Colloids and Surfaces B: Biointerfaces 83 (2011) 122–127

Contents lists available at ScienceDirect

Colloids and Surfaces B: Biointerfaces journal homepage: www.elsevier.com/locate/colsurfb

Preferential adsorption of extracellular polymeric substances from bacteria on clay minerals and iron oxide Yuanyuan Cao a , Xing Wei a , Peng Cai a,b,∗ , Qiaoyun Huang a,b , Xinming Rong b , Wei Liang b a b

State Key Laboratory of Agricultural Microbiology, Huazhong Agricultural University, Wuhan 430070, China Key Laboratory of Subtropical Agricultural Resources and Environment, Ministry of Agriculture, Huazhong Agricultural University, Wuhan 430070, China

a r t i c l e

i n f o

Article history: Received 13 September 2010 Received in revised form 3 November 2010 Accepted 8 November 2010 Available online 17 November 2010 Keywords: EPS Bacteria Mineral Adsorption

a b s t r a c t The adsorption of extracellular polymeric substances (EPS) from Bacillus subtilis on montmorillonite, kaolinite and goethite was investigated as a function of pH and ionic strength using batch studies coupled with Fourier transform infrared (FTIR) spectroscopy. The adsorption isotherms of EPS on minerals conformed to the Langmuir equation. The amount of EPS-C and -N adsorbed followed the sequence of montmorillonite > goethite > kaolinite. However, EPS-P adsorption was in the order of goethite > montmorillonite > kaolinite. A marked decrease in the mass fraction of EPS adsorption on minerals was observed with the increase of final pH from 3.1 to 8.3. Calcium ion was more efficient than sodium ion in promoting EPS adsorption on minerals. At various pH values and ionic strength, the mass fraction of EPS-N was higher than those of EPS-C and -P on montmorillonite and kaolinite, while the mass fraction of EPS-P was the highest on goethite. These results suggest that proteinaceous constituents were adsorbed preferentially on montmorillonite and kaolinite, and phosphorylated macromolecules were absorbed preferentially on goethite. Adsorption of EPS on clay minerals resulted in obvious shifts of infrared absorption bands of adsorbed water molecules, showing the importance of hydrogen bonding in EPS adsorption. The highest K values in equilibrium adsorption and FTIR are consistent with ligand exchange of EPS phosphate groups for goethite surface. The information obtained is of fundamental significance for understanding interfacial reactions between microorganisms and minerals. © 2010 Elsevier B.V. All rights reserved.

1. Introduction Growth of a wide variety of Gram-negative and Gram-positive bacteria is accompanied by the production of extracellular polymeric substances (EPS). EPS are bound to the cell surface (“capsular”), released into solution (“free”) or associated with the hydrated matrix of biofilms [1]. The interactions of EPS with inorganic colloids have significant effects on biofilm formation, the migration of bacterial cells, mineral dissolution, biomineralization and heavy metals accumulation in soil or aquatic environments [2–5]. EPS are a heterogeneous mixture composed dominantly of polysaccharides and proteins, with nucleic acids and lipids as minor constituents. EPS contain various weakly acidic functionalities (carboxyl, phosphoryl, amide, amino, hydroxyl) that ionize in response to changes in environmental pH or ionic strength [6]. The mechanisms of bacterial EPS adsorption on mineral surfaces which

∗ Corresponding author at: State Key Laboratory of Agricultural Microbiology, Huazhong Agricultural University, Shizi Mountain, Hongshan district, Wuhan 430070, China. Tel.: +86 27 87671033; fax: +86 27 87280670. E-mail address: [email protected] (P. Cai). 0927-7765/$ – see front matter © 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.colsurfb.2010.11.018

include hydrophobic, electrostatic, covalent, and polymer–polymer interactions have been studied by several environmental chemists over the past years [7,8]. Adsorption of EPS on goethite increased with the decrease of pH from 9.0 to 3.0 and NaCl concentrations from 100 to 1 mM [9]. Using attenuated total reflectance (ATR) Fourier transform infrared (FTIR) spectroscopy and quantum chemical calculations, Omoike et al. found that EPS from Pseudomonas aeruginosa and Bacillus subtilis binds to Fe centers on goethite via inner-sphere complexation of phosphate-containing macromolecules [10]. The atomic force microscope (AFM) and electronic structure calculations suggested the phosphate-bearing polymers are major components in EPS responsible for the adhesion strength with silica surfaces at low pH and H-bonds and electrostatic interactions are the dominant forces [11]. Quartz crystal microbalance with dissipation (QCM-D) revealed that deposition efficiencies of EPS from four bacterial strains on bare silica surfaces increased with increasing ionic strength in both monovalent and divalent solutions [12]. Although some attempts have been made to elucidate the mechanisms of EPS adsorption on minerals, the interactions between minerals and EPS remain poorly resolved. To our knowledge, comparison of EPS adsorption on various clay minerals and iron oxide has not been reported. Montmorillonite, kaolinite and goethite are

Y. Cao et al. / Colloids and Surfaces B: Biointerfaces 83 (2011) 122–127

123

Table 1 Langmuir parameters for the adsorption of EPS-C, -N and -P on montmorillonite, kaolinite and goethite. Minerals

EPS-C Xm (mg g

Montmorillonite Kaolinite Goethite

EPS-N −1

)

67.62 6.80 11.62

K (L mg

−1

)

0.0004 0.0084 0.1696

EPS-P

R

Xm (mg g

0.99 0.96 0.99

42.54 2.48 4.30

−1

)

K (L mg

−1

)

0.0014 0.0512 0.1366

R

Xm (mg g−1 )

K (L mg−1 )

R

0.98 0.98 0.99

0.39 0.36 1.88

0.2593 0.1308 1.1643

0.94 0.94 0.99

Aliquots of EPS stock solution were also prepared in 10 mmol L−1 NaCl at pH 7.0. Four milliliter (20 mg) of mineral suspension was mixed with EPS solution with final concentrations ranging from 0 to 1 mg mL−1 . The mixture was gently shaken at 25 ◦ C for 2 h and centrifuged at 20,000 × g for 30 min. The total organic C and total N concentrations of EPS in the supernatant were determined using a TOC/TN analyzer (multi N/C 3100, Analytik Jena, Germany). Total content of EPS-P in the supernatant was digested with potassium peroxydisulfate (K2 S2 O8 ) and measured by molybdenum blue spectrophotometric method (GB 11893-89). The amount of EPSC, -N and -P adsorbed was calculated by the difference between the amount of EPS added and that remaining in the supernatant. Adsorption was also conducted in the range of pH from 3.0 to 9.0 in which 20 mg of mineral and 4.0 mg of EPS were employed. The similar experiments were carried out in the presence of 0–100 mM of Na+ or 0–50 mM of Ca2+ .

common and important clay minerals and oxide in soil and sediments. In the present work, adsorption of EPS from B. subtilis on montmorillonite, kaolinite and goethite was investigated using both macroscopic and spectroscopic techniques. The study aimed to investigate preferential adsorption of EPS constituents on different minerals. A further objective was to better understand the nature of interactions between EPS and environment media. 2. Materials and methods 2.1. Bacterial EPS extraction and purification All solutions and suspensions were prepared using ultrapure (Milli-Q) water (18.24 M cm). B. subtilis was cultivated aerobically in Luria broth at 30 ◦ C and 150 rpm to early stationary (24 h) growth phase. The cells were harvested by centrifugation (5000 × g, 15 min, 4 ◦ C) and EPS were isolated from the supernatant solution as described by Omoike and Chorover [9]. Briefly, the supernatant solution was centrifuged at higher force (12,000 × g, 30 min, 4 ◦ C) to remove residual cells. EPS was precipitated from the supernatant solution by adding cold reagent-grade ethanol at a volumetric ratio of 3:1, and the mixture was then stored at 4 ◦ C for 48 h. The precipitate was separated from the ethanol suspension by centrifugation. The pellet obtained after centrifugation was dialysed against MilliQ water using cellulose membranes (3500 MWCO from Spectrum) to remove low molecular weight impurities including ethanol. After dialysis for 72 h against two changes of Milli-Q water per day, the EPS solution was freeze-dried on a vacuum freeze-drier (FreeZone6, Labconco Corporation, U.S.A.).

2.5. FTIR spectroscopy FTIR spectra were obtained on a spectrometer (IFS 66 v/s, Bruker, Karlsruhe, Germany) equipped with a MCT-MIR liquid nitrogen-cooled detector and OPUS 5.5 processing software. For each interferogram 256 scans were taken in the 400–4000 cm−1 with a resolution of 2 cm−1 . The KBr pressed disc technique were used by mixing EPS, minerals or minerals–EPS complexes with KBr powder (around 1:100) and using a press at the pressure of 10 tonnes. 3. Results and discussion

2.2. Minerals

3.1. Equilibrium adsorption of EPS-C, -N and -P on minerals

The preparation of goethite, less than 2 ␮m fractions from kaolinite and montmorillonite, as well as their properties have been described elsewhere [13].

The adsorption isotherms of EPS-C, -N, and -P on minerals at pH 7.0 are shown in Fig. 1. The batch data of EPS-C, -N, -P adsorbed by minerals were fitted to the Langmuir equation which can be described as X = Xm KC/(1 + KC), where X is the amount of EPS-C, -N, -P adsorbed per unit mass of minerals, Xm is the maximum amount of EPS-C, -N, -P that may be adsorbed, K is an empirical affinity constant related to the adsorption energy and C is the equilibrium concentration of EPS-C, -N, -P in aqueous phase. The greater the K value, the higher the affinity between EPS and minerals. As presented in Table 1, the amount of EPS-C adsorption on montmorillonite was 4.8 and 8.9-fold greater than that on goethite and kaolinite, respectively. EPS-N adsorption on montmorillonite was also far greater than that on goethite (8.9-fold) and kaolinite (16.2fold). However, EPS-P adsorbed by goethite was about 5-times greater than that by montmorillonite and kaolinite. Extracted EPS from B. subtilis are a complex mixture of biomacromolecules consisting primarily of polysaccharides (437.59 ± 11.76 mg g−1 ) and

2.3. Biochemical characterization of EPS Polysaccharides were assayed by the phenol–sulfuric acid method [14]. Proteins were measured by bicinchoninic acid (BCA) (Boisynthesis Co., Ltd., Beijing). Nucleic acid was determined with a U-0080D photodiode array spectrophotometer (Hitachi HighTechnologies, Tokyo, Japan). 2.4. Adsorption experiments Equilibrium adsorption was conducted in 10 mmol L−1 NaCl at pH 7.0. Stock mineral suspensions in 10 mmol L−1 NaCl were preequilibrated overnight at pH 7.0 using 0.01 mol L−1 HCl or NaOH. Table 2 Characteristics of EPS extracted from Bacillus subtilis. EPS (mg g−1 )

TOC

TN

TP

PS

PT

NA

316.68 ± 6.89

25.73 ± 0.48

21.52 ± 0.31

437.59 ± 11.76

249.11 ± 3.33

9.87 ± 0.23

Note: TOC, TN, TP, PS, PT and NA refer to total organic carbon, total nitrogen, total phosphorus, polysaccharides, proteins, nucleic acids, respectively. The data are presented as the means ± the standard error of the means (x ± SEM).

124

Y. Cao et al. / Colloids and Surfaces B: Biointerfaces 83 (2011) 122–127

16

(a) Montmorillonite

EPS-C

Kaolinite

14

EPS-N

Goethite

12

Fraction EPS Adsorbed

-1 Adsorbed EPS-C (mg g )

1.0

Montmorillonite

10 8 6 4 2

0.8

EPS-P

0.6

0.4

0.2

0 0

50

100

150

200

250

0.0

300

3

Equilibrium [EPS-C] (mg L-1 )

5

6

7

1.0

Montmorillonite

(b) Kaolinite

EPS-N

Goethite

2.0 1.5 1.0 0.5 0.0

0.8

EPS-P

0.6

0.4

0.2

0.0 0

2

4

6

8

10

12

14

16

18

20

3

4

5

Equilibrium [EPS-N] (mg L-1) 2.0

6

7

1.0

Montmorillonite

Fraction EPS Adsorbed

1.0

0.5

0.0 0

2

4

6

8

10

12

14

16

9

(c) Goethite

EPS-C EPS-N

Goethite

1.5

8

Final pH

Kaolinite

Adsorbed EPS-P (mg g-1)

9

EPS-C

Kaolinite

2.5

8

Final pH

Fraction EPS Adsorbed

-1 Adsorbed EPS-N (mg g )

3.0

4

18

20

22

Equilibrium [EPS-P] (mg L-1) Fig. 1. Adsorption isotherms of EPS-C, -N, -P on montmorillonite, kaolinite and goethite.

proteins (249.11 ± 3.33 mg g−1 ), with smaller amounts of nucleic acids (9.87 ± 0.23 mg g−1 ) (Table 2). Therefore, these results suggest that EPS-C and -N constituents (from proteins) are mainly adsorbed by montmorillonite and EPS-P constituents (from nucleic acids) are predominantly adsorbed by goethite. The K values for the adsorption of EPS-C, -N and -P on the three minerals followed the sequence of goethite > kaolinite > montmorillonite, indicating that goethite shows the highest affinity to EPS among the examined clay minerals and iron oxide. The higher binding affinity of goethite for EPS may be ascribed to the ligand exchange of EPS phosphate groups for surface hydroxyls at Fe metal centers [9]. This explanation was consistent with the calculated K value of goethite for EPS-P which is several orders higher than those in other systems.

0.8

EPS-P

0.6

0.4

0.2

0.0 3

4

5

6

7

8

9

Final pH Fig. 2. The mass fraction of EPS-C, -N and -P adsorbed by (a) montmorillonite, (b) kaolinite and (c) goethite as a function of pH.

3.2. Effects of pH The mass fraction of EPS-C, -N and -P adsorption on montmorillonite, kaolinite and goethite is plotted as a function of pH in Fig. 2. The mass fraction of EPS-N and -P adsorption on clay minerals and iron oxide decreased sharply with the increase of final pH from 3.1 to 8.3, while the mass fraction of EPS-C showed less pH dependence. EPS functional groups are mostly protonated at pH < 2.0, but become progressively negative-charged with increasing pH due to proton dissociation of carboxyl (pH 2.0–6.0), phospholipids (pH 2.4–7.2), phosphodiester (pH 3.2–3.5), hydroxyl (pH 9.0–10.0) and amino (pH 9.0–11.0) groups [15]. The values of PZC (point of zero charge) are 2.5, 3.6 and 8.3 for montmorillonite, kaolinte and goethite, respectively [13]. Zeta potential measurement of miner-

Y. Cao et al. / Colloids and Surfaces B: Biointerfaces 83 (2011) 122–127

1.0

(a) Montmorillonite

EPS-C

EPS-C

1.0

(a) Montmorillonite

EPS-N EPS-P

EPS-P

Fraction EPS Adsorbed

Fraction EPS Adsorbed

EPS-N 0.8

125

0.6

0.4

0.2

0.8

0.6

0.4

0.2

0.0 0

20

40

60

80

100

NaCl (mM) 1.0

0

(b) Kaolinite

10

20

30

EPS-N

EPS-C

EPS-P

1.0

EPS-N

(b) Kaolinite

EPS-P 0.6

Fraction EPS Adsorbed

Fraction EPS Adsorbed

50

CaCl2 (mM)

EPS-C

0.8

40

0.4

0.2

0.0 0

20

40

60

80

0.8

0.6

0.4

0.2

100

NaCl (mM)

0.0 EPS-C

1.0

(c) Goethite

0

10

EPS-N

20

30

40

50

CaCl2 (mM) EPS-C

1.0

(c) Goethite

EPS-N EPS-P

0.6

Fraction EPS Adsorbed

Fraction EPS Adsorbed

EPS-P 0.8

0.4

0.2

0.0 0

20

40

60

80

100

0.8

0.6

0.4

0.2

NaCl (mM) 0.0 Fig. 3. The mass fraction of EPS-C, -N and -P adsorbed by (a) montmorillonite, (b) kaolinite and (c) goethite as a function of NaCl concentrations.

als showed the zeta potentials at higher pH were more negative than those in lower pH [16]. The dramatic decrease of EPS-N and -P adsorbed by minerals with increasing pH was due to the protonation of EPS groups (hydroxyl and amino) which occurred at a low pH of 3.1, and these positively charged groups reacted as cations with negatively charged groups of the clay interfaces. Whereas at high pH 8.3, EPS molecules exposed negatively charged groups such as carboxyl and phosphate, which enhanced the repulsion between EPS and negatively charged minerals. These results also suggest that electrostatic interaction played an important role in EPS adsorption on clay minerals and iron oxide. The mass fraction values of EPS-N were higher than those of EPS-C and -P for montmorillonite and kaolinite, however, the mass fraction values of EPS-P exceeded those of both EPS-C and -N for goethite. These results

0

10

20

30

40

50

CaCl2 (mM) Fig. 4. The mass fraction of EPS-C, -N and -P adsorbed by (a) montmorillonite, (b) kaolinite and (c) goethite as a function of CaCl2 concentrations.

indicate a strong preference for EPS-N adsorption on montmorillonite and kaolinite, while EPS-P exhibited preferential adsorption on goethite. 3.3. Effects of ionic strength Figs. 3 and 4 show the adsorption of EPS-C, -N and -P on montmorillonite, kaolinite and goethite increased in the presence of NaCl and CaCl2 at pH 7.0. The promotive effect of Na+ and Ca2+ on EPS adsorption may be attributed to the following reasons: (1) cations tend to suppress the formation of diffuse electrical double layers

c 1800

1600

1400

1200

1000

1800

Wavenumber (cm-1)

1600

1400

1051 1031 975

1200

1126 1097 1049

1242 1217

c

1414

1457

1540

1409

1549

1657

1642

1782

Absorbance

997

1100

1636

Absorbance

b

1650 1637

975

1045

1089

1242

1409

1549

1657

a

a

b

1131

1051 1031

Y. Cao et al. / Colloids and Surfaces B: Biointerfaces 83 (2011) 122–127

1131

126

1000

Wavenumber (cm-1)

Fig. 5. FTIR spectra of (a) free EPS, (b) montmorillonite and (c) montmorillonite–EPS complexes after reactions.

Fig. 7. FTIR spectra of (a) free EPS, (b) goethite and (c) goethite–EPS complexes after reactions.

around clay particles, leading to increased contact between EPS and mineral surfaces [17]; (2) cations form bridges between functional groups of the EPS molecules and the negatively charged sites of clays. At the same ionic strength, the mass fraction values for EPS in CaCl2 solutions were greater than those in NaCl solutions, which may be ascribed to the less negative zeta potentials of EPS in CaCl2 solutions relative to those in NaCl solutions under the same conditions [12]. At different ionic strengths, the mass fraction values of EPS-N were higher than those of EPS-C and -P for montmorillonite and kaolinite, however, the mass fraction values of EPS-P were higher than those of EPS-C and -N for goethite. It also indicated preferential adsorption of EPS-N on montmorillonite and kaolinite and EPS-P on goethite.

phorylated proteins; 1131 cm−1 , of O–H deformation, or C–O ring vibrations of polysaccharides; 1051 cm−1 , of P O of phosphodiester backbone of nucleic acids, or C–OH stretch of phosphorylated proteins; 1031 cm−1 , of P–O symmetrical stretching; 975 cm−1 , of asymmetric ester O–P–O stretching from nucleic acids. For montmorillonite and kaolinite, the absorption bands at 1636 and 1634 cm−1 were assigned to bending vibrations of adsorbed water. The strong bands at 992, 1045, 1089 and 1124 cm−1 were attributed to Si–O stretching vibrations. No new absorption bands were found on EPS–montmorillonite or kaolinite complexes compared with those on clay minerals. However, the vibrations of water molecules adsorbed on montmorillonite and kaolinite at 1636 and 1634 cm−1 shifted to 1642 and 1647 cm−1 , respectively. These shifts suggested that the water molecules on the clay minerals are involved in EPS adsorption on clay minerals. Additionally, the Si–O stretching vibrations of montmorillonite at 1089 and 1045 cm−1 shifted to 1100 and 997 cm−1 , respectively, on EPS–montmorillonite complexes. Kaolinite–EPS complexes also produced characteristic alternations of the kaolinite spectra (the Si–O stretching at 1124 and 992 cm−1 shifted to 1115 and 1005 cm−1 and the Al–OH deformation shifted from 906 to 913 cm−1 ). These results indicated that hydrogen bonding may be responsible to EPS binding on clay minerals [18]. The FTIR spectrum of unreacted goethite shows two high frequency bands (1782 and 1637 cm−1 ) corresponding to overtones of OH vibrations. EPS bound on goethite resulted in some significant changes in the FTIR spectra compared to free EPS. The native EPS spectrum of amide II at 1549 cm−1 shifted to 1540 cm−1 on EPS bound on goethite, indicating that the proteins of EPS are involved in EPS adsorption on goethite. Furthermore, the adsorption bands of the stretching vibration of PO2 − of free EPS at 1242, 1131, 1051, 1031 cm−1 shifted to 1217, 1126, 1097, 1049 cm−1 , respectively, for EPS bound on goethite. This result suggested that electronic density of the phosphorous atom is weakened upon complexation and inner-sphere bonding of EPS phosphate groups (deriving principally from phosphodiesters of nucleic acids) at goethite surface hydroxyls [19].

3.4. Fourier transform infrared spectra

1031

1051

975

c 1800

913

1005

1647

b

1634

1115

Absorbance

a

906

992

1124

1242

1409

1549

1657

1131

FTIR spectra of EPS, minerals and their complexes are shown in Figs. 5–7. The main absorption bands of EPS are at: 1657 cm−1 , corresponding to C O of amides associated with proteins (amide I); 1549 cm−1 , of N–H and C–N in CO–NH– of proteins (amide II); 1409 cm−1 , of C–O of COO− groups; 1242 cm−1 , of P O of phosphodiester backbone of nucleic acids (DNA and RNA), or phos-

1600

4. Conclusions 1400

1200

1000

Wavenumber (cm-1) Fig. 6. FTIR spectra of (a) free EPS, (b) kaolinite and (c) kaolinite–EPS complexes after reactions.

EPS-N moieties mainly from proteins are adsorbed preferentially on clay minerals and EPS-P moieties predominantly from nucleic acids are adsorbed preferentially on goethite. An increase in the concentration of cations and/or a decrease in the pH favored

Y. Cao et al. / Colloids and Surfaces B: Biointerfaces 83 (2011) 122–127

EPS adsorption on minerals. Hydrogen bonding and the electrostatic interaction are the main forces governing the adsorption of EPS on clay minerals. Besides these interaction forces, chemical bonding interactions (ligand exchange) also contribute to selective fractionation of EPS adsorption on goethite. Acknowledgments The authors are grateful to the National Natural Science Foundation of China (40801095), Doctoral Fund of Ministry of Education of China (200805041060), and a Foundation for the Author of National Excellent Doctoral Dissertation of China. We also thank the National Synchrotron Radiation Laboratory (NSRL) of China for technical assistance with the FTIR. References [1] J. Wingender, T.R. Neu, H.C. Flemming (Eds.), Microbial Extracellular Polymeric Substances, Springer-Verlag, Heidelberg, 1999, p. 258. [2] J.F. Banfield, W.W. Barker, S.A. Welch, A. Taunton, Proc. Natl. Acad. Sci. U.S.A. 96 (1999) 3404.

127

[3] Y.M. Nelson, L.W. Lion, M.L. Shuler, W.C. Ghiorse, Environ. Sci. Technol. 30 (1996) 2027. [4] C.S. Chan, G.D. Stasio, S.A. Welch, M. Girasole, B.H. Frazer, M.V. Nesterova, S. Fakra, J.F. Banfield, Science 303 (2004) 1656. [5] L.C. Fang, Q.Y. Huang, X. Wei, W. Liang, X.M. Rong, W.L. Chen, P. Cai, Bioresour. Technol. 101 (2010) 5774. [6] A. Omoike, J. Chorover, Biomacromolecules 5 (2004) 1219. [7] H.C. van der Mei, S. Meijer, H.J. Busscher, J. Colloid Interface Sci. 205 (1998) 185. [8] S. Tsuneda, H. Aikawa, H. Hayashi, A. Yuasa, A. Hirata, FEMS Microbiol. Lett. 223 (2003) 287. [9] A. Omoike, J. Chorover, Geochim. Cosmochim. Acta 70 (2006) 827. [10] A. Omoike, J. Chorover, K.D. Kwon, J.D. Kubicki, Langmuir 20 (2004) 11108. [11] K.D. Kwon, V. Vadillo-Rodriguez, B.E. Logan, J.D. Kubicki, Geochim. Cosmochim. Acta 70 (2006) 3803. [12] P.T. Zhu, G.Y. Long, J.R. Ni, M.P. Tong, Environ. Sci. Technol. 43 (2009) 5699. [13] P. Cai, Q.Y. Huang, X.W. Zhang, Appl. Clay Sci. 32 (2006) 147. [14] M. Dubois, K.A. Gilles, J.K. Hamilton, P.A. Rebers, F. Smith, Anal. Chem. 28 (1956) 350. [15] R.E. Martinez, D.S. Smith, E. Kulczycki, F.G. Ferris, J. Colloid Interface Sci. 253 (2002) 130. [16] X.M. Rong, Q.Y. Huang, X.M. He, H. Chen, P. Cai, W. Liang, Colloids Surf. B: Biointerfaces 64 (2008) 49. [17] L.A.G. Aylmore, J.P. Quirk, Nature 187 (1960) 1046. [18] Y. Mao, L.N. Daniel, N. Whittaer, U. Saffiottil, Environ. Health Perspect. 102 (1994) 165. [19] B.C. Barja, M.I. Tejedor-Tejedor, M.A. Anderson, Langmuir 15 (1999) 2316.