Fluorescence spectroscopy detects mold in bourbon grains

Ahh, there’s nothing like the rich, balanced flavor of a good bourbon, with strong notes of vanilla, oak and caramel. The harmony between dry and sweet fades quickly as you slowly consume a swig.

Changing consumer preferences is fueling its growing popularity into a billion-dollar boom, accounting for almost 8 percent of all spirits sales. And Kentucky distillers account for 95 percent of the world’s bourbon supply.

For the Jim Beam Distillery in Clermont, Kentucky, maker of several bourbon brands, it means a zealous commitment to meeting the needs of those preferences. But it’s not always a smooth ride.

Just ask Luke Adam, Ph.D., an analytical chemist, and Jim Beam’s Global Director of Scientific Services.

Bourbon is made from grains: greater than 51% is corn with the balance made up typically from wheat or rye, and malted barley. The distilleries producing bourbons as well as other spirits — Beam has 14 in total including three producing bourbon in Kentucky — get 80,000-pound deliveries of grain several times a week.

If the grain has been exposed to fungal species or elevated moisture from improper storage conditions, it might get musty. It can give the finished product a bitter flavor. That mustiness comes from the presence of a chemical called Geosmin. Distillers can often detect the smell in grain deliveries and reject it.

But sometimes the chemical is in minute quantities and difficult to detect by smell. Mold, which is associated with Geosmin fluoresces, so exposing a sample spread on a table to a black light might show its presence to the naked eye if the concentration is high enough.

“If it is musty in smell, that's caused primarily by Geosmin which is volatile and hence will carry through our entire process, through mashing, through distillation, through fermentation,” Adam said.

His ultimate goal is for a quality control process that allows the company to reject those truckloads, using science-based, quantitative methods.

“We’ve always wanted it to be a quick test. So, all the work we’ve done is to put guardrails around it. From the very beginning of our research, it was about looking for a new quality control technique.”

In fact, reducing the number of musty distillate batches that it produces is one of Jim Beam’s key performance indicators.

“If it's really bad. the supplier is doing a really poor job. That’s maybe one out 500 loads where maybe they're doing a bin clean out, or maybe they've done some blending of good material with bad material. They shouldn't do that, but it can happen. And sometimes it comes in so musty that the operator can tell. Those are obvious things that we can reject but are super infrequent.”

“What we're actually trying to detect is where the operator thinks the grain is fine, but it's not. And so, it may pass the odor test. The black light doesn't see any or can't detect any mold growth. It’s really addressing all of the samples that are not acceptable, but in the past, we didn't know until we saw musty distillate.”

And that’s unacceptable.

So, Adam sought out an analytical approach using chemistry to identify levels of Geosmin in the grain deliveries. He needed something sensitive enough to detect low levels.

But he had another problem. The distilleries are dusty, experience temperature fluctuations, and leave little room for a large lab — not an ideal environment to do analytical chemistry. His solution had to fit within those constraints.

Although the black light test worked, the observations were subjective and not sensitive enough. Adam wanted a quantitative way to characterize the grain. Since Geosmin-associated mold material fluoresces, he thought that might be a path to follow.

So, he went to the Pittcon trade show three years ago, and without telling anyone his idea.

“I just went from booth to booth asking people ‘what would you do to detect Geosmin or to detect mold and grain and do it quickly,’ ” he said. “And I got a lot of answers, like, ‘you need to use liquid chromatography combined with mass spectrometry,’ and ‘you need to use a triple quad’ and all of these techniques, which probably would work. But you couldn't put them in a distillery. They would be way too slow, they are not easy to operate and it would be incredibly expensive.”

“So, when I eliminated all of those, then I got down to two spectroscopy techniques; Raman and fluorescence. And so, we did some pilot testing with the two different techniques, and it was the results from the unique fluorescence technique offered by HORIBA that proved to be the most successful for us.”

Jim Beam uses HORIBA’s Aqualog A-TEEM Fluorescence and absorbance spectrometer to test samples of grain. It is small, fast, and inexpensive to use, and sample preparation for these grain samples is simple.

The Aqualog uses proprietary A-TEEM technology that offers simultaneous Absorbance-Transmission fluorescence Excitation and Emission Matrix acquisition. This fast and inexpensive optical technique for quantitative molecular fingerprinting provides high specificity and ultrahigh sensitivity.

One of the key things that makes this successful for Jim Beam is that it is a chemometric technique.

‘It requires our model building so that we can utilize the software (Aqualog’s Solo™ package) to give a pass fail. It's not as simple as ‘here's the peak, it's due to Geosmin.’ It's more about taking high Geosmin samples, medium and low Geosmin samples, and grouping those in three-dimensional space through the chemometrics software. It's the building of the models. And then the software’s front end that the operators use will say it's musty, or it's going to say it’s clean. That's really all we need for them. They don't need to know the Geosmin concentration. We can always determine that in our laboratory with other techniques, but we really want to simply report to the operator that it's musty or it's clean. And if it says it's musty, then they rerun it. If it's still musty, they reject that load.

Testing samples with the Aqualog gives Jim Beam a level of quality control/quality assurance that it found evasive before. It’s also a relatively inexpensive technique with minimal lab demands to roll out to Jim Beam’s other distilleries.

Jim Beam has encapsulated the instrument in a laminar flow cabinet, or tissue culture hood, a carefully enclosed bench designed to prevent contamination of semiconductor wafers, biological samples, or any particle sensitive materials. Air is drawn through a HEPA filter and blown in a smooth, laminar flow towards the user.

“You need a small lap bench on wheels,” Adam said. “The cabinets which sit on that help us protect the instrument inside the cabinet and keep it clean, because it's a very dusty environment.”

Initial tests were conducted in 2020 and 2021. Roll out to all three Kentucky distilleries was delayed because labs had to be set up in the receiving areas of each small grain facility.

Since beginning to work with the Aqualog, Jim Beam has found that corn is the major culprit for the must character, followed by rye and wheat, but not the malted barley. Much of that is due to seasonal fluctuations and wet weather conditions in the growing areas.

“It’s geography,” Adam said. “Is it a wet season? Is it a dry season? Is there mold already growing on the kernels during the harvesting process? Grain cleaning before it’s delivered to us helps tremendously, and we’re trying to make that a common practice. That is consistent with how we're preparing the sample, because we're not grinding the grain and trying to extract all of the Geosmin from it. We're actually doing this water extraction from the surface of the kernels. And that's where the mold is growing, on the surface. And the presence of the bacteria especially is on the surface. That's why that simple sample prep works.”

The must can also originate in the storage bin, where wet or bacterial conditions exist.

Adam’s work at the Jim Beam Distilleries is ahead of the pack and helping to enhance the consistency and quality of its bourbon while reducing rejected batches. It’s a novel quality control application of fluorescence spectroscopy and one Jim Beam finds enhances one of its key performance indicators, rejected batches of bourbon due to musty grain.

Follow Us