More than once in my marriage—in fact, more than once in the last month—my significant other has presented me with a mysterious substance and asked, “Does this taste funny to you?” I generally decline these invitations to explore the unknown, not because I lack the courage, though that is true, but as someone with a science background, I feel I lack the proper instrumentation.
Actually, real scientists are increasingly faced with this same dilemma. According to “The Analytical Cornucopia,” a report issued recently by Los Angeles-based consulting firm Strategic Directions International (SDI), the global market for laboratory instrumentation dedicated to food applications amounted to $2 billion last year. Technology geared toward the detection of pesticides—which really do taste funny—garnered a large stack of those dollars.
“Although simple, non-quantitative test kits can provide information on the presence of pesticides and other residues, the food industry is increasingly turning toward the quantitative analytical solutions like those used by regulatory agencies,” said Michael Tice, vice president of consulting for SDI. He added that the technologies currently experiencing the greatest growth in pesticide analysis are liquid chromatography-mass spectrometry (LC-MS) and gas chromatography-mass spectrometry (GC-MS).
Contributing to that growth is the recent purchase by the Environmental Protection Agency (EPA) of the Acquity UPLC/ Xevo TQ Mass Spectrometry System, from the Waters Corp., Milford, Mass. “We’re continually trying to stay abreast of new technology,” says Charles Stafford, team leader of the analytical chemistry branch of the EPA lab in Fort Meade, Md. “We need machines that can detect lower and lower levels of pesticides, and frankly, we need more machines. All our instruments are booked.”
Since the melamine scare of two years ago, Stafford has noticed not only an uptick in overall activity but also a new emphasis on screening samples for unknowns. “We just weren’t doing things like that before, looking for what shouldn’t be there. The same holds true for novel pesticides. So we’re struggling with approaches and instrumentation which will allow us to screen samples for untargeted compounds.”
That is not to say that Stafford is actually doing most of the testing. His lab primarily develops the analytic methods to support regulatory activities, those that discern the dietary exposure side of the risk-assessment equation. “We help either develop methods or provide maximum residue levels (MRL) data that will help scientists determine whether a chemical is safe.”
The globalized nature of foods with multiple ingredients begs the question of harmonization of “safe” MRLs. “From the perspective of my lab,” says Stafford, “if someone asks me to evaluate data that comes from another source, the most important thing we say is ‘how did they do the sample, what method did they use from start to finish,’ and that includes the detector. That’s not to say that data is bad if generated by using an element-specific detector, but in the back of your mind …”
Certainly there are valid alternatives to GC or LC-MS out there. Any number of element-specific assays and instruments exist, but their efficacy hinges on knowing exactly what you’re looking for, and that knowledge can no longer be assumed.
LC or GC?
“There is reasonably good harmonization within the United States,” says Walter Hammack, environmental manager at the Florida Department of Agriculture, “but it’s more of an issue internationally.” Hammack’s mission is to enforce EPA tolerances, collaborating efforts with the United States Department of Agriculture as well as other agencies. “We’re trying to get an idea what residue pesticides people are exposed to right at the dinner table,” regardless of the source. Hammack receives samples from all over the world.
Hammack also assigns the technology and has long coordinated the activities of his LC and GC spectrometers. “Both have very powerful strengths, LC because a lot of the newer pesticides that we’re looking for are amenable to LC separation.” Historically, however, GCs had the advantage of more sensitive detectors, and a true unknown could be tackled with GC-MS and electron ionization with spectra matched against known libraries. “The limit here is the GC part. The analyte may not survive, whereas the LC is far more gentle.”
Choice of instrument may also depend on the sample source. “If you’re looking for a compound that you suspect may be fat soluble, then the advantage goes to GC. The class of pesticide may also tell you what to use. Older generations of pesticides tended to be chemically nonpolar and, therefore, more stable. Today the focus is on pesticides that do their job and then go away,” said Hammack. “These are far more amenable to LC.” But don’t be complacent, he warns; the older compounds are still very much in use.
In the meantime, pesticide detection is all about stretching the detection limits, and to that end, Hammack is looking to tweak his tools. “We’re looking at tandem MS. This initially pushed the pesticide field toward LC because the GC simply didn’t have those detectors available, but now we’re looking at GCs with the added horsepower we need.” His hope is that GC upgrades will bring more balance to a field currently dominated by LC and its inadvertent limitations.
Selling just such an upgrade is Vivian Watts, PhD, an application chemist at Chromsys in Alexandria, Va., a partner of Agilent Technologies. “Agilent was one of the last players to come up with the GC-MS/MS, and so what we decided to do in 2006 is to develop an upgrade to the existing system,” the so-called Evolution Triple Quad upgrade for the Agilent 5973/5 system. With the addition of Chromsys’ proprietary IonRail collision cell and another high-precision quadrupole Q3, the upgrade results in a state-of-the-art, triple quadrupole GC-MS/MS at a fraction of the cost of a new machine.
It’s a timely idea, and business is good. “As time goes on, more and more pesticides are invented, and countries are in a position of constantly amending watch lists of compounds that can no longer be allowed for consumption,” said Dr. Watts. And better detection limits may be nudging tolerances downward. Can you really detect five parts per billion of whatever in my organic tomato? Well, by all means, let’s consider that. “It does take time for regulations to catch up to the technology. But I can tell you that the EU and Japan already have stricter standards. The U.S. is just now catching up.” Upgrading the Agilent instruments already in play can certainly aid in that effort.
Within every industry, there are troubleshooters. Rachel Linck, PhD, is one of those. She describes her duties as director of technologies at Chemir Analytical Services in Maryland Heights, Mo. “We’re specifically set up to handle non-routine analysis. When companies have issues where they’re not sure if their products are safe for consumption, that’s an answer they need right away.” Dr. Linck will quickly develop a method that facilitates the need of a company’s particular sample matrix. “They tell me of the situations they’re facing, and I design the testing that we’re going to do in the lab to answer the question.”
Often, the protocol required is not on the shelf because, as the song goes, “the times they are a-changing.” “It’s been interesting lately,” Dr. Linck said. “Half the phone calls I get are, ‘We sourced this raw material in China, and we don’t trust the source anymore.’” In essence, the analytical game has become one of expecting the unexpected.
As far as instrumentation is concerned, Dr. Linck recognizes the need for both GC and LC capabilities—the requests she receives demand both. Recently, for example, a meat processing plant had, during a normal application of pesticide to the facility, inadvertently left a freezer door ajar. “So, they knew what they had been spraying. They just needed to know if any residue had reached their product.” Samples were shipped overnight, and the GC-MS quickly did the rest (result: no exposure was detected).
Yet the best tool for the job may not be obvious. “Food products themselves are pretty complicated,” said Dr. Linck. Different types of consumables have different considerations for extraction. An apple? No problem. “But we had a product a while ago—these were baked hams. We went ahead with the extraction, but we had all sorts of problems with all the fats present. Is there helpful literature out there? Certainly. But so much of this is trial and error.”
It would be nice to have a (not so pricey) black box into one end of which you could drop a tomato, getting a list of components out of the other end, but that’s years away, if it ever occurs. As instruments become more sophisticated and at the same time more automated, the demands of the operator seem deceptively simple. But, as wise math teachers say, the answer means nothing if you don’t know how you got it. You must show your work.
Canavan is a freelance writer based in Brooklyn, N.Y. Reach him at email@example.com.