The application of molecular imprinted polymers to the solid phase extraction of drugs

Molecular imprinted polymers (MIPs), prepared to propranolol, were evaluated as solid phase extraction (SPE) phases. Monolithic MIPs were converted to particulate SPE materials by crushing, sieving, and sedimentation to remove fines. A high proportion (85%) of template propranolol was recovered with initial washes of the MIPs. Template leaching diminished with successive washes though low level leachings till compromised analysis of propranolol at trace levels (<5ng/mL). It was possible to retain propranolol on the MIPs following application in aqueous solution, biological matrices and organic solvent. In reversed and normal phase mode it was necessary to include an acid or base to achieve elution of basic compounds. In reversed phase mode, propranolol-imprinted MPs showed a degree of selectivity when using methanol:water:triethylamine but not methanol:water:trifluoroacetic acid elution solvents. This selectivity was related to structure when structurally-diverse compounds were assessed. Selectivity was very limited when assessed using compounds structurally similar to propranolol. In normal phase mode, MIPs retained amino-alcohols more strongly than a non-imprinted polymer though there was no evidence of a structure-related selectivity. A non-imprinted polymer and one prepared to an alternative template (tamoxifen) were assessed as reference polymers by which to assess the selectivity of propranolol-imprinted MIPs. Both materials were suitable references in reversed-phase mode though little selectivity was evident. The non-imprinted polymer was a poor comparator in normal phase mode as it appeared to show lesser non-specific binding than the imprinted MIP. Reversed and normal phase SPE methods were successfully developed for compounds structurally-simila to propranolol from human plasma. A protein precipitation step ahead of SPE was necessary. The MEP-based asays showed no advantage over conventional SPE and solvent extraction assay approaches. Although MIP-based extracts were clean, accuracy was poor at low concentrations (2 and 5 ng/mL) due to interference from template impurities leaching from the MJP.