Preparations of magnetic molecularly imprinted polymer for selective recognition and determination of 4-methylimidazole in soft beverage by high performance liquid chromatography

Abstract

The magnetic molecularly imprinted polymers had been synthesized for the selective extraction and clean-up of 4-methylimidazole in cola, the most popular soft beverage. The magnetic molecularly imprinted polymers were prepared by means of suspension polymerization, using Fe₃O₄ as magnetically component, 4-methylimidazole as template molecule, methacrylic acid as functional monomers, and ethylene glycol dimethacrylate as a cross-linker, initiated by 2,2′-azobisisobutyronitrile. The obtained magnetic molecularly imprinted polymers were characterized by scanning electron microscopy, Fourier transform infrared, X-ray diffraction, and vibrating sample magnetometer. High performance liquid chromatography was used for the analysis of the target analyte. The polymers were evaluated further by batch rebinding experiments. From the derived Freundlich isotherm equation, their binding capacity and binding strength were determined. Structurally similar compound and a reference compound were used for investigating the selective recognition capability of magnetic molecularly imprinted polymers. The recoveries of spiked samples ranged from 90.19 to 104.29%. The prepared magnetic molecularly imprinted polymers could be applied to selectively preconcentrate and determine 4-methylimidazole in cola samples.

Keywords

Magnetic molecularly imprinted polymer 4-methylimidazole cola drinks high performance liquid chromatography binding isotherms

Introduction

Molecular imprinting polymers (MIPs) simulate the natural molecular recognition function and prepare the polymer which has specific selectivity for a particular target molecule. The specific target molecule is considered as the template and combines with the functional monomer which has the complementary structure by means of covalent bonding or noncovalent bond. The functional monomer polymerizes in a certain solution using cross-linking agent. Then the rigid and highly cross-linked polymer formed around the template molecules. After the reaction, the template molecules should be removed from the polymer. Then, the structure which matches completely with the template molecules in space position and binding site, formed in the interior of the polymer. As a result, the polymer has the function of specific molecular identification (Moring et al., 2002). At present, MIPs can be divided into two basic types: covalent molecular imprinting technology imprinting method (assembling) and noncovalent imprinting method (self-assembly). Commonly used noncovalent interaction are hydrogen bonds, electrostatic forces, π-πeffect, etc. of which hydrogen bond is applied most. MIPs can be used for the selective extraction, separation, or concentration of target molecules in a wide area (Olsen et al., 1999; Peng et al., 2001), especially in chromatography (Sellergren, 2001), chemical sensors (Alizadeh et al., 2009; Jenkins et al., 2001; Lin and Yamada, 2001), catalysis (Wulff, 2002), drug delivery (Alvarez-Lorenzo and Concheiro, 2004), and solid-phase extraction (SPE) (Chen et al., 2011; Jiang et al., 2007, 2009). Based on the high specificity and selectivity of molecular recognition mechanism, the molecularly imprinted SPE (MISPE) method has been used for the selective separation or determination of natural products in complicated samples (Chen et al., 2012; Gadzala-Kopciuch et al., 2009). In recent study of MISPE, a novel technique that introduces magnetic molecularly imprinted polymers (MMIPs) into SPE sorbents (Ansell and Mosbach, 1998; Chen et al., 2009; Gu et al., 2010) is widely developed. As an eye-catching technology, magnetic separation has attracted most researchers’ attention for its great potential applications in bioseparation (Ouyang et al., 2010), biomolecular sensing (Jing et al., 2011), drug delivery (Kan et al., 2009), and food safety assessment (Feng et al., 2015). As sorbents in SPE procedure, MMIPs can be dispersed into the solution directly and then easily separated from the matrix by an external magnetic field, which makes the sample pretreatment simple.

Soft drinks, including colas, are widely consumed in China and the United States (Bleich et al., 2009; Wang et al., 2008). A common ingredient in many soft drinks (e.g. colas, root beers, iced teas) is caramel color (Kamuf et al., 2003; United States Food and Drug Administration, 2012), which is produced with ammonium compounds. The use of these compounds to manufacture caramel color can result in the formation of 4-methylimidazole (4-MeI) (Moon and Shibamoto, 2011). In recent years, evidence for the carcinogenicity of 4-MeI has raised concerns about uses of caramel color type III and IV that may expose consumers to 4-MeI and increase cancer risk (Aubrey, 2014; Consumer Reports, 2014). The presence of the minor caramel component in most foods and beverages, however, can be hazardous to humans because of toxicity. 4-MeI is a neurotoxic agent (Patey et al., 1985) and some in vitro studies have shown its capability to inhibit the cytochrome P450 isoenzyme which catalyzes the oxidation of many known or suspected carcinogens of low molecular mass in the human liver (Hargreaves et al., 1994). Furthermore, a recent toxicological study conducted by the National Institute of Environmental Health Sciences of USA (Chan et al., 2008) showed that 4-MeI can induce alveolar/bronchiolar adenoma and carcinoma in male and female mice. Also, TOX-67 test shows that it can cause cancer. According to the Food Chemicals Codex (FCC V-2004, USA), 4-MeI in caramel is limited to no more than 0.025%. Chinese government standard (GB8817-2001) stipulates that 4-MeI is no more than 0.02%. Therefore, to ensure its safe and legal use, it is necessary to establish a simple, direct, and sensitive technique to extract and monitor the trace levels of 4-MeI in beverages.

Several methods have been developed to determine 4-MeI based on thin layer chromatography (Rabe et al., 1988), fluorimetry (Wu et al., 2016), capillary electrophoresis (Petruci et al., 2013), or high performance liquid chromatography (HPLC) coupled with ultraviolet light (UV) (Ratnayake et al., 2015). However, these methods require a labor and time-consuming sample pretreatment and have poor sensitivity. In recent years, more sensitive methods have been published, based mainly on mass spectrometry (MS) as a detection technique, coupled with a chromatographic step either by liquid chromatography (LC) (Schlee et al., 2013) or gas chromatography (GC), the latter after derivatization of the analytes (Cunha et al., 2011). Notwithstanding the high selectivity achieved by this technique, the methods include a previous tedious SPE or supercritical fluid extraction. This work is the first attempt to synthesize MMIPs for recognition of 4-MeI. Fe₃O₄ particles were synthesized by means of the chemical coprecipitation method as magnetic cores. MMIPs were synthesized using 4-MeI as template molecule, methacrylic acid (MAA) as functional monomers, ethylene glycol dimethacrylate (EGDMA) as a cross-linker, and initiated by 2,2′-azobisisobutyronitrile (AIBN). The characterization, adsorption capacity, and selectivity of MMIPs and magnetic nonmolecularly imprinted polymers (MNIPs) were investigated. The polymers were used as SPE sorbents coupled with HPLC for the determination of 4-MeI in cola samples. Good recoveries with low detection limit were obtained.

Experimental

Materials and regents

4-MeI was purchased from Alfa Aesar (Tianjin, China) and the structure was shown in Figure 1. MAA, EGDMA, and AIBN were obtained from Alfa Aesar (Tianjin, China). Ferric chloride (FeCl₃·6H₂O) and ferrous chloride (FeCl₂·4H₂O) were purchased from Fuchen Chemical Reagents Factory (Tianjin, China). Chromatographic grade methanol and acetonitrile were purchased from Merck KGaA (Darmstadt, Germany). Dimethyl sulfoxide, polyethylene glycol (PEG-6000), acetic acid, ammonium hydroxide, and the other chemicals were supplied from Tianjin Chemical Reagent Co. (Tianjin, China). Ultrapure water was prepared by an ultrapurification system. Two cola samples were obtained from a supermarket (Xi’an, China). All solutions used for HPLC were filtered through a 0.45 µm filter before use.

Figure 1.

Structure of 4-MeI.

Instrumentation

The HPLC analysis was performed in a WATERS Series (WATERS Technologies, Milford, MA, USA) LC system, which was equipped with an e2695 Alliance Quaternary Pump, a 2998 Photodiode Array Detector, an Alliance Col Heater column oven and an automatic sampler. The system was controlled by an Empower 2 Personal Single System. Chromatographic separation was performed with a Sino-Chrom ODS-AP column (5 µm, 230 mm × 4.6 mm) (Dalian Elite Analytical Instruments Co., Ltd, Dalian, China). The mobile phase was consisted of methanol and KH₂PO₄ (0.05 mol l⁻¹) in a ratio of 10:90 (v/v) with a flow rate of 1.0 ml min⁻¹. The detection wavelength and column temperature were set at 233 nm and 28℃, respectively, and the loading volume was 20 µl. The pH measurements were performed with a model FE20 Plus pH meter (Mettler-Toledo, Shanghai) equipped with InLab® Micro pH combination electrodes (Mettler-Toledo, Switzerland). The PALL ultrapure water systems were supplied by ELGA LabWater Instrument Co., Ltd (Bucks, UK). X-ray diffraction (XRD) measurements were carried out using a Bruker D8 Advance (XRD, Bruker, Germany). The magnetic properties were measured by a Lake Shore 7307 vibrating sample magnetometer (VSM) (Lakeshore, USA). SEM image was obtained via Hitachi S-4800 field emission scanning electron microscope (Tokyo, Japan). Transmission electron micrographs were obtained from a Hitachi H-7650 transmission electron microscope (TEM, Hitachi, Japan). The FT-IR spectra were obtained via a Nicolet Nexus-670 FT-IR spectrometer. The wave numbers of FT-IR measurement range were controlled from 500 to 4000 cm⁻¹.

Preparation of Fe₃O₄ magnetic particles

The Fe₃O₄ magnetic particles were synthesized by a slightly modified chemical coprecipitation method in which 5.11 g FeCl₃·6H₂O and 1.83 g FeCl₂·4H₂O were dissolved in 80 ml of deoxygenated water. The solution was constantly stirred in a 250 ml three-necked flask. When the temperature increased to 80℃, 60 ml of 5% ammonium hydroxide solution was added dropwise, and the mixture was stirred vigorously for 60 min. The entire reaction process was performed under nitrogen gas protection. When the reaction finished, Fe₃O₄ MNPs were collected with a magnet and washed several times with deionized water. Then the Fe₃O₄ particles were modified with surface modifier. Fe₃O₄ (2.0 g) and PEG (10.0 g) were dissolved in 30 ml of deoxygenated water and stirred for 20 min. After sonicating for 30 min, a homogeneously dispersed solution was obtained.

Preparation of the MMIPs

The MMIPs were prepared with a slight modification according to the literature published (Chen et al., 2012; Feng et al., 2015). 4-MeI (1 mmol) as the template and the MAA (4 mmol) as the functional monomer were dissolved in 50 ml acetonitrile and stored in dark for 18 h in room temperature. Then the prepolymerization solution, PEG-Fe₃O₄ particles, dispersing media (doubly distilled water, 80 ml), cross-linker EGDMA (3.8 ml, 20 mmol), and initiator AIBN (40 mg, 24.4 mmol) were well mixed in a 250 ml three-neck flask. The mixture was degassed in an ultrasonic bath for 15 min. Then, the oxygen-free nitrogen passed through the three-neck flask from the side necks as inlet and outlet. The stirring rod of an electric stirrer stuck into the solution from the hole on the rubber plug which plugged in the middle neck. The diameter of the hole precisely can allow the rod to rotate. As a result, the three-neck flask could be considered to be replenished with nitrogen. And then stirring at 300 r min⁻¹, the polymerization was performed with nitrogen protection at 60℃ for 12 h. MMIPs were collected magnetically and washed by a mixture of methanol/acetic acid (9:1, v/v) to remove the templates and then washed by methanol. Finally, the particles were dried in vacuum. The MNIPs were prepared in the same method as MMIPs but without the addition of template.

Adsorption experiment and selectivity evaluation

In order to evaluate the capacity of MMIPs for recognizing and binding 4-MeI in methanol, the adsorption test was carried out. Twenty milligrams of MMIPs or MNIPs were suspended in 1 ml of various concentrations of 4-MeI solution (5–300 mg l⁻¹) in 2 ml centrifuge tube. After shaken in incubator for 24 h at 25℃, a magnet was deposited outside of the sample tube to separate MMIPs or MNIPs from the solution, and then the supernatant was measured by HPLC. The amount of 4-MeI binding to the MMIPs or MNIPs was calculated by subtracting the amount of free 4-MeI from the amount of 4-MeI initially added. The data of the absorption experiment were further processed according to the Freundlich isotherm (FI) model to estimate the binding parameters of the MMIPs and MNIPs.

Compared with salicylic acid and structure-similar compound benzoic acid, the selectivity of the magnetic imprinted sorbent was investigated in three different concentrations (20, 60, 120 mg l⁻¹). The amount of free analyte was measured by HPLC.

Extraction procedure

Twenty milligrams of MMIPs were added directly to 1 ml cola sample and then shaken at room temperature for 15 min. A magnet was used to separate MMIPs from the sample solution. One milliliter of acetonitrile/formic acid (9:1, v/v) was used to wash the MMIPs by sonication for 15 min. Supernatants (0.5 ml) were evaporated to dryness and dissolved in 0.1 ml of methanol for a further HPLC–UV analysis.

Results and discussion

Preparation and characterization of MMIPs

The schematic procedure of preparing Fe₃O₄ magnetic MIP beads is shown in Figure 2. Due to the anisotropic dipole attraction, unmodified magnetic nanoparticles tend to be reunited into larger clusters, thus losing the monodispersity. The dispersion of Fe₃O₄ magnetic nanoparticles in reaction system plays a very important role for the preparation of composite materials. Without Fe₃O₄ magnetic nanoparticles surface modification, Fe₃O₄ magnetic particles sedimentation occurs during the process of reaction, and structure-inerratic composite materials cannot be obtained (Zhang et al., 2010). Considering their low compatibility with the organic phase, Fe₃O₄ particles were modified using a surface modifier to reveal the hydrophobicity for further functionalization. PEG was used as the surface modifier to introduce a polymeric chain on the surface of the particles.

Figure 2.

Schematic representation of 4-MeI-MMIPs.

Molecular recognition of the template molecule in the imprinted polymers is based on the intermolecular interaction between the template molecule and functional groups in the polymer (Sun et al., 2008). The monomer was the most important factor generating the recognition sites through organized self-assembly with the template 4-MeI. Same amounts of functional monomers (AM) and methacrylic acid were tested for monomer selection. Finally, MAA was chosen for preparing the polymers, because MAA had better molecular recognition in polar conditions. The optimum molar ratio of 1:6:30 (template:MAA:EGDMA) was used to prepare MMIPs. Good homogeneity and density of the resultant beads could then be reproducibly prepared. Figure 3 shows the surface morphologies of MMIPs and MNIPs via SEM. The diameter of the beads ranged from 50 to 400 µm, and majority of the beads were spherical (Figure 3(a)). The surface of the MMIP beads (Figure 3(b)) was more porous and rougher than that of MNIPs (Figure 3(c)). This specific structure is favorable for the adsorption or release of the template molecules from the imprinted beads.

Figure 3.

Scanning electron micrographs of the MMIPs and MNIPs: (a) MMIPs bead, (b) surface of MMIPs, and (c) surface of MNIPs.

XRD (see the supplementary figure online) was used to analyze the components of the magnetic nanoparticles and MMIPs. The presence of crystal planes with cubic crystal structures reveals that the magnetic nanoparticles and MMIPs were composed of Fe₃O₄. The crystalline structure of the Fe₃O₄ nanoparticles and MMIPs was essentially maintained.

FT-IR for Fe₃O₄, MMIPs, and MNIPs was carried out to further ensure the preparation of magnetic 4-MeI-MMIPs. Compared with Fe₃O₄ nanoparticles, the absorption band of Fe-O at 592.1 cm⁻¹ found in MMIPs and MNIPs proved that Fe₃O₄ was embedded in these materials (Figure 4). The typical bands 2925.9 cm⁻¹ can be attributed to the C–H stretching vibrations of MAA. The characteristic peak of MAA was observed at 3442.8 cm⁻¹ due to the O–H stretching of MAA. The weak absorbance peak of C = C at 1602.8 cm⁻¹ demonstrates that most MAA were cross-linked. Only few of MAA remained. In the MMIPs spectrum, the absorbance peak of C = O at 1728.1 cm⁻¹ shows that the MMIPs were synthesized through the polymerization of EGDMA and MAA. Moreover, the peak intensity in the MNIPs was lower than that in MMIPs because of the effect of hydrogen between 4-MeI and the MAA.

Figure 4.

FT-IR spectra of the 4-MeI MMIPs and MNIPs, Fe₃O₄ nanoparticles.

The magnetic properties of the Fe₃O₄ nanoparticles and MMIPs were measured by VSM. The saturation magnetizations were 67.34 and 3.93 emu g⁻¹ for Fe₃O₄ nanoparticles and MMIPs, respectively. The Ms value decreased because of the relatively low amounts of Fe₃O₄ nanoparticles loading on MIPs. The results show that MMIPs sorbents and Fe₃O₄ nanoparticles displayed typical superparamagnetic behaviors, which avoided the aggregation. Due to their relatively high saturation magnetizations, MMIPs could be separated easily from the sample solution by an external magnet.

Binding isotherms

Figure 5(a) shows the isothermal adsorptions of 4-MeI in the MMIPs and MNIPs, as well as their corresponding experimental FI. The binding amount of 4-MeI on both polymers increased along with its initial concentration, and the MMIPs had higher affinity for 4-MeI than MNIPs. The binding data can be analyzed by the FI affinity distribution analysis model.

Figure 5.

(a) 4-MeI adsorption isotherms for MMIPs and MNIPs with the corresponding experimental FI for MMIPs and MNIPs and (b) affinity distributions of MMIPs (—) and MNIPs (…).

The FI model is the most important multilayer adsorption isotherm for heterogeneous surfaces (Rampey et al., 2004)

\log B = m \log F + \log a

(1)

where B and F are the concentrations of bound and free analytes, respectively, and m is the heterogeneity index. The parameter m can range from 1 to 0, increasing with decreasing heterogeneity of the material. Value 1 indicates homogeneous and values approaching zero indicates increasingly heterogeneous (Table 1). The affinity distribution can be calculated by equation (2) and the experimentally derived FI fitting parameters (a and m) (Muk and Narayanaswamy, 2011)

N (k) = 2.303 am (1 - m^{2}) K^{- m}

(2)

where K is the affinity constant (K can be assumed to be equal to 1/F), and N(k) is the number of binding sites with a given affinity.

Table 1.

Freundlich fitting parameters, number of binding sites (N_{k min–k max}), and weighted average affinity (K_{k min–k max}) for 4-methylimidazole MMIPs and MNIPs.

Fitting parameters	MMIPs	MNIPs
N_{k min–k max} (µmol g⁻¹)	2.57	0.22
K_{k min–k max} (l mmol⁻¹)	3.96	0.63
Α	3.47	0.19
m	0.24	0.72

MMIP: magnetic molecularly imprinted polymer; MNIP: magnetic nonmolecularly imprinted polymer.

The number of binding sites per gram of material (N_{k min–k max}; see equation (3)) and the weighted average affinity constant (K_{k min–k max}; see equation (4)) can be calculated (Medina-Castillo et al., 2010; Peng et al., 2011), where the values of m and a are equivalent to the Freundlich parameters (Muk and Narayanaswamy, 2011)

N_{k min - k max} = a (1 - m^{2}) (K_{min}^{- m} - K_{max}^{- m})

(3)

K_{k min - k max} = (\frac{m}{m - 1}) (\frac{K_{min}^{1 - m} - K_{max}^{1 - m}}{K_{min}^{- m} - K_{max}^{- m}})

(4)

The values for these parameters can be calculated from the experimental maximum (F_max) and minimum (F_min) free analyte concentrations and within the limits of K_min and K_max being equal to the corresponding reciprocal concentrations K_min = 1/F_max and K_max = 1/F_min.

The binding affinity distribution of MMIPs and MNIPs is shown in Figure 5(b) in the form of a common semi-log format (N versus log K). This plot also represents the site-energy distribution of the polymers. Compared with MNIPs, the number of sites with any given affinity energy in the MMIPs was higher, across the entire range of tested concentrations. This result confirms the imprinting phenomenon.

The calculated fitting parameters (the number of binding sites, N_{k min–k max}, and the weighted average affinity, K_{k min–k max}) are summarized in Table 1. The weighted average K_{k min–k max} of MMIPs and MNIPs were evaluated to be 3.96 and 0.63 l mmol⁻¹, respectively, whereas the total average N_{k min–k max} were 2.57 and 0.22 µmol g⁻¹, respectively. The total number of binding sites and the affinity constants were higher in the MMIPs than in the MNIPs. This result indicates that the total number of binding sites with geometry and functionality complementary to the template molecule is greater in the MMIPs. Thus, the template molecule significantly displays an important role in the heterogeneity of the MMIPs and confirms the imprinting phenomenon.

Selectivity evaluation of MMIPs and MNIPs

Figure 6 illustrates the adsorption ability of MMIPs and MNIPs for two structurally similar compounds. The experiments were carried out at three concentration levels of the standard solution (20, 60, and 120 mg l⁻¹). The binding capacity of the MMIPs to 4-MeI was evidently greater than that to salicylic acid and benzoic acid. Three parameters were used to evaluate the selectivity of MMIPs: the selectivity coefficient (k), distribution coefficient (K_d), and relative selectivity coefficient (k′) (Peng et al., 2011). The selectivity coefficient (k) is defined as the target-to-competitive molecule ratio of the K_d values. The distribution coefficient (K_d) is defined as the adsorbed-to-unadsorbed concentration ratio. The relative selectivity coefficient (k′) is defined as the target-to-competitive molecule ratio of the k values. K_d indicates the adsorption capacity. The larger the value of K_d is, the stronger the adsorption capability of a substance will be. The parameter k reveals the selectivity between two substances, whereas k′ reflects the selective difference between MMIPs and MNIPs. The larger the value of k′ is, the greater the selectivity of molecular imprinting will be.

Figure 6.

Selective recognition capability of MMIPs and MNIPs to (a) 4-MeI, (b) salicylic acid, and (c) benzoic acid at concentrations of 20, 60, and 120 mg l⁻¹.

The values of the three parameters for the tested compounds are shown in Table 2. No obvious differences were observed in rebinding the three compounds for MNIPs. However, the MMIPs demonstrated obvious adsorption selectivity for 4-MeI. The k value of MMIPs (12.20 or 13.67) for salicylic acid and benzoic acid was 11.86 times (12.90 times) greater than that of MNIPs. This result indicates that the prepared MMIPs exhibited a higher selectivity for 4-MeI than the reference compounds and had no specific site to the compound with significantly different structures.

Table 2.

Recognition properties of MMIPs and MNIPs.^a

	Levels (mg l⁻¹)	K_d (µmmol g⁻¹)			k		k′
	Levels (mg l⁻¹)	K_d1 4-MeI	K_d2 Salicylic acid	K_d3 Benzoic acid	k ₁	k ₂	k′₁	k′₂
MMIPs	120	26.30	1.28	1.17	20.55	22.48	15.19	16.20
	60	18.30	1.50	1.33	12.20	13.76	11.86	12.90
	20	12.40	1.86	1.10	6.67	11.27	7.28	7.94
NMIPs	120	1.61	1.19	1.16	1.35	1.39
	60	1.44	1.40	1.35	1.03	1.07
	20	1.52	1.66	1.07	0.92	1.42

4-MeI: 4-methylimidazole; MMIP: magnetic molecularly imprinted polymer; MNIP: magnetic nonmolecularly imprinted polymer .

K_d , distribution coefficient; k, selectivity coefficient; k₁ = K_d1/K_d2, k₂ = K_d1/K_d3; k′, relative selectivity coefficient; k₁ = k_1MMIP/k_1MNIP , k₂ = k_2MMIP/k_2MNIP .

Optimization of extraction, desorption solvent, and desorption time

Extraction process was investigated in three different volumes (i.e. 1, 3, and 5 ml). The concentration of 4-MeI was 300 ng ml⁻¹. Figure 7(a) shows that the MMIPs reached adsorption equilibrium at approximately 15 min with the three volumes. The bounding amount of targets was obviously affected by the sample volume. Therefore, 1 ml was selected as the adsorption volume at 15 min extraction time in the following experiments.

Figure 7.

(a) Dynamic adsorption isotherms of MMIPs of 4-MeI in three sample volumes (i.e. 1, 3, and 5 ml), (b) optimization of desorption solvent, and (c) effect of desorption time.

To elute the target analyte that adsorbed onto the MMIPs, methanol, acetonitrile, methanol with acetic acid (9:1), and acetonitrile with acetic acid (9:1) were optimized as the common organic elution solvents. Figure 7(b) shows that acetonitrile with acetic acid (9:1) had the best desorption ability for target analyte.

Different time intervals (i.e. 3, 6, 9, 15, 30, and 60 min) were evaluated to obtain the desorption time of the target analyte. Figure 7(c) illustrates that 15 min was sufficient to desorb the analyte completely, and then the MMIPs could be isolated in a short time (approximately 30 s) by an external magnetic field. Therefore, the desorption time was set at 15 min.

Analytical method validation and reusability evaluation

A series of experimental parameters, such as linear range, correlation coefficient, limit of detection (LOD), and limit of quantification (LOQ) were performed to evaluate the proposed method and validate the analytical methodology. The calibration curve was obtained using the linear regression method and peak areas were plotted versus concentrations. The regression equation was y = 227.7 + 89.6 x (r = 0.9998) within the concentration range of 0.7 × 10⁻²–1.4 × 10²µmol l⁻¹ for 4-MeI. LOD and LOQ were defined as three and 10 times the signal-to-noise ratio, respectively. LOD and LOQ were 0.04 and 0.13 µmol l⁻¹, respectively.

The standard addition method was used to evaluate the repeatability, accuracy, and recovery of the MMIP extraction process. Sample solutions were spiked with 4-MeI at three concentration levels. The test samples were prepared by mixing 0.5 ml of sample solutions with 0.5 ml of standard 4-MeI solution. The results are summarized in Table 3. The recoveries of the spiked samples for 4-MeI ranged from 90.19 to 104.29% and the relative standard deviation (RSD) values ranged from 2.15 to 7.30%. These results indicated that the new analytical method was reliable and sensitive.

Table 3.

Accuracy of the method for sample solutions spiked at different concentrates (n = 3).

Analyte	Added (µmol l⁻¹)	Found (µmol l⁻¹)	Recovery (%^a)	Average (%)	RSD (%^b)
4-MI	110.0	111.18	101.07	98.51	2.535
	110.0	105.69	96.08
	110.0	108.21	98.37
	4.3	4.45	104.29	97.96	7.305
	4.3	3.85	90.19
	4.3	4.24	99.38
	0.9	0.89	97.55	98.70	2.158
	0.9	0.92	101.16
	0.9	0.89	97.39

((Found-base)/added) × 100; ^b(relative standard deviation, RSD).

The reusability of prepared MMIPs was evaluated. Six times adsorption–desorption process were tested. The peak areas of 4-MeI in six desorption solution were not found obvious changes. The result (Figure 8) showed that the prepared MMIPs could be reused for at least six times.

Figure 8.

Reusability evaluation of MMIPs.

Analysis of 4-MeI in the real samples

The present study aims to provide a simple, selective, and practical process, which can be applied in the determination of analyte from cola samples by MMIPs. Two cola samples were extracted by MMIPs and MNIPs (“Adsorption experiment and selectivity evaluation” section). Figure 9 represents the chromatogram of direct injection of the two cola samples, the spiked two cola samples extracted by MNIPs and MMIPs. 4-MeI cannot be directly determined from the cola samples by HPLC without enrichment (Figure 9(a)). No related peak was observed in Figure 9(b) in the analysis of the spiked solutions extracted by MNIPs. This result indicated that the spiked 4-MeI cannot be extracted from the cola samples by MNIPs. The selective determination of spiked 4-MeI in the cola samples by the proposed MMIP extraction method was successful (Figure 9(c)). These polymers can be used in the field of selective detection of target analytes in complicated samples. By this efficient method, no obvious 4-MeI peak was obtained in the MMIPs-extracted initial cola samples. Because of the LOQ is below the safe range, the purchased cola samples can be considered safe.

Figure 9.

Chromatograms of two samples: (a) initial solutions, (b) solution extracted with MNIPs, and (c) with MMIPs.

Comparison with published methods

In recent years, some studies have reported methods to extraction and determine 4-MeI in food and drinks. The extraction method covered column-SPE and filter. GC–MS, LC–MS, CE–DAD, and HPLC–UV were involved. Although the LOD of Cunha et al. (2011) and Kim et al. (2013) was pretty low, the extraction process took extremely long time. During derivatization process in Cunha et al. (2011), some by-products may be created and some side reactions may occur. These all may have an influence on the determination. The lower LODs relied on the developed LC or GC–MS. Some studies reported several MS methods to detect the aim compounds. So the current study sought an easy applicable alternative (PDA) to the sophisticated HPLC–tandem MS or GC–MS method in view of the limited financial resources of many laboratories. These are more suitable for most laboratories in China and third-world countries.

As for the centrifuge and filter process in other literatures, the current reported SPE material is magnetically modified. As a result, the material can be collected by the extra magnetic field, and the centrifugal or filtering process can be eliminated. So this method will save a lot of time in the entire experiment.

Table 4 describes the methods used to determine 4-MeI in several matrices in recent years. In comparison, even though the LOD is not the best, this reported method still has the advantages of improved simplicity, good recovery, low cost, recyclable material, and feasible conversion into green analytical techniques.

Table 4.

Comparison of some methods used for determination of 4-MeI.

No.	Matrix	Extraction method	Detection	LOD in literatures	LOD (ng ml⁻¹)	Ref.
1	Soft drinks and dark beer	Extraction and derivatization	GC–MS	0.06 µg l⁻¹	0.06	Cunha et al. (2011)
2	Soy sauce	Two-times SPE column	LC–MS/MS	0.4 ng ml⁻¹	0.4	Yamaguchi and Masuda (2011)
3	Caramel color and processed foods	Centrifuge and filter	HPLC–MS/MS	3 mg kg⁻¹	–	Kim et al. (2013)
4	Foods and beverages	SPE column	HPLC–UV	14.5 ng ml⁻¹	14.5	Klejdus et al. (2006)
5	Caramel colors	Filtered through a cellulose acetate filter disk	CE–DAD	0.16 mg l⁻¹	160	Petruci et al. (2013)
6	Cola	MMIP-SPE	HPLC–PDA	0.04 µmol l⁻¹	3.34	This one

4-MeI: 4-methylimidazole; HPLC–UV: high performance liquid chromatography–ultraviolet light; MMIP: magnetic molecularly imprinted polymer.

Conclusions

The MMIPs were prepared as selective extraction sorbents for the analysis of 4-MeI in real beverage samples. The obtained MMIPs were characterized by SEM, FT-IR, XRD, and VSM. Investigation of selectivity recognition properties showed high adsorption capacity and selectivity of the MMIPs to template molecule. The extraction procedure took a short time to reach the adsorption, and desorption equilibrium and the MMIPs were easily collected by an external magnetic field. The proposed method was used to detect 4-MeI in real cola samples, and the MMIPs displayed a good selectivity and specificity to 4-MeI. The high recovery rate of the cola samples proved that the method was valid for the analysis of 4-MeI in real cola samples. The prepared MMIPs had great potential application in determination of 4-MeI in beverage samples.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by fund from the National Natural Science Foundation of China (No. 81302619) and Teacher Doctoral Research Start-Up Funding of Xi'an University of Technology.

Supplemental Material

The online supplementary figure is available at .

References

Alizadeh

Ganjali

Norouzi

(2009) A novel high selective and sensitive para-nitrophenol voltammetric sensor, based on a molecularly imprinted polymer–carbon paste electrode. Talanta 79: 1197–1203.

Alvarez-Lorenzo

Concheiro

(2004) Molecularly imprinted polymers for drug delivery. Journal of Chromatography B 804: 231–245.

Ansell

Mosbach

(1998) Magnetic molecularly imprinted polymer beads for drug radioligand binding assay. Analyst 123: 1611–1616.

Aubrey A (2014) Potential carcinogen in colas has FDA reviewing data. Available at: http://www.npr.org/blogs/thesalt/2014/01/23/265245172/potential-carcinogen-in-colas-has-fda-reviewing-data (accessed 23 January 2014).

Bleich

Wang

(2009) Increasing consumption of sugar-sweetened beverages among US adults: 1988–1994 to 1999–2004. The American Journal of Clinical Nutrition 89: 372–381.

Chan

Hills

Kissling

(2008) Toxicity and carcinogenicity studies of 4-methylimidazole in F344/N rats and B6C13F1 mice. Archives of Toxicology 82: 45–53.

Chen

Wang

Shi

(2011) Molecularly imprinted polymer microspheres for solid-phase extraction of protocatechuic acid in Rhizoma homalomenae. Journal of Separation Science 34: 2602–2610.

Chen

Wang

Shi

(2012) Molecularly imprinted polymer for the specific solid-phase extraction of kirenol from Siegesbeckia pubescens herbal extract. Talanta 89: 505–512.

Chen

Xie

Shi

(2012) Magnetic molecularly imprinted polymer for the detection of rhaponticin in Chinese patent medicines. Journal of Chromatography A 1252: 8–14.

10.

Chen

Liu

Zeng

(2009) Preparation of magnetic molecularly imprinted polymer for the separation of tetracycline antibiotics from egg and tissue samples. Journal of Chromatography B 1216: 3710–3719.

11.

Consumer Reports (2014) Caramel color: The health risk that may be in your soda. Available at: http://www.consumerreports.org/cro/news/2014/01/caramel-color-the-health-risk-that-may-be-in-your-soda/index.htm (accessed 10 February 2014).

12.

Cunha

Barrado

Faria

(2011) Assessment of 4-(5-)methylimidazole in soft drinks and dark beer. Journal of Food Composition and Analysis 24: 609–614.

13.

Feng

Wei

(2015) Fabrication of Fe₃O₄ nanoparticle-coalesced hydroxylated multi-walled carbon nanotubes for the analysis of strychnine in human serum. Analytical Science 31: 399–406.

14.

Gadzala-Kopciuch

Ricanyova

Buszewskia

(2009) Isolation and detection of steroids from human urine by molecularly imprinted solid-phase extraction and liquid chromatography. Journal of Chromatography B 877: 1177–1184.

15.

Yuan

(2010) Preparation of chlorogenic acid surface-imprinted magnetic nanoparticles and their usage in separation of Traditional Chinese Medicine. Analytica Chimica Acta 675: 64–70.

16.

Hargreaves

Jones

Smith

(1994) Inhibition of p-nitrophenol hydroxylase in rat liver microsomes by small aromatic and heterocyclic molecules. Drug Metabolism and Disposition 22: 806–810.

17.

Jenkins

Yin

Jensen

(2001) Molecularly imprinted polymer sensors for pesticide and insecticide detection in water. Analyst 126: 798–802.

18.

Jiang

Zhao

Chu

(2009) Molecularly imprinted solid-phase extraction for the selective determination of 17β-estradiol in fishery samples with high performance liquid chromatography. Talanta 78: 442–447.

19.

Jiang

Tian

Zhao

(2007) A novel sol–gel-material prepared by a surface imprinting technique for the selective solid-phase extraction of bisphenol A. Talanta 72: 119–125.

20.

Jing

Xia

Guan

(2011) Rapid and selective determination of urinary lysozyme based on magnetic molecularly imprinted polymers extraction followed by chemiluminescence detection. Analytica Chimica Acta 692: 73–79.

21.

Kamuf

Nixon

Parker

(2003) Overview of caramel colors. Cereal Foods World 48: 64–69.

22.

Kan

Geng

Zhao

(2009) Magnetic molecularly imprinted polymer for aspirin recognition and controlled release. Nanotechnology 20: 165601.

23.

Kim

Shin

(2013) Determination of 4-methylimidazole and 2-acetyl-4(5)-tetrahydroxybutylimidazole in caramel color and processed foods by LC-MS/MS. Preventive Nutrition and Food Science 18: 263–268.

24.

Klejdus

Moravcová

Lojková

(2006) Solid-phase extraction of 4(5)-methylimidazole (4MeI) and 2-acetyl-4(5)-(1,2,3,4-tetrahydroxybutyl)-imidazole (THI) from foods and beverages with subsequent liquid chromatographic-electrospray mass spectrometric quantification. Journal of Separation Science 29: 378–384.

25.

Lin

Yamada

(2001) Chemiluminescent flow-through sensor for 1,10-phenanthroline based on the combination of molecular imprinting and chemiluminescence. Analyst 126: 810–815.

26.

Medina-Castillo

Mistlberger

Fernandez-Sanchez

(2010) Novel strategy to design magnetic, molecular imprinted polymers with well-controlled structure for the application in optical sensors. Macromolecules 43: 55–61.

27.

Moon

Shibamoto

(2011) Formation of carcinogenic 4(5)-methylimidazole in Maillard reaction systems. Journal of Agricultural and Food Chemistry 59: 615–618.

28.

Moring

Wong

Stobaugh

(2002) Target specific sample preparation from aqueous extracts with molecular imprinted polymers. Journal of Pharmaceutical and Biomed Analysis 27: 719–728.

29.

Muk

Narayanaswamy

(2011) Molecularly imprinted polymers as optical sensing receptors: Correlation between analytical signals and binding isotherms. Analytica Chimica Acta 703: 226.

30.

Olsen

Martin

Wilson

(1999) Methodology for assessing the properties of molecular imprinted polymers for solid phase extraction. Analyst 124: 467–471.

31.

Ouyang

Lei

(2010) Artificial receptor-functionalized nanoshell: facile preparation, fast separation and specific protein recognition. Nanotechnology 21: 185502.

32.

Patey

Shearer

Knowles

(1985) Ammonia caramels: specifications and analysis. Food Additives and Contaminants 2: 107–112.

33.

Peng

Zhang

(2001) Development of a thickness shear mode acoustic sensor based on an electrosynthesized molecularly imprinted polymer using an underivatized amino acid as the template. Analyst 126: 189–194.

34.

Peng

Wang

Zeng

(2011) Molecularly imprinted polymer for solid-phase extraction of rutin in complicated traditional Chinese medicines. Analyst 136: 756–763.

35.

Petruci

Pereira

Cardoso

(2013) Determination of 2-methylimidazole and 4-methylimidazole in caramel colors by capillary electrophoresis. Journal of Agricultural and Food Chemistry 61: 2263–2267.

36.

Rabe

Seibel

Wuenscher

(1988) Detection of caramel colours and other browning substances in baked products. Getreide Mehl Und Brot 42: 142–148.

37.

Rampey

Umpleby

Rushton

(2004) Characterization of the imprint effect and the influence of imprinting conditions on affinity, capacity, and heterogeneity in molecularly imprinted polymers using the Freundlich isotherm-affinity distribution analysis. Analytical Chemistry 76: 1123–1133.

38.

Ratnayake

Halldorson

Bestvater

(2015) Determination of 4(5)-methylimidazole in carbonated beverages by isotope-dilution liquid chromatography-tandem mass spectrometry. Food Additives and Contaminants Part A—Chemistry 32: 1075–1082.

39.

Schlee

Markova

Schrank

(2013) Determination of 2-methylimidazole, 4-methylimidazole and 2-acetyl-4-(1,2,3,4-tetrahydroxybutyl)imidazole in caramel colours and cola using LC/MS/MS. Journal of Chromatography B 927: 223–226.

40.

Sellergren

(2001) Imprinted chiral stationary phases in high-performance liquid chromatography. Journal of Chromatography A 906: 227–252.

41.

Sun

Schussler

Sengl

(2008) Selective trace analysis of diclofenac in surface and wastewater samples using solid-phase extraction with a new molecularly imprinted polymer. Analytica Chimica Acta 620: 73–81.

42.

United States Food and Drug Administration (2012) Caramel. 21 CFR. 73.85.

43.

Wang

Bleich

Gortmaker

(2008) Increasing caloric contribution from sugar-sweetened beverages and 100% fruit juices among US children and adolescents, 1988–2004. Pediatrics 121: e1604–e1614.

44.

Huang

(2016) Rapid and quantitative detection of 4(5)-methylimidazole in caramel colours: A novel fluorescent-based immunochromatographic assay. Food Chemistry 190: 843–847.

45.

Wulff

(2002) Enzyme-like catalysis by molecularly imprinted polymers. Chemical Review 102: 1–8.

46.

Yamaguchi

Masuda

(2011) Determination of 4(5)-methylimidazole in soy sauce and other foods by LC-MS/MS after solid-phase extraction. Journal of Agricultural and Food Chemistry 59: 9770–9775.

47.

Zhang

(2010) Preparation of magnetic indole-3-acetic acid imprinted polymer beads with 4-vinylpyridine and β-cyclodextrin as binary monomer via microwave heating initiated polymerization and their application to trace analysis of auxins in plant tissues. Journal of Chromatography A 1217: 7337–7344.

Preparations of magnetic molecularly imprinted polymer for selective recognition and determination of 4-methylimidazole in soft beverage by high performance liquid chromatography

Abstract

Keywords

Introduction

Experimental

Materials and regents

Instrumentation

Preparation of Fe3O4 magnetic particles

Preparation of the MMIPs

Adsorption experiment and selectivity evaluation

Extraction procedure

Results and discussion

Preparation and characterization of MMIPs

Binding isotherms

Selectivity evaluation of MMIPs and MNIPs

Optimization of extraction, desorption solvent, and desorption time

Analytical method validation and reusability evaluation

Analysis of 4-MeI in the real samples

Comparison with published methods

Conclusions

Footnotes

Declaration of Conflicting Interests

Funding

Supplemental Material

References

Preparation of Fe₃O₄ magnetic particles