The identity and source of foreign matter can be contentious since it may impact the release of a product, trigger recalls, or result in disputes between suppliers, manufactures, and insurance carriers. Since foreign matter in bulk powders can originate at virtually any stage of production, source attribution can be both complex and challenging. Regardless of the type of investigation, the first goal in any alleged foreign matter incident is to identify it with certainty. Contaminant materials can consist of almost anything you might imagine, and many things you probably couldn’t. Over the past three decades, we have identified thousands of unknown materials that have included rodents, reptiles and insects from around the world, machine parts, glass, metals, botanicals, teeth, cremains, illicit drugs, etc.
An unambiguous identification provides factual information upon which a mitigation strategy can be developed. Common examples of alleged contaminants include: an improperly performing material (e.g., caking or charring), a poorly purified raw ingredient (e.g., wheat germ in white flour), or a true foreign material (e.g. metal, glass, plastic). These demonstrable scientific findings can be applied by an industrial hygiene officer to assess health risks, a claim adjuster to validate a claim, or a plant manager to identify a source.
In many instances, the source of the undesirable material may not be obvious or there may be a dispute over its origin. In such cases, a more in-depth analysis can frequently provide clues than can narrow or eventually identify a true source. In one case, a dispute arose over the origin of fine dark particles that were showing up in a mixture of ingredients. Our lab was brought in as a third-party “referee” after the lab for each supplier identified the other’s product as the source of the specks. One supplier provided spray-dried milk and the other sucrose powder. The preparation and subsequent analysis of a cross section showed that they contained a minute core of pristine, uncharred material. Microscopical and microchemical analysis showed that the core of these particles consisted of calcium, protein, and lactose. These are three of the components of spray-dried milk. This decisive answer brought a definitive resolution to the matter.
Contamination investigations have their roots in traditional criminal forensic science. Since the mid-19th century, scientific investigators have searched crime scenes, clothing, and other substrates for traces of material that don’t belong. Such materials commonly include hair, fibers, wood, soil, glass, and dust, but may be comprised of virtually anything. Such traces were collected, sent to a crime laboratory for identification, and the results were used to direct investigations. When possible sources of these materials were identified, they would be compared back to the crime scene material. Probative results would be conveyed to a court and jury.
In one investigation, the retrial of a convicted murderer, detectives had collected debris from the carpet of the victim’s home. Among the debris from the carpet, the crime laboratory observed microscopic “footballs” that were adhering to fibers (Figure 1). Our lab’s founder, Skip Palenik, identified these tiny “footballs” as droplets of wet spray paint that had hit and then dried onto a fiber. By coincidence, the convicted defendant had petitioned the court to retain his car. A re-investigation of the suspect’s vehicle showed that it too was loaded with chemically identical “footballs.” Detectives found the source of these unusual particles when they looked up and saw that the headliner of the vehicle had been spray painted. A comparison of the fibers and paint shed from the headliner turned out to be indistinguishable from those recovered from the carpet of the victim’s home. This further evidence was used by prosecutors to obtain a re-conviction. Through service to both forensic and industrial clients, we have found that the same principles used in forensic work can be applied to advance the investigation and sourcing of alleged contaminants.
What is it?
Establishing probative facts about the identity and properties of suspected foreign material is paramount to any investigation. This can be challenging since foreign material (i.e., trace evidence) can consist of almost anything that can be broken down into dust. Some of the more commonly encountered categories of material we have encountered as foreign matter are listed in Figure 3.
Some of these, particularly when they are larger, can be recognized without specialized expertise. For example, those involved directly in manufacturing may have “hunches” about the origin of a material based upon color or texture. However, as the particles become smaller, more complex, or more unusual, a specific identification becomes increasingly challenging. For example, on-site quality control labs are designed to routinely make specific measurements. However, few industrial plants are equipped to handle the rigorous identification of unknown substances, particularly when they are small or even microscopic. The particle shown in Figure 3a is a sample of glass discovered in a dry cereal mixture. This particle became the center of a multi-year lawsuit. We were asked to conduct detailed analyses of this piece of glass to place constraints on its possible origin. For example, was it from a bottle, a window, a vehicle, or something else?
Figure 3: (a) An alleged glass particle shows surface features that are suggestive of glass. However, when this particle was placed into a drop of water (b) and started to dissolve, it becomes clear that the material is something other than glass. Follow-up analyses by Raman microspectroscopy confirmed that this was a glassy particle of sucrose (sugar).
After documenting the sample as it appeared under a microscope, it was gently drawn into a drop of water to remove some of the residue on its surface. Upon contact with the water, the particle began to dissolve (Figure 3b). It was quickly removed from the drop and allowed to dry. Analysis of the dissolved substance revealed it was sucrose in its glassy state, such as one encounter in lollipop, rather than a silicate. This indisputable fact brought the litigation to a swift conclusion. This example illustrates the importance of establishing a firm identification early in an investigation. It also demonstrates the way in which definitive results can minimize risk and reduce financial expenditures.
While foreign matter may be composed of only a single component, experience has shown they are often mixtures. These more complex mixtures can provide a great deal of information about the specific origin of the contamination. On the line at a quality control checkpoint, red and black colored fragments of material started appearing periodically. The color and size (millimeter to centimeter) of these particles stood out against the white color of the formulation mix.
An initial analysis by light microscopy showed that these actually consisted of multiple components. The various components were isolated and separated for specific identification by chemical and physical methods. The main ingredient of the unidentified material was identified as polyethylene. The red color of the fragments originated from a pigment used to color the plastic. A closer examination by polarized light microscopy revealed that the particles were not a film or bulk solid, but rather a compressed aggregate of still finer plastic particles. The black particles actually contained the same red polymer. However, the red color was obscured by the fine metal shavings that gave the particles their overall dark coloration. Close inspection of the population of red particles showed fine metal particles were also present in them but at lower a concentration. Elemental analysis of these fine metal particles by energy dispersive x-ray spectroscopy in a scanning electron microscope showed that they were abraded from a source of 316-series stainless steel.
We were able to inform our client that the source of this material was a red polyethylene surface that was being abraded by a source of 316 stainless steel. A piece of rotating machinery containing numerous red plastic polyethylene caps was located. A review of maintenance records found that a recently replaced stainless steel bolt in the vicinity did not have a sufficient clearance from the rotating part. An inspection of the plastic caps showed signs of visible abrasion and darkening. A comparison of exemplar samples from the worn caps showed that they were composed of polyethylene containing the same red pigment.
Over the years, our sourcing investigations have identified contaminants from suppliers, manufacturing equipment, and consumers (both intentional and unintentional). In many cases, these analyses have been able to identify a single, specific source. While high-quality analysis and data is important, it is just as important for the lab and the plant to have open dialog during the investigation. When used in concert, seemingly irrelevant observations in both the lab and at the manufacturing facility often lead to solutions when thoughtfully evaluated.
Early efforts to ensure a sample is preserved will help to maximize the relevant information that can be obtained from it. While many analyses can be performed months or even years after a sample has been collected, a poorly collected sample can irreversibly reduce the amount of information available to an analyst. For example, mold (unrelated to the actual sample) can cover or even consume certain samples. Similarly, there is often a temptation for quality control or management staff to probe a sample themselves. One client smashed a sample with a hammer to test its hardness, and another lit a sample on fire in an attempt to identify the material by its burning odor. While we were eventually able to identify the remains of both of these samples and answer the clients’ questions, the depth of our analyses were limited by these prior decisions.
Collection. We are often asked by our clients what type of sample to collect and how much to collect. The actual quantity of sample is generally not important. A microanalytical laboratory should be able to work with materials smaller than one can see with the unaided eye. The overriding goal is to collect a sample that is representative of the observed problem. If dark specs are observed, be sure that the collected sample contains the dark specs. If potential foreign matter is composed of materials with different shapes, textures, or colors, be sure that the submitted sample contains examples of each.
Packaging. The particle(s) should be packed in an appropriately-sized container. A small particle should be isolated from the product matrix and packaged in a bag that is small enough that the material of interest can be readily located. A small particle packaged in an enormous container can be difficult and, at time, impossible to locate and may be confused with other fine contaminants that may be present in the container itself. Glass is generally a better container than plastic (unless the particle of interest is composed of glass) since static charge in the container, which can cause particles to stick to walls, are reduced in glass. If the material of interest or the matrix in which it is found is not (or may not be) shelf-stable, ensure that it is stored and shipped cold. While it should go without saying, it is critical that the container housing the particle is sealed.
Documentation. Although the complaint material may be obvious to eyes familiar with the problem, it may not be to the scientist who is conducting the analysis. An image will ensure that the laboratory focuses on the correct material. While a photo through an inspection microscope is helpful, even a quick picture taken with a cell phone is generally sufficient.
Control samples. The submission of retain samples, relevant ingredient samples, and product ingredient/formulation is not always necessary. However, such materials and lists can be useful for several reasons. Unaffected control samples can be used to ensure that chemical testing is not providing a false reaction. Retain samples from current or prior lots can be examined to determine the extent and source of a problem. Formulation details are often of value when attempting to identify or decide between potential sources.
Preserve. Ensure that a chain of custody is started and preserved. Samples should be sealed and the seals should be initialed and dated. A chain-of-custody form should be maintained with the sample. The chain-of-custody is intended to memorialize details about the collection and custody of a sample through its history. This is particularly important in cases that involve potential litigation.
The range of samples that can be encountered is immense, and sampling decisions can impact the amount of information that can be obtained. The information provided in this article is intended to serve as a starting point, but is by no means inclusive of all possible sampling considerations. Another article that provides further information on the collection and preservation of samples can be requested from the author (Palenik, 2013). Ultimately, if there are questions about the collection or packaging of a sample, it is best to discuss questions first. Laboratories that specialize in the identification of unknown materials are generally happy to discuss specific details of sampling with their clients.
Christopher Palenik, Ph.D., is vice president of Microtrace, which specializes in the identification and sourcing of unknown materials for both industrial and criminal forensic investigations. Palenik has spent much of his life examining using microscopy to solve problems using scientific analysis. He has carried out research in laboratories around the world, including the German Federal Police Crime Laboratory, the IRS National Forensic Laboratory, and the FBI. For more information, visit www.microtrace.com.
Forensic Files Episode “Paint Ball” https://www.youtube.com/watch?v=iVxpynOWqEc
Palenik, C.S. (2013) “Key Considerations for Consumer Complaint Sample Analysis.” Customer Relationship Magazine, Summer 2013, pp. 15-18.