October 1997 -- QA/QC
By: Raymond Marsili
Low margins and fierce competition in nearly every food-industry sector are forcing food companies to work more efficiently than ever before in every phase of their business operations - including chemical laboratory testing. In response to these pressures, companies have developed new instruments and analytical methods geared toward enhancing testing productivity, and improving test accuracy and sensitivity.
New regulatory requirements have pushed analytical work demand to an all-time high. Modern sophisticated analytical workhorses can obviate the need for hiring new chemists and lab technicians, an expense usually far outweighing instrument-purchase and maintenance costs.
Testing for total fat
Testing products for total fat is one of the most frequently requested tests at food companies - and, surprisingly, one of the most inaccurate. With the emphasis on minimizing total fat and saturated fat in the diet, one might think fat-determining analytical procedures would be well-established, accurate and routine. Not so.
Numerous fat-determining test methods are being used - Soxhlet; Weibull-Stoldt (acid hydrolysis followed by extraction); Roese Gottlieb, or Mojonnier; Gerber; near infrared; and microwave extraction. Unfortunately, the same sample - run by different labs using identical or different methodologies - can result in inconsistent data.
One major problem: Most tests use a gravimetric method - the lipid material is extracted with an organic solvent; the solvent is evaporated off, and the remaining residue is calculated as percent fat. However, while all fats are lipids, not all lipids are fats.
Since all natural fats contain fatty acids, it would seem prudent to use the fatty acid content as the basis for determining amount of fat. The U.S. Food and Drug Administration defines total fat as "total lipid fatty acids, that is, the sum of fatty acids from mono-, di-, and triglycerides, free fatty acids, phospholipid fatty acids, and sterol fatty acids.
... The declaration of total fat (must) be expressed as the amount of triglyceride that would provide the analytically measured amount of total lipid fatty acids in the food."
Not only does a need exist for a uniform, commonly accepted definition of fat, there also is a necessity for a fat-determination method, conforming to the accepted definition of fat.
Currently, gas chromatography (GC) methods determine fat, based on fatty acid content. The fat is extracted, saponified and esterified to prepare the fatty acid methyl esters (FAMEs). The fatty acid esters are separated and quantitated by the GC. This method is effective at determining the amount of saturated and unsaturated fatty acids in the sample, but consumes time. A GC/FAME-based test for the determination of total, saturated, unsaturated and monounsaturated fats in cereal products recently was approved as an official test by AOAC International (Journal of AOAC International, March/April 1997). Fats in the cereal products are released by acid hydrolysis, then extracted with ethyl ether and petroleum ether. FAMEs are then prepared, followed by injection into a capillary GC column.
The time-consuming ether-extraction step makes this test cumbersome. Therefore, some food labs have successfully applied supercritical fluid extraction (SFE) or accelerated solvent extraction (ASE®), a technique developed by Dionex Corporation, Sunnyvale, CA.
Results by Dionex have shown the effectiveness of ASE - compared to Soxhlet extraction methods - in analyzing fats, oils and pesticides from foods and agricultural products. In the extraction of unbound fats from snack foods, such as potato and corn chips, recoveries with ASE equaled those of Soxhlet. However, extraction speeds were significantly faster, and the amount of solvent used was a fraction of that required by the Soxhlet method.
Similar results were obtained with extraction of oil from oil seeds and pesticide residue from foods. With ASE, extractions were performed in minutes, compared to hours using the Soxhlet method, and the quantity of solvent used was less by a factor of 10.
On the same team
Research chemist Jeff Hurst, Ph.D., at Hershey Foods, Hershey, PA, is heavily involved in new methods development for analyzing confectionery products, pasta and many other food products. According to Hurst, some applications are better-suited for ASE, while SFE is more appropriate for others.
Joseph Levy, president of Levytech, Inc., Gibsonia, PA, and a consultant specializing in chemical-sample-preparation techniques, agrees with Hurst. "ASE and SFE should be viewed as members of the same team, that when used cooperatively, and in harmony, do present themselves as a formidable force against the problems associated with analyzing complex matrices in the real world," Levy says. "SFE, for example, could be a viable complement for selective extraction of key targets after an exhaustive extraction with ASE. For more difficult matrices, ASE could be used for the general extraction, while the wide variety of off-line collection strategies, or online ability in SFE, could be used to confirm specific target analytes (such as with nuclear magnetic resonance, mass spectroscopy (MS), or Fourier-transform infrared. Moreover, depending on the analytical objective, a mating of ASE and SFE could be viable for some matrices."
Most industrial labs haven't done a lot with SFE to date, says chemical-analysis consultant Merlin K.L. Bicking, of ACCTA, Inc., Woodbury, MN. "People were disappointed with older instruments and got scared off," he says.
ASE, however, is really starting to take off. Since it's capable of more selective extractions, SFE has an advantage over ASE, Bicking says. This is because it has more parameters to manipulate to facilitate extraction. But, the downside is: More parameters also translates into more options to control, more complexity, and longer method-development times.
Food chemists tend to be more familiar with principles of ASE, which, says Bicking, are more closely related to Soxhlet extraction - at least in terms of chemical principles. But ASE uses far less solvent than Soxhlet; pressurizes extracting solvents so they can be used at higher temperatures; and takes minutes instead of hours.
One SFE manufacturer, Isco-Suprex, Lincoln, NE, offers instruments that can perform SFE and enhanced solvent extraction (ESE), a technique similar to ASE. For example, it offers an SFE-IR (supercritical-fluid extraction-infrared analysis) interface kit. The heart of this kit is a high-pressure, heated IR flow cell, which fits into most commercially available IR spectrometers, without the use of fragile optical fibers. This accessory passes the analyte extract directly from the extractor through the heated transfer line to the heated IR flow cell, which is mounted in the optical bench of the IR spectrometer.
SFE-IR testing of fats and oils has been used to determine the:
- total lipid content of food materials (such as milled rice);
- amount of unsaturated lipids or iodine number;
- amount of unnatural trans-unsaturated fat formed during partial hydrogenation;
- content of free fatty acids.
One sign that SFE is finally catching hold as a viable extraction tool is the appearance of official SFE-based test methods. An American Oil Chemists' Society (AOCS)/AOAC collaborative study, begun two years ago, has resulted in an official AOCS method for supercritical fluid extraction of oil from oilseeds. The two-part method - involving extraction with supercritical carbon dioxide alone, or with supercritical carbon dioxide plus 15 percent ethyl alcohol modifier - is applicable for determination of process-scale extractable oil or determination of total oil.
The SFE method is fast, requiring only a 30-minute extraction to determine process-extractable oil, or 60 minutes for total oil. The SFE method uses no hazardous solvents, so it's well-suited for the process-control environment.
A new player in the SFE market, Leco Corporation, St. Joseph, MI, offers an SFE instrument that extracts samples in parallel rather than in sequence. The basic unit performs three extractions simultaneously, and up to two add-on units may be run at the same time by the master unit, allowing for the analysis of as many as nine samples in just 20 minutes.
Testing for pests
Pesticide testing can also benefit from SFE and ASE. For example, Dionex's ASE 200 meets the requirement of U.S. EPA Method 3545 for extraction of base/neutrals and acids; chlorinated pesticides and herbicides; PCBs; and organophosphorous pesticides.
In one innovative approach, Virica Lopez-Avila, Midwest Research Institute, Mountain View, CA, and co-workers determined the presence of pesticides in foods by using a quick SFE-ELISA (enzyme-linked immunosorbent assay) technique. In this study, ELISA test kits (from Ohmicron) assayed for alachlor; aldicarb; atrazine; carbaryl; carbendazim; carbofuran; cyanazine; 2,4-Dichlorophenoxyacetic acid; and metolachlor were used to test baby foods. Recovery studies of the target pesticides from the spiked matrix were acceptable for most compounds - six pesticides demonstrated recoveries above 70%. The method is significantly faster than conventional solvent-solvent extractions, and uses far less organic solvents.
The advent of lower-cost, reliable benchtop systems has made GC/MS a routine analytical tool. In the last couple of years, even more sophisticated methods - once considered esoteric research techniques - increasingly perform routine analysis, primarily due to development of powerful, reasonably priced, commercial ion-trap MS instruments.
GC/MS-MS (gas chromatography/mass spectrometry-mass spectrometry) is a useful tool for overcoming the presence of unwanted contaminant (high background) peaks in a GC/MS chromatogram. Analytical benefits include:
- greater spectral selectivity;
- lower detection limits;
- reduction in chromatographic requirements and need for prior sample cleanup.
Even if the matrix contained another ion with the same mass-to-charge ratio as the parent ion for the analyte of interest, it's highly unlikely the interfering ion would yield the same daughter ions as the analyte. As a result, quantitation based on MS-MS daughter ions yields accurate results, even with interference.
Terry Sheehan, Ph.D., marketing manager, Varian Chromatography Systems, Walnut Creek, CA, points out how practical and useful GC/MS-MS can be as a tool to reduce sample-preparation time and error. His example: analysis of an ethion pesticide residue in a carrot extract.
On GC/MS chromatogram and mass spectra results for suspected ethion pesticide residue, quantitation and positive confirmation of ethion isn't possible, however, because interfering chemicals are detected along with the ethion. Without GC/MS-MS, the chemist faced with such significant matrix interference problems would undoubtedly go back to the bench and attempt to develop suitable sample cleanup steps - perhaps using solid-phase extraction - to eliminate the coeluting chemicals. This time-consuming process might not eliminate all interfering matrix components. Therefore, GC/MS-MS is a far superior way to resolve the interference problem.
A common multiresidue-screening technique for pesticides in foodstuffs is the Luke II method, in which a crop sample is homogenized with a water-miscible solvent (acetone). Unfortunately, other crop materials interfering with the chromatographic detection are co-extracted with the pesticides, and these substances must be eliminated from the extract with a series of liquid-liquid extractions. Each 100-gram sample prepared using this method requires up to 800 ml of organic solvent for cleanup, and the steps are labor-intensive and time-consuming.
An alternative, solid-phase-extraction membranes (Empore™ from 3M Corporation, St. Paul, MN), shows promise for extracting pesticides in high-moisture foods. After blending a 100-gram sample of each crop with 100 ml of acetone, a portion of the acetone-extract-containing pesticide residues, along with co-extracted interfering compounds, is passed through two extraction disks. Stacking a reverse-phase disk with a carbon disk provides excellent recoveries of polar and nonpolar pesticides (see "Evaluation of SPE Disks in Pesticide Extraction," Food Testing & Analysis, February/March 1997). This sorbent combination efficiently removes pesticides from the blender filtrate, while much of the co-extracted material passes through unretained, and is discarded.
The purified acetone vegetable extract containing the pesticides then is evaporated to dryness, dissolved in a suitable solvent, and injected into a GC equipped with an appropriate detector for pesticides.
Using the extraction disks avoids the laborious clean-up steps involving multiple liquid-liquid extractions; it shortens the analysis time, and drastically reduces solvent consumption.
Another sample-preparation step that's been improved in recent years is solvent-evaporation after extraction, prior to injection into the GC. In the good old days, chemists would place extracts in a beaker of warm water, and evaporate the solvent with a stream of nitrogen gas. Today, several companies offer instruments to speed up and simplify the evaporation process.
For example, Labconco Corporation, Kansas City, MO, offers the RapidVap Evaporation System, an instrument capable of processing multiple samples with individual volumes up to 450 ml. Methylene chloride evaporates at a rate up to 5 ml/minute per tube, and water evaporates at a rate up to 1 ml/minute per tube.
A microprocessor-controlled, mechanically created vortex action forces the liquid sample outward against the tube walls, increasing the surface area for faster evaporation. The vortex action also helps mix the sample, and contain the analytes in the solvent for maximum sample recovery.
A block heater supplies a precisely controlled amount of dry heat to speed evaporation. Unlike water bath heaters, the block heater requires little maintenance. A built-in vacuum controller allows the user to adjust vacuum levels.
Hershey's Hurst uses this instrument for evaporating solvents from ASE and SFE extracts, although not specifically for pesticide testing. He says he likes the RapidVap because it doesn't develop annoying condensate problems during solvent evaporation.
From extracting pesticides, to evaporating solvents and improved GC/MS detection capabilities, pesticide testing has become increasingly simpler and less expensive to perform.
Moldys, but ungoodies
Like pesticides, mycotoxins represent yet another unwanted food component frequently requiring analytical monitoring. Mycotoxins are produced by naturally occurring molds, which may infest agricultural commodities, such as grain, maize, peanuts, dairy and other foods. Hot, humid weather conditions promote mold growth, so contamination is a chronic problem in many parts of the world.
Presence of unacceptable mycotoxin levels in raw or finished products represents a processor's constant concern, and inaccurate analysis can cause serious consequences.
Elevated incidents of liver cancer occur in populations exposed to mycotoxins (such as aflatoxin) through dietary intake, according to the World Health Organization's International Agency for Research on Cancer.
Traditional methods for mycotoxin analysis use extensive sample-extraction and clean-up procedures, followed by thin-layer chromatography or high-performance liquid chromatography (HPLC). While quantitative and sensitive, these methods are lengthy, costly and complicated. New "affinity" columns, consisting of monoclonal antibodies on support beads, selectively extract the mycotoxin of interest from the food matrix, resulting in significantly simpler and faster testing.
New test kits from Vicam, Watertown, MA, use immunoaffinity columns for aflatoxins and fumonisin, as well as for ochratoxin, deoxynivalenol and zearalenone. These kits are approved by AOAC and the U.S. Department of Agriculture.
Most mycotoxin assays, using these columns, are made by direct measurement of fluorescence of the affinity-purified mycotoxins. This is done by collecting the mycotoxin from the affinity column, and placing it in a test tube in the fluorometer.
Affinity columns also can provide excellent cleanup for HPLC. HPLC confirms the presence of the mycotoxins and also allows determination of the individual aflatoxins or fumonisins. One particularly effective analytical strategy is using the affinity column/fluorometer method as a field or QC screening tool, and using the affinity column/HPLC method to confirm the presence of mycotoxins in samples that test positive by QC fluorometer methods.
Figuring out flavor
Solid-phase microextraction (SPME) provides numerous advantages over other sample-preparation methods for food-aroma analysis. SPME allows rapid, solventless extraction, and pre-concentration of volatile and semivolatile organic compounds.
This device partitions organic components between a bulk aqueous or vapor phase, and thin polymeric films coated onto fused silica fibers. It is essentially a modified syringe,with a spring-loaded plunger and a barrel with a detent to allow the plunger to be held in an extended position during the extraction phase and the injection/desorption period. The barrel contains a modified 24-gauge stainless steel needle, which encloses another length of stainless steel tubing fitted tightly to a short piece of solid-core-fused silica fiber.
The bottom 1 cm of the fused silica fiber is coated with a relatively thin film of any of several stationary phases. This film serves as the organic "solvent" during the absorption of the volatile compounds. The needle punctures the septa, sealing both the sample container and the GC injection port. It also protects the fragile silica fiber during storage and use.
Silica fibers, coated with nonpolar polydimethylsiloxane and the more polar polyacrylate, have been commercially available for roughly two years. Supelco recently developed a new fiber material, Carboxen(tm)/polydimethylsiloxane, a highly porous carbon with a surface area of 1,200 m2/g. The high porosity enables the fiber to strongly retain small analytes, and provides a high sample capacity. The small pores enable this fiber to extract analytes at higher orders of magnitude than other SPME fibers.
Two general operation modes exist for SPME: sampling the analyte in the gas phase from the equilibrated headspace in a sealed vial; or immersing the fiber to collect the analyte directly from the liquid portion of the sample.
Major advantages of SPME are that it is:
- easily automated for increased sample throughput;
- fast and solvent-free - equilibration times for sampling onto fiber range from two to 15 minutes;
- capable of being used manually with any GC or GC/MS;
- sensitive, with ppt detection limits for some compounds;
- ideal for quick screening; and
New speed limits
No matter which technique is used to extract flavor-important chemicals in foods and ingredients, one problem plaguing flavor chemists is the lengthy times - sometimes as long as two hours - required for all the peaks to elute in their GC runs.
Instrument makers have responded by developing fast GC techniques, incorporating specialized oven heaters and new types of detectors. Leco Corporation recently introduced Pegasus(r) II, the industry's first fully automated time-of-flight mass spectrometer (TOFMS) for fast gas chromatography. It redefines the time frame for GC/MS analysis - an important tool for flavor chemists.
Combining time-of-flight technology and Leco's proprietary fast-data-collection system yields up to 500 spectra per second. These faster acquisition rates allow the instrument to easily handle the narrow peak widths resulting from today's fast GC separations, without loss of data integrity or system sensitivity. Traditional GC/MS analysis of citrus oils, for example, sometimes takes as long as two hours, but the automated TOFMS takes less than three minutes.
The new instruments and analytical methods now available to food product designers provide benefits that are truly impressive: significantly faster analysis time, compared to traditional methods; low cost per analysis; high precision and accuracy; automated control requiring little, or no, operator attention; minimal usage of flammable organic solvents and toxic reagents (which soon will cost more to dispose of than to buy); and/or advanced, computer-based, data-management capabilities.
Back to top
© 1997 by Weeks Publishing Company
3400 Dundee Rd. Suite #100
Northbrook, IL 60062