Anal. Chem. 1998, 70, 2426-2431
Focused Microwave-Assisted Soxhlet: An Advantageous Tool for Sample Extraction L. E. Garcı´a-Ayuso,† M. Sa´nchez,‡ A. Ferna´ndez de Alba,§ and M. D. Luque de Castro*,†
Department of Analytical Chemistry, Faculty of Sciences, University of Co´ rdoba, E-14004 Co´ rdoba, Spain, Agrarian Laboratory of Co´ rdoba, Junta de Andalucı´a, E-14080 Co´ rdoba, Spain, and Department of Analytical Chemistry, Faculty of Experimental Sciences, University of Almerı´a, E-04120 Almerı´a, Spain
A new type of microwave-assisted Soxhlet is here reported. The device uses conventional Soxhlet glassware for solid sample extraction and a focused-microwave digester for irradiation of the sample cartridge at the required intervals while the fresh solvent (condensed vapors from the distillation flask) drips on and passes through the solid sample. In this way breaking of the analyte-matrix bonds is facilitated by application of the appropriate energy. The new approach has been checked in a comparative study by its application to the extraction of analyte families of different polarity (namely, alkanes, polyaromatic hydrocarbons, and herbicides) from the same soil matrix using dichloromethane as extractant. The reduction of the extraction times (from 8 h to 50-60 min, depending on the polarity of the analytes) with efficiency similar to or even higher than that afforded by the conventional Soxhlet technique supports the suitability of the new approach. In addition, recycling of the solvent during extract preconcentration enables minimal environmental contamination to be achieved. Soxhlet extraction is one of the oldest and most widespread ways for routine extraction of solid samples. Despite the fact that the process is commonly designated as extraction, strictly speaking Soxhlet is a leaching, lixiviation process as it involves solidliquid contact for removal of one or several species from a solid by being dissolved into the liquid phase. In a conventional Soxhlet device the initial solid phase (or a solid-liquid mixture) is placed in a cavity that is gradually filled with extractant liquid phase by condensation of vapors from a distillation flask. When the liquid reaches a preset level, a siphon aspirates the content of the cavity and unloads it back into the distillation flask, by carrying the extracted analytes in the bulk liquid. This operation is repeated until virtually complete separation is achieved and the analytes are all in the flask. Inasmuch as the solvent acts stepwise, the assembly can be considered a batch system; however, since the solvent is recirculated through the sample, the system also bears a continuous character. Hence this is regarded as a mixed operation mode. The most salient advantages of conventional Soxhlet are as follows: (a) the sample phase is repeatedly brought into contact with fresh portions of the solvent, thereby enhancing †
University of Co´rdoba. Junta de Andalucı´a. § University of Almerı´a. ‡
2426 Analytical Chemistry, Vol. 70, No. 11, June 1, 1998
the displacement of the analyte from the matrix; (b) the temperature of the system is higher than the room temperature since the heat applied to the distillation flask reaches the extraction cavity to some extent; (c) no filtration is required. Technical drawbacks of this technique are the inability to provide agitation, which would be of help to accelerate the process, and the long time required for extraction. The latter aspect has promoted both the development of alternative methods for leaching, such as supercritical fluid extraction,1 and the use of energy sources, which accelerate the leaching process. Such is the case of the use of ultrasound, laser, microwaves, etc.,2-4 which dramatically shorten the leaching time. The new alternatives are usually compared with conventional Soxhlet in order to check their performance.5-11 Some of these comparisons provide very surprising results: recoveries higher than 100% have been obtained (e.g., in supercritical fluid extractions12-14) when values from conventional Soxhlet have been considered as 100% recovery. This fact can be explained in terms of matrix-analyte interactions: the fraction of analyte strongly bound to the matrix is not removed even when the process is developed for long periods due to the fact that the energy necessary to break the bond cannot be supplied by the extracting solvent. On the contrary, a supercritical fluid can supply the necessary energy, thus providing a better yield of the step. A pleiad of modifications of conventional Soxhlet have been aimed at improving its performance.15-22 (1) Luque de Castro, M. D.; Valca´rcel, M.; Tena, M. T. Analytical Supercritical Fluid Extraction; Springer-Verlag: Heidelberg, 1994. (2) Luque de Castro, M. D.; da Silva, M. P. Trends Anal. Chem. 1997, 16 (1), 16. (3) Pare´, J. R. J.; Be´langer, J. M. R. Trends Anal. Chem. 1994, 13 (4), 176. (4) Ganzler, K.; Salgo, A.; Valko, K. J. Chromatogr. 1986, 371, 299. (5) Lo´pez-AÄ vila, V.; Young, R.; Beckert, W. F. Anal. Chem. 1994, 66, 1097. (6) Lo´pez-AÄ vila, V.; Young, R.; Teplitsky, N. J. AOAC Int. 1996, 79 (1), 142. (7) David, M. D.; Seiber, J. N. Anal. Chem. 1996, 68, 3038. (8) Bowadt, S.; et al. Anal. Chem. 1995, 67, 2424. (9) Fisher, J.; Scarlett, M. J.; Stott, A. D. Environ. Sci. Technol. 1997, 31, 1120. (10) Dean, J. R.; Barnabas, I. J.; Fowlis, I. A. Anal. Commun. 1995, 32, 305. (11) Barnabas, I. J.; Dean, J. R.; Tomlinson, W. R.; Owen, S. P. Anal. Chem. 1995, 67, 2064. (12) Reighard, T. S.; Olesik, S. V. Anal. Chem. 1996, 68, 3612. (13) Tong, P.; Imagawa, T. Anal. Chim. Acta 1995, 310, 93. (14) Wenclawiack, B. W.; Pasche, T.; Krappe, M. Fresenius’ J. Anal. Chem. 1997, 357, 1128. (15) Prosky, L.; O’Dell, R. J.sAssoc. Off. Anal. Chem. 1973, 56 (1), 226. (16) Cuddeback, J. E.; Burg, W. R. Rev. Sci. Instrum. 1975, 46 (6), 680. (17) Buchanan, W. L.; Eisenbraun, E. J. Chem. Ind. 1977, 1, 35. (18) Pinto, A. F. J. Am. Oil Chem. Soc. 1967, 44 (2), 160. (19) Matusiewicz, H. Anal. Chem. 1982, 54 (11), 1909. S0003-2700(97)01104-9 CCC: $15.00
© 1998 American Chemical Society Published on Web 05/01/1998
Taking into account the advantages and disadvantages of conventional Soxhlet, a new design is here proposed which is aimed at keeping the advantages but circumventing the disadvantages of the former by reducing the leaching time and leading analyte removal to completeness. For this purpose, the conventional Soxhlet extractor has been used for taking advantage of points a-c mentioned in the previous paragraph, and the sample compartment which contains the sample cartridge has been located at the irradiation zone of a focused microwave digester. In this way, the sample is irradiated with the required microwave power at the most appropriate moments and during the required intervals, while the passage of fresh solvent and concentration of the removed analytes in the distillation flask occur in the conventional Soxhlet way. The performance of the new device has been checked by adopting the following criteria: (1) application of the new extraction process to three analyte groups of different polarity, namely, alkanes, PAHs, and herbicides (namely, triazines, organochlorines, and nitroaromatics), used either as pre-emergence or selective pesticides; (2) addition of the analytes to the same matrix in all instances, namely, a clayey soil duly powdered and sieved; (3) usage of the same leaching agent, namely, dichloromethane; (4) comparison with conventional Soxhlet, whose results are considered as 100% recovery (obviously, the leaching agent was also dichloromethane in this case). EXPERIMENTAL SECTION Instrumentation. The instruments and apparatus used for developing the different steps were as follows. Soil Preparation. A rotary shaker, a forced-air oven, and a rotary evaporator (from Selecta, Barcelona, Spain) were used for preparation of the spiked soil. Extraction. A conventional Soxhlet extractor was properly modified in order to make feasible the housing of the sample cartridge compartment in the irradiation zone of a Prolabo Microdigest 301 (200 W of maximum power) device. The latter was also modified: an orifice at the bottom of the irradiation zone enabled connection of the cartridge compartment with the distillation flask through the siphon. The first version of the prototype used for development of the extraction step was provided with two peristaltic pumps equipped with flexible tubes, for solvent aspiration and siphoning in addition to the Soxhlet approach, which was made of glass. The aim of these pumps was to achieve a more strict control of both the contact time between the sample and the fresh solvent (by aspirating the latter at preset intervals) and the introduction of fresh solvent into the cartridge at a preset flow rate. A scheme of the performance of the overall device is given in Figure 1. A Prolabo Megal 500 special thermometer from Prolabo, Briare-le-Canal, France, was used for monitoring the extraction temperature. Three controllers were used for the microwave unit, pumps, and thermometer, and a Prolabo electrical isomantle (with resistance control) was used as a heating source of the distillation flask. Determination. Alkanes: An 8410 Perkin-Elmer gas chromatograph (Norwalk, CT) equipped with a flame ionization detector (20) Giusiani, M.; et al. Arzneim.-Forsch. 1983, 33 (II), 1422. (21) Ndiomu, D. P.; Simpson, C. F. Anal. Chim. Acta 1988, 213, 237. (22) Hengstgstmann, R.; Hamann, R.; Weber, H.; Kettrup, A. Fresenius’ Z. Anal. Chem. 1989, 335 (8), 982.
Figure 1. Scheme of the prototype.
(FID) was used. PAHs: A GC-ITMS Saturn 3 system (Varian, Harbor City, CA), consisting of a 3400 Varian gas chromatograph and a 1093 septum-programmable injector (SPI), was used. Herbicides: A 3400 CX Varian Star gas chromatograph equipped with a 8200 CX autosampler and an electron capture detector (ECD) was used. Standards, Solvents, and Cartridges. All PAHs and alkanes used as analytes were supplied by Sigma Chemical (St. Quentin, Fallavier, France). Serial dilutions were carried out in benzenedichloromethane (DCM) and benzene, respectively, to obtain the required standard concentrations. The Agrarian Laboratory of Co´rdoba kindly donated herbicides. Cellulose disposable cartridges (22 × 88 mm) from Albet (Barcelona, Spain) were also used. Dichloromethane used for extraction was pesticide grade and generously supplied by Prolabo. Others solvents used (benzene, acetone, cyclohexane, ethyl ether, and ethyl acetate) were chromatographic grade and supplied by Merck (Darmstadt, Germany); anhydrous sodium sulfate was from Panreac (Barcelona, Spain). Preparation of Spiked Soil. Soil was sieved to a size smaller than 2 mm. Samples spiked with either alkanes or PAHs were prepared by rotary evaporation of 500 g of soil to which 200 mL of ethyl ether containing the necessary concentrations of the respective analyte mixtures were added. Solvent evaporation was carried out 7 days after the solid was spiked. The clayey soil contaminated with herbicides was provided by the Agrarian Laboratory of Co´rdoba and was prepared as follows: 500 g of clayey soil was spiked with 50 mL of 25:75 acetone-water containing the herbicides; the spiked soil was then shaken for 48 h in a rotatory shaker. The soil was stored at room temperature during 7 days, allowing the adsorption of the herbicides by the clayey matrix. Finally, the soil was dried in a forced-air oven during 48 h and at 35 °C. After that, the three spiked soils were stored at -20 °C until required. Conventional Soxhlet Extraction Procedure. Soxhlet extractions were performed using 3.5 g portions of soil to which 5 g of anhydrous sodium sulfate was added. The mixture was transferred to a prewashed cellulose cartridge and inserted into the 50 mL Soxhlet thimble. The apparatus was fitted with a 100 mL flask containing 70 mL of DCM and a boiling regulator. The assembly was heated and refluxed for 8 h (8-9 cycles/h) using an electrical isomantle. Analytical Chemistry, Vol. 70, No. 11, June 1, 1998
2427
Table 1. Study of Variables
variable irradiatn power, % water content, mL irradiatn time, s (each cycle) no. of cycles
optimum value range studied alkanes PAHs herbicides 10-99 0-7 10-90 4-12
50 1.5 15 8
50 1 15 10
50 1 15 10
Focused Microwave-Assisted Soxhlet Extraction (FMASE) Procedure. Into the distillation flask and open microwave vessel respectively were poured 120 and 30 mL of DCM. Clayey soil (7 g) was weighed in a cellulose cartridge. When wet samples were used in order to favor microwave absorption by the sample, a preset volume of ultrapure water (1.0 or 1.5 mL depending on the type of analytes, PAHs and herbicides or alkanes, respectively) was added to the cartridge and the sample was allowed to soak for 10-15 min. Finally, the thimble was covered with a loose wad of glass wool, inserted into the extraction vessel, which was then placed in the microwave-irradiation zone. The extraction program consisted of a number of cycles depending on the target analytes. Each cycle involved three steps: (1) filling of the extraction vessel, that is, delivery of 50 mL of fresh solvent by pump 1; (2) irradiation for 15 s (50% of the maximum microwave power); (3) unloading of the extraction vessel, that is, delivery of 50 mL of solvent analyte from the extraction vessel to the distillation flask by pump 2 (pump speed 15 mL/min). The duration of a cycle was 6 min. Individual Separation and Determination. Extract Conditioning. After FMASE for 50 or 60 min depending on the analytes (alkanes or PAHs and herbicides, respectively), the extracts were allowed to cool for 25-30 min, dried over anhydrous sodium sulfate, and concentrated to dryness in a rotary evaporator (35 °C). The extracts were recomposed to 2 mL with the appropriate solvent (1:1 cyclohexanes-ethyl acetate for PAHs and benzene for herbicides and alkanes) and placed in 2 mL glass vials. No cleanup step was necessary. Chromatographic-Detection Conditions. Alkanes: Aliquots (1 µL) of the extracts were injected manually on a 2 m length × 1/8 in. i.d. × 0.5 µm film thickness Chromosorb W-HP 80/100 chromatographic column (Hewlett-Packard, Palo Alto, CA) with nitrogen as carrier at a flow rate of 25 mL/min. The injector and detector temperatures were 220 and 275 °C, respectively. The GC oven program was as follows: initial temperature 75 °C, retained for 3 min, increased at a rate of 10 °C/min to 200 °C. The segment acquire time was 20 min. Data acquisition processing and delivery was performed by an LC-100 Perkin-Elmer integrator. PAHs: Aliquots (2 µL) of the extracts were injected in the splitless mode on a 30 m length × 025 mm i.d. × 0.25 µm film thickness DB-5MS capillary column (J & W Scientific, Folsom, CA) with He as carrier gas at a flow rate of 1 mL/min. The injector temperature was held at 120 °C for 0.1 min and then increased at 200 °C/min to a final temperature of 280 °C, where it was held for 20 min. Full-scan spectra were run in the electron impact (EI) mode from m/z 65 to 550, and the segment acquire time was 35 min. The detector temperature was 230 °C, electron energy 30 eV, filament current 35 mA, electron multiplier tube 1500 V, and interface temperature 280 °C. The GC oven program 2428 Analytical Chemistry, Vol. 70, No. 11, June 1, 1998
Figure 2. Effect of the key variables on the recovery of alkanes by FMASE (for details, see text): 2, dodecane; b, tridecane; 9, tetradecane.
was as follows: initial temperature 80 °C, retained for 1.20 min, increased at a rate of 25 °C/min to 180 °C, where it was held for 10 min. Data acquisition, processing, and instrument control were performed by Saturn GC-ITMS workstation software (version 5.2) loaded into a 486 DX, 66 MHz computer. Herbicides: Aliquots (1 µL) of the extracts were injected in the split mode (split ratio 1/ ) on a 30 m length × 0.25 mm i.d. × 0.25 µm film thickness 65 SPB-5 fused silica capillary column (Supelco, Bellefonte, PA) with nitrogen as carrier gas at a flow rate of 1.3 mL/min. The injector and detector temperatures were 240 and 350 °C, respectively. The GC oven program was as follows: initial temperature 80 °C, retained for 1 min, increased at a rate of 30 °C/min to 190 °C, retained for 15 min, increased at a rate of 5 °C/min to 240 °C. Data acquisition, processing, and instrument control were performed by Varian Star GC workstation software (version 4.0) loaded into a 486 DX, 66 MHz computer. RESULTS AND DISCUSSION As the aim of the present research was the comparison between the new and the conventional Soxhlet, the Soxhlet procedure for extraction of PAHs from soils23 was also used for the two other families of analytes (namely, alkanes and herbicides). In this way the behavior of analytes of different polarity is studied under the same extraction conditions. For this reason the recoveries obtained from PAHs and herbicides in conventional Soxhlet are in agreement with the spiked amounts, but not for alkanes, as these analytes require an apolar solvent for optimal extraction. This is not a drawback as all the recoveries are expressed with respect to conventional Soxhlet, which is accepted as providing the highest recoveries (100%) under the working conditions, which are the same in both approaches. In order to quantify the efficiency of the extraction step through the study of the variables affecting it, calibration of the individual separation/determination step for each group of target analytes was previously performed. Optimization Study of the FMASE. The univariate method was used in all instances for optimization of the four parameters, which affect the extraction step: irradiation power and time, (23) USEPA Method 3540, U.S. Government Printing Office, Washington, DC, 1986.
Table 2. Results of the Focused Microwave Extraction from Soil (Using DCM) compound
recoverya (%)
RSDb (%)
RSDSoxhletb (%)
n-dodecane n-tridecane n-tetradecane av benz[a]anthracene benz[e]acephenanthrylene benzo[k]fluoranthene pyrene benzo[a]pyrene benzo[e]pyrene av 2,6-dichloro-benzonitrile (dichlorobenil) 2,6-dinitro-N,N-dipropyl-4-(trifluoromethyl)benzenamine (trifluraline) N3,N3-diethyl-2,4-dinitro-6-(trifluoromethyl)-1,3-benzenediamine (dinitramine) 2-chloro-N-(2,6-diethylphenyl)-N-(methoxymethyl)acetamide (alachlor) 4-amino-6-(1,1-dimethylethyl)-3-(methylthio)-1,2,4-triazin-5(4H)-one (metribuzine) N-(1,1-dimethylethyl)-N′-ethyl-6-(methylthio)-1,3,5-triazine-2,4-diamine (terbutrine) 6-chloro-N,N′-diethyl-1,3,5-triazine-2,4-diamine (simazine) 2,4-dichloro-1-(4-nitrophenoxy)benzene (nitrofen) av
101 96.5 92.7 96.8 99.3 97.4 104 94.0 108 107 102 118 101 104 108 136 91.8 88.5 132 110
6.0 3.2 14 7.7 9.4 11 14 9.3 16 16 13 7.3 10 6.1 13 16 11 5.7 16 11
4.7 8.0 9.9 7.5 4.8 4.1 2.7 7.9 5.6 9.1 5.7 16 7.3 9.2 5.4 9.3 8.6 10 8.1 9.2
a
Results are expressed with respect to 100% recovery for conventional Soxhlet extraction. b RSD: relative standard deviation (n ) 7).
Table 3. Statistical Test compound alkane n-dodecane n-tridecane n-tetradecane PAH benz[a]anthracene benz[e]acephenanthrylene benzo[k]fluoranthene pyrene benzo[a]pyrene benzo[e]pyrene herbicide dichlorobenil trifluraline dinitramine alachlor metribuzine terbutrine simazine nitrofen
S2FMASE
S2Soxhlet
Fcala
Ftab
Sweighb
tcalc
ttabd
40.0 9.30 166
22.3 63.7 98.6
1.7 6.8 1.7
4.3 4.3 4.3
5.6
0.4 1.1 1.2
2.2 2.3 2.2
88.0 115 223 76.7 283 167
23.3 16.8 7.13 62.9 31.2 32.6
3.8 6.8 31 1.2 9.1 5.1
4.3 4.3 4.3 4.3 4.3 4.3
7.5
0.2 0.6 0.7 1.3 1.2 1.3
2.2 2.3 2.5 2.2 2.3 2.3
3.4 1.9 2.1 6.9 5.3 1.4 4.1 5.1
4.3 4.3 4.3 4.3 4.3 4.3 4.3 4.3
13 8.9 7.9
2.7 0.3 1.0 1.4 4.0 1.6 2.7 3.8
2.2 2.2 2.2 2.3 2.3 2.2 2.2 2.3
74.1 103 40.4 202 465 101 25.3 432
249 53.6 84.6 29.2 87.1 74.1 103 85.4
11
8.4
9.4 8.0
a F b 2 2 c cal calculated from the Snedecor test. Ftab ) (R ) 0.05, 6 df, 6 df). Sweighed is applied for the Student test only when S FMASE ) S Soxhlet. tcal calculated from the Student test. d In order to calculate ttab, it is necessary to know the number of degrees of freedom (df). If the variances are not significantly different, df ) n1 + n2 - 2 ) 12; otherwise, df is calculated by means of the Welch test or the Cochran test.
number of cycles, and sample water content. Both the range in which the variables were studied and the optimal values obtained are summarized in Table 1. Irradiation Power (IP). This variable was changed between 10 and 99% (10, 30, 50, 70, 99%) of the nominal value provided by the digester with a constant irradiation time of 15 s. In all three groups of analytes the behavior was similar to that shown in Figure 2A for the alkane family. A plateau of maximum recovery was obtained from an irradiation power of 50%, which did not increase for higher irradiation times. An IP of 50% was selected for subsequent experiments. Irradiation Time (IT). When the time during which the sample was subjected to microwave irradiation was changed within 10-90 s (10, 15, 45, 60, 90 s), the three families of target analytes
behaved in a similar way (see in Figure 2B the effect of this variable on alkanes). The optimum value of this variable was 15 s, decreasing the recovery for longer and shorter ITs. This behavior could be explained as follows: (a) For this IT and 50% irradiation power the microwave energy absorbed by the cartridge is insufficient for reaching the temperature of the phase change, so it is consumed to break the analyte-matrix bonds. With higher IT the boiling point of the solvent is reached and the boiling conditions decrease the sample-solvent contact. (b) The cycle time (i.e., the interval elapsed between two consecutive aspirations of pump 2) is constant. This interval is used in the cartridge for irradiation plus subsequent sample-hot solvent contact time: as the former increases, the second decreases, and the recovery decreases as a result. Analytical Chemistry, Vol. 70, No. 11, June 1, 1998
2429
Figure 3. Chromatograms from alkane extracts (GC-FID separation and detection) under optimal conditions: A, standards; B, conventional Soxhlet; C, FMASE; (1) dodecane, (2) tridecane, (3) tetradecane.
Figure 5. Chromatograms from PAH extracts (GC/MS separation and detection) under optimal conditions: A, standards; B, conventional Soxhlet; C, FMASE; (1) pyrene, (2) benzo[a]anthracene, (3) benzo[e]acephenanthrylene, (4) benzo[k]fluoranthene, (5) benzo[e]pyrene, (6) benzo[a]pyrene.
Figure 4. Chromatograms from herbicide extracts (GC-ECD separation and detection) under optimal conditions: A, standards; B, conventional Soxhlet; C, FMASE; (1) dichlorobenil, (2) trifluraline, (3) simazine, (4) dinitramine, (5) terbutrine, (6) metribuzine, (7) alachlor, (8) nitrofen.
Number of Cycles (NC). This variable was varied from 4 to 12 cycles; 100% recovery for alkanes was achieved within 8 cycles (see Figure 2C), while PAHs and herbicides required 10 cycles for total recovery. This behavior was foreseeable due to the following: (i) the boiling points of alkanes are lower than those of PAHs and herbicides, so their removal from the matrix is easier, and (ii) the low polarity of alkanes gives rise to a weaker interaction with the soil. Water Content (WC). The percentage of water in the soil was studied between 0 and 50% (w/w) (this means addition of 0-7 mL of water to each sample). The maximum recovery of alkanes was obtained in samples that contained 1.5 mL of water/sample (ca. 17.7%), as shown in Figure 2D. Herbicides and PAHs provided the maximum recovery in samples containing 1.0 mL of water (ca. 12.5%). This behavior was foreseeable as a consequence of the polarity of the target group: the polarity of alkanes is the lowest, so alkanes require a higher percentage of water in order to make more polar the cartridge content, which in this way 2430 Analytical Chemistry, Vol. 70, No. 11, June 1, 1998
absorbs more microwave energy. Nevertheless, when the water content surpassed 17.7% (w/w), the extraction efficiency decreased in all instances, probably due to a worse sample-solvent contact. Reproducibility of the Extraction Step. A reproducibility study was carried out on seven samples that were extracted on different days, with both the proposed and the conventional Soxhlet approaches. As can be seen in Table 2, the relative standard deviation (RSD) obtained is similar. The slightly better precision afforded by conventional Soxhlet is a consequence of the smaller number of variables affecting the extraction. Critical parameters in FMASE are sample humidity and initial temperature as a result of the short extraction time. On the other hand, the long time required for the conventional Soxhlet smoothes the effect of small changes of the variables. A statistical comparative study between the conventional Soxhlet and the proposed approach (Table 3) shows that there are no significant differences between the results they provide. The variances of the two methods appear in columns 2 and 3. After calculation of the Fisher F (column 4), the Snedecor test shows significant (Ftab < Fcal) and nonsignificant (Ftab > Fcal) differences for a 95% confidence level. In the latter case, weighed variances were obtained (column 6), and then the Student test was calculated (columns 7 and 8). FOCUSED MICROWAVE-ASSISTED SOXHLET VERSUS CONVENTIONAL SOXHLET In addition to the chromatograms of the standards for the three groups of target analytes (A in Figures 3-5) the chromatograms
obtained with extracts from both the proposed approach (C in Figures 3-5) and the conventional Soxhlet (B in the same figures) are shown comparatively in these figures. As can be seen, the extracts from the new device are usually as clean as those obtained by its conventional counterpart (even better baseline in PAHs), the peaks are higher (better recoveries in almost all instances), and the extraction time is 50-60 min versus the 8 h required by the conventional approach. CONCLUSIONS From the results obtained in the present study the following aspects are worth emphasizing: (i) The proposed approach provides efficiencies similar to those obtained by conventional Soxhlet with extraction times at least 8 times shorter (50-60 min vs 8 h). The quantitativeness of the extraction process is a consequence of the continuous samplefresh solvent contact on which the new approach is based, similar to conventional Soxhlet. This is one of the most remarkable differences between the FMASE here reported and commercial devices based on the use of either focused or multimode microwaves. (ii) Recoveries higher than those provided by conventional Soxhlet could be due to the higher temperature reached in the cartridge, which makes possible the release of strongly adsorbed or bound fractions of the target analytes, difficult to remove at the temperatures reached in the conventional device.
(iii) The extracts obtained by FMASE are similar to those provided by its conventional counterpart as shown in the chromatographs obtained from both. (iv) The new approach can be used for almost all applications of conventional Soxhlet under the same working conditions. Improvements of the experimental setup achieved during the development of the present research have been as follows: (a) The peristaltic pump system is removed, as a better design of the siphon and connections enables a strict control of both the cycle time and the level of solvent in the cartridge at the aspiration time. The overall approach is thus noticeably simplified. (b) An additional valve system allows solvent recycling during the evaporation step for the extract preconcentration. About 75% of the solvent is recycled, thus endowing FMASE with the label of clean method hardly applicable to conventional Soxhlet. ACKNOWLEDGMENT Prolabo (France) is thanked for partially covering the personnel expenses, apparatus, and reagents for development of this research. Comisio´n Interministerial de Ciencia y Tecnologı´a (CICyT) is thanked for financial support (Project PB96-0505).
Received for review October 7, 1997. Accepted March 19, 1998. AC9711044
Analytical Chemistry, Vol. 70, No. 11, June 1, 1998
2431
