Pestic. Sci. 1998, 53, 300È310

for Determining the Vapour Pressure ofMethodsActive Ingredients Used in Crop Protection.

V: Thermogravimetry Combined with SolidPartPhase MicroExtraction (SPME)”

Thomas Kro� hl,* Roland Ka� stel, Werner Ko� nig, Hans Ziegler, Harald Ko� hle &Adolf Parg

BASF Aktiengesellschaft, Crop Protection Product Development, 67114 Limburgerhof, Germany

6 November 1997 ; revised version received 24 February 1998 ; accepted 30 March 1998)

Abstract : The well-established evaporation rate method for determining thevapour pressure of active ingredients in crop protection down to the order of10~12 mbar can be supplemented by the new sample preparation technique ofSolid Phase MicroExtraction (SPME). With this technique, it is possible to iden-tify evaporating substances by analysis after partitioning into the polymercoating of a thin Ðbre in the outlet-gas Ñow of thermogravimetric equipment.

The active ingredients fenpropimorph, kresoxim-methyl, metolachlor, cloma-zone and (Z)-9-dodecenyl acetate were used in this study, which showed that,despite the relatively small amount of collected material, an analytical identiÐca-tion of the evaporating compound by SPME/GC can be successfully achieved. Inparticular, the experiments have demonstrated a clear correlation between thelinearity of the weight loss curve and the evaporation process of a pure com-pound.

In the case of organic compounds that are unstable to heat, the SPME methodcan also be utilized to show whether, and at what temperatures, decompositionof the sample into fragments of higher volatility occurs. For example, the insecti-cide dimethoate showed a clear temperature dependence of both evaporationbehaviour and in the SPME/GC analysis. 1998 SCI(

Pestic. Sci., 53, 300È310 (1998)

Key words : vapour pressure ; evaporation rate ; thermogravimetry ; active ingre-dients in crop protection ; solid phase microextraction (SPME) ; gas chromatog-raphy

1 INTRODUCTION

One of the most important physical parameters ofactive substances in crop protection is their vapour

* To whom correspondence should be addressed.Email address : thomas.kroehl=apd.x400.basf-ag.de” Part IV: Gu� ckel, W. et al., Pestic Sci. 45 (1995) 27È31.

pressure. This parameter belongs to the standardproperties of the physicochemical data base for theregistration of plant protection products. It is the majorauxiliary variable for calculating auxiliary volatility,and thus it is a signiÐcant factor for predicting atmo-spheric concentrations of the substances.

Our method for determining vapour pressure basedon measurements of the evaporation rate at di†erent

3001998 SCI. Pestic. Sci. 0031-613X/98/$17.50. Printed in Great Britain(

V apour pressure of crop protection AI by thermogravimetry and SPME 301

temperatures at ambient pressure is described in earlierpublications.1h3 Using the newer, thermogravimetricmethod (TG method), which is a further development ofprevious procedures, it is now possible to determinevapour pressures of solid and liquid substances as lowas 10~12 mbar (10~10 Pa).4

A disadvantage of the existing method is that only anindirect determination of the evaporated compounds ismade. The method is based on the assumption that alinear decrease in the weight of the sample during iso-thermal measurement corresponds to the evaporation ofthe pure compound. However, a speciÐc analyticaldetermination of the evaporating substances is neededto demonstrate that this assumption is correct.

The well-known gas saturation method5 uses gaschromatography (GC) or gravimetry as the speciÐc ana-lytical method, depending on the particular samplecharacteristics. Commercially available equipment forthe gas saturation method, such as Netzsch Dampf-druckanalysator VPA 434 work, for example, with a gaschromatograph for analysis. In the apparatus VPA 434,a stream of nitrogen gas is passed over the substance insuch a way that it becomes saturated with its vapourand the latter is then collected in a suitable trap (e.g.Tenax). Measurement of the amount of material trans-ported by a known amount of carrier gas is used tocalculate the vapour pressure at a given temperature.However, this apparatus is restricted to vapour pres-sures above 10~5 Pa and, therefore, is not suitable foractive ingredients with lower volatility.6

In recent years, solid phase microextraction (SPME)has developed into a powerful method to preparesamples for chemical analysis. SPME was developed inthe late 1980s by J. Pawliszyn and co-workers at theUniversity of Waterloo, Canada, as a new method toprepare samples without using any solvent and with norequirement for sample work-up.7h10 The main part ofan SPME device is a Ðne fused-silica Ðbre which iscoated with a polymer sorbent that has a high affinityfor organic compounds. The SPME procedure consistsbasically of two steps : (1) extracting analytes by par-titioning them between the sample and the coating ofthe Ðbre and (2) desorption of enriched analytes into ananalytical instrument. In this way, organic compoundscan be concentrated from very dilute aqueous orgaseous samples without using any solvent.11

Applications of SPME most commonly reported inthe chemical literature include the analysis of environ-mental water or air samples,12 the determination ofpesticides,13,14 soil analysis,15,16 and many furtherexamples in the food, pharmaceutical and toxicologicalÐelds. Analysis of pesticides is reported almost exclu-sively for aqueous samples, whereas the detection ofvolatile compounds in the vapour phase is relativelyrare. An interesting example of analysis in biology hasbeen shown for airborne pheromones, where SPME wasused to analyse volatiles emitted by living insects.17

This paper describes an analytical determination ofevaporated chemicals using a combination of our estab-lished TG method with the SPME technique for samplepreparation and gas chromatography as standard ana-lytical method.

2 MATERIALS AND METHODS

2.1 Materials

Fenpropimorph (1), kresoxim-methyl (2), metolachlor(3), clomazone (4), (Z)-9-dodecenyl acetate (5) anddimethoate (6) analytical standards were used for theexperiments in this study (Fig. 1). These are the activeingredients of the commercial products “CorbelÏTM,“DiscusÏTM, “PyracurÏTM, “CommandÏTM, “Rak 1 PlusÏTMand “PerfekthionÏTM, respectively. The properties of theactive ingredients are presented in Table 1. Methanoland acetone (p.a.) were used as solvents for methoddevelopment.

2.2 SPME device

The SPME device (Supelco, Deisenhofen, Germany)used in this study is shown in Fig. 2. The sorbent for

Fig. 1. Active substances used in this study.

302 T homas Kro� hl et al.

TABLE 1Properties of Active Ingredients Used in this Study

Purity m.p.a V apour pressureActive ingredient (%) Rel. mol. mass (¡C) at 20¡C (mbar)b

Fenpropimorph (1) 99É6 303É5 Oil 2É20 ] 10~5Kresoxim-methyl (2) 99É7 313É4 100 2É30 ] 10~8Metolachlor (3) 98É1 283É8 Oil 1É70 ] 10~5Clomazone (4) 99É7 239É7 Oil 1É92 ] 10~4(Z)-9-Dodecenyl acetate (5) 93É5 226É4 Oil 6É50 ] 10~4Dimethoate (6) 99É8 229É3 45 2É90 ] 10~6

a m.p.\ melting point, determined by DSC (di†erential scanning calorimetry).b 1 mbar \ 100 Pa.

extracting the analytes from aqueous or gas phases iscoated onto a fused-silica Ðbre that acts as a carrier rod.The Ðbre is connected to a stainless steel tube toincrease the mechanical strength. A specially designedsyringe is used for taking up the stainless steel rod bywithdrawing it into the syringe needle. The syringemakes the device portable and allows, for example, foran easy accommodation in a gas chromatographinjector.

A variety of sorbent materials can be used for SPMEof di†erent groups of analytes. The most successful sta-

tionary phase is polydimethylsiloxane (PDMS), which isavailable in thicknesses of 7, 30 and 100 lm and is espe-cially suitable for non-polar compounds. Polyacrylateand carbowax are used for polar compounds, whereaspolyimide coatings extract chlorinated hydrocarbons.

For all the active ingredients (see Table 1) in thisstudy, non-polar PDMS-coated Ðbres were used.

To extract analytes from their matrix, the Ðbre has tobe pushed out of the needle by pressing down theplunger (Fig. 2). After an appropriate time of exposureof the coating to the sample, the Ðbre is completely

Fig. 2. Schematic diagram of a SPME device (from Ref. 11).

V apour pressure of crop protection AI by thermogravimetry and SPME 303

TABLE 2Chromatographic Conditions Used in this Study

Active ingredientParameter Condition (cf. T able 1)

GC HP 5890 1Varian 3600 2È6

Column CP-Sil-5 CB, Length 25 m, Diam. 0É32 mm 1HP 5, Length 30 m, Diam. 0É32 mm 2È6

Carrier gas N2 , 30 ml min~1 1H2 , 30 ml min~1 1È6He, 1 bar 2È6Syn. air, 300 ml min~1 1È6

Detector FID 1È6Injector temp. 240¡C 1

280¡C 2È6Detector temp. 280¡C 1È6Oven temp. 100¡C 3 min, 40¡C min~1, 250¡C 5 min 1

200¡C 260¡C 3 min, 30¡C min~1, 220¡C 5.7 min 3, 460¡C 3 min, 30¡C min~1, 220¡C 3.7 min 560¡C 3 min, 30¡C min~1, 150¡C 14 min 6

Split 1 : 5 11 : 4 2È6

Film coating thickness 7 lm PDMSa 2È630 lm PDMS 2È5100 lm PDMS 1È4

a PDMS\ polydimethylsiloxane.

withdrawn into the needle and immediately transferredto a gas chromatograph, where the analytes are ther-mally desorbed from the Ðbre coating in the heatedinjector port for analysis. Further technical details andthe theoretical background of SPME are broadlydescribed in the literature.9,11

2.3 Gas chromatography

HP 5890 or Varian 3600 gas chromatographs were usedfor analysis. Existing GC methods for the active ingre-dients in Fig. 1 were evaluated and modiÐed to givesatisfactory chromatographic results. The chromato-graphic conditions used in this study are shown inTable 2.

Typical chromatograms of compound 1 are presentedin Fig. 3. For each preparation the retention time of6É75 min was found to be reproducible. The retentiontimes for all compounds investigated are listed in Table3.

2.4 Combining evaporation rate measurements withSPME

A schematic diagram of the device used to determineevaporation rates is shown in Fig. 4.4 The samplecarrier plate, hanging on a microbalance in a thermo-regulated casing, is swept by a stream of dry nitrogengas which carries the vaporized molecules. The evapo-

Fig. 3. GC chromatograms for compound 1, obtained underconditions described in Table 2. (A) Acetonic solution of 1(45.2 mg litre~1) ; (B) Aqueous solution of 1 (124.0 mg litre~1).

TABLE 3GC Retention Times for Compounds Investigated

Retention timeaActive ingredient (min)

Fenpropimorph (1) 6É75Kresoxim-methyl (2) 10É67Metolachlor (3) 11É05Clomazone (4) 9É57(Z)-9-Dodecenyl acetate (5) 8É57Dimethoate (6) 16É4

a Determined with standard solutions in acetone (1 g litre~1).

304 T homas Kro� hl et al.

Fig. 4. Schematic diagram of the apparatus used to determinethe evaporation rates of active ingredients, combined with an

SPME device.

ration rate v at a deÐnite temperature is calculated fromthe weight loss *m of the sample plate by

v\ *mFt

(g cm~2 h~1),

where F\ surface area of the sample plate (cm2), andt \ time (h) for a weight loss *m (g).

A sorption unit prevents environmental pollution. Inorder to identify the evaporating material, a small pro-portion of it is collected between the furnace and thesorption unit for subsequent analysis. As shown in Fig.4, a SPME device is installed for this purpose directly atthe outlet of the thermogravimetric furnace. The SPMEis held by a Ðxed adapter with an inner thread intowhich the syringe can be screwed.

3 RESULTS AND DISCUSSION

3.1 Extraction from a saturated vapour phase

Extraction of compounds from their own saturatedvapour phase was carried out in 10-ml ampoules, closedwith a septum, where a small amount of the substancewas deposited on the bottom of the ampoule. Afterpiercing the septum with the SPME needle, the coated

Ðbre was exposed to the sample vapour and absorptioninto the coating took place.

3.1.1 Dependence on extraction timeFigure 5 shows a SPME/GC chromatogram for fenpro-pimorph (1), extracted with a 100-lm PDMS Ðbre withan extraction time of 5 h. The peak of the compoundappears sharply at the same retention time as in thechromatograms of Fig. 3. The dependence of the peakarea on the extraction time is shown in Fig. 6. The peakarea increases linearly with time. This illustrates that, asa consequence of the low vapour pressure of the com-pound at room temperature (Table 1), the equilibriumbetween the sample fraction dissolved in the siloxanephase and that remaining in the sample bulk has notyet been reached. The coated Ðbre acts as a sink,absorbing successively all molecules of the slowlyevaporating analyte present in the vapour phase.

3.1.2 Di†erent Ðbres with the same coating thicknessIn order to determine the reproducibility between Ðbres,the absorption behaviour of three di†erent Ðbres withthe same coating thickness (100 lm PDMS) and withthe same active ingredient (1) was investigated. Resultsof measurements for six extraction times between 10and 60 min are presented in Fig. 7. Although the peakarea at each point of time di†ered in the order of^10%, the rates of absorption for each of the threeÐbres investigated were in good agreement, i.e.repeatable.

Fig. 5. SPME/GC chromatogram for substance 1, extractedwith a PDMS Ðbre (100 lm) at room temperature with an

extraction time of 5 h.

Fig. 6. Dependence of the SPME/GC peak area on theextraction time for the extraction of fenpropimorph (1) from

its own vapour phase at room temperature.

V apour pressure of crop protection AI by thermogravimetry and SPME 305

Fig. 7. Dependence of the SPME/GC peak area on theextraction time for fenpropimorph (1), extracted at room tem-perature with three di†erent Ðbres with the same coating

thickness (100 lm PDMS).

3.1.3 Di†erent thicknesses of the Ðbre coatingMetolachlor (3) and clomazone (4) were extracted fromtheir vapour phase in 10 ml ampoules at 60¡C byplacing the ampoules together with the inserted SPMEdevice in a heating chamber. The higher temperaturewas used to maintain more molecules in the gaseousphase and therefore more vaporized material for collec-tion by the Ðbre in an appropriate time. Results for anextraction time of 1 h and Ðbre coating thicknesses of 7,30 and 100 lm are shown in Fig. 8. Metolachlor (A)

could be detected very well with all three Ðbre thick-nesses, more material being taken up by thicker Ðbrecoatings. In the case of clomazone (B), an increase inpeak broadening with coating thickness was observedwhereby only the 7- and 30-lm Ðbres were suitable forextraction. The release of analyte molecules from the100-lm Ðbre was delayed and an unacceptably broadGC peak was obtained. Obviously, the solubility of clo-mazone in the coating material is higher than that ofmetolachlor and the desorption of clomazone is thusmore delayed.

3.2 Extraction at the outlet of the thermogravimetricsystem

3.2.1 T hermally stable compoundsFenpropimorph (1) was used in the Ðrst experimentswhere evaporation rate determinations had been com-bined with SPME extraction. The results of measure-ments with an extraction time of 1 h at oventemperatures of 80 and 100¡C are presented in Tables 4and 5. After calibration with standard solutions inacetone, the peak areas were converted to extracted

Fig. 8. SPME/GC chromatograms obtained for (A) metolachlor (3) and (B) clomazone (4), extracted from their vapour phases for1 h at 60¡C with three di†erent Ðbre coating thicknesses (7, 30 and 100 lm).

306 T homas Kro� hl et al.

TABLE 4SPME/GC Results with Fenpropimorph (1) at 80¡C.

Evaporated mass SPME-extrac.M

evamass M

ext(mg) GC peak area (ng) M

eva/M

ext

0É88 342 641 1435É8 6130É88 346 847 1453É4 6050É86 353 800 1482É6 5800É82 349 140 1463É1 5600É91 382 494 1602É8 5680É92 355 571 1490É0 617

Mean value 0É88 355 082 1488É0 591Rel. standarddeviation (%) 4É10 4É01 4É01 4É14

amounts in nanograms (ng). Although the absorbedamounts are not very high, their standard deviationsare relatively small, indicating a good repeatability ofthe measurements. An increase of the furnace tem-

perature from 80 to 100¡C increased the amount ofabsorbed material by a factor of about three. The lastcolumns of Tables 4 and 5 show the ratio between theevaporated mass and the SPME-extracted mass. This

TABLE 5SPME/GC Results with Fenpropimorph (1) at 100¡C.

Evaporated mass SPME-extract.M

evamass M

ext(mg) GC peak area (ng) M

eva/M

ext

3É42 1 228 471 5147É8 6643É28 1 104 116 4626É7 7093É56 1 196 112 5012É2 7103É52 1 136 832 4763É8 7393É49 1 159 088 4857É1 7193É28 1 109 079 4647É5 706

Mean value 3É43 1 155 616 4842É5 708Rel. standarddeviation (%) 3É54 4É26 4É26 3É45

Fig. 9. (A) SPME/GC chromatograms obtained for kresoxim-methyl (2) with extraction times of 10È50 min and (B) the calculatedpeak areas ; furnace temperature during evaporation : 120¡C.

V apour pressure of crop protection AI by thermogravimetry and SPME 307

Fig. 10. Weight loss of the test sample kresoxim-methyl (2)during evaporation for 50 min at a furnace temperature of

120¡C.

ratio is very large, which indicates that most of theevaporated material is passing the SPME device at adistance at which it cannot be absorbed by the coating.Nevertheless, the amount of extracted analyte is stillsufficiently high to identify the evaporated substance bySPME analysis.

Kresoxim-methyl (2), a fungicide with a relatively lowvapour pressure (Table 1), had to be extracted at highertemperatures. Results for a furnace temperature of120¡C are shown in Fig. 9, where the absorbed amount

is plotted as a function of the extraction time. As withfenpropimorph (Fig. 6), the absorbed amount increasedlinearly with time, which indicates that the absorbingcapacity of the Ðbre coating was not exceeded (Fig. 9B).

The GC chromatograms in Fig. 9A show the peak ofthe prepared test substance only ; there are no furtherpeaks to indicate impurities, contaminants or decompo-sition products. The weight of the sample, measuredduring the evaporation process, decreased linearly withtime (Fig. 10). Thus, a linear decreasing weight-losscurve corresponds to the evaporation of a purecompound.

Experimental data obtained with metolachlor (3), clo-mazone (4) and (Z)-9-dodecenyl acetate (5) are shown inFig. 11. SPME extractions were carried out with a30-lm Ðbre at a furnace temperature of 80¡C and anextraction time of 1 h (compounds 3 and 4). The morevolatile (Z)-9-dodecenyl acetate (5) was extracted for30 min at 40¡C.

All these active ingredients displayed linear evapo-ration behaviour (Fig. 11A), and SPME/GC analysisdetected only the pure ingredients, indicated by sharpmain GC peaks (Fig. 11B).

Fig. 11. (A) Weight loss curves and (B) SPME/GC chromatograms for metolachlor (3), clomazone (4) and Z9-dodecenylacetate (5).The substances (3) and (4) were extracted for 1 h with a 30-lm Ðbre at a furnace temperature of 80¡C, and (5) for 30 min at 40¡C.

308 T homas Kro� hl et al.

3.2.2 T hermally unstable compoundsThe determination of the vapour pressures of low-volatility, thermally unstable compounds is renderedmore difficult because the measurements have to be per-formed at lower temperatures to avoid thermal decom-position of the substances. Low temperatures lead torelatively slow weight losses of the sample and thus themeasurements are, in some cases, near the resolutionlimit of the thermogravimetric system. For this type ofinvestigation, SPME analysis can help to determinewhether decomposition of the compound has occurred.

The insecticide dimethoate (6) was chosen as a rela-tively low-volatility ingredient which shows instabilityat elevated temperatures (Fig. 12).

The results of evaporation experiments carried out at80, 100, 120 and 140¡C, accompanied by SPME/GCanalysis (7-lm PDMS Ðbre) are presented in Fig. 13.The SPME/GC diagram for 80¡C (Fig. 13A, left) showsthe peak of only dimethoate at a retention time ofabove 16 min. There are no peaks from decompositionproducts. This corresponds to the linear shape of theweight loss curve (Fig. 13A, right). Also at 100¡C (Fig.

Fig. 12. Di†erential scanning calorimetry (DSC) diagram fordimethoate (99.8%), m\ 8É48 mg, heating rate 5 K min~1.

The onset for decomposition appears at about 110¡C.

13B) the GC chromatogram shows no decompositionproducts and the weight loss curve remained linear,which indicates that the studied compound had not yetbegun to decompose.

Increasing the temperature by another 20¡C to 120¡C(Fig. 13C), caused additional peaks to appear in the GC

Fig. 13. GC chromatograms (left) and weight loss curves (right) from evaporation experiments with dimethoate (6) at (A) 80¡C, (B)100¡C, (C) 120¡C and (D) 140¡C.

V apour pressure of crop protection AI by thermogravimetry and SPME 309

chromatogram, mainly at shorter retention timesbetween 5 and 10 min. Together with the downwardsinÑection of the weight loss curve, this is a clear evi-dence that the compound had started to decomposeinto smaller fragments of higher volatility.

Finally, at 140¡C (Fig. 13D), only a small peak ofintact dimethoate could be observed, while the fragmentpeaks dominate the chromatogram. The decreasingslope of the weight loss curve resulted from an increas-ing encrustation of the sample on the glass plate as aconsequence of the decomposition of the compound,and thus to a decreasing evaporation rate of thematerial with time.

Like Tables 4 and 5, Table 6 presents a summary ofevaporation/SPME/GC experiments for all compoundsinvestigated. The results demonstrate the following : 1.As with fenpropimorph (Tables 4 & 5), only a relativelysmall percentage of the evaporated matter is collectedby the SPME device at the outlet of the thermogravi-metric system 2. Because of the di†erences in the chemi-cal structure and in the polarity of the compoundsinvestigated, each has a di†erent affinity to the Ðbre-coating material, as indicated by the di†erent values forthe ratio of evaporated mass to SPME-extracted mass.

4 CONCLUSIONS

The thermogravimetric method to determine the vapourpressure of organic compounds has been combined withthe new technique of solid phase microextraction(SPME). This technique, initially developed as asolvent-free method to extract organic compounds fromaqueous samples, can also be used to extract both vola-

tile and semi-volatile substances from the vapour phase.SPME extractions from saturated vapour phases

were carried out in septum-closed ampoules with threedi†erent active ingredients. In experiments designed toestablish the dependence of the SPME/GC signal on theextraction time it could be shown that, as a conse-quence of the low evaporation rates, an equilibriumbetween the extracted amount and the sample in itsmatrix could not be reached within an appropriate time.The GC peak areas increased linearly with the extrac-tion time, which indicates that each molecule of theslowly evaporating substance was successively absorbedby the Ðbre coating. This would appear to indicate thatthis kind of “non-equilibriumÏ-state is a typical charac-teristic of extractions from vapour phases. As reportedin the literature, this is di†erent from SPME extractionsfrom aqueous matrices, where a steady state is reachedwithin a relatively short period of time (mostly\1 min).10

The thickness of the Ðbre coating is an experimentalparameter whose relevance di†ers, depending on thesolubility of the analyte in the coating material or itsaffinity to the latter. Generally, a widening of the GCpeak was observed with an increase in coating thickness,such that, in some cases (e.g. clomazone), only thinlycoated Ðbres could be used for acceptable GC results.

Installing a SPME device at the outlet of a thermog-ravimetric apparatus makes it possible to collect a smallproportion of the volatiles for subsequent GC analysis.It could be shown for several thermally stable com-pounds that, despite the relatively small percentage ofcollected material, an analytical identiÐcation of theevaporating compound by SPME/GC could be suc-cessfully achieved. In particular, the experiments have

TABLE 6Evaporation/SPME/GC-Results for Compounds 1È6

T emp. T imea Meva

b Mext

cActive ingredient (¡C) (h) (mg) (ng) M

eva/M

ext

Fenpropimorph (1) 80 1 0É88 1488 591d100 1 3É43 4843 708e

Kresoxim-methyl (2) 120 1 1É726 7775 222Metolachlor (3) 80 1 0É887 1243 714

100 1 3É515 4162 845Clomazone (4) 60 1 0É678 425 1595

80 1 2É96 2627 1127(Z)-9-Dodecenyl acetate (5) 30 0É5 0É089 430 206

40 0É5 0É21 3056 69Dimethoate (6) 80 1 0É33 106 3113

100 1 0É75 117 6410

a Time\ evaporation time in the thermogravimetric system \ SPME-extraction time.mass in the thermogravimetric system.b Meva \ evaporated

mass at the outlet of the thermogravimetric system.c Mext\ SPME-extractedd From Table 4.e From Table 5.

310 T homas Kro� hl et al.

demonstrated a clear correlation between the linearityof the weight-loss curve and the evaporation process ofa pure compound.

In the case of organic compounds that are unstableto heat, the SPME/GC method could be utilized toshow whether, and at what temperatures, a decomposi-tion of the compound into fragments of higher volatilityoccurred. For example, the insecticide dimethoateexhibited both a linear evaporation behaviour and asingle GC peak at temperatures up to approximately100¡C, whereas at temperatures of 120¡C and higher,fragments from decomposition processes dominated theGC chromatograms.

In conclusion, evaporation rate measurements can becarried out with a greater degree of certainty concerningthe identity of the evaporating substances, when com-bined with the SPME/GC technique. Using mass spec-tral detection in place of FID could provide even moreinformation regarding the chemical identity of thevaporized species. Experiments along these lines andalso those using a combination of SPME and HPLC,especially for weakly volatile or thermally labile com-pounds, are in progress.

ACKNOWLEDGEMENTS

We thank Mr Anton Ho� rner, Mr Gu� nter Walter, MrMarkus Keller, Mr Bernd Machauer, Mr Michael Sie-becker, Mr Ralf Kreucher and Mr Andreas Bog forexcellent technical assistance and for data evaluation.

REFERENCES

1. Gu� ckel, W., Synnatschke, G. & Ritttig, R., A method fordetermining the volatility of active ingredients used inplant protection. Pestic. Sci., 4 (1973) 137È47.

2. Gu� ckel, W., Synnatschke, G. & Ritttig, R., A method fordetermining the volatility of active ingredients used inplant protection II. Applications to formulated products.Pestic. Sci., 5 (1974) 393È400.

3. Gu� ckel, W., Ka� stel, R., Lewerenz, J. & Synnatschke, G., Amethod for determining the volatility of active ingredientsused in plant protection Part III : The temperature

relationship between vapour pressure and evaporationrate. Pestic. Sci., 13 (1982) 161È8.

4. Gu� ckel, W., Ka� stel, R., Kro� hl, T. & Parg, A., Methods fordetermining the vapour pressure of active ingredients usedin crop protection. Part IV: An improved thermogravi-metric determination based on evaporation rate. Pestic.Sci., 45 (1995) 27È31.

5. EPA Product Properties Test Guideline OPPTS 830.7950Vapor Pressure, EPA 712-C-95-043 (1995).

6. Netzsch Dampfdruckbestimmung VPA 434, TechnischeDaten (1997).

7. Berlardi, R. & Pawliszyn, J., The application of chemicallymodiÐed fused silica Ðbers in the extraction of organicsfrom water matrix samples and their rapid transfer to cap-illary columns. W ater Pollution Research J. of Canada, 24(1989) 179.

8. Arthur, C. & Pawliszyn, J., Solid phase microextractionwith thermal desorption using fused silica optic Ðbers.Anal. Chem., 62 (1990) 2145È8.

9. Arthur, C., Potter, D., Buchholz, K., Motlagh, S. & Paw-liszyn, J., Solid phase microextraction for the directanalysis of water : Theory and practice. L C-GC, 10 (1992)656È61.

10. Louch, D., Mortlaugh, S. & Pawliszyn, J., Dynamics oforganic compounds extraction from water using liquid-coated fused silica Ðbers. Anal. Chem., 64 (1992) 1187È99.

11. Zhang, Z., Yang, M. & Pawliszyn, J., Solid phase micro-extraction. Anal. Chem., 66 (1994) 844È53.

12. Chai, M., Arthur, C., Pawliszyn, J., Berlardi, R. & Pratt,K., Determination of volatile chlorinated hydrocarbons inair and water with solid-phase microextraction. Analyst,118 (1993) 1501È5.

13. Boyd-Boland, A. & Pawliszyn, J., Solid phase micro-extraction of nitrogen-containing herbicides. J. Chromato-gr., 704 (1994) 163È72.

14. Eisert, R. & Levsen, K., Determination of pesticides inaqueous environmental samples via SPME (solid-phasemicroextraction). GIT Fachz. L ab., 39(1) (1995) 25È6.

15. Popp, P., Kalibitz, K. & Opperman, G., Application ofsolid-phase microextraction and gas chromatography withelectron-capture and mass spectrometric detection for thedetermination of hexachlorocyclohexanes in soil solutions.J. Chromatogr., 687 (1994) 133È40.

16. Bengsson, S. & Berglof, T., Extraction of Pesticides bySolid Phase Microextraction (from Soil). Dept. Of Environ-mental Assessment, Swedish University of AgriculturalSciences, PO Box 7050, Uppsala, Sweden.

17. Malosse, C., Ramirez-Lucas, P., Rochat, D. & Morin,J.-P., Solid-phase microextraction, an alternative methodfor the study of airborne Insect pheromones (Metamasiushemipterus, Coleoptera, Curculionidae). J. High Resol.Chromatogr., 18 (1995) 669È70.