Detection of Counterfeit Semiconductors: Nondestructive and Destructive Examples


Reading time ( words)

We need to go beyond visual inspection to track down counterfeit components. Hemant C. Warad, Sakda Sangthamma, Anusorn Sawetwong, and Martin Huehne, Celestica, describe two examples of counterfeit integrated circuits (ICs) received from a broker for assembly on printed cicuit boards (PCBs). The counterfeits were intercepted by a counterfeit detection procedure routinely applied to every purchase of electronic parts from a broker. The successful detection of the counterfeiting was based on optical inspection of the package, X-ray imaging, scanning acoustic microscopy (C-SAM), and optical inspection of the die after decapsulation. The described cases occurred in 2008.

Failures in electronic circuits are not only caused by defects in components or flawed assembly-level processes. The component’s legitimacy is also a key issue. In widely published cases beyond the electronics industry, counterfeiting appears often as imitation of an existing product, sold on the market masquerading under the brand name of the original. The definition of counterfeiting as the “deliberate falsification or modification of any commodity with the main objective to defraud or deceive the purchaser”1 encompasses cases such as the sale of recycled components or diverted rejects, which seem to be more common for semiconductors than the fabrication of counterfeits from raw material. Semiconductor counterfeiting has increased exponentially over the years, impacting the reliability of a wide variety of electronics systems used by consumers, businesses, and governments.2 Annually, billions of dollars are lost in revenue from counterfeit semiconductor devices making their way into each and every sector — from highly dependable military equipment to consumer products.3 The cost multiplies several folds for a company to replace a counterfeit component.

A reasonably safe way to prevent counterfeits from being mounted on PCB assemblies is to acquire all raw parts exclusively through distribution channels authorized by the suppliers.4 However, if the lead time to procure components through the authorized distribution channel does not meet the production schedule, acquiring them from a broker is a viable alternative to delaying production.4 In this case, minimize the reliability risk of using counterfeits by analyzing part samples for signs of counterfeiting.

Experimental

Optical microscopy of the package surface with a stereomicroscope and transmissive X-ray imaging are two of the main nondestructive analysis (NDA) techniques to find external and internal package anomalies.5 At some EMS facilities, these two basic nondestructive tests are mandatory for any type of electronic component acquired through a non-authorized channel. The external visual inspection is carried out by sampling one or more devices from a given lot. Device markings and, in some cases, dimensions are compared to the datasheet to check for authenticity. The package is thoroughly examined for evidence of improper handling or previous use. The lead finish is inspected for presence of solder/flux left by previous use or refinishing. X-ray inspection checks the internal structure and can detect packages without a die, with a different die size, with a different lead frame, or with different wire bond connections than a known-good sample. Severe delamination and internal package cracks, signs that a used part was removed from an assembly and re-sold, may show up in X-ray inspection.

If optical or X-ray inspection raises counterfeiting suspicions, scanning acoustic microscopy (C-SAM) can be used as an additional NDA technique to provide complementary information on package damage such as delamination and internal package cracks. Decapsulation exposes the die for optical inspection under a metallurgical microscope, in particular to check die markings for the design year and other information that can help verify counterfeit/authentic with the help of the original component supplier.

Results

The results described were obtained with the aforementioned methods on commercially available equipment installed in the onsite laboratory.

Figure 1. X-ray inspection showing anomalies of silver epoxy in sample for case A. Inset (left top) showing the sample and the inset (right bottom) shows the die markings. The name of the component manufacturer and related information are blanked out.

Case A. Table 1 briefly describes inspection techniques and the results, which cast doubt on the authenticity of this component. Optical inspection shows that the new series components are in use since “2003” and our component with the same part number has a year code “1996,” which indicates the possibility of a suspected reworked component with new markings applied. 2D X-ray inspection (Figure 1) shows suspicious anomalies in the silver epoxy. Inspection of the decapsulated IC reveals an old design year, which supports the finding from the inspection that this is possibly an improperly reworked part with new markings. 

Table 1. Summary for case A.TechniqueResultOptical inspection The marking “CKJ9613” on the top indicates a very old date code “9613,” and according to the manufacturer’s specification this component is discontinued since April 2003 and replaced with a new series.X-ray imaging The contrast pattern indicates inhomogeneous distribution of silver epoxy with apparently insufficient coverage at corners and at edges under the die and wire bonds, and several irregularly shaped crack-like anomalies.DecapsulationOld design year.

Case B. This case demonstrates that optical inspection of the package alone is not adequate to verify authenticity of a component (Table 2). Optically, the package appears as specified by the supplier and raised no suspicion. The anomalous white lines in X-ray images were not considered conclusive enough to reject the part, but were suspicious enough to proceed with further analysis by scanning acoustic microscopy (C-SAM) and cross-sectioning. The A-scans (Insets in Figure 2) indicate delamination on the die attach paddle and lead fingers as well as on the die surface. The bright lines on the leadframe and at the edges of the die, observed in X-ray and C-SAM, were verified by cross-sectioning to be support material attached above the leadframe. This support material apparently has a lower attenuation for X-rays than the mold compound. This appears to be there by design and is no indication of counterfeiting. However, cross-sectioning revealed cracks at the sides of the die. These cracks are a clear sign of counterfeiting.

Figure 2. C-SAM inspection showing delamination on the die attach paddle and lead fingers in sample #1 and on the die surface, die attach paddle and lead fingers in sample 2 of case B. Inset shows the A-Scan indicating phase inversion to show delamination.

Delamination is usually not present in new wire-bonded components, but it does occur when moisture has been absorbed by the package and the part is exposed without prior baking to high-temperature desoldering. This is the likely cause of the cracks found at the sides of the die when the component was cross sectioned. These cracks indicate that these parts might have been reclaimed from discarded electronics, reworked without sufficient prior baking, and then sold as new components, i.e. as counterfeits.

Table 2. Summary for case B.AnalysisDescriptionOptical inspection No obvious defects.X-ray imaging The contrast pattern with some bright lines was observed on the lead frame and at the edges of the die possibly indicates delamination and package cracks.C-SAMSample 1: Delamination on the entire die attach paddle and on a wide area on the lead fingers.Sample 2: Delamination on the entire die surface, the entire die attach paddle and on a wide area on the leadframe.Cross section The cracks at the sides of the die indicate improper rework. The bars of lighter material on the leadframe and at the edges of the die explain some of the bright contrast in the X-ray image with package design (Figure 3).Conclusion

In the two cases discussed here, the techniques routinely implemented for inspection of parts acquired from a broker were effective in intercepting counterfeit parts before assembly on PCBs. In one case, the comparison of package markings with information from the original supplier indicated the part to be a counterfeit because the device was never fabricated by the original supplier in the observed device name and package year combination. Anomalies in X-ray images indicated that the counterfeiter removed the part from an assembly improperly. Apparently, the counterfeiter covered the package with new markings. In the other counterfeit case, the severe delamination exposed by C-SAM and the cracks exposed in the cross sections strongly indicated that this component, too, had already been assembled before, removed from the board with an improper method detrimental to package integrity, and sold fraudulently as a new part.

Figure 3. Cross-sectional image of sample 1 with package cracks radiating (Insets) from the edges of the die and package cracks at the bottom of the die in sample 2, case B.

REFERENCES:1.  Henry Livingston, "Avoiding Counterfeit Electronic Components", IEEE Transactions on Components and Packaging Technologies, Vol.30, Iss.1, Mar 2007, pp.187-189.2.  Lev Shapiro, "Counterfeit Electronics: Threats, Risks and Prevention Practices", Conference Proceeding, 14th Pan Pacific Microelectronics Symposium, Feb 2009, pp. 28-32.3. Michael Pecht, Sanjay Tiku, “Bogus: Electronic Manufacturing and Consumers Confront a Rising Tide of Counterfeit Electronics” IEEE Spectrum, Vol. 43, no. 5, May 2006, pp. 37-46.4. Kristal Snider and Daniel J. DiMase, "Avoiding and Resolving Disputes Over Unsatisfactory Components", Conference Proceeding, SMTA International, Aug 2008.5. David Bernard and Bob Willis, "A Suggested Process for Detecting Counterfeit Components", Conference Proceeding, SMTA International, Aug 2008.

Hemant C. Warad, Sakda Sangthamma, Anusorn Sawetwong, and Martin Huehne may be contacted at the Regional Failure Analysis & Reliability Laboratory, Celestica (Thailand) Ltd, 49/18 Laem Chabang Industrial Estate, Moo 5, Tungsukhla, Sriracha, Chonburi 20230, THAILAND; wcheman@celestica.com.

SMT, April 2010

Read More about Counterfeit Detection:No Die Code, No OEM Certification, No Service: How One Component Distributor Is Keeping Counterfeit Parts out of Your SMT Line North Shore Components Expands Counterfeit Avoidance Programs More...

Subscribe

Join the PennWell SMT Group on LinkedIn

Become a Fan on SMT's Facebook Page

Post your electronics manufacturing, SMT-related material to the #SMT community on Twitter. Use the #SMT hashtag.

Share

Print



Copyright © 2019 I-Connect007. All rights reserved.