Title
Estimating the reliability of encapsulated wire bonds Werner Engelmaier
Case study: How to assure the long-term reliability of encapsulated wire bonds with insufficient knowledge of material properties.
Estimating the reliability of encapsulated wire bonds
In my last column I discussed the fundamentals of fatigue and its analysis in general [Ref. 1]. I recently ran into a situation with a client that I had never encountered before and to which we developed an interesting solution. Concerns about the long-term fatigue reliability in a temperature-cyclic operating environment arose because of some wire bond failures during high-temperature storage testing. The information in Reference 1 gives the fatigue fundamentals; however in and of themselves they were not sufficient because the loading conditions for the wire bonds in Figure 1 could not be directly quantified. The available
material properties were inadequate for an analytical approach using finite element analysis [see Ref. 2]. The 99Al1Si
bonding wire was given as having a modulus of elasticity of 62 GPa (8.99x106
psi), an elongation of ~2.2 %, and a breaking strength of ~276 MPa (40,000 psi). For the transparent encapsulant, the
only known property was a coefficient of thermal expansion (CTE) of 148 ppm/˚C above its glass transition temperature of -49˚C; the CTEs of the other materials are
ENCAPSULANT! DIE!
SUBSTRATE! Figure 1. Configuration of encapsulated wire bonds.
that the bonding wire was not very ductile; however, the actual ductility needed to be established since for thin samples elongation measurements are an inadequate measure of ductility3
for the ceramic substrate ~6.8 ppm/˚C, of the silicon die ~2.8 ppm/˚C and the aluminum wire ~24 ppm/˚C. The elongation of ~2.2% indicated
h! , and the
99 90
ductility is needed for the estimation of the fatigue life of the bonding wire. The ductility and the fatigue behavior
of the bonding wire had to be established using the method standardized in References 4 and 5 for metallic foils. This was less than straightforward, since preparing the fatigue test samples consisting of 1.25-mil- diameter aluminum wire turned out to be significantly more difficult than working with foil samples of the same or even lesser thicknesses.
A mean ductility of 12.7% with
a standard deviation of 1.1% was determined. In Figure 2, a Manson-Coffin plot resulting from Eq. 4 in Reference 1 and material property values of Df
40,000 psi and E ≅ 8.99x106
= 12.7%, Su psi is shown.
≅
Also shown are the fatigue test results. As is evident, there is excellent correspondence between the derived Manson-Coffin plot
N(50%,ΔT=100°C) ≅480 cycles
50
Failures [%]
10 5
✕!✕!✕✕!! ✕!
@~225 cycles ✕!
First Failure 1
Figure 2. Manson-Coffin plot and mechanical cycling fatigue results for 99Al1Si bonding wire.
100 1,000 Fatigue Life [cycles]
Figure 3. Weibull plot of thermal cycling fatigue results of 99Al1Si wire bonds of configuration in Figure 1 for three different temperature cycles with the product reliability requirement.
10,000 ✕!
✕!✕! ✕!
✕! ✕!✕!
✕!!✕! ✕!✕!✕!✕!
✕!✕! ✕! ✕!
✕! ✕✕!!
✕! ✕!
PRODUCT RELIABILITY: N(10%)=3,650 cycles
N(50%,ΔT=70°C) ≅1,500 cycles
N(50%,ΔT=85°C) ≅880 cycles
N(50%,product) =5,659cycles
50 β=3.0
42 – Global SMT & Packaging – Celebrating 10 Years – December 2010
www.globalsmt.net
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48 |
Page 49 |
Page 50 |
Page 51 |
Page 52 |
Page 53 |
Page 54 |
Page 55 |
Page 56 |
Page 57 |
Page 58 |
Page 59 |
Page 60