Quantitative Performance of a Simple Tenax-GC Adsorption Method For Use in the Analysis of Aroma Volatiles G. OLAFSDOlTIR, J. A. STEINKE, and R. C. LINDSAY ABSTRACT A simple method for obtaining etherextracts of aroma volatiles from foods for gas chromatographic analysis using single-use Tenax-CC collectiontubesis described. The method was characterized quanti- tatively for the collection of a varietyof compounds from modelsys- tems by both dynamic-purging (10 ppb level) and equilibrium displacement techniques (15ppmlevel), and for collection of fractions (250 kg level) from packed gaschromatographic columns.A range of variation in reproducibility of 5-22% was found for recovery of compounds by the dynamic purging system, and less than 10% vari- ation was encountered for ethyl esters using the equilibrium displace- ment system. Collection efficiencies for selected ethyl esters from packed gas chromatography column collections were about 80% of theoretical with a variation of less than 10%. INTRODUCTION CONCENTRATION of trace aroma volatiles present in head- spacegases over foods, beverages,and environmental samples by adsorption onto a variety of porous solids has become an established procedure for preparing samples for analysis by gas chromatography (CC) and mass spectrometry (Teranishi et al., 1971; Jennings et al., 1972; Murray, 1977; Noble et al., 1980; Schaefer, 1981; Nunez et al., 1984; McNally and Grob, 1985a; 198513). Tenax-GC (2,6-diphenylparaphenylene oxide polymer; ENKA, The Netherlands) has emerged as a widely-used porous polymer for these applications. Some of the general properties of adsorptive techniques in these appli- cations have been reported, including retention volumes for different components (Kuo et al., 1977; Novotny et al., 1974; Sakodynskii, 1974; Bertsch et al., 1975; Krost et al., 1982; Kawata et al., 1982) and the effects of sampling times and purging rates on retention of volatiles on Tenax-GC (Buckholz et al., 1980). Other studies have involved assessments of re- coveries of various components from porous polymers (Leahy and Reineccius, 1984; Williams, 1982) as well as reproduc- ibility of analyses (Noble et al., 1980; Jennings et al., 1972; Withycombe and Lindsay, 1972). While earlier studies have employed a variety of collection devices and procedures, most have utilized thermal-desorption and associatedtransfer techniques for recovery of volatiles for analysis subsequently by GC (Bertsch et al., 1975; Murray, 1977; Buckholz et al., 1980; Nitz and Julich, 1984; Sandra et al., 1980). However, complex and sometimescostly collection equipment generally has been employed in the earlier studies. As a part of continuing studies on aroma volatiles in foods, simple Tenax-GC collection proceduresusing readily available materials and ethyl ether elution of adsorbed volatiles have been developed and evaluated in our laboratory. The objectives of this paper are to describe some of these procedures and to provide results of studies on their applications in the estimation of a variety of aroma volatiles. The authors are affiliated with the Dept. of Food Science, Univ. of Wisconsin, Madison, WI 53706. MATERIALS & METHODS Tenax-GC collection tube construction Collection &bcs for Tenax-CC were fabricated from 9-inch Pasteur disposable pipettes (Scientific Products, McGaw Park, IL) as shown in Fig. I for the configuration employed for collection of packed- column CC fractions. When collections of volatiles from slurries of foods by dynamic purging were carried out, the length of the large- diameter glass tubing was increased by about 3 cm to accommodate a larger quantity of Tenax-CC. Phosphoric acid-treated glass wool (Supelco, Inc., Bellefonte, PA) was used for positioning Tenax-GC in tubes. Tenax-CC (ENKA, The Netherlands) was used without special treatment in most instances although sequential washing with acetone and ethyl ether at times proved useful to remove trace contaminants present in some lots. The amount of Tenax-GC employed in each trap varied from 11 mg for collection of equilibrium headspace samples and packed-column CC fractions to 5@75 mg for dynamic gas-purged samples. Assembled collection tubes were rinsed with 20mL ethyl ether (Mallinckrodt, Inc., Paris, KY) that had been fractionally dis- tilled if necessary to exclude impurities. Ether was washed through traps under reduced-pressure using a vacuum flask equipped with a suitably-bored rubber stopper and a water aspirator, and then the con- tents of tubes were brought to dryness with continued aspiration. Dynamic gas-purging apparatus All-glass gas-purging systems for collection of aroma volatiles from beverages or slurries of foods were constructed as shown in Fig. 2. Sizes of round-bottom flask varied from 250-IOOOmL, and the length of the purge-tube was adjusted to accommodate the system and the flask chosen. Swept-surface purgingcould be used for samples that foamed readily (Michael et al., 1980), and bubble-purging was suit- able for samples that did not foam easily. For collection of volatiles, samples were placed into the flask, the internal-standard solution was added, and the purge-collection part of the assembly was put in place. Tenax-CC tubes were fixed into place in the assembly using heat-shrinkable Teflon tubing (5 mm o.d.; Berghof/America, Inc., Raymond, NH), and the blue-tip of a Bunsen burner flame, using care not to scorch the Tenax-GC. Water-pumped nitrogen which had been passed through molecular sieve (Gas purifier, Alltech Associates, Springfield, IL) was then purged through the sys- tem at rates of 2&500 mL/min. Samples were stirred during purging at a moderate rate using a magnetic stirrer (Magnestir; Lab-Line In- struments, Inc., Melrose Park, IL). An aqueous solution containing 10 ppb each of I-butanol, l-penten- 3-01, I-octen-3-01, (E)-2-nonenal, ethyl heptanoate, and ethyl unde- canoate was employed for determination of reproducibility of the pro- cedure, and for determination of the effects of varying purge flow rates and collection times on the recovery of volatiles. For each trial 100 mL of the standard solution was placed in a 250 ml flask, and each was analyzed under prescribed conditions using the bubble-purge configuration of the apparatus while held at 21°C. Comparative studies on the efficiency of recovery of natural volatile aroma compounds from fresh whitefish (Coregonus clupeaformis) (Josephson et al., 1983) were carried out with the dynamic-purge Tenax-CC method. Gutted fish that had been held on ice no longer than 5 days were each washed with approximately 1OOmL saturated sodium Sulfate solution which was selected because it caused less foaming than sodium chloride solution. All of these extracts were pooled to obtain a large sample of the drip/slime mixture, and 100 mL subsamples were taken for analysis by the bubble-purge config- uration of the apparatus using 250 mL flasks. Volume 50 (1985)JOURNAL OF FOOD SCIENCE-1431