Which molecularly imprinted polymer is better?

Molecularly imprinted polymers (MIP) have been successfully synthesized toward many different compounds in the last decades. The mechanistic details of selective binding at binding sites are not yet well understood. For this reason the characterization of MIP binding has been mostly phenomenological and this makes the transfer of results between different laboratories or between different types of applications difficult. In this paper we analyze the relationship between different types of characterization like isotherms, binding site models, chromatographic k and a values, etc. as they relate to different applications like HPLC, solid phase extraction (SPE), binding assays, batch extraction and sensors. It is shown that a values determined by elution chromatography depend on seemingly irrelevant factors as the length and diameter of the column, respectively. The determination of distribution ratios or partition coefficients is proposed as an easily understandable and useful quantity in the characterization of novel MIPs. Data used for the characterization of a MIP should be transferable between different applications but the qualification of MIPs as better or worse will depend on the application in case. (C) 2007 Elsevier B.V. All rights reserved.