Molecularly Imprinted Polymeric Fibers for Solid-Phase

Mar 20, 2007 - Subsequently, analytes can be thermally desorbed directly onto the injection port of a gas chromatograph, eluted with the mobile phase ...
7 downloads 19 Views 253KB Size
Download PDF
Anal. Chem. 2007, 79, 3099-3104

Molecularly Imprinted Polymeric Fibers for Solid-Phase Microextraction E. Turiel, J. L. Tadeo, and A. Martin-Esteban*

Departamento de Medio Ambiente, INIA, Carretera de A Corun˜a km 7.5, E-28040 Madrid, Spain

Solid-phase microextraction (SPME) is widely used in analytical laboratories for the analysis of organic compounds, thanks to its simplicity and versatility. However, the current commercially available fibers are based on nonselective sorbents, making difficult in some cases the final determination of target compounds by chromatographic techniques. Molecularly imprinted polymers (MIPs) are stable polymers with selective molecular recognition abilities, provided by the template used during their synthesis. In the present work, a simple polymerization strategy allowing the obtainment of molecularly imprinted polymeric fibers to be used in SPME is proposed. Such a strategy is based on the direct synthesis of molecularly imprinted polymeric fibers (monoliths) using silica capillaries as molds, with silica being etched away after polymerization. The system propazine:methacrylic acid was used as a model for the preparation of molecularly imprinted fibers, and its ability to selectively rebind triazines was evaluated. Variables affecting polymer morphology (i.e., polymerization time, fiber thickness) and binding-elution of target analytes (i.e., solvents, time, temperature) were studied in detail. The imprinted fiber showing the best performance in terms of selectivity and affinity for triazines was successfully applied to the extraction of target analytes from environmental and food samples. Since solid-phase microextraction (SPME) was introduced by Pawliszyn1 in the early 1990s, its use for sample preparation has been rapidly implemented in analytical laboratories. SPME is based on the partitioning of target analytes between the sample and a stationary phase, which is typically coated to the surface of a fused silica fiber. Subsequently, analytes can be thermally desorbed directly onto the injection port of a gas chromatograph, eluted with the mobile phase on an especially designed SPMEHPLC interface, or eluted in a small volume of a suitable solvent to be further analyzed by chromatographic techniques.2,3 Its simplicity of operation, its solventless nature, and the availability of commercial fibers have made SPME become a tool routinely used for certain applications. * To whom correspondence should be addressed: Phone: 34-91 3478700. Fax: 34-91 3572293. E-mail: [email protected]. (1) Arthur, C. L.; Pawliszyn, J. Anal. Chem. 1990, 62, 2145. (2) Lord, H. L.; Pawliszyn, J. LC-GC Int. 1998, 16, 41. (3) Pawliszyn, J. Solid-Phase Microextraction; Wiley: New York, 1998. 10.1021/ac062387f CCC: $37.00 Published on Web 03/20/2007

© 2007 American Chemical Society

However, the variety of commercially available fibers is rather limited and only covers the scale of polarity. Consequently, the extraction process lacks selectivity, preventing the final determination of target analytes at trace levels in complex samples using chromatographic techniques coupled to common detectors. Thus, recent years have seen an increasing interest in the preparation of tailor-made fibers. Fiber coating procedures, which have been recently reviewed,4 include sol-gel technology, electrochemical methods, and physical deposition. These procedures provide a wide range of homemade coatings which can sort out some of the drawbacks associated with the commercial fibers. However, except for a few certain cases, the majority of the homemade fibers have not been applied to the determination of target analytes in real samples, which makes it difficult to confirm the desired improvement of selectivity. In recent years, molecularly imprinted polymers (MIPs) have proven to be useful materials in several areas of analytical chemistry.5,6 MIPs are cross-linked synthetic polymers obtained by copolymerizing a monomer with a cross-linker in the presence of a template molecule. After polymerization, the template is removed from the porous network by washing, leaving cavities in the polymeric matrix that are complementary in size, shape, and chemical functionality to the template. Thus, the imprinted polymer is able to rebind selectively the analyte (the template) under certain experimental conditions. The inherent selectivity associated with MIPs has made them optimum sorbents to be used in solid-phase extraction,7-9 so-called molecularly imprinted solid-phase extraction (MISPE). Accordingly, the combination of molecular imprinting and SPME would ideally provide a powerful analytical tool with the characteristics of both technologies, simplicity, flexibility, and selectivity. The easiest way for combining both technologies was proposed by Mullet et al. in 2001, which consisted in packing a capillary with the MIP particles for in-tube SPME and was used for the selective determination of propanolol in serum samples.10 The developed method was successfully applied and the advantages of in-tube SPME were obvious (high enrichment factors provided by multiple draw/eject cycles, ease of automation, and fast operation). However, this methodology is not free of some (4) Dietz, C.; Sanz, J.; Ca´mara, C. J. Chromatogr., A 2006, 1103, 183. (5) Mayes, A. G.; Mosbach, K. Trends Anal. Chem. 1997, 16, 321. (6) Haupt, K. Anal. Chem. 2003, 75, 376A. (7) Martin-Esteban, A Fresenius’ J. Anal. Chem. 2001, 370, 795. (8) Lanza, F.; Sellergren, B. Chromatographia 2001, 53, 599. (9) Caro, E.; Marce, R. M.; Borrull, F.; Cormack, P. A. G.; Sherrington, D. C. Trends Anal. Chem. 2006, 25, 143. (10) Mullet, W. M.; Martin, P.; Pawliszyn, J. Anal. Chem. 2001, 73, 2383.

Analytical Chemistry, Vol. 79, No. 8, April 15, 2007 3099

important drawbacks such as the lack of compatibility between the solvent needed to desorb analytes from the MIP and the mobile phase used (typical drawback of online MISPE protocols) and the necessity of extra instrumentation (pump, multiport valves). Thus, the preparation of silica fibers coated with a MIP to perform SPME would be the best option. To the best of our knowledge, only one paper has been published dealing with the preparation and use of a MIP-coated silica fiber.11 In this work, clenbuterol-imprinted fibers were prepared and used in the selective extraction of brombuterol from human urine. The preparation of imprinted fibers was performed by silylation of silica fibers which were subsequently immersed in the polymerization solution composed of clenbuterol, methacrylic acid, ethylene glycol dimethacrylate, and azo(bis)isobutyronitrile dissolved in acetonitrile. Then, polymerization was performed during 12 h at 4 °C under irradiation at 350 nm. According to the authors, fibers with a polymeric film thickness of ∼75 µm were obtained in a reproducible manner. However, in spite of the success of the proposed approach, further papers dealing with the same topic have not been published. In the present paper, a completely different and much more simple and easy approach for the preparation of imprinted fibers is proposed. It is based on the direct synthesis of molecularly imprinted polymeric fibers (monoliths) using silica capillaries as molds, with silica being etched away after polymerization.12 The systempropazine:methacrylicacid:ethyleneglycoldimethacrylate13-15 has been used as a model for the preparation of the imprinted fibers. The optimization of variables affecting both polymer morphology (i.e., polymerization time, fiber thickness) and binding-elution conditions of target analytes (i.e., solvents, time, temperature) will be described in detail. Also, the performance of the new imprinted fibers for the SPME of triazines from environmental and food samples will be assessed. EXPERIMENTAL SECTION Reagents. Desethylatrazine (DEA), desisopropylatrazine (DIA), simazine (SIM), cyanazine (CYA), atrazine (ATR), propazine (PPZ), and terbutylazine (TER) were purchased from Dr. Ehrenstofer (Augsburg, Germany). Stock standard solutions (1 g‚L-1) were prepared in acetonitrile and stored at -20 °C in the dark. Methacrylic acid (MAA), ethylene glycol dimethacrylate (EDMA), and 2,2′-azobis(isobutyronitrile) (AIBN) were purchased from Sigma-Aldrich (Madrid, Spain). Toluene, dichloromethane, HPLCgrade water, acetonitrile, and methanol were purchased from Scharlab (Barcelona, Spain). All other chemicals were of analytical reagent grade. EDMA and MAA were freed from stabilizers by distillation under reduced pressure, and AIBN was recrystallized from methanol prior to use. All other chemicals were used as received. Untreated fused silica capillaries of different inner diameters (0.10, 0.25, 0.32, and 0.53 mm) were obtained from Supelco (Bellefonte, PA). (11) Koester, E. H. M.; Crescenzi, C.; Den Hoedt, W.; Ensing, K.; de Jong, G. J. Anal. Chem. 2001, 73, 3140. (12) Martı´n-Esteban, A.; Turiel, E. Spain Pat. Appl. No. 200700105. (13) Turiel, E.; Martı´n-Esteban, A.; Ferna´ndez, P.; Pe´rez-Conde, C.; Ca´mara, C. Anal. Chem. 2001, 73, 5133. (14) Turiel, E.; Pe´rez-Conde, C.; Martı´n-Esteban, A. Analyst 2003, 128, 137. (15) Cacho, C.; Turiel, E.; Martin-Esteban, A.; Pe´rez-Conde, C.; Ca´mara, C. J. Chromatogr., B 2004, 802, 347.

3100

Analytical Chemistry, Vol. 79, No. 8, April 15, 2007

Apparatus. All chromatographic measurements were performed using an Agilent Technologies 1200 Series HPLC instrument equipped with a quaternary high-pressure pump, a vacuum degasser, an autosampler, and a diode-array detector. A sample volume of 100 µL was injected into a Kromasil 5 ODS (150 mm × 4.6 mm i.d.) analytical column, and analytes were separated by gradient elution from 90% water (A) and 10% acetonitrile (B) to 40% A and 60% B in 25 min. Triazinic herbicides were monitored at 220 nm and quantified by external calibration using peak area measurements. Molecularly Imprinted Polymeric Fiber Preparation. The preparation of MIP fibers is schematically shown in Figure 1. First, fused silica capillaries were cut to approximately 30 cm long pieces and four windows of about 1.0 cm were prepared by burning the protecting polymer layer. Windows were placed at 5 cm intervals, leaving at least 1 cm from the end up to the first window. With the help of a syringe, the capillary was filled with a polymerization mixture, previously degassed with a gentle N2 stream for 5 min, containing template molecule (PPZ, 0.15 mmol), functional monomer (MAA, 0.60 mmol), cross-linker (EDMA, 3 mmol), initiator (AIBN, 0.13 mmol), and porogen (toluene, 0.87 mL), and both capillary ends were closed with two small pieces of rubber. Then, the filled capillaries were introduced in an oven and polymerization took place at 65 °C for a certain period of time. Finally, capillaries were cut and immersed in a 3 M aqueous solution of NH4HF2 for 12 h under agitation with silica walls being etched away. Finally, MIP fibers were immersed in a methanol:acetic acid (1:1, v/v) solution for 2 h to remove the template. Nonimprinted polymeric fibers were also prepared as described above but without the addition of template. The obtained fibers were visually characterized by optical micrographs using an Olympus SZX12 microscope equipped with a digital camera and by scanning electron micrographs using a JEOL JM-6400 (Peabody, MA). Molecularly Imprinted SPME Procedure. Fibers were conditioned during 15 min by immersion in toluene. A volume of 1.7 mL of standard or sample extract solutions in toluene were added into a 2 mL vial containing a 7 × 2 mm Teflon stirring bar. Molecularly imprinted SPME (MI-SPME) was performed by direct immersion of the fiber in the sample for 60 min under low stirring (a vortex just appeared) using a magnetic stirrer. Then, fibers were immersed in toluene for 15 min under stirring in order to remove nonspecific interactions. Finally, the fibers were air-dried for 5 min and desorbed with 200 µL of methanol in a 0.4 mL vial insert by sonication for 5 min. The full methanolic extracts were diluted with 200 µL of water for HPLC analysis. Between samples, fibers were reconditioned in two consecutive steps of 15 min by immersion in acetonitrile and toluene, respectively. Sample Preparation. A volume of 40 mL of acetonitrile was added to 10 g of sample (soil, potato, and pea), and the mixture was sonicated for 30 min. The supernatant was filtered through a 0.45 mm filter, and the extracts were evaporated to dryness by a gentle nitrogen stream and redissolved in 1.7 mL of toluene containing 100 µg L-1 of each triazine for MI-SPME. RESULTS AND DISCUSSION The work presented herein focuses on the preparation of propazine-imprinted polymeric fibers, based on the use of silica capillaries of different inner diameters and lengths, to be used in

Figure 1. Preparation of molecularly imprinted fibers.

Figure 2. (A) Imprinted fibers obtained using capillaries of different inner diameter; (B) imprinted fibers before and after removing silica capillary walls with NH4HF2.

SPME. In this way, theoretically, the resulting imprinted polymer fibers should possess a controlled and homogeneous morphology with optimum characteristics allowing the selective extraction of triazines from soil and vegetable sample organic extracts. Preparation of Imprinted Fibers. Initially, different fibers were prepared using silica capillaries of different inner diameters (0.1, 0.25, 0.32, and 0.53 mm) following the procedure shown in Figure 1. The polymerization mixture consisted of propazine, methacrylic acid, and ethylene glycol dimethacrylate (1:4:20 molar ratio, respectively) dissolved in toluene, and polymerization took place at 65 °C during 24 h. After polymerization, the capillaries were cut down as described in the Experimental Section, and subsequently, the silica walls were etched away leading to imprinted fibers of different thickness. As can be observed in Figure 2A, well-formed fibers were obtained regardless of the inner diameter of the silica capillary used as a mold. Also, the success of the dissolution of silica is clearly observed in Figure 2B, which shows a comparison of two fibers of 0.53 mm thickness before and after the treatment with NH4HF2 (aqueous). These

figures confirm the success of the preparation of polymeric fibers of the desired size following a very simple polymerization procedure, which could be easily performed in any laboratory without the necessity of complex equipment. Moreover, as can be observed in Figure 1, the obtained fibers were flexible and it was possible to bend them to a certain extent. This excellent mechanical property prevents the easy breakage traditionally associated with the commercial coated fused silica fibers. Optimization of MI-SPME Procedure. Loading, Washing, and Elution Solvents. As in any other extraction technique, it is necessary to optimize loading, washing, and elution conditions in order to extract as much analyte as possible minimizing nonspecific interactions. Initially, fibers of 1 cm × 0.53 mm obtained following the procedure previously described (polymerization time ) 24 h) were used to optimize the MI-SPME procedure of standard solutions of propazine at 2 mg‚L-1 concentration level. Initially, a loading time of 1 h at 25 °C was fixed and acetonitrile, dichloromethane, and toluene were tested as loading solvents. In parallel, the same mentioned solvents were tested as washing Analytical Chemistry, Vol. 79, No. 8, April 15, 2007

3101

Figure 3. Relationship between the obtained recoveries of propazine and area (A) or volume (B) of the fiber used.

solutions and acetonitrile, methanol, and methanol:acetic acid mixtures were tested as elution solvents. From this study, it was concluded that toluene was the best solvent to be used during both loading and washing steps and methanol was the optimum as elution solvent. Under those conditions, a maximum recovery of 14.1 ( 2% for the imprinted fiber was obtained, whereas the recovery for the nonimprinted fiber dropped to 0.5 ( 0.1%, which suggests the existence of selective binding sites for propazine within the polymeric network. The optimum solvents are in full agreement with those obtained in previous studies on propazineimprinted polymers carried out by our group,13-15 although the obtained recoveries are much lower. In this sense, it is clear that the low recoveries obtained can only be attributed to the different format (fiber) and/or extraction procedure (SPME) and thus both aspects need to be considered in order to improve these initial results. Fiber Thickness. First, the influence of fiber thickness on the extraction recovery was evaluated. The optimum MI-SPME was followed using fibers of different thicknesses (0.1, 0.25. 0.32, and 0.53 mm), and the recoveries obtained were 0.4, 2.1, 3.6, and 12.2%, respectively. As expected, by increasing the fiber thickness, and thus the amount of sorbent, the obtained recoveries were higher. However, a clear relationship between both parameters was not found. Consequently, the area and the volume of the fibers were calculated, considering the fiber as a cylinder, and represented versus the obtained recoveries. As can be observed in Figure 3, a linear relationship exists between the volume of the fiber and the obtained recoveries which means that not only propazine is retained on the binding sites located on the external surface of the fiber, but also it is able to reach binding sites located in the inner part of the fiber. Figure 4 shows the scanning electron micrographs of an imprinted fiber of 0.53 mm thickness. It can be observed that fibers possess a porous structure, which suggests the existence of pathways for target analytes to reach inner binding sites. Accordingly, those parameters affecting the diffusion of propazine to binding sites (i.e., loading time, loading temperature) and the porosity of the fiber (i.e., polymerization time) were optimized. 3102 Analytical Chemistry, Vol. 79, No. 8, April 15, 2007

Figure 4. Scanning electron micrographs of a propazine-imprinted fiber of 0.53 mm thickness. Magnification: (A) ×4000; (B) ×5000.

Figure 5. (A) Effect of time and temperature of loading step using an imprinted fiber of 0.53 mm thickness after polymerization during 24 h. (B) Effect of polymerization time on the performance of imprinted fibers for the MI-SPME of propazine.

Effect of Time and Temperature of Loading Step. It is well-known that the time and the temperature of the loading step affect the recoveries obtained in SPME. Figure 5A shows the recoveries obtained following the MI-SPME procedure at different loading times and temperatures using a fiber of 0.53 mm thickness (polymerization time ) 24 h). It is evident that, by increasing both the time and temperature, the recoveries reached values closer

Table 1. Recoveries (%) of Triazines after MI-SPME of an Standard Solution of Triazines (0.25 mg‚L-1) Using Three Different Fibers Obtained from the Same Capillarya

desisopropylatrazine desethylatrazine simazine cyanazine atrazine propazine terbutylazine a

fiber A

fiber B

fiber C

20.0 15.9 11.1 10.2 11.6 11.6 9.2

16.1 13.6 13.9 7.0 10.7 10.1 9.7

17.9 14.8 10.9 8.7 9.0 9.5 10.0

Capillary of 0.53 mm i.d. Polymerization time ) 150 min.

to 50%, confirming that propazine mass transfer is favored at longer loading times and higher temperatures. Effect of Polymerization Time. One of the parameters that can affect fiber porosity is the polymerization time, since it controls the degree of cross-linking of the polymeric network. Thus, new fibers were prepared using different polymerization times and used for the MI-SPME. Experiments were carried out with propazine standard solutions, setting the loading time in 1 h at 25 °C, and the results obtained are shown in Figure 5B. At short polymerization times, the obtained recoveries were rather low, likely because binding sites were not properly created yet. At longer polymerization times, low recoveries were obtained as well, but in this case, this result can be attributed to an excessive degree of cross-linking, leading to a less porous structure. Thus, a polymerization time of 150 min was chosen as optimum for the preparation of imprinted fibers. The effect of temperature during the loading step was reoptimized using the new fibers (polymerization time ) 150 min) and the recoveries obtained were 36.5, 40.5, and 43.1% for temperatures of 25, 45, and 60 °C, respectively. Although an increase of the temperature during the loading step had again a positive effect on the recovery obtained, such an increase was not so significant. It seems clear that a polymerization time of 150 min allows the obtainment of porous fibers favoring propazine mass transfer even at 25 °C, being comparable to that obtained with less porous fibers at higher temperatures. Taking into account that working at high temperatures makes a simple technique such as SPME to be a bit more complex, a loading temperature of 25 °C was chosen as optimum for further experiments. Similarly, long loading times are not practical since they affect negatively the total analysis time, and thus, a loading time of 1 h was fixed in subsequent experiments. It is important to stress that the above-mentioned experiments were carried out using 1.7 mL of loading solution of propazine in toluene at 2 mg‚L-1 concentration level. Thus, such a low recovery might be related to a low loading capacity of the fiber. Keeping these comments in mind, a new experiment under optimum conditions was carried out, but a propazine standard solution at 0.25 mg‚L-1 concentration level was used in this case. The calculated recovery was 93.2 ( 5.1% (n ) 3), confirming the low loading capacity of the fiber which was estimated, according to the obtained recoveries, in ∼1.2 µg. This capacity, although low, is enough for the analysis of propazine at the concentration levels typically found in environmental and food samples.

Figure 6. LC-UV chromatograms obtained at 220 nm for (A) a soil sample extract directly injected without any previous cleanup, (B) a soil sample extract enriched with triazines at 0.1 mg‚L-1 concentration level after MI-SPME, (C) a 0.1 mg‚L-1 standard solution of triazines after MI-SPME, and (D) a nonspiked soil sample extract after MI-SPME. Peak numbers: (1) DIA; (2) DEA; (3) SIM; (4) CYA; (5) ATR; (6) PPZ; (7) TER. Chromatographic conditions: see Experimental Section.

Cross-Reactivity Study. Previous studies have demonstrated that propazine-imprinted polymers present cross-reactivity against other triazines and thus have been employed for the simultaneous extraction and cleanup of triazines from different samples.13-15 In this sense, under optimum conditions, three fibers obtained from the same capillary were immersed in 1.7 mL of a mixture of seven triazines (DIA, DEA, SIM, CYA, ATR, PPZ, and TER) in toluene at 0.25 mg‚L-1 concentration level. Table 1 shows the recoveries obtained for the tested compounds for each fiber. It is evident that propazine-imprinted fibers are able to extract simultaneously all the triazines tested although with low recoveries ranging from 20 to 7% depending upon the triazine and the fiber used. As mentioned above, the fiber capacity is rather low and thus the obtained low recoveries can only be attributed to the competition between target analytes for the available binding sites. The slightly higher recoveries obtained for the smallest triazines (DIA and DEA) suggest that these analytes are able to reach binding sites not easily accessible to the bigger triazines. It is clear that molecular size is one of the main parameters influencing the recognition mechanism in MIPs, as has been stated in previous works.13-15 Apart from these theoretical considerations, the obtained results prevent the accurate quantification of target analytes, since the recoveries are dependent on the concentration assayed. However, a good reproducibility was obtained both for fibers obtained from the same capillary (relative standard deviations (RSDs) ) 4-10%, n ) 5) and for fibers obtained using different capillaries prepared on three different days (RSDs ) 7-15%, n ) 3) depending upon the analyte. Thus, the proposed MI-SPME procedure could be used for the screening of target analytes in environmental and food samples at the concentration level required by current legislation. MI-SPME of Triazines from Environmental and Food Sample Extracts. As stated in the introduction, the main goal of Analytical Chemistry, Vol. 79, No. 8, April 15, 2007

3103

fiber or were able to hinder the interaction of target analytes to the binding sites. Finally, according to the chromatogram corresponding to the nonspiked soil sample extract (Figure 6D), it is clear that neither template leaking nor carryover was observed. Figure 7 shows the chromatograms obtained after MI-SPME of potato (Figure 7A) and pea (Figure 7B) sample extracts. As can be observed, a high degree of selectivity was obtained for these complex samples, obtaining chromatograms as clean as that shown in Figure 6C for the standard solution. Finally, it is important to stress that the imprinted fibers were used for more than 30 extractions, including real samples, and no losses in their performance were observed. In addition, as mentioned below, their excellent mechanical properties allow the repeated use of the obtained fibers.

Figure 7. LC-UV chromatograms obtained at 220 nm for potato (A) and pea (B) sample extracts enriched with triazines at 0.1 mg‚L-1 concentration level after MI-SPME. Peak numbers: (1) DIA; (2) DEA; (3) SIM; (4) CYA; (5) ATR; (6) PPZ; (7) TER. Chromatographic conditions: see Experimental Section.

this work is to provide selectivity to SPME of target analytes from complex samples by incorporating selective binding sites to the fibers using molecular imprinting technology. Thus, soil and vegetable samples (potato and pea) were treated as described in the Experimental Section and sample extracts were spiked with a mixture of triazines at a concentration level of 0.1 mg‚L-1 (equivalent to 17 ng‚g-1) and subjected to the proposed MI-SPME procedure. Figure 6 shows the chromatograms obtained for a soil sample extract directly injected without previous cleanup (Figure 6A), a spiked soil sample extract subjected to MI-SPME (Figure 6B), a standard solution of triazines after MI-SPME (Figure 6C), and a nonspiked soil sample extract subjected to MI-SPME (Figure 6D). As can be observed, a high degree of selectivity is obtained by the proposed MI-SPME procedure, allowing the detection of target analytes at very low concentration levels, which would be extremely difficult without performing any cleanup. It is clear that the baseline obtained for the analysis of soil extracts after MI-SPME is as clean as that obtained for the standard solution, demonstrating even more clearly the high selectivity provided by the imprinted fiber. In addition, it is important to stress that no significant differences were found between the recoveries obtained (ranging from 12 to 25% depending upon the analyte) for soil sample extracts and for standard solutions. This result indicates that matrix components did not interact with the

3104

Analytical Chemistry, Vol. 79, No. 8, April 15, 2007

CONCLUSIONS In this work, a successful new procedure for the preparation of imprinted fibers for SPME has been proposed. The fiber synthesis has shown to be easy to perform in the laboratory without the necessity of complex instrumentation. It is important to stress that, although the work was oriented toward the obtainment of imprinted fibers, nonselective fibers like those commercially available could be prepared following the same procedure in a very easy and simple manner. Until now, MIPs had been used in different areas of analytical chemistry but not properly incorporated into a powerful technique such as SPME. This work demonstrates that the combination of molecular imprinting and SPME is possible and thus it opens new areas of research. Obviously, some aspects need some improvements, especially those related to the low capacity of the fibers. In this sense, the use of different cross-linkers and porogens, which affect the final porosity of the fiber and thus the accessibility of target analytes to binding sites, must be further evaluated. Finally, studies on the desorption of target analytes both directly into the injection port of a gas chromatograph and into the commercial SPME-HPLC interface, as well as the preparation of imprinted fibers for other analytes, are underway in our laboratory in order to extend the field of application of the new imprinted fibers which, from our point of view, could become a new tool routinely used in analytical laboratories in the near future. ACKNOWLEDGMENT Comunidad de Madrid and Ministerio de Educacio´n y Ciencia are acknowledged for financial support through Projects Nos. GRMAT-0493-2004 and AGL2005-00905, respectively. E.T. thanks Ministerio de Educacio´n y Ciencia for a Ramon y Cajal contract.

Received for review December 18, 2006. Accepted February 19, 2007. AC062387F