Selective separation and determination of glucocorticoids in cosmetics using dual-template magnetic molecularly imprinted polymers and HPLC

Accepted Manuscript Selective separation and determination of glucocorticoids in cosmetics using dual-template magnetic molecularly imprinted polymers and HPLC Min Liu, Xiaoyan Li, Junjie Li, Zongyuan Wu, Fang Wang, Li Liu, Xuecai Tan, Fuhou Lei PII: DOI: Reference:

S0021-9797(17)30561-1 http://dx.doi.org/10.1016/j.jcis.2017.05.041 YJCIS 22348

To appear in:

Journal of Colloid and Interface Science

Received Date: Revised Date: Accepted Date:

28 December 2016 9 May 2017 14 May 2017

Please cite this article as: M. Liu, X. Li, J. Li, Z. Wu, F. Wang, L. Liu, X. Tan, F. Lei, Selective separation and determination of glucocorticoids in cosmetics using dual-template magnetic molecularly imprinted polymers and HPLC, Journal of Colloid and Interface Science (2017), doi: http://dx.doi.org/10.1016/j.jcis.2017.05.041

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Selective separation and determination of glucocorticoids in cosmetics using dual-template magnetic molecularly imprinted polymers and HPLC Min Liu a, Xiaoyan Li a*, Junjie Li b, Zongyuan Wu a, Fang Wang a, Li Liu a, Xuecai Tan a, Fuhou Lei a. a

School of Chemistry and Chemical Engineering, Guangxi University for Nationalities,

Key Laboratory of Guangxi Colleges and Universities for Food Safety and Pharmaceutical Analytical Chemistry, Guangxi Key Laboratory of Chemistry and Engineering of Forest Products, Guangxi Colleges and Universities Key Laboratory of Utilization of Microbial and Botanical Resources, Nanning, Guangxi 530008, China b

School of Chemistry and Chemical Engineering, Guangxi University, Nanning, Guangxi

530004,China

Abstract Molecularly imprinting polymers (MIPs) are typically prepared using a single template molecule, which allows selective separation and enrichment of only one target analyte. It is not suitable for determination of complex real samples containing multiple analytes. In order to expand the practical application of imprinted polymers, novel dual-template magnetic molecularly imprinted polymers (MMIPs) were synthesized by surface polymerization using hydrocortisone and dexamethasone as the dual-template molecules in this study. The dual-template MMIPs were prepared by copolymerization on the surface of Fe3O4 @ SiO2-NH2, the template molecules, the functional monomer acrylamide (AM), the cross-linking agent ethylene glycol dimethacrylate

(EGDMA),

and

the

initiator

2,2-azobisisobutyronitrile.

The

morphology, magnetic properties and adsorption characteristics of the obtained dual-template MMIPs were studied by field emission scanning electron microscopy, dynamic light scattering, Fourier transform infrared spectroscopy, thermal gravimetric analysis, and vibrating sample magnetometry, and re-binding experiments. The results indicated that dual-template MMIPs had uniform particle size, strong magnetic

* Corresponding authors at: School of Chemistry and Chemical Engineering, Guangxi University for Nationalities, Nanning, Guangxi 530008, China. E-mail addresses: [email protected] (X. Li).

properties, high thermal stability, and good mass transfer rate. To investigate the selectivity of dual-template MMIPs, the template molecules were mixed along with their structural analogs. The dual-template MMIPs revealed a significantly higher adsorption amount for the template molecule than its structure analog. The dual-template MMIPs can be used for the enrichment and determination of hydrocortisone and dexamethasone in cosmetic products with the recoveries of spiked cosmetic samples ranging from 86.8-107.5% and 91.2-104.3%, respectively. The relative standard deviation (RSD) for hydrocortisone was <2.89%, and RSD for

dexamethasone was <2. 62%.

Keywords magnetic dual-template molecularly imprinted polymers · hydrocortisone · dexamethasone · cosmetic samples · HPLC

1. Introduction MIPs are prepared by molecular imprinting technology, and are designed to selectively recognize the target molecules[1]. Compared with traditional separation methods, MIPs offer better selectively for substrates, and are easy to prepare. However, there is often some sample loss when MIPs are used to extract the absorbed target molecules. In order to overcome this drawback, magnetic molecular imprinted polymers (MMIPs) are being increasingly used in sample treatment and separation applications[2,3]. MMIPs not only have excellent selectivity for the guest molecules, but also have magnetic response properties. Thus, after adsorption of the target analyte, the MMIPs can be easily collected in the presence of an external magnetic field without any filtration or centrifugation steps. The use of MMIPs can simplify the experimental steps, save time, and also save costs as the magnetic material can be recycled [4]. Currently, most of the imprinted polymers are prepared with a single template and have recognition sites only for one target molecule, so they cannot exhibit high affinity and selectivity for a family of analytes. For many applications with complex

samples, it is not practical to detect only one analyte at a time. Physically mixing individually imprinted polymers is also not feasible as it requires much time and effort for synthesizing several polymers. To address this drawback, MIPs have been prepared with two or more templates to increase their utility and expand their potential applications [5]. However, to the best of our knowledge, there is no publication till date that reports the simultaneous separation/enrichment of hydrocortisone and dexamethasone based on dual-template MMIPs. Glucocorticoids are an important class of corticosteroids that are widely used in clinical

practice,

because

of

their

broad-spectrum

antipyretic-analgesic,

anti-inflammatory, and anti-allergic properties[6]. Moreover, clinical studies have shown that these compounds can inhibit the proliferation of fibroblasts and helping in treatment of skin diseases such as psoriasis and eczema [7-9]. When glucocorticoids are

included in skin care products, they improve the smoothness and texture of skin, and prevent skin aging [10]. However, the long-term use of these drugs can cause skin thinning, redness and itching, as well as other conditions such as hyperglycemia, hypertension, osteoporosis, fetal abnormalities, diminished immune function and other side effects caused by absorption of the drug through the skin[11, 12]. The Health Ministry of People’s Republic of China Hygienic Standard for Cosmetics (2007) and EU Cosmetic Regulations[13] have clearly specified that the glucocorticoids are prohibited substances in cosmetics. In recent years, despite these regulations, some cosmetics manufacturers continue to add these banned substances in their products for treating skin conditions such as skin itching, seborrheic dermatitis, and even for whitening skin. Therefore, it is important to develop suitable analytical methods to determine glucocorticoids in cosmetics to ensure that the cosmetics products are safe for use. At present, the main methods for detection and determination of hormones in cosmetics are reversed-phase high performance liquid chromatography (RP-HPLC) [14-16], gas chromatography-mass spectrometry (GC-MS) [17], high performance liquid chromatography diode array detector (HPLC-DAD) [18], thin layer chromatography (TLC) [19, 20], ultra-performance liquid chromatography (UPLC)

[21], and high performance liquid chromatography tandem mass spectrometry method (HPLC-MS)[22]. The detection of trace components in a complex matrix is difficult with these methods and requires inclusion of suitable sample pretreatment techniques. Conventional sample preparation methods such as liquid-liquid extraction (LLE) [23-25]and solid-phase extraction (SPE)[26, 27] are sensitiveand accurate. However, these methods are often time consuming, complicated and require large amounts of solvents and reagents. Thus, the selection of an appropriate sample preparation method is key for the efficient analysis of complex samples. In this study, mixed-MMIPs were designed and synthesized using hydrocortisone and dexamethasone as the templates [28]. Figure 1 depicts the preparation process for the dual-template MMIPs, The double template imprinting polymer synthesis conditions were optimized and the resulting products were characterized by field emission scanning electron microscopy (FESEM), Dynamic light scattering (DLS), Fourier transform infrared spectroscopy (FT-IR), thermal gravimetric analysis (TGA), and vibrating sample magnetometry (VSM). The experiment results show that surface polymerization preparation of dual-template MMIPs has larger specific surface area, high the magnetic intensity, and quick absorption rates. The dual-template MMIPs were successfully used for the analysis of actual cosmetics samples and get better effective and feasible.

Fig.1 Schematic illustration of the preparation of dual-template MMIPs

2. Experimental

2.1 Chemicals Ethyl acetate, methanol, ethanol, and acetic acid were purchased from Guangdong Shantou Xilong Chemical Co., Guangdong, China. Hydrocortisone, dexamethasone, triamcinolone acetonide, and mifepristone were purchased from Hubei Xinyinhe Chemical Engineering Co., China. 3-Aminopropyl triethoxy silane (APTES) was purchased from Shanghai Kayon Biological Technology Co., Shanghai, China. Ethylene glycol dimethyl acrylate (EGDMA) was purchased from Fushun Anxin Chemical Co., China. 2,2-Azobisisobutyronitril (AIBN) was purchased from Shanghai No. 4 Reagent & H.V. Chemical Co., Ltd., China. Acetonitrile was purchased from Fisher Scientific (USA).The cosmetic samples were randomly purchased from a supermarket in Nanning (Guangxi, China).

2.2 Instrumentation and HPLC conditions The following instruments were used for the characterizations and sample analyses: UV-Vis 1800 spectrophotometer UV-Vis1800 (Suzhou Shimadzu Corporation); Fourier transform infrared spectrometer (Nicolet company); SU8020 SEM (Li Jing Company, Japan); 7410 vibrating sample magnetometer (Lake Shore Company, USA); Zetasizer Nano ZS(Malvern Instruments,UK); TG209F1 thermogravimetric analyzer (Netzsch, Germany); BS110S electronic balance (Sartorius Instrument Co. Ltd., Beijing, China); 98-1-B electronic temperature regulating electrothermal set (Taisite Instrument Co., Ltd. Tianjin City, China); and Hy-2a display multi-purpose control oscillator (Jiangsu Jintan City Medical Instrument Factory, China).The HPLC analyses were performed using an Agilent 1260 HPLC equipped with a G1314B variable wavelength UV–visible detector, a G1316A column oven, and a G1311C quat pump. A Phenomenex Gemini C18 (5 μm particle size, 250 mm × 4.6 mm) analytical column was used for the separation of analytes. The mobile phase was acetonitrile/H2O (35:65, v/v), and the flow rate was 1.0 mL/min at 35 °C. The injection volume was 20 μL, and the detector was set to monitor at 235 nm. 2.3 Preparation of dual-template MMIPs and MNIPs 2.3.1 Preparation of Fe3O4@ SiO2-NH2 Magnetic Fe3O4 NPs were synthesized by the co-precipitation method[29]. The surface of the as-prepared Fe3O4 NPs was modified using APTES to obtain Fe3O4@ SiO2-NH2. In a typical procedure, 2.0 g of the black magnetic Fe3O4 nanoparticles were dispersed in 350 mL of ethanol water (v/v, 1:1) with ultrasonication in a three-neck flask. Then, 2.5 mL of APTES was added, and the pH was adjusted to 4.0. The solution was stirred for 3 h at 60 °C. The resulting product was collected using an external magnet and washed with highly purified water until the solution pH was neutral. 2.3.2 Preparation of dual- template MMIPs and MNIPs

About 0.1 g of Fe3O4@SiO2-NH2 nanoparticles, 0.0725 g hydrocortisone, and 0.1137 g AM were accurately weighed out in a bottle. In a separate bottle, 0.1 g Fe3O4@SiO2-NH2 nanoparticles, 0.0785 g dexamethasone, and 0.1137 g AM were weighed out. The two mixtures were each dissolved in 75 mL of acetonitrile with ultrasonication for 15 min and incubated for 12 h. The pre-polymerization solution was transferred into a 500 mL three-neck flask and then. 3.962 g EGDMA and 0.06 g AIBN were dispersed into the above solution while stirring at 300 rpm for 15 min at room temperature to form a homogeneous dispersion. The polymerization solution was reacted at 60 °C for 24 h under the protection of nitrogen. After reaction, the product was separated by an external magnetic field from the solution, washed several times with methanol/acetic acid (v/v, 9:1) and finally with methanol to remove the template molecules, and then air-dried. The MNIPs were synthesized in the same manner as above except that the hydrocortisone and dexamethasone were not added. 2.4 Characterizations of molecularly imprinted polymers The particle size and morphology of the polymers were analyzed by field emission scanning electron microscopy (FESEM), the chemical structures of the magnetic polymers were studied by Fourier transform infrared spectroscopy (FT-IR), a vibrating sample magnetometer (VSM) was used for measuring the magnetic intensities of the polymers, and thermal stabilities of the magnetic polymers were studied by thermogravimetric analysis (TGA). 2.5 Binding experiments The binding experiments were performed with both MMIPs and MNIPs for comparison. In a typical procedure, a specific amount of the dual-template MMIPs/MNIPs was weighed out, and added to a certain concentration of hydrocortisone or dexamethasone solution in a conical flask. The mixture was covered and incubated for a certain time. Dual-template MMIPs/MNIPs which had reached adsorption equilibrium were separated from the above solution with the aid of

an external magnetic field. The concentration of hydrocortisone or dexamethasone in the supernatant was determined by UV-Vis measurements at 235nm or 237nm, and the amount of bound of hydrocortisone or dexamethasone was calculated by Equation (1).

Q  (C0  Ce )  V / W

(1)

Where Q is the amount of hydrocortisone or dexamethasone bound to MMIPs/MNIPs at equilibrium, C0 is the initial concentration of hydrocortisone or dexamethasone, C e is the concentration of free hydrocortisone or dexamethasone in the solution at equilibrium, V is the volume of solution, and W is the quality of dual-template MMIPs/MNIPs. The molecular recognition characteristics of dual-template MMIPs were evaluated by the partition coefficients of target molecules between the polymers and solutions. The partition coefficient K can be expressed as follows: K = Cp/Cs

(2)

Where, Cp is the combined amount of target molecules bound with dual-template MMIPs or MNIPs, and Cs is the residual concentration of target molecules in the solutions. The selectivity of values dual-template MMIPs and MNIPs were evaluated by estimating the imprinting factor ().When the value of the imprinting factor is high; it indicates that the binding capacity of MMIPs to the template is stronger than that of the MNIPs. This can be calculated as follows:  = KMIPs/KNIPs

(3)

Where, KMIPs and KNIPs are the partition coefficients of dual-template MMIPs and MNIPs with target molecules, respectively. 2.6 Separation enrichment and determination of hydrocortisone and dexamethasone in cosmetic samples Commercially available cosmetics (creams, lotions, and powders) were purchased from local supermarkets (Nanning, China) for spiked sample analysis. A mixed

standard solution of hydrocortisone and dexamethasone was prepared in acetonitrile with a concentration of 1 mg/mL. Then, 10 μL, 30 μL, 60 μL of the above mixed standard solution was added into 2.0 g of each cosmetic sample to obtain the spiked levels of 5 μg/g, 15 μg/g, and 30 μg/g. Then, 1 mL of saturated sodium chloride solution and 10 mL of acetonitrile were ultrasonically mixed and then added to the spiked sample. The mixture was centrifuged at 10000 rpm for 10 min. Then 5 mL of the supernatant was added to 200 mg of the dual-template MMIPs and shaken for 30 min at room temperature. After the extraction, the dual-template MMIPs were isolated from the extract using an external magnetic field, and the supernatant was discarded from the flask. The dual-template MMIPs were washed with 2 mL of acetonitrile to remove the impurities. The hydrocortisone and dexamethasone were eluted from the dual-template MMIPs with 3 × 2.0 mL of methanol. The obtained supernatant was evaporated to dryness under a stream of nitrogen. The residue was dissolved in 1.0 mL acetonitrile/water (35:65, V / V) and filtered through a 0.45 μm membrane for further HPLC analysis.

3. Results and discussion 3.1 Preparation of dual-template MMIPs and MNIPs The preparation process of dual-template MMIPs is shown in Fig.1.In this study, the Fe3O4 NPs were first synthesized by co-precipitation and then their surfaces were modified with APTES to give Fe3O4 @SiO2-NH2. Thus, the Fe3O4 NPs have polar surfaces. The SiO2-NH2 groups can not only shield the intermolecular forces between magnetite (Fe3O4) NPs, but also increase the dispersion of magnetite NPs in organic solvents. The surface of Fe3O4 was reacted with ATPES to introduce amino groups, which interact with the template molecular (hydrocortisone and dexamethasone) that contain

more

functional

group

such

as

hydroxyl

and

carboxyl

by

hydrogen-bonding.so amino-modified Fe3O4NPs, template molecular, and functional monomer

can

form

complex

through

non-covalent

interactions.

So

the

MMIPs/MMIPs shells were coated on the surface of Fe3O4 by the copolymerization of complex, cross-linking agent (EGDMA), the initiator (AIBN). In this study, the novel dual-template MMIPs were synthesized via the surface polymerization combined with nanotechnology. The Fe3O4@SiO2-NH2 NPs were used as the magnetic carriers and then coated by MIPs using entrapping method. The MIPs were covered on the surface of Fe3O4 @SiO2-NH2 by the copolymerization of a functional monomer (AM), a cross-linking agent (EGDMA), an initiator (AIBN), and the double template molecules (hydrocortisone and dexamethasone). Finally, the templates were removed to give the dual-template MMIPs. The template molecules were extracted from the polymer network by Soxhlet extraction with methanol/acetic acid (v/v, 9:1) until template molecules could not be detected at 235nm by UV-Vis. and finally were washed with pure methanol to remove residual acetic acid. The MNIPs were also prepared using the same process, but without adding template molecules of hydrocortisone and dexamethasone. 3.2 Characterization of dual-template MMIPs The size and shape of the Fe3O4 NPs, Fe3O4@SiO2-NH2, dual-template MMIPs, and MNIPs were examined by FESEM and DLS. From the FESEM images, the average diameter of Fe3O4 NPs was 20 nm (Fig. 2A). After modifying with APTES, the diameter of the resulting Fe3O4@SiO2-NH2 increased to 25 nm (Fig. 2B). The diameters of the MNIPs and dual-template MMIPs increased to 50 nm after the polymerization reaction. The FESEM images in Fig. 2 (A–D) show that Fe3O4, Fe3O4@SiO2-NH2, dual-template MMIPs, and MNIPs have uniform particle size distributions. As shown in the Fig.2, measured average diameter of Fe3O4 NPs, Fe3O4@SiO2-NH2, dual-template MMIPs, and MNIPs by DLS produced different result of FESEM; the most important reason was that DLS measured the hydrodynamic diameter of the particles, which include the solvent layer at the interface [30, 31].The results of FESEM and DLS indicated that the average diameter of the microballoons gradually increased, indicating that APTES was successfully

modified on the surface of Fe3O4. The MMIPs/MNIPs formed a core-shell structure, which helps their ability to bind to dual-template molecules in solution.

(A) Fe3O4（50K×）

(C) dual-template MMIPs（100K×）

(B) Fe3O4@SiO2-NH2（50K×）

(D) MNIP s（100K×）

Fig.2 FESEM and DLS images of (A) Fe3O4 NPs, (B) Fe3O4@SiO2-NH2, (C) dual-template MMIPs, and (D) MNIPs.

The FT-IR spectra of Fe3O4, Fe3O4@SiO2-NH2, dual-template MMIPs, and MNIPs were obtained using KBr disc, and are shown in Fig. 3. The peak at 580 cm1 can be attributed to the Fe–O stretching vibrations, and the peak at 3400 cm1 can be attributed to the O–H stretching vibration on the Fe3O4 surface (Fig. 3A). The stretching vibrations of Si-O-Si groups were observed at 1055 cm1. The peak at 1628 cm-1 caused by the bending vibrations of the N-H and - OH groups appeared as a peak at 1628 cm-1 (Figure 3B) and broadening of absorption peak at 3400 cm1 was also

observed. This indicates that the APTES was successfully modified on the surface of Fe3O4 NPs. Figs. 3C and 3D show the peaks of C=O stretching vibrations at 1750 cm1 and the C–H stretching and bending vibrations of methyl and methylene groups at 2925 cm1 and 2852cm1 respectively. These indicate that an AM-EGDMA layer was successfully formed on the surface of Fe3O4 @SiO2-NH2. Moreover, dual-template MMIPs (Fig. 3C) and MNIPs (Fig. 3D) showed almost the same characteristic peaks, indicating the complete removal of templates.

(A) 3400

(B)

1628

1055 580

(C) 1735

2925

630

(D) 2926

4000

3500

3000

1735

2500

2000

1500

1000

500

-1

Wavenumber(cm )

Fig.3 FT-IR spectra of (A) Fe3O4 NPs, (B) Fe3O4@SiO2-NH2, (C) dual-template MMIPs, and (D) MNIPs.

Vibrating sample magnetometry (VSM) was used to study the magnetic properties.

The

magnetic hysteresis

loops in Fig.

4 show that

Fe3O4,

Fe3O4@SiO2-NH2, and dual-template MMIPs are centrally symmetric and both the remanence and coercivity of the magnetic materials are zero, indicating that all the samples possessed superparamagnetism. The saturation magnetization values were 76.0, 57.3, and50.6 emu/g for Fe3O4, Fe3O4@SiO2-NH2, and dual-template MMIPs, respectively. The decrease in the magnetization values was attributed to the

modification of Fe3O4 surface with APTES and the imprinted polymer layer. Moreover, the inset of Fig. 4 shows the dual-template MMIPs suspended in solution without the external magnetic field (left) and separated by an external magnet (right). Even in the presence of an external magnetic field, the dual-template MMIPs binding target molecules could be easily separated.

Fig.4 Hysteresis loops of Fe3O4 (A), Fe3O4@SiO2-NH2 (B), and dual-template MMIPs(C). The insets show the particle separation and redispersion of the dual-template MMIPs

Thermogravimetric analysis (TGA) was performed under inert atmosphere (N2) in the temperature range from ambient temperature to 600 °C at a rate 10 °C/min [32]. The TGA curves for Fe3O4 and dual-template MMIPs are shown in Fig.5. For the Fe3O4, the loss of weight was 7.52% from room temperature to 600 °C. The weight loss of the dual-template MMIPs (1.95%) were not pronounce over the temperature range of 0−110 °C, which was caused by the remaining water or zero-gravity value of the solvent. The weight loss of dual-template MMIPs was 10.10% over the thermal degradation temperature range of 110−600 °C, which can be attributed to the thermal degradation of the imprinted polymers layer. The TGA analysis revealed that the weight loss of the dual-template MMIPs was more than that of Fe3O4, because of the

Fe3O4 particles were encapsulated by imprinted polymers.

Fig.5 TGA curves for Fe3O4 (A) and dual-template MMIPs (B).

3.3 Study of binding properties The binding properties of dual-template MMIPs towards hydrocortisone and dexamethasone were investigated by the adsorption of solvents, adsorption kinetics, adsorption thermodynamics and Scatchard analyses. The types and characteristics of adsorption solvents have a significant influence on the adsorption properties of the MMIPs. In the adsorption test, several types of solvents were investigated such as methanol, methanol-ethyl acetate (v/v, 9:1) and acetonitrile with hydrocortisone or dexamethasone solutions of 0.5 mg/mL concentration. The results are shown in Fig. 6. The adsorption capacity of dual-template MMIPs gradually decreased with increasing solvent polarity, because the hydroxyl functional group of methanol interfered with the hydrogen bond between the polymer functional groups and template molecules. Thus, the least polar acetonitrile was selected as the adsorption solvent in this experiment.

Fig.6 Adsorption of hydrocortisone and dexamethasone onto dual-template MMIPs and MNIPs in (A) methanol, (B) methanol: ethyl acetate (V: V, 9:1), and (C) acetonitrile.

The

binding

experiments

of

the

dual-template

MMIPs/MNIPs

with

hydrocortisone and dexamethasone were performed as described in the experimental section. Based on the UV-Vis measurements, adsorption capacities of the dual-template MMIPs/MNIPs were calculated using Equation (1). These data were then used to plot the adsorption isotherm of the MMIPs/MNIPs at different initial concentrations of hydrocortisone or dexamethasone, as shown in Figure 7. It can be seen that as the initial concentration of hydrocortisone or dexamethasone increased, up to 0.5 mg/mL, the adsorption capacity of MMIPs increased rapidly. When the initial concentration was greater than 0.5 mg/mL, the adsorption curve reached a relatively stable stage and the adsorption capacity of MMIPs reached saturation. The experimental results demonstrate that the adsorption capacity of MMIPs for the template molecule was significantly higher than that of MNIPs at the same initial concentration.

Fig.7 Adsorption isotherm of hydrocortisone and dexamethasone on dual-template MMIPs and MNIPs

To further investigate the binding characteristics of dual-template MMIPs towards hydrocortisone or dexamethasone, we used the Scatchard equation to analyze the adsorption isotherm data. The Scatchard equation is expressed as follows:

Q Ce  Qmax  Q K d Where Q is the adsorption capacity of dual-template MMIPs for template molecule at equilibrium, Q max is the apparent maximum adsorption capacity, and Ce is the free hydrocortisone or dexamethasone concentration at equilibrium. K d and Qmax can be calculated from the slope and intercept of the linear line, in the plot of Q / Ce versus Q. The results are shown in Figure 8.

A

Fig.8 Scatchard plots of Q/Ce versus Q

B

It can be seen in Figure 8A and 8B that there is a non-linear relationship between Q/Ce and Q for hydrocortisone or dexamethasone. However, the Scatchard curves of the MMIPs for hydrocortisone or dexamethasone can be divided into two parts that showed a good linear relationship. Fig. 8A is the Scatchard plot of the dual-template MMIPs for hydrocortisone. The left part of the curve can be expressed as Q/Ce =－12.2Q+364.9(r2=0.9999). Then, the K and Qmax values were calculated to be 0.0820 mg/mL and 29.9 mg/g from the slope and intercept. The right part of the curve is Q/Ce =－3.3Q+194.0 (r2=0.9481), and the corresponding K and Q max values were calculated to be 0.303 mg/mL and 58.7 mg/g using the slope and intercept. Fig. 8B is the Scatchard plot of the dual-template MMIPs for dexamethasone. The left part of the linear regression equation can be expressed as Q/C e =－ 14.0Q+424.8 (r2=0.9945), and the K and Qmax values were calculated to be 0.0715 mg/mL and 30.2 mg/g as before. The right part of the linear regression equation is Q/Ce =－2.58Q+193.2 (r2=0.9232), and the K and Qmax values were calculated to be 0.388 mg/mL and 75.4 mg/g. The results show that the adsorption isotherms of MMIPs conformed well to the Scatchard adsorption model. Hydrocortisone or dexamethasone could undergo adsorption on different sites of the MMIPs. Fig. 9 shows the adsorption kinetics curves of dual-template MMIPs and MNIPs for adsorption of hydrocortisone or dexamethasone at 0.5 mg/mL. The amount of hydrocortisone or dexamethasone adsorbed on the dual-template MMIPs/MNIPs showed a steep increase in the initial 20 min, indicating rapid adsorption. The dual-template MMIPs took 30 min to reach adsorption equilibrium while the MNIPs needed only 20 min. This difference is due to the spatial cavity structure of the dual-template MMIPs with some information of the target molecule and large specific surface area. The adsorption experiments showed that MMIPs possessed a higher initial adsorption rate and a higher saturated adsorption amount than MNIPs.

Fig.9 Adsorption kinetics of dual-template MMIPs and MNIPs

3.4 Selective recognition ability of dual-template MMIPs and MNIPs The specificity of the dual-template MMIPs was evaluated by the partition coefficient and the imprinting factor of the target molecule. To investigate the selectivity of the mixed-MMIPs towards hydrocortisone and dexamethasone, triamcinolone acetonide and mifepristone were used as the structural analogs. The adsorption capacity, distribution coefficient and imprinting factor values were determined for the dual-template MMIPs and MNIPs with a mixed solution of hydrocortisone, dexamethasone,

triamcinolone

acetonide,

and

mifepristone

of 0.5

concentration in 10 mL of acetonitrile. The results are shown in Table 1.

mg/mL

hydrocortisone

triamcinolone acetonide

dexamethasone

mifepristone

Fig.10 Structures of hydrocortisone, dexamethasone, triamcinolone acetonide, and mifepristone

From Fig. 10 and Table 1, we can see that the binding capacity and imprinting factor of the dual-template MMIPs toward the templates of hydrocortisone and dexamethasone were significantly higher than those of the structural analogs, triamcinolone acetonide and mifepristone. This is mainly because the dual-template MMIPs have high affinity and selectivity towards the template molecule. Fig. 10 shows the chemical structures of the template molecules and their structural analogs. The chemical structures of the four kinds of drugs show that they have a similar steroid ring containing functional groups in their structures, but the main difference between their structures is that mifepristone contains two benzene rings. Thus, mifepristone cannot easily enter the target molecule cavity in the dual-template MMIPs, likely due to the high steric hindrance and difficulty in coordinating with the adopted conformation of the molecule. Moreover, the dual-templates were imprinted on the surface of the magnetic core during preparation of the dual-template MMIPs. After removal of the template molecules, the complementary shape, size, spatial distribution and binding sites of imprinted cavities were formed in the dual-template

MMIPs. These results further verified the excellent imprinting efficiency of the present method.

Table.1 Adsorption capacities, partition coefficients, and imprinting factors of hydrocortisone, dexamethasone, triamcinolone acetonide, and mifepristone for the MMIPs and MNIPs QMMIPs mg/g

QMNIPs mg/g

KMMIPs mL/g

KMNIPs mL/g

hydrocortisone

24.4

8.6

59.3

18.8

3.15

dexamethasone

19.2

6.4

47.5

13.7

3.46

triamcinolone acetonide

15.4

8.6

36.4

18.8

1.93

mifepristone

1.4

1.2

2.84

2.43

1.17

analysis

α

The standard calibration curve was measured using standard solutions of hydrocortisone and dexamethasone in the concentration range from 0.5 to 15.0 mg/mL. A linear line was obtained in the plot for peak areas versus mass concentration for hydrocortisone and dexamethasone. The linear regression equations for hydrocortisone and dexamethasone can be expressed as S=39.622C+2.0058 and S=39.17C+6.0054, respectively. The corresponding correlation coefficient (r) values were 0.9973 and 0.9991. The limit of detection (LOD) and the limit of quantification (LOQ) were determined based on analyte concentration producing a signal-to-noise ratio of 3:1. The LOD and LOQ values for hydrocortisone and dexamethasone were 0.015 and 0.01 μg/mL, respectively. The accuracy and practicability of the method were verified by spiked recovery experiments, in which the mixed-MMIPs were used for selective separation and enrichment of hydrocortisone and dexamethasone in cosmetics. Cosmetic samples from different manufacturers purchased from different local markets were analyzed. About 2.0 g of lotion, toner, or mask was weighed out accurately and added with 10 μL, 40 μL, and 80 μL of standard acetonitrile mixed solution of hydrocortisone and

dexamethasone with the concentration of 1 mg/mL with three different spiked levels (5, 15, and 30 μg/g). Table 2 shows the recoveries of hydrocortisone and dexamethasone bound to dual-template MMIPs for spiked cosmetics. The recoveries of hydrocortisone and dexamethasone in lotion were in the range of 86.8 - 105.4% and 100.5 - 102.3%, respectively. The corresponding values for the toner were 86.8 105.4% and 100.5 - 102.3%, and values for the mask were 91.2-104.3% and 96.88 105.1%, respectively, while the relative standard deviation (RSD) was less than 2.62%. The results indicate a suitable improvement in precision and accuracy, and suggest that MMIPs can be directly used for selective adsorption and detection of hydrocortisone and dexamethasone in cosmetics. Table.2 Recoveries of hydrocortisone and dexamethasone bound to dual-template MMIPs in spiked cosmetic samples (n = 5) 5μg/g sample

analysis

Recovery （%）

30 μg/g

15μg/g RSD （%）

Recovery （%）

RSD

Recovery

RSD

（%）

（%）

（%）

hydrocortisone

94.9

0.95

107.5

1.32

92.2

2.89

dexamethasone

95.6

1.86

99.8

2.37

93.3

1.81

hydrocortison

86.8

1.25

105.4

0.89

100.1

0.54

dexamethasone

102.3

2.34

100.5

1.51

101.5

1.25

hydrocortison

96.8

1.34

105.1

1.58

98.4

1.22

dexamethasone

91.2

2.62

1. 42

100.5

2.57

lotion

toner

mask 104.3

The chromatograms of the spiked toner samples at a concentration of 5µg/g are shown in Fig. 11. The elution of mixed-MMIPs adsorbed with hydrocortisone and dexamethasone was performed with methanol, and displayed in Fig. 11. Fig. 11a, b and c represent the chromatograms of a blank cosmetic sample, a spiked cosmetic sample with a concentration of 5 μg/g, and the methanol elution of hydrocortisone and dexamethasone adsorbed by dual-template MMIPs, respectively. In Figure 11a, the

peak for the matrix appeared at 2~ 4 min in the HPLC chromatograms but the peaks for hydrocortisone and dexamethasone were not observed. The matrix peak did not change its position and the chromatographic peaks for hydrocortisone and dexamethasone appeared at 7.0 min and 11.0 min, respectively, as seen from Figure 11b with 5 μg/g spiked cosmetic sample. The peak of sample matrix between 2 ~ 4 min disappeared from Figure 11c, and the chromatographic peaks of hydrocortisone and dexamethasone were significantly enhanced compared to the 5 μg/g spiked cosmetic sample. The results indicated that dual-template MMIPs have the ability to purify target compounds. This study has demonstrated that the developed dual-template MMIPs method is a reliable approach to selectively enrich hydrocortisone and dexamethasone in cosmetics samples.

Fig. 11 Chromatograms of (a) blank cosmetic sample, (b) cosmetic sample spiked with hydrocortisone and dexamethasone at a concentration of 5 g/g, and (c) elution of hydrocortisone and dexamethasone on dual-template MMIPs washed with methanol.

4. Conclusions In this work, a novel method has been presented for the synthesis of dual-template MMIPs

with

uniform

core–shell

structure

by

using

hydrocortisone

and

dexamethasone as template molecules. The resulting mixed-MMIPs showed uniform particle size, good thermal stability, enhanced mass transfer rate of molecular targets, a highly improved imprinting effect, fast adsorption kinetics (adsorption equilibrium within 30 min) and high adsorption capacity. The method was applied successfully to extract hydrocortisone and dexamethasone in cosmetics samples. The selective separation and enrichment of hydrocortisone and dexamethasone from cosmetics samples and good recovery after a reasonably mild elution demonstrated that mixed-MMIPs can be applied for the selective separation and enrichment of hydrocortisone and dexamethasone at low concentration levels in cosmetic samples. Therefore, mixed-MMIPs coupled with HPLC is a promising method for the selective separation and determination of trace hydrocortisone and dexamethasone. Conflict of Interest The authors declare that they have no conflict of interest. Acknowledgements This work was supported by the National Natural Science Foundation of China (Nos. 21165003, 21545011) and the high-level-innovation team (guijiaoren[2014]7) and outstand-ing scholar project of Guangxi Higher Education Institutes.

References

[1]

X. Li, M. Li, J. Li, F. Lei, X. Su, M. Liu, P. Li, X. Tan, Synthesis and characterization of molecularly imprinted polymers with modified rosin as a cross-linker and selective SPE-HPLC detection of basic orange II in foods, Analytical Methods, 6 (2014) 6397-6406.

[2]

X. Su, X. Li, J. Li, M. Liu, F. Lei, X. Tan, P. Li, W. Luo, Synthesis and characterization of core–shell magnetic molecularly imprinted polymers for solid-phase extraction and determination of Rhodamine B in food, Food chemistry, 171 (2015) 292-297.

[3]

S.Wei, J. Li, Y. Liu, J. Ma, Development of magnetic molecularly imprinted polymers with double templates for the rapid and selective determination of amphenicol antibiotics in water, blood, and egg samples, Journal of Chromatography A, 1473 (2016) 19–27.

[4]

M. Liu, X.Y. Li, J.J. Li, X.M. Su, Z.Y. Wu, P.F. Li, F.H. Lei, X.C. Tan, Z.W. Shi, Synthesis of magnetic molecularly imprinted polymers for the selective separation and determination of metronidazole in cosmetic samples, Analytical and bioanalytical chemistry, 407 (2015) 3875-3880.

[5]

T. Jing, Y. Wang, Q. Dai, H. Xia, J. Niu, Q. Hao, S. Mei, Y. Zhou, Preparation of mixed-templates

molecularly imprinted polymers and investigation of the recognition ability for tetracycline antibiotics, Biosensors & bioelectronics, 25 (2010) 2218-2224. [6]

P.J. Barnes, Glucocorticosteroids: current and future directions, British journal of pharmacology, 163 (2011) 29-43.

[7]

E. Bornhövd, W.H. Burgdorf, A. Wollenberg, Macrolactam immunomodulators for topical treatment of inflammatory skin diseases, Journal of the American Academy of Dermatology, 45 (2001) 736-743.

[8]

D.T. Boumpas, G.P. Chrousos, R.L. Wilder, T.R. Cupps, J.E. Balow, Glucocorticoid therapy for immune-mediated diseases: basic and clinical correlates, Annals of internal medicine, 119 (1993) 1198-1208.

[9]

M. Schäfer-Korting, W. Mehnert, H.-C. Korting, Lipid nanoparticles for improved topical application of drugs for skin diseases, Advanced drug delivery reviews, 59 (2007) 427-443.

[10]

S. Schoepe, H. Schäcke, E. May, K. Asadullah, Glucocorticoid therapy‐induced skin atrophy, Experimental dermatology, 15 (2006) 406-420.

[11]

D. Huscher, K. Thiele, E. Gromnica-Ihle, G. Hein, W. Demary, R. Dreher, A. Zink, F. Buttgereit, Dose-related patterns of glucocorticoid-induced side effects, Annals of the Rheumatic Diseases, 68 (2009) 1119-1124.

[12]

H. Schäcke, W.-D. Döcke, K. Asadullah, Mechanisms involved in the side effects of glucocorticoids, Pharmacology & therapeutics, 96 (2002) 23-43.

[13]

O. Santoni, EU Cosmetics Regulation, Workshop on Natural Cosmetics–University of Bangor, 2015, pp. 2.

[14]

I. Baranowska, P. Markowski, J. Baranowski, Simultaneous determination of 11 drugs belonging to four different groups in human urine samples by reversed-phase high-performance liquid chromatography method, Analytica chimica acta, 570 (2006) 46-58.

[15]

J. Qu, Y. Qu, R.M. Straubinger, Ultra-sensitive quantification of corticosteroids in plasma samples using selective solid-phase extraction and reversed-phase capillary high-performance liquid chromatography/tandem mass spectrometry, Analytical chemistry, 79 (2007) 3786-3793.

[16]

A. Singh, C. Roopkishora, R. Gupta, S. Kumar, M. Kumar, Development and validation of reversed-phase high performance liquid chromatographic method for hydroxychloroquine sulphate, Indian journal of pharmaceutical sciences, 77 (2015) 586.

[17]

L. Amendola, F. Garribba, F. Botrè, Determination of endogenous and synthetic glucocorticoids in human urine by gas chromatography–mass spectrometry following microwave-assisted derivatization, Analytica chimica acta, 489 (2003) 233-243.

[18]

H. Gao, B. Li, J. Wang, Application of HPLC-DAD-MS/MS to the screening and quantification of 14 adrenocortical hormones adulterated in Chinese patent drugs, Chinese Journal of Pharmaceutical Analysis, 35 (2015) 1458-1465..

[19]

L. Gagliardi, D. De Orsi, M.R. Del Giudice, F. Gatta, R. Porrà, P. Chimenti, D. Tonelli, Development of a tandem thin-layer chromatography–high-performance liquid chromatography method for the identification and determination of corticosteroids in cosmetic products, Analytica Chimica Acta, 457 (2002) 187-198.

[20]

R. Xia, S. Dong, B. Che, Identification of corticosteroids in traditional Chinese medicine by means of TLC method. Chinese Journal of Pharmaceutical Analysis,

[21]

28 (2008) 470-471.

M. Gros, S. Rodríguez-Mozaz, D. Barceló, Fast and comprehensive multi-residue analysis of a broad range of human and veterinary pharmaceuticals and some of their metabolites in surface and

treated waters by ultra-high-performance liquid chromatography coupled to quadrupole-linear ion trap tandem mass spectrometry, Journal of Chromatography A, 1248 (2012) 104-121. [22]

X. Liu, J. Feng, X. Sun, Y. Li, G. Duan, Three-layer structure graphene/mesoporous silica composites incorporated with C8-modified interior pore-walls for residue analysis of glucocorticoids in milk by liquid chromatography–tandem mass spectrometry, Analytica chimica acta, 884 (2015) 61-69.

[23]

V.A. Frerichs, K.M. Tornatore, Determination of the glucocorticoids prednisone, prednisolone, dexamethasone, and cortisol in human serum using liquid chromatography coupled to tandem mass spectrometry, Journal of Chromatography B, 802 (2004) 329-338.

[24]

Y. Luo, C. Uboh, L. Soma, F. Guan, J. Rudy, D. Tsang, Simultaneous analysis of twenty‐one glucocorticoids in equine plasma by liquid chromatography/tandem mass spectrometry, Rapid communications in mass spectrometry, 19 (2005) 1245-1256.

[25]

M. Mazzarino, X. de la Torre, F. Botrè, A screening method for the simultaneous detection of glucocorticoids, diuretics, stimulants, anti-oestrogens, beta-adrenergic drugs and anabolic steroids in human urine by LC-ESI-MS/MS, Analytical and bioanalytical chemistry, 392 (2008) 681-698.

[26]

B. Shao, X. Cui, Y. Yang, J. Zhang, Y. Wu, Validation of a solid-phase extraction and ultra-performance liquid chromatographic tandem mass spectrometric method for the detection of 16 glucocorticoids in pig tissues, Journal of AOAC International, 92 (2009) 604-611.

[27]

S. Tong, Q. Liu, Y. Li, W. Zhou, Q. Jia, T. Duan, Preparation of porous polymer monolithic column incorporated with graphene nanosheets for solid phase microextraction and enrichment of glucocorticoids, Journal of Chromatography A, 1253 (2012) 22-31.

[28]

C. Baggiani, P. Baravalle, C. Giovannoli, L. Anfossi, G. Giraudi, Molecularly imprinted polymers for corticosteroids: Analysis of binding selectivity, Biosensors and Bioelectronics 26 (2010) 590–595.

[29]

B.Wang, F. Zhang, J. Qiu, X. Zhang, H. Chen, Y. Du, P. Xu, Preparation of Fe3O4

Superparamagnetic Nanocrystals by Coprecipitation with Ultrasonic Enhancement and Their Characterization. Acta Chimica Sinica, 67 (2009) 1211-1216. [30] O.U. Rahman, S.C. Mohapatra, S. Ahmad. Fe 3O4, inverse spinal super paramagnetic nanoparticles, Materials Chemistry & Physics, 132(2012):196–202. [31] J. Wang, G. Xu, F. Wei, J. Yang, P. Zhou, Q. Hu, A novel Fe3O4/CdTe fluorescence probe for sialic acid detection based on a phenylboronic acid–sialic acid recognition system, Rsc Advances, 6 (2015) 481-488.

[32] J. Pan, B. Wang, J. Dai, X. Dai, H. Hang, H. Ou, Y. Yan, Selective recognition of 2,4,5-trichlorophenol by temperature responsive and magnetic molecularly imprinted polymers based on halloysite nanotubes. Journal of Materials Chemistry, 22 (2012) 3360-3369.

Selective separation and determination of glucocorticoids in cosmetics using dual-template magnetic molecularly imprinted polymers and HPLC Min Liu a, Xiaoyan Li a*, Junjie Li b, Zongyuan Wu a, Fang Wang a , Li Liu a, Xuecai Tan a, Fuhou Lei a. a. School of Chemistry and Chemical Engineering, Guangxi University for Nationalities, Key Laboratory of Guangxi Colleges and Universities for Food Safety and Pharmaceutical Analytical Chemistry, Guangxi Key Laboratory of Chemistry and Engineering of Forest Products, Guangxi Colleges and Universities Key Laboratory of Utilization of Microbial and Botanical Resources, Nanning, Guangxi 530008, China b.School of Chemistry and Chemical Engineering, Guangxi University, Nanning, Guangxi 530004,China

Corresponding Author ∗

Telephone:

+86-771-3260558. Fax: +86-771-3260558.

E-mail: [email protected]

Graphical

Abstract