Preparation of metallic pivot-based imprinted monoliths with a hydrophilic macromonomer

Abstract

A new metallic pivot-based molecularly imprinted polymer (MIP) was developed to enhance the imprinting effect of a water-soluble template. The hydrophilic macromonomer oligo(ethyleneglycol) methyl ether methacrylate (OEG), bearing a small oligoethylene glycol side chain, was introduced into the MIP matrix in order to achieve good selectivity and less hydrophobic character. In a ternary porogenic system of dimethyl sulfoxide–dimethylformamide–1-butyl-3-methylimidazolium tetrafluoroborate, an imprinted monolithic column with high porosity and good permeability was synthesized using a mixture of gallic acid (template), 4-vinylpyridine (4-VP), ethylene glycol dimethacrylate, and nickel acetate. The effects of some polymerization variables, such as the ratio of OEG/4-VP and the ratio of template to nickel ions, on the imprinting of the resulting MIP monoliths were systematically investigated. The greatest imprinting factor of 11.53 was achieved on the water compatible MIP monolith with optimized polymerization parameters. In addition, Freundlich analyses indicated that the Ni²⁺-mediated OEG MIP had homogeneous affinity with a heterogeneity index of 0.87, compared with the Ni²⁺-mediated OEG-free MIP (0.65) and Ni²⁺-free OEG MIP (0.67).

---

Introduction

Molecularly imprinted polymers (MIPs) have been proven to be synthetic materials with highly specific molecular recognition abilities.^1,2 The preparation of MIPs involves using a target molecule as a template, which directs the self-assembly of functional monomers that are subsequently co-polymerized in the presence of cross-linking monomers. With their tailored selectivity, easy preparation, and chemical robustness, MIPs can be employed as specific affinity matrices for target templates. Impressive progress has been made over the past few years in the production of materials for various analytical and separation applications^3,4 due to a better understanding of the mechanisms of forming imprints. Recent developments in the use of MIPs for chromatographic stationary phases,^5–7 solid-phase extraction,^8,9 drug release,^10,11 catalysis,^12,13 and biosensing^14–16 have been reported.

Preparing MIPs for polar compounds, such as water-soluble phenolic acid, is much more difficult than for non-polar templates. Optimally, the preparative conditions are the same as the re-binding conditions because the same compositions of both preparative and re-binding media can favor the recognition of the template.¹⁷ Concerning this matter, the preparation of MIPs for water-soluble templates in aqueous solution is better than in organic solution. However, the protocol of MIP preparation for organic compounds has generally been based on hydrogen bonding interactions in non-polar solvents, which is weakened in aqueous solution due to the competition of water. Furthermore, non-selective adsorption is often observed on a conventional MIP, originating from hydrophobic interactions on the surface of the polymer matrix.¹⁷ Consequently, a novel method for preparing MIPs containing hydrophilic units is needed.

The use of a highly aqueously solvated cross-linker, such as pentaerythritol triacrylate¹⁸ or N,N-methylenebisacrylamide,¹⁹ to prepare hydrophilic MIPs is one option to solve the problems above. Nevertheless, the interactions between these hydrophilic cross-linkers and the solvent in the rebinding procedure may cause the deformation of the imprinting cavities or even collapse between the functional monomer and template, and this worsens the recognition properties of MIPs. Recently, a new synthetic strategy of copolymerization of the functional monomer with a macromonomer has been developed.²⁰ The hydrophilic monomer bears a small oligoethylene glycol side chain and can increase the hydrophilicity of the imprinted cavities while maintaining a conventional concentration of the usual cross-linkers. However, the introduction of macromonomer segments causes a decrease in the capacity factors of the polymers since the analytes are retained less by nonspecific hydrophobic interactions and pass through the polymer network faster due to the increased hydrophilicity by such segments. Consequently, a novel method for preparing hydrophilic MIPs that have higher retention and affinity is desired.

For MIPs created via noncovalent imprinting, monomers that can undergo noncovalent interactions are brought together with the template molecule and monomer to form well-defined complexes. Thus, the achievable selectivity of the resultant MIPs is governed by the nature and stability of these complexes. A generally applicable approach for stabilizing the complex is to design a particular functional monomer capable of forming strong interactions with the template. For example, a functional monomer forming strong pre-polymerization complexes with the template in a stoichiometric ratio^21,22 can generate imprinted polymers with binding sites of higher affinity and increased retention. In addition, a number of works have demonstrated that the interactions between the template molecule and functional monomers could be stabilized by a macromolecular crowding agent to increase the retention of MIPs.^23–25

The use of a metallic pivot for self-assembly has been revealed to effectively produce highly specific MIPs.^26–29 Using this strategy, the weak linkage between the monomer and template, such as a hydrogen bond or Coulomb force, is replaced by stronger coordination bonding. This stabilization originates from a metal ion-mediated self-organized architecture and leads to a decrease in the thermal and mechanical motions of the monomers or oligomer–template. In other words, when assembling with a metal ion as the pivot, monomers are regularly positioned around the template via coordinative bridges, which largely restrains the relative motion of the monomers or oligomer–template. As a result, a higher fidelity of the imprint can be achieved using this approach.

In view of the facts above, it is intriguing for us to investigate whether assembly with a metal ion as the pivot can be utilized to prepare hydrophilic MIPs with enhanced affinity. In this work, we prepared OEG-based imprinted monoliths with nickel ions as the pivot for the first time. The monolithic format of the MIPs is desired because the general problems of MIP preparation using conventional bulk polymerization can be avoided.⁵ Gallic acid (GA) was selected as a model water-soluble template and 4-vinylpyridine (4-VP) as a functional monomer. The effect of polymerization parameters such as the ratio of OEG/4-VP and ratio of the template to nickel ions on the imprinting of this new MIP monolith was investigated. The binding characteristics and homogeneity of the imprinting sites on the metal ion-mediated OEG MIP were studied in detail. Furthermore, the selectivity of the novel MIP was evaluated by comparing the retention properties of structural analogues of GA (Fig. 1) on the Ni²⁺-mediated OEG MIP, Ni²⁺-mediated OEG-free MIP and Ni²⁺-free OEG MIP.

Fig. 1 Structures of GA and analogues tested.

Results and discussion

Preparation of metallic pivot-based OEG imprinted monoliths

This work is an effort to improve the imprinting effect of MIPs with good water compatibility. The enhanced molecular recognition towards the imprint species is expected to be achieved by reducing nonspecific hydrophobic interactions using a hydrophilic macromonomer, based on the strategy of a metallic pivot. To investigate the effect of assembly with a metal ion as the pivot on enhancing the affinity of the resulting hydrophilic MIPs, we prepared OEG-based MIPs in a monolithic format (Table 1). In the present work, porogen formulation is crucial to prepare metallic pivot-based imprinted monoliths with a hydrophilic macromonomer. On the one hand, the porogenic solvent can dissolve the polar template molecule, hydrophilic OEG and metal ion. Next, the porogen should produce large pores to assure good flow-through properties of the resultant MIP. On the other hand, the porogenic solvent should avoid the disturbance caused by the polar solvent during the polymerization, in addition to its influence on the polymer morphology. It was found that a previously developed porogenic system,³⁰ a ternary mixture of DMSO, DMF and [BMIM]BF₄, can solve the problems above (Table 1). [BMIM]BF₄ was found to be a unique IL, affording good permeability for the resulting monolithic MIP. Other imidazolium-based ILs with varying cation alkyl chain length (C4–C16) containing the same anion (BF₄⁻) led to MIP monoliths with very high back pressure, which thus can not be evaluated further. Anion-type ionic liquids with fixed cations ([BMIM]⁺) were also used as a component of the porogenic solvent. The miscibility of the cation-type of ionic liquid ([BMIM]PF₆ or [BMIM]HSO₄) with the pre-polymerization mixture limited its use in this study.

Table 1 Preparation protocol for OEG MIP monoliths

Column	GA (mg)	Ni(Ac)2 (mg)	4-VP (μL)	OEG (μL)	EDMA (μL)	DMF (μL)	DMSO (μL)	[BMIM]BF4 (μL)	Time (h)
C1-MIP	17.01	24.88	64	343	453	240	1200	2468	18
C2-MIP	17.01	24.88	64	343	453	240	1200	2468	24
C3-MIP	17.01	24.88	64	343	453	240	1200	2468	24
C4-NIP	—	24.88	64	343	453	240	1200	2468	24
C5-MIP	17.01	—	64	343	453	240	1200	2468	24
C6-NIP	—	—	64	343	453	240	1200	2468	24
C7-MIP	17.01	12.44	64	343	453	240	1200	2468	24
C8-NIP	—	12.44	64	343	453	240	1200	2468	24
C9-MIP	17.01	24.88	64	343	339	240	1200	2468	24
C10-NIP	—	24.88	64	343	339	240	1200	2468	24
C11-MIP	17.01	24.88	64	343	630	240	1200	2468	18
C12-NIP	—	24.88	64	343	630	240	1200	2468	18
C13-MIP	17.01	24.88	64	343	1018	240	1200	2468	18
C14-NIP	—	24.88	64	343	1018	240	1200	2468	18
C15-MIP	17.01	24.88	64	—	630	240	1200	2468	18
C16-NIP	—	24.88	64	—	630	240	1200	2468	18
C17-MIP	17.01	24.88	64	171	630	240	1200	2468	18
C18-NIP	—	24.88	64	171	630	240	1200	2468	18
C19-MIP	17.01	24.88	64	511	630	240	1200	2468	18
C20-NIP	—	24.88	64	511	630	240	1200	2468	18
C21-MIP	17.01	24.88	64	426	630	240	1200	2468	18
C22-NIP	—	24.88	64	426	630	240	1200	2468	18
C23-MIP	17.01	24.88	64	255	630	240	1200	2468	18
C24-NIP	—	24.88	64	255	630	240	1200	2468	18

---

In a first attempt to prepare a GA-imprinted monolith in the presence of OEG, MIP monolith C5 was made in the absence of a metal ion. The resulting MIP did not show any recognition ability, perhaps due to the high polarity of the solvent used in the polymerization system (DMSO, DMF and [BMIM]BF₄)³⁰ affecting the formation of the template–monomer complex. The interaction between the functional monomer and template in a non-covalent MIP may include hydrogen bonding, hydrophobic interaction, charge transfer, or other forms. However, the formation of the template–monomer complex in a polar solvent cannot be efficient enough, because the hydrogen bonding interaction between the GA and monomers can be interrupted by the polar solvent. In order to avoid the disturbance caused by the polar solvent during the polymerization, metal ions have been introduced as a mediator during the pre-polymerization to form a stronger template–metal ion–monomer complex.^26–29 In the present study, a greater imprinting effect (IF = 8.63) was obtained with the Ni²⁺-mediated OEG MIPs (C11) (Fig. 2). The selectivity factors (α) of GA on MIP C11 for its analogues were all increased in comparison with the corresponding NIP. As demonstrated in Fig. 3, the higher IF showed that Ni²⁺ played a vital role in the formation of the MIP materials. The retention factor of the MIP without the involvement of metal ions was almost the same as that of the NIP. In other words, the IF value of the MIP was close to 1, showing little imprinting effect. In this case, when assembling with a metal as a pivot, the monomer and the template are bridged through a coordination bond. This effect might be due to the fact that a more stable ternary complex of monomer–metal ion–template can be formed in polar solvents before polymerization, in which GA and 4-VP could strongly chelate with Ni²⁺ through their carboxyl groups and pyridine groups, respectively. Thus, the crucial role of the metal ion in achieving complete self-assembly for effective imprinting was demonstrated for the polar template in the polar porogen.

Fig. 2 Selectivity evaluation of the MIPs and NIPs. C11, C15, C5: MIPs; C12, C16, C6: NIPs. Mobile phase: acetonitrile–NaAc/HAc buffer (50 mmol L⁻¹, pH 5.0), 70/30 (v/v); detection wavelength: 271 nm; flow rate: 0.5 mL min⁻¹; injection volume: 20 μL; temperature: 25 °C.

Fig. 3 Schematic representation of the preparation of an OEG-based MIP using a metal ion as a pivot, and of the molecular recognition by the MIP.

The morphologies of the Ni²⁺-mediated OEG MIP, Ni²⁺-free OEG MIP and Ni²⁺-mediated OEG-free MIP were observed by SEM (Fig. 4). An agglomerate of microspheres with a coarse surface, fused into a continuous structure, and a typical bimodal pore-size distribution (4–6 μm macropores) were visible for the Ni²⁺-mediated OEG MIP. In contrast, regarding the texture of the Ni²⁺-mediated OEG-free MIP, microglobules of relatively uniform size were agglomerated into larger clusters and greater sizes were observed. In addition, the micrograph of the Ni²⁺-free OEG MIP showed microglobules of smaller size. The results indicated that the existence of the template or Ni²⁺ would have remarkable influence on the size of the microglobules.

Fig. 4 Scanning electron micrographs of the following samples: (a) Ni²⁺-mediated OEG MIP (C11); (b) Ni²⁺-mediated OEG-free MIP (C15); (c) Ni²⁺-free OEG MIP (C5).

To ensure that the superior retention observed for the selected MIPs is not simply a surface effect, multipoint BET measurements for the three MIPs were performed to obtain pore characterization. As shown in Fig. 5a, all the monoliths display “type IV” isotherms which are usually related to meso–macroporous materials.³¹ The hysteresis loops resemble H3 types with a desorption branch leading to the closure point at a non-zero P/P₀ value, suggesting a specific structure of slit-shaped pores. The BET surface area of the OEG MIP C11 calculated from the adsorption data was 13.41 m² g⁻¹. In addition, a maximum corresponding to a pore diameter of about 26.55 nm was observed for the OEG MIP in the desorption-based distribution curve (Fig. 5b), indicating larger mesopores than the other MIP monoliths.^30,32 The BET surface areas of the corresponding MIPs without OEG or Ni²⁺ were 10.69 m² g⁻¹ and 28.32 m² g⁻¹, respectively. Apparently, the presence of OEG or Ni²⁺ has little effect on the pore structure of the resulting MIP monolith. Thus, the contribution of the pore structure and morphology to the imprinting effect of the OEG-based MIP was minor compared with the interaction between the functional monomer, metallic pivot and template. This result precludes the assumption that the highest retention during chromatographic separation is merely a consequence of an increased overall surface area. Therefore, further investigation is required to establish precisely how the pre-organization of monomer–template–metal ion affects the progression of polymerization and leads to differences in the imprinting factors.

Fig. 5 (a) Nitrogen adsorption–desorption isotherms for Ni²⁺-mediated OEG MIP, Ni²⁺-free OEG MIP and Ni²⁺-mediated OEG-free MIP at 77 K; (b) differential pore size distribution curves of Ni²⁺-mediated OEG MIP, Ni²⁺-free OEG MIP and Ni²⁺-mediated OEG-free MIP.

Effect of polymerization variables on molecular recognition

Effect of OEG/4-VP molar ratio. Considering that conventional polymerization parameters of metallic pivot-based MIPs have been investigated in detail previously,^27,29 we focused on three polymerization variables related to the OEG-based MIP in the present study. One of the variables is the molar ratio of OEG/4-VP in the polymerization mixture. It seemed that the interaction between OEG and the template was not involved in the recognition of the resulting MIP. To examine this assumption, an NMR study was conducted with a pseudo-pre-polymerization mixture consisting of OEG, 4-VP, and GA. The initiator AIBN was omitted because of its insignificant involvement in complex formation. The crosslinker EDMA was also omitted because its carbonyl group could not be involved in hydrogen bonding with the template, avoiding the system becoming too complicated in order to observe the complexation between 4-VP or OEG and the template. The concentrations of 4-VP and the template were the same as those used in the polymerization. As shown in Fig. S1,† with the addition of OEG to the GA solution, the peak derived from a carboxyl proton of GA was not shifted downfield, suggesting no formation of hydrogen bonds between GA and OEG.

To achieve optimized imprinting results, the stoichiometric ratios of OEG/4-VP for the MIP preparations studied were set at 2.5 : 1, 2 : 1, 1.5 : 1 and 1 : 1 (Table S1†). When a low level of 4-VP was used, an undesirable effect on imprinting was observed. This may be attributed to a relative excess of the template, which leads to the loss of site integrity due to coalescence of binding sites derived from the template self-association.³³ The maximum imprinting factor was observed at a OEG/4-VP ratio of 2 : 1 (Fig. 6).

Fig. 6 Retention factor and imprinting factor of Ni²⁺-mediated OEG MIPs prepared with different ratios of OEG to 4-VP. Mobile phase: acetonitrile–acetate buffer (50 mmol L⁻¹, pH 5.0) (70/30, v/v); flow rate: 0.5 mL min⁻¹; detection wavelength: 271 nm; injected volume: 20 μL; temperature: 25 °C.

Effect of the molar ratio of cross-linking monomer to functional monomer. For the non-covalent approach, the relationship between the degree of cross-linking of the polymers and their recognition properties is rather complicated. In a few cases, high levels of crosslinker caused an increase in selectivity of the resulting MIPs.³⁴ In another case, it was observed that the selectivity reached to a maximum at a lower degree of cross-linking.³⁵ In the present work, it was observed that high levels of crosslinker resulted in a notable decrease in the imprinting factor from 8.63 to 1.16 (Fig. 7). This may be explained by the severely increased stiffness of the polymer network, thus significantly decreasing the accessibility of the cavities.¹ The optimal degree of crosslinking was found to be 65% in terms of the imprinting factor, which is lower than the conventional 80%. A possible reason is the balance of site stability, integrity and accessibility at the level of an intermediate amount of crosslinker due to the pre-organization of the GA–Ni²⁺–4-VP complex in the preparation of MIP monoliths.

Fig. 7 Retention factor and imprinting factor of Ni²⁺-mediated OEG MIPs prepared with different levels of EDMA. Mobile phase: acetonitrile–acetate buffer (50 mmol L⁻¹, pH 5.0) (70/30, v/v); flow rate: 0.5 mL min⁻¹; detection wavelength: 271 nm; injected volume: 20 μL; temperature: 25 °C.

Effect of the molar ratio of template to nickel ions. It should be noted that not all the OEG-based MIPs with Ni²⁺ participation have larger IFs. It can be inferred from the results of Table S2† that other factors such as the amount of the functional monomers and template molecules also have an impact on the molecular imprinting effect of the MIP. To study the effect of the mediation of nickel ions on the interaction between GA and 4-VP, we prepared a number of Ni²⁺-mediated MIP monoliths by setting the ratio of GA to 4-VP to 1 : 6 and 4-VP to EDMA to 1 : 4. The retention behaviors of the template molecule with the nickel ion-mediated MIP monoliths were evaluated by using a mobile phase of acetonitrile–acetate buffer (pH 3.6) (90/10, v/v). As shown in Table S2,† when the molar ratio of Ni²⁺ to GA decreased from 1 : 1 to 1 : 2, the retention factor of GA on the imprinted monolith decreased from 1.96 to 0.60. The corresponding IF value shifted from 11.53 to 0.34. In contrast, when no nickel ions were used, the retention factor of GA on the MIP (k = 4.16) and NIP (k = 4.39) did not vary much, with an IF value of 0.95. Apparently, a non-stoichiometric ratio of Ni²⁺ to GA adversely affects the specific binding, perhaps due to a decrease in the amount of imprinting complex of 4-VP–Ni²⁺–GA at non-stoichiometric ratios. The stoichiometric ratio-dependent phenomena indicated that for the ion-mediated imprinting system here there are interactions existing bilaterally in the monomer–template, monomer–metal and template–metal, which are subsequently saturated by setting a stoichiometric amount of monomer and template.

Recognition mechanism of imprinted monolith

Effect of pH of mobile phase on retention property. The rebinding of GA to the Ni²⁺-mediated OEG MIP is strongly dependent on the mobile phase used. In this study, acetonitrile–acetate buffer (70/30, v/v) with a pH range from 3.0 to 7.0 was used as the mobile phase to evaluate effect of pH on the imprinting factor. The impact of pH on the recognition of the template is shown in Fig. 8; the binding of the Ni²⁺-mediated OEG MIP (C11) is strongly influenced by electrostatic interactions. The maximum imprinting factor was observed at an eluent pH of 5.0, a value close to the pK_a (= 5.3) of GA. At one pH unit above the pK_a value, a certain percentage of the molecule will be deprotonated (negatively charged). As the pH nears the pK_a value, the number of molecules negatively charged decreases (it will be 50% at the pK_a value). Thus, this indicates that the retention is controlled by an ion-exchange process.³⁶ A further increase in pH of the mobile phase led to a peak split of the GA, and the measurement of the retention factor and thus imprinting factor was impossible. It should be noted that similar retention factors of GA can be observed for all three NIPs but the retention on the Ni²⁺-mediated OEG-free MIP (C15) or Ni²⁺-free OEG MIP (C5) was totally different from the Ni²⁺-mediated OEG MIP (C11). In addition, the analogs of GA showed varying retention behaviors on the Ni²⁺-mediated OEG-free MIP (C15) or Ni²⁺-free OEG MIP (C5), except for MG (Fig. S2 and S3†).

Fig. 8 Influence of the pH value of the mobile phase on the retention factors of GA and imprinting factor of (a) Ni²⁺-mediated OEG MIP (C11), (b) Ni²⁺-mediated OEG-free MIP (C15), (c) Ni²⁺-free OEG MIP (C5) and respective NIPs. HPLC conditions: column temperature, 25 °C; mobile phase: acetonitrile–acetate buffer (50 mmol L⁻¹) (70/30, v/v); flow rate: 0.5 mL min⁻¹; detection wavelength: 271 nm; injected volume: 20 μL.

Influence of organic phase composition on retention property. The influence of the organic modifier in a mixture of acetonitrile–acetate buffer on the retention factor of GA and its analogues were studied using the Ni²⁺-mediated OEG MIP (C11) and non-imprinted monolith (C12). A mixture of acetonitrile–acetate buffer solution (50 mmol L⁻¹, pH 3.6) was used as the mobile phase, with the content of acetonitrile ranging from 20% to 90%. As shown in Fig. 9, the retention factors of the template decreased with decreasing amount of acetonitrile from 90% to 50%. When the amount of acetonitrile decreased from 50% to 20%, the retention factors increased. These results implied a change in the retention mode from an electrostatically driven mode in the low-water system to desolvation retention at higher water contents on the Ni²⁺-based imprinted column.³⁷ In contrast, similar retention behaviors cannot be observed on the Ni²⁺-mediated OEG-free MIP (C15) and Ni²⁺-free OEG MIP (C5) in spite of same trend in retention on all three NIPs. It should be noted that MG is a molecule strictly resembling the template, characterized by the sole absence of a methyl. A greater difference in retention properties for the template was observed on the ion-mediated OEG MIP, the Ni²⁺-mediated OEG-free MIP and Ni²⁺-free OEG MIP (data not shown), which showed the good selectivity of the GA-imprinted polymer. Furthermore, the tested structural analogues, such as SA, MHA, PHA, EA and DHA, showed different retention properties on the Ni²⁺-mediated OEG-free MIP (C15) and Ni²⁺-free OEG MIP (C5) (Fig. S4 and S5†).

Fig. 9 Influence of organic phase composition on the retention factors of GA and imprinting factor of (a) Ni²⁺-mediated OEG MIP (C11), (b) Ni²⁺-mediated OEG-free MIP (C15), (c) Ni²⁺-free OEG MIP (C5) and respective NIPs. HPLC conditions: column temperature, 25 °C; mobile phase: acetonitrile–acetate buffer (50 mmol L⁻¹, pH 3.6); flow rate: 0.5 mL min⁻¹; detection wavelength: 271 nm; injected volume: 20 μL.

Binding characteristics of MIP monolith

The binding characteristics between the template and the imprinted polymer can be determined using frontal chromatography (data not shown). The affinity of the imprinted polymers was determined using Scatchard–Rosenthal analysis (Fig. S6†) and nonlinear profiles were observed. The assessment was therefore conducted, paying particular attention to a partially linear section observed at a range of 1–12 μM where relatively high-affinity binding sites of each polymer can be estimated. The dissociation constants K_d of the Ni²⁺ mediated OEG MIP (C11), Ni²⁺ mediated OEG-free MIP (C15) and Ni²⁺-free OEG MIP (C5) were 9.26 × 10⁻⁴, 5.49 × 10⁻³, and 1.24 × 10⁻³ mol g⁻¹, respectively (Table 2). A noteworthy improvement in affinity of the MIP was also marked with the use of OEG, compared to the imprinting system using Ni²⁺ only. The enhanced imprinting effect may be attributed to decreased nonspecific hydrophobic interactions since the hydrophile segments of OEG can result in increased hydrophilicity and thus less retention of analytes.²⁰ In addition, the results suggest improved accessibility of the imprinted cavities due to the fact that crosslinking density is decreased by introducing OEG segments into the matrix of the MIP. This can further be supported by comparing the number of binding sites of the OEG MIP (C11) and OEG-free MIP (C15). The corresponding numbers of selective binding sites of the Ni²⁺-mediated OEG MIP, Ni²⁺ mediated OEG-free MIP and Ni²⁺-free OEG MIP were 32.2, 50.8, and 64.8 μmol g⁻¹, respectively. This means that the number of binding sites of high affinity in the Ni²⁺ mediated OEG MIP was about 2 times less than that in the Ni²⁺-free OEG MIP. In addition, the number of non-selective binding sites in the former was about 2 times more than that in the latter. This leads us to a conclusion that Ni²⁺ plays a crucial role in increasing the specificity and reducing the amount of non-selective sites.

Table 2 Adsorption parameters of different MIPs

Column	Freundlich fitting			Scatchard–Rosenthal analysis
	n	Kf (mmol L^−1)	R^2	Kd (mol g^−1)	Q (μmol g^−1)	R^2
C11-MIP	0.87	0.452	0.998	Kd1 = 9.26 × 10^−4	Q1 = 32.2	0.993
				Kd2 = 5.03 × 10^−3	Q2 = 114.5	0.975
C15-MIP	0.65	0.0028	0.990	Kd1 = 5.49 × 10^−3	Q1 = 50.8	0.994
				Kd2 = 7.14 × 10^−3	Q2 = 65.0	0.990
C5-MIP	0.67	0.793	0.988	Kd1 = 1.24 × 10^−3	Q1 = 64.8	0.996
				Kd2 = 4.63 × 10^−3	Q2 = 208.6	0.999

---

To describe the surface heterogeneity of the adsorption process, Freundlich isotherm is often used to analyze the data of adsorption on MIPs.³⁸ The equation is frequently used in a linear form as

lg Q_e = n lg C_e + lg K_f (1)

where C_e is the equilibrium concentration (mmol L⁻¹), Q_e is the amount of GA adsorbed at equilibrium (mmol L⁻¹), K_f is the isotherm constant (mmol L⁻¹), and n is the heterogeneity index. In most cases, the imprinted polymer has a higher degree of heterogeneity, i.e., a lower heterogeneity index, n, than its corresponding non-imprinted control.³⁸ In our study, the heterogeneities of the imprinted polymers were compared by the fitting constant K_f and n of the MIPs with the linear form of eqn (1) (Fig. 10). The results showed that the Ni²⁺-mediated OEG MIP had a higher heterogeneity index (n = 0.87), suggesting the affinity sites of the MIP are more homogeneous. In contrast, the Ni²⁺ mediated OEG-free MIP and the Ni²⁺-free OEG MIP showed lower heterogeneity indices with values of 0.65 and 0.67, respectively.

Fig. 10 Freundlich analysis for Ni²⁺-mediated OEG MIP (C11), Ni²⁺-mediated OEG-free MIP (C15) and Ni²⁺-free OEG MIP (C5).

Conclusions

A new strategy of combining a metallic pivot and hydrophilic macromonomer was successfully developed to prepare GA-imprinted monoliths. The greatest molecular recognition ability towards the imprint species can be achieved at a molar ratio of OEG to 4-VP of 2 : 1, with good water compatibility. It was demonstrated that the affinity of the resulting MIP is a function of the molar ratio of metal ion to GA. The study of the binding characteristics of the OEG MIP monolith showed that the number of affinity sites can be significantly shifted with the introduction of metal ions. Moreover, a higher heterogeneity index indicated that the affinity sites of the metal ion-mediated OEG MIP monolith were more homogeneous. In conclusion, the approach presented here may be an effective method to prepare MIPs for water-soluble templates with both good selectivity and less hydrophobic character.

Experimental

Reagents and chemicals

Gallic acid (GA), salicylic acid (SA), m-hydroxybenzoic acid (MHA), and p-hydroxybenzoic acid (PHA), were obtained from Shanghai Guangtuo Chemical Reagent (Shanghai, China). Methyl gallate (MG) was obtained from Beijing Bailingwei Chemical Reagent (Beijing, China). 4-Vinylpyridine (4-VP), ethylene glycol dimethacrylate (EDMA) and oligo(ethyleneglycol) methyl ether methacrylate (OEG) (M_n = 300 gmol⁻¹, mean degree of polymerization 4–5) were purchased from Sigma (St. Louis, MO, USA). Dimethyl sulfoxide (DMSO) and dimethylformamide (DMF) were obtained from Tianjin Jiangtian Chemical Industry Reagent (Tianjin, China). 1-Butyl-3-methylimidazolium tetrafluoroborate ([BMIM]BF₄) was purchased from Shanghai Chengjie Chemical Reagent (Shanghai, China). Nickel acetate and 2,2-azobisisobutyronitrile (AIBN) were supplied by Kermel Chemical Reagent (Tianjin, China). HPLC-grade acetonitrile (ACN) was obtained from Tianjin Biaoshiqi Chemical Reagent (Tianjin, China). Other analytical reagents were from Tianjin Chemical Reagent Ltd. Co. (Tianjin, China).

Preparation of MIP monolithic columns

The preparation of the GA-imprinted monoliths was carried out as follows: a pre-polymerization mixture was prepared by mixing GA, 4-VP, nickel acetate, EDMA, DMSO, DMF, [BMIM]BF₄ and AIBN, as shown in Table 1. The pre-polymerization mixture was sonicated for 15 min and introduced into a stainless steel column (100 mm × 4.6 mm). The ends of the column were sealed and the column was submerged in a 60 °C water bath for 18 or 24 h. After polymerization, the column was flushed with acetonitrile to remove any unreacted reagents. Subsequently, the resulting monolithic column was washed with a mixture of methanol and acetic acid (9 : 1, v/v) until no template molecules were detected in the extraction solvent. A non-imprinted column (NIP) was prepared similarly in the absence of GA.

Chromatographic evaluation

High performance liquid chromatography was performed on an Agilent 1100 series chromatographic system consisting of a G1311A pump, a G131513 DAD detector, a Rheodyne 7225 injector with a 20 μL loop, and a Vertex VT4820 temperature controller. Data processing was carried out using a HPCORE workstation. The detection was performed at 271 nm with a flow rate of 0.5 mL min⁻¹. All of the mobile phases were filtered through a 0.22 μm membrane from Millipore before use. The column void volume was measured by injecting 20 μL of acetone (0.1%, v/v) in a mobile phase of acetonitrile–acetate buffer (pH 3.6) (70/30, v/v).

The retention factor (k) is calculated as (t_R − t₀)/t₀, where t_R is the retention time of the eluted substance and t₀ the retention time of the void marker. The imprinting factor (IF) is calculated as IF = k_MIP/k_NIP, where k_MIP is the retention factor of the template molecule eluted from the imprinted polymer and k_NIP is the retention factor of the template molecule eluted from the non-imprinted polymer.²⁷

Frontal analysis

The binding capacities of the imprinted and non-imprinted monoliths were investigated using the frontal chromatography method. The elution was monitored at 271 nm and the mobile phase was methanol with a flow-rate of 1.0 mL min⁻¹. A series of different concentrations of GA (0.1, 0.15, 0.2, 0.3, 0.4, 0.5, 0.6 and 0.7 mmol L⁻¹) were prepared in the mobile phase and loaded onto the imprinted monolithic columns at 25 °C. The break-through curves were first generated and the retention time was obtained from the time at half-height of the break-through curve. The number of binding sites (L_t) and the dissociation constant (K_d) are calculated using the following equation:³⁰

1/{[A]₀(V − V₀)} = K_d/([A]₀L_t) + 1/L_t (2)

where [A]₀ is the concentration of the analyte. V and V₀ are the elution volumes of the analyte and the void marker, respectively, which are calculated from the void time (acetone) and retention time of template multiplied by the flow rate. L_t and K_d can be calculated from the intercepts on the ordinate and the slope based on the plots of 1/{[A]₀(V − V₀)} versus 1/[A]₀.

The adsorption quantity of equilibrium in the frontal analysis, Q, is calculated according to:³⁰

where C is the sample concentration, V_equ is the volume when the adsorption is balanced, and V_a is the stationary phase volume.

Acknowledgements

This work was supported by the High Technology Research and Development Program of Xinjiang (no. 201217149).

References

1. E. Saridakis and N. E. Chayen, Trends Biotechnol., 2013, 31, 515.
2. Y. Ge, B. Butler, F. Mirza, S. Habib-Ullah and D. Fei, Macromol. Rapid Commun., 2013, 34, 903.
3. B. Tóth and G. Horvai, Top. Curr. Chem., 2012, 325, 267.
4. A. Rahmani, A. H. Mohammadpour, A. Sahebnasagh and S. A. Mohajeri, Anal. Bioanal. Chem., 2014, 406, 7729.
5. C. Zheng, Y. P. Huang and Z. S. Liu, Anal. Bioanal. Chem., 2013, 405, 2147.
6. B. C. Iacob, E. Bodoki and R. Oprean, Electrophoresis, 2014, 35, 2722.
7. X. N. Wang, R. P. Liang, X. Y. Meng and J. D. Qiu, J. Chromatogr. A, 2014, 1362, 301.
8. X. Shen, L. Zhu, N. Wang, L. Ye and H. Tang, Chem. Commun., 2012, 48, 788.
9. M. M. Moein, D. Jabbar, A. Colmsjö and M. Abdel-Rehim, J. Chromatogr. A, 2014, 1366, 15.
10. A. Tieppo, C. J. White, A. C. Paine, M. L. Voyles, M. K. McBride and M. E. Byrne, J. Controlled Release, 2012, 157, 391.
11. F. Puoci, G. Cirillo, M. Curcio, O. I. Parisi, F. Iemma and N. Picci, Expert Opin. Drug Delivery, 2011, 8, 1379.
12. G. Wulff and J. Liu, Acc. Chem. Res., 2012, 45, 239.
13. A. Yarman and F. W. Scheller, Angew. Chem., Int. Ed. Engl., 2013, 52, 11521.
14. M. P. Tiwari and A. Prasad, Anal. Chim. Acta, 2015, 853, 1.
15. Y. Fuchs, O. Soppera and K. Haupt, Anal. Chim. Acta, 2012, 717, 7.
16. C. Malitesta, E. Mazzotta, R. A. Picca, A. Poma, I. Chianella and S. A. Piletsky, Anal. Bioanal. Chem., 2012, 402, 1827.
17. G. Wulff, Angew. Chem., Int. Ed. Engl., 1995, 34, 1812.
18. T. Kubo, K. Hosoy, M. Nomachi, N. Tanaka and K. Kaya, Anal. Bioanal. Chem., 2005, 382, 1698.
19. J. Cao, X. Zhang, X. He, L. Chen and Y. Zhang, Chem.–Asian. J., 2014, 9, 526.
20. A. Pardo, L. Mespouille, P. Dubois, B. Blankert and P. Duez, Chem.–Eur. J., 2014, 20, 3500.
21. J. L. Urraca, M. C. Moreno-Bondi, A. J. Hall and B. Sellergren, Anal. Chem., 2007, 79, 695.
22. P. Manesiotis, Q. Osmani and P. McLoughlin, J. Mater. Chem., 2012, 22, 11201.
23. J. Matsui, S. Goji, T. Murashima, D. Miyoshi, S. Komai, A. Shigeyasu, T. Kushida, T. Miyazawa, T. Yamada, K. Tamaki and N. Sugimoto, Anal. Chem., 2007, 79, 1749.
24. X.-X. Li, L.-H. Bai, H. Wang, J. Wang, Y.-P. Huang and Z.-S. Liu, J. Chromatogr. A, 2012, 1251, 141.
25. L. N. Mu, X. H. Wang, L. Zhao, Y. P. Huang and Z. S. Liu, J. Chromatogr. A, 2011, 1218, 9236.
26. S. J Li, C. Liao, W. Li, Y. Chen and X. Hao, Macromol. Biosci., 2007, 7, 1112.
27. L. Zhao, L. Ban, Q. W. Zhang, Y. P. Huang and Z. S. Liu, J. Chromatogr. A, 2011, 1218, 9071.
28. G. Qu, S. Zheng, Y. Liu, W. Xie, A. Wu and D. Zhang, J. Chromatogr. B: Anal. Technol. Biomed. Life Sci., 2009, 877, 3187.
29. D.-D. Zhong, Y.-P. Huang, X.-L. Xin, Z.-S. Liu and H. A. Aisa, J. Chromatogr. B: Anal. Technol. Biomed. Life Sci., 2013, 934, 109.
30. X. Sun, C.-Y. Zhao, X.-H. Wang, Y.-P. Huang and Z.-S. Liu, Anal. Bioanal. Chem., 2014, 406, 5359.
31. K. S. W. Sing, Pure Appl. Chem., 1982, 54, 2201.
32. Y. P. Huang, S. J. Zhang, X. Wu, Q. W. Zhang and Z. S. Liu, Chromatographia, 2009, 70, 691.
33. H. S. Andersson, J. G. Karlsson, S. A. Piletsky, A.-C. Koch-Schmidt, K. Mosbach and I. A. Nicholls, J. Chromatogr. A, 1999, 848, 39.
34. G. Wulff, Chem. Rev., 2002, 102, 1.
35. A. G. Mayes and K. Mosbach, Anal. Chem., 1996, 68, 3769.
36. B. Sellergren and K. J. Shea, J. Chromatogr. A, 1993, 635, 31.
37. B. Sellergren, J. Chromatogr. A, 2001, 906, 227.
38. A. M. Rampey, R. J. Umpleby II, G. T. Rushton, J. C. Iseman, R. N. Shah and K. D. Shimizu, Anal. Chem., 2004, 76, 1123.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra02699g

---