- Email: [email protected]
Membrane Foulants Characterization in a Drinking Water Production Unit
MEMBRANE FOULANTS CHARACTERIZATION IN A DRINKING WATER PRODUCTION UNIT B. Doume`che1, L. Galas2, H. Vaudry2 and P. Di Martino1, 1
Laboratoire ERRMECe (EA1391), Universite´ de Cergy-Pontoise, Pontoise, France. Plate-forme re´gionale de recherche en imagerie cellulaire, INSERM U413 – IFRMP 23, Universite´ de Rouen, Mont-Saint-Aignan, France.
2
Abstract: We characterized the composition and organization of membrane foulants in a French drinking water production plant between March and September 2005 by the combination of Attenuated Total Reflection Fourier Transform Infrared (ATR-FTIR) spectroscopy, lectin-binding analysis, epifluorescence microscopy and Confocal Laser Scanning Microscopy (CLSM). The comparison of ATR-FTIR spectra of clean and fouled membranes revealed a qualitatively homogenous deposition of biological matter onto the membrane surface, a high heterogeneity in the composition of the fouling material and a high temporal stability of this composition. Characteristic signals of mainly polysaccharides (900 –1200 cm21) and proteins (1300 –1700 cm21) were observed. Fluorescence microscopy observations after nucleic acid staining with DAPI and polysaccharides staining with lectins labelled with fluorescein isothiocyanate or tetramethylrhodamine isothiocyanate indicated a high spatial heterogeneity inside the foulant matter. The microbial cells, mainly composed of bacteria, were localized in the superficial layer of the fouling material and were organized as microcolonies interspersed at the membrane surface. Polysaccharide residues were found in areas where microcolonies were present and in areas devoid of microbial cells. High staining with Peanut and Bandeiraea simplicifolia agglutinins (PNA, BS-1, respectively) revealed high occurrence of galactosides residues in the polysaccharide components of the foulants. The BS-1 lectin staining pattern indicated a high degree of spatial organization with the observation of long and entangled fibres. Wheat germ agglutinin staining showed short fibres and cloud stained areas. PNA and Concanavalin A lectin staining were more interspersed. In conclusion, the combination of ATR-FTIR spectroscopy with lectin-binding analysis was a very informative tool for the characterization (composition and spatial organization) of the foulants that cover the surface of the membranes. Correspondence to: Dr P. Di Martino, Laboratoire ERRMECe UFR Sciences et Techniques Universite´ de CergyPontoise, 2 avenue Adolphe Chauvin BP222 – 95302 Pontoise cedex France. E-mail: patrick.di_martino@ u-cergy.fr
DOI: 10.1205/fbp06020 0960–3085/07/ $30.00 þ 0.00 Food and Bioproducts Processing Trans IChemE, Part C, March 2007 # 2007 Institution of Chemical Engineers
Keywords: fouling; biofilm; membrane; water; lectin
INTRODUCTION
leading to biofilm development, referred to as biofouling, is a major component of membrane fouling in water separation applications (Bos et al., 1999). In drinking water plants, biofouling leads to higher operating pressures, increases the frequency of chemical cleanings, and leads to membrane deterioration (Flemming, 2002). Moreover, biofilms and other materials accumulated on the membrane surface, which cannot be removed by cross flow, backflushing, or backpulsing, can result in permanent permeability loss. One characteristic of a biofilm is the production of an extracellular matrix that envelops the attached cells (Allison, 2003). This viscous layer is generally composed of water and microbial polymers and provides a complex array of microenvironments surrounding the attached cells. These extracellular biopolymers usually include exopolysaccharides, proteins, nucleic acids, and
Fouling is a commonly encountered problem during food and drink processing (Changani et al., 1997; James et al., 2003; Mittelman, 1998; Scha¨fer et al., 2000). Membrane fouling can result from the accumulation of multivalent ions, organic matter and microbial cells in suspension and from the development of biomass at the membrane surface. This matter accumulation leads to a decrease of the process efficiency and then to an increase in the production cost. In the drinking water industry, membrane fouling is known to depend on several parameters such as the composition of the feed water, the physicalchemical properties of membranes and also the hydrodynamic conditions (AWWA Membrane Technology Research Committee, 2005). Irreversible adherence and growth of microbial cells on the membrane surface 42
Vol 85 (C1) 42–48
MEMBRANE FOULANTS CHARACTERIZATION IN A DRINKING WATER PRODUCTION UNIT phospholipids. The exopolysaccharides (EPS) support the membrane surface colonization by microbial cells, facilitate the arrangement of different species within the biofilm, and protect the sessile community against toxic compounds in solution. Moreover, polysaccharide organic matter is responsible for the evolution of irreversible fouling (Kimura et al., 2004). A better knowledge of the fouling material composition and organisation can help in the development of strategies for fouling prevention and control (Flemming et al., 1997; Verran et al., 2001; Violleau et al., 2005; Xu et al., 2006). There are many methods described for assessing the fouling of surfaces by microorganisms or by organic material but a relatively simple method for assessment of the components in combination is not readily available (Verran, 2002). Some infrared spectroscopic analyses can be used to elucidate the nature of foulants of predominantly organic and biological origin (Her et al., 2000; Howe et al., 2002; Kimura et al., 2004; Nivens et al., 1993; Rabiller-Baudry et al., 2002; Suci et al., 1997). Fluorescently labelled lectins have also been described in combination with confocal laser scanning microscopy (CLSM) to allow the visualization and characterization of carbohydrate-containing extracellular polymers in biofilms (Neu et al., 2001). A combination of infrared spectroscopy and fluorescence microscopy associated with fluorescently labelled lectins has never been used before to investigate membrane fouling. The primary objective of this research was to identify the major components of the fouling material from filtration membrane surfaces and to determine their seasonal variations. We performed three autopsies of filtration membranes in March, June and September 2005, respectively, and analysed the foulant matter by the combination of attenuated total reflection Fourier transform infrared (ATR-FTIR) spectroscopy, epifluorescence and CLSM observations.
METHODS AND MATERIALS Membrane Autopsy The membranes used in the plant were composed of an ultra thin top layer made of polypiperazine amide and a polysulfone microporous support. Membranes were in service for at least 5 years before our experiments. In March, June and September 2005, some membrane modules were extracted from the drinking water plant, partially drained of excess liquid, disassembled, stored at 48C and analysed within 24 h in the laboratory. Some samples were air-dried and analysed by ATR-FTIR and other wet samples were fixed in paraformaldehyde (PFA) before staining and fluorescence microscopy analysis, as described below.
Analysis of Membrane Foulants by ATR-FTIR Samples of air-dried fouled membrane were analysed by ATR-FTIR. ATR-FTIR spectra were recorded using a Tensor 27 IR spectrophotometer with a 458 diamond/ZeSe flat plate crystal and an average depth penetration of 2 mm. Each spectrum presented is the result of 32 scans obtained with a resolution of 2 cm21 with air as the background.
Lectin Staining of Foulants Samples of the fouled membranes were treated with paraformaldehyde (4%, v/v) for 60 min prior to lectin application in
43
order to fix the foulant matter. Fluorescently labelled lectins (Sigma, Saint Quentin Fallavier, France) conjugated with fluorescein isothiocyanate (FITC) or tetramethylrhodamine isothiocyanate (TRITC) were used for the study of their binding to polysaccharides contained in the fouling layer (Table 1). Stock solutions of lectins (1 mg ml21) were prepared by adding filter-sterilized (0.2 mm) water and stored frozen. Prior to use, a thawed portion was diluted with filter-sterilized water and centrifuged to remove crystals and debris. Double staining with two lectins, one FITC- and one TRITC-labelled were done with a mixture of Bandeiraea simplicifolia agglutinin (BS-1) and concanavalin A (ConA) or a mixture of peanut agglutinin (PNA) and wheat germ agglutinin (WGA). 100 ml of a mixture of two lectins (final concentration of 100 mg ml21 each) were carefully applied directly on top of the membrane. After incubation for 30 min in the dark at room temperature, unbound lectins were removed by washing with filter-sterilized water. The fluorescent DNA-binding stain 40 ,6-diamidino-2phenyindole dihydrochloride (Sigma, Saint Quentin Fallavier, France) was used to visualize cell distribution in combination with lectin staining. Following the lectin staining step, the fouled membranes were treated with 100 ml of a solution of DAPI (1 mg l21) in filter-sterilized water. After incubation for 30 min in the dark at room temperature, unbound DAPI was removed by washing the membrane surface with filtersterilized water. The stained preparations were then mounted with Mowiol (Calbiochem, Meudon, France) and stored at 48C in the dark. For the negative control, the lectins were replaced by filter-sterilized water.
Microscopy and Image Analysis The fouled membranes stained with DAPI and fluorescently labelled lectins were examined with a Leica epifluorescence microscope (MPS 60) and with a Leica SP2 upright confocal laser scanning microscope (DM RAX-UV) equipped with the Acousto-Optical Beam Splitter (AOBS) system and using 63, N.A. 1.32, oil immersion objective (Leica microsystems, Rueil-Malmaison, France). For epifluorescence images, DAPI was excited at 364 nm, FITC was excited at 494 nm, and TRITC was excited at 550 nm. For confocal images, DAPI was excited at 405 nm and observed from 410 to 600 nm, FITC was excited at 488 nm and observed from 505 to 540 nm, TRITC was excited at 543 nm and observed from 560 to 600 nm. The selection of spectral emission window for each fluorophore has been determined through l scan analysis on single stained membrane fragments. The gain and offset for each photomultiplier have been adjusted to optimize nucleic acid and lectins detection.
Table 1. Carbohydrate-binding specificities of lectins employed for staining of membrane foulants. Lectin (abbreviation)
Conjugate
Main specificity
Bandeiraea simplicifolia agglutinin (BS-1) Concanavalin A (ConA) Peanut agglutinin (PNA) Wheat germ agglutinin (WGA)
TRITC
a-gal, a-galNAc
FITC TRITC FITC
a-man, a-glc b-gal(1!3)galNAc (glcNAc)2, NeuNAc
TRITC, tetramethylrhodamine isothiocyanate; FITC, fluoresceinisothio cyanate.
Trans IChemE, Part C, Food and Bioproducts Processing, 2007, 85(C1): 42–48
44
DOUME`CHE et al.
Images of the CLSM observations (1024 1024 pixels) have been acquired through a sequential mode, between stacks, to exclude spectral crosstalk from the data. 400 Hz scan speed was used and signal/noise ratio has been increased through frame average. Overlay and maximum projection of the z-stacks files have been performed with post acquisition Leica Confocal Software (LCS) functions to obtain the presenting snapshots. Original z-stack Leica files have then been imported into Imaris 4.0 software (Bitplane AG, Zu¨rich) to obtain snapshots illustrating xz representation and 3D modelling. Geometric shapes have been used to represent the microbial microcolonies and the biofilm matrix inside the foulant matter in 3D.
RESULTS, ANALYSES AND DISCUSSION Analysis of Membrane Foulants by ATR-FTIR The FTIR spectra of the clean membrane and fouled membranes extracted in March, June and September 2005 are compared in Figure 1. Different spectra illustrating heterogeneity among the foulant layer are presented for each fouled membrane. Several spectral features were observed in the 1700 –900 cm21 range and in the 1700–700 cm21 range for the fouled and clean membranes, respectively. For the fouled membranes, IR bands indicative of biomass, as previously defined (Nivens et al., 1993), were detected near 1650 cm21 (amide I), 1550 cm21 (amide II), 1450 cm21 (due in part to C-H deformation), 1400 cm21 (due in part to symetric stretch for the carboxylate ion), and 1250 cm21
(P5 5O and C2 2O2 2C stretching and/or amide III). The broad complex region from 1250 to 900 cm21 has been characterized in the literature as the region where both DNA-RNA and polysaccharides appear (Suci et al., 1997). Vibrations in this region are characterized by C2 2O2 2C, C2 2O, ringstretching vibrations and the P5 5O stretch of phosphodiesters. The FTIR profiles were very similar for the fouled membranes analysed at the three different seasons. Thus, the foulant matter was overall stable in nature between March and September. None of the peaks of the clean and new membrane were observed in the spectra obtained in March or June, indicating the total coverage of the membrane by the fouling layer. For the fouled membrane extracted in September, IR bands corresponding to the membrane were detected near 700, 850, 1150 and 1250 cm21 for one of the four presented spectra, thus underlining local heterogeneity within the fouling layer [Figure 1(c); dotted spectrum].
Analysis of Membrane Foulant by Lectin Staining and Fluorescence Microscopy Polymeric membranes can be subjected to autofluorescence emission after excitation at the wavelengths used in histology. The background level of fluorescence from polymers depends on their content in aromatic structures and conjugated double bonds in their molecular structure (Shadpour et al., 2006). As shown in Figure 2, the background level of autofluorescence of the filtration membranes extracted from the plant did not prevent observing the
Figure 1. ATR-FTIR spectra of fouled membranes in March (a), June (b) and September (c). The different spectra observed at one time differ from their localization onto the membrane surface. The spectra of clean membrane are given as reference (d). Trans IChemE, Part C, Food and Bioproducts Processing, 2007, 85(C1): 42–48
MEMBRANE FOULANTS CHARACTERIZATION IN A DRINKING WATER PRODUCTION UNIT
45
Figure 2. Interactions between fluorescently-labelled lectins and fouled membranes, as visualized by epifluorescence microscopy. Shown are fouled membranes extracted in March [(a) to (f)], June [(g) to (l)) and September [(m) to (r)] 2005, stained with: DAPI þ PNA-TRITC þ WGA-FITC [(a), (b), (c), (g), (h), (i), (m), (n), (o)]; DAPI þ BS-1-TRITC þ ConA-FITC [(d), (e), (f), (j), (k), (l), (p), (q), (r)]. The panels coloured in blue [(a), (d), (g), (j), (m), (p)], green [(b), (e), (h), (k), (n), (q)] and red [(c), (f), (i), (l), (o), (r)] correspond to observations after illumination at 364, 494, and 540 nm respectively. Magnification,630.
Trans IChemE, Part C, Food and Bioproducts Processing, 2007, 85(C1): 42–48
46
DOUME`CHE et al.
microbial cells and exopolysaccharides after staining with DAPI and fluorophore-conjugated lectins. Epifluorescence observations after DAPI staining showed that the microbial part of the fouling layer was mainly composed of bacteria organized as microcolonies interspersed at the membrane surface [Figure 2(a), (d), (g), (j), (m), (p)]. Some algae were also present, as shown by autofluorescence properties of these cells. We did not observe any extracellular DNA (eDNA) stained with DAPI. The eDNA has been previously shown to be involved in biofilm stability during the early stages of biofilm growth in vitro (Whitchurch et al., 2002). The eDNA abundance inside biofilms has been shown to be strain dependent and the accumulation of eDNA in dualspecies biofilms is unpredictable (Steinberger and Holden, 2005). Moreover, the presence of eDNA in natural and industrial multiple-species environments is still unknown. The biomass attached at the membrane surface increased from March to June and seemed to be unchanged from June to September. The application of lectin-binding-analysis to define glycoconjugate composition of the foulants was used. As shown in Figure 2, epifluorescence observations revealed two types of polymeric carbohydrate structures: those located on cell surfaces and those located extracellularly throughout the biofilm matrix. Lectin staining was not only concentrated in areas where microcolonies were present but also extended in areas devoid of microbial cells. This
binding pattern of the lectins has been previously observed with environmental biofilms grown in vitro with river water as the sole source of carbon and nutrients (Neu et al., 2001). During the study, the dominant lectin binding was with BS-1 revealing high occurrence of galactosides residues in the polysaccharide part of the foulants. Nevertheless, the polysaccharide composition of the fouling layer changed from March to September. Lectin staining increased from March to September for all the lectins used. The lectin-binding changes with time may be linked to an increase of the biomass attached at the membrane surface and to changes among the populations of attached cells. Staining with BS-1 increased constantly in March, June and September [Figure 2(f), (l), and (r)]. A high increase of binding with PNA, and ConA was observed between March and June, but the binding of these two lectins did not change between June and September [Figure 2(c), (i), (o), and (e), (k), (q), respectively]. Staining with the WGA was weak in March and June and was higher in September [Figure 2(b), (h), and (n)]. Nutrients, oxygen level and the concentration of metals have been shown to influence the exopolymer abundance of environmental model biofilms grown in vitro with river water as the sole source of carbon and nutrients (Lawrence et al., 2004). The modification of these parameters has been shown to lead to a shift in the glycoconjugate makeup of the biofilms. In our study, the glycoconjugate composition of the foulants at the membrane
Figure 3. Interactions between fluorescently-labelled lectins and fouled membranes, as visualized by CLSM. Shown are fouled membranes extracted in June stained with: DAPI þ PNA-TRITC þ WGA-FITC (a); DAPI þ BS-1-TRITC þ ConA-FITC (b). Colour allocation: blue—DAPI, green—FITC-labelled lectin and red—TRITC-labelled lectin. (1) Magnification 630. (2) Enlargement of the dotted square zones. White arrows indicate the orientation of the foulant matter from the bottom to the top. Trans IChemE, Part C, Food and Bioproducts Processing, 2007, 85(C1): 42–48
MEMBRANE FOULANTS CHARACTERIZATION IN A DRINKING WATER PRODUCTION UNIT
47
Figure 4. Three-dimensional representation of the surface of a fouled membrane stained with DAPI þ PNA-TRITC þ WGA-FITC obtained with the Imaris 4.0 software. Shown are the entire fields corresponding to the observation of the membrane at a magnification of 630 (a), and the enlargement of the dotted squared part of the field (b). DAPI, TRITC, and FITC fluorescence appears in blue, red and green, respectively.
surface was relatively stable. Further studies are needed to determine whether this stability of biofilm is linked to a relative stability of nutrients, oxygen level and metals in the surface water of the river. Confocal laser scanning microscopy was used to study the spatial organization and distribution of the fouling layer at the membrane surface. Multiple stains were applied and recorded simultaneously: DAPI þ PNA-TRITC þ WGA-FITC and DAPI þ BS-1-TRITC þ ConA-FITC. Examples of these multiple staining experiments are shown in Figure 3(a) and (b), respectively. The post acquisition Leica Confocal Software (LCS) allowed the elimination of the membrane autofluorescence background. CLSM analysis confirmed that most of the analysed polysaccharides (stained with the lectins used) were located extracellularly throughout the biofilm matrix. A high degree of spatial organization and high heterogeneity were observed within the fouling material: WGA binding showed short fibres and cloud stained areas; BS-1 binding revealed long and entangled fibres; PNA and ConA staining were more dispersed. Confocal images showing the Z-axis revealed heterogeneity in depth: the stained part of the fouling matter varied in depth from 6 to 27 mm (Figure 4). The microbial cells, mainly organized as microcolonies interspersed at the membrane surface, were localized in the superficial layer of the fouling deposit (Figure 4). These staining patterns suggested that the fouling layer was organized as an interpenetrated network mainly composed of exopolysaccharides supporting microbial cells adhesion.
CONCLUSION In conclusion, the combination of ATR-FTIR spectroscopy with DAPI staining and lectin-binding-analysis was efficient to study the composition, the spatial organization and the seasonal variations of the fouling material present at the
membrane surface used for drinking water production. Data collected by this combined approach gave important new information about the membrane foulants, extending previous observations of foulant speciation related to membrane filtration efficiency (Violleau et al., 2005; Xu et al., 2006). Epifluorescence microscopy allowed determining the qualitative and quantitative evolutions of the foulant matter. The more sophisticated and expensive CLSM gave the opportunity to observe the spatial organization and 3D structure of the foulants. The fouling material exhibited the same characteristics as a biofilm. These biofilms developed at the surface of membranes in a drinking water plant may result from the presence of microbial cells and nutrients coming from feed water and also from material that has not been removed even after successive cleaning and production processes. A better knowledge of the composition and seasonal variations of the foulants will help us in the development of more efficient cleaning processes of membranes.
REFERENCES Allison, D.G., 2003, The biofilm matrix, Biofouling, 19(2): 139 –150. AWWA Membrane Technology Research Committee, 2005, Recent advances and research needs in membrane fouling, Journal AWWA, 97(8): 79–89. Bos, R., Van Der Mei, H.C. and Busscher, H.J., 1999, Physicochemistry of initial microbial adhesive interactions—its mechanisms and methods for study, FEMS Microbiol Rev, 23: 179– 230. Changani, S.D., Belmar-Beiny, M.T. and Fryer, P.J., 1997, Engineering and chemical factors associated with fouling and cleaning in milk processing, Exp Thermal Fluide Sci, 14(4): 392– 406. Flemming, H.C., Schaule, G., Griebe, T., Schmitt, J. and Tamachkiarowa, A., 1997, Biofouling—the achilles heel of membrane processes, Desalination, 113: 215–225. Flemming, H.C., 2002, Biofouling in water systems-cases, causes, and countermeasures, Appl Microbiol Biotechnol, 56(6): 629– 640.
Trans IChemE, Part C, Food and Bioproducts Processing, 2007, 85(C1): 42–48
48
DOUME`CHE et al.
Her, N., Amy, G. and Jarusutthirak, C., 2000, Seasonal variations of nanofiltration (NF) foulants: identification and control, Proceedings of the Conference in Membranes in Drinking and Industrial Water Production, 2: 257–274. Howe, K.J., Ishida, K.P. and Clark, M.M., 2002, Use of ATR/FTIR spectrometry to study fouling of microfiltration membranes by natural waters, Desalination, 147: 251– 255. James, B.J., Jing, Y. and Chen, X.D., 2003, Membrane fouling during filtration of milk—a microstructural study, J Food Engineering, 60(4): 431–437. Kimura, K., Hane, Y., Watanabe, Y., Amy, G. and Ohkuma, N., 2004, Irreversible membrane fouling during ultrafiltration of surface water, Water Res, 38: 3431– 3441. Lawrence, J.R., Chenier, M.R., Roy, R., Beaumier, D., Fortin, N., Swerhone, G.D.W., Neu, T.R. and Greer, C.W., 2004, Microscale and molecular assessment of impacts of nickel, nutrients, and oxygen level on structure and function of river biofilm communities, Appl Environ Microbiol, 70: 4326– 4339. Mittelman, M.W., 1998, Structure and functional characteristics of bacterial biofilms in fluid processing operations, J Dairy Sci, 81(10): 2760–2764. Neu, T.R., Swerhone, G.D.W. and Lawrence, J.R., 2001, Assessment of lectin-binding analysis for in situ detection of glycoconjugates in biofilm systems, Microbiology, 147: 299–313. Nivens, D.E., Chambers, J.Q., Anderson, T.R., Tunlid, A., Smit, J. and White, D.C., 1993, Monitoring microbial adhesion and biofilm formation by attenuated total reflection/Fourier transform infrared spectroscopy, J Microbiol Methods, 17: 199 –213. Rabiller-Baudry, M., Le Maux, M., Chaufer, B. and Begoin, L., 2002, Characterisation of cleaned and fouled membrane by ATR-FTIR and EDX analysis coupled with SEM: application to UF of skimmed milk with a PES membrane, Desalination, 146: 123– 128. Scha¨fer, A.I., Fane, A.G. and Waite, T.D., 2000, Fouling effects on rejection in the membrane filtration of natural waters, Desalination, 131: 215– 224.
Shadpour, H., Musyimi, H., Chen, J. and Soper, S.A., 2006, Physiochemical properties of various polymer substrates and their effects on microchip electrophoresis performance, J Chromatography, 1111: 238–251. Steinberger, R.E. and Holden, P.A., 2005, Extracellular DNA in single- and multiple-species unsaturated biofilms, Appl Env Microbiol, 71(9): 5404–5410. Suci, P.A., Siedlecki, K.J., Palmer, Jr, R.J., White, D.C. and Geesey, G.G., 1997, Combined light microscopy and attenuated total reflection fourier transform infrared spectroscopy for integration of biofilm structure, distribution, and chemistry at solid-liquid interfaces, Appl Environ Microbiol, 63(11): 4600– 4603. Verran, J., Boyd, R.D., Hall, K. and West, R.H., 2001, Microbiological and chemical analyses of stainless steel and ceramics subjected to repeated soiling and cleaning treatments, J Food Prot, 64(9): 1377–1387. Verran, J., 2002, Biofouling in food processing: biofilm or biotransfer potential? Trans IChemE, 80C: 292 –298. Violleau, D., Essis-Tome, H., Habarou, H., Croue´, J.P. and Pontie´, M., 2005, Fouling studies of a polyamide nanofiltration membrane by selected natural organic matter: an analytical approach, Desalination, 173: 223– 238 Whitchurch, C.B., Tolker-Nielsen, T., Ragas, P.C. and Mattick, J.S., 2002, Extracellular DNA required for bacterial biofilm formation, Science, 295: 1487. Xu, P., Drewes, J.E., Kim, T.U., Bellona, C. and Amy, G., 2006, Effect of membrane fouling on transport of organic contaminants in NF/ RO membrane applications, J Memb Science, 279: 165– 175.
ACKNOWLEDGEMENT The authors thank Solange Salzard for the access to the ATR-FTIR spectrophotometer. The manuscript was received 24 April 2006 and accepted for publication after revision 1 September 2006.
Trans IChemE, Part C, Food and Bioproducts Processing, 2007, 85(C1): 42–48
Laboratoire ERRMECe (EA1391), Universite´ de Cergy-Pontoise, Pontoise, France. Plate-forme re´gionale de recherche en imagerie cellulaire, INSERM U413 – IFRMP 23, Universite´ de Rouen, Mont-Saint-Aignan, France.
2
Abstract: We characterized the composition and organization of membrane foulants in a French drinking water production plant between March and September 2005 by the combination of Attenuated Total Reflection Fourier Transform Infrared (ATR-FTIR) spectroscopy, lectin-binding analysis, epifluorescence microscopy and Confocal Laser Scanning Microscopy (CLSM). The comparison of ATR-FTIR spectra of clean and fouled membranes revealed a qualitatively homogenous deposition of biological matter onto the membrane surface, a high heterogeneity in the composition of the fouling material and a high temporal stability of this composition. Characteristic signals of mainly polysaccharides (900 –1200 cm21) and proteins (1300 –1700 cm21) were observed. Fluorescence microscopy observations after nucleic acid staining with DAPI and polysaccharides staining with lectins labelled with fluorescein isothiocyanate or tetramethylrhodamine isothiocyanate indicated a high spatial heterogeneity inside the foulant matter. The microbial cells, mainly composed of bacteria, were localized in the superficial layer of the fouling material and were organized as microcolonies interspersed at the membrane surface. Polysaccharide residues were found in areas where microcolonies were present and in areas devoid of microbial cells. High staining with Peanut and Bandeiraea simplicifolia agglutinins (PNA, BS-1, respectively) revealed high occurrence of galactosides residues in the polysaccharide components of the foulants. The BS-1 lectin staining pattern indicated a high degree of spatial organization with the observation of long and entangled fibres. Wheat germ agglutinin staining showed short fibres and cloud stained areas. PNA and Concanavalin A lectin staining were more interspersed. In conclusion, the combination of ATR-FTIR spectroscopy with lectin-binding analysis was a very informative tool for the characterization (composition and spatial organization) of the foulants that cover the surface of the membranes. Correspondence to: Dr P. Di Martino, Laboratoire ERRMECe UFR Sciences et Techniques Universite´ de CergyPontoise, 2 avenue Adolphe Chauvin BP222 – 95302 Pontoise cedex France. E-mail: patrick.di_martino@ u-cergy.fr
DOI: 10.1205/fbp06020 0960–3085/07/ $30.00 þ 0.00 Food and Bioproducts Processing Trans IChemE, Part C, March 2007 # 2007 Institution of Chemical Engineers
Keywords: fouling; biofilm; membrane; water; lectin
INTRODUCTION
leading to biofilm development, referred to as biofouling, is a major component of membrane fouling in water separation applications (Bos et al., 1999). In drinking water plants, biofouling leads to higher operating pressures, increases the frequency of chemical cleanings, and leads to membrane deterioration (Flemming, 2002). Moreover, biofilms and other materials accumulated on the membrane surface, which cannot be removed by cross flow, backflushing, or backpulsing, can result in permanent permeability loss. One characteristic of a biofilm is the production of an extracellular matrix that envelops the attached cells (Allison, 2003). This viscous layer is generally composed of water and microbial polymers and provides a complex array of microenvironments surrounding the attached cells. These extracellular biopolymers usually include exopolysaccharides, proteins, nucleic acids, and
Fouling is a commonly encountered problem during food and drink processing (Changani et al., 1997; James et al., 2003; Mittelman, 1998; Scha¨fer et al., 2000). Membrane fouling can result from the accumulation of multivalent ions, organic matter and microbial cells in suspension and from the development of biomass at the membrane surface. This matter accumulation leads to a decrease of the process efficiency and then to an increase in the production cost. In the drinking water industry, membrane fouling is known to depend on several parameters such as the composition of the feed water, the physicalchemical properties of membranes and also the hydrodynamic conditions (AWWA Membrane Technology Research Committee, 2005). Irreversible adherence and growth of microbial cells on the membrane surface 42
Vol 85 (C1) 42–48
MEMBRANE FOULANTS CHARACTERIZATION IN A DRINKING WATER PRODUCTION UNIT phospholipids. The exopolysaccharides (EPS) support the membrane surface colonization by microbial cells, facilitate the arrangement of different species within the biofilm, and protect the sessile community against toxic compounds in solution. Moreover, polysaccharide organic matter is responsible for the evolution of irreversible fouling (Kimura et al., 2004). A better knowledge of the fouling material composition and organisation can help in the development of strategies for fouling prevention and control (Flemming et al., 1997; Verran et al., 2001; Violleau et al., 2005; Xu et al., 2006). There are many methods described for assessing the fouling of surfaces by microorganisms or by organic material but a relatively simple method for assessment of the components in combination is not readily available (Verran, 2002). Some infrared spectroscopic analyses can be used to elucidate the nature of foulants of predominantly organic and biological origin (Her et al., 2000; Howe et al., 2002; Kimura et al., 2004; Nivens et al., 1993; Rabiller-Baudry et al., 2002; Suci et al., 1997). Fluorescently labelled lectins have also been described in combination with confocal laser scanning microscopy (CLSM) to allow the visualization and characterization of carbohydrate-containing extracellular polymers in biofilms (Neu et al., 2001). A combination of infrared spectroscopy and fluorescence microscopy associated with fluorescently labelled lectins has never been used before to investigate membrane fouling. The primary objective of this research was to identify the major components of the fouling material from filtration membrane surfaces and to determine their seasonal variations. We performed three autopsies of filtration membranes in March, June and September 2005, respectively, and analysed the foulant matter by the combination of attenuated total reflection Fourier transform infrared (ATR-FTIR) spectroscopy, epifluorescence and CLSM observations.
METHODS AND MATERIALS Membrane Autopsy The membranes used in the plant were composed of an ultra thin top layer made of polypiperazine amide and a polysulfone microporous support. Membranes were in service for at least 5 years before our experiments. In March, June and September 2005, some membrane modules were extracted from the drinking water plant, partially drained of excess liquid, disassembled, stored at 48C and analysed within 24 h in the laboratory. Some samples were air-dried and analysed by ATR-FTIR and other wet samples were fixed in paraformaldehyde (PFA) before staining and fluorescence microscopy analysis, as described below.
Analysis of Membrane Foulants by ATR-FTIR Samples of air-dried fouled membrane were analysed by ATR-FTIR. ATR-FTIR spectra were recorded using a Tensor 27 IR spectrophotometer with a 458 diamond/ZeSe flat plate crystal and an average depth penetration of 2 mm. Each spectrum presented is the result of 32 scans obtained with a resolution of 2 cm21 with air as the background.
Lectin Staining of Foulants Samples of the fouled membranes were treated with paraformaldehyde (4%, v/v) for 60 min prior to lectin application in
43
order to fix the foulant matter. Fluorescently labelled lectins (Sigma, Saint Quentin Fallavier, France) conjugated with fluorescein isothiocyanate (FITC) or tetramethylrhodamine isothiocyanate (TRITC) were used for the study of their binding to polysaccharides contained in the fouling layer (Table 1). Stock solutions of lectins (1 mg ml21) were prepared by adding filter-sterilized (0.2 mm) water and stored frozen. Prior to use, a thawed portion was diluted with filter-sterilized water and centrifuged to remove crystals and debris. Double staining with two lectins, one FITC- and one TRITC-labelled were done with a mixture of Bandeiraea simplicifolia agglutinin (BS-1) and concanavalin A (ConA) or a mixture of peanut agglutinin (PNA) and wheat germ agglutinin (WGA). 100 ml of a mixture of two lectins (final concentration of 100 mg ml21 each) were carefully applied directly on top of the membrane. After incubation for 30 min in the dark at room temperature, unbound lectins were removed by washing with filter-sterilized water. The fluorescent DNA-binding stain 40 ,6-diamidino-2phenyindole dihydrochloride (Sigma, Saint Quentin Fallavier, France) was used to visualize cell distribution in combination with lectin staining. Following the lectin staining step, the fouled membranes were treated with 100 ml of a solution of DAPI (1 mg l21) in filter-sterilized water. After incubation for 30 min in the dark at room temperature, unbound DAPI was removed by washing the membrane surface with filtersterilized water. The stained preparations were then mounted with Mowiol (Calbiochem, Meudon, France) and stored at 48C in the dark. For the negative control, the lectins were replaced by filter-sterilized water.
Microscopy and Image Analysis The fouled membranes stained with DAPI and fluorescently labelled lectins were examined with a Leica epifluorescence microscope (MPS 60) and with a Leica SP2 upright confocal laser scanning microscope (DM RAX-UV) equipped with the Acousto-Optical Beam Splitter (AOBS) system and using 63, N.A. 1.32, oil immersion objective (Leica microsystems, Rueil-Malmaison, France). For epifluorescence images, DAPI was excited at 364 nm, FITC was excited at 494 nm, and TRITC was excited at 550 nm. For confocal images, DAPI was excited at 405 nm and observed from 410 to 600 nm, FITC was excited at 488 nm and observed from 505 to 540 nm, TRITC was excited at 543 nm and observed from 560 to 600 nm. The selection of spectral emission window for each fluorophore has been determined through l scan analysis on single stained membrane fragments. The gain and offset for each photomultiplier have been adjusted to optimize nucleic acid and lectins detection.
Table 1. Carbohydrate-binding specificities of lectins employed for staining of membrane foulants. Lectin (abbreviation)
Conjugate
Main specificity
Bandeiraea simplicifolia agglutinin (BS-1) Concanavalin A (ConA) Peanut agglutinin (PNA) Wheat germ agglutinin (WGA)
TRITC
a-gal, a-galNAc
FITC TRITC FITC
a-man, a-glc b-gal(1!3)galNAc (glcNAc)2, NeuNAc
TRITC, tetramethylrhodamine isothiocyanate; FITC, fluoresceinisothio cyanate.
Trans IChemE, Part C, Food and Bioproducts Processing, 2007, 85(C1): 42–48
44
DOUME`CHE et al.
Images of the CLSM observations (1024 1024 pixels) have been acquired through a sequential mode, between stacks, to exclude spectral crosstalk from the data. 400 Hz scan speed was used and signal/noise ratio has been increased through frame average. Overlay and maximum projection of the z-stacks files have been performed with post acquisition Leica Confocal Software (LCS) functions to obtain the presenting snapshots. Original z-stack Leica files have then been imported into Imaris 4.0 software (Bitplane AG, Zu¨rich) to obtain snapshots illustrating xz representation and 3D modelling. Geometric shapes have been used to represent the microbial microcolonies and the biofilm matrix inside the foulant matter in 3D.
RESULTS, ANALYSES AND DISCUSSION Analysis of Membrane Foulants by ATR-FTIR The FTIR spectra of the clean membrane and fouled membranes extracted in March, June and September 2005 are compared in Figure 1. Different spectra illustrating heterogeneity among the foulant layer are presented for each fouled membrane. Several spectral features were observed in the 1700 –900 cm21 range and in the 1700–700 cm21 range for the fouled and clean membranes, respectively. For the fouled membranes, IR bands indicative of biomass, as previously defined (Nivens et al., 1993), were detected near 1650 cm21 (amide I), 1550 cm21 (amide II), 1450 cm21 (due in part to C-H deformation), 1400 cm21 (due in part to symetric stretch for the carboxylate ion), and 1250 cm21
(P5 5O and C2 2O2 2C stretching and/or amide III). The broad complex region from 1250 to 900 cm21 has been characterized in the literature as the region where both DNA-RNA and polysaccharides appear (Suci et al., 1997). Vibrations in this region are characterized by C2 2O2 2C, C2 2O, ringstretching vibrations and the P5 5O stretch of phosphodiesters. The FTIR profiles were very similar for the fouled membranes analysed at the three different seasons. Thus, the foulant matter was overall stable in nature between March and September. None of the peaks of the clean and new membrane were observed in the spectra obtained in March or June, indicating the total coverage of the membrane by the fouling layer. For the fouled membrane extracted in September, IR bands corresponding to the membrane were detected near 700, 850, 1150 and 1250 cm21 for one of the four presented spectra, thus underlining local heterogeneity within the fouling layer [Figure 1(c); dotted spectrum].
Analysis of Membrane Foulant by Lectin Staining and Fluorescence Microscopy Polymeric membranes can be subjected to autofluorescence emission after excitation at the wavelengths used in histology. The background level of fluorescence from polymers depends on their content in aromatic structures and conjugated double bonds in their molecular structure (Shadpour et al., 2006). As shown in Figure 2, the background level of autofluorescence of the filtration membranes extracted from the plant did not prevent observing the
Figure 1. ATR-FTIR spectra of fouled membranes in March (a), June (b) and September (c). The different spectra observed at one time differ from their localization onto the membrane surface. The spectra of clean membrane are given as reference (d). Trans IChemE, Part C, Food and Bioproducts Processing, 2007, 85(C1): 42–48
MEMBRANE FOULANTS CHARACTERIZATION IN A DRINKING WATER PRODUCTION UNIT
45
Figure 2. Interactions between fluorescently-labelled lectins and fouled membranes, as visualized by epifluorescence microscopy. Shown are fouled membranes extracted in March [(a) to (f)], June [(g) to (l)) and September [(m) to (r)] 2005, stained with: DAPI þ PNA-TRITC þ WGA-FITC [(a), (b), (c), (g), (h), (i), (m), (n), (o)]; DAPI þ BS-1-TRITC þ ConA-FITC [(d), (e), (f), (j), (k), (l), (p), (q), (r)]. The panels coloured in blue [(a), (d), (g), (j), (m), (p)], green [(b), (e), (h), (k), (n), (q)] and red [(c), (f), (i), (l), (o), (r)] correspond to observations after illumination at 364, 494, and 540 nm respectively. Magnification,630.
Trans IChemE, Part C, Food and Bioproducts Processing, 2007, 85(C1): 42–48
46
DOUME`CHE et al.
microbial cells and exopolysaccharides after staining with DAPI and fluorophore-conjugated lectins. Epifluorescence observations after DAPI staining showed that the microbial part of the fouling layer was mainly composed of bacteria organized as microcolonies interspersed at the membrane surface [Figure 2(a), (d), (g), (j), (m), (p)]. Some algae were also present, as shown by autofluorescence properties of these cells. We did not observe any extracellular DNA (eDNA) stained with DAPI. The eDNA has been previously shown to be involved in biofilm stability during the early stages of biofilm growth in vitro (Whitchurch et al., 2002). The eDNA abundance inside biofilms has been shown to be strain dependent and the accumulation of eDNA in dualspecies biofilms is unpredictable (Steinberger and Holden, 2005). Moreover, the presence of eDNA in natural and industrial multiple-species environments is still unknown. The biomass attached at the membrane surface increased from March to June and seemed to be unchanged from June to September. The application of lectin-binding-analysis to define glycoconjugate composition of the foulants was used. As shown in Figure 2, epifluorescence observations revealed two types of polymeric carbohydrate structures: those located on cell surfaces and those located extracellularly throughout the biofilm matrix. Lectin staining was not only concentrated in areas where microcolonies were present but also extended in areas devoid of microbial cells. This
binding pattern of the lectins has been previously observed with environmental biofilms grown in vitro with river water as the sole source of carbon and nutrients (Neu et al., 2001). During the study, the dominant lectin binding was with BS-1 revealing high occurrence of galactosides residues in the polysaccharide part of the foulants. Nevertheless, the polysaccharide composition of the fouling layer changed from March to September. Lectin staining increased from March to September for all the lectins used. The lectin-binding changes with time may be linked to an increase of the biomass attached at the membrane surface and to changes among the populations of attached cells. Staining with BS-1 increased constantly in March, June and September [Figure 2(f), (l), and (r)]. A high increase of binding with PNA, and ConA was observed between March and June, but the binding of these two lectins did not change between June and September [Figure 2(c), (i), (o), and (e), (k), (q), respectively]. Staining with the WGA was weak in March and June and was higher in September [Figure 2(b), (h), and (n)]. Nutrients, oxygen level and the concentration of metals have been shown to influence the exopolymer abundance of environmental model biofilms grown in vitro with river water as the sole source of carbon and nutrients (Lawrence et al., 2004). The modification of these parameters has been shown to lead to a shift in the glycoconjugate makeup of the biofilms. In our study, the glycoconjugate composition of the foulants at the membrane
Figure 3. Interactions between fluorescently-labelled lectins and fouled membranes, as visualized by CLSM. Shown are fouled membranes extracted in June stained with: DAPI þ PNA-TRITC þ WGA-FITC (a); DAPI þ BS-1-TRITC þ ConA-FITC (b). Colour allocation: blue—DAPI, green—FITC-labelled lectin and red—TRITC-labelled lectin. (1) Magnification 630. (2) Enlargement of the dotted square zones. White arrows indicate the orientation of the foulant matter from the bottom to the top. Trans IChemE, Part C, Food and Bioproducts Processing, 2007, 85(C1): 42–48
MEMBRANE FOULANTS CHARACTERIZATION IN A DRINKING WATER PRODUCTION UNIT
47
Figure 4. Three-dimensional representation of the surface of a fouled membrane stained with DAPI þ PNA-TRITC þ WGA-FITC obtained with the Imaris 4.0 software. Shown are the entire fields corresponding to the observation of the membrane at a magnification of 630 (a), and the enlargement of the dotted squared part of the field (b). DAPI, TRITC, and FITC fluorescence appears in blue, red and green, respectively.
surface was relatively stable. Further studies are needed to determine whether this stability of biofilm is linked to a relative stability of nutrients, oxygen level and metals in the surface water of the river. Confocal laser scanning microscopy was used to study the spatial organization and distribution of the fouling layer at the membrane surface. Multiple stains were applied and recorded simultaneously: DAPI þ PNA-TRITC þ WGA-FITC and DAPI þ BS-1-TRITC þ ConA-FITC. Examples of these multiple staining experiments are shown in Figure 3(a) and (b), respectively. The post acquisition Leica Confocal Software (LCS) allowed the elimination of the membrane autofluorescence background. CLSM analysis confirmed that most of the analysed polysaccharides (stained with the lectins used) were located extracellularly throughout the biofilm matrix. A high degree of spatial organization and high heterogeneity were observed within the fouling material: WGA binding showed short fibres and cloud stained areas; BS-1 binding revealed long and entangled fibres; PNA and ConA staining were more dispersed. Confocal images showing the Z-axis revealed heterogeneity in depth: the stained part of the fouling matter varied in depth from 6 to 27 mm (Figure 4). The microbial cells, mainly organized as microcolonies interspersed at the membrane surface, were localized in the superficial layer of the fouling deposit (Figure 4). These staining patterns suggested that the fouling layer was organized as an interpenetrated network mainly composed of exopolysaccharides supporting microbial cells adhesion.
CONCLUSION In conclusion, the combination of ATR-FTIR spectroscopy with DAPI staining and lectin-binding-analysis was efficient to study the composition, the spatial organization and the seasonal variations of the fouling material present at the
membrane surface used for drinking water production. Data collected by this combined approach gave important new information about the membrane foulants, extending previous observations of foulant speciation related to membrane filtration efficiency (Violleau et al., 2005; Xu et al., 2006). Epifluorescence microscopy allowed determining the qualitative and quantitative evolutions of the foulant matter. The more sophisticated and expensive CLSM gave the opportunity to observe the spatial organization and 3D structure of the foulants. The fouling material exhibited the same characteristics as a biofilm. These biofilms developed at the surface of membranes in a drinking water plant may result from the presence of microbial cells and nutrients coming from feed water and also from material that has not been removed even after successive cleaning and production processes. A better knowledge of the composition and seasonal variations of the foulants will help us in the development of more efficient cleaning processes of membranes.
REFERENCES Allison, D.G., 2003, The biofilm matrix, Biofouling, 19(2): 139 –150. AWWA Membrane Technology Research Committee, 2005, Recent advances and research needs in membrane fouling, Journal AWWA, 97(8): 79–89. Bos, R., Van Der Mei, H.C. and Busscher, H.J., 1999, Physicochemistry of initial microbial adhesive interactions—its mechanisms and methods for study, FEMS Microbiol Rev, 23: 179– 230. Changani, S.D., Belmar-Beiny, M.T. and Fryer, P.J., 1997, Engineering and chemical factors associated with fouling and cleaning in milk processing, Exp Thermal Fluide Sci, 14(4): 392– 406. Flemming, H.C., Schaule, G., Griebe, T., Schmitt, J. and Tamachkiarowa, A., 1997, Biofouling—the achilles heel of membrane processes, Desalination, 113: 215–225. Flemming, H.C., 2002, Biofouling in water systems-cases, causes, and countermeasures, Appl Microbiol Biotechnol, 56(6): 629– 640.
Trans IChemE, Part C, Food and Bioproducts Processing, 2007, 85(C1): 42–48
48
DOUME`CHE et al.
Her, N., Amy, G. and Jarusutthirak, C., 2000, Seasonal variations of nanofiltration (NF) foulants: identification and control, Proceedings of the Conference in Membranes in Drinking and Industrial Water Production, 2: 257–274. Howe, K.J., Ishida, K.P. and Clark, M.M., 2002, Use of ATR/FTIR spectrometry to study fouling of microfiltration membranes by natural waters, Desalination, 147: 251– 255. James, B.J., Jing, Y. and Chen, X.D., 2003, Membrane fouling during filtration of milk—a microstructural study, J Food Engineering, 60(4): 431–437. Kimura, K., Hane, Y., Watanabe, Y., Amy, G. and Ohkuma, N., 2004, Irreversible membrane fouling during ultrafiltration of surface water, Water Res, 38: 3431– 3441. Lawrence, J.R., Chenier, M.R., Roy, R., Beaumier, D., Fortin, N., Swerhone, G.D.W., Neu, T.R. and Greer, C.W., 2004, Microscale and molecular assessment of impacts of nickel, nutrients, and oxygen level on structure and function of river biofilm communities, Appl Environ Microbiol, 70: 4326– 4339. Mittelman, M.W., 1998, Structure and functional characteristics of bacterial biofilms in fluid processing operations, J Dairy Sci, 81(10): 2760–2764. Neu, T.R., Swerhone, G.D.W. and Lawrence, J.R., 2001, Assessment of lectin-binding analysis for in situ detection of glycoconjugates in biofilm systems, Microbiology, 147: 299–313. Nivens, D.E., Chambers, J.Q., Anderson, T.R., Tunlid, A., Smit, J. and White, D.C., 1993, Monitoring microbial adhesion and biofilm formation by attenuated total reflection/Fourier transform infrared spectroscopy, J Microbiol Methods, 17: 199 –213. Rabiller-Baudry, M., Le Maux, M., Chaufer, B. and Begoin, L., 2002, Characterisation of cleaned and fouled membrane by ATR-FTIR and EDX analysis coupled with SEM: application to UF of skimmed milk with a PES membrane, Desalination, 146: 123– 128. Scha¨fer, A.I., Fane, A.G. and Waite, T.D., 2000, Fouling effects on rejection in the membrane filtration of natural waters, Desalination, 131: 215– 224.
Shadpour, H., Musyimi, H., Chen, J. and Soper, S.A., 2006, Physiochemical properties of various polymer substrates and their effects on microchip electrophoresis performance, J Chromatography, 1111: 238–251. Steinberger, R.E. and Holden, P.A., 2005, Extracellular DNA in single- and multiple-species unsaturated biofilms, Appl Env Microbiol, 71(9): 5404–5410. Suci, P.A., Siedlecki, K.J., Palmer, Jr, R.J., White, D.C. and Geesey, G.G., 1997, Combined light microscopy and attenuated total reflection fourier transform infrared spectroscopy for integration of biofilm structure, distribution, and chemistry at solid-liquid interfaces, Appl Environ Microbiol, 63(11): 4600– 4603. Verran, J., Boyd, R.D., Hall, K. and West, R.H., 2001, Microbiological and chemical analyses of stainless steel and ceramics subjected to repeated soiling and cleaning treatments, J Food Prot, 64(9): 1377–1387. Verran, J., 2002, Biofouling in food processing: biofilm or biotransfer potential? Trans IChemE, 80C: 292 –298. Violleau, D., Essis-Tome, H., Habarou, H., Croue´, J.P. and Pontie´, M., 2005, Fouling studies of a polyamide nanofiltration membrane by selected natural organic matter: an analytical approach, Desalination, 173: 223– 238 Whitchurch, C.B., Tolker-Nielsen, T., Ragas, P.C. and Mattick, J.S., 2002, Extracellular DNA required for bacterial biofilm formation, Science, 295: 1487. Xu, P., Drewes, J.E., Kim, T.U., Bellona, C. and Amy, G., 2006, Effect of membrane fouling on transport of organic contaminants in NF/ RO membrane applications, J Memb Science, 279: 165– 175.
ACKNOWLEDGEMENT The authors thank Solange Salzard for the access to the ATR-FTIR spectrophotometer. The manuscript was received 24 April 2006 and accepted for publication after revision 1 September 2006.
Trans IChemE, Part C, Food and Bioproducts Processing, 2007, 85(C1): 42–48