Welcome to our solutions for nanoanalysis podcast brought to you by Bruker Nanoanalytics. We look forward to bringing you a new podcast regularly. My name is Cody Morton. I'm a marketing communications specialist at Bruker Nanoanalytics and an information enthusiast. If you like to learn from specialists in their field and hear what technologies are solving their problems, you will enjoy this podcast.

Every session, we will focus on a different problem that our colleagues face in the lab and in the field. Some of the solutions will be a variation of ideas you may have heard before or even worked with. We will bring you these topics in a new and interesting way and introduce you to updated and thought provoking results. We will talk about how the problem was dealt with in the past and what we're doing to solve the problem now and perhaps even envision future solutions. Join us as we talk solutions with a variety of scientists and technicians in many different industries in the Solutions for Nanoanalysis podcast.

Today, we're talking foreign objects found in food. I was introduced to the idea after, high school reading the book, The Jungle. This book written by Upton Sinclair in nineteen o six was presented to president Theodore Roosevelt. After he read Sinclair's novel, he was motivated to act. Roosevelt commissioned Charles p Neil, commissioner of the Bureau of Labor, and James Bronson Reynolds, a social worker, to investigate the meat packing industry in Chicago.

After reading the Neil and Reynolds report, Roosevelt wrote a message to congress. In part, it said, it is imperatively necessary in the interest of health and of decency that they should be the rules should be radically changed. Under the existing law, it is wholly impossible to secure satisfactory results. As a result, two bills, the Pure Food and Drugs Act of nineteen o six and the Federal Meat Inspection Act of nineteen o six, moved quickly through congress. Roosevelt wasted no time signing both into law.

The new legislation made it illegal to manufacture, sell, or transport harmful food and assured that livestock was processed in a sanitary fashion. A hundred years later, we are still talking about food safety. We've come a long way since then, but sensational news stories are still in the headlines. The more sensational food stories we've heard about are usually fraudulent, such as the case of a woman who found a severed finger in her chili back in twenty o five. It was discovered that she actually planted it in order to sue the company for a substantial amount of money.

Even though her case was fraudulent and she was sentenced to prison for her actions, it cost the company millions of dollars in lost revenue from the bad press. That's clearly a dramatic and newsworthy case. But today, we're gonna find out about some of the more typical foreign objects found in food and how handheld XRF helps manufacturers solve issues around this very real problem. Our speaker today is Kim Russell, who is market segment manager for food safety and agriculture at Bruker Nanoanalytics. Her product solution focus is handheld XRF, TXRF, and micro XRF.

Kim, thank you so much for speaking with us today about this problem. Before we talk solutions, tell us a little about your background. What made you interested in working for an analytical instrument manufacturer?

Hi, Cody. Well, thank you for this opportunity. So after completing my Master's degree in chemistry with a focus on analytical instrumentation, I was hired to set up and manage a technical service lab at a radiopharmaceutical company. The objective of the lab was to help with both production and customer problems. So once I determined the type of equipment we needed, I began to invite analytical instrument manufacturers to come to the lab to demonstrate their products to me.

During this process I learned about the job of an applications scientist. They traveled a lot, they got to visit all sorts of companies and find out how they did things, and they helped the people in these companies do their jobs better. The applications people I met said no day was ever the same and they were never bored because there was always something new to learn. I'm a curious person by nature and began to consider changing jobs if the opportunity presented itself.

Well, does sound exciting, but how did you end up working for an instrument company?

Well, once I'd been with that company for a year, I was given permission to go to my first technical conference and to take a short course. They wanted me to go to PitCon to check out new equipment. That was way back in 1982 when it was in Atlantic City. I decided to take a short course on something I didn't know anything about. I chose a two day Introduction to X-ray Fluorescence Spectrometry a short course taught by two of the leading applications scientists at the time, Ron Jenkins and John Croak.

I really enjoyed it and I hoped I could put it to work one day. I knew I was going to be at Pitcon for a full week, so I decided to take a look at the employment opportunities because there weren't many problems for the tech service lab I set up to solve and I was getting bored. There was actually a job posted for an applications chemist close to my home. My big challenge in terms of qualifications was that it was for XRF and they wanted two years experience. I only had two days worth that weekend's workshop, none of which was hands on.

So I decided to give it a try anyway because the worst they could say was no, but the best they could say was you're hired. And that's what happened. I got hired and I began my career working in XRF as an applications chemist for a wavelength dispersive XRF manufacturer.

Is there much difference in the kind of XRF you work with now compared to back in the early eighties? I would think so.

Yes. Well, like all elemental analysis techniques, there have been numerous improvements since back then. In general, XRF identifies elements in a sample by the peaks in a spectrum that occur when the elements are excited by a source. XRF can determine the concentration of those elements using information from the height of their peaks. Wavelength dispersive peaks are in wavelength units while energy dispersive XRF peaks are in energy units.

I was working with very expensive state of the art wavelength instrumentation back then but the analyzers were as large as today's minivans are, and they had very unforgiving software. For each sample application, I had to determine all the optimum measurement conditions myself and manually enter them into the microprocessor. One typo and I had to start all over again. The samples had to be very carefully prepared and placed in a sample chamber in the analyzer. And I had to use very large reference books full of wavelength tables to look up which peaks represented which elements.

I had to hand enter all of the data for reference samples and develop a calibration curve for every element in a sample set and then carefully enter that into the microprocessor. Now wavelength analyzers have come a long way since then in terms of a reduction in size and cost as well as ease of use. There are no more unforgiving microprocesses to deal with and you don't need to use the big reference books with wavelength tables anymore. That information is all part of the analyzer now. They are also more sensitive in terms of measuring lighter elements and lower concentrations in samples.

I actually started working with handheld XRF in 02/2003, after they had been on the market for about ten years. The very early handheld XRFs focused on heavy metal analysis because elements like lead were very straightforward to measure in soil, paint, and other flat surfaces with a radioactive isotope as an excitation source. The commercial use of X-ray tubes as an excitation source in the early 2000s really increased the value of handheld XRF by opening up the range of applications. And that included metals, alloys, plastics, geological materials, art objects and just so much more. When I first started using them, I was simply amazed that something the size of a handheld power drill could essentially analyze elements like the large units I learned on back in the eighties.

Amazingly, measurement conditions were already optimized for various sample types, calibrations were built into the analyzer, minimal or no sample preparation was required, some measurements only took seconds and handheld XRF's were considerably less expensive. In many cases handheld XRF is simply a point and shoot measurement. A key feature about XRF as an elemental analysis technique is that it's nondestructive. You still have your sample after the analysis. So if you need it again for any future reason, you still have it.

It sounds a lot like me trying to explain cell phones technology getting better to my kids who have never seen a rotary phone. I would hate to think about all of those reentering data that you must have had to do at the time.

Yes. It was painful.

What is the primary use of handheld XRF's now?

Well, the analysis of metals and alloys is the largest market. It's very nuanced, but in general it covers the identification of scrap metal for purchase or sale, the evaluation of precious metals, confirmation that parts are fabricated with the correct alloys, and positive material identification of alloy grades or that's sometimes referred to as PMI. So PMI of alloys is critical in numerous industries. For example, oil refineries are typically located in areas of high temperature and humidity and the piping used for processing oil is on a very, very large scale and the process itself can create an environment of extreme heat, humidity, pressure, and vibration. So it's absolutely critical that they're using the correct alloys formulated to stand up to those conditions.

And it's not just the piping, it's the connectors, screws, bolts and weld materials. All of the components must be made from the correct alloy grades or serious problems both for production and safety can occur. Handheld XRF is the ideal tool for PMI inspectors who are responsible for ensuring conformity to design specs. Not only does a handheld XRF have an alloy calibration built into the analyzer but it also has a library with information on over a thousand alloy grades. So when the handheld XRF is analyzing an alloy sample not only does it determine the concentration of that sample but it also compares it to the onboard library to identify which alloy grade it is.

And the measurement identifies the alloy grade within seconds. Essentially, fast answers with handheld XRF can prevent big problems, and the return on investment can be almost immediate.

Well, Kim, I know that we'd said we were going to talk about food manufacturers, and it sounds like there are a lot of different ways that handheld XRF can help in all sorts of industries. But how do food manufacturers use handheld XRF to help with the problem of foreign objects found in their products?

Good question. Alright. When a foreign object is found, the product containing it is removed. Sometimes the production line even has to be shut down, is very costly. The challenge is to keep production going, but to make sure it's producing safe and high quality product.

The two primary questions they have are: What is that object? And Where did it come from? The sooner those two questions are answered, the sooner safe production can resume and the less costly the situation is. So handheld XRF can reduce the time it takes to identify the physical contaminant found and determine its source. Typical foreign objects found during production are not fingers.

They are bits of ceramic, glass, plastic or one of the most common being metal in the form of shards or shavings. Food manufacturers already have multiple measures in place to prevent these objects from getting into the product. They have large in process metal detectors or x-ray inspection systems to detect foreign objects and in some cases they have large magnets to remove the ferrous metals. The most likely reason these detected objects enter the process is from wear and tear of processing equipment and occasionally from starting materials. When a metal piece is found, handheld XRF can quickly identify it using the procedure I mentioned earlier about alloy identification.

The piece is measured, the alloy grade is quickly identified using the Envoy library. Once they know the alloy grade is a metal object, they can begin to determine its source. For instance, if it's a three sixteen stainless, they can investigate which components of the process are made of that alloy grade and then inspect them for any wear and tear to find the source.

I have questions about the onboard library. We're gonna talk about that in a minute, aren't we?

Yes. Yes. Of course. That that's

that's the fingerprinting I've heard about. Right?

Yes.

Okay. Good. Because it sounds really fast once you have all of that. Is it always that simple?

Well, I wish it were, and I'm sure they do too. But other factors can be involved. In a way, their environmental conditions are similar to those of the oil refinery ones I mentioned earlier. Most equipment is constantly moving and frequently exposed to high temperatures, humidity and vibration. For food production, the material passing through can change form and that would mean it could be powders, solids or liquids.

Now many of the equipment metal contaminant sources, which are typically augers, conveyors, grinders or cutters, mixers, roller mills, and sorters, or packagers, are made of the same or similar alloy grades. They are selected for cleanliness as a food contact material and to operate functionally in that kind of environment. These alloy grades include three zero four, three 16, and four thirty steels. In large operations, or those which have multiple components with the same alloy grade, it takes a little more time and effort to determine the source of the contaminant. But it's not complicated.

The good thing is that even if components are made of the same alloy grade, the one used for a particular component does have a unique spectral fingerprint which can usually be identified by matching it to a spectral fingerprint from the library of production line equipment.

That sounds like a criminal fingerprint ID system we see on CSI shows. But what does a spectral fingerprint look like, and does processing equipment come with one?

Well, you know, many people refer to this as the FBI application, an acronym for foreign body investigation. So thinking of it as a CSI type of thing makes perfect sense. You remember earlier in our conversation, I told you how XRF identifies elements in a sample. It detects the peaks in the form of a spectrum that occur when elements are excited by a source. And it can determine the concentration of those elements using information from the height of their peaks.

That is the spectral fingerprint. So when you're taking an XRF measurement of a component sample on the production line, even for less than thirty seconds, you are automatically collecting and storing its spectral fingerprint. Now I'm not aware of any components that come with spectral fingerprints when purchased. However, our customers do collect their own spectral fingerprints and create their own production libraries. It's a significant investment in time and planning but it pays off quickly when it matters.

In fact, one of our newer customers said that their first handheld XRF purchase paid for itself in the first four months of receiving the unit in downtime alone by pointing them in the direction of the equipment that was damaged.

So how do you collect the fingerprints and create a library?

Well, the procedure of collecting spectral fingerprints for a library is straightforward. An XRF measurement is taken of all food contact devices and components on a given production line. That's the part that takes time, because there are typically many, many parts. Even though they are only measuring each piece for about thirty seconds, it adds up. But before any measurements are taken, the library needs to be planned.

For instance, the naming convention of the parts needs to make sense to the particular process. For example, you could use names like air deflector cooked production cooler or part c tempered wheat screw conveyor or shingle cooked production cooler or skin a continuous cooker. Then you have to be consistent with the naming convention. By the way, there should also be consistency in the naming convention of contaminants, usually something which incorporates the type it is, whether it's metal, plastic, ceramic, the date, the time, and the production line where it was found. The organization or structure of the library also requires planning, and again it needs to make sense to a given process.

You want the library to contain all the XRF spectral fingerprints collected, and you also want to be able to store the spectral fingerprints of any foreign objects found. To start, a main folder should be created in a PC documents folder to contain all the test data. Next, if there is more than one production line, subfolders for each line should be created. Then each production line folder should contain a subfolder for processing equipment fingerprints and another one for physical contaminant fingerprints. These can even be further subdivided for various physical contaminant types if you want, like metals, plastics, ceramics, or glass.

How do the customers figure out which spectral fingerprint matches the foreign object they found?

Well, matching fingerprints is straightforward, but it takes a little training. We have a software program called ARTACS, which is an advanced spectral viewing, matching, and data analysis software package. It's used to both store and link information. RTX makes the management of large datasets like food production spectral fingerprint libraries easy. So when a physical contaminant is found on a production line that lines library of spectral fingerprints is transferred to an RTACS project folder.

The physical contaminant spectral fingerprint is then transferred to the same project folder which contains all the possible sources on that production line. The contaminants fingerprint is then compared using the RTAC's MATCH program to find the most likely source of the contaminant in that project folder. You basically just click on MATCH and then RTAC searches through the library to find any hits that match. It then gives a probability that the match is correct. In some cases there can be like a hundred hits with a correlation of 90% or better, but there are just a few with a correlation of 99% or better.

So it's a great matching program. If needed there's a step that can be used for further confirmation. You can overlay the spectral fingerprint of the physical contaminant and the suspect source and zoom in on all features of the spectra to make a visual comparison. All of that takes up to about five minutes to do. So it saves considerable time in determining the source of a physical contaminant.

Now some of our customers with smaller production capacity only have one handheld XRF, but our larger ones have multiple handhelds at multiple facilities. And they typically have a quality person who is kind of considered their guru of the process who others within the company look to for help, to either set up the process or to train new users. Since this kind of information is typically highly proprietary to a company, they don't want our assistance. They want an internal person.

I like the term suspect source. How small of a physical contaminant piece can XRF measure to match to a suspect source?

That's a really good question. Handheld XRF works well for small samples, even down to a millimeter in size. But sometimes physical contaminants are even smaller than that, especially those that have been found in end products by customers. In those cases, Micro XRF technology is ideal. Like handheld XRF, micro XRF identifies elements by the peaks in their spectra and determines concentrations using information from the height of the element's peaks.

The big difference is micro XRF is used for very small samples or small spots on the micron level, as the name would imply. Bruker has a benchtop MicroXRF, which can analyze samples as small as 100 micrometers and a lab MicroXRF, which can analyze sample spots at less than 20 micrometers. Consequently, micro XRF is used for identifying very small micron level physical contaminants found in food.

Okay. Can you give me an example?

Well, sure. In one case, a glass library was developed with the Micro XRF software of the various jars and bottles used by a manufacturer. Each jar had a different compositional makeup. Therefore, when a very small glass particle was found, they could measure that piece and compare it to their library to determine if the glass contaminant was from one of the jars or came from the processing equipment. The same thing can be done with alloys.

A real value bonus to using micro XRF is that elemental distribution maps of objects can be created for further visual inspection. For example, maps of shards of metal embedded in potatoes from mechanical harvesters, conveyors, or sorters can be investigated to look at patterns to further help determine the source. The downsides to using micro XRF in comparison to handheld XRF are that it takes much longer to collect data on the order of hours compared to seconds and the purchase price can be up to 10 times out of a handheld XRF. Still, they provide high value, critical information when it counts and are worth the investment.

Okay. Does Bruker have any other technologies that help identify foreign objects?

Yes. Our benchtop FTIR spectrometers complement our XRF technologies in that they work well for identifying unknown plastics, rubber and other organic materials. And our FTIR microscopes are ideal for analyzing micro particles such as microplastics found in foods. All of Bruker's product solutions essentially help people do their jobs better. Whichever technologies are used, the bottom line is our product solutions enhance physical contaminant quality programs.

They help with preventative maintenance of production equipment. They help reduce future delays in production. They help increase confidence in suppliers. They help assist with false claim investigations. And they help optimize production's overall risk management.

Thanks, Kim. What would you recommend a listener do next to get more information on these Bruker solutions for identifying and sourcing foreign materials found in food?

Well, we have multiple resources for this. We have a two part video on the Bruker Nanoanalytics YouTube channel. One is titled physical contaminant ID in food production, and the other is physical contaminant spectral fingerprinting for food production. We have a webinar on July 1 entitled Quickly Identify and Source Foreign Material Contaminants Found in Food Products with Handheld XRF, and that will be available later on demand. We also have application notes on using handheld XRF and micro XRF for this application.

And, of course, we have application and product descriptions on all of Bruker Solutions on our website.

Terrific. We'll make this information available in our show notes for the podcast. Kim, thanks for speaking with us today about how Bruker Nanoanalytics product solutions help solve problems associated with foreign objects found in food. I know that this is a topic that's difficult for some of our customers to really talk about due to concerns of food safety is always at the top of their minds. But I do hope our listeners will check out our website and some of the other sources you mentioned.

It's been great talking to you and look forward to talking to you again soon.

My pleasure, Cody. Thanks again for the opportunity.

Thank you to our speakers today. If you would like more information about today's topic or submit a topic idea, please email info.bna.us@Bruker.com. You can also check out more information in today's show notes. Join us next time as we look at a new solution with more scientists and technicians in all sorts of industries.