- Email: [email protected]
Sorptive tape extraction in the analysis of the volatile fraction emitted from biological solid matrices
Journal of Chromatography A, 1148 (2007) 137–144
Sorptive tape extraction in the analysis of the volatile fraction emitted from biological solid matrices C. Bicchi a,∗ , C. Cordero a , E. Liberto a , P. Rubiolo a , B. Sgorbini a , P. Sandra b a
Dipartimento di Scienza e Tecnologia del Farmaco, Universit`a degli Studi di Torino, Via P. Giuria 9, I-10125 Torino, Italy b Research Institute for Chromatography, Kennedypark 20, B-8500 Kortrijk, Belgium Received 1 December 2006; received in revised form 26 February 2007; accepted 5 March 2007 Available online 12 March 2007
Abstract Sorptive tape extraction (STE) is a recent sorption-based sampling technique in which a flexible polydimethylsiloxane (PDMS) tape is used to recover analytes at the surface of a solid matrix by direct contact as well as from the headspace in equilibrium with it. Solutes thus enriched on the inert PDMS material can be recovered either by solvent desorption or by thermo-desorption. The concentration capability of both direct contact and headspace STE was evaluated by sampling (a) aromatic plants to study the reaction of a vegetable matrix submitted to stress, and (b) fruits at the surface of the pulp or inside the pulp; the composition of the volatile fraction released from the skin when a perfume is sprayed on the back of the hand was also studied. The concentration capability of direct contact and headspace STE was compared to that of HSSE with a 20 L PDMS twister and HS-SPME with a PDMS 100 m fibre, by determining the relative abundances (RA) of the characterizing components of the aromatic plants under investigation. Repeatability and influence of tape surface on STE recovery were also evaluated. © 2007 Elsevier B.V. All rights reserved. Keywords: Sorptive tape extraction; STE; Direct contact (DC); Headspace (HS); PDMS tape; Biological solid matrices
1. Introduction The chemical composition of fractions sampled from solid living organisms belonging to the vegetable or animal kingdoms, as well as that of the volatile fractions released from them, may be taken as an important biosensor diagnostic of changes taking place in the metabolism of living organisms. Detection of variations in the composition of such fractions, identification and quantitation of marker(s), monitoring reaction dynamics at the living surface of an organism submitted for instance to stress, all require an analytical set-up in which analyte recovery over time is maximized while analysis time is reduced to a minimum. Several solventless sampling techniques have recently been introduced to study these problems: the best known among them are solid phase microextraction (SPME) [1] and stir bar sorptive extraction (SBSE) [2], both of which have successfully been applied to in-solution and headspace sampling [3–5]. Analyte ∗ Corresponding author at: Dipartimento di Scienza e Tecnologia del Farmaco, Universit`a di Torino, Via Pietro Giuria 9, I-10125 Torino, Italy. Tel.: +39 011 670 7661/2; fax: +39 011 670 7687. E-mail address: [email protected] (C. Bicchi).
0021-9673/$ – see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.chroma.2007.03.007
recovery with both SPME and SBSE using polydimethylsiloxane (PDMS) as accumulation polymer is based on analyte partition between liquid or vapor phases and the polymer: this phenomenon is also known as sorption. Neither technique can successfully be applied to sampling a biological solid matrix by direct contact, because of the limited exchange surface of both fibers and twisters with the matrix itself. Bruheim et al. compared a thin sheet of a PDMS membrane to SPME for the extraction of semivolatile analytes through both in-solution and headspace modes [6]. They found that better extraction efficiency and sensitivity were achieved with the PDMS membrane than with the PDMS thick film fiber because of the larger surface area/extraction phase volume ratio. Membrane extraction was adopted by Pawliszyn and coworkers [7] in combination with a sorbent interface to sample dynamically the headspace of Eucaliptus dunnii Maiden [8,9]. This technique is known with the acronym of MESI (membrane extraction sorbent interface). In a study aiming to evaluate the influence of sampling conditions on HSSE recovery carried out on six highto-medium-volatility compounds with different KO/W , Bicchi et al. [10] recently showed, among others, that HSSE recovery was not only related to the volume of PDMS coating the twister, but
138
C. Bicchi et al. / J. Chromatogr. A 1148 (2007) 137–144
also to its surface in contact with the vapor phase. Very recently, Sandra et al. [11] introduced sorptive tape extraction (STE) to study the effect on sebum composition before and after cosmetic treatment through in vivo sampling at the human skin surface. In this STE application, sebum components were sampled by direct contact of a thin flexible PDMS tape with the skin surface for a fixed time to study the effect of a matifying product on skin shininess. The sampled analytes were then recovered by either thermal or solvent desorption and analyzed on-line by GC or GC–MS. The PDMS tapes used were 15 mm × 4 mm × 0.5 mm for thermal desorption and 15 mm × 12 mm × 0.5 mm for liquid desorption. The wide exchange surface with headspace offered by PDMS tape and its ability to be applied directly to the surface of a living matrix makes STE very interesting as a sampling technique to study the chemical messages emitted from plants or animals when a variation in their metabolism occurs. This article reports the preliminary results of direct contact and headspace STE sampling applied to aromatic plants, fruits, and the release of perfumes from human skin. The results obtained with aromatic plants were also compared to those obtained by HS-SPME and HSSE. The concentration capability of HS-STE, DC-STE and HSSE, taking HS-SPME with a PDMS 100 m fiber as reference standard, was evaluated by determining the relative abundances (RA) of the characterizing components of the aromatic plants under investigation. 2. Experimental Commercially-available live spearmint (Mentha spicata L.) and rosemary (Rosmarinus officinalis L.) plants, perfumes and apples were analyzed. 2.1. Aromatic plant sampling 2.1.1. STE sampling system STE sampling was done using the special assemblies shown in Fig. 1. Two different vials were used; with both of them the PDMS tape was kept in contact with the biological surface, or in a position suitable for headspace sampling, by means of the stainless steel holder shown in Fig. 1, which was fixed to the body of the vial by one of its lateral arms. Vial A was especially designed for sampling from live plants. One of the stoppers of the vial and its teflon-coated gasket were cut, leaving a narrow cleft through which one or more leaves can be introduced into the vial without removing them from the stem, thus affording sampling from live plant organisms. Vial B was designed for samplings from the skin. It consists of a 40 mL glass bell that is placed on the skin surface, at the top of which the STE holder is fixed. 2.1.2. Sampling conditions The headspace resulting from one leaf of spearmint (whether whole or crushed) corresponding to 50 mg of plant material in the 40 mL vial A described above was sampled for 1 h at 25 ◦ C by (a) HS-SPME with a PDMS 100 m fiber (PDMS surface: 10 mm2 , volume 0.6 L), (b) HSSE with a conventional
Fig. 1. Tools for HS-STE and DC-STE samplings of solid biological matrices (for details see text).
twister (PDMS surface 89.8 mm2 , volume: 20 L, length: 1 cm, thickness: 0.5 mm), and (c) HS-STE with a PDMS tape (PDMS surface 60 mm2 because the upper surface was covered by the STE holder, volume: 30 L, length: 15 mm, width: 4 mm, thickness: 0.5 mm); sampling was at 1 cm from the leaf surface. The same leaves were also submitted to DC-STE sampling (d) with the same PDMS tape used for (c). The headspace of two freshly-cut rosemary leaves (20 mg) in a 7 mL vial was also sampled for 1 h at 25 ◦ C by STE at 1 cm from the leaf surface. The results of samplings using the PDMS tape as such (i.e. 120 mm2 ) or when its surface was halved by covering one side of it with silver paper (60 mm2 ) were compared. The PDMS tape was rested/supported on a simple 1 mm e.d. stainless steel wire support kept in position by inserting it in the tefloncoated septum of the vial. After sampling, PDMS twisters and tapes were removed from the sampling vial, inserted into a glass tube and then introduced into a thermodesorber (TDU, Gerstel, M¨ulheim a/d Ruhr, Germany) from where the analytes were recovered and analyzed by GC and GC–MS (see Section 2.4). HS-SPME fiber was thermally desorbed in a split/splitless GC injector and analyzed by GC and GC/MS. Each experiment was repeated three times. 2.2. DC-STE apple sampling A commercially available red apple was cut in half and the resulting pulp surface was sampled by DC-STE with a PDMS tape (length: 15 mm, width: 4 mm, thickness: 0.5 mm) for 30 min at room temperature. Another PDMS tape was inserted into the inside of the pulp of the same apple for 1.2 cm and left to sample, again for 30 min. The PDMS tapes were then thermally desorbed and the recovered samples analyzed on-line by GC and GC–MS.
C. Bicchi et al. / J. Chromatogr. A 1148 (2007) 137–144
2.3. Perfume released from skin
carvone, 10: cis-dihydrocarveol, 11: carvyl acetate, 12: transcalamenene, 13: trans-carveol, 14: cis-carveol Rosemary – Column: FSOT SE 52 (df 0.25 m, i.d 0.25 mm, length 30 m) (Mega, Legnano (Milan), Italy). Temperature program: from 50 ◦ C (1 min) to 220 ◦ C (5 min) at 3 ◦ C/min. Marker components: 1: ␣-pinene, 2: camphene, 3: 1-octen-3-ol, 4: -myrcene, 5: p-cymene, 6: 1,8-cineole, 7: limonene, 8: ␥terpinene, 9: trans-sabinene hydrate, 10: crysanthenone, 11: camphor, 12: borneol, 13: verbenone, 14: bornyl acetate, 15: -caryophyllene. Analysis of apple pulp – Column: FSOT Carbowax 20 M (df 0.25 m, i.d 0.25 mm, length 30 m) (Mega, Legnano (Milan), Italy). Temperature program: from 50 ◦ C (1 min) to 230 ◦ C (5 min) at 3 ◦ C/min. Analysis of perfume released from the skin – Column: FSOT SE 52 (df 0.25 m, i.d 0.25 mm, length 30 m) (Mega, Legnano (Milan), Italy). Temperature program: from 0 ◦ C (1 min) to 70 ◦ C at 80 ◦ C/min then to 230 ◦ C (5 min) at 3 ◦ C/min. Marker compounds: 1: dihydromyrcenol, ␣-i-methyl ionone, 3: MW 236, 4: dihydromethyl jasmonate, 5: benzyl salicylate. MSD conditions – MS operated in EI mode (70 eV), full scan with a mass range from 35 to 450 amu.
A limited part of the surface of the back of one hand of a volunteer, large enough to run 10 non-overlapping DCSTE and HS-STE samplings, was sprayed uniformly with a commercially-available perfume (Fig. 2). In DC-STE the PDMS tape (length: 15 mm, width: 4 mm, thickness: 0.5 mm) assembled on the STE holder inserted in vial B was rested on the sprayed surface of the hand for half an hour at 25 ◦ C. For both DC-STE and HS-STE, the holder (and as a consequence vial B) was correctly positioned by fixing it to a stainless steel bar support through a jointed clamp. DC-STE was carried out every 90 min (for a total of four samplings) and in different positions of the sprayed surface of the back of the hand. On the other hand, HS-STE was carried out every 45 min (six samplings) keeping the STE holder at 0.5 cm from the surface of the back of the hand. Each series of experiments was repeated twice. 2.4. Analysis conditions Analyte thermal desorption were carried out with a TDU unit from Gerstel (Gerstel, M¨ulheim a/d Ruhr, Germany) installed on an Agilent 6890 GC unit coupled to an Agilent 5973N MSD (Agilent, Little Falls, DE, USA). For TDU the following parameters were used: desorption program: from 30 ◦ C to 250 ◦ C (5 min) at 60 ◦ C/min; flow mode: splitless, transfer line: 250 ◦ C. A Gerstel CIS-4 PTV injector was used to focus cryogenically the analytes thermally desorbed from the PDMS stir bars or tapes. The PTV was cooled to −50 ◦ C using liquid CO2 ; injection: PTV; injection temperature: from −50 ◦ C to 250 ◦ C (5 min) at 12 ◦ C/s. The inlet was operated in the split mode for spearmint, rosemary and apple (split ratio 1:10) and perfume released from the skin (split ratio: 1:50). Chromatographic conditions: Helium was used as carrier gas at a flow rate of 1 mL/min. Spearmint – Column: FSOT Carbowax 20 M (df 0.25 m, i.d 0.25 mm, length 30 m) (Mega, Legnano (Milan), Italy). Temperature program: from 50 ◦ C (1 min) to 230 ◦ C (5 min) at 3 ◦ C/min. Marker components: 1: -bourbonene, 2: ␣gurjunene, 3: linalool, 4: -caryophyllene, 5: dihydrocarvone, 6: ␣-humulene, 7: dihydrocarvyl acetate, 8: germacrene D, 9:
Fig. 2. Diagram of the sampling strategy used to study the composition of the volatile fraction at the skin surface, and of that released from the skin over time, by DC-STE and HS-STE.
139
3. Results and discussion The potential of STE is illustrated through some applications in different fields. 3.1. DC-STE and HS-STE on living plants PDMS tapes were used to sample the volatile fraction of rosemary and spearmint leaves. The STE concentration capability in both direct contact and headspace modes was compared to that of HS-SPME with a PDMS 100 m fiber and to HSSE with a 20 L PDMS twister. Fig. 3 reports the GC–MS profiles of the headspace of a leaf of spearmint after HS-SPME (a), HSSE (b) and HS-STE (c) samplings and of the volatile fraction sampled by DC-STE (d). Fig. 4 reports the relative abundance of a group of spearmint marker compounds (listed in Table 1) calculated Table 1 Average area and RSD% of a group of spearmint marker compounds #
Compounds
Average area
RSD%
1 2 3 4 5 6 7 8 9 10 11 12 13 14
-Bourbonene ␣-Gurjunene Linalool -Caryophyllene Dihydrocarvone ␣-Humulene Dihydrocarvylacetate Germacrene D Carvone cis-Dihydrocarveol Carvylacetate trans-Calamenene trans-Carveol cis-Carveol
1035248 3507206 6787618 13499666 10094891 1962006 34575154 419949 715000000 48678554 66932855 3044209 38576087 30449050
8.5 6.7 9.1 5.8 10.3 8.3 6.3 4.6 4.1 5.6 6.2 6.5 4.8 5.6
140
C. Bicchi et al. / J. Chromatogr. A 1148 (2007) 137–144
Fig. 3. GC–MS profiles of the headspace of a leaf of spearmint after HS-SPME (a), HSSE (b) and HS-STE (c) samplings and of the volatile fraction sampled by DC-STE (d). Marker components: 1: -bourbonene, 2: ␣-gurjunene, 3: linalool, 4: -caryophyllene, 5: dihydrocarvone, 6: ␣-humulene, 7: dihydro-carvyl acetate, 8: germacrene D, 9: carvone, 10: cis-dihydrocarveol, 11: carvyl acetate, 12: cis-calamenene, 13: trans-carveol, 14: cis-carveol.
through their absolute area versus that obtained by HS-SPME, taken as reference and set at 100. Fig. 5 reports the GC–MS profiles of the headspace of a leaf of spearmint of the same plant cut in half after HS-SPME (a), HSSE (b) and HS-STE (c) samplings and of the volatile fraction sampled along the cut by DC-STE (d). Leaves were very carefully cut in order to reduce breakage of the essential oil secretory structures to a minimum. These results show that STE may be extremely useful to study variations in the
Fig. 4. Relative abundance of a group of marker compounds of spearmint calculated vs. HS-SPME taken as reference standard. For peak identification see Section 2 and the legend to Fig. 3.
composition of the volatile fraction due to the reaction at the surface of a leaf when submitted to stress. In particular, STE can be used to compare compositions at different distances from the site of stress, by direct contact sampling, as well as the composition of the emitted fraction, by headspace sampling. Although GC profiles look very different, at least in quantitative terms, in consideration of the limited number of analyses carried out to date, it is impossible to hypothesize any explanation of the biological phenomena involved. In-depth experiments on the chemical composition of the volatile fractions sampled by both DC-STE and HS-STE are under way to study whether it is possible to detect chemical markers monitoring plant stress. Sandra et al. [11] have shown that PDMS tape is highly reproducible when direct contact sampling on the skin is carried out. Our experiments gave comparable results. On the other hand, the intermediate precision of STE in headspace sampling was determined by sampling the headspace at a distance of 1 cm from three leaves of almost the same weight cut off the plant in a 7 mL vial, three times and with three different PDMS tapes from the same batch. Table 1 reports the average area and RSD% of a group of spearmint marker compounds: these results show that HS-STE intermediate precision is high, ranging between 4.1% for carvone and 10.3% for dihydrocarvone. The next group of experiments aimed at evaluating the influence of the sampling surface on recovery, by submitting rosemary leaves to two sets of experiments. In the first, sampling was done by using PDMS tape as such, in the second,
C. Bicchi et al. / J. Chromatogr. A 1148 (2007) 137–144
141
Fig. 5. GC–MS profiles of the headspace of a leaf of spearmint from the same plant, cut in half, sampled by HS-SPME (a), HSSE (b) and HS-STE (c) and of the volatile fraction sampled along the cut by DC-STE (d). For peak identification see Section 2 and the legend to Fig. 3.
the tape surface was halved by covering one of its sides with silver paper while keeping the PDMS volume constant (30 L). Fig. 6 reports the GC-FID profiles of the rosemary headspace when sampled with PDMS tapes with a sampling surface of 60 mm2 (a) and 120 mm2 (b), respectively. Table 2 reports the
mean absolute areas from three experiments of the rosemary marker compounds after STE headspace sampling using a PDMS tape with an effective sampling surface of 60 mm2 (one side) and 120 mm2 (two sides), respectively, and their two sides/one side area ratio. These results clearly show that
Table 2 Mean absolute areas from three experiments of the rosemary marker compounds after STE headspace sampling using a PDMS tape with an effective sampling surface of 60 mm2 (one side) and 120 mm2 (two sides), respectively, and their two sides/one side area ratio #
Ret. index
Compounds
Absolute area two sides
Absolute area one side
Ratio two sides/one side
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
931 943 965 982 1011 1020 1021 1047 1052 1096 1117 1147 1177 1264 1401
␣-Pinene Camphene 1-Octen-3-ol -Myrcene p-Cymene 1,8-Cineole Limonene ␥-Terpinene trans-Sabinene hydrate Crysanthenone Camphor Borneol Verbenone Bornyl acetate -Caryophyllene
355089 94496 94494 128115 13968 311314 113654 13177 20771 110434 309227 295487 307310 26053 44943
97437 29694 23024 39523 3836 80269 36176 4166 8498 43489 94038 123046 96385 25043 27938
3.64 3.18 4.10 3.24 3.64 3.88 3.14 3.16 2.44 2.54 3.29 2.40 3.19 1.04 1.61
142
C. Bicchi et al. / J. Chromatogr. A 1148 (2007) 137–144
Fig. 6. GC–MS profiles of rosemary headspace sampled with PDMS tapes with sampling surface of 60 mm2 (a) and 120 mm2 (b), respectively. Marker components: 1: ␣-pinene, 2: camphene, 3: 1-octen-3-ol, 4: -myrcene, 5: p-cymene, 6: 1,8-cineole, 7: limonene, 8: ␥-terpinene, 9: trans-sabinene hydrate, 10: crysanthenone, 11: camphor, 12: borneol, 13: verbenone, 14: bornyl acetate, 15: -caryophyllene.
analyte recovery is strongly influenced by the tape surface, even under rigorously static headspace conditions, i.e. recovery of most components was about three times higher by doubling the surface.
3.2. Apple reaction to stress This application aimed to show how effective a PDMS tape operating in direct contact mode can be to evaluate the reaction
Fig. 7. GC–MS profiles obtained after DC-STE sampling (a) of the cut pulp surface of a red apple cut in half, and of the inner pulp when the PDMS tape is inserted into it (b). 1: 2,3 dihydro-3,5 dihydroxy-6-methyl-4H pyran-4-one, 2: 5-hydroxy-methyl furfural.
C. Bicchi et al. / J. Chromatogr. A 1148 (2007) 137–144
of a vegetable matrix submitted to stress. A PDMS tape was used to sample the volatile fraction at the surface of the pulp of a red apple cut in half. In parallel, another PDMS tape was inserted into the pulp of the same apple perpendicularly to the previously-sampled cut surface for 1.2 cm and left to sample for 30 min. Fig. 7 reports the GC–MS profiles obtained after DCSTE sampling of the cut surface of the pulp and of the inside of the pulp. The two profiles are qualitatively similar but the areas of both 2,3-dihydro-3,5-dihydroxy-6-methyl-4H-pyran-4-one and of 5-hydroxy-methyl furfural are greatly increased. One of the possible explanations is that this increase might be related to hydrolysis of the corresponding glycosides, because of the stress caused by introducing the PDMS tape into the apple pulp. 3.3. Perfume released from the skin The composition of the volatile fraction at the skin surface and of that released from the skin over time were also investigated. HS-STE and DC-STE samplings were carried out on a defined surface on the back of the hand that was uniformly sprayed with a commercially-available perfume. Vial B was used to fix the sampling volume and to prevent pollution from the surrounding environment. Fig. 2 also shows the distribution of the sampling sites. Fig. 8 reports the GC–MS profiles of the volatiles sampled by DC-STE (a) and of those emitted from the skin sampled by HS-STE (b) after one hour’s sampling. Fig. 9 reports the decrease over time of the average peak areas, calculated on two experiments, of five components characteristic of the perfume investigated, on the skin surface sampled by DC-STE and on the emitted fraction sampled by HS-STE. Both figures show
143
how complementary and useful the two STE approaches can be when used to study the “life” of a perfume applied to the skin. In particular, STE enables those components that mainly influence a perfume’s olfactory properties to be detected, together with their evolution over time, their interaction with the skin surface and above all the differing release of different components as a consequence of their volatility, lipophilicity and/or the skin composition and structure. In-depth studies on this topic are under way. 3.4. General considerations Although the results reported here are only preliminary, STE looks very promising to sample the volatile fractions from a living matrix directly, and very useful to compare the composition at the emitting surface to that of the vapor phase in equilibrium with it (i.e. its headspace). HS-STE sampling from live solid matrices is purely based on the static headspace approach, and as a consequence its sampling rate (i.e. the amount of analyte recovered over time) is limited, since it cannot be increased by reducing the influence of the barriers to mass transport, due to both headspace/extraction phase and sample/headspace interfaces, by modifying parameters such as stirring, temperature or matrix volume, as was shown by Bruheim et al. [6]. In live solid matrices the only HS-STE parameters that can be varied are sampling time (i.e. equilibrium or non-equilibrium headspace sampling) and surface and volume of the sorption phase. Similar considerations may be made for DC-STE. In spite of this, the abundances of the spearmint marker compounds relative to SPME show that STE in both DC and HS modes provides recoveries higher than
Fig. 8. GC–MS profiles of the volatiles sampled by DC-STE (a) after spraying perfume onto the skin, and of those emitted by the skin, sampled by HS-STE (b) after one hour’s sampling. Marker compounds: 1: dihydromyrcenol, ␣-i-methyl ionone, 3: MW 236, 4: dihydromethyl jasmonate, 5: benzyl salicylate.
144
C. Bicchi et al. / J. Chromatogr. A 1148 (2007) 137–144
Fig. 9. Decrease over time of the peak areas of five components characteristic of the perfume investigated on the skin surface sampled by DC-STE (a) and on the emitted fraction sampled by HS-STE (b).
SPME and comparable to HSSE, as reported in Fig. 4. PDMS areas and volumes of both tapes and twisters used for these experiments are comparable and much higher than that of the SPME fiber, being 10 mm2 and 0.6 L for the PDMS 100 fiber, 89.8 mm2 and 20 L for the twister and 60 mm2 when just one side is exposed to the vapor phase or 120 mm2 both sides of the tape are exposed, and 30.0 L for the PDMS tape. The results of the rosemary experiments show that the exchange surface is the parameter that chiefly conditions recovery. STE by both direct and headspace samplings were able to detect analytes at a level of fractions of ppm achieving a sensitivity comparable to that reported for DC-STE by Sandra et al. [11] in their original publication. An in-depth study comparing the effectiveness of recovery over time and sensitivity from a standard solution of analytes with different volatilities and Ko/w by different sorptive sampling tools (fibers, twisters and tapes) in static mode will be the object of a forthcoming publication [12]. STE offers further advantages when a solid biological matrix has to be sampled, namely (a) the possibility (in particular for DC-STE) to place the PDMS tape in different positions on the surface of a biological matrix in order to evaluate the effect of distance from the point of stress on the composition of the volatile fraction emitted, and (b) the possibility to choose the most suitable position to maximize the exchange surface for HS sampling.
Acknowledgements CB, CC, EL, PR and BS are indebted to Centro di Eccellenza per la Biosensoristica Vegetale e Microbica (CEBIOVEM) of the University of Torino. References [1] C.L. Arthur, J. Pawliszyn, Anal. Chem. 62 (1990) 2145. [2] E. Baltussen, P. Sandra, F. David, C. Cramers, J. Microcol. Sep. 11 (1999) 737. [3] Z. Zhang, J. Pawliszyn, Anal. Chem. 65 (1993) 1843. [4] B. Tienpont, F. David, C. Bicchi, P. Sandra, J. Microcol. Sep. 12 (2000) 577. [5] C. Bicchi, C. Cordero, C. Iori, P. Rubiolo, P. Sandra, J. High Resolut. Chromatogr. 23 (2000) 539. [6] I. Bruheim, X. Liu, J. Pawliszyn, Anal. Chem. 75 (2003) 1002. [7] A. Segal, T. Gorecki, P. Mussche, J. Lips, J. Pawliszyn, J. Chromatogr. A 873 (2000) 13. [8] L. Wang, H. Lord, R. Morehead, F. Dorman, J. Pawliszyn, J. Agric. Food Chem. 50 (2002) 6281. [9] X. Liu, R. Pawliszyn, L. Wang, J. Pawliszyn, Analyst 129 (2004) 55. [10] C. Bicchi, C. Cordero, E. Liberto, P. Rubiolo, B. Sgorbini, P. Sandra, J. Chromatogr. A 1071 (2004) 111. [11] P. Sandra, S. Sisalli, A. Adao, M. Lebel, I. Le Fur, LC–GC Europe 19 (2006) 33. [12] C. Bicchi, C. Cordero, E. Liberto, P. Rubiolo, B. Sgorbini, P, Sandra, in preparation.
Sorptive tape extraction in the analysis of the volatile fraction emitted from biological solid matrices C. Bicchi a,∗ , C. Cordero a , E. Liberto a , P. Rubiolo a , B. Sgorbini a , P. Sandra b a
Dipartimento di Scienza e Tecnologia del Farmaco, Universit`a degli Studi di Torino, Via P. Giuria 9, I-10125 Torino, Italy b Research Institute for Chromatography, Kennedypark 20, B-8500 Kortrijk, Belgium Received 1 December 2006; received in revised form 26 February 2007; accepted 5 March 2007 Available online 12 March 2007
Abstract Sorptive tape extraction (STE) is a recent sorption-based sampling technique in which a flexible polydimethylsiloxane (PDMS) tape is used to recover analytes at the surface of a solid matrix by direct contact as well as from the headspace in equilibrium with it. Solutes thus enriched on the inert PDMS material can be recovered either by solvent desorption or by thermo-desorption. The concentration capability of both direct contact and headspace STE was evaluated by sampling (a) aromatic plants to study the reaction of a vegetable matrix submitted to stress, and (b) fruits at the surface of the pulp or inside the pulp; the composition of the volatile fraction released from the skin when a perfume is sprayed on the back of the hand was also studied. The concentration capability of direct contact and headspace STE was compared to that of HSSE with a 20 L PDMS twister and HS-SPME with a PDMS 100 m fibre, by determining the relative abundances (RA) of the characterizing components of the aromatic plants under investigation. Repeatability and influence of tape surface on STE recovery were also evaluated. © 2007 Elsevier B.V. All rights reserved. Keywords: Sorptive tape extraction; STE; Direct contact (DC); Headspace (HS); PDMS tape; Biological solid matrices
1. Introduction The chemical composition of fractions sampled from solid living organisms belonging to the vegetable or animal kingdoms, as well as that of the volatile fractions released from them, may be taken as an important biosensor diagnostic of changes taking place in the metabolism of living organisms. Detection of variations in the composition of such fractions, identification and quantitation of marker(s), monitoring reaction dynamics at the living surface of an organism submitted for instance to stress, all require an analytical set-up in which analyte recovery over time is maximized while analysis time is reduced to a minimum. Several solventless sampling techniques have recently been introduced to study these problems: the best known among them are solid phase microextraction (SPME) [1] and stir bar sorptive extraction (SBSE) [2], both of which have successfully been applied to in-solution and headspace sampling [3–5]. Analyte ∗ Corresponding author at: Dipartimento di Scienza e Tecnologia del Farmaco, Universit`a di Torino, Via Pietro Giuria 9, I-10125 Torino, Italy. Tel.: +39 011 670 7661/2; fax: +39 011 670 7687. E-mail address: [email protected] (C. Bicchi).
0021-9673/$ – see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.chroma.2007.03.007
recovery with both SPME and SBSE using polydimethylsiloxane (PDMS) as accumulation polymer is based on analyte partition between liquid or vapor phases and the polymer: this phenomenon is also known as sorption. Neither technique can successfully be applied to sampling a biological solid matrix by direct contact, because of the limited exchange surface of both fibers and twisters with the matrix itself. Bruheim et al. compared a thin sheet of a PDMS membrane to SPME for the extraction of semivolatile analytes through both in-solution and headspace modes [6]. They found that better extraction efficiency and sensitivity were achieved with the PDMS membrane than with the PDMS thick film fiber because of the larger surface area/extraction phase volume ratio. Membrane extraction was adopted by Pawliszyn and coworkers [7] in combination with a sorbent interface to sample dynamically the headspace of Eucaliptus dunnii Maiden [8,9]. This technique is known with the acronym of MESI (membrane extraction sorbent interface). In a study aiming to evaluate the influence of sampling conditions on HSSE recovery carried out on six highto-medium-volatility compounds with different KO/W , Bicchi et al. [10] recently showed, among others, that HSSE recovery was not only related to the volume of PDMS coating the twister, but
138
C. Bicchi et al. / J. Chromatogr. A 1148 (2007) 137–144
also to its surface in contact with the vapor phase. Very recently, Sandra et al. [11] introduced sorptive tape extraction (STE) to study the effect on sebum composition before and after cosmetic treatment through in vivo sampling at the human skin surface. In this STE application, sebum components were sampled by direct contact of a thin flexible PDMS tape with the skin surface for a fixed time to study the effect of a matifying product on skin shininess. The sampled analytes were then recovered by either thermal or solvent desorption and analyzed on-line by GC or GC–MS. The PDMS tapes used were 15 mm × 4 mm × 0.5 mm for thermal desorption and 15 mm × 12 mm × 0.5 mm for liquid desorption. The wide exchange surface with headspace offered by PDMS tape and its ability to be applied directly to the surface of a living matrix makes STE very interesting as a sampling technique to study the chemical messages emitted from plants or animals when a variation in their metabolism occurs. This article reports the preliminary results of direct contact and headspace STE sampling applied to aromatic plants, fruits, and the release of perfumes from human skin. The results obtained with aromatic plants were also compared to those obtained by HS-SPME and HSSE. The concentration capability of HS-STE, DC-STE and HSSE, taking HS-SPME with a PDMS 100 m fiber as reference standard, was evaluated by determining the relative abundances (RA) of the characterizing components of the aromatic plants under investigation. 2. Experimental Commercially-available live spearmint (Mentha spicata L.) and rosemary (Rosmarinus officinalis L.) plants, perfumes and apples were analyzed. 2.1. Aromatic plant sampling 2.1.1. STE sampling system STE sampling was done using the special assemblies shown in Fig. 1. Two different vials were used; with both of them the PDMS tape was kept in contact with the biological surface, or in a position suitable for headspace sampling, by means of the stainless steel holder shown in Fig. 1, which was fixed to the body of the vial by one of its lateral arms. Vial A was especially designed for sampling from live plants. One of the stoppers of the vial and its teflon-coated gasket were cut, leaving a narrow cleft through which one or more leaves can be introduced into the vial without removing them from the stem, thus affording sampling from live plant organisms. Vial B was designed for samplings from the skin. It consists of a 40 mL glass bell that is placed on the skin surface, at the top of which the STE holder is fixed. 2.1.2. Sampling conditions The headspace resulting from one leaf of spearmint (whether whole or crushed) corresponding to 50 mg of plant material in the 40 mL vial A described above was sampled for 1 h at 25 ◦ C by (a) HS-SPME with a PDMS 100 m fiber (PDMS surface: 10 mm2 , volume 0.6 L), (b) HSSE with a conventional
Fig. 1. Tools for HS-STE and DC-STE samplings of solid biological matrices (for details see text).
twister (PDMS surface 89.8 mm2 , volume: 20 L, length: 1 cm, thickness: 0.5 mm), and (c) HS-STE with a PDMS tape (PDMS surface 60 mm2 because the upper surface was covered by the STE holder, volume: 30 L, length: 15 mm, width: 4 mm, thickness: 0.5 mm); sampling was at 1 cm from the leaf surface. The same leaves were also submitted to DC-STE sampling (d) with the same PDMS tape used for (c). The headspace of two freshly-cut rosemary leaves (20 mg) in a 7 mL vial was also sampled for 1 h at 25 ◦ C by STE at 1 cm from the leaf surface. The results of samplings using the PDMS tape as such (i.e. 120 mm2 ) or when its surface was halved by covering one side of it with silver paper (60 mm2 ) were compared. The PDMS tape was rested/supported on a simple 1 mm e.d. stainless steel wire support kept in position by inserting it in the tefloncoated septum of the vial. After sampling, PDMS twisters and tapes were removed from the sampling vial, inserted into a glass tube and then introduced into a thermodesorber (TDU, Gerstel, M¨ulheim a/d Ruhr, Germany) from where the analytes were recovered and analyzed by GC and GC–MS (see Section 2.4). HS-SPME fiber was thermally desorbed in a split/splitless GC injector and analyzed by GC and GC/MS. Each experiment was repeated three times. 2.2. DC-STE apple sampling A commercially available red apple was cut in half and the resulting pulp surface was sampled by DC-STE with a PDMS tape (length: 15 mm, width: 4 mm, thickness: 0.5 mm) for 30 min at room temperature. Another PDMS tape was inserted into the inside of the pulp of the same apple for 1.2 cm and left to sample, again for 30 min. The PDMS tapes were then thermally desorbed and the recovered samples analyzed on-line by GC and GC–MS.
C. Bicchi et al. / J. Chromatogr. A 1148 (2007) 137–144
2.3. Perfume released from skin
carvone, 10: cis-dihydrocarveol, 11: carvyl acetate, 12: transcalamenene, 13: trans-carveol, 14: cis-carveol Rosemary – Column: FSOT SE 52 (df 0.25 m, i.d 0.25 mm, length 30 m) (Mega, Legnano (Milan), Italy). Temperature program: from 50 ◦ C (1 min) to 220 ◦ C (5 min) at 3 ◦ C/min. Marker components: 1: ␣-pinene, 2: camphene, 3: 1-octen-3-ol, 4: -myrcene, 5: p-cymene, 6: 1,8-cineole, 7: limonene, 8: ␥terpinene, 9: trans-sabinene hydrate, 10: crysanthenone, 11: camphor, 12: borneol, 13: verbenone, 14: bornyl acetate, 15: -caryophyllene. Analysis of apple pulp – Column: FSOT Carbowax 20 M (df 0.25 m, i.d 0.25 mm, length 30 m) (Mega, Legnano (Milan), Italy). Temperature program: from 50 ◦ C (1 min) to 230 ◦ C (5 min) at 3 ◦ C/min. Analysis of perfume released from the skin – Column: FSOT SE 52 (df 0.25 m, i.d 0.25 mm, length 30 m) (Mega, Legnano (Milan), Italy). Temperature program: from 0 ◦ C (1 min) to 70 ◦ C at 80 ◦ C/min then to 230 ◦ C (5 min) at 3 ◦ C/min. Marker compounds: 1: dihydromyrcenol, ␣-i-methyl ionone, 3: MW 236, 4: dihydromethyl jasmonate, 5: benzyl salicylate. MSD conditions – MS operated in EI mode (70 eV), full scan with a mass range from 35 to 450 amu.
A limited part of the surface of the back of one hand of a volunteer, large enough to run 10 non-overlapping DCSTE and HS-STE samplings, was sprayed uniformly with a commercially-available perfume (Fig. 2). In DC-STE the PDMS tape (length: 15 mm, width: 4 mm, thickness: 0.5 mm) assembled on the STE holder inserted in vial B was rested on the sprayed surface of the hand for half an hour at 25 ◦ C. For both DC-STE and HS-STE, the holder (and as a consequence vial B) was correctly positioned by fixing it to a stainless steel bar support through a jointed clamp. DC-STE was carried out every 90 min (for a total of four samplings) and in different positions of the sprayed surface of the back of the hand. On the other hand, HS-STE was carried out every 45 min (six samplings) keeping the STE holder at 0.5 cm from the surface of the back of the hand. Each series of experiments was repeated twice. 2.4. Analysis conditions Analyte thermal desorption were carried out with a TDU unit from Gerstel (Gerstel, M¨ulheim a/d Ruhr, Germany) installed on an Agilent 6890 GC unit coupled to an Agilent 5973N MSD (Agilent, Little Falls, DE, USA). For TDU the following parameters were used: desorption program: from 30 ◦ C to 250 ◦ C (5 min) at 60 ◦ C/min; flow mode: splitless, transfer line: 250 ◦ C. A Gerstel CIS-4 PTV injector was used to focus cryogenically the analytes thermally desorbed from the PDMS stir bars or tapes. The PTV was cooled to −50 ◦ C using liquid CO2 ; injection: PTV; injection temperature: from −50 ◦ C to 250 ◦ C (5 min) at 12 ◦ C/s. The inlet was operated in the split mode for spearmint, rosemary and apple (split ratio 1:10) and perfume released from the skin (split ratio: 1:50). Chromatographic conditions: Helium was used as carrier gas at a flow rate of 1 mL/min. Spearmint – Column: FSOT Carbowax 20 M (df 0.25 m, i.d 0.25 mm, length 30 m) (Mega, Legnano (Milan), Italy). Temperature program: from 50 ◦ C (1 min) to 230 ◦ C (5 min) at 3 ◦ C/min. Marker components: 1: -bourbonene, 2: ␣gurjunene, 3: linalool, 4: -caryophyllene, 5: dihydrocarvone, 6: ␣-humulene, 7: dihydrocarvyl acetate, 8: germacrene D, 9:
Fig. 2. Diagram of the sampling strategy used to study the composition of the volatile fraction at the skin surface, and of that released from the skin over time, by DC-STE and HS-STE.
139
3. Results and discussion The potential of STE is illustrated through some applications in different fields. 3.1. DC-STE and HS-STE on living plants PDMS tapes were used to sample the volatile fraction of rosemary and spearmint leaves. The STE concentration capability in both direct contact and headspace modes was compared to that of HS-SPME with a PDMS 100 m fiber and to HSSE with a 20 L PDMS twister. Fig. 3 reports the GC–MS profiles of the headspace of a leaf of spearmint after HS-SPME (a), HSSE (b) and HS-STE (c) samplings and of the volatile fraction sampled by DC-STE (d). Fig. 4 reports the relative abundance of a group of spearmint marker compounds (listed in Table 1) calculated Table 1 Average area and RSD% of a group of spearmint marker compounds #
Compounds
Average area
RSD%
1 2 3 4 5 6 7 8 9 10 11 12 13 14
-Bourbonene ␣-Gurjunene Linalool -Caryophyllene Dihydrocarvone ␣-Humulene Dihydrocarvylacetate Germacrene D Carvone cis-Dihydrocarveol Carvylacetate trans-Calamenene trans-Carveol cis-Carveol
1035248 3507206 6787618 13499666 10094891 1962006 34575154 419949 715000000 48678554 66932855 3044209 38576087 30449050
8.5 6.7 9.1 5.8 10.3 8.3 6.3 4.6 4.1 5.6 6.2 6.5 4.8 5.6
140
C. Bicchi et al. / J. Chromatogr. A 1148 (2007) 137–144
Fig. 3. GC–MS profiles of the headspace of a leaf of spearmint after HS-SPME (a), HSSE (b) and HS-STE (c) samplings and of the volatile fraction sampled by DC-STE (d). Marker components: 1: -bourbonene, 2: ␣-gurjunene, 3: linalool, 4: -caryophyllene, 5: dihydrocarvone, 6: ␣-humulene, 7: dihydro-carvyl acetate, 8: germacrene D, 9: carvone, 10: cis-dihydrocarveol, 11: carvyl acetate, 12: cis-calamenene, 13: trans-carveol, 14: cis-carveol.
through their absolute area versus that obtained by HS-SPME, taken as reference and set at 100. Fig. 5 reports the GC–MS profiles of the headspace of a leaf of spearmint of the same plant cut in half after HS-SPME (a), HSSE (b) and HS-STE (c) samplings and of the volatile fraction sampled along the cut by DC-STE (d). Leaves were very carefully cut in order to reduce breakage of the essential oil secretory structures to a minimum. These results show that STE may be extremely useful to study variations in the
Fig. 4. Relative abundance of a group of marker compounds of spearmint calculated vs. HS-SPME taken as reference standard. For peak identification see Section 2 and the legend to Fig. 3.
composition of the volatile fraction due to the reaction at the surface of a leaf when submitted to stress. In particular, STE can be used to compare compositions at different distances from the site of stress, by direct contact sampling, as well as the composition of the emitted fraction, by headspace sampling. Although GC profiles look very different, at least in quantitative terms, in consideration of the limited number of analyses carried out to date, it is impossible to hypothesize any explanation of the biological phenomena involved. In-depth experiments on the chemical composition of the volatile fractions sampled by both DC-STE and HS-STE are under way to study whether it is possible to detect chemical markers monitoring plant stress. Sandra et al. [11] have shown that PDMS tape is highly reproducible when direct contact sampling on the skin is carried out. Our experiments gave comparable results. On the other hand, the intermediate precision of STE in headspace sampling was determined by sampling the headspace at a distance of 1 cm from three leaves of almost the same weight cut off the plant in a 7 mL vial, three times and with three different PDMS tapes from the same batch. Table 1 reports the average area and RSD% of a group of spearmint marker compounds: these results show that HS-STE intermediate precision is high, ranging between 4.1% for carvone and 10.3% for dihydrocarvone. The next group of experiments aimed at evaluating the influence of the sampling surface on recovery, by submitting rosemary leaves to two sets of experiments. In the first, sampling was done by using PDMS tape as such, in the second,
C. Bicchi et al. / J. Chromatogr. A 1148 (2007) 137–144
141
Fig. 5. GC–MS profiles of the headspace of a leaf of spearmint from the same plant, cut in half, sampled by HS-SPME (a), HSSE (b) and HS-STE (c) and of the volatile fraction sampled along the cut by DC-STE (d). For peak identification see Section 2 and the legend to Fig. 3.
the tape surface was halved by covering one of its sides with silver paper while keeping the PDMS volume constant (30 L). Fig. 6 reports the GC-FID profiles of the rosemary headspace when sampled with PDMS tapes with a sampling surface of 60 mm2 (a) and 120 mm2 (b), respectively. Table 2 reports the
mean absolute areas from three experiments of the rosemary marker compounds after STE headspace sampling using a PDMS tape with an effective sampling surface of 60 mm2 (one side) and 120 mm2 (two sides), respectively, and their two sides/one side area ratio. These results clearly show that
Table 2 Mean absolute areas from three experiments of the rosemary marker compounds after STE headspace sampling using a PDMS tape with an effective sampling surface of 60 mm2 (one side) and 120 mm2 (two sides), respectively, and their two sides/one side area ratio #
Ret. index
Compounds
Absolute area two sides
Absolute area one side
Ratio two sides/one side
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
931 943 965 982 1011 1020 1021 1047 1052 1096 1117 1147 1177 1264 1401
␣-Pinene Camphene 1-Octen-3-ol -Myrcene p-Cymene 1,8-Cineole Limonene ␥-Terpinene trans-Sabinene hydrate Crysanthenone Camphor Borneol Verbenone Bornyl acetate -Caryophyllene
355089 94496 94494 128115 13968 311314 113654 13177 20771 110434 309227 295487 307310 26053 44943
97437 29694 23024 39523 3836 80269 36176 4166 8498 43489 94038 123046 96385 25043 27938
3.64 3.18 4.10 3.24 3.64 3.88 3.14 3.16 2.44 2.54 3.29 2.40 3.19 1.04 1.61
142
C. Bicchi et al. / J. Chromatogr. A 1148 (2007) 137–144
Fig. 6. GC–MS profiles of rosemary headspace sampled with PDMS tapes with sampling surface of 60 mm2 (a) and 120 mm2 (b), respectively. Marker components: 1: ␣-pinene, 2: camphene, 3: 1-octen-3-ol, 4: -myrcene, 5: p-cymene, 6: 1,8-cineole, 7: limonene, 8: ␥-terpinene, 9: trans-sabinene hydrate, 10: crysanthenone, 11: camphor, 12: borneol, 13: verbenone, 14: bornyl acetate, 15: -caryophyllene.
analyte recovery is strongly influenced by the tape surface, even under rigorously static headspace conditions, i.e. recovery of most components was about three times higher by doubling the surface.
3.2. Apple reaction to stress This application aimed to show how effective a PDMS tape operating in direct contact mode can be to evaluate the reaction
Fig. 7. GC–MS profiles obtained after DC-STE sampling (a) of the cut pulp surface of a red apple cut in half, and of the inner pulp when the PDMS tape is inserted into it (b). 1: 2,3 dihydro-3,5 dihydroxy-6-methyl-4H pyran-4-one, 2: 5-hydroxy-methyl furfural.
C. Bicchi et al. / J. Chromatogr. A 1148 (2007) 137–144
of a vegetable matrix submitted to stress. A PDMS tape was used to sample the volatile fraction at the surface of the pulp of a red apple cut in half. In parallel, another PDMS tape was inserted into the pulp of the same apple perpendicularly to the previously-sampled cut surface for 1.2 cm and left to sample for 30 min. Fig. 7 reports the GC–MS profiles obtained after DCSTE sampling of the cut surface of the pulp and of the inside of the pulp. The two profiles are qualitatively similar but the areas of both 2,3-dihydro-3,5-dihydroxy-6-methyl-4H-pyran-4-one and of 5-hydroxy-methyl furfural are greatly increased. One of the possible explanations is that this increase might be related to hydrolysis of the corresponding glycosides, because of the stress caused by introducing the PDMS tape into the apple pulp. 3.3. Perfume released from the skin The composition of the volatile fraction at the skin surface and of that released from the skin over time were also investigated. HS-STE and DC-STE samplings were carried out on a defined surface on the back of the hand that was uniformly sprayed with a commercially-available perfume. Vial B was used to fix the sampling volume and to prevent pollution from the surrounding environment. Fig. 2 also shows the distribution of the sampling sites. Fig. 8 reports the GC–MS profiles of the volatiles sampled by DC-STE (a) and of those emitted from the skin sampled by HS-STE (b) after one hour’s sampling. Fig. 9 reports the decrease over time of the average peak areas, calculated on two experiments, of five components characteristic of the perfume investigated, on the skin surface sampled by DC-STE and on the emitted fraction sampled by HS-STE. Both figures show
143
how complementary and useful the two STE approaches can be when used to study the “life” of a perfume applied to the skin. In particular, STE enables those components that mainly influence a perfume’s olfactory properties to be detected, together with their evolution over time, their interaction with the skin surface and above all the differing release of different components as a consequence of their volatility, lipophilicity and/or the skin composition and structure. In-depth studies on this topic are under way. 3.4. General considerations Although the results reported here are only preliminary, STE looks very promising to sample the volatile fractions from a living matrix directly, and very useful to compare the composition at the emitting surface to that of the vapor phase in equilibrium with it (i.e. its headspace). HS-STE sampling from live solid matrices is purely based on the static headspace approach, and as a consequence its sampling rate (i.e. the amount of analyte recovered over time) is limited, since it cannot be increased by reducing the influence of the barriers to mass transport, due to both headspace/extraction phase and sample/headspace interfaces, by modifying parameters such as stirring, temperature or matrix volume, as was shown by Bruheim et al. [6]. In live solid matrices the only HS-STE parameters that can be varied are sampling time (i.e. equilibrium or non-equilibrium headspace sampling) and surface and volume of the sorption phase. Similar considerations may be made for DC-STE. In spite of this, the abundances of the spearmint marker compounds relative to SPME show that STE in both DC and HS modes provides recoveries higher than
Fig. 8. GC–MS profiles of the volatiles sampled by DC-STE (a) after spraying perfume onto the skin, and of those emitted by the skin, sampled by HS-STE (b) after one hour’s sampling. Marker compounds: 1: dihydromyrcenol, ␣-i-methyl ionone, 3: MW 236, 4: dihydromethyl jasmonate, 5: benzyl salicylate.
144
C. Bicchi et al. / J. Chromatogr. A 1148 (2007) 137–144
Fig. 9. Decrease over time of the peak areas of five components characteristic of the perfume investigated on the skin surface sampled by DC-STE (a) and on the emitted fraction sampled by HS-STE (b).
SPME and comparable to HSSE, as reported in Fig. 4. PDMS areas and volumes of both tapes and twisters used for these experiments are comparable and much higher than that of the SPME fiber, being 10 mm2 and 0.6 L for the PDMS 100 fiber, 89.8 mm2 and 20 L for the twister and 60 mm2 when just one side is exposed to the vapor phase or 120 mm2 both sides of the tape are exposed, and 30.0 L for the PDMS tape. The results of the rosemary experiments show that the exchange surface is the parameter that chiefly conditions recovery. STE by both direct and headspace samplings were able to detect analytes at a level of fractions of ppm achieving a sensitivity comparable to that reported for DC-STE by Sandra et al. [11] in their original publication. An in-depth study comparing the effectiveness of recovery over time and sensitivity from a standard solution of analytes with different volatilities and Ko/w by different sorptive sampling tools (fibers, twisters and tapes) in static mode will be the object of a forthcoming publication [12]. STE offers further advantages when a solid biological matrix has to be sampled, namely (a) the possibility (in particular for DC-STE) to place the PDMS tape in different positions on the surface of a biological matrix in order to evaluate the effect of distance from the point of stress on the composition of the volatile fraction emitted, and (b) the possibility to choose the most suitable position to maximize the exchange surface for HS sampling.
Acknowledgements CB, CC, EL, PR and BS are indebted to Centro di Eccellenza per la Biosensoristica Vegetale e Microbica (CEBIOVEM) of the University of Torino. References [1] C.L. Arthur, J. Pawliszyn, Anal. Chem. 62 (1990) 2145. [2] E. Baltussen, P. Sandra, F. David, C. Cramers, J. Microcol. Sep. 11 (1999) 737. [3] Z. Zhang, J. Pawliszyn, Anal. Chem. 65 (1993) 1843. [4] B. Tienpont, F. David, C. Bicchi, P. Sandra, J. Microcol. Sep. 12 (2000) 577. [5] C. Bicchi, C. Cordero, C. Iori, P. Rubiolo, P. Sandra, J. High Resolut. Chromatogr. 23 (2000) 539. [6] I. Bruheim, X. Liu, J. Pawliszyn, Anal. Chem. 75 (2003) 1002. [7] A. Segal, T. Gorecki, P. Mussche, J. Lips, J. Pawliszyn, J. Chromatogr. A 873 (2000) 13. [8] L. Wang, H. Lord, R. Morehead, F. Dorman, J. Pawliszyn, J. Agric. Food Chem. 50 (2002) 6281. [9] X. Liu, R. Pawliszyn, L. Wang, J. Pawliszyn, Analyst 129 (2004) 55. [10] C. Bicchi, C. Cordero, E. Liberto, P. Rubiolo, B. Sgorbini, P. Sandra, J. Chromatogr. A 1071 (2004) 111. [11] P. Sandra, S. Sisalli, A. Adao, M. Lebel, I. Le Fur, LC–GC Europe 19 (2006) 33. [12] C. Bicchi, C. Cordero, E. Liberto, P. Rubiolo, B. Sgorbini, P, Sandra, in preparation.