Desalination 316 (2013) 85–90

Contents lists available at SciVerse ScienceDirect

Desalination

j ourna l homepage: www.e lsev ie r .com/ locate /desa l

The impact of cell metabolic activity on biofilm formation and flux decline duringcross-flow filtration of ultrafiltration membranes

Amir Mohaghegh Motlagh, Santosh Pant, Cyndee Gruden ⁎Department of Civil Engineering, The University of Toledo, 2801 W Bancroft St, Toledo OH 43606, USA

H I G H L I G H T S

► Physiological state of biofoulant impacted biofoulant activity.► Biofoulant activity level did not correlate to biofilm activity level.► Biofilm activity level was not a good indicator of flux decline.► Stationary phase samples had more and unique EPS.► Inactive cells resulted in the least amount of flux decline.

⁎ Corresponding author. Tel.: +1 419 530 8128; fax:E-mail address: cgruden@eng.utoledo.edu (C. Grude

0011-9164/$ – see front matter © 2013 Elsevier B.V. Allhttp://dx.doi.org/10.1016/j.desal.2013.02.001

a b s t r a c t

a r t i c l e i n f o

Article history:Received 30 October 2012Received in revised form 28 January 2013Accepted 1 February 2013Available online 1 March 2013

Keywords:Metabolic activityAdenosine triphosphateUltrafiltration membraneFluxBiofouling

One significant challenge to membrane filtration technologies is membrane fouling causing pressure drop,flux decline and eventually significant cost of membrane replacement. The objective of this research was todetermine the impact of metabolic activity as measured in adenosine triphosphate (ATP) concentration ofthe pure culture of biofoulants on the membrane biofilm metabolic activity, biofilm formation rate, and op-erational flux decline. Our results showed that after 10–12 h of filtration, the biofilm ATP levels reach anequilibrium concentration (avg. 8 amol/cell) and do not appear to be related to biofoulant ATP levels fromcells harvested in the late exponential growth phase regardless of initial ATP level. However, the bacterialgrowth phase affected the ATP activity of cells and membrane biofilms formed from biofoulants in the lagand stationary phases of growth contained similar levels of activity, and the exponential phase cells resultedin significant higher activity. Flux decline does not appear to be related to metabolic activity of the biofoulantor biofilm following 24 h of filtration but notably, there was much less flux decline when the biofoulant cellswere inactive.

© 2013 Elsevier B.V. All rights reserved.

1. Introduction

Membrane separation is an attractive option for drinking waterproduction since it produces biologically safe and consistently highquality drinking water and is more sustainable compared to conven-tional water treatment due to fewer additions of chemicals to rawwater in the water treatment process and a smaller process footprint[1]. However, a serious impediment to more widespread implemen-tation is membrane fouling, causing a significant decrease in fluxwith a corresponding increase in operational energy costs to maintainthe design flow rate and, ultimately, membrane replacement. Amongthe different types of fouling (i.e. crystalline fouling, organic fouling,colloidal fouling, microbiological fouling), biofouling is considered amajor problem in membrane separation systems [2] since othertypes of fouling can be controlled by reduction of foulant concentra-tion in the feed water [3]. Biofouling is hard to control due to the

+1 419 530 8116.n).

rights reserved.

ability of microorganisms present even at very small concentrationsin nearly all water systems to colonize almost any surface and to sur-vive under extreme conditions [4]. Furthermore, compared to plank-tonic cells, bacteria embedded in a biofilm are up to 1000-times moreresistant to conventional approaches to mitigating biofouling includ-ing disinfection [5].

In order to control and develop effective anti-biofouling strategies inmembrane filtration, it is essential to understand the mechanisms ofbiofouling. This work focuses on early stages of biofilm formationsince the first step or immediate attachment of bacteria onto the mem-brane surface is a reversible process [6]. Thereafter, the biofilm growthprocess is dominated by the production of extracellular polymeric sub-stances (EPS)whichmakes the attachmentmore firmly adhered and ir-reversible [7]. The hypothesis in this study is that the activity ofplanktonic cells in the feed water (biofoulant) will affect membranebiofilm cell activity, EPS formation, and consequently membrane per-formance (i.e. flux decline) in early stage biofilm formation. The activityof the biofoulant is likely to directly impact not only cellular activity inthe biofilm but also the amount and composition of extracellular

86 A.M. Motlagh et al. / Desalination 316 (2013) 85–90

polymeric substances (EPS) formed [8,9], and consequently the mem-brane flux [10].

Adenosine triphosphate (ATP) has been used to determine the in-tracellular metabolic activity of biomass [11]. ATP is a fundamentalbiomolecule present in all viable cells whether their metabolism isaerobic or anaerobic and is degraded immediately after cell death[12]. Furthermore, ATP is stable in many studied organisms formingbiomass [13]. Prediction of biofilm production can be achieved byanalysis of biofilm formation rate [14], which is based on ATP analysisof the biofilm in monitoring biofilm and extraction and measurementof biomass ATP activity [15]. The quantity of ATP depends on thephysiological state of the organisms and consequently on the biomassgrowth rate [16,17].

Therefore, the objective of this studywas to investigate the effects ofbiofoulant activity in ultrafiltration membrane biofouling and the con-sequent effects on permeate flux. Short-term biofouling experimentswith a mono-culture biofilm were carried out using a laboratory-scaleultrafiltration membrane filtration test unit and the activity of thepure culture of biofoulant, feed water, and biofilm accumulated on themembrane surface were investigated with ATP measurement. In addi-tion, the knowledge of the metabolic activity of the membrane biofilmmay assist in understanding the biofilm accumulation mechanisms toapply appropriate countermeasures and control ofmembrane biofouling.

2. Material and methods

2.1. Cross-flow filtration experiments

Cellulose acetate ultrafiltration (UF) membranes with a molecularweight cut-off of 20,000 Da (General Electric water and process technol-ogy,Minnetonka,MN)were used for the cross-flow filtration. Cross-flowexperiments were carried out using membrane sheets mounted on thecell membrane (Osmonics Sepa CF, Minnetonka, MN) with an effectivemembrane area of 155 cm2. A schematic diagram of the membrane fil-tration system has been shown in Fig. 1 [18].

The feedwaterwas circulated from the reservoir to the cellmembraneand both the retentate (concentrate) and permeate outlets were recycledto the reservoir. The cross-flow experiment was performed for differentfiltration durations (4, 11, and 24 h) to understand changes in the activityof the formed biofilm (biofilm activity) during the first day of filtration.Biofoulant activity, defined as the cellular activity of model organismspresent in the feed solution, was determined by collecting samples ofthe feed water following inoculation with bacteria at the beginning ofthe experiment. In addition, in order to represent a variety of possiblegrowth conditions and physiological states of biofoulants that couldbe present in a real membrane-based water treatment system, the

Fig. 1. Diagram of cross-flow membrane filtration system.Modified from Good et al. [18].

cross-flow filtration was carried out in 24 h with biofoulant in differentgrowth phases (i.e. lag, late exponential and stationary phases).

Cross-flow filtration was operated at a constant pressure of172.36 kPa (25 psi) and a constant water temperature of 27 °C usinga cooling fan. All membranes were pre-compacted for 12 h to reach aconstant flux before starting the filtration period. The permeate flux(J) wasmeasured during the filtration experiment and the final perme-ate fluxwas evaluated as normalizedfluxwith respect to initialflux (J0).The initial flux for all of the experiments was adjusted to 26 L/m2-h. Allof the cross-flowfiltration experimentswere carried out using syntheticfeed water composed of deionized (DI) water and buffering chemicalsas described in Zaky et al. [19]. During the filtration experiments usinginactive bacteria, 0.2 mMsodiumazide (NaN3)was added insteadof so-dium chloride to prevent bacterial cell growth [10]. In this study, bio-fouling was investigated under consistent biological (i.e. dissolvedorganics, type of carbon and energy source); physical (i.e. cross-flowve-locity, applied pressure, temperature, initial permeate flux); and chem-ical (i.e. membrane type, pH, ionic strength) conditions and the effectsof cell deposition and biofilm growth on themembranewere character-izedwith different levels of biofoulant activity. Cell concentrations in thefeed water and in the biofilm following 24 h of filtration as a function ofbiofoulant ATP activity were compared using spectrophotometery(OD600). It should be noted that all of the experiments includingthe biofilm activity in different filtration durations, growth curveand ATP standard curve (data not shown) were carried out in tripli-cate set of tests to have consistent results.

2.2. Bacterial strain and growth conditions

A Gram-negative aerobic bacterium, Pseudomonas fluorescensMigula(ATCC # 12842) was used as the microorganism to produce the mem-brane biofouling. This rod-shaped bacterium has the optimal growthtemperature of 27 °C, pH 7.0, and use glucose as its carbon source.Mem-bers of Pseudomonas genus are one of the most ubiquitous bacterial spe-cies in the environment and water systems and P. fluorescens is wellknown to be good biofilm producer due to its short generation timeand resistance to temperature fluctuation [20,21]. The bacterial growthcurve (cell counts vs. incubation time) was established at 27 °C inLuria–Bertani (LB) broth to determine the lag, exponential and station-ary growth phases of the organism.

2.3. Cellular ATP determination

The intracellular adenosine triphosphate (ATP) concentration wasmeasured using an ATP colorimetric/fluorometric assay kit (Biovision,Mountain View, CA) as the kit was previously reported [22,23]. Tomeasure the amount of ATP, cell solutions were lysed in the assaybuffer based on the cell concentration (1×106 cells in 100 μL of ATPassay buffer). ATP measurements were performed by extracting50 μL of the sample and adding 50 μL of the ATP extract mixture, in-cubating at room temperature in the dark for 30 min and the amountof light produced was measured by the spectrophotometer in relativefluorescence units (RFU) which is proportional to the amount of ATP[24].

The ATP activity and cell numberweremeasured as a function of incu-bation time. Growth phaseswere determined as follows: in the beginningof lag phase (t=2 h), beginning of exponential phase (t=9 h), mid-lateof exponential phase which is the time that the cells were used in themembrane filtration experiments (t=21 h), the transition between theexponential phase and the start of stationary phase (t=27 h), and inthe middle of the stationary phase (t=36 h). ATP concentrations rangedfrom 4 amol/cell in the lag phase, up to 12 amol/cell in the late exponen-tial growth phase, and then reducing to approximately 6 amol/cell in thestationary phase (data not shown). Dead (fixed) cells had an ATP concen-tration of approximately 2 amol/cell. In order to investigate the influencesof bacterial cell activity on themembrane biofilm activity andmembrane

Fig. 2. Influence of (A) active and (B) inactive (fixed) biofoulant on biofilm activitylevel and membrane permeate flux during different filtration durations (4, 11, and24 h).

87A.M. Motlagh et al. / Desalination 316 (2013) 85–90

performance in this study, the filtration experiments were conductedwith active bacterial cells in their late-exponential phase, dead bacterialcells with minimal extracellular polymeric substances (EPS) content,and bacterial cells in different growth conditions.

2.4. Surface biofilm metabolic activity analysis

For the membrane biofilm analysis, the biomass was removed fromthe membrane surface. A section of the biofouled membrane with anarea of 2 cm2 (10 mm×20 mm) was cut and the biomass was scrapedoff the membrane. In order to counteract aggregation of the bacterialcells, 0.1% (V/V) tween-80 was added to DI water and the membranewas immersed in the tween-80 solution and agitated for 5 min at25 rpm. Bacteria thatwere still attached to themembranewere scrapedoff themembrane by using a sterilized utility knife. To avoid light inter-ference, all procedures were carried out in a poorly lit room and all ofthe materials and solutions were stored in the dark. Furthermore, inorder to minimize light interference of adjacent wells in the spectro-photometer, black microplates were used. Excitation and emissionwavelengths were 535 nm and 587 nm, respectively.

2.5. FTIR spectroscopy: Biofilm chemical characterization

FTIR spectroscopy analysis uses measurements of vibrated spectrato identify the chemical structure of materials. It can be used to assessthe presence of different bands of the biofouling layers such as poly-saccharides, proteins, and fatty acids [25,26]. Measurements wereperformed using a Digilab UMA 600 FT-IT microscope with a PikeHATR adapter and an Excalibur FTS 400 spectrometer. Membranesheets (2 cm in diameter) of each (clean and biofouled samples),were analyzed using FTIR in attenuated total reflectance (ATR)mode. Clean and fouled membrane sheets at the permeate outletand flow inlet were analyzed using FTIR in attenuated total reflec-tance (ATR) mode with a five scan iterations of 256 in the range of700 cm−1 to 4000 cm−1 and infrared spectra of fouled membraneswere compared. These measurements will allow differentiation ofEPS produced from biofoulant harvested at different phases of thelife cycle (lag, exponential, stationary). Examples of how IR peaksshift can be quantified in the presence of a biofouling layer on a mem-brane surface which can be reviewed elsewhere [26–28].

3. Results and discussion

3.1. Effect of the presence of active biofoulant on biofilm activity and fluxdecline

This experiment investigated the influence of biofoulant activity onbiofilm formation and formed biofilm activity level by introducing ac-tive or inactive (fixed) bacteria in the feed water prior to filtration andmeasuring biofilm activity after 4, 11, and 24 h. Biofoulant and biofilmATP activity levels as well as membrane performance in terms of fluxdecline are shown in Fig. 2A (active biofoulant) and B (inactivebiofoulant). During filtration with active biofoulant, formed biofilm ac-tivity, as measured by ATP, reached an equilibrium concentration of ap-proximately 8 amol/cell after 10–12 h of filtration, regardless of initialbiofoulant ATP activity levels which varied between less than 6 amol/cell and more than 9 amol/cell. The normalized flux decreased with in-cubation time to 0.91. The inactive (fixed) biofoulant cells remained atapproximately 2 amol ATP/cell in the biofilm. As the bacterial cellswere pre-fixed with formaldehyde, ATP activity of the membrane bio-film did not change significantly during different durations of filtration(up to 24 h). In addition, the normalized flux only reached 0.98.

Biofilm growth has been attributed to flux loss in the filtrationprocess. It has been purported that the primary adhesion of microor-ganisms is an abiotic process driven by physicochemical processes[29], and that there might be no physiological or chemical differences

between active or inactive biofilm. However, our results indicated asignificant difference in terms of flux decline for active and inactive(fixed) biofilms. Whereas, our previous findings verified no statisti-cally significant difference between biofilm coverage on the surfaceof the membrane for active or inactive biofoulant cells following24 h of filtration [19]. These results are consistent with previous stud-ies on live and dead biofilms [10] that reported reduction in flux de-cline in the absence of live, active cells.

The equilibrated level of ATP activity in the biofilm formed thatwas observed in these experiments may be attributed to limitationsto the penetration of nutrients through the biofilm due to diffusionlimitations. Hence, there may be a certain threshold level, whichmight result from a maximum active biofilm thickness determinedby diffusivity of the available nutrients [30]. There is also the possibil-ity that the biofilm cells reached their carrying capacity over time. Adefined and consistent supply of nutrients can only support a fixedlevel of cell activity (biofilm) [31]. Based on these findings, it wouldappear that the biofilm activity level might be independent of thebiofoulant activity level and that the biofilm activity level might di-rectly affect membrane flux. However, these findings prompted addi-tional questions related to the impact of more widely varying activitylevels or physiological states of the biofoulant, rather than cellsharvested in exponential growth phase and fixed cells as demonstrat-ed in Fig. 2A and B.

3.2. Effect of biofoulant growth phase on biofilm activity

The biofoulant (P. fluorescens) was harvested at different phases ofgrowth (lag, late exponential, and stationary) in order to represent var-ied initial ATP levels for the filtration experiments and to investigate theimpact of initial biofoulant activity level on biofilm formation. The

88 A.M. Motlagh et al. / Desalination 316 (2013) 85–90

varied biofoulant ATP levels did provide divergent results, as measuredby the ATP activity of the membrane biofilm formed after 24 h of filtra-tion and flux declines (Fig. 3).

Initial biofoulant ATP activity levels were as expected with thelowest being the lag phase cells (6.8 amole/cell) and the highestbeing the exponential phase cells (11.2 amole/cell). However, unlikein the previous experiment, the formed biofilm ATP levels variedwidely after 24 h of filtration. The ATP activity levels of the lag andstationary phase biofilms were 1.4 and 2.1 amol/cell, respectively,which is significantly lower than the ATP level of the exponentialphase biofilm (10.2 amol/cell). Interestingly, the feed water ATP ac-tivity level after 24 h of filtration was lowest for the exponentialphase biofoulant while the lag and stationary phase feed watersafter 24 h showed no statistically significant difference. The lag andstationary phase biofoulant samples showed a downward trend inATP activity from initial biofoulant, to feed water ATP activity levelafter 24 h, and finally ATP activity on the formed biofilm. The expo-nential phase biofoulant samples did not reflect this trend.

In the case of the exponential phase biofoulant, the ATP activitylevel of the biofoulant decreased dramatically (from 11 to 2 amolATP/cell) in the feed water after 24 h of filtration but the activity onthe membrane biofilm is similar to the biofoulant activity that wasinitially present in the feed water. This finding might suggest that ac-tive cells tend to adhere to the membrane due to accumulation of nu-trients in the EPS matrix on the surface, which contributes to theirbacterial growth rate and metabolic activity [32]. On the contrary,the lag phase and stationary phase cultures produced similar out-comes to one another, resulting in higher levels of ATP activity inthe feed water as compared to in the biofilm. This suggests thatthese cells are more competitive in planktonic condition rather thanin the biofilm during the 24 h of this experiment.

3.3. Effect of biofoulant growth phase on flux decline

There was no direct relationship between biofilm ATP activitylevel and normalized flux as was previously observed and describedin Fig. 2A and B (Table 1). It was anticipated that the highest flux de-cline would be measured for exponential phase biofoulants, sincethey produce the most ATP active biofilm. Unexpectedly, the mini-mum normalized flux (normalized flux=0.765) was observed duringthe filtration experiment with stationary phase biofoulant, which hada biofilm activity level of 2.1 amol/cell. Whereas, the exponential andlag phase biofoulants resulted in normalized fluxes of 0.876 and

Pure CultureFeed Water

Membrane

Lag

Exponential

Stationary

0

2

4

6

8

10

12

AT

P A

ctiv

ity

(am

ol/c

ell)

Fig. 3. ATP activity changes in different growth phases in biofoulant, feed water after24 h of filtration, and biofilm scraped from the membrane surface. Standard deviationsof ATP concentration were ≤10% of the mean value.

0.923, respectively. This finding is likely attributed to the amount ofbiomass present on the membrane or the quantity and type of EPSin the biofilm, which may hinder membrane flux.

To investigate this, the bacterial concentration in the feed waterand the biomass comprising the membrane biofilm was considered(Table 1). The concentration of bacteria on the membrane is highestfor the stationary phase experiment (OD600=0.245), likely contrib-uting to cake formation and the maximum flux decline (normal-ized flux=0.76). However, the exponential phase experimentyielded the highest concentration of bacteria in the feed water(OD600=0.26) and the lowest amount of biofilm formed on themembrane (OD600=0.043), while the normalized flux was sec-ond lowest at 0.876. So, the normalized flux results could notbe simply explained by the ATP activity of the biofilm or by theconcentration of bacteria on the membrane.

It was observed that the biomass embedded in the biofilm was in-creased when bacterial cells were in stationary phase. This could sug-gest the presence of a signal substance, present within stationaryphase cultures, which promotes attachment to membrane surfaceand biofilm formation [33]. In addition, the difference in biofilm for-mation could be due to different types of EPS produced by biofoulantsharvested during the different phases of the growth cycle. In order toevaluate this possibility, FTIR analysis was carried out in attenuatedtotal reflectance (ATR) mode to study the chemical nature of biofilmspresent on the membranes at lag, exponential and stationary phases(Fig. 4). Peak analyses of FTIR spectra for the membranes were ana-lyzed using Table 2 [26,34–36].

The peak at 1040 cm−1 at the lag and exponential phases shows theC\O stretching of polysaccharides present in EPS on the membranes.Lag and exponential phases have another peak at 1220 cm−1 which isassociated with the P_O of the phosphodiester of nucleic acid (likely at-tributed to cells in the biofilm). Peaks at 1640 cm−1 and 1530 cm−1 forthe exponential and stationary phases show the amide I band andamide II band of protein. There is no peak for the lag phase sample at1640 cm−1 and 1530 cm−1, which indicates that protein was presentonly at the exponential and stationary phases suggesting the formationof additional EPS as protein can also be a component of EPS [8]. Thereare unique peaks at 1080 cm−1 and 1400 cm−1 for stationary phasesamples indicating C\O\C ring vibration of polysaccharide and asym-metric deformation of CH3 and CH2 or symmetric stretch of C\O ofCOO− groups of protein, respectively. FTIR analysis at different stagesshowed varied transmittance intensities showing different quantitiesand compositions of EPS. There were more peaks measured and alower normalized flux for samples from stationary and exponentialphases, which is consistent with the smaller quantity of overall EPS vi-sually observed in lag phase biofilm samples. Furthermore, uniquepeaksmeasured for stationary phase samples as compared to exponen-tial phase samples indicate different chemical compositions of EPS. TheFTIR result showed variation in the chemical composition of EPS at dif-ferent growth phases, which could have dramatically changed the bio-mass content on the membrane.

It has been found previously that Pseudomonas produces two verydissimilar polysaccharides that compose the EPS. The first type that isproduced only in the exponential phase of growth contains sugarsglucose, galactose, glucuronic acid, and galacturonic acid. The secondtype of EPS that is produced only in the stationary phase

Table 1Concentration of bacteria in the feed water and on the membrane as well as operation-al flux decline corresponding to different growth phases.

Phase Concentration of bacteria (OD600) Flux decline

Feed water Membrane

Lag 0.046 0.151 0.923Exponential 0.265 0.043 0.876Stationary 0.060 0.245 0.765

Fig. 4. Representative ATR spectra (2000–600 cm−1) of biofilm formed from lag, lateexponential, and stationary phase bacteria.

89A.M. Motlagh et al. / Desalination 316 (2013) 85–90

contains N-acetylglucosamine, 2-keto-3-deoxyoctulosonic acid,and an unidentified 6-deoxyhexose [37]. This difference in the polysac-charides can result in the dramatic variation observed in themembranebiomass content since certain chemical substances can encourage bio-film formation by serving as a readily degradable food source or by pro-viding a “sticky” surface. In addition, these substances might result invarying flux declines due to their composition (size). For example,amino sugar N-acetylglucosamine forms several hundred to severalthousand repeating units to form the final EPS [38].

4. Conclusions

The results indicated that the bacterial growth phase affects met-abolic activity as determined by ATP concentration, and the maxi-mum value occurs during late exponential growth phase. The initialbiofoulant activity is not directly correlated to the activity of themembrane biofilm as measured by ATP. However, inactivating (fix-ing) the biofoulants can be a great help in a filtration system to reducethe chance of biofilm formation and flux decline. There is much lessflux decline with fixed (inactive) cells as compared to an active bio-film, as measured by ATP.

The growth phase of the initial biofoulant determines the ATP ac-tivity in the feed water and the biofilm. Lag phase and stationaryphase had similar and low levels of ATP activity in the biofilm(b2.5 amol/cell). However, the exponential phase cells resulted insignificant activity (>10 amol/cell) in the biofilm formed. The lagphase resulted in the least amount of flux decline. Although the activ-ity level in the biofilm created via stationary phase cells was lower,this experiment resulted in the most flux decline and most biomassformed on the membrane, due to quantity and composition of EPSon the membrane.

Table 2FTIR functional group location and corresponding region.

Wavenumber Group Region

~970 C\O stretch Polysaccharide~1052 C\O or C\O\C stretch Polysaccharide~1085 Ring vibration Polysaccharide~1242 P_O of the phosphodiester Nucleic acid~1398 Asymmetric deformation of CH3 and CH2

and symmetric stretch of C\O of COO−

groups

Protein

~1455 Asymmetric deformation of CH3 and CH2 Protein1652–1648 Amide I (C_O) Protein1550–1548 Amide II (C\N, N\H) Protein~1735 C_O stretch of esters Fatty acids

PolysaccharidesLipopolysaccharides

Lastly, it can be concluded that sole measure of activity is notenough, and that quantifying the cell activity is not adequate to pre-dict flux decline. More information is needed as demonstrated by var-iable flux declines with similar levels of activity in the growth phaseexperiment.

Acknowledgments

This work was supported with the grant # 4614 Ohio Water De-velopment Authority: State-of-the-art membrane characterization to-ward biofouling control and improved membrane performance.

References

[1] G.K. Pearce, UF/MF pre-treatment to RO in seawater and wastewater reuseapplications: a comparison of energy costs, Desalination 222 (2008) 66–73.

[2] Recent advances and research needs in membrane fouling. AWWA MembraneTechnology Research Committee report, I.C. Escobar, E.M. Hoek, C.J. Gabelich,F.A. Digiano, Y.A. Le Gouellec, P. Berube, K.J. Howe (Eds.), J. AWWA 97 (2005)79–89.

[3] H.C. Flemming, Reverse osmosis membrane biofouling, Exp. Therm. Fluid Sci. 14(1997) 382.

[4] In: W.G. Characklis, K.C. Marshall (Eds.), Biofilms, Publ. John Wiley, New York,1990.

[5] C.A. Fux, P. Stoodley, L. Hall-Stoodley, J.W. Costerton, Bacterial biofilms: a diag-nostic and therapeutic challenge, Expert Rev. Anti Infect. Ther. 1 (2003) 667–683.

[6] I. Ofek, R.J. Doyle, Bacterial Adhesion to Cells and Tissues, Chapman & Hall, NewYork, 1994.

[7] I.B. Beech, Corrosion of technical materials in the presence of biofilms—currentunderstanding and state-of-the art methods of study, Int. Biodeter. Biodegr. 53(2004) 177–183.

[8] H.C. Flemming, J. Wingender, Relevance of microbial extracellular polymeric sub-stances (EPSs)—part I: structural and ecological aspects, Water Sci. Technol. 43(2001) 1–8.

[9] N.J. Horan, C.R. Eccles, Purification and characterization of extracellular polysac-charide from activated sludges, Water Res. 20 (1986) 1427–1432.

[10] M. Herzberg, M. Elimelech, Biofouling of reverse osmosis membranes: role ofbiofilm-enhanced osmotic pressure, J. Membr. Sci. 295 (2007) 11–20.

[11] H.S.C. Eydal, K. Pedersen, Use of an ATP assay to determine viablemicrobial biomassin Fennoscandian Shield groundwater from depths of 3–1000 m, J. Microbiol.Methods 70 (2007) 363–373.

[12] D.C. White, R.J. Bobbie, J.S. Herron, J.D. King, S.J. Morrison, Biochemical measure-ments of microbial mass and activity from environmental samples, native aquaticbacteria: enumeration, activity, and ecology, Am. Soc. Test. Mater. (1979) 69–81.

[13] J. Patternson, P.L. Brezonik, H.D. Putnam, Measurement and significance of aden-osine triphosphate in activated sludge, Environ. Sci. Technol. 4 (1970) 569–575.

[14] F.A. Hammes, T. Egli, New method for assimilable organic carbon using flow-cytometric enumeration and a natural microbial consortium as inoculum, Envi-ron. Sci. Technol. 39 (2005) 3289–3294.

[15] L. Fiksdal, L. Bjorkoy, Development of a toolbox for identifying and quantifyingmembrane biofouling in drinking water treatment, Protocol Report, Techneau,2007.

[16] S.Y. Chiu, I.C. Kao, L.E. Erickson, L.T. Fan, ATP pools in activated sludge, WaterPollut. Control Fed. 45 (1973) 1746–1758.

[17] R.M. Atlas, Enumeration and estimation of microbial biomass, Experimental Mi-crobial EcologyBlackwell Scientific, 1982, pp. 84–102.

[18] K. Good, I. Escobar, X. Xu, M. Coleman, M. Ponting, Modification of commercialwater treatment membranes by ion beam irradiation, Desalination 146 (2002)259–264.

[19] A. Zaky, I. Escobar, A.M. Motlagh, C. Gruden, Determining the influence of activecells and conditioning layer on early stage biofilm formation using cellulose ace-tate ultrafiltration membranes, Desalination 286 (2012) 296–303.

[20] M.O. Pereira, M.J. Vieira, Effects of the interactions between glutaraldehyde andthe polymeric matrix on the efficacy of the biocide against Pseudomonasfluorescens biofilms, Biofouling 17 (2001) 93–101.

[21] M. Simoes, L.C. Simoes, M.J. Vieira, Physiology and behavior of Pseudomonasfluorescens single and dual strain biofilms under diverse hydrodynamics stresses,Int. J. Food Microbiol. 128 (2008) 309–316.

[22] Y. Su, J. Hong, Betanodavirus B2 causes ATP depletion-induced cell death via mi-tochondrial targeting and complex II inhibition in vitro and in vivo, J. Biol. Chem.285 (2010) 39801–39810.

[23] H. Wei, Z. Zhang, A. Saha, S. Peng, G. Chandra, Z. Quezado, A.B. Mukherjee, Disrup-tion of adaptive energy metabolism and elevated ribosomal p-S6K1 levels con-tribute to INCL pathogenesis; partial rescue by resveratrol, Hum. Mol. Genet. 20(2011) 1–11.

[24] K. Venkateswaran, N. Hattori, M.T. La Duc, T. Kern, ATP as a biomarker of viable mi-croorganisms in clean-room facilities, J. Microbiol. Methods 52 (2003) 367–377.

[25] J. Pink, D. Chisholm, T.J. Beveridge, D.A. Pink, An FTIR study of Pseudomonasaeruginosa PAO1 biofilm development: interpretation of ATR–FTIR data in the1500–1180 cm−1 region, Biofilms 2 (2005) 165–175.

90 A.M. Motlagh et al. / Desalination 316 (2013) 85–90

[26] J. Schmitt, D. Nivens, D.C. White, H.C. Flemming, Changes of biofilm properties inresponse to sorbed substances — an FTIR–ATR study, Water Sci. Technol. 32(1995) 149–155.

[27] R. Hausman, T. Gullinkala, I.C. Escobar, Development of copper-charged polypro-pylene feedspacers for biofouling control, J. Membr. Sci. 358 (2010) 114–121.

[28] L. Marcotte, G. Kegelaer, C. Sandt, J. Barbeau, M. Lafleur, An alternative infraredspectroscopy assay for the quantification of polysaccharides in bacterial samples,Anal. Biochem. 361 (2007) 7–14.

[29] K.C. Marshall, B. Blainey, Role of bacterial adhesion in biofilm formation andbiocorrosion, in: H.-C. Flemming, G.G. Geesey (Eds.), Biofouling and Biocorrosionin Industrial Water Systems, Springer, Berlin Heidelberg New York, 1991, pp. 28–45.

[30] E.J. LaMotta, Kinetics of growth and substrate uptake in a biological film system,Appl. Environ. Microbiol. 31 (1976) 286–293.

[31] D.G. Allison, B. Ruiz, C. SanJose, A. Jaspe, P. Gilbert, Extracellular products as me-diators of the formation and detachment of Pseudomonas fluorescens biofilms,FEMS Microbiol. Lett. 167 (1998) 179–184.

[32] M.C. van Loosdrecht, J. Lyklema, W. Norde, A.J. Zehnder, Influence of interfaces onmicrobial activity, Microbiol. Mol. Biol. Rev. 54 (1990) 75–87.

[33] A.W. Decho, Microbial biofilms in intertidal systems: an overview, Cont. Shelf Res.20 (2000) 1257–1273.

[34] M. Kanziz, P. Heraud, B. Wood, F. Burden, J. Beardall, D. McNaughton, Fouriertransform infrared microspectroscopy and chemometrics as a tool for the dis-crimination of cyanobacterial strains, Phytochemistry 52 (1999) 407–417.

[35] D. Osiro, L.A. Colnago, A.M.M.B. Otoboni, E.G.M. Lemos, d.S.A. Alves, C.F.H. Della,M.A. Machado, A kinetic model for Xylella fastidiosa adhesion, biofilm formation,and virulence, FEMS Microbiol. Lett. 236 (2004) 313–318.

[36] H.Y. Cheung, G.K.L. Chan, S.H. Cheung, S.Q. Sun,W.F. Fong, Morphological and chem-ical changes in the attached cells of Pseudomonas aeruginosa as primary biofilms de-velop on aluminum and CaF2 plates, J. Appl. Microbiol. 102 (2007) 701–710.

[37] B.E. Christensen, J. Kjosbakken, O. Smidsrod, Partial chemical and physical charac-terization of two extracellular polysaccharides produced by marine, periphyticPseudomonas sp. strain NCMB 2021, Appl. Environ. Microbiol. 50 (1985) 837–845.

[38] B. Degeest, F. Vaningelgem, A.P. Laws, L. De Vuyst, UDP-N-Acetylglucosamine4-epimerase activity indicates the presence of N-acetylgalactosamine inexopolysaccharides of Streptococcus thermophilus strains, Appl. Environ.Microbiol. 67 (2001) 3976–3984.