Detecting Non-Destructive Faults with Electro Optical Terahertz Pulse Reflectometry

In the following article, how electro optical terahertz pulse reflectometry (EOTPR) can be used to quickly and non-destructively isolate faults in 2.5D packages will be discussed. Case studies to demonstrate how EOTPR can unambiguously differentiate between faults located in the C4 bump, TSV, RDL, and micro-bump of a 2.5D package are also presented.

Introduction

The numerous complex IC packages with their high density of critical components present new challenges for the failure analysis community. Complexities crop up because of shrinking interconnect sizes and packaging footprints, caused by growing functionality requirements as heterogeneous integration of electronic devices through 3D system-in-package (SiP) becomes more widespread. The 2.5D package is one such SiP that has lately gained traction in the semiconductor sector. Locating defects in 2.5D packages rapidly with good accuracy using traditional electrical fault isolation (EFI) methods is becoming increasingly more difficult.

Electro optical terahertz pulse reflectometry (EOTPR) is a novel terahertz time domain reflectometry method that offers the ability to rapidly and non-destructively isolate faults in advanced IC packages to an accuracy of less than 10 μm [1-8]. An EOTPR system has two photo conductive switches (PCSs), one functioning as an electrical pulse source and the other as an immediate current detector triggered by a mode-locked near-infrared laser. The emitted pulse is coupled into the DUT via a co-axial cable and a high frequency probe, while the reflected pulse is directed into a photoconductive detector.

An optical delay stage in the detector beam path sweeps the relative delay between the emission and detection timings (Figure 1a), producing a continuous voltage-time waveform. Figure 1b illustrates an example waveform from the open end of a high frequency probe.

This broadband method has a very low time base jitter and high temporal resolution [1]. Here, the fault isolation method is applied to a 2.5D package to illustrate how it can easily locate and differentiate between faults in C4 bumps, RDLs, TSVs and micro-bumps.

Schematic diagram of an EOTPR system.

Figure 1. a) Schematic diagram of an EOTPR system. (b) Typical raw EOTPR waveform from the open end of a high frequency circuit probe.

Results and Discussion

2.5D Package Details

Figure 2 shows a schematic of the 2.5D package used in this research. It consists of an application-specific integrated circuit (ASIC) die and high bandwidth memories (HBMs) assembled onto an interposer and then mounted on an organic substrate. During assembly, the various components undergo several cycles of solder reflow and underfill curing. Owing to the disparity of the materials’ coefficients of thermal expansion (CTE) within the 2.5D stack, thermo-mechanical stresses build up in the package. Consequently, failures can show up in the form of cracks (C4 bumps, μbump, dielectric) and delamination at the weakest interface.

The ability to identify the failure location non-destructively with good accuracy will be advantageous when deciding how to continue with physical failure analysis (PFA). This article will describe the application of EOTPR in two 2.5D packages with open failures.

Schematic of the 2.5D unit used in this study.

Figure 2. Schematic of the 2.5D unit used in this study.

Case Study 1

In case study 1, an open failure was identified when biasing a HBM signal pin in relation to the ground pin by curve trace. Waveforms acquired from time-domain reflectometry (TDR) largely indicated that the open failure lay between the substrate and interposer terminations as shown in Figure 3.

Schematic of suggested defect location identified by TDR.

Figure 3. Schematic of suggested defect location identified by TDR.

2D Real Time X-ray (RTX) did not expose any anomalies at the related signal pin interconnects as illustrated in Figure 4. Moreover, acoustic imaging did not expose any abnormalities at the μbump and C4 interface because of high signal losses through the HBM stack.

2D RTX image at defect location.

Figure 4. 2D RTX image at defect location.

Therefore, EOTPR was utilized to attain a more accurate defect localization. Figure 5 illustrates the EOTPR waveform recorded from the failing unit (red curve). Also illustrated are waveforms from a known good unit (green curve), a bare PCB substrate that terminates at the C4 bump pad (black curve), and a PCB substrate with interposer, which terminates at the μbump pad (blue curve). The failed unit waveform evidently shows an open fault, indicated by the positive peak in the waveform.

This happens at the same location as the interposer reference sample termination, revealing that the fault in the failed unit is situated in the μbump. Subsequent PFA exposed the open fault was due to a crack in the μbump (see inset of Figure 5), confirming the non-destructive EOTPR results.

EOTPR results from case study 1. The EOTPR waveform from the failed unit (red line) clearly show an open fault indicated by the positive peak in the waveform. This occurs at the same location as the interposer reference sample termination (blue curve), showing that the fault in the failed unit is located in the μbump. Top and Inset: SEM images of a cross section through the failed pin μbumps, showing a crack through the bumps.

EOTPR results from case study 1. The EOTPR waveform from the failed unit (red line) clearly show an open fault indicated by the positive peak in the waveform. This occurs at the same location as the interposer reference sample termination (blue curve), showing that the fault in the failed unit is located in the μbump. Top and Inset: SEM images of a cross section through the failed pin μbumps, showing a crack through the bumps.

Figure 5. EOTPR results from case study 1. The EOTPR waveform from the failed unit (red line) clearly show an open fault indicated by the positive peak in the waveform. This occurs at the same location as the interposer reference sample termination (blue curve), showing that the fault in the failed unit is located in the μbump. Top and Inset: SEM images of a cross section through the failed pin μbumps, showing a crack through the bumps.

Case Study 2

In case study 2, an open failure was noticed when biasing a specific signal pin with regards to the ground pin by curve trace. Likewise, waveforms attained from TDR again broadly indicated the open lay between the interposer and substrate (refer to Figure 3) while Scanning Acoustic Microscopy (SAM) and RTX did not divulge any anomalies at the associated failure pin interconnects as illustrated in Figure 6.

CSAM imaging at signal pin C4 bump

Figure 6. (a) Zoom in CSAM imaging at signal pin C4 bump and (b) 2D RTX at associated μbump and TSV location.

Again, EOTPR was used to attain a more accurate defect localization. Comparing the waveforms attained in Figure 7, the defect was found to be mid-way between the substrate termination and interposer termination, placing it near the bottom of the TSV. Subsequent PFA using Ga FIB exposed a thin layer of passivation material between the TSV and RDL to be the cause of the open (see inset of Figure 7).

EOTPR results from Case Study 2. The EOTPR waveform from the failed unit (red line) clearly shows an open fault indicated by the positive peak in the waveform. This occurs mid-way between the substrate (black curve) and interposer (blue curve) terminations, showing that the fault in the failed unit is located near the bottom of the TSV. Top and Inset: SEM images of a cross section through the failed unit TSV showing a thin layer of passivation between the TSV and RDL Cu.

EOTPR results from Case Study 2

Figure 7. EOTPR results from Case Study 2. The EOTPR waveform from the failed unit (red line) clearly shows an open fault indicated by the positive peak in the waveform. This occurs mid-way between the substrate (black curve) and interposer (blue curve) terminations, showing that the fault in the failed unit is located near the bottom of the TSV. Top and Inset: SEM images of a cross section through the failed unit TSV showing a thin layer of passivation between the TSV and RDL Cu.

Conclusions

To conclude, the article has dealt with how EOTPR can be applied to non-destructively isolate faults in 2.5D packages. EOTPR can unambiguously differentiate between failures at the key device features - namely the TSVs, C4 bumps, and μbumps - making it a robust tool for fast, non-destructive fault isolation in advanced packages.

Acknowledgments

The authors would like to acknowledge Advanced Micro Devices, Inc. (Singapore) for collaborating in this project. AMD, the AMD Arrow logo, and combinations thereof are trademarks of Advanced Micro Devices, Inc. Other product names used in this publication are for identification purposes only and may be trademarks of their respective companies. ©2017 Advanced Micro Devices, Inc. All rights reserved.

References

[1] Y. Cai, Z. Wang, R. Dias, and D. Goyal, "Electro Optical Terahertz Pulse Reflectometry - an Innovative Fault Isolation Tool", Proceedings of the 60th Electronic Components and Technology Conference (ECTC), Las Vegas, Nevada, USA, June 1-4, pp. 1309-1315, 2010

[2] M. Y. Tay, L. Cao, M. Venkata, L. Tran, W. Donna, W. Qiu, J. Alton, P. F. Taday, and M. Lin, "Advanced Fault Isolation Technique Using Electro-optical Terahertz Pulse Reflectometry", Proceedings of the 19th International Symposium on the Physical and Failure Analysis of Integrated Circuits (IPFA), Singapore, July 2-6, 2012.

[3] L. Cao, M. Venkata, M. Y. Tay, W. Qiu, J. Alton, P. F. Taday, and M. Igarashi, "Advanced Fault Isolation Technique Using Electro-optical Terahertz Pulse Reflectometry (EOTPR) for 2D and 2.5D Flip-chip Package", Proceedings of the 38th International Symposium for Testing and Failure Analysis (ISTFA), Phoenix, Arizona, USA, November 11-15, pp. 21-25, 2012.

[4] S. Barbeau, J. Alton, and M. Igarashi, "Electro Optical Terahertz Pulse Reflectometry - a fast and highly accurate non-destructive fault isolation technique for 3D Flip Chip Packages", Proceedings of the 39th International Symposium for Testing and Failure Analysis (ISTFA), San Jose, California, USA, November 3-7, 2013.

[5] K.C. Lee, J. Alton, M. Igarashi, and S. Barbeau, "Feature Based Non-Destructive Fault Isolation in Advanced IC Packages", Proceedings of the 40th International Symposium for Testing and Failure Analysis (ISTFA), Houston, Texas, USA, November 9-13, pp. 214-218, 2014.

[6] E. Kowalczuk, A. Bhattacharya, K. C. Lee, J. Alton, M. Igarashi, and S. Barbeau, “Fault Localisation of defects using Electro Optical Terahertz Pulse Reflectometry and 3D EM modelling with Virtual Known Good Device”, Proceedings of the IEEE 3D System Integration Conference, 2014, Kinsale, Cork, Ireland, December 1-3 2014.

[7] K.C. Lee, J. Alton, M. Igarashi, and S. Barbeau, " Fast Feature Based Non-Destructive Fault Isolation in 3D IC Packages Utilizing Virtual Known Good Device ", Proceedings of the 41st International Symposium for Testing and Failure Analysis (ISTFA), Portland, Oregon, USA, November 1-5, 2015.

[8] C. Schmidt, P. S. Pichumani, J. Alton, M. Igarashi, L. Chan, and E. Principe, " Advanced Package FA flow for next-gen packaging technology using EOTPR, 3DXRAY & Plasma FIB", Proceedings of the 42nd International Symposium for Testing and Failure Analysis (ISTFA), Fort Worth Convention Center; Fort Worth , Texas, USA, November 6-10, 2016.

TeraView Ltd

This information has been sourced, reviewed and adapted from materials provided by TeraView Ltd.

For more information on this source, please visit TeraView Ltd.

Tell Us What You Think

Do you have a review, update or anything you would like to add to this article?

Leave your feedback
Submit