**3. Molecularly imprinted polymers (MIPs)**

Molecularly imprinted polymers (MIPs) are cross-linked, synthetic polymers with an artificially generated three-dimensional network able to specifically rebind a target analyte, or a class of structural analogues [1]. The principle is that a polymer network is obtained by polymerizing functional and cross-linking monomers around a template molecule. Subsequent removal of the template leaves a cavity with specific recognition sites complementary in shape, size and functional groups to the target analyte (**Figure 3**). These recognition sites can specifically bind target compounds in a similar way that antibodies specifically bind to an antigen, with the advantages of being very selective without suffering from stability problems associated to biological receptors [49]. All these aspects, together with the fact that MIPs synthesis is also relatively easy and cheap when compared with the purification procedure of natural antibodies, have led to a considerable growth of interest in the use of MIPs in several analytical techniques and applications.

Over the last 15 years, MIPs have been successfully applied as stationary phase on liquid chromatography, solid-phase extraction, micro-extraction, capillary electrochromatography, immunoassay determinations, and chemical sensing, with an almost exponential increase in the number of publications [50]. However, it should be pointed out that even if the interest in the area is relatively new, the concept itself has a long history. The earliest documents describing conceptually similar approaches had first been published in the early 1930s [51]. Nonetheless, today's concept of molecular imprinting technology started back in 1972 when the groups of Klotz and Wulff independently presented the first examples of synthetic organic polymers with predetermined ligand selectivities. In both of these studies, MIP synthesis was based on a covalent linkage of the template molecule to the monomers prior to polymerization. Later on, in the early 1980s, the group of Mosbach has reported for the first time a general noncovalent approach for producing organic imprinted polymers [52,53]. This important development has broadened the scope of molecularly imprinting polymers, improving considerably the versatility and the number of possible applications for this type of materials.

**Figure 3.** Schematic representation of non-covalent molecular imprinting procedures: (1) complex formation between the template (methacrylic acid) and the functional monomers (estriol), (2) polymerization, (3) template extraction, (4) analyte rebinding.

## **3.1. Applications of MIPs to SPE**

138 Chromatography – The Most Versatile Method of Chemical Analysis

order of few nanograms per liter (0.3-2.1 ng L−1).

**3. Molecularly imprinted polymers (MIPs)** 

the use of MIPs in several analytical techniques and applications.

Gallart-Ayala et al. [48], used an automated on-line SPE fast LC–MS/MS method for the simultaneous analysis of bisphenol A (BPA), bisphenol F (BPF), bisphenol E (BPE), bisphenol B (BPB) and bisphenol S (BPS) in canned soft drinks without any previous sample treatment. SPE on-line pre-concentration was performed by using a C18 cartridge. The analysis of all compounds was accomplished in 3 min. Quality parameters of the method were established and the authors obtained a simple, fast, reproducible (RSD values lower than 10%) and accurate (trueness higher than 93%) method for the analysis of bisphenols in

Finally, in the current year, Vega-Morales et al. [2] used an on-line SPE-UHPLC-MS/MS to characterize 27 endocrine disrupting compounds (norethindrone, norgestrel, 17-alphaethinyloestradiol, etc.) in sewage samples. SPE treatment was performed by using Oasis HLB columns (mixed-mode sorbent). The complete analysis of each sample required less than 4 min and provided satisfactory recoveries (72–110%) and limits of detection in the

Molecularly imprinted polymers (MIPs) are cross-linked, synthetic polymers with an artificially generated three-dimensional network able to specifically rebind a target analyte, or a class of structural analogues [1]. The principle is that a polymer network is obtained by polymerizing functional and cross-linking monomers around a template molecule. Subsequent removal of the template leaves a cavity with specific recognition sites complementary in shape, size and functional groups to the target analyte (**Figure 3**). These recognition sites can specifically bind target compounds in a similar way that antibodies specifically bind to an antigen, with the advantages of being very selective without suffering from stability problems associated to biological receptors [49]. All these aspects, together with the fact that MIPs synthesis is also relatively easy and cheap when compared with the purification procedure of natural antibodies, have led to a considerable growth of interest in

Over the last 15 years, MIPs have been successfully applied as stationary phase on liquid chromatography, solid-phase extraction, micro-extraction, capillary electrochromatography, immunoassay determinations, and chemical sensing, with an almost exponential increase in the number of publications [50]. However, it should be pointed out that even if the interest in the area is relatively new, the concept itself has a long history. The earliest documents describing conceptually similar approaches had first been published in the early 1930s [51]. Nonetheless, today's concept of molecular imprinting technology started back in 1972 when the groups of Klotz and Wulff independently presented the first examples of synthetic organic polymers with predetermined ligand selectivities. In both of these studies, MIP synthesis was based on a covalent linkage of the template molecule to the monomers prior to polymerization. Later on, in the early 1980s, the group of Mosbach has reported for the first time a general noncovalent approach for producing organic imprinted polymers [52,53]. This important development has broadened the scope of molecularly imprinting polymers, improving considerably the versatility and the number of possible applications for this type of materials.

canned soft drinks at the ng L−1 level using matrix-matched calibration.

Out of all the MIPs applications, the use of MIPs as selective sorbents for solid-phase extraction (MIP-SPE) represents the most important application area in the field of analytical separation sciences [54]. Solid-phase extraction (SPE) is a well-established method routinely used for clean-up and pre-concentration of analytes in a wide range of environmental, pharmaceutical, agricultural and food analysis [1]. Nevertheless, sorbents used in conventional SPE often lack selectivity resulting in co-extraction of interfering matrix components. Therefore, specificity, selectivity and sensitivity together with high extraction efficiency can be obtained using sorbents based on molecularly imprinted polymers (MIPs) [8].

To assess the potential of MIPs in terms of selectivity, we have compared the ability of MIP-SPE for selective extraction of zearalenone from cereal sample extracts with that of a commercial immunoaffinity column (IAC). **Figure 4** shows the similarity of the behavior of these two types of selective sorbents, resulted in high degrees of clean-up. In both cases, very reliable baselines and similar recoveries were obtained, proving that the high selectivity of immunoaffinity sorbents also can be achieved with molecularly imprinted polymers SPE. Furthermore, previous studies have found MIP-SPE to have a similar selectivity but a higher capacity than commercial IAC columns [49,55].

Current Trends in Sample Treatment Techniques for Environmental and Food Analysis 141

HPLC-PDA 89.1 to 101.3% LOD ≥ 0.095 µg L-1 [64]

HPLC-UV 79.9 to 132.7 % -- [65]

GC-MS 94.7 to 101.9% LOD ≥ 5.49 ng L-1 [67]

HPLC-PDA 94 to 99 % LOD ≥ 80 ng L-1 [68]

HPLC-UV 82.3 to 94.7% -- [69]

HPLC-UV 80 to 90 % LOD < 1 ng g-1 [70]

Samples HPLC-FLD 62 to 102% LOD ≥ 1 ng L-1 [7]

**Table 2.** Some recent applications of MIP-SPE in food and environmental analysis. LOD= Limit of

and selectivity, each step of the extraction procedure must be properly optimized.

Regarding the analytical method, MIP-SPE procedure is based on the same main four steps as conventional SPE such as pre-conditioning of the sorbent, sample loading, interferences wash step and elution of the target compounds. Therefore, to obtain optimal recovery rates

MIP-SPE can be basically used in both the reversed phase and normal phase modes. In the normal phase approach, the sample is usually percolated though the MIP-SPE column using the same solvent that was used as porogen for the MIP synthesis. Under this condition, the target analyte develops specific interactions with the monomer residues present in the polymer cavities, resulting in selective adsorption and molecular recognition by MIP due to

GC-NPD 95.4 to 96.1 % LOD ≥ 10 ng L-1

LOQ ≥ 14.9 ng L-1 [8]

LOQ ≥ 55.3 ng L-1 [63]

LOD ≥ 3.8 ng g-1 [66]

MS 48 to 106 % LOD ≥ 4.50 ng L-1

Seawater GC-ECD 86.4 to 96.0% LOD ≥ 16.6 ng L-1

Natural and Synthetic Estrogens

Pyrethroid Insecticides

Water-Soluble Acid Dyes

Levonorgestrel

Methamidophos

Dibutyl Phthalate

Atrazine Herbicide

Parabens

Chlorsulfuron

Fluoroquinolone antimicrobials

River and Tap

Aquaculture

Wastewater and Soft Drink Samples

River and WWTP influent and effluent samples

Soil Samples, Tap and River

Water Samples

Aqueous Environment Samples

Aqueous Environment Samples

Water, Soil, and Wheat Plant Samples

Soil and Sediment Samples

Water

the well-known solvent "memory" effect [54].

detection; LOQ= Limit of quantification

UHPLC-MS-

Water Samples

**Figure 4.** HPLC-FLD chromatograms of wheat sample spiked with ZON at level of 100 µgkg−1 after extract clean-up with (**—**) MIP-SPE (AFFINIMIP® SPE ZEARALENONE; Polyintell) and (**—**) immunoaffinity column (IAC).

These aspects are highly attractive for matrix clean-up, enrichment and selective extraction of analytes in difficult samples that are very common to food and environmental analyses. Hence, several examples of MIP-SPE applications have been described in the literature, as exemplified in **Table 2**, which presents a selection of the most recently published scientific research.



**Compounds Sample Analysis Recovery Rates** 

Sudan I Chilli Sauce HPLC-UV 87.5 to 103.4 % LOD ≥ 3.3 µg kg−<sup>1</sup>

HPLC–PDA–

Antibiotics Egg Samples HPLC-PDA 91.6 to 107.6% LOQ ≥ 0.8 ng g-1 [6]

immunoaffinity column (IAC).

research.

*Food Analysis* 

Mycophenolic

Tetracycline

Thiamphenicol

Bisphenol A

Ochratoxin A Wheat

Domoic acid Seafood

Zearalenone Cereal

*Environmental Analysis* 

Catechins Tea, Cocoa,

Grape

acid Maize HPLC-MS-

Milk and Honey Samples

Ultrapure, Tap, Drinking, River Water Samples

**Figure 4.** HPLC-FLD chromatograms of wheat sample spiked with ZON at level of 100 µgkg−1 after extract clean-up with (**—**) MIP-SPE (AFFINIMIP® SPE ZEARALENONE; Polyintell) and (**—**)

These aspects are highly attractive for matrix clean-up, enrichment and selective extraction of analytes in difficult samples that are very common to food and environmental analyses. Hence, several examples of MIP-SPE applications have been described in the literature, as exemplified in **Table 2**, which presents a selection of the most recently published scientific

Samples MISPE-FLD 92.1 to 104 % LOD ≥ 1.2 ng mL-1 [57]

Samples HPLC-PDA 93.4 to 96.7% LOQ ≥ 0.1mg L−1 [58]

Samples HPLC-UV 82 to 90% -- [49]

MS 49 to 84% LOD ≥ 0.17 µg kg-1

HPLC-PDA 92.9 to 99.3% LOD≥ 0.003 µg mL-1

HPLC-FLD 84.7 to 93.8% LOD ≥ 2.50 pg mL-1

FL 50 to 100% -- [59]

**(%) Analytical features Reference** 

LOQ ≥ 0.57 µg kg-1 [60]

LOD ≥ 0.002 µg g-1 [61]

LOQ ≥ 8.33 pg mL-1 [62]

[56]

**Table 2.** Some recent applications of MIP-SPE in food and environmental analysis. LOD= Limit of detection; LOQ= Limit of quantification

Regarding the analytical method, MIP-SPE procedure is based on the same main four steps as conventional SPE such as pre-conditioning of the sorbent, sample loading, interferences wash step and elution of the target compounds. Therefore, to obtain optimal recovery rates and selectivity, each step of the extraction procedure must be properly optimized.

MIP-SPE can be basically used in both the reversed phase and normal phase modes. In the normal phase approach, the sample is usually percolated though the MIP-SPE column using the same solvent that was used as porogen for the MIP synthesis. Under this condition, the target analyte develops specific interactions with the monomer residues present in the polymer cavities, resulting in selective adsorption and molecular recognition by MIP due to the well-known solvent "memory" effect [54].

Nonetheless, in some common situations, a loading step based on direct percolation of aqueous sample through the MIP-SPE cartridge is highly desirable since most environmental or biological samples exist in an aqueous matrix [71]. Under this reversedphase condition, the target analytes as well as non-polar interfering compounds are mainly retained by non-specific hydrophobic interactions. Thus, to generate specific interactions between the target compounds and the MIP and to disrupt the non-specific interactions between the residual monomers located at the surface of the polymer and matrix components, a selective washing step using low-to-medium polarity organic solvents, such as dichloromethane, chloroform, toluene or acetonitrile, is usually required [49]. It should be pointed out, however, that in some cases this selective washing step can be problematic because of the low polarity of common solvents used, which may give rise to miscibility problems and/or losses of the analyte [54]. Consequently, a drying step prior to this organic washing process becomes mandatory [8].

Current Trends in Sample Treatment Techniques for Environmental and Food Analysis 143

The chromatogram obtained (red line) clearly illustrates the efficiency of the MIP-SPE procedure (extraction rate of about 90%) and the advantages of both the concentration and sample clean-up with very low background and no interferences close to the retention time of 17β-estradiol. As a result, the use of such kind of selective sorbent allowed the successful detection of estradiol present at very low concentration without the need for a more selective and sensitive method of data acquisition such as mass spectrometry (MS)

Regarding different operation modes, MIP-SPE has been used both in on-line and off-line modes prior to various detection techniques. Most of the MIP-SPE applications reported so far have been developed in the off-line mode because of its simplicity, ease-of-use and high flexibility. In addition, the drying step required in most of the off-line procedures is not compatible with on-line operations. Nonetheless, in the last years, there has been a considerably increase in the number of applications that use MIPs as sorbent for on-line SPE because provides higher pre-concentration factors with reduced analysis time and sample

In the off-line mode, MIP–SPE has been applied for the selective extraction and preconcentration of a wide range of analytes, such as mycotoxins in cereal samples [55], thiamphenicol in milk and honey samples [61], domoic acid from seafood [58], ofloxacin and lomefloxacin in chicken muscle [72], pyrethroid insecticides in aquaculture seawater [63] and parabens in soil and sediment samples [70]. Furthermore, MIP-SPE has been applied not only for the extraction of single target analyte but also for the simultaneous isolation of a class of structurally related compounds such as catechins or estrogens from real samples [8,59]. Luo et al. [64], developed a sensitive and selective off-line MIP-SPE based method to determine five water-soluble acid dyes in wastewater and soft drink samples. The precision and accuracy of the method were satisfactory and it gave average recoveries between 89.1 and 91.0%. In a similar way, Qi et al. [65], prepared a MIP by conventional bulk polymerization for the extraction of levonorgestrel from water samples. The synthesized MIPs not only displayed high specific recognition for levonorgestrel (recoveries > 79%), but also showed high cross-reactivity values for structurally related contraceptive drugs,

suggesting that MIPs could be used as broad specific recognition absorbent.

drinking and fish farm water samples, respectively.

On-line MIP-SPE protocol has been successfully used to extract benzimidazole fungicides in water samples [73], ochratoxin A in wheat samples [57], and bisphenol A in environmental water samples [74]. An automated on-line SPE using microspherical monodispersed molecularly imprinted particles coupled to HPLC-fluorescence detector was successfully applied to the simultaneous multi-residue analysis of six fluoroquinolone antimicrobials (enrofloxacin, ciprofloxacin, norfloxacin, levofloxacin, danofloxacin, and sarafloxacin) in water samples [7]. In this work, polymer particles prepared via precipitation polymerization were used as SPE sorbent. High recoveries with good precision (RSDs <5%) were obtained for the different fluoroquinolones tested, with values ranging from 91 to 102% in drinking and fish farm water samples. The detection limits were between 1-11 and 1-12 ng L-1 for

detection.

manipulation.

Once matrix interferences are removed, the analytes can be eluted from the column with a pure solvent, solvent that contains a small amount of modifier such as acetic acid or a combination of solvents with different polarities that must possess an elution strength sufficiently high to disrupt the specific interactions of the target analytes with the polymer, in minimal elution volumes.

As example of the successful application of a MIP-SPE compatible with aqueous samples, **Figure 5** shows the HPLC-FLD chromatogram (red line) corresponding to the injection of the elution fraction obtained after the purification of 100mL of Seine river water spiked with 0.5 ng mL−1 of 17β-estradiol.

**Figure 5.** HPLC-FLD chromatograms obtained after extracts clean-up with MIP-SPE (AFFINIMIP® SPE Estrogens; Polyintell) of 100mL of Seine water spiked at 0.5 ng mL-1 with 17β-estradiol (**—**) and before MIP clean-up (—).

The chromatogram obtained (red line) clearly illustrates the efficiency of the MIP-SPE procedure (extraction rate of about 90%) and the advantages of both the concentration and sample clean-up with very low background and no interferences close to the retention time of 17β-estradiol. As a result, the use of such kind of selective sorbent allowed the successful detection of estradiol present at very low concentration without the need for a more selective and sensitive method of data acquisition such as mass spectrometry (MS) detection.

142 Chromatography – The Most Versatile Method of Chemical Analysis

washing process becomes mandatory [8].

in minimal elution volumes.

0.5 ng mL−1 of 17β-estradiol.

MIP clean-up (—).

Nonetheless, in some common situations, a loading step based on direct percolation of aqueous sample through the MIP-SPE cartridge is highly desirable since most environmental or biological samples exist in an aqueous matrix [71]. Under this reversedphase condition, the target analytes as well as non-polar interfering compounds are mainly retained by non-specific hydrophobic interactions. Thus, to generate specific interactions between the target compounds and the MIP and to disrupt the non-specific interactions between the residual monomers located at the surface of the polymer and matrix components, a selective washing step using low-to-medium polarity organic solvents, such as dichloromethane, chloroform, toluene or acetonitrile, is usually required [49]. It should be pointed out, however, that in some cases this selective washing step can be problematic because of the low polarity of common solvents used, which may give rise to miscibility problems and/or losses of the analyte [54]. Consequently, a drying step prior to this organic

Once matrix interferences are removed, the analytes can be eluted from the column with a pure solvent, solvent that contains a small amount of modifier such as acetic acid or a combination of solvents with different polarities that must possess an elution strength sufficiently high to disrupt the specific interactions of the target analytes with the polymer,

As example of the successful application of a MIP-SPE compatible with aqueous samples, **Figure 5** shows the HPLC-FLD chromatogram (red line) corresponding to the injection of the elution fraction obtained after the purification of 100mL of Seine river water spiked with

**Figure 5.** HPLC-FLD chromatograms obtained after extracts clean-up with MIP-SPE (AFFINIMIP® SPE Estrogens; Polyintell) of 100mL of Seine water spiked at 0.5 ng mL-1 with 17β-estradiol (**—**) and before

Regarding different operation modes, MIP-SPE has been used both in on-line and off-line modes prior to various detection techniques. Most of the MIP-SPE applications reported so far have been developed in the off-line mode because of its simplicity, ease-of-use and high flexibility. In addition, the drying step required in most of the off-line procedures is not compatible with on-line operations. Nonetheless, in the last years, there has been a considerably increase in the number of applications that use MIPs as sorbent for on-line SPE because provides higher pre-concentration factors with reduced analysis time and sample manipulation.

In the off-line mode, MIP–SPE has been applied for the selective extraction and preconcentration of a wide range of analytes, such as mycotoxins in cereal samples [55], thiamphenicol in milk and honey samples [61], domoic acid from seafood [58], ofloxacin and lomefloxacin in chicken muscle [72], pyrethroid insecticides in aquaculture seawater [63] and parabens in soil and sediment samples [70]. Furthermore, MIP-SPE has been applied not only for the extraction of single target analyte but also for the simultaneous isolation of a class of structurally related compounds such as catechins or estrogens from real samples [8,59]. Luo et al. [64], developed a sensitive and selective off-line MIP-SPE based method to determine five water-soluble acid dyes in wastewater and soft drink samples. The precision and accuracy of the method were satisfactory and it gave average recoveries between 89.1 and 91.0%. In a similar way, Qi et al. [65], prepared a MIP by conventional bulk polymerization for the extraction of levonorgestrel from water samples. The synthesized MIPs not only displayed high specific recognition for levonorgestrel (recoveries > 79%), but also showed high cross-reactivity values for structurally related contraceptive drugs, suggesting that MIPs could be used as broad specific recognition absorbent.

On-line MIP-SPE protocol has been successfully used to extract benzimidazole fungicides in water samples [73], ochratoxin A in wheat samples [57], and bisphenol A in environmental water samples [74]. An automated on-line SPE using microspherical monodispersed molecularly imprinted particles coupled to HPLC-fluorescence detector was successfully applied to the simultaneous multi-residue analysis of six fluoroquinolone antimicrobials (enrofloxacin, ciprofloxacin, norfloxacin, levofloxacin, danofloxacin, and sarafloxacin) in water samples [7]. In this work, polymer particles prepared via precipitation polymerization were used as SPE sorbent. High recoveries with good precision (RSDs <5%) were obtained for the different fluoroquinolones tested, with values ranging from 91 to 102% in drinking and fish farm water samples. The detection limits were between 1-11 and 1-12 ng L-1 for drinking and fish farm water samples, respectively.

On-line MIP-SPE pre-concentration methodology has also recently been used by Jing et al. [6], for the determination of trace tetracycline antibiotics (TCs) in egg samples. This approach affords high-throughput analysis (18 min per sample), and also provides high sensitivity and selectivity with recoveries ranging between 91.6 and 107.6%, showing that efforts should continue to be made in this promising research area.

Current Trends in Sample Treatment Techniques for Environmental and Food Analysis 145

The idea of the QuEChERS procedure was to reduce complicated, laborious and timeconsuming multi-residue sample treatment methods that required high amount of solvents and were therefore expensive. Moreover, some basic, acidic and very polar compounds cannot be satisfactorily extracted with common multi-residue methods. Thus, in order to cover all these analytes, laboratories have to perform specific analysis, and as a consequence,

The first two steps of a typical QuEChERS procedure consist in weighing an appropriate amount of sample previously processed and homogenized (for instance 10 g) in a 50 mL Teflon tube (Step 1) and the addition of a solvent for the extraction (Step 2), in general acetonitrile, although the use of other organic solvents such as acetone, THF or ethyl acetate

Then, an extraction-partitioning step takes place by the addition of magnesium sulfate alone or in combination with other salts, generally sodium chloride (Step 3). Acetonitrile is the recommended solvent for QuEChERS because, upon the addition of salts, it is easily separated from water than, for instance, acetone. Ethyl acetate has the advantage of a partial miscibility with water but it can also co-extract lipids and provides lower recoveries during the dispersive SPE. The extraction of lipophilic materials is lower with acetonitrile but this solvent can form two phases with water when samples with high sugar content are manipulated [75]. The addition of salts in Step 3 helps to induce the phase separation. This salting-out effect also influences analyte partition, which of course is also dependent upon the solvent used for extraction. The concentration of salt can also influence the percentage of water in the organic phase and can play an important role in adjusting its polarity. Magnesium sulfate acts as a drying salt to reduce the water phase, thereby helping to improve recoveries by promoting partitioning of the pesticides (or other target compounds) into the organic layer while sodium chloride helps to control the polarity of the extraction solvent. With this, a single extraction-partitioning step is carried out (similarly to an "online" approach) which simplifies the necessity of multiple partitioning steps required in other multi-residue methods. Moreover, this extraction-partitioning step is produced by shaking vigorously for a few minutes, thus preventing more time-consuming steps such as sample blending. At this point, internal standards can be added to the system if necessary (Step 4), followed by shaking again the solution and a centrifugation step that help to separate salts. The use of internal standards can minimize the error generated in the multiple steps of the QuEChERS method. Sometimes the use of more than one internal standard is recommended especially with samples with high fat content because the excessive fat can form an additional layer into which the analytes can also partition [82]. Another advantage of QuEChERS procedure is the fact that, once the extraction-partitioning step is carried out, an aliquot of the extract is used for the next steps, minimizing also the separation or the transfer of the entire extracts frequently employed in other multi-residue

Then, a dispersive solid-phase extraction (d-SPE) clean-up procedure is carried out with an aliquot (for instance 1 mL) of the extract which is placed in a vial containing again magnesium sulfate and small amounts of bulk SPE sorbent materials (Step 5). The vial is

some of these compounds were not being monitored.

have been described [81].

methods.
