Development of ion-imprinted polymers for the selective extraction of Cu(II) ions in environmental waters

Highlights

• Benefit of extending the complexation time between template ion and ligands.

• First study to date on the effect of acid treatment on NIP.

• Fine optimization of the washing step of the SPE protocol.

• Remarkable specificity towards Cu(II) with respect to interfering ions.

• Determination of Cu(II) at trace levels in environmental waters.

Abstract

Several ion-imprinted polymers (IIPs) were synthesized via bulk polymerization with Cu(II) as template ion, methacrylic acid as functional monomer, ethylene glycol dimethacrylate as crosslinking agent, and azobisisobutyronitrile as initiator in acetonitrile or methanol as porogen solvent. Non-imprinted polymers (NIPs) were similarly synthesized but without Cu(II). After grounding and sieving, the template ions were removed from IIPs particles through several cycles of elimination in 3 M HCl. All NIPs were equally subjected to this acid treatment with the exception of one NIP, called unwashed NIP. The resulting IIP/NIP particles were packed in solid phase extraction (SPE) cartridges for characterization. The SPE protocol was designed by optimizing a washing step following the sample percolation to eliminate potential interfering ions prior to the elution of Cu(II), all fractions analyzed by inductively coupled plasma mass spectrometry. The best IIP showed a high specificity (recovery of Cu(II) vs. interfering ions) and a good selectivity (retention on IIP vs. NIP). Its adsorption capacity was determined to be 63 μg g⁻¹. Then, a volume of 50 mL was percolated with 30 mg of IIP, thus giving rise to an enrichment factor of 24. Finally, applications to real samples (mineral and sea waters) were successfully performed. In addition, Brunauer-Emmett-Teller analyses showed that the surface area of the washed NIP was almost double that of the unwashed one (140.70 vs. 74.49 m² g⁻¹), demonstrating for the first time that the post-treatment of a NIP after its synthesis may have a significant impact on its porous structure, and thus need to be more precisely detailed by authors in the future papers.

Introduction

Copper, present in drinking, fresh and sea water, has an impact not only on the human health [1,2] but also on marine ecosystems [3]. Currently, the determination of copper at trace levels can be carried out by various techniques, whereas the direct determination in real samples suffers from strong matrix interferences, especially with the most highly sensible techniques such as inductively coupled plasma mass spectrometry (ICP-MS) and electrothermal atomic absorption spectrometry [4]. For this reason, a sample treatment prior to the determination of Cu(II) is frequently required for the removal of matrix components or the enrichment of target analytes [5], and the most widely used technique is solid phase extraction (SPE). Its performance largely depends on the nature of the sorbent, whereas conventional sorbents like ion-exchange resins with or without the functionalization with metal-complexing agents lack selectivity. To address this drawback, ion-imprinted polymers (IIPs) which possess selective cavities for ions were first introduced by Nishide et al., in 1976 [6] and are of continuing interest up to now.

Synthesis of IIPs starts with the complexation between template ions and one or several appropriate functional monomers. After the copolymerization with a crosslinking agent, the resultant complexes are then immobilized in a highly cross-linked polymer matrix. Finally, template ions must be removed from the polymer so as to leave tailored cavities that are complementary to target ions in size and coordination geometries. This step is commonly performed using a strong acid from 0.1 M to 6.0 M as a leacher, such as HCl [[7], [8], [9]], HNO₃ [[10], [11], [12]] or H₂SO₄ [[13], [14], [15]], in order to disrupt the specific interactions between template ions and binding cavities. In order to evaluate the imprinting effect of IIP, a non-imprinted polymer (NIP) is synthesized as a control polymer by using the same conditions but without template ions. However, one point remaining unclear so far is whether the NIP should be subjected to the same treatment used to remove the template ions from the IIP, as no template ion was introduced for the NIP synthesis. To our knowledge, studies devoted to the influence of post-synthesis treatment of NIP have not yet been reported in literature.

After synthesis, IIPs can be used in SPE with cartridge or in batch (so-called dispersive SPE, dSPE). dSPE is based on an adsorption equilibrium between sorbent particles and analytes under stirring for a sufficiently long incubation time. In practice, a follow-up operation of filtration, centrifugation, or decantation is necessary to recover the subsequent solution for its analysis, which is time-consuming and labor-intensive. By contrast, SPE with cartridge offers many advantages since particles are immobilized allowing the direct and rapid percolation of samples and recovery of solutions. Thanks to that convenience and simplicity, one or several washing steps can be applied to eliminate potential interfering ions before the final elution of target ions.

Nowadays many studies have been reported on the preparation of IIPs dedicated to Cu(II) [16]. However, almost all of those IIPs were used in dSPE rather than SPE. To the best of our knowledge, there are only five papers on Cu(II)-targeting IIPs in SPE including selectivity study [[17], [18], [19], [20], [21]] and two of them did not validate their analytical methods with real samples [18,19]. Moreover, none of them implemented an optimized washing step, to eliminate interfering ions and thus optimizing the specificity, which is considered as one of the fundamental advantages of SPE as described previously.

In this work, Cu(II), methacrylic acid (MAA), ethylene glycol dimethylacrylate (EGDMA), and azo-N, N′-diisobutyronitrile (AIBN) were selected as template ion, ligand, crosslinking agent, and initiator, respectively. Methanol and acetonitrile (ACN) were tested as porogen solvent. Several IIPs were synthesized via bulk polymerization by varying the nature and volume of porogen solvent as well as the time of complexation between Cu(II) and MAA. After grounding and sieving, the template ions were removed by washing IIP particles with 3 M HCl. All related NIPs were subjected to the same acid treatment for being rigorous references, while one NIP remained unwashed for investigating the effect of acid treatment. The resulting IIPs/NIPs were characterized in SPE cartridges by involving a washing step following the sample percolation prior to the elution of Cu(II) ions. Using ICP-MS for the analysis of the different resulting SPE fractions, each SPE step was optimized for promoting selectivity (retention on IIP vs. NIP) and specificity (recovery of target ion vs. interfering ions). The most promising IIP was thereupon selected. Its capacity, breakthrough volume, and enrichment factor were determined under the optimum SPE conditions. Finally, applicability to real samples (i.e. mineral and sea waters) as well as reusability were assessed.