Preparation and evaluation of molecularly imprinted polymer for selective recognition and adsorption of gossypol

Abstract

Molecularly imprinted polymers (MIPs) were designed and prepared via bulk thermal polymerization with gossypol as the template molecule and dimethylaminoethyl methacrylate as the functional monomer. The morphology and microstructures of MIPs were characterized by scanning electron microscope and Brunauer‐Emmett‐Teller surface areas. Static adsorption tests were performed to evaluate adsorption behavior of gossypol by the MIPs. It was found that adsorption kinetics and adsorption isotherms data of MIPs for gossypol were fit well with the pseudo‐second‐order model and Freundlich model, respectively. Scatchard analysis showed that heterogeneous binding sites were formed in the MIPs, including lower‐affinity binding sites with the maximum adsorption of 252 mg/g and higher‐affinity binding sites with the maximum adsorption of 632 mg/g. Binding studies also revealed that MIPs had favorable selectivity towards gossypol compared with non‐imprinted polymers. Furthermore, adsorption capacity of MIPs maintained above 90% after 5 regeneration cycles, indicating MIPs were recyclable and could be used multiple times. These results demonstrated that prepared MIPs could be a promising functional material for selective adsorption of gossypol.

1 INTRODUCTION

Gossypol (1,1′,6,6′,7,7′‐hexahydroxy‐3,3′‐dimethyl‐5,5′‐diisopropyl‐(2,2′‐binaphthalene)‐8,8′‐dicarboxaldehyde, C₃₀H₃₀O₈) is a polyphenolic aldehydic compound extracted from cotton plants. Numerous studies have indicated that gossypol is toxic to monogastric animals as well as young ruminants.1-4 Its poisonous side effects occur in reproductive diseases,5 growth depression,6 and intestinal organ abnormalities.7 China is one of the largest cotton producers worldwide. Cotton seed meal, as one of the main byproduct of cotton industry, is an important part of animal protein feed resources due to its high nutritive value. However, the highly toxic gossypol in cotton seed poses a big hurdle for its practical application. Hence, to remove gossypol from cotton seed products is crucial. For this purpose, different methods have been developed, including solvent extraction treatment,8 ferrous sulfate treatment,9 microbes fermentation,10 and adsorption treatment.11, 12 Among them, adsorption approach has some advantageous characteristics, such as simple operation and low cost. Along this line, alumina, silica, and synthetic magnesium silicates were studied as adsorbents target gossypol.11, 12 However, functional materials based on molecularly imprinted polymers (MIPs) with high specificity and selectivity has been rarely reported.

Molecularly imprinted polymers, which mimic nature's enzyme and antibody, are tailored materials with high affinity and selectivity for target molecules.13-19 Molecularly imprinted polymers are prepared by polymerization of functional monomer and cross‐linker in the presence of template molecule. Specific recognition sites are formed by non‐covalent or covalent interactions between the functional monomer and the template molecule in the cross‐linking polymerization process. After the template molecule is removed, generates 3‐dimensional cavities which are complementary to the template in terms of shape, size, and functional groups. These cavities can rebind template with high specificity. The MIPs have unique advantages, such as facile preparation, low cost, reusability, and excellent stability compared with natural receptors. The MIPs have been extensively reported in solid‐phase extraction,20 chromatographic analysis,21 enantiomeric separation,22-24 sensors,25 drug delivery,26 and enzyme‐like catalysts.27

Along these lines, it is highly desirable and practically useful to design and synthesize functional materials as adsorbent which feature high affinity and specificity for gossypol. In the work reported here, MIPs were designed and prepared using gossypol as the template, dimethylaminoethyl methacrylate (DMAEMA) as the functional monomer, and ethylene glycol dimethacrylate (EGDMA) as the cross‐linker by bulk thermal polymerization. To evaluate specific recognition and adsorption performance of the MIPs for gossypol, adsorption kinetics, adsorption isotherms, adsorption selectivity, and reusability were investigated in detail. Results suggested that the MIP had high adsorption capacity and improved selectivity for gossypol via molecular imprinting.

2 EXPERIMENTAL SECTION

2.1 Materials and regents

Gossypol from cotton seeds was purchased from Sigma‐Aldrich. Dimethylaminoethyl methacrylate, 2‐aminoethyl methacrylate hydrochloride, diethylaminoethyl methacrylate, 4‐vinyl pyridine, EGDMA, divinylbenzene, 1,1′‐bi‐2‐naphthol, and 2,2′‐azobis(isobutyronitrile) were purchased from Adamas Reagent Co., Ltd (Shanghai, China). Hexaphenol was purchased from WAKO Pure Chemical Industry, Ltd (in Japan). Dichloromethane, choloroform, methaol, and tetrahydrofuran were Chromatographic Grade. Water was purified by a Millipore (Billerica, Massachusetts) Milli‐Q gradient system to high‐performance liquid chromatography grade.

2.2 Preparation and characterization of MIPs and NIPs

2.2.1 Synthesis of MIPs and NIPs

Gossypol (0.083 mmol, 43 mg), DMAEMA (1 mmol, 157 mg), EGDMA (5 mmol, 990 mg), and 2,2′‐azobis(isobutyronitrile) (0.27 mmol, 44 mg) were dissolved in 4.0‐mL dichloromethane in a screw‐capped vial. This mixture was sonicated for 5 min and degassed for 15 min under nitrogen. The vial was sealed then immersed in a water bath at 65°C for 24 h. As a blank, a non‐imprinted control polymer (NIP) was synthesized following the same protocol but without template. The resulting polymer monoliths were crushed and ground with a mortar and pestle to a fine powder. The powder was sieved using 200 mesh sieve. Particle size fraction below 75 μm was collected.

2.2.2 Removal of template molecules

For MIPs, the unreacted species and templates were removed by Soxhlet extraction with methanol for 24 h and then washed with 10.0 mmol/L NaOH aqueous solutions for 2 h per cycle until no template molecules were detected by a UV‐Vis spectroscopy. Finally, the MIPs were washed with water up to be neutral. For NIPs, the particles were washed by Soxhlet extraction with methanol for 24 h to remove the unreacted substances. Finally, both MIP and NIP were dried at 60°C using the vacuum drying oven overnight.

2.3 Characterization techniques

Binding capacity of the MIPs and NIPs was measured by UV‐Vis spectroscopy (UV‐2600, SHIMADZU, Japan). The surface morphology analysis of MIPs and NIPs were characterized by field‐emission scanning electron microscopy (SUPRA 55VP, Zeiss, Germany). The Brunauer‐Emmett‐Teller (BET) specific surface areas of MIPs and NIPs were obtained with an Autosorb Chemisorption/Physisorption Analyzer (Quantachrome, Florida) by nitrogen adsorption at 77.3 K using the BET method to calculate the specific surface areas.

2.4 Static adsorption

2.4.1 Adsorption kinetic experiments

Molecularly imprinted polymers (9 samples, 10 mg each) and NIPs (9 samples, 10 mg each) samples were added to 5 mL of 300 mg/L gossypol methanol solution. The mixtures were incubated in a shaker and were sampled at time intervals of 30 min, 60 min, 140 min, 240 min, 370 min, 540 min, 680 min, 770 min, and 900 min. Each time sample was passed through a filter (0.2‐μm polytetrafluoroethylene syringe filters) to separate particles from supernatant. Residual concentrations of gossypol in filtrate were quantified by measuring the UV absorbance at 373 nm. The adsorption capacity can be calculated according to the following equation (Equation 1):

(1)

where Q _t (mg/g) is adsorption capacity of MIPs and NIPs at different time intervals; C ₀ (mg/L) and C _t (mg/L) are the initial and residual concentrations of gossypol, respectively; V (L) is the volume of gossypol solution; and m (g) is the mass of the absorbent.

2.4.2 Adsorption isotherms

Molecularly imprinted polymers (10 samples, 10 mg each) and NIPs (10 samples, 10 mg each) samples were added to 5 mL of gossypol methanol solution with an initial concentration from 200 to 2000 mg/L (C ₀, mg/L: 200, 300, 400, 500, 600, 800, 1000, 1400, 1700, 2000). The mixtures were shaken continually for 12 h. Then the samples were passed through a filter (0.2‐μm polytetrafluoroethylene syringe filters) to separate particles from supernatant. The equilibrium concentrations of gossypol (C _e, mg/L) were measured by absorbance change at 373 nm. The equilibrium adsorption capacity (Q _e, mg/g) was calculated according to the following equation (Equation 2):

(2)

where V (L) is the volume of gossypol solution and m (g) is the mass of MIP and NIP.

2.5 Adsorption selectivity studies

For selectivity test, 2 different analogues, namely, 1,1′‐bi‐2‐naphthol and hexaphenol were chosen as control. The MIPs (10 mg) and NIPs (10 mg) samples were added to 5 mL of 1.0 mmol/L gossypol and 1,1′‐bi‐2‐naphthol and hexaphenol solutions, respectively. After 12 h under shaking, the samples were filtered. The equilibrium concentrations (C_e, mg/L) of gossypol and 2 analogues were determined. Distribution coefficients (K _d) (Equation 3), selectivity coefficients (k ) (Equation 4), and relative selectivity coefficients (k ′) (Equation 5) of gossypol, 1,1′‐bi‐2‐naphthol, and hexaphenol can be calculated according to the following equations:

(3)

where K _d represents the distribution coefficient (L/g); C ₀ (mg/L) and C _e (mg/L) are initial and equilibrium concentrations of gossypol and control molecules, respectively; V (L) is the volume of the solution; and m (g) is the mass of the absorbent.

(4)

In Equation 4, k is the selectivity coefficient, K _d (gossypol) and K _d (C) are the distribution coefficients of gossypol and the 2 structural analogues, respectively.

(5)

In Equation 5, k ′ is the relative selectivity coefficient; k _MIP and k _NIP are the selectivity coefficients of MIPs and NIPs, respectively.

2.6 Desorption and regeneration of MIPs

Reusability is an important factor for functional absorption materials. The reusability of MIPs was investigated through 5 desorption‐adsorption cycles. Fifteen milligrams of MIPs was added into 5‐mL gossypol solution in methanol with an initial concentration of 400 mg/L in the centrifuge tube. After shaking for 12 h, MIP particles were centrifuged and supernatants were passed through a filter. The equilibrium concentration of gossypol was measured by UV‐Vis spectrophotometer. The used MIPs were washed with 10 mmol/L NaOH aqueous solution to remove residual templates and washed with water up to be neutral and then washed with methanol. Finally, the used MIP particles were dried under vacuum at 60°C. The recovered MIPs were tested for gossypol adsorption for multiple times.

3 RESULTS AND DISCUSSION

3.1 Preparation of MIPs and NIPs

The template molecule, gossypol, containing 6 phenolic hydroxyl functional groups, is an acidic compound with pKa 28 of about 6.5. The acidity of gossypol is between phenols (pKa  ~ 10) and carboxylic acids (pKa  ~ 5).29 Theoretically, the strong acid‐base interaction can be formed between gossypol and basic functional monomer. Therefore, acid‐base interaction was chosen as working mechanism for designing organic functional material for specific recognition and adsorption of gossypol (Scheme 1). Along this line, several functional monomers, including DMAEMA, 2‐aminoethyl methacrylate hydrochloride, diethylaminoethyl methacrylate, and 4‐vinyl pyridine bearing basic amine groups are tested. The structures of functional monomers are shown in Figure 1. According to binding capacity and imprinting effect, DMAEMA was found to be desirable and optimal and was used as functional monomer in subsequent studies. Cross‐linkers were also screened using EGDMA and divinylbenzene. Ethylene glycol dimethacrylate was found to be superior. Chloroform, dichloromethane, and tetrahydrofuran were tested as imprinting solvents, and dichloromethane was found to be optimal.

Scheme 1

Schematic illustration of designing organic functional material for specific recognition and adsorption of gossypol based on molecular recognition and imprinting

Figure 1

Structures of template gossypol, cross‐linkers, and functional monomers in molecularly imprinted polymers synthesis

Under optimized conditions, DMAEMA was chosen as the functional monomer; MIPs were prepared by bulk thermal polymerization method. Dimethylaminoethyl methacrylate containing tertiary amino groups is basic with pKa 30 about 8.4. In this work, the basic working mechanism is acid‐base interaction. Dimethylaminoethyl methacrylate provides tertiary amino groups, which could strongly interact with phenolic hydroxyl groups on the template to form specific recognition sites. Therefore, the template gossypol can be selectively bound through reversible ionic interaction to the MIP matrix. Scheme 2 depicts the mechanism of acid‐base interaction and the molecular imprinting process of gossypol within DMAEMA/EGDMA system.

Scheme 2

Schematic illustration of ionic interaction between functional monomer DMAEMA and template gossypol and the synthesis process of molecularly imprinted polymer by bulk polymerization

3.2 Morphology analysis

The surface morphology of MIPs and NIPs particles were characterized with scanning electron microscope and were shown in Figure 2. As can be seen in Figure 2A and 2B, both NIPs (Figure 2A) and MIPs (Figure 2B) particles were irregular with a broad size distribution which was due to that the polymer particles were grinded with a mortar and pestle before sieving. Compared with surface morphology of NIPs (Figure 2C and 2E), MIPs (Figure 2D and 2F) was rough, loose, and porous with many cavities, which were resulting from imprinting process.

Figure 2

Scanning electron microscope images of (A, C, and E) NIPs and (B, D, and F) MIPs under different magnifications

3.3 Binding properties of the MIP

Binding capacity was measured by UV‐Vis spectroscopy. Figure 3 showed UV spectrum of 0.05 mmol/L gossypol solutions in methanol. The adsorption peak at 373 nm was chosen for the quantitative analysis. The standard curve with a linear fitting (the inset in Figure 3) was established to determine the concentration of gossypol solution before and after binding studies.

Figure 3

The UV spectrum of 0.05 mmol/L gossypol solution in methanol; the inset shows standard curve of gossypol solution; a linear relationship between absorbance and concentration is established

3.3.1 Adsorption kinetics

Adsorption kinetic experiments for MIPs and NIPs were performed to investigate the binding process in detail. Figure 4A presented the adsorption kinetic curves for 300‐mg/L gossypol. Results showed that for MIP, the adsorption capacity increased rapidly in the first 60 min, then slowed down, and eventually reached adsorption equilibrium in 12 h. The observation was consistent with slow‐binding kinetics of bulk molecular imprinting process which was probably due to the relatively large size of gossypol compared with small template molecules.31-33 Figure 4A also showed that the amount of gossypol bound by MIPs was 15% higher than that of NIPs. The higher binding capacity was attributed to specific binding sites in MIPs and thus stronger affinity toward gossypol than that of NIPs.

To further study the adsorption kinetics, the pseudo‐first‐order model34 (Equation 6) and pseudo‐second‐order model35 (Equation 7) were used to fit the data of adsorption kinetics to investigate mass transfer process. The pseudo‐first‐order rate equation is given as follows:

(6)

where Q _e (mg/g) and Q _t (mg/g) are the binding capacity of gossypol adsorbed at equilibrium and at time t (min), respectively, and k ₁ (min⁻¹) is pseudo‐first‐order adsorption rate constant.

The pseudo‐second‐order equation based on equilibrium adsorption capacity is expressed in (Equation 7).

(7)

where k ₂ (mg/g∙min) is the rate constant of pseudo‐second‐order adsorption model.

Figure 4

A, Adsorption kinetic curves for gossypol of MIPs and NIPs. B, Two kinetic models fit for gossypol MIPs

The binding data of MIPs was fitted to 2 kinetic models and was shown in Figure 4B. The results of kinetic parameters and correlation coefficients (R ²) were listed in Table 1. The R ² values for second‐order were higher than that of first‐order, the Q _e calculated by pseudo‐second‐order model (138.6 mg/g) were very close to the experimental Q _e (139 mg/g). These results suggested that the pseudo‐second‐order mechanism was predominant.

Table 1. Kinetic parameters for the adsorption of gossypol on molecularly imprinted polymer

Kinetic models	Q e, mg/g	k	R ^2
Pseudo‐first‐order	126.8	0.01871 min^−1	0.9006
Pseudo‐second‐order	138.6	0.0001843 mg/g∙min	0.9633

3.3.2 Adsorption isotherms

To estimate the adsorption capability of the MIPs and NIPs, the equilibrium adsorption isotherm studies were conducted with the initial concentration of gossypol range of 200 to 2000 mg/L in methanol in the presence of 10 mg of the MIPs and NIPs. As shown in Figure 5A, with increasing concentration of gossypol, the gossypol adsorption capacity at the time of adsorption equilibrium of MIPs and NIPs both gradually increased. The MIPs outperformed NIPs (13%‐15% higher) on adsorption capacity of gossypol. Results indicated that specific binding sites were generated for gossypol in MIPs which was complementary to the gossypol in both shape and functionality. To rule out the possibility that the binding differences between MIPs and NIPs were due to morphology, BET surface areas were measured and were shown in Table 2. Results showed that MIPs had a surface area that was 7 times lower than NIPs (13.9 m²/g vs 91.1 m²/g). If it were all nonspecific interactions during binding process, the binding capacity of MIPs would be 7 times lower than that of NIPs. However, results showed that adsorption capacity of MIPs were 13% to 15% higher than that of NIPs providing strong evidence that this was due to molecular imprinting process.

Figure 5

A, Adsorption isotherm of molecularly imprinted polymers (MIPs) and non‐imprinted polymers (NIPs) for gossypol. B, Binding isotherm of MIPs fitted to Langmuir and Freundlich models

Table 2. Surface area and average pore diameter of non‐imprinted polymer (NIP) and molecularly imprinted polymer (MIP)

Samples	Surface area, m^2/g	Average pore diameter, nm
NIP	91.1	6.57
MIP	13.9	40.4

Two well‐established isothermal models, Langmuir (Equation 8) and Freundlich (Equation 9) and the Scatchard equation (Equation 10) were adopted to further evaluate the binding process and adsorption parameters of the MIPs. Langmuir model is based on monolayer adsorption on surface with finite number of identical sites.36 Freundlich model is valid for adsorption on a heterogeneous surface.37 Scatchard equation is known to be sensitive to heterogeneous and cooperative binding effect. They are given as follows:

(8)

(9)

(10)

where Q _e (mg/g) and C _e (mg/L) are the equilibrium adsorption capacity and the equilibrium concentration of gossypol, respectively; Q _m (mg/g) is the maximum adsorption capacity of the MIP; K _L (L/mg) is Langmuir constant, which is related to affinity of the binding sites; K _F and n are both Freundlich constants indicating adsorption capacity and intensity; K _D (mg/L) is the equilibrium dissociation constant.

As shown in Figure 5B, the Langmuir and Freundlich adsorption plots for MIPs were obtained by using nonlinear regression and the results of adsorption isotherm parameters listed in Table 3. The R ² values indicated that fit was better using Freundlich model (R ² = 0.9913) than with Langmuir model (R ² = 0.9527). It implied multilayer adsorption was occurring on heterogeneous surface. And value of Freundlich constant n is 2.75, suggesting the favorable adsorption process.

Table 3. Adsorption isotherm parameters of MIP for gossypol

Isotherm models	K	Q m, mg/g	n	R ^2
Langmuir	0.00869, L/mg	606.9	—	0.9527
Freundlich	49.78, mg/g	—	2.751	0.9913

The isotherm data were further analyzed by using Scatchard model which was shown in Figure 6. Scatchard plot for MIPs was not a single straight line, indicating that the binding sites in the MIPs were heterogeneous.38-40 The Scatchard plot contained 2 different linear regression lines, suggesting 2 types of binding sites so called higher‐affinity and the lower‐affinity binding sites, were generated in MIPs. Linear regression parameters of the Scatchard plot were summarized in Table 4. As can be seen, the higher‐affinity binding sites dominated at lower concentrations range 200 to 500 mg/L. While the lower‐affinity binding sites prevailed at higher concentrations range 500 to 2000 mg/L. The Q _max values at higher concentrations were 632 mg/g. The experimental Q _e values at the highest concentration were 564 mg/g. Such high binding capacity also demonstrated that so prepared MIPs had outstanding binding affinity toward the target template gossypol.

Figure 6

Scatchard plot of molecularly imprinted polymers (MIPs) for gossypol

Table 4. Adsorption parameters from Scatchard analysis of molecularly imprinted polymers

C 0, mg/L	Regression equation	K D, mg/L	Q m mg/g	R ^2
200‐500	Q e/C e = 19.6 − 0.0776Q	12.9	252	0.9761
500‐2000	Q e/C e = 4.56 − 0.00722Q	138.5	632	0.8971

3.4 Adsorption selectivity of the MIPs

The selective adsorption study of MIPs was performed by gossypol and other two structural analogues, namely 1,1′‐bi‐2‐naphthol and hexaphenol (Figure 7). A comparison of binding capacity and the data of the distribution coefficients K _d, selectivity coefficients k and relative selectivity coefficients k ′ were obtained and summarized in Figure 8 and Table 5. MIP imprinted with gossypol demonstrated much higher adsorption capacity for gossypol compared with structural analogues. The k ′ values were 1.69 and 2.15, indicating that molecular imprinting process was indeed successful. The high selectivity for gossypol could be resulted from the differences in chemical structures among gossypol, 1,1′‐bi‐2‐naphthol, and hexaphenol. It was probably due to specific binding sites generated by imprinting and those sites matched well with gossypol in terms of size, shape and three‐dimensional functionality.

Figure 7

Two control molecules used for selectivity study for gossypol imprinted polymers

Figure 8

Selectivity study for gossypol imprinted polymers using 2 structural analogues as control. MIP, molecularly imprinted polymer; NIP, non‐imprinted polymer

Table 5. Distribution coefficient (K _d), selectivity coefficient (k ), and relative selectivity coefficients (k' ) for molecularly imprinted polymers (MIPs) and non‐imprinted polymers (NIPs)

	MIPs			NIPs			k ′
	Q MIP, mg/g	K d, L/g	k	Q NIP, mg/g	K d, L/g	k
Gossypol	226	3.37	—	199	1.66	—	—
Bi‐2‐naphthol	3.65	0.0131	257	3.05	0.0109	152	1.69
Hexaphenol	13.1	0.0384	87.8	13.8	0.0406	40.9	2.15

3.5 Regeneration and reusability of MIPs

As a promising adsorption material, desorption and regeneration ability are very important for practical application. To test the reusability of MIPs, the adsorption‐desorption cycles were repeated 5 times using the same MIP particles. Figure 9 showed that there was only a slight decrease (6.5%) in binding capacity after 5 regeneration cycles. These results indicated the MIP held great application potential and could be used multiple times without significant decrease in binding capacity.

Figure 9

Regeneration and reusability studies of gossypol imprinted molecularly imprinted polymers

4 CONCLUSION

The gossypol imprinted MIPs were successfully designed and prepared with DMAEMA as the functional monomer by bulk thermal polymerization. Adsorption performances of MIPs were evaluated according to adsorption kinetics, adsorption isotherms, adsorption selectivity, and reusability studies. The so‐prepared MIPs showed excellent recognition ability for gossypol with the binding capacity as high as 564 mg/g, improved selectivity, and excellent reusability. Scatchard analysis showed that 2 classes of binding sites existed in the gossypol‐imprinted polymers. The binding sites in the MIP were heterogeneous; higher‐affinity and lower‐affinity binding sites were both produced in MIPs.

ACKNOWLEDGEMENT

This work was financially supported by the Natural Science Foundation of Xinjiang, China (2015211A046).