die edge is.

Figure 11

Comparison of the interfacial peeling stress (S22) at the edge of the die.

Figure 12

FE simulated distribution of the interfacial stresses along the die/underfill interface when the edge fillet of underfill is considered. (The B1 and C1 curves are very close to each other in the figures.)

## Discussions

Some tools below are only available to our subscribers or users with an online account.

### Related Content

Customize your page view by dragging and repositioning the boxes below.

Related Journal Articles
Related Proceedings Articles
Related eBook Content
Topic Collections