J. High Resol. Chromatogr. 2000, 23, (9) 547–553 i WILEY-VCH Verlag GmbH,D-69451 Weinheim 2000 0935-6304/2000/0909–0547$17.50+.50/0 547 Considerations on Static and Dynamic Sorptive and Adsorptive Sampling to Monitor Volatiles Emitted by Living Plants Joeri Vercammen, Pat Sandra* University of Gent, Department of Organic Chemistry, Krijgslaan 281 S-4, B-9000, Gent, Belgium; e-mail: pat.sandra@rug.ac.be Erik Baltussen Eindhoven University of Technology, Laboratory of Instrumental Analysis, P.O. Box 513, 5600 MB Eindhoven, The Netherlands Tom Sandra, Frank David Research Institute for Chromatography, Kennedypark 20, B-8500, Kortrijk, Belgium Ms received: April 7, 2000; accepted: July 6, 2000 Key Words: Capillary GC; static and dynamic headspace sampling; sorptive and adsorptive extraction; plant volatiles Summary Static and dynamic headspace sampling have been applied for the enrichment of volatiles emitted by living plants. For solid phase microextraction (SPME) the sorptive fibers polydimethylsiloxane (PDMS) and polyacrylate (PA) have been compared and, in accor- dance with the like-like principle, polar compounds exhibit more affi- nity for the PA fiber while apolar solutes favor the PDMS fiber. For dynamic sampling, tubes packed with PDMS particles show greater inertness than Tenax; some Tenax decomposition products, e.g. ben- zaldehyde and acetophenone, interfere with the analyses. With PDMS particles operated in the breakthrough mode, the obtained profiles are similar to those obtained by SPME on the PA fiber. Recoveries relative to a packed PDMS bed are 85% for Tenax, 2.4% for SPME-PDMS, and 6.2% for SPME-PA. 1 Introduction The analysis of volatile organic compounds (VOCs) emitted by living plants is an interesting and challenging research domain for various reasons. Breeders can characterize differ- ent varieties of the same plant or check its condition by mon- itoring the volatiles surrounding the plant, i.e. the headspace [1–3]. On the other hand, plant physiologists, biochemists, and biotechnologists presently focus their research on changes in the headspace profile induced by stress. A knowl- edge of headspace modification can help us to unravel the underlying biosynthetic pathways or to elucidate possible communications with other organisms [4–6]. Mechanical wounding, a general stress condition, leads to a response which has already been intensively investigated [7]. The emitted compounds, namely C 6 saturated and unsaturated alcohols and aldehydes, are formed by enzymatic degradation of fatty acids from damaged cell walls. Some of these VOCs exhibit antimicrobial activity and protect the wounded plant against infections by pathogens [8, 9]. Other compounds, like methyl salicylate, are synthesized de novo after infection. Diffusion into uninfected leaves leads to the expression of defense related genes which results in an increase in resis- tance. This is a form of intercommunication between plants [10]. Plant-insect communication is another interesting study field. The olfactory receptors are situated on the insect’s antennae. The insects use the emitted chemical signals for host-finding purposes like feeding or ovipositioning [11]. Herbivore attack can be detrimental for a plant. The plant indirectly eliminates this hostile situation by synthesizing specific volatiles, triggered by oral secretions of the insect, which attract the insect’s natural enemies [12–14]. A wide variety of chemically different compounds, some of them with very high polarity, can be emitted by plants and it is of utmost importance to have a reliable sampling technique which, moreover, shows no chemical reactivity towards the emitted VOCs. A commonly applied method is dynamic headspace. The volatiles are enriched on an adsorbent such as activated charcoal [15, 16], carbon molecular sieves [17], or porous polymers, e.g. Tenax [8, 10, 18, 19] followed by liquid or thermal desorption. Presently, thermal desorption is the method of choice because it can be coupled on-line to the analytical system which guarantees the highest sensitivity and less risks for contamination. For thermal desorption, adsorbents are characterized by a number of disadvantages, the most important of which are that high molecular weight and polar solutes can only be desorbed at relatively high tem- peratures. This can result, on the one hand, in irreversible adsorption and, on the other hand, in catalytic induced modi- fication of adsorbed solutes [20, 21]. These shortcomings can be overcome by using sorptive instead of adsorptive enrich- ment. In the mid-1980s several groups, e.g. Burger and Munro [22], Roeraade and Blomberg [23], and Bicchi et al. [2, 24, 25], illustrated the advantages of enrichment of gas- eous samples onto the GC stationary phase polydimethylsil- oxane (PDMS) in which partitioning (sorption or absorption) is the retention mechanism. Initial experiments were carried out on open tubular traps (OTT) coated with thick layers of PDMS but this approach suffered from low sample capacities and low maximal allowable flow rates (a 10 mL/min) which limited their use and general acceptance. Recently, tubes with similar dimensions to adsorption tubes, but packed with 100% PDMS particles were introduced [26]. Compared to OTT, they can be used at high flow rates, possess large sam- ple capacity and volume loadability, and can moreover be used in the breakthrough and in the equilibrium mode [27]. Advantages compared to tubes filled with adsorbent-type trapping materials are the possibility to assign degradation