THE FINAL FRONTIERS IN MODELING AND DESIGNING MOLECULARLY IMPRINTED POLYMERS

Despite ongoing experimental and theoretical progress, molecular imprinting remains a field of controversy in materials science in particular related to the question whether templating a polymer matrix is based on statistical effects, or whether this process may be considered the deliberately controlled entailment of specific binding sites for a particular template molecule. While a wealth of experimental papers since the seminal works of Wulff and Mosbach have been published indicating more - or sometimes less - selectivity, an absolute proof for this open question has not been shown to date.

In this presentation, a breakthrough approach will be shown to potentially answer this pertinent question once and forever. Teaming up with one of the leading groups on large-scale molecular dynamics (MD) simulations of complex systems, we have developed a simulation concept that for the first time provides an MD strategy for generating ’virtually imprinted polymers’ (VIPs), testing such VIPs within a second set of MD simulations enabling ’virtual chromatography experiments’, and – last but not least - applying local density of states (LDOS) calculations for the first time in classifying and selecting suitable functional monomers.

Within these truly fundamental studies, we introduce the first theoretical demonstration for molecular templating processes in atomistic detail for the example of virtually imprinted polymers for 17-β-estradiol, and demonstrate selective binding via virtual chromatography experiments even vs. its enantiomer, which is in accordance with previously obtained experimental results. The obtained simulation results unambiguously prove the existence of a distinct imprinting effect similar to the concepts commonly described in the MIP literature. Achieving selective retainment in a virtual imprinting/chromatography experiment and even generating virtual chromatograms clearly disproves the hypothesis that molecular imprinting is a statistical effect. Furthermore, we visualize for the first time that the presence of the target molecule during the imprinting procedure is mandatory, as the spatial structure of the selective binding sites determined by the arrangement of the functional monomers is crucial for entailing selective binding properties. In addition, a comprehensive strategy towards rationally screening and selecting suitable functional monomers from a library based on LDOS calculations complements this study.

In conclusion, the fundamental concepts discussed in this presentation provide a solid step towards truly predictive modeling of molecular imprinting strategies that could be expanded in future to almost any level of complexity.