8960
Langmuir 2006, 22, 8960-8965
Encapsulation and Selective Recognition of Molecularly Imprinted Theophylline and 17β-Estradiol Nanoparticles within Electrospun Polymer Nanofibers Ioannis S. Chronakis,*,† Alexandra Jakob,† Bengt Hagstro¨m,† and Lei Ye*,‡ IFP Research, Swedish Institute for Fiber and Polymer Research, Box 104, SE 431 22 Mo¨lndal, Sweden, and Pure and Applied Biochemistry, Center for Chemistry and Chemical Engineering, Lund UniVersity, Box 124, SE 22 100 Lund, Sweden ReceiVed May 17, 2006. In Final Form: August 3, 2006
Molecularly imprinted nanoparticles are cross-linked polymer colloids containing tailor-made molecular recognition sites. In this study, molecularly imprinted nanoparticles were easily encapsulated within polymer nanofibers using an electrospinning technique to produce a new type of molecular recognition material. Poly(ethylene terephthalate) (PET) was used as the supporting nanofibers matrix to encapsulate theophylline and 17β-estradiol imprinted nanoparticles. The composite nanofibers had an average diameter of 150-300 nm, depending on the content of molecularly imprinted nanoparticles. For the theophylline and 17β-estradiol imprinted polymers, an optimal loading of molecularly imprinted nanoparticles was 25-37.5 wt % based on PET. The composite nanofibers prepared under these conditions had a well-defined morphology and displayed the best selective target recognition. Our approach of electrospinning-formolecularly imprinted nanoparticles-encapsulation has unique advantages and opens new application opportunities for molecularly imprinted nanoparticles and electrospun nanofibers.
Introduction There is currently great research interest in nanomaterials as their small physical size often offers much improved and even new functions that cannot be achieved with bulk materials. In this respect, nanomaterials bearing tailor-made molecular binding properties are very attractive because they can potentially be used to substitute biological receptor molecules to afford high specific target recognition in different applications, e.g. affinity separations, substrate and enantioselective catalysis, assays, and development of biomimetic sensors. Among the different synthetic strategies that are presently studied, molecular imprinting is probably the mostly straightforward for the purpose of producing nano- and microstructured materials that have predesigned molecular recognition capabilities. Typically, molecular imprinting involves free radical polymerization of functional monomer and cross-linker in the presence of a template molecule. During the radical polymerization process, the functional monomer interacts with the template to form a self-assembled complex, which is further stabilized by crosslinking of the polymer backbone. After polymerization, the template is removed from the solid polymer matrix to furnish an empty cavity, which can rebind the original template molecule with high specificity. During the past few years some interesting studies have been directed toward downsizing molecularly imprinted polymers (MIPs) into the sub-micrometer and nanometer range.1-5 Compared to the traditional MIPs produced as * To whom correspondence should be addressed. Tel.: + 46 31 706 63 00 (I.S.C.); + 46 46 222 95 60 (L.Y.). Fax: + 46 31 706 63 63 (I.S.C.); + 46 46 222 46 11 (L.Y.). E-mail:
[email protected] (I.S.C.);
[email protected] (L.Y.). † Swedish Institute for Fiber and Polymer Research. ‡ Lund University. (1) Perez, N.; Whitcombe, M. J.; Vulfson, E. N. J. Appl. Polym. Sci. 2000, 77, 1851. (2) Biffis, A.; Graham, N. B.; Siedlaczek, G.; Stalberg, S.; Wulff, G. Macromol. Chem. Phys. 2001, 202, 163. (3) Vaihinger, D.; Landfester, K.; Krauter, I.; Brunner, H.; Tovar, G. E. M. Macromol. Chem. Phys. 2002, 203, 1965.
bulk monolith or large particles, molecularly imprinted microspheres and nanoparticles have already shown superior features such as faster binding kinetics and higher target binding capacities.6,7 In addition, we have demonstrated previously that, by limiting the physical size of MIP beads, specific binding sites can be confined in close proximity of copolymerized reporter molecules so that the specific binding event can be directly monitored by simply following the optical response from the reporter.8 Electrospinning is being increasingly used to produce nanofiber materials. During electrospinning, a high voltage is applied to a polymer solution to produce a polymer jet. Accompanied with the fast evaporation of solvent, the charge density in the polymer jet is increased, which results in the formation of nanofibers. The electrospinning process allows (cost-effectively) continuous production of polymer fibers with diameters between a few nanometers to several micrometers. This technique has been used to prepare nanofibers from a wide range of polymer materials.9-19 Furthermore, encapsulation of transition metal (4) Yang, H.-H.; Zhang, S.-Q.; Yang, W.; Chen, X.-L.; Zhuang, Z.-X.; Xu J-G.; Wang, X.-R. J. Am. Chem. Soc. 2004, 126, 4054. (5) Choi, K. M.; Rogers, J. A.; Shea, K. J. Mater. Res. Soc. Symp. Proc. 2005, 872, 137. (6) Ye, L.; Cormack, P. A. G.; Mosbach, K. Anal. Commun. 1999, 36, 35. (7) Ye, L.; Weiss, R.; Mosbach, K. Macromolecules 2000, 33, 8239. (8) Ye, L.; Surugiu, I.; Haupt, K. Anal. Chem. 2002, 74, 959. (9) Reneker, D. H.; Chun, I. Nanotechnology 1996, 7, 216. (10) Reneker, D. H.; Yarin, A. L.; Fong, H.; Koombhonge, S. J. Appl. Phys. 2000, 87, 4531. (11) Deitzel, J. M.; Kleinmeyer, J.; Harris, D.; Tan, N. C. B. Polymer 2001, 42, 261. (12) Frenot, A.; Chronakis, I. S. Curr. Opin. Colloid Interface Sci. 2003, 8, 64. (13) Dzenis, Y. Science 2004, 304, 1917. (14) Li, D.; Xia, Y. AdV. Mater. 2004, 16, 1151. (15) He, J.-H.; Wan, Y.-Q. Polymer 2004, 45, 6731. (16) Jayaraman, K.; Kotaki, M.; Zhang, Y.; Mo, X.; Ramakrishna, S. J. Nanosci. Nanotechnol. 2004, 4, 52. (17) Subbiah, T.; Bhat, G. S.; Tock, R. W.; Parameswaran, S.; Ramkumar, S. S. J. Appl. Polym. Sci. 2005, 96, 557. (18) Dersch, R.; Steinhart, M.; Boudriot, U.; Greiner, A.; Wendorff, J. H. Polym. AdV. Technol. 2005, 16, 276.
10.1021/la0613880 CCC: $33.50 © 2006 American Chemical Society Published on Web 09/19/2006
Molecularly Imprinted Nanoparticle Encapsulation
Langmuir, Vol. 22, No. 21, 2006 8961 Table 1. Preparation of Molecularly Imprinted Nanoparticles
Figure 1. Chemical structure of the polyester (PET) used for electrospinning and of the template compounds used for molecular imprinting.
nanoparticles and carbon nanotubes into composite nanofibers has been realized.14,20-28 In a recent work we have demonstrated for the first time that electrospun template directed molecular imprinting is a viable method for creating robust, molecularly imprinted nanofibers that can selectively rebind the target molecule. Molecular recognition sites in polymer nanofibers were prepared with a simple electrospinning method using a combination of structural and functional polymers as the starting material. When the template, 2,4-dichlorophenoxyacetic acid (2,4-D) was removed by solvent extraction, imprinted binding sites were left in the nanofiber materials that are capable of selectively rebinding the target molecule.29 This simple method requires that the condition of forming imprinted binding sites to be compatible with the electrospinning process, making the method applicable only to certain types of template compounds. The aim of the present work is to utilize the electrospinning technique to encapsulate premade MIP nanoparticles into composite polymer nanofibers. The composite nanofibers, when collected on a flat surface, form well-controlled filtration matrixes that can be used in affinity separations, e.g., in solid-phase extraction in analytical applications. We envisaged that the scalable electrospinning process would allow mass production of composite nanofiber materials. As a first proof-of-principle, we chose to encapsulate nanoparticles imprinted against theophylline and 17β-estradiol, two interesting drug molecules that we have previously used as model compounds to study molecular imprinting (Figure 1). Experimental Section Materials. Poly(ethylene terephthalate) (PET) was purchased from Wellman International Ltd. (Ireland). Analytical grade trifluoroacetic acid (TFA) and dichloromethane (DCM) were purchased from SigmaAldrich and used without further purification. [8-3H]Theophylline (specific activity, 18.5 Ci/mmol) was from Sigma, and [2,4,6,73H(N)]estradiol (specific activity, 72.0 Ci/mmol) was from NEN Life Science Products, Inc. (Boston, MA). Scintillation liquid, Ecoscint O, was purchased from National Diagnostics (Manville, NJ). Methacrylic acid (MAA, >99%) and trimethylolpropane trimethacrylate (TRIM, >90%) were from Merck (Darmstadt, Germany) and used as supplied. Azobisisobutyronitrile (AIBN, (19) Chronakis, I. S. J. Mater. Process. Technol. 2005, 167, 283. (20) Senecal, K. J.; Ziegler, D. P.; He, J.; Mosurkal, R.; Schreuder-Gibson, H.; Samuelson, L. A. Mater. Res. Soc. Symp. Proc. 2002, 788, 285. (21) Pedicini, A.; Farris, R. J. J. Polym. Sci., Polym. Phys. Ed. 2004, 42, 752. (22) Yang, Q. B.; Li, D. M.; Hong, Y. L.; Li, Z. Y.; Wang, C.; Qiu, S. L.; Wei, Y. Synth. Met. 2003, 137, 973. (23) Ko, F.; Gogotsi, Y.; Ali, A.; Nagiub, N.; Ye, H.; Yang, G.; Li, C.; Willis, P. AdV. Mater. 2003, 15, 1161. (24) Salalha, W.; Dror, Y.; Khalfin, R. L.; Cohen, Y.; Yarin, A. L.; Zussman, E. Langmuir 2004, 20, 9852. (25) Seoul, C.; Kim, Y. T.; Baek, C.-K. J. Polym. Sci., Polym. Phys. Ed. 2003, 41, 1572. (26) Sen, R.; Zhao, B.; Perea, D.; Itkis, M. E.; Hu, H.; Love, J.; Bekyarova, E.; Haddon, R. C. Nano Lett. 2004, 4, 459. (27) Hou, H.; Jun, Z.; Reuning, A.; Schaper, A.; Wendorff, J. H.; Greiner, A. Macromolecules 2002, 35, 2429. (28) Hou, H. Reneker, D. H. AdV. Mater. 2004, 16, 69. (29) Chronakis, I. S.; Milosevic, B.; Frenot, A.; Ye, L. Macromolecules 2006, 39, 357.
polymer
template
MIP-TH NIP-TH MIP-ES NIP-ES
theophylline/0.511 mmol 17β-estradiol/0.734 mmol
polymerization conditiona
template binding in acetonitrile (%)
UV 350 nm, 20 °C UV 350 nm, 20 °C 60 °C 60 °C
38b 6.3b 44c 11c
a The template was dissolved in a solution of MAA (3.454 mmol) and TRIM (1.485 mmol) in 40 mL of anhydrous acetonitrile prior to the addition of AIBN (85.2 µmol). b Polymer loading, 3.2 mg. c Polymer loading, 30 mg.
Figure 2. SEM image of PET nanofiber mats. The PET nanofiber was electrospun from a 10% PET solution (wt) in a 1:1 (v/v) mixture of TFA and DCM. The scale bar is 10 µm. >98%) from Merck was recrystallized from methanol prior to use. Anhydrous acetonitrile used for polymer syntheses was from LabScan (Stillorgan, Co., Dublin, Ireland). Preparation of Molecularly Imprinted Nanoparticles. The molecularly imprinted nanoparticles were prepared with theophylline (TH) and 17β-estradiol (ES) as the template compounds using a precipitation polymerization method (Table 1). Due to its limited solubility, the template 17β-estradiol was dissolved in the prepolymerization mixture containing methacrylic acid, trimethylolpropane trimethacrylate, and acetonitrile at 60 °C prior to addition of the initiator. Details of the synthetic procedure can be found in previous publications.6,7 Two types of non-imprinted nanoparticles (NIP-TH and NIP-ES) were prepared under the same conditions as those used to prepare molecularly imprinted nanoparticles (MIP-TH and MIP-ES) except that no template was added to the reaction mixture. After polymerization, the nanoparticles were collected by ultracentrifugation, washed repetitively in methanol:acetic acid (90: 10, v/v) until no template could be detected from the washing solvent with an UV-vis spectrophotometer. The nanoparticles were finally washed in acetone and dried in a vacuum. Scanning electron micrographs clearly indicate discrete nanosphere morphology with average diameters of 200-300 nm (see below). Electrospinning. Electrospinning of PET fiber was carried out at room temperature at a high voltage of 20 kV (HV Power Supply, Gamma High Voltage Research, Ormond, FL). The syringe used in the experiments had a capillary tip with a diameter of 0.8 or 0.9 mm. A copper wire was mounted in the capillary tip and used as the positive electrode. A grounded aluminum foil was used as the counter electrode and mounted at a distance of 20 cm from the capillary tip. Continuous PET fibers were collected on the aluminum foil in the form of a fibrous mat. After electrospinning, the nanofiber mats were placed in a vacuum chamber for at least 24 h to remove organic solvent. To encapsulate MIP nanoparticles, the imprinted polymers were suspended in 1 mL of DCM and sonicated for 20 min. To the suspension was added TFA (1 mL), whereafter the mixture was stirred vigorously for 10 min. Finally, PET (200 mg) was added to the mixture and stirred for 3 h until complete dissolution. The suspension obtained contained 10% PET (w/w), and the nanoparticle
8962 Langmuir, Vol. 22, No. 21, 2006
Chronakis et al.
Figure 3. SEM images of MIP-ES nanoparticles (a) and the electrospun PET nanofibers containing different amounts of MIP-ES: (b) 25, (c) 37.5, (d) 50, and (e) 75%. The scale bar is 1 µm for the nanoparticle and 10 µm for the nanofiber images. content varied from 10 to 75% based on the PET weight. This was used as the feeding material for electrospinning the composite fibers. To ensure that all the TFA solvent was removed, the composite nanofiber mats were treated with methanol in an accelerated extractor and dried in a vacuum before they were characterized by radioligand binding analysis. The nanofiber mats were extracted in methanol at 60 °C and 2000 psi for 30 min using a Dionex ASE200 accelerated solvent extractor (Sunnyvale, CA). The purified nanofiber mats were finally dried in a vacuum chamber for at least 24 h. Scanning Electron Microscopy. The morphology and diameter of PET nanofibers were determined with a scanning electron microscope (SEM; JEOL JSM-T300). A small section of the fiber mat was placed on the SEM sample holder and sputter-coated with gold prior to the analysis. Radioligand Binding Analysis. Polymer nanoparticles were added into a solution of [8-3H]theophylline (20 nM) or [2,4,6,7-3H(N)]-
estradiol (0.54 pM) prepared in different solvents. The samples were gently stirred on a rocking table at 20 °C for 16 h. After the incubation, the polymer nanoparticles were separated by centrifugation at 14 000 rpm for 5 min. The supernatant (500 µL) was mixed with 10 mL of scintillation liquid and the radioactivity measured with a model 2119 Rackbeta β-radiation counter from LKB Wallac (Sollentuna, Sweden). The imprinted and reference nanofiber mats were cut into 3 × 0.5 cm strips and placed in 2 mL polypropylene vials. The PET sample strip (3 × 0.5 cm) had an average mass of 2.0 mg. The amount of nanoparticles encapsulated in the sample strip thus varied from 0.18 to 0.86 mg. To each vial was added 1.5 mL of radioisotope labeled theophylline (20 nM in toluene) and 17β-estradiol (0.54 pM in n-heptane). The samples were gently stirred on a rocking table at 20 °C for 4 h. Finally, 0.65 mL of supernatant was withdrawn, from which the amounts of unbound theophylline and 17β-estradiol
Molecularly Imprinted Nanoparticle Encapsulation
Langmuir, Vol. 22, No. 21, 2006 8963
Figure 4. SEM images of electrospun PET nanofibers containing 37.5% of MIP-ES at different magnifications: 2000 (a), 10 000 (b), 20 000 (c), and 40 000 (d). were quantified by radioactivity measurement using the same Rackbeta liquid scintillation counter.
Results and Discussions Electrospinning of Composite PET Fibers for Nanoparticle Encapsulation. To optimize the electrospinning condition, we started from neat PET polymer solution to prepare electrospun nanofiber mats. The choice of PET as the supporting material for MIP encapsulation was based on PET being easily fabricated into well-defined ultrafine fibers with electrospinning. As seen in Figure 2, it was possible to obtain a continuous nanofibrous structure from a 10% PET solution dissolved in a 1:1 (v/v) mixture of DCM and TFA. The electrospun PET nanofibers formed an evenly distributed fibrous mat on the aluminum foil. The fibers exhibited a cylindrical morphology with an average fiber diameter of approximately 150 nm. PET polymer nanofibers containing different amounts of MIP nanoparticles and the corresponding control, nonimprinted nanoparticles, were prepared with the optimized electrospinning condition. As illustrated in Figures 3-5, MIP nanoparticle domains indeed remained clearly visible under the scanning electron microscope. The nanoparticles encapsulated in the PET support were distributed evenly along the nanofiber structures. The PET nanofibers could accommodate up to 75 wt % of MIPES and NIP-ES. For MIP-TH and NIP-TH, the maximum amount of nanoparticles that could be encapsulated in PET was 62.5 wt %. The two imprinted polymers had been prepared under different reaction conditions6,7 and were expected to have slightly different surface properties. This may have affected their compatibility with PET during the electrospinning process. In general, the
present electrospinning conditions allowed encapsulation of 2070 wt % of MIP nanoparticles in the PET nanofiber without any obvious particle loss. Despite the high particle content, most of the fibers had a regular morphology containing isolated nanoparticles, all encapsulated in the fibrous PET material. When the nanoparticle content was further increased, it became difficult to electrospin the PET/nanoparticle mixture, presumably because of increased nanoparticle aggregation. The diameter of the PET nanofibers became larger with increased nanoparticle content. Figure 6 illustrates the change in fiber diameter when different amounts of MIP-TH and MIPES nanoparticles were added to the starting suspension. At the highest nanoparticle content, the average diameter of PET nanofibers was almost twice that of the pure PET nanofibers. The MIP-TH had a greater effect on fiber diameter than MIP-ES, presumably because the former more easily tended to form an aggregated structure, especially at a high particle load (Figure5). Molecular Recognition of the Imprinted Nanofiber Mats. The composite PET nanofiber mats displayed good stability in different solvents. The composite nanofiber mats could be kept in toluene, acetonitrile, n-heptane, and water for prolonged periods (more than 1 week) without losing the encapsulated nanoparticles. To ensure that all the TFA solvent was removed, the composite nanofiber mats were treated with methanol in an accelerated extractor and dried in a vacuum before they were characterized by radioligand binding analysis. The favorable ligand binding selectivity of the two kinds of imprinted nanoparticles has been demonstrated in our previous studies (Table 1).6,7 To investigate if the imprinted binding sites in the encapsulated nanoparticles
8964 Langmuir, Vol. 22, No. 21, 2006
Chronakis et al.
Figure 6. Average diameter of PET fibers as a function of nanoparticle content: 17β-Estradiol imprinted nanoparticles (MIPES) (open circle); theophylline imprinted nanoparticles (MIP-TH) (solid square).
Figure 5. SEM images of MIP-TH nanoparticles (a), and the electrospun PET nanofibers containing different amounts of MIPTH: (b) 37.5 and (c) 62.5%. The scale bar is 1 µm for the nanoparticle and 10 µm for the nanofiber images.
are still accessible for target recognition, radioligand binding experiments were carried out with the different composite nanofibers. Figure 7a depicts the binding of labeled theophylline on PET nanofibers containing different amounts of MIP-TH and NIPTH nanoparticles. A control experiment indicated that the background theophylline binding to pure PET nanofiber was less than 5%. As shown in Table 1 and Figure 7a, the imprinted nanoparticles indeed maintain, qualitatively, an excellent binding selectivity after encapsulation. The optimal load of theophyllineimprinted nanoparticles in the composite nanofibers was found to be 25-37.5% based on the weight of PET. Under these conditions template bounding with the imprinted composites is 2-3-fold of that with the nonimprinted composites, indicating
Figure 7. (a) Binding of labeled theophylline on theophylline imprinted polymer nanofiber (solid square) and nonimprinted polymer nanofiber (open circle). (b) Effect of solvent composition on theophylline binding to 2 mg of the imprinted nanoparticles (MIPTH, solid square) and the nonimprinted nanoparticles (NIP-TH, open circle).
clearly the specific molecular recognition capability of the encapsulated nanoparticles. To evaluate the selectivity of the imprinted nanoparticles themselves under similar condition, MIPTH and NIP-TH were incubated with tritium-labeled theophylline in toluene containing different amounts of acetonitrile. We noticed that, in neat toluene, the polymer nanoparticles formed large aggregates that were difficult to be accurately dispensed with a pipet. Therefore, acetonitrile was added as a cosolvent to give stable nanopaticle suspension easy to handle. As shown in Figure 7b, the imprinted nanoparticles (MIP-TH) bound more theophylline than the nonimprinted nanoparticles (NIP-TH), although an increased nonspecific binding was observed in the most toluene-rich solvent. These results suggest that after the physical encapsulation into PET nanofibers, the imprinted binding sites of the nanoparticles remained accessible for specific analyte binding. For more, the PET nanofibers seem to provide an efficient
Molecularly Imprinted Nanoparticle Encapsulation
Langmuir, Vol. 22, No. 21, 2006 8965
nanoparticle loading may not be surprising: when the amount of PET was not sufficient to provide effective encapsulation, part of the nanoparticles were exposed to the bulk, nonpolar solvent, in which nonspecific adsorption became dominant. In fact, when tested in neat n-heptane, estradiol binding to MIP-ES and NIP-ES was almost quantitative and indistinguishable (data not shown). However, when the binding experiment was carried out in n-heptane:THF:acetic acid (40:10:0.1, v/v/v), the imprinted nanoparticles displayed much higher template binding than the nonimprinted nanoparticles (Figure 8b). Similar to the theophylline imprinted composite nanofibers, the estradiol recognition sites of the encapsulate MIP-ES nanoparticles remained accessible even under altered solvent conditions.
Conclusions
Figure 8. (a) Binding of labeled estradiol on estradiol imprinted polymer nanofiber (solid square) and nonimprinted polymer nanofiber (open circle). (b) Binding of labeled estradiol on 2 mg of polymer nanoparticles MIP-ES (solid square) and NIP-ES (open circle) in n-heptane:THF:acetic acid (40:10:0.1, v/v/v).
surface modification for the encapsulated nanoparticles, making them applicable in many different solvent systems. The PET nanofiber containing MIP-ES showed an interesting binding profile (Figure 8a). Only the fiber containing 37.5 wt % of MIP-ES showed specific binding of 17β-estradiol in n-heptane. At this point the imprinted nanofiber bound four times more 17β-estradiol than the nonimprinted nanofiber. When no nanoparticle was encapsulated, the fraction of 17β-estradiol bound to PET nanofiber was less than 10%. At present it is not clear why the nonimprinted composite material bound slightly less estradiol at 37.5% than at 25% particle content. Actually, at these nanoparticle contents estradiol binding to the composite nanofibers was close to that achieved by the PET nanofiber itself. The fact that a high nonspecific binding occurred at high
We have demonstrated that molecularly imprinted nanoparticles can easily be encapsulated within nanofiber materials using an electrospinning technique. The electrospinning of composite fibers containing MIP nanoparticles produced a new type of molecular recognition material, which has potential applications in affinity separation, sensing, and diagnostics. In the present work, PET was used as the supporting matrix to encapsulate theophylline and 17β-estradiol imprinted nanoparticles. The composite nanofibers had an average diameter of 150-300 nm, depending on the content of MIP nanoparticles. More importantly, the composite nanofibers maintained a favorable molecular recognition capability. For the theophylline and 17β-estradiol imprinted polymers, an optimal loading of MIP nanoparticles was 25-38 wt % based on PET. The composite nanofibers prepared under these conditions had a well-defined morphology and displayed the best target recognition. Our approach of electrospinning-forMIP-encapsulation has unique advantages: it opens new application opportunities for MIP nanoparticles and electrospun nanofibers, for example in solid-phase extraction for sample preparation and in industrial filtration, where a fast flow is desired. This approach may also be used to provide a simple surface modification of MIPs to comply with altered application environments, for example to deal with aqueous samples. Acknowledgment. The financial support of the Swedish Agency for Innovation Systems (VINNOVA) and the European Commission (MASMICRO Integrated Project) to IFP Research is gratefully acknowledged. LA0613880
