Adsorption of catechin onto cellulose and its mechanism study: Kinetic models, characterization and molecular simulation

Accepted Manuscript Adsorption of catechin onto cellulose and its mechanism study: Kinetic models, characterization and molecular simulation

Yujia Liu, Danyang Ying, Luz Sanguansri, Yanxue Cai, Xueyi Le PII: DOI: Reference:

S0963-9969(18)30491-5 doi:10.1016/j.foodres.2018.06.044 FRIN 7711

To appear in:

Food Research International

Received date: Revised date: Accepted date:

19 March 2018 17 June 2018 20 June 2018

Please cite this article as: Yujia Liu, Danyang Ying, Luz Sanguansri, Yanxue Cai, Xueyi Le , Adsorption of catechin onto cellulose and its mechanism study: Kinetic models, characterization and molecular simulation. Frin (2018), doi:10.1016/j.foodres.2018.06.044

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.

ACCEPTED MANUSCRIPT Adsorption of catechin onto cellulose and its mechanism study: kinetic models, characterization and molecular simulation

Department of Applied Chemistry, South China Agricultural University, Guangzhou

SC

1

RI

PT

Yujia Liu1,2,3, Danyang Ying2, Luz Sanguansri2, Yanxue Cai2,4, Xueyi Le1,*

510642, China

School of Chemical Engineering and Energy Technology, Dongguan University of

MA

3

CSIRO Agriculture & Food, 671 Sneydes Road, Werribee, Vic 3030, Australia

NU

2

Technology, Dongguan 523808, China College of Food Sciences, South China University of Technology, Guangzhou

PT E

D

4

510640, China

CE

*Corresponding author: Prof. Xueyi Le

AC

No. 483, Wushan Road, Tianhe District, Guangzhou 510642, China. Tel: +86 020 85280319. Fax: +86 020 85280319 E-mail: [email protected]

ACCEPTED MANUSCRIPT Abstract: Catechin, an important component of flavan-3-ol, and dietary fiber are both important ingredients with many associated health benefits. The adsorption of catechin onto various dietary fiber has been studied widely, most of the researches focus on the

PT

adsorption capacities of catechin under different fibers and the adsorption types by

RI

using adsorption models. However, little is known on the dynamic adsorption process

SC

and mechanism, including the adsorption sites, interaction types, and participant molecules. In this study, the adsorption behavior and mechanism of catechin onto

NU

cellulose were examined by the time function in combination with molecular simulation.

MA

The adsorption capacities of cellulose for catechin were 2.70 and 2.82 mg/g at pH 2.0 and 7.0, respectively. The adsorption process was fitted by three stage models (rapid

D

adsorption, saturation, and equilibrium). The features of cellulose and catechin were

PT E

characterized by FTIR to identify the functional groups in the adsorption. Molecular simulation revealed that the catechin was adsorbed onto the hydrophilic surface of

CE

cellulose rather than hydrophobic one, and that the total binding energy was -8.57

AC

kcal/mol of the hydrophilic surface, which was due to Van der Waals' force and H-bond more than electrostatic force. Furthermore, the studies on isothermal adsorption combined with adsorption at various pH illustrated the main interaction between cellulose and catechin for the binding. This work assisted understanding of the adsorption of polyphenols on to insoluble dietary fiber and has the potential of applications in functional foods. Keywords: cellulose; catechin; adsorption process; adsorption mechanism; kinetic

ACCEPTED MANUSCRIPT models; molecular simulation 1. Introduction Dietary fiber is important for human health. It has a role in balancing gastrointestinal microbes (Habibi, Lucia & Rojas, 2010), lowering blood cholesterol

PT

and triglycerides (Lin et al., 2015), and reducing the risk of gastrointestinal diseases

RI

(Trompette et al., 2014). Dietary fiber is a class of polysaccharide, which cannot be

SC

digested and absorbed in gastrointestinal tract and does not produce energy because human body lacks β-1,4-glucosidase, which is needed to break down the cellulose and

NU

fiber.

MA

Dietary fiber has good water and oil adsorption ability and helps to maintain weight and lower blood triglycerides (Slavin, 2005). However, dietary fiber also binds

D

to some bioactive molecules, such as vitamin E and C, as well as some mineral elements

PT E

(Faure, Koppenol & Nyström, 2015; Macagnan, 2016; Sangnark & Noomhorm, 2003). Moreover, it also has an excellent adsorption capacity for polyphenols (Çelik &

CE

Gökmen, 2014; Saura-Calixto, 2010; Tomas et al., 2018). Polyphenols are functional

AC

components in food with many health benefits, but their bioavailability may be affected by the gastrointestinal components (Cervantes-Paz et al., 2017; Tuohy, Conterno, Gasperotti & Viola, 2012) and the presence of other food ingredients (Saura-Calixto, 2010). The influence of adsorption by fibers on polyphenols activity attracts much attention of researchers. The dietary fiber could adsorb polyphenols to improve their stability in the gastrointestinal tract (Bohn, 2014; Wu, Sanguansri & Augustin, 2014), which is beneficial to human health. However, on the other hand, the fibers are likely

ACCEPTED MANUSCRIPT to cause detrimental effects on polyphenol bioavailability and activity (Bandyopadhyay, Ghosh & Ghosh, 2012; Tamura, Iwami, Hirayama & Itoh, 2009), mainly due to factors such as physical entrapment (Bohn, 2014) and excretion with feces. Despite the growing interest of polyphenols and dietary fiber, most studies mainly

PT

focused on the adsorption capacity of polyphenols onto different types of fibers, and

RI

estimated the adsorption effect using the adsorption models. Polyphenols, such as gallic

SC

acid, catechin, and cyanidin-3-glucoside, bound to bacterial cellulose by alkali treatment with a spontaneously and rapidly process (Phan et al., 2015). The phenolic

NU

compounds, catechin, caffeic acid, and ferulic acid, have been examined to find out the

MA

possible interferences by the adsorption of cellulose and xylan. The results indicated that the amount of catechin adsorbed was affected by pH (Costa, Rogez & Pena, 2015)

D

and temperature (Wu, Melton, Sanguansri & Augustin, 2014). There were diverse

PT E

models to describe the isothermal adsorption, where the Langmuir and Freundlich models were the most commonly used to explain the adsorption of compounds and the

CE

adsorption mechanism (Shi et al., 2015). The adsorption is a process of mass transfer

AC

in which one or more compounds present in a fluid phase adhering to a solid surface (Geankoplis, 2003). For the food matrix system, it is a dynamic process containing the building-up and breaking-down of interaction between macromolecule (solid surface) and polyphenols (compounds). Currently, there is rare research about the interaction types and participant molecular dynamic behavior between dietary fiber and polyphenols, however, which is important to understand the digestion of dietary fiber by human body and building new functional food. So the study of the dynamic

ACCEPTED MANUSCRIPT adsorption process as a function of time and the molecular simulation of catechin onto the surface of fibers can contribute to the understanding of the interaction between the polyphenols and fibers. Cellulose is a major insoluble dietary fiber from cereal and vegetables. Catechin

PT

is the chosen polyphenol, which has the basic chemical unit structure of flavan-3-ol

RI

component, and could be found from teas as well as so many other fruits and their

SC

derivatives, such as apples, cacao, grapes, and wines (Ragone et al., 2015). In this study, the cellulose and catechin were selected as the model molecules to investigate the

NU

adsorption kinetics and mechanism. The functional group changes of participant

MA

molecules were examined by fourier transform infrared spectroscopy (FTIR) and adsorption of catechin on the cellulose surface was studied by fluorescence microscopy.

D

In addition, molecular simulation and docking were used to study the possible

PT E

adsorption sites and the types of interaction at various pH. This work would help to understand the interaction between polyphenols and insoluble dietary fiber for further

AC

CE

studies on the adsorption process in vivo and design of functional foods.

2. Materials and methods 2.1. Materials

(+)-Catechin (≥98%, HPLC, powder), cellulose (medium fibers), methyl-cellulose, and ethyl-cellulose were purchased from Sigma-Aldrich Chemical Co. (St. Louis, MO, USA). All other chemicals and reagents used were of analytical grade. Purified water was obtained by Milli-Q water purification system (Millipore, MA, USA).

ACCEPTED MANUSCRIPT 2.2. HPLC analysis The concentration of catechin in the aqueous solutions was determined by HPLC (Surveyor system with diode array detector and MS pump, Shimadzu, Kyoto, Japan) according to the reported method (Soetaredjo et al., 2013) with slight modification. The

PT

column used YMC C18 column (250 length×4.6 mm inner diameter, 5 μm particle

RI

dimension) and the UV-vis detector wavelength was set at 280 nm. HPLC conditions

SC

were as follows: mobile phase A: water (0.3% acetic acid); mobile phase B: acetonitrile; column temperature of 20 °C; injection volume of 10 μL. The elution program was set

NU

as isocratic elution by mobile phase A:B=6:4 at the flow rate of 0.4 mL/min. For

MA

quantification of catechin, peak areas were compared to a calibration curve (y=18097x217773, R2=0.99) by catechin standard solutions at 10, 25, 50, 100, 200, 400, and 600

D

mg/L (Figure S1).

PT E

2.3. Adsorption and kinetics study

Cellulose of 5.0 g was added to 100 mL catechin solution (200 mg/L) in two

CE

conical flasks at pH 2.0 (glycine-HCl buffer, 0.05 M) and pH 7.0 (PBS buffer, 0.2 M),

AC

respectively. The two conical flasks were sealed by plastic wrap and placed in a shaking water bath at 120 rpm at 37 °C, respectively. At specific time intervals (10,30, 60, 90, 120, 240, 360, 480, and 1440 min), aliquots (1.0 mL) were taken by syringes, filtered (0.22 μm), and subjected to the HPLC analysis. The amount of catechin adsorbed at time t, Qt (mg/g), was calculated with the following equation (Eq. 1): Qt=(C0-Ct)V/W

(1)

where C0 and Ct (mg/L) are the catechin concentrations of the liquid phase initially

ACCEPTED MANUSCRIPT and at time t, respectively. V (L) is the volume of the solution, and W (g) is the cellulose weight. The experiment was carried out twice. In addition, the pseudo-first-order (Eq. 2) and pseudo-second-order (Eq. 3) were used to fit the experimental data so as to evaluate the performance of the adsorbents. Weber-Moris interparticle diffusion models

PT

(Wu, Sanguansri & Augustin, 2015) (Eq. 4) were used to describe the mechanism of

RI

the adsorption process.

Qt = kit1/2+C

NU

t/Qt = 1/k2(Qe)2+t/Qe

SC

ln(Qe-Qt) = lnQe-k1t

(2) (3) (4)

MA

Where Qt (mg/g) were the amounts of catechin adsorbed at different times t (min) and Qe (mg/g) were the equilibrium adsorbed amount of catechin; k1 (1/min), k2 (g/mg

D

min), and ki (min1/2 mg/g) were the rate constants of the pseudo-first-order, pseudo-

2.4. FTIR analysis

PT E

second-order, and Weber-Morris interparticle diffusion models, respectively.

CE

Pure catechin and catechin-cellulose binding samples spectra were recorded in the

AC

range of 600~4000 cm−1 by FTIR (8400S, Shimadzu, Kyoto, Japan). The binding sample was prepared as follows: 0.01 g cellulose was added into 5 mL of 400 mg/L catechin solution with deionized water in a 10 mL centrifuge tube, which was put in a shaking water bath for 24 h at 120 rpm and 37 °C. The binding samples were dried by airing at the room temperature (~22 °C). And then 5.0 mg pure catechin or binding samples were placed on the silicon platform for scanning at a resolution of 4 cm−1 for 90 times to calculate the mean values.

ACCEPTED MANUSCRIPT 2.5. Fluorescence microscopy Pure catechin, cellulose, and catechin-cellulose binding samples were prepared so as to observe the adsorption effect of catechin onto cellulose using fluorescence microscopy (DM600-B, Leica Microsystems(Switzerland) Ltd., Sankt Gallen, Swiss).

PT

The images were captured and processed using Leica application suite V4.0.0 software.

RI

Visible and fluorescence lights were provided by 12V/100W halogen and 100W

SC

mercury lamps, respectively. The LP 425 suppression filter was used for the blue excitation range (340~380 nm), and the LP 515 suppression filter was used for the green

NU

excitation range (450~490 nm). The binding sample preparation was same as the

MA

adsorption study in this work. Briefly, 5.0 g cellulose was added to 100 mL catechin solution at the concentration of 200 mg/L at pH 7.0. After 24 h equilibration, the

D

cellulose was drawn onto the slides and the superfluous water was removed for

PT E

observation. Pure catechin and cellulose were selected as control. Excess powder samples were prepared on the glass slide and wetted with deionized water to retain their

CE

crystal structures for observation.

AC

2.6. Molecular simulation The model for cellulose was obtained from the X-ray crystal structure of cellulose Iβ (Nishiyama, Sugiyama, Chanzy & Langan, 2003), consisting of 18 glucan chains of 10 glycosylic residues linked via β-1,4 linkage in 7 layers (Mohanta, Madras & Patil, 2014) as shown in supplementary Figure S2. There were six surfaces, named (100), (110), (010), (200), (120), and (020), respectively; of these, (100) and (200) were the hydrophobic surfaces, and (010) and (020) were the comparatively hydrophilic surfaces

ACCEPTED MANUSCRIPT with a relatively more hydrogen bond acceptors and donors. The catechin molecule was optimized by Materials Studio 4.4 for the initial distorted molecular structure. The docking was carried out by the AutoDock 4.0 with MGL-tools 1.5.6. Briefly, the free water molecules were removed at first; then, all

PT

atoms, including cellulose and catechin, were distributed in grids by AutoGrid4 and a

RI

grid box was set up with a size of 60×60×60 Å with a grid spacing of 0.375 Å for the

SC

preparation of docking running. The docking calculations were performed with 100 times runs by Genetic Algorithm and the output were sorted by Lamarckian GA module.

NU

Other options were using the default parameters. The best conformation of complex

2.7. Isothermal adsorption study

MA

was selected through the minimum energy scoring from all docking results.

D

In order to understand the adsorption discrepancy depending on different

PT E

functional groups among cellulose, methyl-cellulose, and ethyl-cellulose, the adsorption isotherm test was carried out using a method with slight modification (Wu,

CE

Sanguansri & Augustin, 2015). Briefly, solutions of 5.0 mL of catechin with various

AC

concentrations of 25, 50, 100, 200, and 400 mg/L were prepared at pH 6.0 (PBS buffer, 0.2 M). They were placed into 10 mL centrifuge tube with 0.1 g cellulose, and then all samples were sealed with plastic wrap and shaken (120 rpm) in a water bath at 37 °C for 24 h. Finally, these tubes were centrifuged at 10000 rpm at 4 °C for 10 min, and 1 mL of supernatants were filtered (0.22 μm) for HPLC analysis. The methyl-cellulose and ethyl-cellulose were treated using the same method. The Langmuir model (Eq. 5) and Freundlich model (Eq. 6) were used to describe

ACCEPTED MANUSCRIPT the adsorption isotherm (Wu, Sanguansri & Augustin, 2015): Ce/Qe = 1/QmaxKL+Ce/Qmax

(5)

Where, Ce (mg/L) is the equilibrium concentration and Qe (mg/g) is the amount of adsorbed catechin per unit sample. The Langmuir parameters KL (L/mol) is related to

PT

the adsorption rate and Qmax (mg/g) is the adsorption capacity. (6)

RI

lnQe = 1/n·lnCe+lnKF

SC

The Freundlich constant (n) is related to the heterogeneity of the adsorption and

2.8. Adsorption study at different pH

NU

KF (mg/g(mL/mg)1/n) is the adsorption capacity of the adsorbent.

MA

The effect of pH on the adsorption of catechin was monitored in order to understand the electrostatic interaction between cellulose and catechin. Various pH

D

buffers were prepared, including pH 2.0 (glycine-HCl buffer, 0.05 M), pH 5.0-8.0 (PBS

PT E

buffer, 0.2 M), and pH 8.5-10.0 (glycine-NaOH buffer, 0.05 M). Solutions of 5 mL of catechin (400 mg/mL) was placed into 10 mL centrifuge tube with 0.1 g cellulose at

CE

different pH. Subsequently, all tubes were sealed with plastic wrap and shaken (120

AC

rpm) in a water bath at 37 °C for 24 h. All samples were centrifuged at 10000 rpm at 4 °C for 10 min and supernatants of 1.0 mL were filtered (0.22 μm) for HPLC analysis. 2.9. Statistical analysis All adsorption experiments were carried out in duplicate and the data were analyzed and fitted by linear and nonlinear regression models by Origin 9.0 (Origin Lab Inc., MA, USA). Statistics on a completely randomized design were determined using the one-way analysis of variance (ANOVA) procedure by SPSS 17.0 software, at a level

ACCEPTED MANUSCRIPT of significance set at p≤0.05.

3. Results and discussion 3.1. The adsorption of catechin onto cellulose

PT

The adsorption capacity of catechin onto cellulose was measured to investigate the

RI

adsorption behaviour as a time function at pH 2.0 and 7.0, which pH were used to

SC

correspond the physiological pH environment of gastrointestinal tract. As shown in Fig. 1a, the adsorption capacity increased with time. The adsorption process could be

NU

divided into two stages: it was fast in the first 300 min and, thereafter, slowed down as

MA

it approached equilibration at 24 h. According to adsorption theory, if the adsorption of natural chemical component onto adsorbents almost reaches equilibrium in a relatively

D

short time from several minutes to hours, this suggests that the process likely happens

PT E

via intermolecular forces (Pignatello & Xing, 1995). In the initial stage, there were adequate binding sites for catechin on the surface of cellulose, accounting for the fast

CE

adsorption rate. Subsequently, the rate was decreased as the equilibrium was

AC

approached because of the reduced number of sites for adsorption and the reduced concentration of catechin. At pH 7.0, the adsorption rate started to reduce slowly at the point of about 240 min, which was shorter than that of pH 2.0 at about 300 min. After 24 h, the adsorption capacity of cellulose reached 2.70 ± 0.34 mg/g (pH 2.0) and 2.82 ± 0.30 mg/g (pH 7.0), respectively, but without significant difference (p=0.54). The reason is because the pKa of catechin is 8.77 (Herrero Martínez, Sanmartin, Rosés, Bosch & Ràfols, 2005) and

ACCEPTED MANUSCRIPT hence at pH 2.0 and 7.0, catechin is only partially dissociated, implying the effect of ionization of catechin on adsorption capacity less than the adsorption rate. The faster adsorption rate at pH 7.0 (Fig. 1a) was also corroborated when either the pseudo-first-order and pseudo-second-order models were used to fit the

PT

experimental adsorption data (Fig. 1b and 1c). The kinetic parameters, correlation

RI

coefficient (R2), and relative error were summarized in Table 1. The fitting of the

calculated values (Qe,cal) when the

SC

models showed that the experimental adsorption capacity (Qe,exp) was closer to the pseudo-second-order kinetic model was used.

NU

There were similar findings from studies on the adsorption of (-)-epigallocatechin

MA

gallate onto apple pomace (Wu, Sanguansri & Augustin, 2015) or rice bran (Shi, Yang, Jin, Huang, Ye & Liang, 2015), indicating that adsorption of catechin onto cellulose

D

may have similar adsorption mechanisms, implying low selectivity for target adsorbates

PT E

and a physical adsorption process.

Then, Weber & Morris intra-particle diffusion models have been used to

CE

completely define the adsorption process. Generally, the adsorption of adsorbate

AC

molecules from a liquid phase onto a solid adsorbent involves three phases: film diffusion, intra-particle diffusion, and balance (Ding, Deng, Wu & Han, 2012). Based on this models, the adsorption process has been separated into two parts by linear fitting as shown in Fig. 1d. At the first stage, the adsorption capacity was increased rapidly, and then, the adsorption process turned to the next stage at about 225 min at pH 7.0. However, the turning point was delayed under the condition of pH 2.0, meaning that it takes a longer time to reach saturation with a slower adsorption rate at pH 2.0. But,

ACCEPTED MANUSCRIPT there is no significant difference between their adsorption capacities after the adsorption reached the equilibrium, implying that pH had much higher effect on adsorption rate than adsorption capacity. Furthermore, the fitted line did not pass through the origin, indicating that intra-particle diffusion was not the rate limiting factor during the process

PT

and the adsorption mainly occurred on the surface of cellulose crystals due to its non-

RI

microporous structures.

SC

Previous studies usually focused on the adsorption capacity of polyphenols by various cellulose subtypes (Costa, Rogez & Pena, 2015; Phan, Netzel, Wang, Flanagan,

NU

D Arcy & Gidley, 2015) and their adsorption isothermal kinetics under different

MA

environmental conditions (Shi, Yang, Jin, Huang, Ye & Liang, 2015), which was useful for studying the relationship between polyphenols and dietary fiber. Those study results

D

were often obtained with a prolonged adsorption until reaching saturation, however, it

PT E

was difficult to ensure that the adsorption was saturated in the actual food intake and food production process. So it was more important to study their adsorption status at

CE

different stages. The present study results suggested that adsorption of catechin onto

AC

cellulose surface took three stages: the adsorption was rapid at first stage, and then the adsorption rate decreased after adsorption saturation, and it reached equilibrium at last (Fig. 2).

3.2. Physicochemical characterization of catechin adsorption onto cellulose FTIR spectroscopy revealed the type of the molecular interactions between catechin and cellulose, which was also useful to explain the change in the environment

ACCEPTED MANUSCRIPT of functional groups. The FTIR spectra of pure catechin and catechin-cellulose are shown in Fig. 3. A broad absorption band in the range of 3000~3600 cm-1 was observed in both spectra due to the overlapping of hydroxyl group from the absorption of water. The peaks at 1519 and 1469 cm-1 were assigned to C=C aromatic ring and C-H alkanes

PT

of benzene ring, respectively (Harwansh et al., 2016). The wavelength of 1000~1400

RI

cm-1 was enlarged, showing that red shift has been found from 1279 to 1285 cm-1 (C-O

SC

alcohols), from 1140 to 1146 cm-1 (-OH aromatic), and from 1108 cm-1 to 1112 cm-1 (CO ethers) which is due to the bond length and bond angle change owing to hydrogen

NU

bonds formation between catechin and cellulose moieties, involving these groups. In

MA

addition, the peaks in the range of 600~1000 cm-1 were regarded as the benzene ring substitution position and fingerprint peaks, which were not changed because no

D

chemical reaction occurred. The FTIR results implied the changes in the environment

PT E

of functional groups induced by the binding of catechin and cellulose. The -OH on the benzene ring of catechin may be the main functional groups binding to the cellulose

CE

molecules by intermolecular interactions but no chemical reaction during the adsorption

AC

process really happened.

Fluorescence microscopy was used to observe the adsorption phenomenon (Fig. 4). The pure catechin displayed blue and green fluorescence (Fig. 4A-2 and 3) due to its natural characteristics. The cellulose showed a weak blue fluorescence (Fig. 4B-2) and no green fluorescence (Fig. 4B-3). For the cellulose-catechin binding samples, an obvious blue and green fluorescence were observed (Fig. 4C-2 and 3), especially, the green fluorescence only appeared on the cellulose surface. The results suggested that

ACCEPTED MANUSCRIPT the green fluorescence was due to catechin adsorbed onto the cellulose surface because of its dense and compact microstructure of cellulose. Similar results that the adsorption of some polyphenols molecules onto bacterial cellulose surface were proposed in previous literature reports, which implied that the binding of polyphenols to rigid

PT

cellulose was relatively reversible (Adt, Flanagan, D'Arcy & Gidley, 2017; Phan,

RI

Netzel, Wang, Flanagan, D Arcy & Gidley, 2015). These image observations support

NU

onto cellulose was a dynamic reversible process.

SC

the previous results of time function models in this study that the adsorption of catechin

MA

3.3. Molecular simulation of catechin docking to cellulose Molecular simulation was used to further illustrate the experiments of the

D

interaction between catechin and cellulose. The cellulose molecule, including six

PT E

surfaces, was built by repeated cellulose Iβ units obtained from the X-ray crystal structure. For the cellulose molecule structure which has β-1,4 linked glucose residues,

CE

the structure constructed with a large number of hydrogen bonds between glucose units

AC

or glucose chains inside the molecule form a very stable structure (Habibi, Lucia & Rojas, 2010). Therefore, there is rare chemical reaction between cellulose and other molecules at room temperature and natural pH, so the interaction between cellulose and polyphenols is mainly due to the intermolecular interaction. The catechin molecule was docked onto the cellulose hydrophobic surface (100) and hydrophilic surface (010), the main surface in the cellulose structure as shown in Fig. 5a. The catechin molecule was more suited to the gully of the hydrophilic surface. Furthermore, the total binding

ACCEPTED MANUSCRIPT energy of the hydrophilic surface was -8.57 kcal mol−1, which was much lower than that of the hydrophobic surface (-4.71 kcal mol−1). In addition, the H-bonds number on the hydrophilic and hydrophobic surface was six and four, respectively (Fig. 5b and c), indicating a more stable combined structure on the former case. Others studied the

PT

binding of xylans (Busse Wicher et al., 2014) and curcumin (Mohanta, Madras & Patil,

RI

2014), and suggested that the hydrophilic surface was the main binding sites on the

SC

cellulose surface, which was consistent with the results obtained from our study. From the results of bonding energy (Table 2), Van der Waals forces and H-bonds gave greater

NU

contribution to the binding compared to electrostatic interaction, which implying that

MA

the electrostatic interaction was not the main influencing factors for binding mechanism. The electrostatic interaction was affected by the changes from the surface potential of

D

cellulose and the ionization of catechin owing to pH values. So the bonding energy

PT E

results conduce to explain that the adsorption capacity of cellulose for catechin was similar at pH 2 and 7. Molecular simulation results help to understand the intuitional

CE

binding sites and the intermolecular forces types, furthermore, those also evidence the

AC

adsorption between cellulose and catechin belonged to physical adsorption process due to intermolecular forces without chemical bonds formation, which was consistent with the adsorption kinetics results in this study.

3.4. The chemical groups difference effect on adsorption Isothermal adsorption is generally used to study the relationship between adsorbent adsorption capacity (Qe) and adsorbate concentration in the liquid phase (Ce)

ACCEPTED MANUSCRIPT in dynamic equilibrium at fixed temperature. The isothermal adsorption of catechin onto methyl-cellulose and ethyl-cellulose was compared with catechin adsorption onto cellulose. The curves showed that the adsorption capacity increased with the length of the

PT

side chain (Fig. 6). Presumably, the altered binding sites for catechin contributed to

RI

different adsorption mechanisms. The Freundlich empirical formula was used to

SC

describe the adsorption process, and the Langmuir isotherm model was applied to explain the monolayer adsorption theory of catechin on the surface of the cellulose. The

NU

experimental data were fitted by these two isothermal adsorption models (Fig. 7a and

MA

7b). The results showed that the order of calculated maximum adsorption Qmax was ethyl-cellulose>methyl-cellulose>cellulose (Table 3), a trend which was also found in

D

the adsorption curve (Fig. 6). This was interpreted as more hydrogen bonding sites for

PT E

catechin to improve the adsorption capacity due to the increase in the number of hydrogen bond donors and acceptors from the methyl and ethyl groups. It is known that

CE

the binding of polyphenols with cellulose is mediated via hydrophobic interaction,

AC

hydrogen binding or a combination of these (Tang, Covington & Hancock, 2003). Catechin is hydrophilic in nature with water solubility of about 2.26 g/L at 298.75 K (Srinivas, King, Howard & Monrad, 2010) and log P value, octanol-water partition coefficient, of 0.86 (Yang, Kotani, Arai & KUSU, 2001). So it is capable of binding to biopolymers dominated by hydrogen bonds, such as chitosan (Hu et al., 2008), βcyclodextrin (Yan, Xiu, Li & Hao, 2007), and poly(lactic acid)-poly(hydroxybutyrate) (Arrieta et al., 2014).

ACCEPTED MANUSCRIPT KL and KF, corresponding to the parameters of Freundlich model and Langmuir model, respectively, were related to the adsorption rate and adsorption capacity. The parameter ‘‘1/n’’ was the Freundlich constant which was related to heterogeneity of the adsorption, ranging from 0.49 to 0.78 (Table 3), indicating that the adsorption was

PT

heterogenous (Wu, Sanguansri & Augustin, 2015). Overall, the results indicated that

RI

the adsorption of catechin onto cellulose could be described by these two models with

SC

high goodness of fit (R2=0.99), and also revealed the adsorption process was controlled by the initial concentration of catechin and the unoccupied adsorption sites on the

3.5. The pH effect on adsorption

MA

NU

surface of adsorbents (Benjamin & Leckie, 1981).

D

Generally speaking, some bioactive molecules, under a certain pH, would ionize

PT E

and lost some protons to present an ionic state. The pKa, the negative logarithm of ionization constant, of catechin was 8.77 (Herrero Martínez, Sanmartin, Rosés, Bosch

CE

& Ràfols, 2005). According to the definition, when the pH of a solution is equal to the

AC

pKa of a compound in solution, 50% of the molecules of the solute are in an ionic state and dissociated. When the pH is greater than pKa, the dissociation of -OH groups increases and the molecular form decreases (Srivastava, Swamy, Mall, Prasad & Mishra, 2006). Therefore, the catechin will rarely ionize in gastric juice (about pH=2.0) and partially ionize in intestinal fluid (about pH=7.6). In this work, the adsorption of catechin onto cellulose at pH=2.0 and 5.0-10.0 were determined to study its ionization effect on adsorption, and the chemical structure of catechin was considered stable under

ACCEPTED MANUSCRIPT the condition of those pHs (Herrero Martínez, Sanmartin, Rosés, Bosch & Ràfols, 2005). As shown in Fig. 8, the adsorption capacities showed regular fluctuations from pH 7.0 to pH 10.0 and got the biggest adsorption capacity at pH 9.0, indicating that catechin lost protons and had more negative sites at high pH. This means there were much more

PT

lone-pair electrons on the bare oxygen atom, making it easier to interact with cellulose

RI

surface, which may also be the reason for the increase of adsorption capacity from pH

SC

7.0 to pH 9.0. However, the adsorption of catechin was not significantly changed under the condition of pH 2.0 and pH 7.0 (p=0.827), as well as pH 2.0 and pH 7.5 (p=0.452),

NU

which may be due to an incomplete ionization of catechin at these pH (physiological

MA

pH environment of gastrointestinal tract), thus the contribution of electrostatic interaction for catechin adsorption was relatively less under these conditions. Phan et

D

al. (Phan, Netzel, Wang, Flanagan, D Arcy & Gidley, 2015) also reported that the native

PT E

charge of polyphenols was the secondary factor in the interactions between polyphenols

CE

and cellulose.

AC

4. Conclusion

This study investigated the adsorption behavior and adsorption process of catechin onto cellulose. Cellulose could adsorb catechin on the surface with three stages: the adsorption was rapid at first several hours, and then its rate decreased when the adsorption was saturated, at last, the adsorption reached equilibrium. In addition, the molecular simulation in combination with isothermal adsorption experiments and adsorption study at various pH, suggested that the adsorption may occur more on the

ACCEPTED MANUSCRIPT hydrophilic surface of cellulose than the hydrophobic one mainly by Van der Waals forces and H-bonds interactions type. Gathered findings can be useful to develop future application in designing functional foods. The binding of various phenolic compounds to other dietary fibers, and the release from fiber carriers in vitro are recommended for

PT

the further investigations to bring better understand of the interaction between

SC

RI

polyphenols and dietary fibers and their adsorption and release process.

Acknowledgements

NU

This work was supported by the Program of Guangdong Provincial Science &

MA

Technology (2017A020208038) and High Level University Construction Project of South China Agricultural University, China (No.215309). The authors would like to

D

acknowledge the support from South China University of Technology for molecular

PT E

simulation. Yujia Liu would also like to acknowledge the Graduate Student Overseas

AC

CE

Study Program from SCAU for the financial support to study in CSIRO.

ACCEPTED MANUSCRIPT

References: Adt, P., Flanagan, B. M., D'Arcy, B. R., & Gidley, M. J. (2017). Binding selectivity of dietary polyphenols to different plant cell wall components: Quantification and

PT

mechanism. Food Chemistry, 233, 216-227.

RI

Arrieta, M. P., Castro-López, M. D. M., Rayón, E., Barral-Losada, L. F., López-

SC

Vilariño, J. M., López, J., & González-Rodríguez, M. V. (2014). Plasticized poly (lactic acid)-poly (hydroxybutyrate)(PLA-PHB) blends incorporated with catechin

NU

intended for active food-packaging applications. Journal of Agricultural and Food

MA

Chemistry, 62, 10170-10180.

Bandyopadhyay, P., Ghosh, A. K., & Ghosh, C. (2012). Recent developments on

D

polyphenol-protein interactions: effects on tea and coffee taste, antioxidant

PT E

properties and the digestive system. Food & Function, 3, 592-605. Benjamin, M. M., & Leckie, J. O. (1981). Multiple-site adsorption of Cd, Cu, Zn, and

AC

209-221.

CE

Pb on amorphous iron oxyhydroxide. Journal of Colloid and Interface Science, 79,

Bohn, T. (2014). Dietary factors affecting polyphenol bioavailability. Nutrition Reviews, 72, 429-452. Busse Wicher, M., Gomes, T. C., Tryfona, T., Nikolovski, N., Stott, K., Grantham, N. J., Bolam, D. N., Skaf, M. S., & Dupree, P. (2014). The pattern of xylan acetylation suggests xylan may interact with cellulose microfibrils as a twofold helical screw in the secondary plant cell wall of Arabidopsis thaliana. The Plant Journal, 79, 492-

ACCEPTED MANUSCRIPT 506. Çelik, E. E., & Gökmen, V. (2014). Investigation of the interaction between soluble antioxidants in green tea and insoluble dietary fiber bound antioxidants. Food Research International, 63, 266-270.

PT

Cervantes-Paz, B., Ornelas-Paz, J. D. J., Ruiz-Cruz, S., Rios-Velasco, C., Ibarra-

RI

Junquera, V., Yahia, E. M., & Gardea-Béjar, A. A. (2017). Effects of pectin on lipid

SC

digestion and possible implications for carotenoid bioavailability during preabsorptive stages: A review. Food Research International, 99, 917-927.

NU

Costa, T. D. S., Rogez, H., & Pena, R. D. S. (2015). Adsorption capacity of phenolic

MA

compounds onto cellulose and xylan. Food Science and Technology (Campinas), 35, 314-320.

D

Ding, L., Deng, H., Wu, C., & Han, X. (2012). Affecting factors, equilibrium, kinetics

PT E

and thermodynamics of bromide removal from aqueous solutions by MIEX resin. Chemical Engineering Journal, 181-182, 360-370.

CE

Faure, A. M., Koppenol, W. H., & Nyström, L. (2015). Iron (II) binding by cereal beta-

AC

glucan. Carbohydrate Polymers, 115, 739-743. Geankoplis,

C.

J.

(2003).

Transport

processes

and

separation

process

principles:(includes unit operations): Prentice Hall Professional Technical Reference. Habibi, Y., Lucia, L. A., & Rojas, O. J. (2010). Cellulose nanocrystals: chemistry, selfassembly, and applications. Chemical Reviews, 110, 3479-3500. Harwansh, R. K., Mukherjee, P. K., Kar, A., Bahadur, S., Al-Dhabi, N. A., &

ACCEPTED MANUSCRIPT Duraipandiyan, V. (2016). Enhancement of photoprotection potential of catechin loaded nanoemulsion gel against UVA induced oxidative stress. Journal of Photochemistry and Photobiology B: Biology, 160, 318-329. Herrero Martínez, J. M., Sanmartin, M., Rosés, M., Bosch, E., & Ràfols, C. (2005).

PT

Determination of dissociation constants of flavonoids by capillary electrophoresis.

RI

Electrophoresis, 26, 1886-1895.

SC

Hu, B., Pan, C., Sun, Y., Hou, Z., Ye, H., Hu, B., & Zeng, X. (2008). Optimization of fabrication parameters to produce chitosan-tripolyphosphate nanoparticles for

NU

delivery of tea catechins. Journal of Agricultural and Food Chemistry, 56, 7451-

MA

7458.

Lin, L., Peng, C., Yeh, X., Huang, B., Wang, H., Chen, K., & Peng, R. Y. (2015).

D

Antihyperlipidemic bioactivity of Alpinia officinarum (Hance) Farw Zingiberaceae

PT E

can be attributed to the coexistance of curcumin, polyphenolics, dietary fibers and phytosterols. Food & Function, 6, 1600-1610.

CE

Macagnan, F. T. (2016). Dietary fibre: The scientific search for an ideal definition and

AC

methodology of analysis, and its physiological importance as a carrier of bioactive compounds. Food Research International, 85, 144-154. Mohanta, V., Madras, G., & Patil, S. (2014). Layer-by-layer assembled thin films and microcapsules of nanocrystalline cellulose for hydrophobic drug delivery. ACS Applied Materials & Interfaces, 6, 20093-20101. Nishiyama, Y., Sugiyama, J., Chanzy, H., & Langan, P. (2003). Crystal structure and hydrogen bonding system in cellulose Iα from synchrotron X-ray and neutron fiber

ACCEPTED MANUSCRIPT diffraction. Journal of the American Chemical Society, 125, 14300-14306. Phan, A. D. T., Netzel, G., Wang, D., Flanagan, B. M., D Arcy, B. R., & Gidley, M. J. (2015). Binding of dietary polyphenols to cellulose: Structural and nutritional aspects. Food Chemistry, 171, 388-396.

PT

Pignatello, J. J., & Xing, B. (1995). Mechanisms of slow sorption of organic chemicals

RI

to natural particles. Environmental Science & Technology, 30, 1-11.

SC

Ragone, R., Crupi, P., Piccinonna, S., Bergamini, C., Mazzone, F., Fanizzi, F. P., Schena, F. P., & Antonacci, D. (2015). Classification and chemometric study of

NU

southern italy monovarietal wines based on NMR and HPLC-DAD-MS. Food

MA

Science & Biotechnology, 24, 817-826.

Sangnark, A., & Noomhorm, A. (2003). Effect of particle sizes on in-vitro calcium and

PT E

International, 36, 91-96.

D

magnesium binding capacity of prepared dietary fibers. Food Research

Saura-Calixto, F. (2010). Dietary fiber as a carrier of dietary antioxidants: an essential

CE

physiological function. Journal of Agricultural and Food Chemistry, 59, 43-49.

AC

Shi, M., Yang, Y. P., Jin, J., Huang, L. Y., Ye, J. H., & Liang, Y. R. (2015). Using defatted rice bran as a bioadsorbent for carrying tea catechins. Journal of Food Science, 80, C2134-C2139. Slavin, J. L. (2005). Dietary fiber and body weight. Nutrition, 21, 411-418. Soetaredjo, F. E., Ismadji, S., Santoso, S. P., Ki, O. L., Kurniawan, A., & Ju, Y. (2013). Recovery of catechin and epicatechin from sago waste effluent: Study of kinetic and binary adsorption isotherm studies. Chemical Engineering Journal, 231, 406-413.

ACCEPTED MANUSCRIPT Srinivas, K., King, J. W., Howard, L. R., & Monrad, J. K. (2010). Solubility of gallic acid, catechin, and protocatechuic acid in subcritical water from (298.75 to 415.85) K. Journal of Chemical & Engineering Data, 55, 3101-3108. Srivastava, V. C., Swamy, M. M., Mall, I. D., Prasad, B., & Mishra, I. M. (2006).

PT

Adsorptive removal of phenol by bagasse fly ash and activated carbon: equilibrium,

RI

kinetics and thermodynamics. Colloids and Surfaces a: Physicochemical and

SC

Engineering Aspects, 272, 89-104.

Tamura, M., Iwami, T., Hirayama, K., & Itoh, K. (2009). High fiber diet supplemented

NU

with rice bran hemicellulose may reduce daidzein absorption in mice. Food Science

MA

and Technology Research, 15, 141-146.

Tang, H. R., Covington, A. D., & Hancock, R. A. (2003). Structure–activity

D

relationships in the hydrophobic interactions of polyphenols with cellulose and

PT E

collagen. Biopolymers, 70, 403-413. Tomas, M., Beekwilder, J., Hall, R. D., Simon, C. D., Sagdic, O., & Capanoglu, E.

CE

(2018). Effect of dietary fiber (inulin) addition on phenolics and in vitro

AC

bioaccessibility of tomato sauce. Food Research International, 106, 129-135. Trompette, A., Gollwitzer, E. S., Yadava, K., Sichelstiel, A. K., Sprenger, N., NgomBru, C., Blanchard, C., Junt, T., Nicod, L. P., & Harris, N. L. (2014). Gut microbiota metabolism of dietary fiber influences allergic airway disease and hematopoiesis. Nature Medicine, 20, 159-166. Tuohy, K. M., Conterno, L., Gasperotti, M., & Viola, R. (2012). Up-regulating the human intestinal microbiome using whole plant foods, polyphenols, and/or fiber.

ACCEPTED MANUSCRIPT Journal of Agricultural and Food Chemistry, 60, 8776. Wu, L., Melton, L. D., Sanguansri, L., & Augustin, M. A. (2014). The batch adsorption of the epigallocatechin gallate onto apple pomace. Food Chemistry, 160, 260-265. Wu, L., Sanguansri, L., & Augustin, M. A. (2014). Protection of epigallocatechin

PT

gallate against degradation during in vitro digestion using apple pomace as a carrier.

RI

Journal of Agricultural and Food Chemistry, 62, 12265-12270.

SC

Wu, L., Sanguansri, L., & Augustin, M. A. (2015). Processing treatments enhance the

of Food Engineering, 150, 75-81.

NU

adsorption characteristics of epigallocatechin-3-gallate onto apple pomace. Journal

MA

Yan, C., Xiu, Z., Li, X., & Hao, C. (2007). Molecular modeling study of β-cyclodextrin complexes with (+)-catechin and (-)-epicatechin. Journal of Molecular Graphics

D

and Modelling, 26, 420-428.

PT E

Yang, B., Kotani, A., Arai, K., & Kusu, F. (2001). Relationship of electrochemical oxidation of catechins on their antioxidant activity in microsomal lipid peroxidation.

AC

CE

Chemical and pharmaceutical bulletin, 49, 747-751.

ACCEPTED MANUSCRIPT Figure Captions Fig. 1. (a)Adsorption of catechin onto cellulose at pH 2.0 and pH 7.0 as a function of time at 37 °C. Fitting of data to a pseudo-first-order kinetic model (b), Pseudo-secondorder kinetic model (c) and the Weber & Morris intraparticle diffusion (d) plot for

PT

catechin adsorption onto cellulose.

RI

Fig. 2. Schematic diagram of the adsorption process of catechin onto cellulose.

SC

Fig. 3. FTIR spectra of pure catechin (a) and cellulose/catechin mixture (b). Fig. 4. Fluorescence micrographs of pure catechin (A), cellulose (B) and

NU

cellulose/catechin binding samples (C) under the visible light and fluorescence light

MA

(scale was 100 μm).

Fig. 5. Models of catechin interactions with cellulose. (a) The docking sites onto the

D

hydrophobic surface (left) and hydrophilic surface (right) of cellulose. The hydrogen

PT E

bond formation sites and bond length on the hydrophobic surface (b) and hydrophilic surface (c).

CE

Fig. 6. Adsorption of catechin onto cellulose, methyl-cellulose and ethyl-cellulose at

AC

37 °C. (Qe = adsorption capacity; Ce = equilibrium concentration of catechin). Fig. 7. Langmuir isotherms (a) and Freundlich isotherms (b) for catechin adsorption onto cellulose, methyl-cellulose and ethyl-cellulose. Fig. 8. The adsorption capacity of catechin onto cellulose at pH 2.0 & 5.0~10.0. Figures Figure 1

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

AC

CE

PT E

D

Figure 2

ACCEPTED MANUSCRIPT

NU

SC

RI

PT

Figure 3

AC

CE

PT E

D

MA

Figure 4

ACCEPTED MANUSCRIPT

MA

NU

SC

RI

PT

Figure 5

AC

CE

PT E

D

Figure 6

ACCEPTED MANUSCRIPT

NU

SC

RI

PT

Figure 7

AC

CE

PT E

D

MA

Figure 8

ACCEPTED MANUSCRIPT Tables

Table 1. Kinetic parameters in the catechin adsorption onto cellulose at pH 2.0 and 7.0. Qe,cal

Relative error

(mg/g)

(mg/g)

(%)

pH 2.0

0.0039

0.99

2.70

pH 7.0

0.0054

0.93

2.82

k2 (g/mg min)

R2

pH 2.0

0.0047

0.99

pH 7.0

0.0059

0.99

31.11

1.91

32.32

Qe,exp

Qe,cal

Relative error

(mg/g)

(mg/g)

(%)

2.70

2.81

4.36

2.82

2.93

4.20

SC

Pseudo-second-order model

1.86

RI

R2

PT

Qe,exp

k1 (1/min)

Pseudo-first-order model

MA

NU

Note: Qe,exp, experiment data; Qe,cal, calculated form the model; Relative error = 100×(|Qe,exp-Qe,cal|)/Qe,exp.

Table 2. Binding energies on hydrophobic and hydrophilic surfaces. Vdw & H-bond

kcal/mol

kcal/mol

D

Electrostatic

PT E

Adsorption surface

H-bond number

Total Binding energy kcal/mol

-0.05

-6.31

4

-4.71

Hydrophilic surface

-0.1

-10.26

6

-8.57

CE

Hydrophobic surface

AC

Table 3. Isotherm parameters for catechin onto cellulose/methyl-cellulose/ethylcellulose. Langmuir

Freundlich

Qmax (mg/g)

KL (L/mol)

1/n

KF (mg/g(mL/mg)1/n)

Cellulose

2.06

4014.99

0.49

5.04

Methyl-cellulose

8.93

1658.34

0.71

15.26

Ethyl-cellulose

13.76

1209.10

0.78

22.69

ACCEPTED MANUSCRIPT Highlights  Adsorption of catechin onto cellulose took place in three stages.  The adsorption was a physical and dynamic reversible/equilibrium process.

AC

CE

PT E

D

MA

NU

SC

RI

PT

 Catechin was adsorbed onto cellulose hydrophilic surface by intermolecular forces.