A Novel SPE Method with Two MIPs by Two Steps for Improving

May 27, 2019 - Blank image Log In · logo Journal Logo. PREVIOUS .... A Novel SPE Method with Two MIPs by Two Steps for Improving Selectivity of MIPs ...
0 downloads 0 Views 1MB Size
Article pubs.acs.org/ac

Cite This: Anal. Chem. 2019, 91, 8436−8442

A SPE Method with Two MIPs in Two Steps for Improving the Selectivity of MIPs Yike Huang,† Jingmiao Pan,† Yi Liu,† Min Wang,† Suya Deng,† and Zhining Xia*,† †

School of Pharmaceutical Sciences and Innovative Drug Research Centre, Chongqing University, Chongqing, 401331, P. R. China

Downloaded via NOTTINGHAM TRENT UNIV on July 24, 2019 at 00:57:53 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

S Supporting Information *

ABSTRACT: Molecular imprinted polymers (MIPs) have been widely applied in the separation of compounds in complex matrixes due to the high selectivity for molecular recognition. However, MIPs not only adsorb the targeted molecule but also adsorb structurally similar analogues, which leads to some loss of selectivity. In this work, for improvement of selectivity of MIPs, a novel solid-phase extraction method with two MIPs in two steps (TMIPs-TSPE) was established. As a demonstration, two MIPs were prepared by using quercetin as the template and 4-vinylypyridine (4VP) and acrylamide (AM) as representative functional monomers, respectively. The adsorption properties and kinetic characteristics of the two MIPs showed that they had a distinct adsorption capacity and adsorption mechanism, which is the basis for establishment of TMIPs-TSPE. The TMIPs-TSPE method first used one of the two MIPs as adsorbent to extract molecules from a solution mixture containing quercetin and three analogues. Then the other MIP was used to achieve a second extraction of the extracted molecules from the first step. The results showed that the unique targeted molecule quercetin was extracted, which illustrates that TMIPs-TSPE improved the specificity of the MIPs. The process of molecular recognition can be influenced by the intensity of binding sites between MIPs and molecules. Moreover, it may also depend on the spatial orientation of molecules entering the cavities of MIPs, which deserves more attention as one important property for the development of molecular imprinting. These results demonstrated that the novel TMIPs-TSPE method contributes to the improved selectivity of MIPs.

M

olecular imprinting, established in 1972 by Wulff and Sarhan,1 is one significant approach for molecular recognition.2−4 Molecularly imprinted polymers (MIPs) as adsorbents have been widely applied in the separation of compounds from complex mixtures with high selectivity.5−9 MIPs are produced by a copolymerization process involving functional monomers, cross-linkers, and a template molecule. Based on Pauling’s theory of bioimprinting10 and the lock-andkey concept,11,12 MIPs form a highly three-dimensional imprint cavity with a defined shape and multiple recognition sites.9,13 However, MIPs not only adsorb imprinted molecules but can also adsorb analogues having a molecular weight, electric density, and spatial structure similar to that of the template.14−17 Therefore, the selectivity of MIPs for targeted molecules is not yet satisfactory. The recognition mechanism of MIPs to imprinted molecules is mainly associated with the matching degree on binding sites and the spatial structure. In other words, the selectivity of MIPs depends on two factors: (1) the intensity of binding sites between template and monomer; (2) the shape and rigidity of template. For better understanding, Nicholls18 et al. suggest that the physical mechanisms underlying formation and recognition events to MIPs−ligand are ruled by thermodynamic theory as described in the following equation. © 2019 American Chemical Society

ΔG bind = ΔGt + r + ΔGr + ΔG h + ΔGvib + ∑ ΔGp + ΔGconf + ΔGvdw

where the Gibbs free energy changes are as follows: ΔGbind, complex formation; ΔGt+r, translational and rotational; ΔGr, restriction of rotors upon complexation; ΔGh, hydrophobic interactions; ΔGvib, residual soft vibrational modes; ∑ΔGp, the sum of interacting polar group contributions; ΔGconf, comformational changes; ΔGvdw, van der Waals interactions. The Gibbs free energy changes are influenced by various energetic contributions to binding. Thus, various MIPs that recognize targeted molecules have different Gibbs free energies. Meanwhile, these MIPs that recognize the analogues also have different Gibbs free energies, which is the basis to develop a novel method for improving selectivity of MIPs. On the one hand, in the imprinting process, the monomer− template complex is formed through covalent or noncovalent interactions.19 Noncovalent bonds mainly include hydrogen bonds, ionic bonds, metal coordination bonds, hydrophobic interactions, and van der Waals interactions. The stability of the complex depends on both the amount of the binding sites and the type of binding interaction.13 Thus, MIPs with Received: March 21, 2019 Accepted: May 27, 2019 Published: May 27, 2019 8436

DOI: 10.1021/acs.analchem.9b01453 Anal. Chem. 2019, 91, 8436−8442

Article

Analytical Chemistry

acrylamide (AM) as monomers, respectively. The adsorption properties of the two MIPs were revealed by comparing their molecular recognition ability. The recognition mechanism was analyzed by using the Scatchard equation. Considering the different absorption abilities and possible mechanism, the TMIPs-TSPE method used the designed MIPs as adsorbents to extract molecules in the matrix mixture containing quercetin and its structural analogues (i.e., rutin, luteolin, and kaempferol shown in Figure S1). Two steps of solid-phase extraction (SPE) were involved in this method. First, one of the two MIPs (MIP-1) was employed to extract molecules from the solution mixture through SPE. Subsequently, the other MIP (MIP-2) was used to further extract the molecules eluted from MIP-1. The eluent was analyzed by high performance liquid chromatography (HPLC). In addition, the effect of the concentrations of the solution mixture and the order of exaction on the extraction efficiency were studied to understand the recognition mechanism of TMIPs-TSPE.

different monomers will have different adsorption mechanisms to separate species, which offers the potential for improving the selectivity of imprinted molecules. On the other hand, during the synthesis of MIPs, the threedimensional cavity formed depends on the steric structure of the template itself. The shape of the cavity and the precise spatial arrangement of functional groups in MIPs are complementary to the spatial structure of the template.20 During recognition, it is significant whether or not molecules can enter the cavity through an available orientation (such as quercetin in Figure 1). In other words, different orientations of



EXPERIMENTAL SECTION Preparation of the MIPs and Nonmolecular Imprinted Polymers (NIPs). MIPs for quercetin were prepared by a noncovalent approach through bulk polymerization by using 4-vinylpyridine (4VP) and acrylamide (AM) as functional monomers, respectively, ethylene glycol dimethacrylate (EGDMA) as the cross-linker, and 2,2′-azobis(2-methylpropionitrile) (AIBN) as the initiator in tetrahydrofuran (THF). Briefly, 1 mmol of quercetin and 6 mmol of 4VP or AM were dissolved in 5 mL of THF solvent for prepolymerization. After oscillation for 4 h, 25 mmol of EGDMA and 10 mg of AIBN were added. The mixture was degassed in a sonicating bath for 15 min, filled with nitrogen for 10 min, and then thermally polymerized at 60 °C for 24 h. After that, hard, light yellow, solid polymers were formed, ground and sieved through a 200mesh filter. The powder was washed with MeOH/HAc (8/2, v/v) to remove quercetin and other reagents until there was no UV absorption at 371 nm. The collected polymers were dried in a vacuum oven at 60 °C. As experimental references, NIPs without template molecules were prepared under the same procedure by using the same above-mentioned method. Characterization of MIPs and NIPs. The morphology of MIPs was observed using a Zeiss Merlin VP compact scanning electron microscope (SEM, ZEISS, Germany). The FTIR spectra of the MIPs and NIPs were obtained using a FTIR spectrometer (Tensor 27, Bruker, Billerica, MA). All samples were mixed thoroughly with KBr and pressed into tablets. The pellets were measured in the range of 4000−400 cm−1. Static and Dynamic Adsorption Determination of MIPs and NIPs for Quercetin. In the static adsorption experiment, 10 mg of MIPs or NIPs was added into 10 mL tubes, respectively, and mixed with 3 mL of methanol solution containing known concentrations of quercitin. Then the mixtures were shaken at 90 rpm for 6 h. The mixtures were centrifuged at 8000 rpm for 15 min, and the supernatant solutions of quercetin were determined by using absorbance of 371 nm.26 The equilibrium adsorption capacity (Q, μg/g) was calculated as follows:

Figure 1. Diagram of the template molecule and its absorption on MIPs. The computer model of quercetin observed from front view and top view (A1 and B1) and the virtual graph of the structure of quercetin and its analogues entering the cavities of the MIPs in different spatial orientations.

functional groups of binding sites inside cavities can affect the recognition ability of MIPs. According to Wulff’s two-point binding theory,20 there are two bindings between templates and monomers in the imprinted cavity. The template is first bound through one functional group of the monomer only (one-point binding), which then changes to stable two-point binding. Only the template molecule can fit well in the cavity. In addition, the arrangement of the functional groups of the binding sites within the cavity is also a decisive factor for molecular recognition.21 At present, the property of spatial orientation of molecules has been demonstrated with chiral molecules and enantiomers,22−25 while little has been accomplished with nonchiral molecules to improve the specificity of molecules. Thus, we hypothesize that various functional monomers on MIPs have different orientations and locations inside the cavities or the interspace of MIPs, which provides the basis for templates to enter cavities at different orientations. We suggest that the energy of MIPs recognition of molecules has a vector property. On the basis of the above theory, we emphasize two essential properties: (1) the intensity of interaction between monomers and template; (2) the spatial effects inside the cavity of MIPs. Meanwhile, we expect more precise selectivity for molecular recognition such as in a “fingerprinting lock”. To verify the aforementioned theoretical hypothesis and realize the expectation in this work, two MIPs with different monomers are considered. They have differences in interaction intensity with molecules and the expected orientations of template binding to MIPs. Thus, a novel method based on two steps of SPE with two MIPs as adsorbents (TMIPs-TSPE) is developed. Specifically, two MIPs were synthesized by using quercetin as template (Figure S1) and 4-vinylypyridine (4VP) and

Q=

(C0 −Ce) × V m

where C0 (μg/mL) and Ce (μg/mL) are the initial and equilibrium concentration of quercetin, respectively, V (mL) is 8437

DOI: 10.1021/acs.analchem.9b01453 Anal. Chem. 2019, 91, 8436−8442

Article

Analytical Chemistry

Figure 2. Procedure of the TMIPs-TSPE method.

the volume of quercetin solution, and m (g) is the mass of the adsorbent used (MIPs materials). The Scatchard plot equation was used for further evaluation of the binding property of MIPs.

sample and the two elution solutions were dried, redissolved with 1 mL of MeOH, and analyzed using HPLC. The control groups (MIP-1-MIP-1 and MIP-2-MIP-2) were also tested by using the same MIP as adsorbent in the two step SPE. The procedure for the control groups followed the above process. The relative peak area percentages of these compounds were calculated by the following equation:

Q −Q Q = max C Kd

percentage of relative peak area =

where Q is the adsorption capacity at the adsorption equilibrium, C is the equilibrium concentration, Qmax is the apparent maximum adsorption capacity, and Kd is the dissociation constant. In the dynamic adsorption experiment, 10 mg of MIPs or NIPs was added into 10 mL tubes, respectively, and mixed with 3 mL of 15 μg/mL quercetin methanol solution. The solutions were shaken at 90 rpm for 0.25, 0.5, 1, 2, 4, 6, and 8 h, respectively. The following procedures were the same as for the static adsorption experiment. Two Steps of SPE with Two MIPs (TMIPs-TSPE) Procedure for Templates. To demonstrate the efficiency of our novel method, the procedure called TMIPs-TSPE, for the extraction of quercetin from the solution mixture, was designed as follows (Figure 2): A mixture containing quercetin and its structurally similar molecules (rutin, luteolin, and kaempferol) was dissolved in methanol solution. First, 100 mg of one MIP (MIP-1) was added into tubes with the aforementioned mixture at concentrations of 0.0125 or 0.05 mM and then shaken at 90 rpm for 4 h at 30 °C. After that, MIP-1 was washed by using 3 mL of methanol at 90 rpm for 30 min and then the solution was removed. Second, 3 mL of MeOH/HAc (8/2, v/v) solution was added and shaken at 90 rpm three times (1 h/time) to obtain the compounds which had interacted with functional monomers in the MIPs. The samples in the tubes were centrifuged, and the supernatant solution was dried by nitrogen and redissolved using 3 mL of MeOH solution as elution solution 1 (E1) which was used as adsorption solution for the next SPE. The following procedure using another MIP (MIP-2) was the same as above. After the MIP-2 adsorption process, the eluent using MeOH solution was named as elution solution 2 (E2). Finally, the initial

Ax × 100% A1 + A 2 + A3 + A4

where A is the peak area, x is one of the compounds (quercetin, rutin, luteolin, and kaempferol), and 1, 2, 3, and 4 are quercetin (QC), rutin (RT), luteolin (LTL), and kaempferol (KPF), respectively.



RESULTS AND DISCUSSION Preparation and Characterization of MIPs. Quercetin was dissolved in THF which was the usual porogen with better dissolving capacity.27,28 The two selected monomers (4VP and AM) were supposed to have different binding sites and intensities with the template and spatial orientations inside the cavity of the MIPs after polymerization. The SEM morphologies (Figure S2) show that MIP-4VP and MIP-AM had similar regular spherical particles with homogeneous distributions. These particles were tightly connected and formed irregular clusters, and overall the MIPs had some porous structures, similar to that previously reported.29 In addition, the average diameters for the two MIPs particles seemed to be slightly different, which suggests some influences of the different monomers on the particle growth during the polymerization process.30 FTIR spectra of the MIPs and NIPs are shown in Figure S3. Generally, the two kinds of MIPs and NIPs had some similarities in the fingerprint region. The stretching vibration peak of methyl or methylene (C−H) was found at 2950 and 2962 cm−1 for MIP-4VP and MIP-AM, respectively. The peaks at 1718 and 1726 cm−1 were attributed to the stretching vibration absorbance of CO, which originated from EGDMA and monomers of 4-VP and AM. The peak at 1670 cm−1 was attributed to C−C stretching vibration in MIP-AM. 8438

DOI: 10.1021/acs.analchem.9b01453 Anal. Chem. 2019, 91, 8436−8442

Article

Analytical Chemistry

concentration. This is because the NIPs have no special spatial structure, leading to nonspecific adsorption capacity. Moreover, through observing the adsorption curve (Figure 3, MIP-4VP and MIP-AM), the differences in recognition molecule capacity between the two MIPs were compared. The maximum equilibrium adsorption capacity of QC on MIP-4VP and MIP-AM was 397 and 283 μg/g, respectively. These results demonstrated that the prepared MIPs with different functional monomers had different adsorption capacities at the same QC concentration. In addition, MIP-4VP showed an adsorption capacity for QC higher than that of MIP-AM, which reflected better specificity.30 The monomer 4VP is an alkaline molecule with a pyridine ring which can interact with the acidic groups of quercetin to form a stable complex.27,33 In MIP-AM, the functional monomer AM is alkalescent. The amino group with electron absorption capacity can form a relatively weak bond with the phenolic hydroxyl group of QC, which lowers the binding intensity of MIP-AM compared with that of MIP-4VP. Thus, different monomers may have distinct binding intensities with the template. To understand the binding sites of adsorbents and theoretical binding amounts for quercetin,34 the Scatchard plots for the two MIPs were analyzed in Figure 4 and Table S1.

The characteristic signals of the NIPs were similar to those of the MIPs, which indicated that EGDMA and monomers probably formed the polymers without specific structure. There were some differences in the FTIR spectra of the two MIPs. MIP-4VP showed a broad peak at 3427 cm−1 which was assigned to the asymmetric stretching vibration of N−H formed by the interaction between the pyridine ring of 4VP and the phenolic hydroxyl group of quercetin. The peaks at 1633 and 1670 cm−1 were attributed to the CC stretching vibration of the pyridine ring of 4VP. This demonstrated that EGDMA and 4VP had grafted on the MIPs. MIP-AM showed a peak at 3462 cm−1 attributed to N−H. The peak at 1396 cm−1 was the flexural vibration of O−H in OC-NH2, which indicates that the functional monomers of acrylamide had been conjugated in the polymers. The results obviously confirmed that there were functional groups which interacted with the template, namely the acylamide group in the MIPs.28 Adsorption Properties of MIPs to Quercetin. Adsorption determination can illuminate molecular recognition behavior of the synthesized polymers; thus, the experimental conditions of SPE were investigated. Static adsorption experiments were conducted by using different concentrations (from 2.5 to 30 μg/mL) of quercetin in methanol. Adsorption isotherms of MIPs (Figure 3) showed that the amounts of quercetin bound to the MIPs at equilibrium

Figure 4. Scatchard analysis of MIP-4VP (A) and MIP-AM (B). Figure 3. Adsorption isotherm of MIP/NIP-4VP (A) and MIP/NIPAM (B).

As shown for MIP-4VP, there are two distinct sections within the plots regarded as straight lines, which indicates that MIP4VP possesses two binding sites (high and low affinity sites).35 Meanwhile, the Kd (0.74 and 2.42 μg/mL) and Qmax (26.48 and 66.27 μg/g) values of MIP-4VP (Table S1) demonstrated that the binding sites in MIP-4VP were heterogeneous with respect to the affinity of QC.36 In addition, the slope of plot a1 (−1.35) was lower than that of plot a2 (−0.42), which indicated that the amounts of high affinity sites were higher in MIP-4VP,37 while the Scatchard plot for MIP-AM was only a single straight line. The linear regression equation was Q/C = −0.42Q + 19.54 (R2 = 0.94, b1), suggesting that homogeneous recognition sites for quercetin were formed in MIP-AM.28,38

increased with the rise in concentration until it reached a saturation level (about 20 μg/mL of QC). The results indicated that specific binding was the main factor due to a limited number of binding sites in the cavity. The adsorption amounts for MIPs (MIP-4VP and MIP-AM) were much more for various concentrations of QC compared with that of NIPs (NIP-4VP and NIP-AM), respectively, suggesting that the MIPs with specific cavities showed higher affinity.31,32 Further, the adsorption amounts for NIP-4VP and NIP-AM gradually reached saturated adsorption with the increase in QC 8439

DOI: 10.1021/acs.analchem.9b01453 Anal. Chem. 2019, 91, 8436−8442

Article

Analytical Chemistry Kd and Qmax were calculated to be 5.14 μg/mL and 78.84 μg/g, respectively (MIP-AM, Table S1 b1). Comparison of the parameters of the Scatchard equation of the two MIPs showed different bindings, which indicated that each MIP may have a different absorption mechanism. Especially, the Kd values of MIP-4VP in a1 and MIP-AM in b1 were 2.43 and 2.38 μg/mL, respectively, which indicated that the two MIPs may depend on hydrogen bonding between functional monomers and quercetin, facilitated by the five phenolic hydroxyl groups in the quercetin molecule.39 Other binding sites were shown in MIP-4VP, which indicates that other recognition mechanisms may exist. In addition, electrostatic and π−π interactions were expected to form better recognition properties.40 Some weak interactions between functional monomers and quercetin such as van der Waals interaction may influence the recognition.28 The dynamic curves for the adsorption of quercetin onto the MIPs were evaluated in Figure 5, where Q was the amount of

only quercetin but also its analogues, similar to that reported previously,42 and thus indicated unsatisfactory specificity for the target molecule. When using MIP-4VP or MIP-AM as adsorbents for the first SPE, the relative peak area percentages of quercetin were 37.5% and 14.5%, respectively, which showed that the absorption capacity of MIP-4VP was better than that of MIP-AM. This indicated that the intensity of binding sites in the recognition process of MIPs was influenced by different functional monomers on the MIPs. Meanwhile, the different template−monomer complexes (QC-4VP and QCAM) formed may have different Gibbs free energies.39 However, the absorption capacity of MIPs has been significantly improved after the second extraction in TMIPsTSPE by using the other MIP, which indicated that some unknown factors may exist. At the beginning of recognition of MIPs (under the predictions of noncovalent binding and excess monomers with different orientations on MIPs), molecules with the same main structure (rutin and luteolin) randomly entered cavities and freely interacted with various oriented functional monomers on the MIPs (see Figure 1), which may lead these molecules to enter the cavities in incorrect spatial orientations and cause (according to Wulff’s two-point binding theory) specific recognition to fail. Only QC could fit in the cavity of MIPs under the correct spatial orientation. In contrast, the control group (Figure S4, Table S2) revealed that using the same two MIPs in TMIPs-TSPE did not bring an improvement in selectivity. Thus, we deem that the energy of MIPs recognition has a vector property, which can realize better selectivity such as in a “fingerprinting lock”. The desired results have been obtained when using MIP4VP as adsorbent for the first SPE and MIP-AM as adsorbent for the second SPE. To fully explore the outstanding advantages and influence factors of sequence properties in the TMIPs-TSPE method, the extraction order of the method was considered. As shown in Figure 6c and 6f, by changing the order of TSPE from MIP-AM to MIP-4VP, the relative percentages of peak area of quercetin were 100% and 66.6% (Table 1), respectively, which indicated that the order of TMIPs-TSPE could influence the selective recognition for quercetin. When using MIP-4VP (Figure 6b), no rutin was observed because of the large group substituted at the γ-OH position of quercetin, which conveys steric hindrance and precludes a fit in the cavity.27 In addition, 4-VP as the best monomer on MIP-4VP may help to interact with quercetin specially. Other analogues (luteolin and kaempferol) were also adsorbed through interaction with the cavity of MIPs owing to the steric effect of their similar main structure.43 After the second step using MIP-AM, only QC was retained, which may be based on the difference of interaction intensities between monomer and template and the orientations for cavity entrance. When exchanging the extraction order of MIP-4VP and MIP-AM, it was observed that MIP-AM could adsorb quercetin and all analogues at the first extraction step (Figure 6e). This may due to the different recognition mechanisms for each molecule. Once these molecules interacted with monomers in MIP-AM, nonspecific adsorption would increase. Thus, the order of extraction in TMIPs-TSPE should be one vital factor for improving the selectivity of MIPs. Because the adsorption capacity of MIPs was usually within a range of concentrations,44,45 the concentrations of the mixture needed to be considered. According to the results of the static adsorption, 0.0125 mM was applied as the low

Figure 5. Dynamic adsorption curves of MIP-4VP and MIP-AM.

adsorbed quercetin. It was observed that the adsorbed amounts of quercetin by the two MIPs showed increasing trends with an increase in time. Meanwhile, Q nearly reached saturation level at 2 h, which was similar to that previously reported by Zhang et al.41 Therefore, 6 h was enough time for the static adsorption experiment. Moreover, the equilibrium adsorption ability of MIP-4VP for quercetin was stronger than that of MIP-AM, which was in accord with static adsorption results. TMIPs-TSPE with Improved Selectivity. On the basis of the previous adsorption properties, differences were obtained for adsorption capacity, adsorption concentrations, adsorption equilibrium time, and adsorption mechanism of the two MIPs for quercetin. We designed a TMIPs-TSPE method based on MIPs with different adsorption properties. The key aspect of the TMIPs-TSPE method focused on the two steps of SPE. Thus, we chose two MIPs as adsorbents whose functional monomers (4VP and AM) had different acid−base properties, which would cause more differences in adsorption capacity and recognition mechanism. The procedure for TMIPs-TSPE is described in Experimental Section. Further, the factors of molecular recognition mechanism were investigated by changing the extraction order and mixture concentrations. As shown in Figure 6c, the highest peak in the HPLC chromatogram of elution 2 after two steps of extraction was quercetin and the relative peak area percentage was 100%, which was about 2.6 times higher than that in elution 1 (37.5%, Figure 6b). The results indicated that the selective adsorption capacity of TMIPs-TSPE was much better than that of MIPs-SPE, which is a common method through one step of extraction. As shown in Figure 6b and 6e, MIPs adsorbed not 8440

DOI: 10.1021/acs.analchem.9b01453 Anal. Chem. 2019, 91, 8436−8442

Article

Analytical Chemistry

Figure 6. HPLC chromatograms of samples subjected to different extraction orders and different concentrations of sample mixtures for TMIPsTSPE. The numbers 1, 2, 3, and 4 represent rutin, quercetin, luteolin, and kaempferol.

Table 1. Percentages of Relative Peak Area of the Four Compounds in HPLC Chromatograms for MIP-4VP and MIP-AMa a, b, c sample eluent 1 eluent 2

d, e, f

g, h, i

1

2

3

4

1

2

3

4

1

2

3

4

33.7 6.2

23.2 37.5 100.0

20.3 12.8

22.8 43.5

33.7 30.6 14.7

23.2 14.5 66.6

20.3 18.1 9.4

22.8 36.8 9.3

30.4 9.6 8.0

24.8 23.8 73.1

20.8 19.2 7.5

23.9 47.3 11.4

a

The letters (from a to i), respectively, represent each HPLC chromatogram shown in Figure 6.



CONCLUSIONS In this work, a novel TMIPs-TSPE method has been successfully developed. This method can largely eliminate structural analogues in a complex matrix and achieve more adsorption of targeted molecules. Compared with single MIPsSPE, the percentage of relative peak area of QC was 100% in the TMIPs-TSPE method, which shows more precise selectivity of the MIPs such as in a “fingerprint lock”. The advantage of the TMIPs-TSPE method is derived from the differences in the adsorption ability and adsorption mechanism of the two MIPs. Meanwhile, a different extraction order of the method can influence the molecular recognition. The molecular recognition mechanism may rely on two crucial factors: (1) the intensity of binding sites between functional monomers and imprinted molecules; (2) the spatial

concentration, and 0.05 mM (which was nearly the saturation concentration of absorption) was used as the high concentration condition. In Figure 6b and 6h, the change in relative peak area of the four compounds (RT, QC, LTL, and KPF) was examined and found to be similar at the two concentrations, which indicated that these concentrations were feasible after the first SPE. Compared with the HPLC chromatogram of the low concentration sample, there were some peaks of the adsorbed analogues in the HPLC chromatogram of the high concentration sample (Figure 6i). The high concentration mixture could possibly increase competitive adsorption between QC and its analogues, leading to more nonspecific binding. Therefore, it is necessary to choose a relatively lower concentration of solution mixture for the TMIPs-TSPE method. 8441

DOI: 10.1021/acs.analchem.9b01453 Anal. Chem. 2019, 91, 8436−8442

Article

Analytical Chemistry

(15) Wang, D.-D.; Gao, D.; Xu, W.-J.; Li, F.; Yin, M.-N.; Fu, Q.-F.; Xia, Z.-N. Talanta 2018, 184, 307−315. (16) Chen, F.-F.; Xie, X.-Y.; Shi, Y.-P. Talanta 2013, 115, 482−489. (17) Hiratsuka, Y.; Funaya, N.; Matsunaga, H.; Haginaka, J. J. Pharm. Biomed. Anal. 2013, 75, 180−185. (18) Nicholls, I. A.; Adbo, K.; Andersson, H. S.; Andersson, P. O.; Ankarloo, J.; Hedin-Dahlstrom, J.; Jokela, P.; Karlsson, J. G.; Olofsson, L.; Rosengren, J.; Shoravi, S.; Svenson, J.; Wikman, S. Anal. Chim. Acta 2001, 435, 9−18. (19) Wulff, G. Microchim. Acta 2013, 180, 1359−1370. (20) Wulff, G. Angew. Chem., Int. Ed. Engl. 1995, 34, 1812−1832. (21) Wulff, G.; Schauhoff, S. J. Org. Chem. 1991, 56, 395−400. (22) Liu, B.; Lian, H.; Chen, L.; Wei, X.; Sun, X. Anal. Biochem. 2019, 574, 39−45. (23) Duan, D.; Yang, H.; Ding, Y.; Li, L.; Ma, G. Electrochim. Acta 2019, 302, 137−144. (24) Rutkowska, M.; Plotka-Wasylka, J.; Morrison, C.; Wieczorek, P. P.; Namiesnik, J.; Marc, M. TrAC, Trends Anal. Chem. 2018, 102, 91− 102. (25) Xing, R.; Wang, S.; Bie, Z.; He, H.; Liu, Z. Nat. Protoc. 2017, 12, 964−987. (26) Silva, L. F. d.; Guerra, C. C.; Czermainski, A. B. C.; Ferrari, L.; Bergold, A. M. Pesqui. Agropecu. Bras. 2017, 52, 335−343. (27) Ersoy, S. K.; Tutem, E.; Baskan, K. S.; Apak, R.; Nergiz, C. J. Chromatogr. B: Anal. Technol. Biomed. Life Sci. 2016, 1017, 89−100. (28) Song, X.; Li, J.; Wang, J.; Chen, L. Talanta 2009, 80, 694−702. (29) Chrzanowska, A. M.; Poliwoda, A.; Wieczorek, P. P. Mater. Sci. Eng., C 2015, 49, 793−798. (30) Castro Lopez, M. d. M.; Cela Perez, M. C.; Dopico Garcia, M. S.; Lopez Vilarino, J. M.; Gonzalez Rodriguez, M. V.; Barral Losada, L. F. Anal. Chim. Acta 2012, 721, 68−78. (31) Wang, X.; Huang, P.; Ma, X.; Wang, H.; Lu, X.; Du, X. Talanta 2017, 166, 300−305. (32) Hu, Y.; Wang, C.; Li, X.; Liu, L. Electrophoresis 2017, 38, 2462− 2467. (33) Li, N.; Ng, T. B.; Wong, J. H.; Qiao, J. X.; Zhang, Y. N.; Zhou, R.; Chen, R. R.; Liu, F. Food Chem. 2013, 139, 1161−1167. (34) Matsui, J.; Miyoshi, Y.; Doblhoffdier, O.; Takeuchi, T. Anal. Chem. 1995, 67, 4404−4408. (35) Funaya, N.; Haginaka, J. J. Chromatogr. A 2012, 1248, 18−23. (36) Qu, Q.; Zhu, T. Anal. Methods 2017, 9, 6525−6533. (37) Asfaram, A.; Arabi, M.; Ostovan, A.; Sadeghi, H.; Ghaedi, M. New J. Chem. 2018, 42, 16144−16153. (38) Zhang, Z.; Yun, Y.; Li, C. Desalin. Water Treat. 2013, 51, 3914− 3919. (39) Yu, H.; Chen, Z.; Fu, Y.; Kang, L.; Wang, M.; Du, X. Polym. Int. 2012, 61, 1002−1009. (40) Saad, E. M.; Madbouly, A.; Ayoub, N.; El Nashar, R. M. Anal. Chim. Acta 2015, 877, 80−89. (41) Zhang, H.; Zhao, S.; Zhang, L.; Han, B.; Yao, X.; Chen, W.; Hu, Y. J. Sep. Sci. 2016, 39, 3098−3104. (42) Cheng, Y.; Nie, J.; Li, J.; Liu, H.; Yan, Z.; Kuang, L. Food Chem. 2019, 287, 100−106. (43) Hu, Y.; Xia, Q.; Huang, W.; Hou, X.; Tian, M. Microchim. Acta 2018, 185. DOI: 10.1007/s00604-017-2608-2 (44) Sun, X.; Wang, M.; Yang, L.; Wen, H.; Wang, L.; Li, T.; Tang, C.; Yang, J. J. Chromatogr. A 2019, 1586, 1−8. (45) Heravizadeh, O. R.; Khadem, M.; Nabizadeh, R.; Shahtaheri, S. J. Chem. Pap. 2018, 72, 3057−3068.

orientation of molecules when entering the cavities of MIPs. Both factors contribute to the improvement of selectivity. Thus, these results can support our hypothesis that the property of molecular spatial orientation can be applied. In addition, two factors should be considered in the TMIPs-TSPE method: First, there was noncovalent binding between templates and functional monomers in MIPs; second, the adsorption capacity of MIPs should not reach saturation level. In view of this method’s shortcomings, the recognition mechanism of TMIPs-TSPE should be further analyzed.



ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.analchem.9b01453. Materials and reagents, apparatus, structural formula of quercetin and its analogues, SEM images of MIP-4VP and MIP-AM particles at a 500 and 100 nm scale, FT-IR spectra of MIPs and NIPs prepared by different monomers of 4-VP and AM, HPLC chromatograms of the control groups subjeced to MIP-4VP-MIP-4VP and MIP-AM-MIP-AM for TMIPs-TSPE, linear regression data of MIP-4VP and MIP-AM, and percentages of relative peak area of the four compounds in HPLC chromatograms for the control groups (PDF)



AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Tel: +86 138 830 77188/Fax: +23 656 780 06. ORCID

Min Wang: 0000-0002-5756-1015 Zhining Xia: 0000-0002-2064-5880 Notes

The authors declare no competing financial interest.

■ ■

ACKNOWLEDGMENTS This work was supported by the National Natural Science Foundation of China (21705012, 21675016). REFERENCES

(1) Wulff, G.; Sarhan, A. Angew. Chem., Int. Ed. 1972, 11, 341−343. (2) Chen, L.; Wang, X.; Lu, W.; Wu, X.; Li, J. Chem. Soc. Rev. 2016, 45, 2137−2211. (3) Cheong, W. J.; Yang, S. H.; Ali, F. J. Sep. Sci. 2013, 36, 609−628. (4) Komiyama, M.; Mori, T.; Ariga, K. Bull. Chem. Soc. Jpn. 2018, 91, 1075−1111. (5) Xu, W.; Wang, Y.; Wei, X.; Chen, J.; Ni, R.; Meng, J.; Zhou, Y.; Xu, P. Anal. Chim. Acta 2019, 1048, 1−11. (6) Speltini, A.; Scalabrini, A.; Maraschi, F.; Sturini, M.; Profumo, A. Anal. Chim. Acta 2017, 974, 1−26. (7) Kubo, T.; Otsuka, K. J. Pharm. Biomed. Anal. 2016, 130, 68−80. (8) Schirhagl, R. Anal. Chem. 2014, 86, 250−261. (9) Martin-Esteban, A. TrAC, Trends Anal. Chem. 2013, 45, 169− 181. (10) Pauling, L. J. Am. Chem. Soc. 1940, 62, 2643−2657. (11) Pauling, L.; Campbell, D. H.; Pressman, D. Physiol. Rev. 1943, 23, 203−219. (12) Dickey, F. H. Proc. Natl. Acad. Sci. U. S. A. 1949, 35, 227−229. (13) Lofgreen, J. E.; Ozin, G. A. Chem. Soc. Rev. 2014, 43, 911−933. (14) Gao, D.; Wang, D.-D.; Fu, Q.-F.; Wang, L.-J.; Zhang, K.-L.; Yang, F.-Q.; Xia, Z.-N. Talanta 2018, 178, 299−307. 8442

DOI: 10.1021/acs.analchem.9b01453 Anal. Chem. 2019, 91, 8436−8442