Previous Page  38 / 66 Next Page
Information
Show Menu
Previous Page 38 / 66 Next Page
Page Background

edfas.org

ELECTRONIC DEVICE FAILURE ANALYSIS | VOLUME 18 NO. 4

38

no bump pad can be seen on the top die side, which

confirms a die-die misalignment as the most likely root

cause. By using EOTPR, x-ray, and plasma FIB, this FA case

was able to be closed within 5 to 8 h, with sufficient data

information for follow-up questions and requirements.

CONCLUSION

In this article, the authors have attempted to reflect

on the emerging issue of how high-quality package FA

can be provided for increasingly complex 2-D, 2.5-D, and

3-D package products. After identifying pressing issues

and potential bottlenecks with the state-of-the-art FA

flow, the methods of EOTPR, 3-D x-ray, and plasma FIB

have been introduced. An integration flow for specific FA

cases has beenproposed anddemonstratedon twodiffer-

ent case studies. Hereby, the FA results confirm the high

potential of all three techniques and additionally enable

the opportunity for a fully nondestructive FA approach.

From a technical point of view, the following advantages

can be observed:

• EOTPR allows for early detection of the defect

z

-posi-

tion with the sample fully functional and unprepared.

This enables further control over product-specific FA

flow optimization.

• 3-D x-ray helps indicate the defect mechanism before

physical failure analysis. Using the virtual cross-sec-

tioning feature, the defect was isolated by 3-D, which

allowed optimum navigation for the cross section. In

addition, a combination of EOTPR and 3-D x-ray with

a defect database opens the opportunity for a fully

nondestructive package FA.

• Plasma FIB over conventional cross-sectioning tech-

niques removes the risk of potential preparation

artifacts such as cracks, foreign materials, or damage

(slurry, acid, etc.). It also significantly reduces the

turnaround time compared to Ga-FIB.

Considering these methods from a high-volume per-

spective, further optimization is required. In all cases,

standard setups allow only one sample investigation at

a time. While a multitool option is both cost and mainte-

nance extensive, further development focus is needed on

measurement automation.

REFERENCES

1. “3D IC Integration and TSV Interconnects—2010 Market Analysis,”

Yole Development, 2010.

2. J. Oswald et al.: “2.5D Packaging Solution—From Concept to Plat-

form Qualification,”

17th Electron. Packg. Technol. Conf.,

2015,

pp. 1-7.

(a)

(b)

Fig. 10

3-D x-ray results showing a misalignment between microbumps and TSV features

Fig. 11

Plasma FIB cut through x-ray-investigated area for (a) reference and (b) failing devices. High-magnification scanning

electron microscopy imaging confirmed missing die-die interconnection as the root cause.