Selective solid phase extraction of lanthanides from tap and river waters with ion imprinted polymers.

For the first time, an ion imprinted polymer (IIP) able to selectively extract simultaneously all the lanthanide ions was successfully synthesized in acetonitrile using Nd3+ as a template ion, methacrylic acid as a complexing monomer, and ethylene glycol dimethacrylate as a cross-linker. A non-imprinted polymer (NIP) was synthesized under the same conditions as those of the IIP, but in the absence of the template ion. After the removal of the template ions, grounding and sieving, the IIP particles were packed in solid phase extraction (SPE) cartridges. The selectivity of the IIP was evaluated by comparing its behavior with the one of the NIP. Each SPE step (percolation, washing, and elution) was optimized in order to find the best compromise between the selectivity and the extraction recoveries. Using the optimized SPE conditions, the extraction recoveries of eight lanthanide ions representative of the lanthanide family were higher than 77% with an average value of 83% with the IIP, whereas, in the case of the NIP, they ranged between 14 and 36% and they were below 3% for the interfering ions from alkali, transition, and post-transition metal families with the IIP. A first evaluation of the reproducibility of the SPE profiles was carried out by performing statistical tests on the data obtained with several cartridges filled with particles obtained from two different IIP and NIP syntheses. Promising results were obtained. The specific capacity, i. e. the adsorption capacity of Nd3+ ions by the specific cavities of the imprinted polymer, was about 9 mg of Nd3+ per gram of IIP (60 µmol g-1), which is more than enough for the extraction of the lanthanide ions at trace levels. The breakthrough volume was about 1 mL per mg of IIP, leading to an enrichment factor of 15, which allows not only to selectively extract the lanthanides but also to concentrate them. Finally, the imprinted polymer was successfully used to selectively extract lanthanides from tap and river waters spiked at 1 µg L-1.

Potential of ion imprinted polymers synthesized by trapping approach for selective solid phase extraction of lanthanides.

Ion imprinted polymers (IIPs) specific to lanthanides were synthesized using neodymium ions (Nd3+) as template ions. Nd3+ ions form binary complex ions with 5,7-dichloroquinoline-8-ol (DCQ) or vinylpyridine (VP), or ternary complex ions with both DCQ and VP in 2-methoxyethanol, before copolymerization in the presence of styrene and divinylbenzene as monomer and cross-linker, respectively. DCQ was expected to be trapped in the synthesized polymers pores. The template ion removal was then optimized. For the first time, the DCQ leakage was determined by HPLC-UV during the template removal and the sedimentation steps before solid-phase extraction (SPE) packing. It was observed that the trapped DCQ was unfortunately lost in significant amounts, up to 51%, and that this amount varied from one synthesis to another. The grinded and sieved polymers were next packed in SPE cartridges. The study of the SPE profiles obtained with the IIPs synthesized either with the binary or the ternary complex confirmed the prominent role of DCQ on the selectivity of an IIP by comparison with a non-imprinted polymer (NIP), i.e. a polymer synthesized under the same conditions as those of the IIP but without template ions. The influence of the porogenic solvent on the selectivity was also investigated by replacing 2-methoxyethanol by acetonitrile or dimethylsulfoxyde (DMSO). The polymers synthesized in DMSO led to the most repeatable results when elution solutions with a gradual decrease in pH were percolated through the cartridge. This is why DMSO was used to optimize the SPE protocol in order to maximize the difference of extraction yield between the IIP and the NIP, i.e. promoting a selective retention on the IIP. A value of about 30% was obtained for La3+, Ce3+, Nd3+, and Sm3+. Nevertheless, with the optimized SPE protocol, IIPs from different syntheses did not have the same SPE behavior, which may result from different random leakages of DCQ. This demonstrates for the first time the main limitation of the IIPs synthesized in bulk with the trapping approach for their use in SPE.