Every morning in the failure analysis lab holds the potential for a new challenge. A board from a missile guidance system, an integrated circuit from the latest cell phone or video game console, or pieces of a high tech neural implant may be but a few of the many different devices that analysts may find waiting on their desks in the morning (after, of course, a requisite stop at the coffee pot – like many other engineering fields, xanthic alkaloids are one of the cornerstones of a healthy analyst’s diet).

Though there is a vast range of device types that may cross an electronic failure (FA) analyst’s desk, there are similarities between every FA project that can be examined; regardless of the unique circumstances of a given electronic device, there are still a handful of standard steps that come together to make up a typical day in the electronic failure analysis lab.

In many cases, an electronic failure analysis engineer may not even lay eyes upon a failing device during the first step of the electronic failure analysis process – before beginning any work with the samples, it is often necessary to perform a little preliminary research. This initial review may consist of conversations with a customer, poring over data sheets, or reading through a device history, all in order to better understand the failing device – the conditions it was subjected to, the environment in which it was used, the amount of use the device received before failing, and so on. This initial investigation is crucial, as it allows the analyst to begin formulating hypotheses about potential failure mechanisms – which, in turn, can help to determine the course of the analysis going forward, by eliminating unnecessary tests and identifying techniques that present the best chance for uncovering the failure.

Once an electronic failure analysis lab analyst has researched the project sufficiently, the next step is to begin non-destructive testing of the failed electronic device. A course of non-destructive testing may be comprised of many different aspects – almost always, an in-depth optical inspection is performed to look for any obvious physical defects (e.g. cracked leads or encapsulant, burned circuit board traces, etc.). The optical inspection will often be supplemented by a high-resolution x-ray inspection to look for anomalies that may be hidden from view, like fused bond wires or voided solder. For failing integrated circuits, acoustic imaging may also be beneficial, allowing an analyst to detect delamination or other packaging anomalies that could be related to a failure. Just like the preliminary research, non-destructive testing often provides an analyst with crucial information that may dictate the entire course of the analysis. For example, many large microprocessors may have hundreds of connections between the package and the silicon die; if x-ray inspection can reveal that one of these bond wires has been fused, an analyst can immediately identify an area for further testing without needing to perform electrical characterization of the multitude of various inputs and outputs of the device, and to create a theory about the most likely cause of failure (since fusing a bond wire requires very high current, electrical overstress immediately shoots up to candidate #1).

After completing the non-destructive testing, the electronic failure analysis engineer must characterize and isolate the failure. The analyst must first confirm that the device is, in fact, failing in the reported fashion, by assembling an electrical test setup and exercising the part in the same way as reported by the customer. After the failure is confirmed, additional electrical characterization may be performed in an attempt to define the problem more narrowly (e.g. a problem reported as “output stuck high” may be translated into “output shorted to the power supply through 100 ohms” with properly directed electrical characterization). With the device characterized, attempts to isolate the failure can begin – oftentimes, these tests are destructive in nature to varying degrees, since integrated circuits must be decapsulated, PCBs must have parts removed or traces cut, and so on. The techniques used to isolate the failure may be as technologically involved as thermal imaging or light emission microscopy, or as (relatively) simple as tracing a signal path using hand-held probe tips. The ultimate goal of the isolation step is to identify a site that can be investigated to reveal the root cause of electronic device failure.

Finally, armed with the knowledge of an isolated location, the electronic failure analysis lab engineer can proceed to the final steps of the project. Depending on the type of failure, these steps may be wildly different – a cross-section might be necessary to identify damage between the layers of a printed circuit board, while an integrated circuit may require parallel deprocessing to uncover defects at the silicon substrate. The final steps of the analysis need not be destructive; often, elemental analysis is indicated, to determine the nature of a contaminant on a sample. Finally, with all analysis concluded, the analyst can write up the report, closing out their day… only to begin again anew tomorrow, with an entirely different set of challenges.

In many cases, it is necessary to isolate a single defect amidst a vast array of circuitry, singling out a single leaky gate or overdriven transistor from among billions, in order to perform a successful failure analysis. Without some visual way to pluck the single defective device out from the lineup of identical looking circuit elements, an analyst cannot properly target the more destructive steps in the analysis, like cross-section or deprocessing. While some tools, like thermal imaging or other heat-sensitive techniques, can be successful in isolating an area for further investigation, in some cases they aren’t enough; the defect may not be generating enough heat to be detected. In these cases, a different approach, in which one takes the time to understand a device more completely by contrasting some sort of characteristic signature of malfunctioning devices against those that are properly functioning, may be able to isolate the failure. Emission microscopy is one such method of characterizing devices, and offers an excellent picture of many different types of failure upon which to build an analysis.

Emission microscopy (often referred to as light emission microscopy, photoemission microscopy, or by the trade name EMMI – EMission MIcroscopy) uses a high-gain camera to detect the infinitesimally small amounts of light emitted by some semiconductor devices and defects. The device under test is placed in a completely dark enclosure that houses the microscope, then is put under electrical bias in the failing condition. The camera system begins acquiring an image (essentially counting photons), mathematically integrating data over times as long as several hours. At the end of the integration time, the system overlays a map of emitted light on top of an image of the failing chip, showing areas of dense photoemission with brightly contrasting bursts of color.

There can be several different causes of photoemission: transistors that have been damaged by electrical overstress or those with gate oxide pinholes, for example, will often photoemit. To confuse things further, there are some semiconductor devices that photoemit even when operating properly, simply as part of their normal operation; aside from the obvious example of optoelectronics like LEDs, memories will often photoemit when being read or written. While it may seem like emission microscopy has the potential to offer false positives, limiting its utility as an analytical tool, it is actually its ability to provide a sort of functional map of a device that makes it so useful.

While emission microscopy can certainly be used successfully to locate many of the same types of defects that can be found with thermally-sensitive techniques – the damaged diffusions that result from an electrical overstress event at an ESD protection diode, for example, is often a great photoemitter – the real benefit comes from the ability to analyze a failing unit with respect to a known good device. By first analyzing a functional device in its normal operating condition, an analyst can generate a reference against which to judge the failing device. With a photoemission image of the reference device in hand, emission microscopy analysis of the failing unit is reduced to a childhood game of “find the differences”, with malfunctioning transistors and damaged diffusions standing in for the typical whimsical fare of tulips or buttons. Another benefit of emission microscopy is that it opens up an alternative method for analyzing very dense, cutting-edge chips, simply by virtue of the nature of the camera used to acquire photoemission images.

Typically, emission microscopy is performed following the chemical decapsulation of a part, after the plastic encapsulant material has been dissolved away from the microchip that lies at the heart of the packaged device. In more modern semiconductor devices, the junctions and transistors of the device (the most likely sources of photoemission) are buried beneath multiple layers of densely packed metal, hiding potential emission sites from the view of the camera.  Fortunately, due to the emission microscope camera’s sensitivity to short-wave infrared light, there is a way to image a failing device and completely bypass the problem of the interceding metal layers – by imaging through the back side of the device. First, the device is thinned and polished to a mirror finish, to help facilitate the transmission of photons through the silicon substrate. Because silicon is moderately transparent to infrared light, some photoemission makes it all the way through the substrate and is gathered by the camera; when used in conjunction with a source of infrared illumination, the emission microscope can generate images that are very similar to those taken conventionally, from the top side of the device. By imaging from the back side, the dense metallization that shrouds the emission sites from the camera can be removed from the equation; the active layer is, from the camera’s point of view, the uppermost layer, with nothing but semitransparent silicon substrate in between to intercede.

Emission microscopy can be an invaluable tool for rooting out defects on a failing semiconductor device. It is important to remember, however, that the emission microscope only generates a picture; it is the job of a well-trained failure analyst to take that picture and translate it into data that can be used to root out defects on a malfunctioning device.

There are many hurdles that must be overcome when attempting to introduce a new electronic gadget to the market. The trials and tribulations of creating a prototype and developing a unique, compelling solution to a consumer problem are only the first step in a long series of trials; with a working prototype in hand, a manufacturer must perform extensive testing on their new product in order to ensure reliability over its lifespan, a process that often leads to several costly design revisions before the product is even released for general consumption. Even after a reliable product has been produced, the qualification process for the new device is not over; unless the manufacturer is making a type of device that is specifically exempted, the new product must undergo RoHS certification or be barred from sale in the vast majority of markets.

Implemented in July of 2006, the Restriction on Hazardous Substances Directive (RoHS) is a European Union edict that places stringent limitations on the use of six substances found in many electronic devices. These substances were chosen to be restricted, since they are the greatest contributors to environmentally hazardous electronic waste. The heavy metals, like lead and mercury, in “e-waste” can contribute to groundwater poisoning, while other materials like the polychlorinated biphenyls found in scrap electronics are highly carcinogenic. Worse still, undeveloped countries often import large quantities of e-waste, then recycle it using crude methods that release untold quantities of carcinogens and neurotoxins into the air. Remnants are washed away into rivers and eventually to the ocean.

RoHS aims to remedy this highly toxic situation; however, there are additional ramifications to this directive for producers of electronic components, since the onus falls upon them to ensure that their devices are in compliance with the new restrictions. The RoHS certification process is one way to demonstrate this type of compliance; this process often involves extensive analytical work to ensure that none of the prohibited materials have unintentionally made their way into the end product.

Performing a RoHS certification involves a deep-dive into the composition of the constituent parts of a device – extensive analysis must be performed to ensure that the types of solder used to mount components onto the board, the fire-retardant material in the board itself, and even the protective coating used to shield delicate electronics from harsh environments are all free from proscribed elements.

This analysis can take many forms – x-ray fluorescence is often used to examine plating on components to look for traces of lead, while more precise energy dispersive spectroscopy techniques might be employed to characterize solder joints. These tests are sufficient for the elemental substances restricted by the RoHS directive; for the molecular compounds, it is often necessary to use techniques like Fourier transform infrared spectroscopy, which uses a single-wavelength infrared light source to excite the atoms in a molecule, causing bonds to vibrate and oscillate in characteristic patterns that can be used to positively identify a given material at the molecular level.

Though ensuring that the restricted materials are absent from a given part may be the primary goal of an RoHS certification, a thorough analysis doesn’t stop there; changing to “environmentally friendly” materials may have a detrimental impact on product reliability, a possibility that should also be explored during the certification process.

The substitute materials used for creating RoHS compliant devices (lead-free solder being chief amongst them) often have different characteristics than those they replace. Reflow times and temperatures must be changed to accommodate the different demands of the new materials, potentially causing additional stress that may drastically reduce the lifespan of the product. These materials also react differently to prolonged thermal stress, making more robust cooling systems necessary for RoHS compliant products that generate large amounts of heat. Further, improperly formulated solder or component plating may lead to the formation of “tin whiskers”, monocrystalline extrusions of conductive tin that can break free and create short circuit conditions between the leads of a device, potentially resulting in a catastrophic failure.

Due to the increased risks inherent in making a transition into RoHS certification, it is often necessary to perform additional characterization of the device to avoid succumbing to any of the aforementioned pitfalls. Fortunately, many labs that can perform RoHS certification inspections also have the capability to aid with the sorts of reliability analysis (thermal stressing, cross-section, and so on), thereby providing a complete picture of the impact of a RoHS transition.

Though RoHS certification may seem intimidating, it is an absolute necessity for any company wishing to be competitive in the electronics industry; after all, without RoHS certification, the doors of countless markets will remain closed to any new product.  Fortunately, there are IC failure analysis labs that can offer in-depth, comprehensive analysis of a new product, providing a helpful guide through the tangles of the RoHS auditing process.

In part one of this series of tips for outsourcing or hiring an electronics failure analysis service, we examined the wide variety of information that should be gathered before sending a failing part out for analysis. The construction of a detailed packet of data, including a problem description, a background or history of the failing device, and any auxiliary documents like layouts or schematics that may be necessary in chasing down the root cause of failure of a device is an involved process – but, once such a dataset has been assembled, the struggles of choosing a lab to entrust it with can begin in earnest. Just as one would not want to drop an expensive supercar off with any random shadetree mechanic, a one-of-a-kind failure should be sent to a lab with the  best (and most relevant) capabilities, experience, and a proven track record, in order to help ensure the best results.

When determining which lab has equipment and capabilities that best match up with a given sample, it is often necessary to consider the unique features of the failing device. Analysis of a high speed, low noise amplifier, fabricated using III-V semiconductor technology and used for wireless communications, requires a completely different set of test equipment and techniques than the disassembly of a microelectromechanical system (or MEMS) used for measuring acceleration as part of the accident detection system in a car. To elaborate on the prior example, a lab that has focused on the equipment necessary to test wireless technology – high frequency oscilloscopes, expensive active probing systems, and the like – may not have invested the same amount of money into the precision grinding, polishing, and chemical containment systems that are required for MEMS analysis. Similarly, a lab that primarily deals with counterfeit inspection of legacy electronics may not have the deprocessing or high resolution imaging capability to tackle defects at the more modern technology nodes with features measuring 90 nanometers or smaller. To that end, it will often be necessary to speak with FA engineers at multiple potential suppliers, conducting a sort of survey of the types of techniques that may be available that are applicable to a given failure.

Though electronic failure analysis equipment is important when hiring an electronics failure analysis service, it is not the sole consideration; equally important is determining whether the lab personnel have the experience not only to run the equipment, but to correctly interpret the results. The issue of proper interpretation can become especially pronounced when sending parts to overseas test or FA facilities; though, inevitably, their services may be much cheaper, and some may indeed provide a quality service, their work often fails to go into the level of detail necessary for a truly beneficial failure analysis. Keeping this in mind, it is absolutely necessary to vet the background of a lab’s employees; fortunately, many of the best labs will publish the curriculum vitae of their key employees, allowing potential customers to easily see the breadth and depth of products that the FA engineers have had direct, in-depth exposure to. As an example, a lab boasting an employee who has published several papers on inspecting electronic components for counterfeiting will almost certainly be an excellent source for performing authenticity assessments on samples from a suspect supplier.

Finally, another consideration in evaluating a lab when outsourcing electronics failure analysis is examining their track record of successes and failures. Naturally, coming out and directly asking the marketing staff of the lab is an exercise in futility (unless of course the particular marketing rep is in a particularly disgruntled mood); however, one of the best ways to determine a lab’s history is through word of mouth. Given the vast array of manufacturers and the relatively small number of electronics failure analysis labs, there is a good likelihood that a colleague or other professional contact has used a given lab at some point in the past, and can give a frank evaluation of their capabilities. Indeed, this effect helps to ensure that independent FA labs perform all due diligence and explore all failures to the extent of their capabilities – the last thing any lab wants is to be associated with bad results.

Finally, once the painstaking process of evaluating and choosing a lab has borne fruit, the failing sample can be sent off to be torn down into its constituent parts, deconstructed in detail in order to determine the root cause of the given failure and allowing constructive corrective action to be taken to prevent any further fallout. Choosing a lab for outsourcing electronics failure analysis services is certainly a daunting task, especially given all the various elements that must be considered; fortunately, some labs have taken much of the effort out of the equation, collecting a great deal of the necessary information in one place for the convenience of any potential customers.

Inevitably, in any product’s life cycle, there will arise an obstacle that may seem insurmountable: products may experience unexpected levels of inexplicable malfunctions after hitting store shelves, low production yields may wipe out any hope of profitability, or any of a number of other issues can rear their heads. When faced with such gremlins, manufacturers often struggle to find the best approach for solving their woes – without being able to pin down the problem, finding a solution is impossible. External failure analysis services can often be invaluable in such situations; however, the task of choosing a lab – and providing them with the information needed to ensure their success – can be difficult as well. Fortunately there are some tips that can help in the process of hiring an electronics failure analysis service, to ensure that the necessary results are obtained.

The first step that must be taken whenever outsourcing an FA job – even before a lab has been selected – is the construction of a detailed background information set describing the problem. Designers and manufacturers have access to a wealth of information: design specifications, drawings and computer models of the device’s layout, in-depth schematics, and simulation data are all things that a manufacturer may take for granted that are unavailable to external labs, yet may be key to uncovering the root cause of the problem. Similarly, manufacturers generally have the “whole picture” of how a device may have failed – they know how long the sample was in service, whether the failure of the device appears to be an isolated incident or representative of an epidemic, and so forth. Like any other scientific pursuit, failure analysis does not occur in a vacuum (except for the rare cases calling for failure analysis on vacuum cleaners); this type of background knowledge can be vital in determining the course an analysis may take, as an analyst may be cued to look for different root causes depending on the history of a device. As an example, environmental contamination may be a relatively unlikely cause for failure for a device that broke down in a tightly-controlled cleanroom; however, for a cheap consumer device like a remote control or an MP3 player, there are innumerable sources for potential contaminants to force their way into the circuits of a device.

Just as an in-depth history is vital for an analyst to quickly and efficiently drive a failure to resolution, an accurate, concise description of the issue plaguing a device is absolutely necessary. Just like taking a car into the mechanic, a failure analyst can diagnose an issue more efficiently if the problem description is more in-depth than “funny noise when driving”. Any problem reported by the client will, at some point, need to be translated into a testable condition by the FA team – providing the most complete description of the problem possible will allow an analyst to properly design a test program to isolate the root cause of the failure. In the same vein, it is often beneficial to provide a sample that is working properly along with the failing sample; by giving a golden unit to compare against, it is possible to employ techniques that greatly increase sensitivity to small defects, by allowing analysts to separate normally-occurring phenomena from those stemming from the failure of the device. For so-called “functional failures”, in which cases a device is still mostly operational (i.e. is not short- or open-circuited), but may not give a correct output, the use of a correlation sample is almost mandatory, since the normal operation of such a device will often create many of the indicators (photoemission, thermal hot-spots, et cetera) that analysts use to isolate failure sites – in order to find the true failure, they must have some way of filtering out those sites that are inherent to the normal operation of the device.

Once the initial data about a failure has been collected and filtered for any unnecessary or sensitive data in order to ensure the best results from a given failure analysis job, the next step is to evaluate and choose a lab to handle the project. In part two of this series, we will discuss the sorts of things to look for when choosing a lab, in order to provide the greatest probability that a given issue will be resolved successfully.