Effect of polymerisation by microwave on the physical properties of molecularly imprinted polymers (MIPs) specific for caffeine

Molecularly Imprinted Polymers (MIPs) are a class of polymeric materials that exhibit highly specific recognition properties towards a chosen target. These “smart materials” offer robustness to work in extreme environmental conditions and cost effectiveness; and have shown themselves capable of the affinities/specificities observed of their biomolecular counterparts. Despite this, in many MIP systems heterogeneity generated in the polymerisation process is known to affect the performance. Microwave reactors have been extensively studied in organic chemistry because they can afford fast and well-controlled reactions, and have been used for polymerisation reactions; however, their use for creating MIPs is limited. Here we report a case study of a model MIP system imprinted for caffeine, using microwave initiation. Experimental parameters such as polymerisation time, temperature and applied microwave power have been investigated and compared with polymers prepared by oven and UV irradiation. MIPs have been characterised by BET, SEM, DSC, TGA, NMR, and HPLC for their physical properties and analyte recognition performance. The results suggest that the performance of these polymers correlates to their physical characteristics. These characteristics were significantly influenced by changes in the experimental polymerisation parameters, and the complexity of the component mixture. A series of trends were observed as each parameter was altered, suggesting that the performance of a generated polymer could be possible to predict. As expected, component selection is shown to be a major factor in the success of an imprint using this method, but this also has a significant effect on the quality of resultant polymers suggesting that only certain types of MIPs can be made using microwave irradiation. This work also indicates that the controlled polymerisation conditions offered by microwave reactors could open a promising future in the development of MIPs with more predictable analyte recognition performance, assuming material selection lends itself to this type of initiation