Preparation and application of new sorptive polymers applied to wine with a focus on removal of 3-isobutyl-2-methoxypyrazine

Abstract

Synthetic polymers have been used as solid phase extraction (SPE) sorbents for analysis of food and wine components due to their versatile and stable sorptive properties. Moreover, polymers such as polyvinylpolypyrrolidone (PVPP) can be used as fining agents in the winemaking process and polymeric materials have been investigated as a remedial treatment to remove undesirable compounds from wines. In order to decrease the loss of desirable components and improve selectivity, tailor-made sorbents known as molecularly imprinted polymers (MIPs) have also been produced and broadly studied. MIPs have been widely used in a range of areas where selective binding is of importance, such as immunoassays, sensors, and analytical chemistry and biochemistry (sample clean-up and pre-concentration). MIPs, also known as plastic antibodies, can selectively recognise and bind with target molecules by binding at recognition sites formed with a template molecule during preparation of polymer. Based on the selective binding character, it would be beneficial to apply MIPs in winemaking to help remove off-flavours including elevated methoxypyrazines (MPs) without affecting other wine components. Excessive MPs can lead to unbalanced aroma, yielding wines with strong vegetative and herbaceous notes that suppress the pleasant fruity and floral bouquets. Early harvest grapes to make lower alcohol wines, or grapes from cool climate regions, may contain higher levels of MPs and could produce wines with ‘unripe’ characters. Since grapederived MPs are relatively stable during general winemaking procedures, investigation of MIPs as a post-harvest treatment to specifically remove MPs from grape juice or wine is warranted. 3-Isobutyl-2-methoxypyrazine (IBMP), reminiscent of green capsicum aroma, has been chosen as a target compound in this project. A range of putative imprinted polymers along with non-imprinted polymers (NIPs) were synthesised through bulk polymerisation. Additionally, putative imprinted magnetic polymers (PIMPs) were prepared analogously to the putative imprinted polymers with the addition of magnetic nanoparticles. In this way, PIMPs could be separated by an external magnet compared to non-magnetic polymers that require separation through filtration or used in a packed column. Adsorption tests and physical characterisations were carried out to assess the various polymers. Several fundamental challenges in the molecular imprinting process that could lead to experimental artefacts were observed during trials and relative adjustments were made to avoid the pitfalls. Adsorption tests were first carried out in model wine solution, and the adsorption of IBMP on putative imprinted polymers (magnetic/nonmagnetic) could not be differentiated from non-imprinted counterparts (magnetic/non-magnetic), and neither could their isotherms. Further adsorption test in original porogen solvent showed that there was no difference between the imprinted and the non-imprinted counterparts, meaning the adsorption was dominated by hydrophobic interactions without specific binding in model wine solution. During the preliminary investigation of imprinting methods, microwave-assisted polymerisation was adopted as comparison. Results showed that microwave synthesis did not introduce differences to the polymers regarding adsorption isotherms and physical characters (determined by scanning electron microscopy and Fourier-transform infrared spectroscopy) compared to the thermal synthesis. The practical usage of magnetic polymers in winemaking was investigated with putative imprinted magnetic polymer (PIMP) and non-imprinted magnetic polymer (NIMP) added to IBMP-spiked Cabernet Sauvignon grape must, pre- and post-fermentation. Polylactic acid-based (PLA) film was adopted as a post-fermentation treatment to compare. Olfactory sensory analysis revealed that wines treated with PIMP and NIMP were found to have trace ‘green’ characters, whereas, PLA treatments turned out to have noticeable ‘green’ characters and were most similar to the untreated control wines. The sensory results were consistent with the results from chemical analysis of volatiles. IBMP concentrations were decreased from 21 ng/L (in the control wines) to 7 ng/L and 5 ng/L (i.e., below the detection threshold of IBMP in red wine) by PIMP and NIMP post-fermentation addition, respectively, in comparison to the PLA wine having 17 ng/L. The difference between PIMP and NIMP arose with the timing of addition, where pre-fermentation treatments had less sorption of IBMP and other wine volatiles than the post-fermentation treatments, with the same trend observed for colour properties. In spite of the sorption of volatile compounds other than IBMP, the overall aroma intensity and fruity characters were not different according to sensory analysis. The character impact role of IBMP in wine aroma matrix was observed as well. As the key for imprinting relies on the reaction between templates and monomers in the proper porogen environment, a series of newly prepared polymers (non-magnetic) with various template, functional monomer and porogen solvent combinations were trialled to achieve specific binding. Although further study is still required to improve specificity, polymers (non-imprinted) with different functional monomer and porogen solvent combinations were found to have affinities towards different groups of compounds in wine, and could potentially be used as sorbents for fining of wine, SPE or as stationary phases for liquid chromatography. Physicochemical properties (surface area, pore size distribution, and polarity) of polymeric sorbents could be affected by the compositions and production methods and the relation between physicochemical properties and sorption properties in wines was investigated, along with that of four commercial sorbents (C18, Oasis HLB, Strata SDB-L, and PVPP). In terms of production methods, thermally synthesised polymers were found to have larger surface area, on the other hand, microwave synthesised polymers were observed to have narrower pore size distributions with abundance of micro- and mesopores. Polarities of the sorbents could be differentiated by copolymerising various hydrophilic monomers. Thermally synthesised sorbent with double the amount of methacrylic acid as co-monomer in acetonitrile may be more polar and has the potential for carbonyl compounds extraction. Microwave-produced sorbent with 4-vinylbenzoic acid in dichloromethane was found to have higher retention of non-polar, small molecular weight compounds (e.g., various volatiles) in white and red wines. Sorbents prepared by microwave in dichloromethane with methacrylic acid and acrylic acid as co-monomers, respectively, were observed to have higher affinity towards moderately polar, small to large molecular weight compounds (e.g., pigments and other phenolics) in wines. As for the four commercial sorbents, C18 was found to have high affinity towards volatiles and low affinity towards phenolics in white and red wines; PVPP, on the contrary, was observed to have high sorption of phenolics and low retention of volatile compounds. Oasis HLB and Strata SDB-L were found to have higher sorptive ability among the commercial sorbents, with Oasis HLB having higher affinity towards moderately polar compounds than that of Strata SDBL. With similarities to the commercial sorbents regarding hydrophobic and hydrophilic characters, our synthesised polymers showed potential use as selective sorbents. A series of polymers (magnetic, putative imprinted) were successfully synthesised and characterised. Results from this project revealed some fundamental challenges faced with the molecular imprinting process and some pitfalls that could be avoided in future work. The usage of magnetic polymers in winemaking provided a feasible post-harvest option to remediate wines with elevated levels of MPs. The possibilities of applying selective polymeric sorbents in wine for analytical and remedial purposes have been investigated. Improvement of the specificity of the imprinted polymers applied in wine should also be the focus of future work.