Chemosphere 84 (2011) 191–198

Contents lists available at ScienceDirect

Chemosphere

journal homepage: www.elsevier .com/locate /chemosphere

Chitosan/albumin/CaCO3 as mimics for membrane bioreactor fouling: Genesis ofstructural mineralized-EPS-building blocks and cake layer compressibility

Miguel Herrera-Robledo a,⇑, Concepción Arenas b, Juan M. Morgan-Sagastume a, Víctor Castaño b,Adalberto Noyola a

a Grupo de Investigación en Procesos Anaerobios, Instituto de Ingeniería, Universidad Nacional Autónoma de México, Apartado Postal 70-472, 04510, Ciudad Universitaria, Coyoacán,México, DF, Mexicob Departamento de Ingeniería Molecular de Materiales, Centro de Física Aplicada y Tecnología Avanzada, Universidad Nacional Autónoma de México Campus Juriquilla,Apartado Postal 73-230, Boulevard Juriquilla 3001, Juriquilla, Qro, Mexico

a r t i c l e i n f o a b s t r a c t

Article history:Received 17 February 2011Received in revised form 20 April 2011Accepted 21 April 2011Available online 19 May 2011

Keywords:Membrane bioreactorsBiofoulingExtracellular polymeric substances (EPS)Mineralized building blocksBiomineralization

0045-6535/$ - see front matter � 2011 Elsevier Ltd. Adoi:10.1016/j.chemosphere.2011.04.059

⇑ Corresponding author. Tel.: +52 55 5623 3600; faE-mail address: mherrerar@iingen.unam.mx (M. H

Membrane bioreactor biofouling is usually described as an extracellular matrix in which biopolymers,inorganic salts and active microbes co-exist. For that reason, biomineralization (BM) models can be usefulto describe the spatial organization and environmental constraints within the referred scenario. BM argu-ments were utilized as background in order to (1) evaluate CaCO3 influence on flux decline; pore blockingand cake layer properties (resistance, permeability and compressibility) in a wide range of Chitosan/Bovine serum albumin (BSA) mixtures during step-pressure runs and, (2) perform membrane autopsiesin order to explore the genesis of mineralized extracellular building blocks (MEBB) during cake layerbuild up. Using low molecular weight chitosan (LC) and BSA, 2 L of 5 LC/BSA mixtures (0.25–1.85 ratio)were pumped to an external ultra filtration (UF) membrane (23.5 cm2, hydrophobic, piezoelectric,100 kDa as molecular weight cut-off). Eight different pressure steps (40 ± 7 to 540 ± 21 kPa) were applied.Each pressure step was held for 900 s. CaCO3 was added to LC/BSA mixtures at 0.5, 1.5 and 3 mM in orderto create MEBB during the filtration tests. Membrane autopsies were performed after the filtration testsusing thermo gravimetric, scanning microscopy and specific membrane mass (mg cm�2) analyses. Bio-polymer-CaCO3 step-pressure filtration created compressible cake layers (with inner voids). The forma-tion of an internal skeleton of MEBB may contribute to irreversible fouling consolidation. A hypothesis forMEBB genesis and development was set forth.

� 2011 Elsevier Ltd. All rights reserved.

1. Introduction

Anaerobic Membrane BioReactors (AnMBR: the combinationbetween high-rate anaerobic digesters and membrane separation)are getting more acceptance in the rapid increasing market of mu-nicipal wastewater reclamation. Although their fouling problemscan be partially controlled using either on-line cake layer evalua-tion (Jeison and van Lier, 2006) or membrane chemical modifica-tion (Asatekin et al., 2006; Drews, 2009; Meng et al., 2009)fundamental knowledge (i.e. microscale interactions, biofoulingarchitecture) dealing with the development of cake layer duringAnMBR performance is nowadays a current matter of research.

Recent studies focus on macro molecular analysis and AnMBRoperation, pointing out the central role of polysaccharides and sol-uble proteins during the cake layer build up (Rosenberger et al.,2006; Arabi and Nakhla, 2008; Huang et al., 2008) since their sorp-tion on membrane usually produce surface conditioning. On this

ll rights reserved.

x: +52 55 56162164.errera-Robledo).

regard, molecular weight distribution of soluble microbial prod-ucts (SMP) of anaerobic effluents has been already described (Aqui-no et al., 2006). Moreover, the effect of anaerobic bioreactoroperating conditions (hydraulic residence –HRT- and sludge reten-tion times –SRT–) over SMP production, as carbohydrates (C) andproteins (P), was explored in membrane fouling assays (Huanget al., 2008) before considering the C/P ratio as an experimentalvariable (Arabi and Nakhla, 2008).

Electrostatic (supramolecular) interactions between organicand inorganic matter have been identified as a significant factorduring fouling layer development in water treatment performedby membranes: Wei et al. (2007) found that scaling (inorganicdeposition over membranes) was a direct consequence of biologi-cally-induced mineralization between biopolymers and salts dur-ing reverse osmosis (RO) operations; Kim et al. (2008) based onresults from a membrane autopsy, noticed cell-activity andin situ extracellular polymeric substances (EPS) release duringultrafiltration (UF) performed through polymeric membranes. Also,Teychene et al. (2008) proposed that supramolecular interactionsbetween COO� groups and colloidal inorganic matter were respon-

192 M. Herrera-Robledo et al. / Chemosphere 84 (2011) 191–198

sible for the vast amorphous Ca deposition within cake layer dur-ing membrane bioreactor (MBR) operation. Within the MBR bio-fouling, the numerous electrostatic interactions betweenbiopolymers (EPS) and inorganic constituents (i.e. CaCO3) pro-duced hybrid macro structures that usually remain after conven-tional cleaning procedures (Houari et al., 2010) labeled asmineralized EPS-building blocks (MEBB) (Herrera-Robledo et al.,2010). These structures currently constitute the target for novelcleaning procedures in RO and high-pressure separations.

Considering that MBR biofouling is typically described as anextracellular matrix in which biopolymers, inorganic salts and ac-tive microbes co-exist, biomineralization models can be useful todescribe the spatial organization and environmental constraintswithin this extracellular milieu. For that reason, boundary-orga-nized biomineralization (BOB) (Mann, 1986) and organic phaseas reinforced fibrous network (Silyn-Roberts and Sharp, 1986)arguments were utilized as background in order to (1) evaluateCaCO3 influence on flux decline, pore blocking and cake layer prop-erties (resistance, permeability and compressibility) in a widerange of Chitosan/BSA (Bovine serum albumin) mixtures (ratiosfrom 0.25 to 1.85) during step-pressure runs, and (2) performmembrane autopsies in order to explore the genesis of MEBB dur-ing cake layer build up.

2. Materials and methods

2.1. Reactants and suspensions

Chitosan (LC: 350 kDa, number of glucosamine monomers =1953; 15–20% acetylating, 448869, Sigma–Aldrich, MO) and BSA(66.4 kDa, 583 amino acids residues, pI (isoelectric point) = 4.9–5.3; secondary structure �72%, 9048-46-8, Sigma–Aldrich, MO),were selected as model polymers. Biopolymer mixtures were la-beled as ‘‘LC/BSA’’ and utilized at different ratios (0.25–1.85, Table 1).Each mixture had 233 mg as total mass in 2 L using distilled water;pH of mixtures ranged from 6.24 to 7.16. Electric charge density(ECD) was estimated for each mixture according to De la Torreet al. (2009). ECD associated to chitosan was estimated consideringfree-hydroxyl negative (two) and amine (single) chemical residuesthat surround each glucosamine monomer. BSA primary structureis acidic (�21) and thus ECD estimation was performed consideringthat each mol of BSA contributes to 21 negative charges. Particle sizewas measured via dynamic light scattering in a Mastersizer 2000(Malvern, UK) and ranged from 0.22 to 0.41 lm. Table 1 shows bio-polymer suspensions physicochemical properties.

CaCO3 (C6763, Sigma–Aldrich, MO) was utilized as inorganicmodel in three different concentrations (0.5, 1.5 and 3 mM) inorder to evaluate it contribution to flux decrease and pore-blockage. Chemical potential (DGMix) was calculated according to

Table 1Physicochemical characterization of biopolymer mixtures.

LC/BSAa

[LC]b [BSA]b �Electric chargeslinked to LCc

+Electric chargeslinked to LCd

0.25 1.19E�05 1.41E�06 4.63E�02 2.32E�020.42 1.78E�05 1.23E�06 6.95E�02 3.48E�020.75 2.56E�05 1.00E�06 9.98E�02 4.99E�021.49 3.58E�05 7.00E�07 1.40E�01 6.99E�021.85 3.86E�05 6.10E�07 1.51E�01 7.54E�02

a Weight to weight.b M.c Number of negative electric charges along chitosan, two per each glucosamine mond Number of positive electric charges along chitosan, one per each glucosamine monoe Number of negative electric charges along BSA.f Electric charge density, [ECD] = # charges mol g�1.g The sum of negative ECD of both polymers.

Chang (2005), in order to predict biopolymer-CaCO3 thermody-namics in an isothermal multi-component system (See Supple-mentary material (SM), Table SM-1).

2.2. UF unit and bench-scale AnMBR

Each biopolymer-CaCO3 mixture (2 L) was filtered using frag-ments of commercially available polyvinylidene fluoride mem-brane tubes (ABCOR by Koch, MA). Single membrane tubes(diameter 1.1 cm, length 5.8 cm, 23.5 cm2 as effective surface area,100 kDa molecular weight cut-off –MWCO–, polymeric, hydropho-bic, piezoelectric) were placed in an acrylic housing and assembledin a data acquisition system with a pressure transmitter (0–2100 kPa, 68001 Cole Parmer, IL) and USB card (Ni6009, NationalInstruments, TX), both controlled by a LabVIEW application.Cross-flow velocity was applied by means of a positive displace-ment pump (Moyno 33201, Baldor, AR). Transmembrane pressure(TMP) was controlled by two (1.27 cm internal diameter) globevalves. High-performance Goodyear tubing (Pliovic plus, diame-ter = 0.95 cm, USA) was utilized to guarantee system stability.

Two up-flow anaerobic sludge bed (UASB) effluents (HRT: 4 and8 h, SRT = 180 d), coming from a bench-scale UASB reactor (Plexi-glas made, volume = 12 L) fed with synthetic wastewater were alsotested in UF unit for comparison proposes.

2.3. Fouling layer hydraulic-structural parameters, membranepermeability and pore-blockage measurements

Eight different pressure steps (40 ± 7–540 ± 21 kPa) were uti-lized in order to assess fouling layer properties during the UF ofbiopolymer-CaCO3 mixtures. Each pressure step was held for900 s and the permeate volume was measured every minute. Fordata analysis, the resistances-in-series model (Choi et al., 2006)was utilized in order to investigate fouling characteristics in termsof two filtration resistances (cake layer and membrane). Thefollowing Eqs. (1) and (2) were used to evaluate experimental data:

Jv ¼DPlRT

ð1Þ

RT ¼ Rm þ Rc ð2Þ

where Jv is the permeate volume flux (L m�2 h�1), DP the TMP(kPa), l the viscosity of permeate (Pa s) and RT the total hydraulicresistance. Rm and Rc are membrane and cake layer hydraulic resis-tances (m�1) respectively. For data analysis, each Rc value wascompared in terms of relative value to Rm.

Cake layer compressibility (n) was estimated by using Eq. (3),where Rc0 is the initial cake resistance and n is the compressibility

�Electric charges linkedto BSAe

ECDffor [LC] (66%negative)

ECDffor[BSA]

Overall negativeECDg

2.95E�05 3.88E�04 2.61E�07 2.56E�O42.57E�05 5.82E�04 2.27E�07 3.84E�042.10E�05 8.36E�04 1.85E�07 5.52E�O41.46E�05 1.17E�03 1.29E�07 7.73E�O41.29E�05 1.26E�03 1.14E�07 8.33E�04

omer.mer.

Table 2Pore-blocking progression along UF tests. White spheres represent the best fit to standard blocking model. Black ones denote the best fit to cake filtration model.

Trans membrane pressure step (kPa)

40 ± 7 75 ± 7 135 ± 7 190 ± 14 260 ± 14 310 ± 14 408 ± 14 540 ± 21

Mixtures0.25 s s � � s � � �0.25 + 0.5 mMCaCO3 s s s � � � s �0.25 + 1.5mMCaCO3 s s s Both � � � �0.25 + 3.0 mMCaCO3 � Both s � � � � �0.42 s s � s � � � �0.42 + 0.5 mMCaCO3 s � � � � � � �0.42 + 1.5mMCaCO3 s s s Both � s � �0.42 + 3.0 mMCaCO3 s s Both � � � s �0.75 s s � � � � � �0.75 + 0.5 mMCaCO3 s � � � � � � s

0.75 + 1.5mMCaCO3 s s � s s � � �0.75 + 3.0 mMCaCO3 s � s s � � � �1.49 s � � � s � � �1.49 + 0.5 mMCaCO3 s s � � � � s �1.49 + 1.5mMCaCO3 s s Both � � � s s

1.49 + 3.0 mMCaCO3 � � � � � � � �1.85 s s � � � � � �1.85 + 0.5 mMCaCO3 s � � � � � � s

1.85 + 1.5 mMCaCO3 s s � � � � � �1.85 + 3.0 mMCaCO3 � � � s � � � �

0

2

4

6

8

10

12

14

16

18

20

Rc/R

m

0

0.02

0.04

0.06

0.08

0.1

0.12

0.14

0.25 0.42 0.75 1.49 1.85

KR

LC/BSA

A

B

Fig. 1. (A) Resistances ratio at the end point of the UF trials and (B) Remnant (final) permeability (K/Kw) (Rm = 6.9 � 1011 m�1 and Kw = 5.9 L m�2 h�1 kPa�1).

M. Herrera-Robledo et al. / Chemosphere 84 (2011) 191–198 193

A

B

Fig. 2. (A) Cake layer compressibility indexes and (B) Specific membrane mass at the end of the UF trails.

Fig. 3. Organic to inorganic matter ratios (o/i) in fouling layers.

194 M. Herrera-Robledo et al. / Chemosphere 84 (2011) 191–198

index. When n goes to zero, the cake layer is more uncompressibleand vice versa.

Rc ¼ Rc0TMPn ð3Þ

Permeability (K) values were estimated by using the Jv/TMPratios along the conducted tests. Remnant (final) permeability(KR) was calculated in each test by using the experimental purewater permeability (Kw) as basis (5.9 ± 2.4 L m�2 h�1 kPa�1).

Fouling layer

Matrix vesicles

(a) (b)

(c)

Matrix vesicles

Fig. 4. SEM exploring in the highest ECD (LC/BSA = 1.85) fouling layers. (a) Examination on LC/BSA + CaCO3 0.5 mM fouling layer. (b) Close-up on fouling layer developed afterLC/BSA + 1.5 mM CaCO3 UF trial. (c) Fouling layer developed after UASB effluent UF (HRT = 8 h). Scale bars represent a micrometer.

M. Herrera-Robledo et al. / Chemosphere 84 (2011) 191–198 195

Pore-blocking mechanisms were estimated during each pressure-step by using the four classical filtration laws at constant pres-sure (Hermia, 1982), adapted for EPS filtration (Ye et al., 2005).

Solid-free water (Millipore ultra pure water, pH = 7.1) was uti-lized as blank and filtered at the referred TMP values in order toobtain Kw, Rm (6.9 � 1011 m�1) and pure water initial flux(Jv = 695 ± 285 L m�2 h�1).

2.4. Membrane autopsy

In order to estimate solids deposition over fouled membranes,they were cut into small squares (0.5 � 0.5 cm) and their dry massmeasured (number of samples = 20 for each membrane). In addi-tion, 20 randomly selected squares were used as membrane blanksafter an in vitro chemical cleaning procedure performed as follow:Membrane pieces and abrasive glass particles (425–600 lm; Sig-ma, MO) were gently shaken for 24 h inside tubes containing NaOH(pH 10) and then transferred to tubes with HCl (pH 2) in order toremove biological and inorganic foulants. Both operations consid-ered the pH exposure limits of the membrane as reported by themanufacturer. The resultant chemically cleaned membranes(CCM) also were dried at 105 �C for 2 h previous to mass and sur-face measurement. Specific membrane mass (SMM: mg cm�2) ofCCM was then obtained (36.1 ± 0.3) and used as basis for specificbiofouling mass (SBM: mg cm�2) estimation for UASB effluentbiofouling.

Dynamic thermo gravimetric analysis (TGA) was used to deter-minate organic and inorganic content (organic to inorganic ratio, o/i), within cake layers by using a TGA Q500 V6.3 Build 189 (0–900 �Cand 2 �C min�1, TA Instruments, USA). Virgin membrane spectrawere utilized as blank for organic and inorganic content estima-tion. SEM imaging was performed over fouling-bearing samples

in a Jeol JSM 5310LV SEM without performing any previouspretreatment.

3. Results and discussion

3.1. UF of biopolymer-CaCO3 mixtures

Taking into consideration the molecular weight differences be-tween model foulants (66 and 350 kDa for albumin and chitosanrespectively) and membrane (100 kDa MWCO) a combined pore-blocking mechanism was expected: first, standard blocking (inter-nal) by BSA, then cake layer formation due to LC fibers. After theapplication of empirical models for pore-blocking estimation (Yeet al., 2005, Table 2) either the referred combined blocking or animmediate cake layer development were noticed during step-pres-sure runs. At the end of each UF trial, cake layer (black spheres inTable 2) constituted the main pore-blocking mechanism in practi-cally all the tested conditions. In accordance to Table 2 outcomes,there is no apparent relation between pore blockage and CaCO3

concentration.

3.2. Electrostatic interactions between biopolymers and CaCO3

Considering that interactions between the tested biopolymersand CaCO3 are spontaneous (the partial molar Gibbs energies arenegative for all tested reactions, see Table SM-1), and the fact thatsharp increases in cake layer resistances were detected after CaCO3

addition to EPS mixtures (Fig. 1a and b), an assumption can bemade linking the differences observed in experimental parameters(resistances ratios and KR) to the electrostatic (supramolecular)interactions between biopolymers and the inorganic salt. Thenon-linear Rc response to CaCO3 possibly was a consequence of

I

II

III

IV

V

BSA-CaCO3

LC-CaCO3

Organic-inorganic hybrids

Mineralized-EPS-building block

Fig. 5. MEBB genesis and development schematically represented. Fine points aredescribed in Section 3.5. (For interpretation of the references to color in this figurelegend, the reader is referred to the web version of this article.)

196 M. Herrera-Robledo et al. / Chemosphere 84 (2011) 191–198

the dynamic behavior of cake layer development during cross-flowUF, wherein erosion and sloughing are ordinary (Klausen et al.,2003; Yeon et al., 2009).

3.3. Biopolymer-CaCO3 and AnMBR biofouling layers

There is no clear difference between the observed compressibil-ity indexes (Fig. 2a) and the UASB operating conditions (n = 0.93and 0.91 for HRT = 8 and 4 h respectively). However, these valuesare comparable with those observed for aerobic MBR biofouling(0.61 and 0.95) for two different Ca2+ concentrations according toZhang et al. (2009). The values of n observed after biopolymer-CaCO3 UF analysis would represent the full spectrum of compress-ibility for MBR biofouling (0.6–1.2) and constitute valuable data forbiofouling modeling.

SMM (mg cm�2) information is presented in Fig. 2b. In broadterms, for the lower LC/BSA ratios (0.25 and 0.42), SMM valuesdiminish as CaCO3 concentration increases and for the higher LC/BSA ratio (1.85) SMM augment. For the first group, decreases inSMM values can be linked to increases in BSA solubility (salting-in effect via CaCO3) causing as effect lower BSA accumulation onmembrane; and for the last condition associated with higher bio-polymer-CaCO3, complexation. The o/i ratio for the 5 experimentalconditions and the 2 UASB effluents are presented in Fig. 3. Themineral fraction increases with CaCO3 at lowest and higher LC/BSA ratio. Both cases reflect different cake layer structures in whichthe insoluble chitosan scaffolding is crucial, considering that in alltested LC/BSA ratios, CaCO3 addition resulted in increases in cakelayer resistance and decreases in permeability (Fig. 1). SMM values(Fig. 2b) in the lower LC/BSA ratios add arguments for debating ifthe accumulated mass is more important than the structure/archi-tecture within the MBR fouling layers, as several reports have beenpointed out (Ahn et al., 2008; Houari et al., 2010).

SMM values linked to UASB effluents (53.7 and 42.3 for 8 and4 h HRT) can be used for SBM estimation by subtracting the CCMvalue (36.1 ± 0.3) from SMM. As a result, a higher SBM (17.6) wasobtained for the UASB reactor effluent at usual operating condi-tions for full-scale UASB applications (HRT = 8 h) than the observedSBM from an over-loaded condition (HRT = 4 h, 6.2 mg cm�2). Theconsiderable differences between UASB effluent SBM and the asso-ciated to mimic mixtures can be associated to deposition and sorp-tion by active microbial communities (Wei et al., 2007; Volesky,2007; Kim et al., 2008; Chan et al., 2009).

3.4. Biomineralization in the highest electric charge density

UF of biopolymer/CaCO3 created massive fouling on membraneeither for protein or for chitosan-rich conditions. However, the ex-pected mineralized structures were better observed in the lastcase. Micrographs of cake layers developed with a LC/BSA ra-tio = 1.85, the most spontaneous process and also with the highestelectric charge density, are presented in Fig. 4. The layers appar-ently were comprised by regular and irregular assemblies of thecomponents used as mimics with an estimated thickness of 2 lm(data not shown). Either UASB effluents or their mimics producedsimilar fouling layers (Fig. 4a and c) in which apparent matrix ves-icles (MV) suggest biomineralization. MV (20–200 nm in diameter)are extracellular structures, spherical-shaped and delimitated by alipid membrane, that usually are associated with early biomineral-ization in cartilage, bone and teeth (Blandford et al., 2003; Golub,2009). The tested cake layers, therefore, could be assumed as amineralized tissue.

LC/BSA/CaCO3 accumulation on membrane caused sharp de-creases in membrane permeability since biopolymer-CaCO3 com-plexation is thermaldynamically spontaneous and favors theconsolidation of hybrid extracellular building blocks, situation that

enhances cake layer development as TMP increase. The fact thatbiopolymer-CaCO3 step-pressure UF created compressible cakelayers (presumably with internal voids) suggests the formation ofan internal skeleton of MEBB. In these structures, biomineraliza-tion may take place along voids boundaries according to BOB pre-mises (Mann, 1986) contributing to irreversible foulingconsolidation (development of mineralized tissue).

3.5. Hypothesis for MEBB development during cross-flow UF inpolymeric membranes

The following hypothesis (Fig. 5) is supported on the fact thatinteractions between polymers and CaCO3 are thermodynamicallyspontaneous (Table SM-1) and also considers that: (1) chitosan fi-bers (350 kDa) are able to develop a hydrated matrix over mem-brane surface (100 kDa MWCO) and (2) BSA molecules(hydrophobic, 66 kDa) adsorb either over or inner membrane andin addition, they are capable of providing nucleation sites for min-eral deposition.

I. Biopolymer-CaCO3 complexation. Spontaneous coordinationbetween Ca2+ and biopolymers can be amplified as response tooperational pressure and turbulence. II. Structural shift in BSA.Meanwhile, chitosan remained as fibers after the Ca2+ loading;the albumin structure may perform an N-F transition (a-helix

M. Herrera-Robledo et al. / Chemosphere 84 (2011) 191–198 197

segments turn to b-sheets) (Addadi and Weiner, 1985; Huntingtonand Stein, 2001), favoring either biomineralization via a templateeffect (He et al., 2009; Wan et al., 2009) or BSA–chitosan associa-tions. III. BSA linking. The continuous cation addition to BSA wouldlead to H-bond fracture and protein denaturalization. As result, su-pra molecular concatenation of 50 albumin molecules occurs,according to the conclusion of calcium biomineralization mediatedby ovoalbumin (Pipich et al., 2008). Taking into account that eachBSA molecule has 66 kDa, the referred concatenation would pro-duce BSA–Ca2+ hybrids ranging from 3000 to 3500 kDa. Consider-ing that inside extracellular matrixes, high-molecular weightmolecules are common and perform structural roles linked toself-assembly capacities (Engel et al., 1981; Bruch et al., 1989) itis suitable to postulate that macro BSA–Ca2+ hybrids carry outstructural roles (i.e. building blocks) on membrane surface, favor-ing cell attachment. As consequence of their retention over mem-brane (due to their weight) BSA–Ca2+ hybrids are continuouslyre-circulated to tank, in which interactions with LC and Ca2+ couldraise their mass and volume. IV. Calcium nucleation. Some albu-mins perform the biological function of calcium storage (Hunting-ton and Stein, 2001) and are able to create amorphous calciumcarbonate particles (ACC), by means of calcium stabilization insideprotein aggregates (Balz et al., 2005a and Balz et al., 2005b). Onthis scenario, fibrous LC-Ca2+ hybrids would stabilize the macro-structures either by template effect or by bridge associations withBSA–Ca2+ spots, in accordance to BOB premises for biomineraliza-tion (Mann, 1986). V. Chitosan fibers as reinforced scaffolding.Throughout phase transitions as result of TMP or local flux in-creases, ACC particles can crystallize, meanwhile biopolymers re-main inside the CaCO3 polymorphs (Silyn-Roberts and Sharp,1986; Pipich et al., 2008) acting as fibrous reinforced network. Ascorollary, mineralized EPS building blocks will cause either irre-versible standard blocking or superficial fouling and even, both.

4. Conclusions

LC/BSA fouling layers over polymeric membranes apparentlywere favored by CaCO3 in a non-linear way. Electrostatic interac-tions between polymers and inorganic salts during UF were ther-modynamically spontaneous and produced a compressiblenetwork of organic–inorganic hybrids (MEBB) within fouling lay-ers. The hypothesis for MEBB genesis and development needs tobe deciphered by means of supplementary experiments, in whichhydrodynamic radii measurements would provide the elementsto accomplish MEBB structural metamorphosis throughout foulingprogression.

Acknowledgements

CONACyT (083670) and DGAPA-UNAM (PAPIIT grant IN115308)supported this work. L Santiago was on charge to develop LabViewapplication. V Franco, M Salmón, L Hernández, C Velasco and VSaucedo provided fruitful discussion. P Altuzar, A del Real, M Cisn-eros and R Briones gave valuable technical support. This researchwas conceived during NiA (Nano-Ingeniería Ambiental) sessions(NiA-1).

Appendix A. Supplementary material

Supplementary data associated with this article can be found, inthe online version, at doi:10.1016/j.chemosphere.2011.04.059.

References

Addadi, L., Weiner, S., 1985. Interactions between acidic proteins and crystals:stereo chemical requirements in biomineralization. Proc. Natl. Acad. Sci. 82,4110–4114.

Ahn, W.-Y., Kalinichev, A., Clark, M., 2008. Effects of background cations on thefouling on PES membranes by NOM: experimental and molecular modelingstudy. J. Membr. Sci. 309, 128–140.

Aquino, S., Hu, A., Akram, A., Stuckey, D., 2006. Characterization of dissolvedcompounds in submerged anaerobic membrane bioreactors (SAMBRs). J. Chem.Technol. Biotechnol. 81, 1804–1904.

Arabi, S., Nakhla, G., 2008. Impact of protein/carbohydrate ratio in the feedwastewater on the membrane fouling in membrane bioreactors. J. Membr. Sci.324, 142–148.

Asatekin, A., Mennitti, A., Kang, S., Elimelech, M., Morgenroth, E., Mayes, A., 2006.Antifouling nanofiltration membrane bioreactors from self-assembling graftcopolymers. J. Membr. Sci. 285, 81–89.

Balz, M., Barriau, E., Istratov, V., Frey, H., Tremel, W., 2005a. Controlledcrystallization of CaCO3 on hyperbranched polyglycerol adsorbed to self-assembled monolayers. Langmuir 21, 3987–3991.

Balz, M., Therese, H., Li, J., Gutmann, J., Kappl, M., Nasdala, L., Hofmeiser, W., Butt, H.,Tremel, W., 2005b. Crystallization of vaterite nanowires by the cooperativeinteraction of tailor-made nucleation surfaces and polyelectrolites. Adv. Funct.Mater. 15, 683–688.

Blandford, N., Sauer, G., Genge, B., Wu, L., Wuthier, R., 2003. Modeling of matrixvesicle biomineralization using large unilamellar vesicles. J. Inorg. Biochem. 94,14–27.

Bruch, M., Landwehr, R., Engel, J., 1989. Dissection of laminin by cathepsin G into itslong-arm and short-arm structures and localization of regions involved incalcium dependent stabilization and self-association. Eur. J. Biochem. 185, 271–279.

Chan, C., Fakra, S., Edwards, D., Emerson, D., Banfield, J., 2009. Iron oxyhydroxymineralization on microbial extracellular polysaccharides. Geochim.Cosmochim. Acta 73, 3807–3818.

Chang, R., 2005. Physical Chemistry for the Biosciences, first ed. University SciencesBooks, Mill Valley, USA.

Choi, H., Zhang, K., Dionysiou, D., Oerther, D., Sorial, G., 2006. Effect of activatedsludge properties and membrane operation conditions on foulingcharacteristics in membrane bioreactors. Chemosphere 63, 1699–1708.

De la Torre, T., Harff, M., Lesjan, B., Drews, A., Kraume, M., 2009. Characterization ofpolysaccharide fouling of an ultra filtration membrane using model solutions.Desalinat. Water Treat. 8, 17–23.

Drews, A., 2009. Membrane fouling in membrane bioreactors – characterization,contradictions, cause and cures. J. Membr. Sci. 363, 1–28.

Engel, J., Odermatt, E., Engel, A., Madri, J., Furthmayr, H., Timpl, H., 1981. Shapes,domain organizations and flexibility of laminin and fibronectin, twomultifuncional proteins of the extracellular matrix. J. Mol. Biol. 150, 97–120.

Golub, E.E., 2009. Role of matrix vesicles in biomineralization. Biochim. Biophys.Acta 1790, 1592–1598.

He, W., Zhou, W., Wang, Y., Zang, X., Zhao, H., Li, Z., Yan, S., 2009. Biomineralizationof iron phosphate nanoparticles in yeast cells. Mater. Sci. Eng. C 29, 1348–1350.

Hermia, J., 1982. Constant pressure blocking filtration law: application to power lawnon-Newtonian fluids. Trans. I. Chem. E 60, 183–186.

Herrera-Robledo, M., Morgan-Sagastume, J.M., Noyola, A.A., 2010. Biofouling andpollutant removal during long-term operation of an AnMBR treating municipalwastewater. Biofouling 26, 23–30.

Houari, A., Seyer, D., Couquard, F., Kecili, K., Democrate, C., Heim, V., Di Martino, P.,2010. Characterization of the biofouling and cleaning efficiency ofnanofiltration membranes. Biofouling 26, 15–21.

Huang, Z., Ong, S., Ng, H., 2008. The effect of HRT and SRT on performance ofsubmerged anaerobic membrane bioreactor (SAMBR) for low-strengthwastewater treatment. In: Proceedings of the IWA North American MembraneResearch Conference. University of Massachusetts, Amherst.

Huntington, J., Stein, P., 2001. Structure and properties of ovalbumin. J. Chromatogr.B 756, 189–198.

Jeison, D., van Lier, J., 2006. On-line cake layer management by trans-membranepressure steady-state assessment in anaerobic membrane bioreactors forwastewater treatment. Biochem. Eng. J. 29, 204–209.

Kim, J., DiGiano, F., Reardon, R., 2008. Autopsy of high-pressure membranes tocompare effectiveness of MF and UF pretreatment. Water Res. 42, 697–706.

Klausen, M., Aaes-Jørgensen, A., Molin, S., Tolker-Nielsen, T., 2003. Involvement ofbacterial migration in the development of complex multicellular structures inPseudomonas aeruginosa biofilms. Mol. Microbiol. 50, 61–68.

Mann, S., 1986. On the nature of boundary-organized biomineralization (BOB). J.Inorg. Biochem. 28, 363–371.

Meng, F., Chae, S.-R., Drews, A., Kraume, M., Shin, H.-S., Yang, F., 2009. Recentadvances in membrane bioreactors (MBRs): membrane fouling and membranematerial. Water Res. 43, 1489–1512.

Pipich, V., Balz, M., Wolf, S., Tremel, W., Schwahn, D., 2008. Nucleation and growthof CaCO3 mediated by the egg-white protein albumin: a time-resolved in situstudy using small-angle neutron scattering. J. Am. Chem. Soc. 130, 6879–6892.

Rosenberger, S., Laabs, C., Lesjan, B., Gnirss, R., Amy, G., Jekel, M., Schrotter, J.C.,2006. Impact of colloidal and soluble organic material on membraneperformance in membrane bioreactors for municipal wastewater treatment.Water Res. 40, 710–720.

198 M. Herrera-Robledo et al. / Chemosphere 84 (2011) 191–198

Silyn-Roberts, H., Sharp, R.M., 1986. Crystal growth and the role of the organicnetwork in eggshell biomineralization. Proc. R. Soc. Lond. B 227, 303–324.

Teychene, B., Guigui, C., Cabassud, C., Amy, G., 2008. Toward a better identificationof foulant species in MBR process. Desalination 231, 27–34.

Volesky, B., 2007. Biosorption and me. Water Res. 41, 4017–4029.Wan, P., Zhao, Y., Tong, H., Yang, Z., Zhu, Z., Shen, X., Hu, J., 2009. The inducing effect

of lecithin liposome organic template on the nucleation and crystal growth ofcalcium carbonate. Mater. Sci. Eng. C 29, 222–227.

Wei, M., Zhao, Y., Wuang, L., 2007. The pre-treatment with enhanced coagulationand UF membrane for seawater desalination with reverse osmosis. Desalination203, 256–259.

Ye, Y., Le-Clech, P., Chen, V., Fane, A., Jefferson, B., 2005. Fouling mechanisms ofalginate solutions as model extracellular polymeric substances. Desalination175, 7–20.

Yeon, K.-M., Cheong, W.-S., Oh, H.-S., Lee, W.-N., Hwang, B.-K., Lee, C.-H., Beyenal, H.,Lewandowski, Z., 2009. Quorum sensing: a new biofouling control paradigm inmembrane bioreactor for advanced wastewater treatment. Environ. Sci.Technol. 43, 380–385.

Zhang, H., Xia, J., Yang, Y., Wang, Z., Yang, F., 2009. Mechanism of calcium mitigatingmembrane fouling in submerged membrane bioreactors. J. Environ. Sci. 21,1066–1073.