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

Analytica Chimica Acta 963 (2017) 44e52

Contents lists available at ScienceDirect

Analytica Chimica Acta journal homepage: www.elsevier.com/locate/aca

Selective solid phase extraction of lanthanides from tap and river waters with ion imprinted polymers
Ndiaye a, Thomas Pinta a, Vale
rie Pichon a, b, Manel Moussa a, b, Massamba Mbacke Thomas Vercouter c, Nathalie Delaunay a, d, * a

Laboratory of Analytical, Bioanalytical Sciences and Miniaturization (LSABM), UMR CBI 8231, ESPCI Paris, PSL Research University, Paris, France UPMC, Sorbonne Universit
es, Paris, France CEA Saclay, DEN, DANS, DPC, SEARS, LANIE, Gif-sur-Yvette, France d CNRS, UMR CBI 8231, Paris, France b c

h i g h l i g h t s

g r a p h i c a l a b s t r a c t


An Ion Imprinted polymer (IIP) was synthesized for the selective SPE of lanthanides.
After optimizing the SPE procedure, IIP has a high selectivity for all lanthanides.
A specific capacity of 60 mmol g1 and an enrichment factor of 15 were obtained.
Selective extractions with tap and river water were successfully performed.

a r t i c l e i n f o

a b s t r a c t

Article history: Received 28 October 2016 Received in revised form 7 February 2017 Accepted 9 February 2017 Available online 24 February 2017

For the first time, an ion imprinted polymer (IIP) able to selectively extract simultaneously all the lanthanide ions was successfully synthesized in acetonitrile using Nd3þ 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 Nd3þ ions by the specific cavities of the imprinted polymer, was about 9 mg of Nd3þ per gram of IIP (60 mmol g1), 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,

Keywords: Ion imprinted polymer Lanthanides Solid-phase extraction (SPE) Tap water River water

Abbreviations: AIBN, 2,20 -Azobis(2-methylpropionitrile); DAP, 2,6-diacetylpyridine; Bis-Tris, 2,2-Bis(hydroxymethyl)-2,20 ,200 -nitrilotriethanol; EDTA, ethylenediaminetetraacetic acid; EGDMA, ethylene glycol dimethacrylate; HREE, heavy rare earth element; IIP, ion imprinted polymer; Ln, lanthanide; LREE, light rare earth element; MAA, methacrylic acid; NIP, non-imprinted polymer; PAN, 1-(2-pyridylazo)-2-naphthol; REE, rare earth element. * Corresponding author. Laboratory of Analytical, Bioanalytical Sciences and Miniaturization (LSABM), UMR CBI 8231, ESPCI Paris, 10 rue Vauquelin, 75231 Paris cedex 05, France. E-mail address: [email protected] (N. Delaunay). http://dx.doi.org/10.1016/j.aca.2017.02.012 0003-2670/© 2017 Elsevier B.V. All rights reserved.

M. Moussa et al. / Analytica Chimica Acta 963 (2017) 44e52

45

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 mg L1. © 2017 Elsevier B.V. All rights reserved.

1. Introduction The lanthanides (Ln) comprise 15 elements, whose atomic numbers range from 57 to 71. Together with Y and Sc, which have similar properties, they form the rare earth element (REE) group. REEs are often divided into “Heavy Rare Earth Elements” (HREEs) and “Light Rare Earth Elements” (LREEs). Based on electronic configurations, LREEs are La, Ce, Pr, Nd, Sm, Eu, and Gd, and HREEs include Y, Tb, Dy, Ho, Er, Tm, Yb, and Lu [1]. With some exceptions, the lanthanides are commonly found in the trivalent oxidation state Ln3þ. Their 4f electronic shell leading to interesting optical and magnetic properties, the lanthanides have seen an increased use in both modern and green technologies, with the consequence of an increased human exposure. Indeed, the anthropogenic presence of lanthanides has recently been reported in natural waters and even in tap water [2e4]. The analysis of the REEs is thus important to evaluate their presence in our environment. Inductively coupled plasma-mass spectrometry (ICP-MS) has become the most powerful technique used for the determination of inorganic elements owing to its high sensitivity and multi-element capability. However, ICP-MS suffers from spectral and non-spectral interferences in addition to the clogging of the sample introduction system when the percentage of dissolved salts is greater than 0.2% (w/v) [5]. To avoid matrix effects, many sample pretreatment techniques are applied before the ICPMS determination of the REEs in environmental samples, such as liquid-liquid extraction (LLE) [6,7], solid phase extraction (SPE) [6e8], and co-precipitation [9e12]. Currently, SPE is the most used sample preparation method due to the large availability of sorbents, the use of small volumes of organic solvents as compared to LLE, its simple and rapid implementation and the ability to be automated and combined on-line with analytical techniques [13e15]. Many SPE sorbents have been used for the matrix removal and/or preconcentration of REEs, such as an octadecyl silica modified with 1(2-pyridylazo)-2-naphthol (PAN) [16], Amberlite XAD-4 functionalized with 2,6-diacetylpyridine (DAP) [17], a polyaminopolycarboxylic acid group-based resin [18], a polyhydroxamic acid-based resin [19], an alkylphosphonic acid based resin [20] or a monolith modified with iminodiacetic acid [21]. However, these ligands are not specific to REEs. For example, a C18 silica cartridge loaded with PAN was used to extract Cu, Fe, Ni, Mn, Pb, and Zn from environmental water samples [22], an iminodiacetic acid chelating resin was used to extract In [23], and DAPfunctionalized Amberlite XAD-4 was used to preconcentrate Cd, Co, Cu, Mn, Ni, Pb, U, and Zn [24]. The ion imprinted polymers (IIPs) represent a new class of sorbents based on imprinting technique, which results in specific recognition sites within a polymer matrix [25,26]. The principle of ion imprinting is based on the formation of a complex between a template ion and appropriate ligands, followed by polymerization in the presence of a cross-linker. Finally, the template ions are removed to strip the specific cavities. In order to evaluate the selectivity of the IIP, a non-imprinted polymer (NIP) is synthesized under the same conditions as those of the IIP, but in the absence of the template ion. According to Prasada Rao, there are four different methods to

synthesize an IIP: (1) cross-linking of linear chains of polymers which contain complexing groups, (2) chemical immobilization of a complex composed by polymerizable ligands and template ions in the presence of a cross-linker, (3) surface imprinting involving a ligand having or not a vinyl group with an aqueous-organic interface, and (4) trapping of a non-vinylated ligand complexing the template ion inside the polymer network [25]. Although the three latter methods were used to synthesize IIPs specific to lanthanides (Ln-IIP) [27e43], the trapping approach was the most extensively used [27e35,43]. Nevertheless, for the first time, we have recently shown that a loss of the “trapped” non-vinylated ligand occurs and that this loss is not repeatable from one IIP-based SPE cartridge to another [44]. This is why the trapping approach has currently a strong limitation for SPE applications. Therefore, the chemical immobilization approach seems to be an alternative. It was already used to synthesize IIPs with Gd3þ [36e38,42] and Nd3þ [39]. Acrylic acid is however the only commercially available monomer that has been used [37,42]. In the other cases, the ligands (ethylenediaminetetraacetic acid (EDTA) and diethylenetriaminepentaacetic acid [36e38,42] or folic acid [39]) were functionalized with vinyl moieties by their users, which considerably increases the cost of the synthesis. Ln-IIPs were all characterized by a batch process [27e43], which is based on mixing an amount of solid sorbents with a solution containing the tested ions and stirring the solution and the sorbent until the adsorption equilibrium is reached. The adsorbed fraction is then determined either by analysis of the supernatant or after separation of the sorbent from the solution by filtration, centrifugation or decantation followed by desorption of the ions with a suitable solution which is then analyzed. Thus, the batch process is based on equilibrium conditions and is therefore time-consuming and labor-intensive; they are also risks of contamination and loss of sorbent when different steps are involved. Moreover, a washing step to suppress the interfering ions retained by non-specific interactions is almost never implemented. If it is, this step further increases the risk of contamination and polymer loss and the sample pretreatment time. IIP particles packed in a cartridge associated with a SPE protocol involving a washing step definitely constitutes a better alternative to batch techniques for routine analysis. It improves selectivity and offers a lower consumption of sample and reagents, faster exchange kinetics, higher reproducibility, no risk of loss of the sorbent and analytes, reduced contamination, possible automation and on-line coupling with an analytical step, and higher throughput. In the case of the Ln-IIPs, only two papers described the implementation of the imprinted particles in a SPE precolumn; however, the characterization of the IIP properties and the optimization of the adsorption and elution media were carried out in batch [35,43]. In both cases, no application with real samples was performed, explaining may be why no washing step was required. In this paper, for the first time, an IIP dedicated to lanthanide ions was synthesized with the chemical approach with the commercially available methacrylic acid (MAA) monomer. It was demonstrated that this IIP was able to selectively extract all the lanthanide ions since they can be simultaneously present in real environmental samples. Indeed, until now, only one given

46

M. Moussa et al. / Analytica Chimica Acta 963 (2017) 44e52

lanthanide ion was targeted either from a lanthanide mixture or from a mixture containing interfering cations from other chemical families. The IIP was synthesized with Nd3þ, MAA, ethylene glycol dimethacrylate (EGDMA), and 2,20 -Azobis(2-methylpropionitrile) (AIBN), as template ion, functional monomer, cross-linker, and initiator, respectively. The IIP particles were packed in cartridges and characterized as SPE sorbent for the first time. The SPE protocol involving the three usual steps was optimized in order to promote selective extractions on the IIP, which implies the lowest possible extraction recoveries with the NIP. Lanthanide quantification following extraction was carried out by ICP-MS. A first preliminary study about the repeatability of the syntheses of the Ln-IIP-based SPE cartridges was investigated. Indeed, even if this investigation is necessary, this kind of results have never been published until now with Ln-IIPs. Finally, for the first time, SPE applications with real samples (tap and river waters) were carried out with IIP dedicated to lanthanide ions.

conditions but without template ion. After polymerization of the IIPs and the NIPs, the vials were crushed and both kinds of polymers were grounded and stirred for 20 h in 3 M HCl, which allows the removal of the template ions from the IIPs. Then, the polymer particles were filtrated under vacuum using a Millipore glass filter holder assembly and cellulose filter papers with pore size of 2.5 mm (Whatman, Sigma-Aldrich). The particles were washed with water until the filtrate collected in the flask was neutral (followed by pH paper). The particles were then dried in an oven at 60  C, ground with a ball mill, and manually sieved. The 25e36 mm particle fraction was collected, sedimented in MeOH/water (8/2 (v/v)) (3 times for 30 min) to eliminate the residual fine particles, and then dried. 30 mg of each polymer were packed between two polyethylene frits (Sigma Aldrich) into polypropylene 1-mL cartridges (Interchim).

2. Material and methods

The analysis of the lanthanides was performed using an Agilent 7700x ICP-MS system with the following parameters: RF power, 1550 W; sampling depth, 8 mm; plasma gas flow rate, 15 L min1; auxiliary gas flow rate, 0.9 L min1; carrier gas flow rate, 0.99 L min1. The measurements were carried out with five replicates with 100 sweeps per replicate with an integration time of 1 s. The following isotopes of the lanthanides were monitored: 139La, 140 Ce, 146Nd,147Sm,157Gd,163Dy,166Er, and 175Lu. The oxide and the double-charged ion formation was followed by monitoring the CeOþ/Ceþ and the Ce2þ/Ceþ ratio couple during the daily performance check. Measurements were executed only when these two ratios were both less than 2%. 59Co, 7Li, 89Y, and 205Tl were analyzed with the same conditions.

2.1. Reagents HPLC-grade acetonitrile was purchased from Carlo Erba Reactifs-SDS (Val De Reuil, France). Neodymium (III) nitrate hexahydrate (99.9%), HCl (32%), MgSO4 (>98%), NaCl (99%), EGDMA (98%), and MAA (99%) were from Sigma-Aldrich (St Quentin Fallavier, France). AVS Titrinorm NaOH 1 M, ultra-pure grade 2,2Bis(hydroxymethyl)-2,20 ,200 -nitrilotriethanol (Bis-Tris), EDTA disodium salt, and nitric acid (65%) for analysis Emsure® were purchased from VWR (Strasbourg, France). AIBN was purchased from Acros Organics (Noisy-le-Grand, France). Ultra-pure water was obtained using a Milli-Q purification system (Millipore, Molsheim, France). 200 mL of EGDMA were washed twice with 100 mL of 10% (v/v) NaOH and then washed twice with 100 mL of water. EGDMA was dried with 100 mL of saturated NaCl solution. Residual water was finally removed with MgSO4 (>98%). Washed EGDMA and MAA were distilled under argon and stored at 4  C. Individual SPEX stock solutions of La, Ce, Nd, Sm, Gd, Dy, Er, and Lu in 2% (v/v) HNO3 at 1 g L1 each were provided by SCP SCIENCE (Courtaboeuf, France). A mixture of the 8 lanthanides at a concentration of 30 mg L1 each was prepared in 2% (v/v) HNO3. Calibration solutions for ICP-MS with concentrations from 0.01 to 1 mg L1 were prepared by diluting this mixture in 2% (v/v) HNO3. A tuning solution for ICP-MS from Agilent Technologies (Les Ulis, France) containing Ce, Co, Li, Mg, Tl, and Y at 1 mg L1 in 2% (v/v) HNO3. was also used.

2.3. Analysis of the lanthanides

2.4. SPE procedure The IIPs and their NIPs were conditioned with 3 mL of Bis-Tris buffer pH 6.5 (220 mM). 1 mL of the same buffer containing some tens of ng of one or several lanthanide ions was then percolated. The washing step consisted in 5 mL of deionized water, followed by 2.5 mL of HCl at pH 3.8 (corresponding to 0.16 mM). Finally, the elution was done with 3 mL of 1 M HCl. After each use, the cartridges were washed with water until neutralization before being stored in water. Each SPE fraction was diluted before ICP-MS analysis using 2% HNO3 in order to have a final lanthanide concentration that fit the concentration range used for calibration. The recovery of each SPE fraction was determined by comparing the amount of a given lanthanide present in the fraction to the amount of this lanthanide initially percolated.

2.2. Synthesis of the IIPs 2.5. Real sample pretreatment The IIPs were synthesized with a template/monomer/crosslinker molar ratio of 1/5/25 in acetonitrile. The ratio 1/5 of the template ion and the ligand was selected to mimic the stochiometry of the complex used commonly for the synthesis of an Ln3þ-IIP with the trapping approach, involving 1 lanthanide ion, 3 molecules of 5,7-dichloroquinoline-8-ol, and 2 molecules of 4vinylpyridine, i. e. 5 molecules of ligand for 1 ion [27e35]. The classically used ratio 1/5 of the ligand and the cross-linker was selected. First, neodymium(III) nitrate hexahydrate and MAA were dissolved in acetonitrile. The complex formation was carried out under magnetic stirring during 2 h. Afterwards, EGDMA and AIBN (1% of AIBN moles with respect to the total number of moles of polymerizable double bonds) were added. The polymerization mixture was placed in an ice bath and was degassed with nitrogen for 10 min. The vial was next sealed and heated at 60  C during 24 h under magnetic stirring. The NIP was synthesized under the same

The tap water was collected in the laboratory and the river water was from the Seine River (Paris, France). They were collected in polyethylene terephthalate bottles and filtered with a 0.45 mm polytetrafluoroethylene filter (Sigma-Aldrich). Before percolation, the Bis-Tris salt was directly dissolved in each water sample at a concentration of 220 mM. The pH was then adjusted to 6.5 with 2 M HCl. The samples were next spiked with La, Ce, Nd, Sm, Er, Gd, Dy, Er, and Lu at 1 mg L1 each. 3. Results and discussion 3.1. Optimization of a selective SPE protocol In order to have a selective SPE protocol, each SPE step has to be optimized in order to maximize the difference in extraction

M. Moussa et al. / Analytica Chimica Acta 963 (2017) 44e52

recovery between the IIP and its NIP. Indeed, the target analytes can be retained on the IIP in the specific cavities but also by non-specific interactions with the surface. A comparison with the NIP where the retention is only due to non-specific interactions is thus necessary to evaluate the contribution of the non-specific interactions to the retention of the lanthanides on the IIP.

3.1.1. Optimization of the percolation step The key parameters of the percolation step are the pH, the nature, and the volume of the percolated solution. As the target samples are different kinds of waters, an aqueous media was selected. As the breakthrough volume and the capacity of the IIP were unknown at the beginning of this study, the pH of the percolation solution was optimized for a small percolated volume (1 mL) and a low amount of Ln3þ ions (50 ng). Sm3þ ions were selected as test ions to prevent from quantitation error if leaching of some residual template ions, Nd3þ, occurred. The studied pH range was 5.5e7.5 in order to favor a high ionization rate of the acidic moieties present in the IIP cavities and consequently a high retention of the Ln3þ ions. Indeed, if the pKa of the MAA monomer is 4.65 at 25  C, the pKa of polymeric acids are different from their monomer due to the additional bonds formed between the monomer units, the additional Van der Waals and electrostatic interactions, and the potential hydrogen bonds formed among the functional groups [45]. Qian et al. found that the pKa values of the dimer and the trimer of MAA are higher by about 0.8 units. A BisTris buffer was selected since it is a cationic buffer to limit its potential complexation ability with Ln3þ ions and it has a strong buffer effect in the studied pH range (pKa value of 6.5). The ionic strength of the Bis-Tris buffer was fixed at 20 mM, which corresponds to a total concentration of 220 mM at pH 7.5. The percolated solutions at pH 6.5, 6.0, and 5.5 were obtained by acidifying the solution of Bis-Tris pH 7.5 (220 mM) with 3 M HCl. Fig. 1 presents the extraction recoveries determined after elution with 3 mL of 3 M HCl with the IIP and its NIP prepared by a first synthesis (S1) according to the pH of the percolation solution. At pH 6.5, 7.0, and 7.5, about 100% of the Sm3þ ions were retained and next eluted from the IIP and its NIP. At pH 6, the extraction recovery with the IIP decreased to 93% and then to 62% at pH 5.5. This lack of retention may stem from a low ionization rate of the acidic moieties present in the IIP cavities or at the polymer surface.

Fig. 1. Effect of the pH of the percolated solution (Bis-Tris 220 mM) on the extraction recovery of 50 ng of Sm3þ on the IIP and its NIP obtained with a first synthesis (S1) (n ¼ 3). Percolation, 1 mL of 220 mM Bis-Tris adjusted at indicated pH. Elution, 3 mL of 3 M HCl.

47

A pH value of 6.5 was chosen in further experiments to reach a good compromise between the extraction recoveries and the risk of precipitation of the lanthanide ions. 3.1.2. Optimization of the elution step The elution step was optimized before the washing step because all the IIP-based SPE cartridges were reused when optimizing the SPE process, which means that it is necessary to determine how to desorb the retained compounds to clean the cartridge before the next use. The elution can be performed either using mineral acids, complexing agents or a mixture of both [25,26]. In this work, the effect of the volume of 1 M HCl on the elution of Sm3þ ions fixed after their percolation in 1 mL of Bis-Tris buffer at pH 6.5 was investigated. With 1, 2, and 3 mL of 1 M HCl, the extraction recoveries were 89, 99, and 100%, respectively. Indeed, at this very low pH, the acidic moieties present in the IIP cavities or at the polymer surface are fully protonated and thus neutral, which should reduce the interactions with Sm3þ ions. Moreover, there is also a very high concentration of Hþ that may contribute to release the retained Sm3þ ions. To ensure a complete elution, a volume of 3 mL of 1 M HCl was selected. A high concentration of a good complexing agent of lanthanide ions such as EDTA should also enable the elution of the Sm3þ ions by competition with the complexation phenomenon in the cavities. A 0.1 M EDTA solution at pH 10 was tested. With a volume of 1 mL, 80% of the Sm3þ ions were eluted and 2 mL allowed a quantitative desorption. However, it was observed that the rinsing of the EDTA solution from the SPE cartridges was a tedious task, which required a large volume of water (30 mL). Nevertheless, the total removal of EDTA was crucial to ensure retention during the following SPE cycle. For this reason, 3 mL of 1 M HCl were preferred as elution conditions. 3.1.3. Optimization of the washing step It can be seen on Fig. 1 that with an SPE protocol without washing step, there is no difference between the extraction recoveries of Sm3þ ions on the IIP and its NIP where only non-specific interactions can occur. A washing step is thus mandatory to eliminate the non-specific interactions. A good washing solution has to be strong enough to disrupt the non-specific interactions, but without eluting the analytes retained in the specific cavities of the IIP. Therefore, an optimized washing step should give a maximal difference between the extraction recoveries of the targeted ions on the IIP and its NIP. In the literature, when a washing step was implemented in seldom cases, the authors used the same solution that the one used for percolation [46e49] or water [50e55]. In this work, a washing with water was first carried out, just after the percolation of the Sm3þ ions in Bis-Tris at pH 6.5. The pH of the water at the output of the cartridges was followed with pH paper and 5 mL were necessary to reach a neutral pH. The water fractions were analyzed by ICP-MS and no Sm3þ ion was detected. A washing protocol using successive washing solutions at pH 10 containing an increasing concentration of EDTA from 0.5 to 10 mM was next evaluated with the IIP and its NIP (Fig. S1 in Supplementary Data). There was no difference between the resulting SPE profiles of the IIP and its NIP. This may be explained by the fact that the affinity of Sm3þ for EDTA is too high. An adjustment of the pH of the washing solution was next investigated. Indeed, the pH affects the ionization state of the acidic moieties present in the cavities, that should give numerous and strong interactions with several well-positioned complexing groups, but also at the surface of the polymers, the ones leading to the non-specific interactions that are expected to be weaker. 1 mL of HCl solution at pH 4.5, 4, 3.8 or 3.5 (corresponding to 0.03, 0.1,

48

M. Moussa et al. / Analytica Chimica Acta 963 (2017) 44e52

0.16, and 0.32 mM) was passed through the IIP and NIP cartridges (Fig. 2A). As it appears on Fig. 2A, with 1 mL of HCl at pH 4 or 4.5, more than 96% of Sm3þ ions were recovered in the elution step and almost no difference was observed between the IIP and its NIP. However, using 1 mL of HCl at pH 3.8, the extraction recovery of Nd was 97% for the IIP but only 82% for the NIP. By decreasing again the pH to 3.5, the difference between both supports increased from 15 to 22%, but the extraction recovery of Sm3þ ions on the IIP was below 70%. The best compromise between selectivity and extraction recovery was thus a pH of 3.8. The volume of HCl at pH 3.8 was next optimized (Fig. 2B). It was observed that 2.5 mL was the optimal washing volume ensuring a difference of about 65% between the extraction recoveries of the Sm3þ ions obtained with the IIP and its NIP. 3.2. Selectivity and specificity study The optimized SPE protocol led to a high selectivity of the IIP for the Sm3þ ions as the average extraction recoveries were (23 ± 4)% with the NIP and (87 ± 4)% with the IIP (n ¼ 3). The selectivity for other lanthanides was next investigated. At this point of the study, no Nd3þ ions were detected in the SPE fractions, so that the study of the behavior of Nd3þ ions on the synthesized polymers with the

optimized SPE procedure became possible. Thus, La3þ, Ce3þ, Nd3þ, and Sm3þ were selected as LREEs and Gd3þ, Dy3þ, Er3þ and Lu3þ as HREE. As can be seen in Fig. 3A, all lanthanides were selectively extracted with recoveries higher than 77% in the elution fraction with the IIP whereas, in the case of the NIP, recoveries ranged between 14% and 36% for La3þ and Lu3þ, respectively. Relative standard deviations were below 5% (n ¼ 3) for the extraction recoveries observed with the IIP. It is interesting to note that the extraction recoveries of Sm3þ and Nd3þ were (87 ± 4)% and (83 ± 2)% respectively with the IIP, which confirms that the Sm3þ ion was a good analogue of the Nd3þ ion. It can also be noticed that the kinetics was rapid for all the lanthanide ions, as the percolation flow rate was about 0.4 mL min1. The specificity of the IIP for the lanthanide ions was studied. Co2þ, Liþ, and Tlþ were selected because they belong to the transition metal, alkali metal, and post-transition metal groups, respectively. Y3þ was also tested since it belongs to HREEs [1]. Fig. 4 presents the resulting SPE profiles. The extraction recoveries of Liþ, Tlþ, and Co2þ were close to 0% with both the IIP and the NIP, whereas the extraction recoveries of Y3þ were similar to those of Ce3þ and Nd3þ for both the IIP and the NIP, which demonstrates the high specificity of the IIP for the lanthanide ions plus yttrium, i. e. for the RRE family. This should allow the simultaneous extraction of all the REEs from a given sample. The repeatability of the synthesis of the MIPs was already demonstrated [56e58]. However, in the case of IIPs, the same study had yet to be carried out. A second synthesis of polymers (IIP S2 and NIP S2) was performed on another day by a different operator. The optimized SPE procedure was next applied (Fig. 3B). The repeatability of the extraction recoveries with the IIP S2 and its NIP is good since the relative standard deviations were lower than 10%. If we compare Fig. 3A and B, the extraction recoveries obtained on the two IIPs resulting from different syntheses were very similar. An ANalysis Of VAriance (ANOVA) was performed to investigate more carefully the repeatability of the synthesis [59]. This test showed that for almost all lanthanide ions, i. e. 7 out of 8, the repeatability of the IIP synthesis was excellent. Indeed, the effect of the synthesis was significant only in the case of Lu3þ where the p-value was 0.03, which is inferior to 0.05. This constitutes the first promising results about the repeatability of the synthesis of IIPs as SPE sorbents. 3.3. Study of the capacity

Fig. 2. Optimization of (A) the pH and (B) the volume of the HCl washing solution (n ¼ 3). Percolation of 50 ng of Sm3þ ions in 1 mL of Bis-Tris buffer (220 mM) pH 6.5 on the IIP and its NIP. Elution with 3 mL of 1 M HCl.

The evaluation of the bounding capacity of the IIP is important since it represents the maximum amount of the analytes that can be retained. Nevertheless, the IIP S1 was not only used for the optimization of the SPE procedure and the study of the selectivity but also for many other SPE cycles. A decrease in the recoveries and in the capacity was observed after around 100 cycles, which shows a very slow degradation of the support with time. This degradation may stem from the use of highly acidic solutions during the elimination of the template ion and the elution steps. Consequently the determinations of the IIP capacity and, next, of the breakthrough volumes were performed with the polymers S2 during their initial uses. The capacity was determined by measuring the extraction recoveries on the IIP and its NIP of increasing amounts of Nd3þ percolated in 1 mL of Bis-Tris buffer at pH 6.5. The optimized washing step was performed to promote a specific retention on the IIP. The curve obtained with the IIP shows two different parts (Fig. 5). The first one presents a linear shape and corresponds to low amounts of Nd3þ. Its slope corresponds to a constant extraction recovery of 90%. For higher percolated amounts, a second linear curve is observed with a lower slope that is close to the one obtained with the NIP, where only non-specific interactions occur.

M. Moussa et al. / Analytica Chimica Acta 963 (2017) 44e52

49

Fig. 3. SPE profiles of 8 Ln3þ on (A) IIP S1 and NIP S1 from a first synthesis and (B) IIP S2 and NIP S2 from a second synthesis (30 mg of sorbent) (n ¼ 3). Percolation: 8 Ln3þ (30 ng each) in 1 mL of Bis-Tris (220 mM) at pH 6.5. Washing: 2.5 mL of HCl pH 3.8. Elution: 3 mL of 1 M HCl. 1, La3þ; 2, Ce3þ; 3, Nd3þ; 4, Sm3þ; 5, Gd3þ; 6, Dy3þ; 7, Er3þ; 8, Lu3þ.

Fig. 4. SPE profiles on IIP S1 and NIP S1 (30 mg of sorbent). Percolation: Liþ, Co2þ, Tlþ, Y3þ, Nd3þand Ce3þ (30 ng each) in 1 mL of Bis-Tris (220 mM) pH 6.5. Washing: 2.5 mL of HCl pH 3.8. Elution: 3 mL of 1 M HCl (n ¼ 3).

These lower slope values are consistent with interaction sites of lower energy, as they are non specific. Therefore, it is possible to consider that the point of intersection of the two linear parts of the

IIP curve is the maximum amount of Nd3þ that is retained in the specific cavities, since the non-specific interactions are negligible at this point (about 3%). The capacity can be estimated at 270 mg of

50

M. Moussa et al. / Analytica Chimica Acta 963 (2017) 44e52

Fig. 5. Capacity curves obtained with the percolation of increasing amounts of Nd3þ on IIP S2 and NIP S2. The amounts of Nd3þ recovered in the elution fraction were found after applying the optimized extraction procedure (see Experimental section). A Zoom was done for 0.4, 0.6, and 1 mg of Nd3þ.

Nd3þ for 30 mg of sorbent, which corresponds to 8.8 mg g1 or 60 mmol g1. This capacity value seems to be lower than those reported with Ln-IIPs in the literature. But, many parameters control the capacity of an IIP such as the polymerization method [60] and the natures of the porogen [61], the monomer, and the cross-linker [29]. By taking into account only Ln-IIPs synthesized by chemical immobilization, the capacities were around 100e150 mmol g1 [38,39]. However if, when possible, we estimate the specific capacity by subtracting the capacity values obtained with the IIP from those obtained with its NIP, the value is rather equal to 10 mmol g1 [38]. Moreover, since no washing step was performed, this estimated capacity value is probably overstated. In our work, the performing of a key washing step combined with the comparison with the NIP behavior allowed us to think that the amount of specific cavities should not be overestimated. In any case, this value is more than sufficient to extract traces of lanthanide ions from environmental water samples.

3.4. Study of the breakthrough volume The study of the breakthrough volume was performed with 30 ng of La3þ, Ce3þ, Nd3þ, Sm3þ, Er3þ, Gd3þ, Dy3þ, Er3þ, and Lu3þ in different volumes (from 1 to 40 mL) of the percolation solution (Fig. 6). As can be seen, the extraction recoveries were higher than 70% for all the lanthanides when the percolation volume was 1, 10, 20, and 30 mL. However, with a volume of 40 mL, the extraction recoveries of La3þ, Ce3þ, and Nd3þ, i. e. 3 LREEs, were 58, 61, and 65%, respectively, whereas the extraction recoveries were still high for the other lanthanide ions. These results were unexpected since the IIP was synthesized with Nd3þ. However, the polymerization was achieved in acetonitrile, while the extraction was performed in Bis-Tris buffer. In those two media, the solvated radius of Nd3þ ions and the swelling of the polymeric matrix are likely to vary, which could explain this behavior. We consider that 30 mL is the breakthrough volume. As seen previously, 2 mL of 1 M HCl are sufficient to desorb 95% of the fixed Sm3þ during percolation, and this volume was finally preferred in order to improve the enrichment factor. In this case, the extraction recoveries of lanthanide ions ranged between 79 and 93%, which led to enrichment factor values between

12 and 14, which are in the same order of magnitude as the values found in the literature with Ln-IIPs [28,31,62]. 3.5. Determination of lanthanide ions in real waters Tap water was first used to evaluate the IIP-SPE procedure with real samples. The percolation volume was 30 mL and the elution volume was reduced to 2 mL. It was observed that 2.5 mL of HCl pH 3.8 was not sufficient to eliminate the non-specific interactions in the case of tap water because the difference in the extraction recoveries of the Ln3þ ions between the IIP and its NIP was only about 30%. The volume of the washing step was increased to 5 mL to improve the suppression of the non-specific interactions. This new SPE procedure was next used with tap water, but also with river water as it is an even more complex sample, and the resulting extraction recoveries of the 8 lanthanides are presented in Table 1. The recoveries of the Ln3þ ions from tap water were between 82 and 96% with the IIP and between 9 and 16% with the NIP, which demonstrates a selective extraction for all lanthanides. The relative standard deviations were below 3% for a given IIP cartridge, showing once again a good SPE repeatability. In the case of river water, the recoveries of La3þ, Ce3þ, Nd3þ, Sm3þ, and Gd3þ were higher than 74%. However, the recoveries of Er3þ and Lu3þ were 64 and 47%, respectively. Nevertheless, the recoveries with the NIP were between 13 and 22%. This demonstrates that despite the decrease in retention of Er3þ and Lu3þ on the IIP, the selectivity is still present even with a complex sample. In fact, the lanthanide ions were lost during the percolation step, perhaps because their breakthrough volumes in river water might have been exceeded. In this case, a decrease in the percolation volume or an increase in the mass of IIP introduced into the cartridge would improve their extraction recoveries. 4. Conclusion Imprinted polymers were synthesized by radical polymerization of Nd3þ, MAA, EGDMA, and AIBN as template ion, monomer, crosslinker, and initiator, respectively, in acetronitrile. The imprinted polymers showed a good selectivity toward both LREEs and HREEs

M. Moussa et al. / Analytica Chimica Acta 963 (2017) 44e52

51

Fig. 6. Extraction recoveries of each lanthanide (30 ng) percolated on IIP S2 as a function of the volume of the percolation solution (Bis-Tris buffer pH 6.5). Washing: 2.5 mL of HCl pH 3.8. Elution: 3 mL of 1 M HCl (n ¼ 3).

Table 1 Extraction recoveries of the 8 Ln3þ ions from tap and river waters (n ¼ 3). Percolation: 30 mL of water sample adjusted at pH 6.5 with Bis-Tris buffer and spiked with 8 Ln3þ (30 ng each). Washing: 5 mL of HCl pH 3.8. Elution: 2 mL of 1 M HCl. Extraction recoveries ± standard deviations (%)

Tap water River water

IIP NIP IIP NIP

La3þ

Ce3þ

90 ± 2 9±1 74 ± 6 13 ± 4

96 10 79 14

± ± ± ±

Nd3þ 2 1 6 3

93 11 80 13

± ± ± ±

Sm3þ 3 2 3 3

compared to the non-imprinted polymers once the SPE procedure was optimized. For the first time, the repeatability of syntheses was investigated and satisfying results were obtained and confirmed with statistical tests. A specific capacity of 60 mmol of Nd3þ per g of sorbent and an enrichment factor of 15 were obtained for the IIP. Selective extractions with tap water and river water were next successfully performed. Other applications with real sample containing lanthanides and Certified Reference Materials constitute the main perspectives of this study to lead to a fully validated analytical method. It would also be very interesting to deepen the characterization of the IIP solid and to study the impact of the polymerization conditions (template/monomer/cross-linker ratio and nature of the porogen) on the structural and SPE properties. This study will also be carried out with the objective of directly synthesizing the IIP in a capillary or a channel of a chip in order to miniaturize the SPE step. Acknowledgments
gion Ile-de-France The authors gratefully acknowledge the Re and the Dim Analytics for their financial support. Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.aca.2017.02.012. References [1] V. Zepf, Rare earth elements: what and where they are, in: Rare Earth Elem., Springer, Berlin Heidelberg, 2013, pp. 11e39. [2] S. Kulaksız, M. Bau, Anthropogenic gadolinium as a microcontaminant in tap water used as drinking water in urban areas and megacities, Appl. Geochem

82 10 78 17

± ± ± ±

2 1 1 1

Gd3þ 85 11 76 16

± ± ± ±

Dy3þ 1 1 2 1

94 12 74 16

± ± ± ±

Er3þ 1 1 1 2

97 15 64 19

± ± ± ±

Lu3þ 0 1 1 6

82 16 47 22

± ± ± ±

1 1 1 7

26 (2011) 1877e1885. [3] G. Merschel, M. Bau, L. Baldewein, E.L. Dantas, D. Walde, B. Bühn, Tracing and tracking wastewater-derived substances in freshwater lakes and reservoirs: anthropogenic gadolinium and geogenic REEs in Lake Parano
a, Brasilia, Comptes Rendus Geosci. 347 (2015) 284e293. [4] S. Kulaksız, M. Bau, Anthropogenic dissolved and colloid/nanoparticle-bound samarium, lanthanum and gadolinium in the Rhine River and the impending destruction of the natural rare earth element distribution in rivers, Earth Planet. Sci. Lett. 362 (2013) 43e50. [5] C. Agatemor, D. Beauchemin, Matrix effects in inductively coupled plasma mass spectrometry: a review, Anal. Chim. Acta 706 (2011) 66e83. [6] M.B. Shabani, T. Akagi, H. Shimizu, A. Masuda, Determination of trace lanthanides and yttrium in seawater by inductively coupled plasma mass spectrometry after preconcentration with solvent extraction and back-extraction, Anal. Chem. 62 (1990) 2709e2714. [7] C.W. Noack, D.A. Dzombak, A.K. Karamalidis, Determination of rare earth elements in hypersaline solutions using low-volume, liquideliquid extraction, Environ. Sci. Technol. 49 (2015) 9423e9430. [8] Y. Zhu, A. Hioki, K. Chiba, Development of an automatic pH-adjustment system for solid phase extraction prior to the determination of REEs in seawater by ICP-MS, J. Anal. At. Spectrom. 28 (2013) 883e889. [9] D. Taicheng, C. Hangting, Z. Xianjin, Determination of rare and rare earth elements in soils and sediments by ICP-MS using Ti(OH)4eFe(OH)3 coprecipitation preconcentration, J. Anal. At. Spectrom. 17 (2002) 410e413. [10] N. Freslon, G. Bayon, D. Birot, C. Bollinger, J.A. Barrat, Determination of rare earth elements and other trace elements (Y, Mn, Co, Cr) in seawater using Tm addition and Mg(OH)2 co-precipitation, Talanta 85 (2011) 582e587. [11] L. Qi, M.-F. Zhou, J. Malpas, M. Sun, Determination of rare earth elements and Y in ultramafic rocks by ICP-MS after preconcentration using Fe(OH)3 and Mg(OH)2 coprecipitation, Geostand. Geoanalytical Res. 29 (2005) 131e141. [12] Y. Li, W. Guo, Z. Wu, L. Jin, Y. Ke, Q. Guo, S. Hu, Determination of ultra-trace rare earth elements in high-salt groundwater using aerosol dilution inductively coupled plasma-mass spectrometry (ICP-MS) after iron hydroxide coprecipitation, Microchem. J. 126 (2016) 194e199. [13] D. Das, M. Dutta, M.L. Cervera, M. de la Guardia, Recent advances in on-line solid-phase pre-concentration for inductively-coupled plasma techniques for determination of mineral elements, TrAC Trends Anal. Chem. 33 (2012) 35e45. [14] M.-C. Hennion, Solid-phase extraction: method development, sorbents, and coupling with liquid chromatography, J. Chromatogr. A 856 (1999) 3e54. [15] K. Pyrzynska, A. Kubiak, I. Wysocka, Application of solid phase extraction

52

[16]

[17]

[18]

[19]

[20]

[21]

[22]

[23] [24]

[25] [26] [27]

[28]

[29]

[30]

[31]

[32] [33]

[34] [35]

[36]

[37] [38]

[39]

M. Moussa et al. / Analytica Chimica Acta 963 (2017) 44e52 procedures for rare earth elements determination in environmental samples, Talanta 154 (2016) 15e22. N. Bahramifar, Y. Yamini, On-line preconcentration of some rare earth elements in water samples using C18-cartridge modified with l-(2-pyridylazo) 2naphthol (PAN) prior to simultaneous determination by inductively coupled plasma optical emission spectrometry (ICPeOES), Anal. Chim. Acta 540 (2005) 325e332. C. Karadas¸, D. Kara, A. Fisher, Determination of rare earth elements in seawater by inductively coupled plasma mass spectrometry with off-line column preconcentration using 2,6-diacetylpyridine functionalized Amberlite XAD-4, Anal. Chim. Acta 689 (2011) 184e189. Y. Zhu, A. Itoh, E. Fujimori, T. Umemura, H. Haraguchi, Determination of rare earth elements in seawater by ICP-MS after preconcentration with a chelating resin-packed minicolumn, J. Alloys Compd. 408e412 (2006) 985e988. S.A. Kumar, S.P. Pandey, N.S. Shenoy, S.D. Kumar, Matrix separation and preconcentration of rare earth elements from seawater by poly hydroxamic acid cartridge followed by determination using ICP-MS, Desalination 281 (2011) 49e54. Q. Fu, L. Yang, Q. Wang, On-line preconcentration with a novel alkyl phosphinic acid extraction resin coupled with inductively coupled plasma mass spectrometry for determination of trace rare earth elements in seawater, Talanta 72 (2007) 1248e1254. L. Zhang, B. Chen, M. He, B. Hu, Polymer monolithic capillary microextraction combined on-line with inductively coupled plasma MS for the determination of trace rare earth elements in biological samples, J. Sep. Sci. 36 (2013) 2158e2167. R.N.C.S. Carvalho, G.B. Brito, M.G.A. Korn, J.S.R. Teixeira, F. de S. Dias, A.F. Dantas, L.S.G. Teixeira, Multi-element determination of copper, iron, nickel, manganese, lead and zinc in environmental water samples by ICP OES after solid phase extraction with a C18 cartridge loaded with 1-(2pyridylazo)-2-naphthol, Anal. Methods 7 (2015) 8714e8719. M.C.B. Fortes, A.H. Martins, J.S. Benedetto, Selective separation of indium by iminodiacetic acid chelating resin, Braz. J. Chem. Eng. 24 (2007) 287e292. D. Kara, A. Fisher, S.J. Hill, Preconcentration and determination of trace elements with 2,6-diacetylpyridine functionalized Amberlite XAD-4 by flow injection and atomic spectroscopy, Analyst 130 (2005) 1518e1523. T.P. Rao, R. Kala, S. Daniel, Metal ion-imprinted polymers-Novel materials for selective recognition of inorganics, Anal. Chim. Acta 578 (2006) 105e116. C. Branger, W. Meouche, A. Margaillan, Recent advances on ion-imprinted polymers, React. Func. Polym. 73 (2013) 859e875. V.M. Biju, J.M. Gladis, T.P. Rao, Ion imprinted polymer particles: synthesis, characterization and dysprosium ion uptake properties suitable for analytical applications, Anal. Chim. Acta 478 (2003) 43e51. V.M. Biju, J.M. Gladis, T.P. Rao, Effect of Y-irradiation of ion imprinted polymer (IIP) particles for the preconcentrative separation of dysprosium from other selected lanthanides, Talanta 60 (2003) 747e754. R. Kala, V.M. Biju, T.P. Rao, Synthesis, characterization, and analytical applications of erbium(III) ion imprinted polymer particles prepared via Y-irradiation with different functional and crosslinking monomers, Anal. Chim. Acta 549 (2005) 51e58. K. Ramakrishnan, T.P. Rao, Ion imprinted polymer solid phase extraction (IIPSPE) for preconcentrative separation of erbium(III) from adjacent lanthanides and yttrium, Sep. Sci. Technol. 41 (2006) 233e246. R. Kala, J. Mary Gladis, T.P. Rao, Preconcentrative separation of erbium from Y, Dy, Ho, Tb and Tm by using ion imprinted polymer particles via solid phase extraction, Anal. Chim. Acta 518 (2004) 143e150. J. Guo, J. Cai, Q. Su, Ion imprinted polymer particles of neodymium: synthesis, characterization and selective recognition, J. Rare Earths 27 (2009) 22e27. P.G. Krishna, J.M. Gladis, T.P. Rao, G.R. Naidu, Selective recognition of neodymium (III) using ion imprinted polymer particles, J. Mol. Recognit. 18 (2005) 109e116. X. Lai, Y. Hu, Y. Fu, L. Wang, J. Xiong, Synthesis and characterization of Lu(III) ion imprinted polymer, J. Inorg. Organomet. Polym. Mater 22 (2012) 112e118. S. Shirvani-Arani, S.J. Ahmadi, A. Bahrami-Samani, M. Ghannadi-Maragheh, Synthesis of nano-pore samarium (III)-imprinted polymer for preconcentrative separation of samarium ions from other lanthanide ions via solid phase extraction, Anal. Chim. Acta 623 (2008) 82e88. O. Vigneau, C. Pinel, M. Lemaire, Solid-liquid separation of lanthanide/ lanthanide and lanthanide/actinide using ionic imprinted polymer based on a DTPA derivative, Chem. Lett. (2002) 202. R. Garcia, C. Pinel, C. Madic, M. Lemaire, Ionic imprinting effect in gadolinium/ lanthanum separation, Tetrahedron Lett. 39 (1998) 8651e8654. O. Vigneau, C. Pinel, M. Lemaire, Ionic imprinted resins based on EDTA and DTPA derivatives for lanthanides(III) separation, Anal. Chim. Acta 435 (2001) 75e82. _ Dolak, R. Keçili, D. Hür, A. Erso €z, R. Say, Ion-imprinted polymers for selective I. recognition of neodymium(III) in environmental samples, Ind. Eng. Chem. Res. 54 (2015) 5328e5335.

[40] C.X. Li, J.M. Pan, J. Gao, Y.S. Yan, G.Q. Zhao, An ion-imprinted polymer supported by attapulgite with a chitosan incorporated solegel process for selective separation of Ce(III), Chin. Chem. Lett. 20 (2009) 985e989. [41] X. Zhang, C. Li, Y. Yan, J. Pan, P. Xu, X. Zhao, A Ce3þ-imprinted functionalized potassium tetratitanate whisker sorbent prepared by surface molecularly imprinting technique for selective separation and determination of Ce3þ, Microchim. Acta 169 (2010) 289e296. [42] R. Garcia, O. Vigneau, C. Pinel, M. Lemaire, Solideliquid lanthanide extraction with ionic-imprinted polymers, Sep. Sci. Technol. 37 (2002) 2839e2857. [43] R. Kala, T. Prasada Rao, Ion imprinted polymer particles for separation of yttrium from selected lanthanides, J. Sep. Sci. 29 (2009) 1281e1287. [44] M. Moussa, V. Pichon, C. Mariet, T. Vercouter, N. Delaunay, Potential of ion imprinted polymers synthesized by trapping approach for selective solid phase extraction of lanthanides, Talanta 161 (2016) 459e468. [45] H. Dong, H. Du, X. Qian, Theoretical prediction of pKa values for methacrylic acid oligomers using combined quantum mechanical and continuum solvation methods, J. Phys. Chem. A 112 (2008) 12687e12694. [46] J. Otero-Romani, A. Moreda-Pineiro, P. Bermejo-Barrera, A. Martin-Esteban, Ionic imprinted polymer for nickel recognition by using the bi-functionalized 5-vinyl-8-hydroxyquinoline as a monomer: application as a new solid phase extraction support, Microchem. J. 93 (2009) 225e231. [47] J. Otero-Romani, A. Moreda-Pineiro, P. Bermejo-Barrera, A. Martin-Esteban, Inductively coupled plasma optical emission spectrometry/mass spectrometry for the determination of Cu, Ni, Pb and Zn in seawater after ionic imprinted polymer based solid phase extraction, Talanta 79 (2009) 723e729. [48] J. Otero-Romani, A. Moreda-Pineiro, P. Bermejo-Barrera, A. Martin-Esteban, Synthesis, characterization and evaluation of ionic-imprinted polymers for solid-phase extraction of nickel from seawater, Anal. Chim. Acta 630 (2008) 1e9. [49] N. Garcia-Otero, C. Teijeiro-Valino, J. Otero-Romani, E. Pena-Vazquez, A. Moreda-Pineiro, P. Bermejo-Barrera, On-line ionic imprinted polymer selective solid-phase extraction of nickel and lead from seawater and their determination by inductively coupled plasma-optical emission spectrometry, Anal. Bioanal. Chem. 395 (2009) 1107e1115. [50] V. Yılmaz, O. Hazer, S¸. Kartal, Synthesis, characterization and application of a novel ion-imprinted polymer for selective solid phase extraction of copper(II) ions from high salt matrices prior to its determination by FAAS, Talanta 116 (2013) 322e329. [51] A.-S. Chauvin, J.-C.G. Bünzli, F. Bochud, R. Scopelliti, P. Froidevaux, Use of dipicolinate-based complexes for producing ion-imprinted polystyrene resins for the extraction of Yttrium-90 and heavy lanthanide cations, Chem. Eur. J. 12 (2006) 6852e6864. [52] J. Wang, F. Liu, Synthesis and application of ion-imprinted interpenetrating polymer network gel for selective solid phase extraction of Cd2þ, Chem. Eng. J. 242 (2014) 117e126.
pez, A. Moreda-Pin ~ eiro, [53] M.C. Barciela-Alonso, V. Plata-García, A. Rouco-Lo P. Bermejo-Barrera, Ionic imprinted polymer based solid phase extraction for cadmium and lead pre-concentration/determination in seafood, Microchem. J. 114 (2014) 106e110. [54] M. Saraji, H. Yousefi, Selective solid-phase extraction of Ni(II) by an ionimprinted polymer from water samples, J. Hazard. Mater 167 (2009) 1152e1157. [55] M. Saraji, H. Yousefi, S. Meghdadi, Preparation and evaluation of an ion imprinted solegel material for selective solid-phase extraction of Ni(II), Int. J. Environ. Anal. Chem. 89 (2009) 305e317.
kova
, V. Pichon, B. group, Synthesis of a [56] P. Svoboda, A. Combes, J. Petit, L. Nova ^-Nmolecularly imprinted sorbent for selective solid-phase extraction of a methylamino-l-alanine, Talanta 144 (2015) 1021e1029. [57] M. Lopez-Nogueroles, S. Lordel-Madeleine, A. Chisvert, A. Salvador, V. Pichon, Development of a selective solid phase extraction method for nitro musk compounds in environmental waters using a molecularly imprinted sorbent, Talanta 110 (2013) 128e134. [58] S. Lordel, F. Chapuis-Hugon, V. Eudes, V. Pichon, Development of imprinted materials for the selective extraction of nitroaromatic explosives, J. Chromatogr. A 1217 (2010) 6674e6680. [59] D.L. Massart, B.G.M. Vandeginste, L.M.C. Buydens, S. De Jong, P.J. Lewi, J. Smeyers-Verbeke, Chapter 6 analysis of variance, in: Data Handl. Sci. Technol., Elsevier, 1998, pp. 121e150. [60] S. Daniel, P. Prabhakara Rao, T. Rao, Investigation of different polymerization methods on the analytical performance of palladium(II) ion imprinted polymer materials, Anal. Chim. Acta 536 (2005) 197e206. [61] J.M. Gladis, T.P. Rao, Effect of porogen type on the synthesis of uranium ion imprinted polymer materials for the preconcentration/separation of traces of uranium, Microchim. Acta 146 (2004) 251e258. [62] N. Zhang, B. Hu, C. Huang, A new ion-imprinted silica gel sorbent for on-line selective solid-phase extraction of dysprosium(III) with detection by inductively coupled plasma-atomic emission spectrometry, Anal. Chim. Acta 597 (2007) 12e18.