Mycotoxins are contaminants produced by various fungal species and can be present in diverse food and feed matrices. These secondary metabolites can induce several toxic effects in both humans and animals upon contamination, although they are often present in low ppbppt concentrations. The mycotoxin issue is a worldwide concern which resulted already in several national and European guidelines, directives and regulations regarding maximum mycotoxin levels. This legislation is absolutely necessary since economic losses due to contamination should not be underestimated. As an example, regulations exist concerning maximum levels of ergot alkaloid sclerotia and deoxynivalenol (DON) in several food and feed matrices. Consequently, rapid, accurate and sensitive mycotoxin analysis is required to be able to identify and quantify the mycotoxin content in samples. Mycotoxin analysis consists of rapid screening and confirmation methods in which the general interest moved toward multimycotoxin detection. On the one hand, the present rapid tests often use antibodies as specific recognition elements but are also characterized by some important drawbacks. On the other hand, sample purification prior to LC-MS/MS confirmation analysis is often based on nonselective interactions. (Chapter 1) In order to meet the existing need for more performant analytical tools and methods for rapid, sensitive, robust and selective multi-mycotoxin analysis, a new type of solid phase extraction (SPE) sorbents based on porous scaffolds was proposed. These carriers were produced by the BioplotterTM technology and subsequently loaded with recognition elements. Next to antibodies which are the gold standard in mycotoxin analysis, the popular synthetic alternative molecularly imprinted polymers (MIP) were used to immobilize onto poly-ε- caprolacton (PCL) scaffolds. Therefore, sub-micrometer sized spherical MIP particles were targetted. After production and characterization of the particles, the optimal MIP immobilization parameters were investigated on 2D PCL surfaces, as well as for antibodies, and followed by transfer to the more complex 3D format for SPE applications. (Chapter 2) In the first part of this research, MIP were produced as recognition elements for the detection of the six major ergot alkaloids and their epimers. MIP represent synthetic recognition elements capable of interacting with a target analyte to obtain selective binding and/or detection and/or removal thereof. These particles can be produced by several polymerization mechanisms and techniques. MIP for ergot alkaloids were produced by using metergoline as template, methacrylic acid as functional monomer and trimethylolpropane trimethacrylate as crosslinker in precipitation polymerization with acetonitrile as solvent. Various particle characterization techniques were explored and evaluated. A first group concerns morphological characterization methods such as electron microscopy and particle size analysis. By using this type of techniques, the production and drying processes of the desired MIP particles were optimized resulting in spherical sub-micrometer sized particles in which aggregation was an important issue. A 1/6/24 ratio of template/functional monomer/crosslinker was selected for further experiments. Secondly, techniques such as nuclear magnetic resonance spectroscopy and thermogravimetric analysis have evolved for investigating chemical structure characteristics and thermal features respectively. Finally, selective target binding is mostly studied by liquid chromatography coupled to different detectors. More specifically, equilibrium experiments by LC-MS/MS clearly indicated a higher metergoline rebinding capacity of MIP compared to the non-imprinted polymers (NIP). However, this observation was not confirmed for the rebinding of the ergot alkaloids in acetonitrile. Consequently, several mixtures of acetonitrile with buffers were tested in which the highest recovery percentages were found when the ACN/[KH2PO4/HCl]-buffer pH3 (80/20) was used as loading solvent. MIP recoveries ranged from 45.06 ± 2.65 % for Ecr(n) to 48.93 ± 6.73 % for Em(n) which were significantly higher compared to the corresponding NIP. These recovery percentages were not as high as expected compared to the initial target which could possibly be related to particle aggregation. In summary, Chapter 3 shows that sub-micrometer sized spherical MIP particles were obtained which could be applied for immobilization onto surfaces and 3D structures, although functionality for ergot alkaloid recognition was not optimal. Chapter 4 describes the immobilization of MIP particles on 2D spincoated PCL surfaces. Hereto, Pluronic® F127 bismethacrylate (PF127-BMA) building blocks were applied. In general, successful immobilization of MIP structures on surfaces such as immobilized particles and films are frequently investigated by electron microscopy, atomic force microscopy, static contact angle measurements, X-ray photoelectron spectroscopy (XPS), quartz crystal microbalance and surface plasmon resonance. Several protocols were tested for the immobilization of MIP for ergot alkaloids by ranging solvent ratios and the hydrogel building block concentration. Due to aggregation of MIP particles in MilliQ water, the hydrogel building block-MIP solution was finally prepared in 50/50 acetonitrile/MilliQ. Furthermore, within the examined concentration range the 7.5 w% PF127-BMA was selected for successful immobilization of MIP with low risk of ‘kin formation’and subsequent drop in rebinding capacity. Additionally, the 7.5 w% PF127-BMA network was not disturbed by the presence of the MIP particles. Consequently, these optimized conditions were transferred towards the 3D MISPE application. Immobilization of MIP particles on 3D PCL porous scaffolds by the 7.5 w% PF127-BMA was shown to be successful by microscopy and μCT analysis. However, the presence of hydrogel was in some cases also observed inside the pores of the scaffolds which could possibly influence the final perfusion during SPE. The highest specific rebinding percentage for a 1 μM metergoline solution was found after performing only one washing step and contained 44.87 ± 8.30 % which was not as high as targeted. Furthermore, ergot alkaloids were not specifically retained by the MIP sorbent. However, MIP particles were not lost during SPE since their presence after this process was confirmed by SEM analysis. Several hypotheses were formulated which could clarify the low binding capacity of the new type of SPE columns. Consequently, optimization of the final multi-ergot alkaloid MISPE scaffold application was not achieved in Chapter 5. The second part of this research applied antibodies as recognition elements onto the PCL scaffolds as SPE sorbent for detection of DON. Hereto, primary amine functions were introduced onto the PCL surface by the 2-aminoethyl methacrylate (AEMA) grafting technology. Prior to the immobilization of anti-DON antibodies, secondary antibodies were immobilized onto 2D PCL substrates. Subsequently, immobilization of secondary antibodies was optimized through application of different models and concentrations. In this respect, both an ionic and covalent model was used to immobilize secondary antibody solutions of 0.0021 μg/μl, 0.0042 μg/μl, 0.0084 μg/μl and 0.0168 μg/μl. After characterization by XPS and radiolabeling analysis, it was concluded that within this application the ionic model did not result in a sufficient secondary antibody deposition. On the contrary, the 0.0084 μg/μl concentration was determined as the optimal secondary antibody dilution within the covalent immobilization model and transferred to the 3D antibody SPE format. (Chapter 6) In Chapter 7, the specific covalent immobilization of a 0.0084 μg/μl secondary antibody solution on 3D PCL scaffolds was confirmed by radiolabeling and μCT-SPECT analysis. A total amount of 11.74 ± 0.10 μg secondary antibody was immobilized on PCL AEMA grafted scaffolds which equaled a yield of 41.93 %. Furthermore, the edges of the scaffolds were characterized by higher amounts of immobilized antibodies which could be related to the immobilization protocol. However, secondary antibodies within the center of the scaffolds were grafted with an acceptable homogeneity, variability and reproducibility. Additionally, the antibodies were present on the pores which created positive expectations for the final SPE application. Despite these promising results, no selective binding of DON was obtained. Similar to the MIP 3D sorbents, several potential causes were stated to address the poor affinity of the new sorbents towards DON. Consequently, further optimization is still required to use this type of sorbents in new SPE applications for the detection of mycotoxins. In conclusion, although the recognition elements were successfully immobilized on the 3D scaffolds, no optimal SPE conditions could be obtained resulting in specific binding of ergot alkaloids and DON by respectively MIP and antibodies immobilized onto PCL scaffolds. In future research, the proposed new types of antibody and MIP loaded porous sorbents require optimization before they can be used in SPE applications prior to LC-MS/MS analysis and rapid screening tests combined with quantum dots as labels. Additionally, stability and reusability studies need to be performed, together with real sample testing and validation of the new methods. (Future perspectives)