Damage Deconstruction
Failure Analysis of Burned Printed Circuit Board Assemblies
When it comes to contamination analysis, things are not always as they appear. by Claire Brennan, Ph.D.

In the failure analysis of electronics assemblies, we are often asked to perform a failure analysis on hardware that has undergone a significant thermal event. Hardware might be burned, melted or covered in debris. Determining a root cause for failure can be extremely difficult when the hardware itself is so damaged that much of the evidence has been destroyed. So, what can you do? Like many things, it depends. The success of the failure analysis depends on the overall degree of damage, the amount and type of secondary damage, and the history of the part. Over the years, we have developed some tools and techniques to get the most out of these challenging failure analysis requests.

The first step in these types of investigations is to manage expectations. Most customers will understand that much of the evidence was destroyed during the thermal event failure and that root cause analysis will be very difficult. It is important to discuss what types of information can be gained, however, and what may not be possible. It is also critical to get as much information as possible about the history of the part and any details about the failure itself. This proactive discussion will help lead the investigation in the “right” direction and avoid going down a path that will not yield useful information. For example, if some of the metallic hardware is corroded, it is important to know the storage environment of the unit, not just temperature and humidity, but also the amount of time the unit was stored and its relative orientation. The product history information is useful to separate damage caused by the failure versus damage that occurred before or after the failure.

As with all types of failure analysis, it is critical to document everything with high-resolution photos and/or optical microscopy at every stage of the investigation. If disassembly needs to be done, it is critical photos are taken before and after the disassembly. In some cases, it can be helpful to take a video of specific disassembly steps. This will help capture “intangible” information such as if the part made a certain noise, or if certain parts were easy or difficult to move. Ensure photos are taken before the part is moved, since heavily damaged hardware can cause pieces to become loose, shift, or fall out during movement. When opening compartments, such as a chassis, make sure to place a plastic bag underneath the part to catch any debris and/or loose parts.

The initial photo-taking stage is usually the best time to gather material samples for analysis. This analysis can confirm the presence of certain materials in the construction (for example, was the correct adhesive used on a particular component?) or to identify possible contaminants that may have contributed to the failure. It is good practice to label each sample with the location where it was taken, often including a corresponding photo. Even if some of the debris appears to have an obvious origin (e.g., copper from the internal layers of the PCB or solder ball from a component’s solder joint), it is good to confirm these assumptions. When it comes to contamination analysis, things are not always as they appear. This debris analysis can also give clues about the temperatures reached during the failure based on the softening point and melting points of various compounds and alloys.

Depending on the type of debris, analysis can be carried out through x-ray fluorescence (XRF), scanning electron microscopy with energy dispersive x-ray spectroscopy (SEM/EDS), Fourier transform infrared spectroscopy (FTIR), or x-ray diffraction (XRD). Each of these analytical techniques is suited for a specific type of sample and will provide different information. TABLE 1 is a short summary of the more common techniques for contamination analysis of solid materials, including the capabilities and limitations of each technique. With heavily thermally damaged electronics hardware, the contamination analysis process will likely be complicated by the large amount of carbon-based debris. Since the high heat of the failure can alter some materials, complementary techniques might be needed, such as SEM/EDS and FTIR, to determine the different materials. For example, if a dark-colored, metallic-looking material is collected on the failed hardware, it can first be analyzed with the SEM/EDS. SEM may show the material charges when exposed to the electron beam, indicating it is not purely metallic, and EDS analysis may show significant carbon and oxygen. If the material is then analyzed using FTIR, it could be found to resemble the FTIR spectrum for an adhesive material used on the assembly. By running a “new” sample of the adhesive against the unknown material from the failed hardware, it could be determined the temperature during the failure was hot enough to degrade the adhesive material, or perhaps an incorrect adhesive was used. The wide variety of materials used in the PCBA necessitates multiple analytical tools for identifying different types of debris.

Table 1. Summary of Analytical Tools Used in Contamination Analysis
Table 1. Summary of Analytical Tools Used in Contamination Analysis
Figure 1. 3-D CT scan images of several layers of a PCBA from the top-down are shown in a) - d)
Figure 1. 3-D CT scan images of several layers of a PCBA from the top-down are shown in a) – d). 3-D CT scan image in e) shows a sideview of the PCBA, where damage and delamination can be seen in several layers. Dark contrast in the images indicates lower x-ray density material, and the dark regions within the PCBA above show excessive heat damage to the PCB.
Figure 3. Metallograph cross-sectional images of a failed PCBA in a) and b), showing internal damage to the layers of the board
Figure 2. Metallograph cross-sectional images of a failed PCBA in a) and b), showing internal damage to the layers of the board. This technique permits visualization and characterization of the damage to the board, including damage to the glass fibers and copper layers as shown here.

After the initial documentation of the as-received hardware and collection of samples, the team should have a short meeting to reevaluate expectations and develop an analysis plan. Typically, the next analysis step will be some sort of nondestructive testing or nondestructive examination (NDT/NDE). This could include 2-D x-ray, 3-D computed tomography (CT), acoustic microscopy, thermal imaging, various types of electrical testing, and optical microscopy (bright field, dark field, polarized light microscopy, etc.) (FIGURE 1). X-ray radiography (both 2-D and 3-D CT scanning) is a powerful tool for examining the internals of failed electronic components, as well as pinpointing and quantifying the amount of damage in a printed circuit board assembly. For example, for heavily damaged PCBs, documenting how many layers are damaged and which layers have the most severe damage provides useful information, since this can help pinpoint where the failure started (FIGURE 2). Identifying internal damage to individual electronic components, such as surface mount components, can be difficult since some of this damage might be secondary to the failure. It is important not to focus too much on damage, which is secondary to the primary failure, but documentation of this type of damage is still necessary to get a complete picture of the failure.

Acoustic microscopy and thermal imaging are other tools used to image and document internal and external damage. Acoustic microscopy has the added advantage that it can pinpoint where the damage is located in a specific device. Depending on the extent of the damage, electrical testing may not be needed. However, for multiple damaged components on a failed printed circuit board, electrical testing can be used to map the extent of damage on the board, as well as diagnose the type of electrical damage that occurred (overstress, overcurrent, short, open, etc.). It is important to discuss any findings from nondestructive testing before moving on to any destructive analysis.

Destructive analysis can encompass a variety of activities, including physical decapsulation of components, cutting/sectioning in regions of interest, cross-sectioning printed circuit boards, and other techniques. Decapsulation and cross-sectioning may be used to verify the components or PCBs were made according to the applicable specification or drawing. For example, cross-sectioning can verify PCB layer thicknesses, as well as other features such as layer finishes and plating. Care should be exercised during this phase of the investigation, taking the necessary precautions to not introduce any artifacts that would confound or confuse the destructive analysis results.

Table 2. Summary of Tools Used in Decapsulation
Table 2. Summary of Tools Used in Decapsulation

Decapsulation of specific components is often a slow and tedious process, depending on the encapsulation/potting materials used and the chemicals used to remove them. Often, commercially available decapsulation chemicals can be used at room or elevated temperature. Alternatively, acids (such as sulfuric, nitric, etc.) can be used alone or mixed at elevated temperature to remove the encapsulation materials. Another tool that might be used in conjunction with chemical techniques is plasma etching or ion milling. Depending on the equipment used, these techniques can be focused or broad in nature, but are typically very slow. A summary of the techniques/tools used in decapsulation of electronic components is listed in TABLE 2. Each technique has its advantages and drawbacks, but one of the most important considerations is always how the decapsulation process alters the internals of the component. For example, some solder alloys are subject to chemical attack from sulfuric acid, so it is important to keep that in mind when analyzing the internals of the part.

Although failure analysis of thermally damaged PCBAs can be challenging, a variety of techniques can be used to get critical information about the failure. A combination of visual inspection, material/contamination analysis, nondestructive evaluation and destructive techniques can achieve probable root cause for the failure. Even if root cause cannot be determined, it might be possible to rule out several scenarios based on the hardware. Throughout the process, documentation of the condition of the hardware and communication with the customer will be needed to develop a complete picture of the failure and manage expectations. It is critical to help the customer gain as much information about the failure as possible, so this information can then be used to make future design and manufacturing decisions/recommendations.

Claire Brennan, Ph.D. , is staff engineer, Materials & Process Engineering, at Collins Aerospace (collins.com).