University of Wollongong University of Wollongong

Research Online Research Online

Faculty of Engineering and Information Sciences - Papers: Part A

Faculty of Engineering and Information Sciences

1-1-2015

The role of forward osmosis and microfiltration in an integrated osmotic-The role of forward osmosis and microfiltration in an integrated osmotic-

microfiltration membrane bioreactor system microfiltration membrane bioreactor system

Wenhai Luo University of Wollongong, wl344@uowmail.edu.au

Faisal I. Hai University of Wollongong, faisal@uow.edu.au

Jinguo Kang University of Wollongong, jkang@uow.edu.au

William E. Price University of Wollongong, wprice@uow.edu.au

Long D. Nghiem University of Wollongong, longn@uow.edu.au

See next page for additional authors

Follow this and additional works at: https://ro.uow.edu.au/eispapers

Part of the Engineering Commons, and the Science and Technology Studies Commons

Recommended Citation Recommended Citation Luo, Wenhai; Hai, Faisal I.; Kang, Jinguo; Price, William E.; Nghiem, Long D.; and Elimelech, Menachem, "The role of forward osmosis and microfiltration in an integrated osmotic-microfiltration membrane bioreactor system" (2015). Faculty of Engineering and Information Sciences - Papers: Part A. 3957. https://ro.uow.edu.au/eispapers/3957

Research Online is the open access institutional repository for the University of Wollongong. For further information contact the UOW Library: research-pubs@uow.edu.au

The role of forward osmosis and microfiltration in an integrated osmotic-The role of forward osmosis and microfiltration in an integrated osmotic-microfiltration membrane bioreactor system microfiltration membrane bioreactor system

Abstract Abstract This study investigates the performance of an integrated osmotic and microfiltration membrane bioreactor (O/MF-MBR) system for wastewater treatment and reclamation. The O/MF-MBR system simultaneously used microfiltration (MF) and forward osmosis (FO) membranes to extract water from the mixed liquor of an aerobic bioreactor. The MF membrane facilitated the bleeding of dissolved inorganic salts and thus prevented the build-up of salinity in the bioreactor. As a result, sludge production and microbial activity were relatively stable over 60 days of operation. Compared to MF, the FO process produced a better permeate quality in terms of nutrients, total organic carbon, as well as hydrophilic and biologically persistent trace organic chemicals (TrOCs). The high rejection by the FO membrane also led to accumulation of hydrophilic and biologically persistent TrOCs in the bioreactor, consequently increasing their concentration in the MF permeate. On the other hand, hydrophobic and readily biodegradable TrOCs were minimally detected in both MF and FO permeates, with no clear difference in the removal efficiencies between two processes.

Disciplines Disciplines Engineering | Science and Technology Studies

Publication Details Publication Details Luo, W., Hai, F. I.., Kang, J., Price, W. E., Nghiem, L. D. & Elimelech, M. (2015). The role of forward osmosis and microfiltration in an integrated osmotic-microfiltration membrane bioreactor system. Chemosphere, 136 125-132.

Authors Authors Wenhai Luo, Faisal I. Hai, Jinguo Kang, William E. Price, Long D. Nghiem, and Menachem Elimelech

This journal article is available at Research Online: https://ro.uow.edu.au/eispapers/3957

1

The role of forward osmosisand microfiltration in an integrated1

osmotic-microfiltration membranebioreactor system2

Revisedmanuscript submitted to Chemosphere3

April 20154

WenhaiLuo a, FaisalI. Hai a, JinguoKangb, William E. Priceb, LongD. Nghiema*,5

MenachemElimelechc6

a StrategicWaterInfrastructureLaboratory, Schoolof Civil Mining andEnvironmental7

Engineering,Universityof Wollongong,Wollongong,NSW2522,Australia8

b StrategicWaterInfrastructureLaboratory,Schoolof Chemistry,Universityof Wollongong,9

Wollongong,NSW2522,Australia10

c Departmentof ChemicalandEnvironmentalEngineering,YaleUniversity,NewHaven,CT11

06520-8286,UnitedStates12

* Correspondingauthor:longn@uow.edu.au; Ph:+61(2) 42214590.

2

Abstract13

This study investigates the performanceof an integrated osmotic and microfiltration14

membranebioreactor(O/MF-MBR) systemfor wastewatertreatmentand reclamation. The15

O/MF-MBR systemsimultaneouslyusedmicrofiltration (MF) and forward osmosis(FO)16

membranesto extract water from the mixed liquor of an aerobic bioreactor.The MF17

membranefacilitatedthe bleedingof dissolvedinorganicsaltsandthuspreventedthe build-18

up of salinity in the bioreactor.As a result, sludgeproductionand microbial activity were19

relatively stableover 60 daysof operation. Comparedto MF, the FO processproduceda20

betterpermeatequality in termsof nutrients,total organiccontent,aswell ashydrophilicand21

biologically persistentTrOCs. The high rejection of the FO membranealso led to the22

transportof severalhydrophilic and biologically persistent TrOCs to the MF permeate. On23

the other hand,hydrophobicand readily biodegradableTrOCs were minimally detectedin24

bothMF andFO permeates,with no cleardifferencein theremovalefficienciesbetweentwo25

processes.26

Key words: Osmoticmembranebioreactor(OMBR); Forwardosmosis(FO); Microfiltration27

(MF); Traceorganicchemicals(TrOCs);Salinity build-up;28

3

1. Introduction29

Water reuseis an importantmeasureto tackle water scarcityand environmentalpollution,30

which arekey factorshamperingeconomicdevelopmentandthreating the nautralecosystem31

(Wintgens et al., 2008; Hochstrat et al., 2010). Safe and reliable water reuse requires32

adequateremovalof salts,nutrients,pathogenicagents,andtraceorganicchemicals(TrOCs)33

from the reclaimed effluent. TrOCs are a diverserangeof emergingorganic chemicals of34

eitheranthropogenicor natural origin. They occurubiqituouslyin munucipalwastewaterat35

concentrations in the rangeof a few nanogramsper lit re (ng/L) to severalmicrogramsper36

lit re (µg/L) (Luo et al., 2014). TheseTrOCspresentarguably the mostvexing challengeto37

practicalpotablewater reuse(Wintgenset al., 2008; Lampardet al., 2010; Dreweset al.,38

2013; Luo etal., 2014).39

Adequate removal of TrOCs is also essentialto facilitate water reuse for agriculture40

production. It has been demonstratedthat the occurrenceof pharmaceuticals, such as41

carbamazepineandtriclocarban, in reclaimedwastewater(Tanoueet al., 2012) andbiosolids42

(Wu et al., 2012) usedto grow fruits and vegetablescan bio-accumulatein ediblepartsof43

theseproduces.Therefore, a major technicalchallengefor the water industry is to develop44

new treatmentprocessesthat can reliably and cost-effectively removethese TrOCs during45

waterreuse.46

Recentefforts in wastewatertreatmentand reusehave led to the emergenceof a novel47

osmoticmembranebioreactor(OMBR) process(Achilli et al., 2009; Cornelissenet al., 2011;48

Nawazet al., 2013), which integratesforward osmosis(FO) with the conventionalactivated49

sludgetreatmenttechnology.In the OMBR system,the osmoticpressuredifferencebetween50

the mixed liquor and draw solution (e.g. NaCl) induceswater diffusion through a semi-51

permeable FO membrane. The FO membrane can effectively retain small organic52

contaminantsin the bioreactor, therebyfacilitating their subsequentbiodegradation(Alturki53

et al., 2013; Codayetal., 2014). Indeed,recentstudieshaveshowntheexcellentperformance54

of OMBR for TrOC removal, particularly the compoundswith relatively large molecule55

weightand/or featuredwith negativecharge(Alturki et al., 2012; Lay et al., 2012; Holloway56

et al., 2014). Thus,OMBR canpotentiallyproducehigh quality reclaimedwaterfor potable57

reuse,irrigation,or directdischargein environmentallysensitiveareas.58

Despitethepotentialof OMBR, salinitybuild-up in thebioreactorcausedby high rejectionof59

theFOmembraneandreversetransportof thedrawsolutionremainsa technicalchallengefor60

4

its further development(Van der BruggenandPatricia,2015). The high bioreactorsalinity61

canreducethedriving force for watertransport(Lay et al., 2010). Sludgecharacteristics and62

microbial community can also be altered with the elevated bioreactor salinity and63

subsequentlyworsenthe biological treatmentand membraneperformance(Qiu and Ting,64

2013). A shortsludgeretentiontime (SRT) is expectedto control the build-up of salinity in65

the bioreactor. However, in an OMBR systemwith an operatingSRT of 10 days, the66

bioreactorsalinity still increasedsubstantially, exertinginhibition on the microbial activity67

(Wanget al., 2014a). The shortSRT could alsoadverselyaffect the biological performance68

(Grelier et al., 2006) and increasethe cost for wastesludgedisposal.Severalstudieshave69

recentlyproposedthe integrationof an microfiltration (MF) or ultrafiltration (UF) process70

with OMBR to bleedout inorganicsaltsfrom the bioreactor(Holloway et al., 2014, 2015;71

Wanget al., 2014b). By applying theapproach,Holloway et al. (2014,2015)showeda stable72

operationof a pilot UFO-MBR treating raw domesticwastewaterover a period of four73

months. Removal to below the detection limit was reported for 15 out of 20 TrOCs74

investigatedin their study in 2014usinga pilot reverseosmosisprocessfor drawsolutionand75

cleanwaterrecoveries(Hollowayet al.,2014).76

Building uponthe existingliteratureon this topic, we aimedto evaluatethe performanceof77

an integratedosmoticandmicrofiltration membranebioreactor(O/MF-MBR) by specifically78

comparingpermeatequalities betweenthe FO and MF processes and examining sludge79

stability in the bioreactor.Thesystemperformancewasalsoassessedin termsof waterflux,80

bioreactorsalinity,andmembranefouling. TrOC removalwasrelatedto their hydrophobicity81

andmolecularstructures to mechanisticallyelucidatetheir fate within the integratedO/MF-82

MBR system.The interactionbetweenFO andMF in the integratedsystemwith regardsto83

thefateandremovalof TrOCswasalsodiscussed.84

2. Materials and methods85

2.1 Representativetraceorganicchemicals86

A stocksolutioncontaining30 representativeTrOCs(TableS1, SupplementaryData)were87

preparedin pure methanoland storedat -18 °C in the dark. The stock solution was used88

within less than a month. These TrOCs were selectedto representfour major groupsof89

chemicalsof emergingconcern– pharmaceuticaland personalcare products,endocrine90

disruptingcompounds, pesticides,andindustrialchemicals– thatareubiquitousin municipal91

wastewater. They have a diverserangeof properties,including hydrophobicity,molecular92

5

weight,andfunctionalgroups(TableS1,SupplementaryData). Hydrophobicityof anorganic93

compoundcan be measuredby Log D, which is the effective octanol-water partition94

coefficientata givensolutionpH (NghiemandColeman,2008). Basedon their Log D values95

at pH of 7, the selectedTrOCs can be classifiedas hydrophilic (i.e. Log D pH 7 < 3) or96

hydrophobic(i.e.Log D pH 7 > 3).97

2.2 FO andMF membranes98

A flat-sheet,cellulosebasedmembranesuppliedby HydrationTechnologyInnovations(HTI,99

Albany, USA) was usedin the FO process.The FO membraneis composedof a cellulose100

triacetate active (CTA) layer reinforced by a polyester mesh for mechanicalsupport101

(McCutcheonand Elimelech,2008). It is noteworthythat thin film composite(TFC) FO102

membraneswith embeddedpolyesterscreensupporthavealso beenreleasedby HTI and103

severalothermanufacturesin recentyears. Both CTA andTFC membraneshavetheir own104

positive attributes.Findingsfrom this study are specific to the OMBR processratherthan105

specificmembranepropertiesandthusapplicableto all typesof FOmembranes.106

A hollow fibre, polyvinylidene fluoride MF membranemodule from Mitsubishi Rayon107

Engineering(Tokyo, Japan) wassubmergedin thebioreactor.Theeffectivesurfaceareaand108

nominalporesizeof theMF membranewere740cm2 and0.4µm, respectively.109

2.3 Experimentalsystem110

The integratedO/MF-MBR systemusedin this study was composedof a cross-flow FO111

configuration,a submergedMF membranemodule,anda 10 L aerobicbioreactor(Fig. 1). An112

electricalair pump(Heilea,Ningbo,China)wasusedto continuouslyaeratethe reactorvia a113

coarsediffuser.A Masterflexperistalticpump(Cole-Parmer,VernonHills, USA) wasusedto114

draw permeatethrough the MF membranewith an operationon/off time of 14/1 min.115

Transmembranepressure(TMP) of theMF membranewascontinuouslymonitoredby a high116

resolution(±0.1kPa)pressuresensor(ExtechInstruments,Nashua,USA).117

A detaileddescriptionof the cross-flow FO configurationis availableelsewhere(Alturki et118

al., 2012). Briefly, theFOconfigurationcomprisedtwo semi-cellsmadeof acrylicplasticand119

a drawsolutiondeliveryandcontrol equipment.TheFO membranewasplacedbetweentwo120

semi-cellsto sealthefeedanddrawsolutionchannelswith a length,width, anddepthof 145,121

95, and 2 mm, respectively.The effective membranesurfaceareawas 138 cm2, with the122

active layer facing the feedchannel(i.e. FO mode).The mixed liquor in the bioreactorwas123

6

circulatedto the feedchannelby a Masterflexperistalticpump(Cole-Parmer,VernonHills,124

USA). On theotherside, a gearpump(Micropump,Vancouver,USA) wasusedto circulatea125

draw solution to the draw solution channel.The circulation flow rateof both the feed and126

draw solutionswas 1 L/min (i.e. a cross-flow velocity of 9 cm/s)monitoredby rotameters127

(Cole-Parmer, Vernon Hills, USA). The draw solution reservoirwas placedon a digital128

balance connectedto a computer. During the experimentalperiod, the draw solution129

concentrationwas kept constantby a conductivity controller equippedwith a conductivity130

probeanda Masterflexperistalticpumpto automaticallydosea concentrateddrawsolutionto131

the draw solution reservoir.The controller accuracywas 0.1 mS/cm (i.e. 0.05 g/L NaCl).132

Both theconcentratedandworking drawsolutionreservoirswereplacedon the samedigital133

balanceto avoidexperimentalerrorsby theconcentrationcontrolequipment.134

[FIGURE 1]135

2.4 Experimentalprotocol136

A submergedMF-MBR systemwasfirst initiatedto seedthebioreactorwith activatedsludge137

from the Wollongong WastewaterTreatmentPlant (Wollongong, Australia). The initial138

mixed liquor suspendedsolid (MLSS) concentrationin the bioreactorwasapproximately5139

g/L. Syntheticwastewaterwas used to simulate medium strengthmunicipal sewageand140

consistedof 100 mg/L glucose,100 mg/L peptone,17.5mg/L KH2PO4, 17.5mg/L MgSO4,141

10 mg/L FeSO4, 225 mg/L CH3COONa and 35 mg/L urea. The MF-MBR systemwas142

stabilized in a temperature-controlled room (22 ± 1 °C) at a working volume of 6 L, a143

hydraulicretentiontime (HRT) of 24 h, anda dissolvedoxygenconcentration(DO) of 5 ± 1144

mg/L. Comparedto a typical MBR system,a longerHRT wasusedin thisstudyto maintaina145

relativelylow waterflux to minimizethemembranefouling. Therelativelyhigh aerationrate146

of 8 L/min usedhere to preventsludgesettlementand scour the membranesurfacealso147

resultedin a higher DO concentrationthan that in a typical MBR system.No sludgewas148

wasted(exceptfor weeklysamplingof 90 mL mixed li quor) to systematicallyinvestigatethe149

build-up of salinity in the bioreactor.Stability of the bioreactorwas determinedby sludge150

production,biomassactivity, andremovalof organicmatterandnutrients.In practice,regular151

sludgewithdrawalcanalleviatesalinity build-up to someextent.152

Once stabilized, the cross-flow FO processwas connectedto the bioreactorto form an153

integratedO/MF-MBR system.At thesametime, theTrOC stocksolutionwasspikedto the154

influent to obtain5 µg/L of eachof the 30 compounds.The integratedsystemwasoperated155

7

continuouslyfor 60 daysundertheconditionsasmentionedabove.To minimizethebiosolids156

blockage in the narrow feed channel of the cross-flow FO system, the initial MLSS157

concentrationin thebioreactorwasadjustedto 2 g/L. Giventheunstablewaterflux of theFO158

process,thepermeateflux of MF wasadjusteddaily to maintaina constantHRT of 24 h. The159

drawsolutionandconcentrateddrawsolutionwere58.5and351g/L NaCl, respectively. The160

drawsolutionwasreplacedeveryday to avoidoverflow andcontaminantaccumulation.The161

concentrateddraw solution was also addedmanuallyon a daily basis.Membranecleaning162

wasnot conductedduringthis study.163

2.5 Analyticalmethods164

Total organiccarbon (TOC) andtotal nitrogen(TN) of theinfluent,mixed liquor supernatant,165

MF and FO permeateswere analysedusing a TOC/TN-VCSH analyser(Shimadzu,Kyoto,166

Japan). Orthophosphate(PO43-) was measuredby a Flow Injection Analysis system167

(QuichChem8500,Lachat,USA). MLSS andmixedliquor volatile suspendedsolid (MLVSS)168

concentrationswere determinedfollowing the StandardMethods for the Examinationof169

WaterandWastewater.Specific oxygenuptakerate(SOUR)of the sludgewastested based170

on the techniquedescribedby Choi et al. (2007). Mixed liquor pH and conductivity were171

measuredusing an Orion 4-Star Plus pH/conductivitymeter(ThermoScientific, Waltham,172

USA).173

TrOC concentrationsin the feed,mixed liquor supernatant,MF permeate,anddrawsolution174

were determinedweekly using an analytical methoddescribedby Hai et al. (2011). The175

methodinvolved solid phaseextractionand derivation, followed by gas chromatography-176

massspectrometry(GC-MS) analysisusinga ShimadzuGC-MS system(Kyoto, Japan).177

In this study,the MF andFO processeswereoperated simultaneously to extractwaterfrom178

the bioreactor.Permeatesamples could thusbe obtainedseparatelyfrom the MF-MBR and179

OMBR channels(i.e. bioreactor-MF and -FO streams, respectively). Against the feed180

contaminantconcentration,the removal efficiency through the MF-MBR channel was181

definedas:182

%100)1(e

���edF

MF

CC

R (1)183

where, CFeed and CMF were contaminantconcentrationsin the feed and MF permeate,184

respectively.Unlike the MF process,contaminants permeatedthrough the FO membrane185

8

were diluted by the draw solution. The dilution factor (DF) was calculatedusing a mass186

balance:187

FO

DS

VV

DF � (2)188

where,VDS and VFO were draw solution and FO permeatevolumesuntil samplingtime. As189

notedabove,to avoid solution overflow and contaminantaccumulation,the draw solution190

wasreplacedeveryday. Thus,the overall removalthroughthe OMBR channelwasdefined191

as:192

%100)1(eed

��� DFCC

RF

DS (3)193

whereCDS was contaminantconcentrations in thedrawsolutionreservoir.194

In this study, TrOC accumulationin biosolidswas not consideredfor removalassessment195

becauseonly compoundsin the aqueousphasecould transportthrough the MF and FO196

membranes.It is alsonoteworthythatTrOC removalhereonly indicatesthedisappearanceof197

parent moleculesbut not necessarilycompletemineralization.Indeed, biodegradationof198

certain TrOCs would produce stable intermediates/metabolites in the bioreactor and199

permeates.However,detaileddiscussionof theseaspectsis beyondthescopeof this study.200

3. Resultsand discussion201

3.1 Processperformance202

3.1.1 Salinity build-up, waterflux, andmembranefouling203

The integrationof the MF membraneinto OMBR preventedthe build-up of salinity in the204

bioreactor, becausedissolvedinorganic salts were readily permeablethrough the micro-205

porous membrane(Fig. 2). After a small increasein the first week, the mixed liquor206

conductivity stabilizedat approximately700 µS/cm (i.e. a salinity of 0.4 g/L NaCl). The207

result comparesfavourably with our previousstudy where a rapid increasein the mixed208

liquor conductivity from 268 to 8270 µS/cm was observedwithin sevendays using the209

similar experimentalconfiguration and conditions without housing the submergedMF210

membranein thebioreactor(Alturki etal., 2012).211

[FIGURE 2]212

Two distinctstagesof waterflux declinecouldbeobservedin theFO processwith time (Fig.213

2). The water flux decreasedrapidly from 6.5 to 3.4 L/m2h within the first week mainly214

9

becauseof salinity build-up in thebioreactorandmembranefouling. With thedecreasein the215

bioreactorsalinity, the waterflux of the FO processdecreasedslightly andthenstabilizedat216

approximately1.7 L/m2h from day 45 onward. The elevatedsalinity could increasethe217

osmotic pressurein the mixed liquor side and thus reduce the driving force for water218

transport.On theotherhand,high salinity could leadto doublelayercompressionandreduce219

electrostaticinteractionamong the macromoleculefunctional groups,resulting in a thicker220

and more compactfouling layer (Nghiem et al., 2005). Indeed,a thick cake layer was221

observedon themembranesurfaceat a feedcross-flow velocity of 9 cm/sin this study(Fig.222

S1,SupplementaryData).The fouling layer could increasethe hydraulicresistanceto water223

permeationand cause severe concentrationpolarization adjacent to membranesurface,224

therebyreducingthewaterflux (HoekandElimelech,2003; Boo et al.,2012).225

It is noteworthythat the stablewaterflux of approximately1.7 L/m2h wasmuchlower than226

that observedby Holloway et al. (2015). The different flux behavioursbetweenthe two227

studiescouldbeattributedto thedifferencein hydrodynamicsadjacentto membranesurface228

betweenthesubmergedandcross-flow FOsystems.In ourcross-flow FOsystem,particulates229

in mixed liquor wereproneto adhereto the membranesurfacein the narrowfeedchannel,230

particularlyata low feedcross-flow velocityof 9 cm/s.231

The TMP value of the MF membraneonly increasedto 5 kPa (0.05 bar) by the endof the232

experiment(Fig. S2, SupplementaryData), indicating a negligible membranefouling. The233

low membranefouling could be attributedto the small water flux and high aerationrate234

appliedin this study.Over 60 daysof experiment,the water flux of MF wasadjustedfrom235

1.6 to 2.6 L/m2h. By consideringthe gradual flux decline in the FO process,this flow236

adjustmentwas necessaryto keepa constantHRT of 24 h during the entire experimental237

period.On theotherhand,the low MLSS concentrationin thebioreactor(2 – 3.3 g/L) could238

alsominimizethemembranefouling.239

3.1.2 Biological performance240

Biological performanceof the integratedO/MF-MBR systemwas assessedwith regardsto241

the removal of basic contaminants(i.e. TOC, TN, and PO43--P), sludgeproduction,and242

biological activity. The removalof basiccontaminantswasstableafter a short-term salinity243

build-up in the bioreactor(Fig. 3). The stableremovalcanalsobe determinedby the small244

standarddeviation of thesecontaminantconcentrationsin different units of O/MF-MBR245

during thecourseof theexperiment(TableS2,SupplementaryData).246

10

Dueto thehigh rejectionof theFOmembrane,permeatequality of FOwassuperiorto thatof247

MF, particularly in termsof TN andPO43--P concentrations(Fig. 3). The removalof TOC248

from theOMBR channelwasover98%during the entireexperimentalperiod(Fig. 3a).The249

result is consistentwith that reportedby Hancocket al. (2013). Given the excellentremoval250

of TOC from the bioreactor (indicated by low TOC concentrationin the mixed liquor251

supernatant),the benefitsof FO over MF werenot significant.However,the removalof TN252

throughthe MF-MBR channelonly varied in the rangeof 20 – 65%, with relatively high253

concentrationin the permeate(Fig. 3b). Since the removal of TN in aerobicbioreactors254

occursmainly via assimilationto the biomass(Hai et al., 2014), it was not surprisedto255

observethe relatively low andunstableremoval.By contrast,TN removalfrom the OMBR256

channelrangedfrom 60 to 90%, althoughtherewasa smalldeclinefrom day40 onward.This257

declinewaslikely dueto theincompleterejectionof NH4+-N andaccumulatedNOx

--N by the258

FO membrane(Irvine et al., 2013; Luo et al., 2015). A small andvariableremovalthrough259

the MF-MBR channelwas also observedfor PO43- -P (Fig. 3c), possibly due to the low260

biomassassimilationand/or phosphorusprecipitationunderthe nearlyneutralpH condition261

in thebioreactor(Qiu andTing, 2014). Nevertheless,PO43--P couldnot bedetectedin theFO262

permeate.Indeed,the FO membranecanalmostcompletelyretainPO43- -P dueto the large263

hydratedradiusandnegativechargeof theorthophosphateions(Holloway et al.,2007).264

[FIGURE 3]265

The MLSS concentrationgradually increasedwith time after a slight decreasein the first266

week(Fig. 4). Thesmall decreasein the MLSS concentrationat the beginningwaspossibly267

due to the inhibitory effects of the elevatedbioreactorsalinity on microbial mass. This268

inhibition wasalsoevidencedby a reductionin biomassactivity asindicatedby theSOURof269

the sludge (Fig. S3, SupplementaryData). With the bioreactorsalinity stabilizing at a270

relatively low level (0.4 g/L NaCl), the sludgeconcentrationin the bioreactorincreased271

graduallywith theMLVSS/MLSSratio of 0.75± 0.05from day7 onward.At thesametime,272

theSOURof thesludgealsoincreasedandsubsequentlylevelledoff at 4.5mg O2/g MLVSS273

h. This stableSOURvalueis in goodagreementwith thatreportedpreviouslyin conventional274

MBRs(Hanetal., 2005; Choi et al.,2007).275

[FIGURE 4]276

3.2 Removalof traceorganicchemicals277

11

The removal of most TrOCs selectedhere was stable during the entire courseof the278

experiment (Fig. 5). There are only six exceptions,namely, clofibric acid, atrazine,279

carbamazepine,propoxur,diclofenacand fenoprop. The removalof thesesix compoundsis280

shownas a function of time in Fig. 6. During biological treatment,TrOC removalcan be281

evaluatedusinga qualitativepredictiveframeworkdevelopedby Tadkaewet al. (2011)based282

on their molecularproperties,suchashydrophobicityandfunctionalgroups.Accordingto the283

scheme,TrOCs investigatedin this studycould be generallyclassifiedas hydrophobic(i.e.284

Log D pH 7 > 3) or hydrophilic(i.e.Log D pH 7 < 3) (section2.1).285

[FIGURE 5]286

[FGIURE 6]287

3.2.1HydrophobicTrOCs288

Of 30 TrOCsselectedin this study,all elevenhydrophobiccompoundscould be effectively289

removed(> 85%) from both OMBR andMF-MBR channels(Fig. 5). Previousstudieshave290

demonstratedtheexcellentremovalof thesehydrophobicTrOCs duringbiological treatment291

(Radjenovi� et al., 2009; Nguyen et al., 2012). Due to the high hydrophobicityof these292

compounds,they can easily absorb on the activatedsludge and thereby facilitate their293

biodegradation(transformation)in the bioreactor(Tadkaewet al., 2011). As a result,apart294

from bisphenolA andoctocrylene,therewasno cleardifferencein theconcentrationof these295

hydrophobicTrOCs betweenthe MF and FO permeates(Fig. 5). It is noteworthy that296

bisphenolA andoctocryleneconcentrationsin theFO permeatewerehigherthanthosein the297

MF permeate.Their high concentrationsin the FO permeatewere possibly due to cake-298

enhancedconcentrationpolarizationcausedby the foulant layer on the membranesurface299

(Vogel et al., 2010). These two compoundsare hydrophobic.Thus, their accumulation300

adjacentto the membranesurfacedue to cake-enhancedconcentrationpolarizationcould301

enhancetheir transportacrosstheFO membranevia hydrophobicinteractions(Nghiemet al.,302

2004). Furtherstudiesarenecessaryto ascertainthe effectsof the sludgecakelayer on the303

rejectionof TrOCs,particularlythehydrophobiccompounds,in theFOprocess.304

3.2.2HydrophilicTrOCs305

Significantvariationin theremovalof hydrophilicTrOCswasobservedfrom bothMF-MBR306

andOMBR channels. By accountingfor therelativelylargeporesof theMF membrane, their307

removalthroughtheMF-MBR channelwasmainly governedby theactivatedsludge.Indeed,308

previous studieshave shown a large variation in the removal of hydrophilic TrOCs in309

12

conventionalMBRs, which was determinedby their intrinsic biodegradabilitydue to their310

weakadsorptionontobiosolids(Tadkaewet al., 2011). In this study,the removalof six very311

hydrophilic TrOCs(i.e. Log D pH 7 < 1), including salicylic acid,metronidazole,ketoprofen,312

naproxen,primidone, and ibuprofen,was higher than 85% throughthe MF-MBR channel.313

The excellent removal of thesecompoundscould be attributedto the presenceof strong314

electrondonatingfunctionalgroups,suchasamineandhydroxyl groups,in their molecular315

structures(Table S1). Containingthesefunctional groupsallowed compoundseasily to be316

electrophilically attacked by oxygenasesfrom the aerobicbacteria.The oxygenasesarekey317

reactantsresponsiblefor biodegradationof organic compounds(Kanazawaet al., 2003;318

Tadkaewet al., 2011). Sincethesehydrophilic TrOCscould be effectively removedin the319

bioreactor,the benefitsof FO over MF were not significant (Fig. 5). It is noted that the320

removalof salicylic acid from the OMBR channelwasslightly lower than that throughthe321

MF-MBR channel. Theexactreasonis still unclearbut it couldbeattributedto theeffectsof322

cake-enhancedconcentrationpolarizationin theFOprocessasnotedabove.323

Dueto thehigh rejectionof the FO membrane,the removalthroughtheOMBR channelwas324

moreeffectivethanthat from the MF-MBR channelfor the six hydrophilicTrOCsshownin325

Fig. 6. The removalof thesecompoundswaslow andhighly variablethrough theMF-MBR326

channelbecauseof their resistanceto biodegradation.Tadkaewet al. (2011)haveattributed327

their low biodegradationto the presenceof one or more strong electron withdrawing328

functionalgroup(e.g.chlorine,amideandnitro groups)and/ortheabsenceof strongelectron329

donatingfunctional groupsin their molecularstructures. Despitethe low removalof these330

compoundsin the bioreactor,their high rejection by the FO membraneensuredexcellent331

removalfrom theOMBR channel.Thebenefitsof theFOmembranefor TrOC rejectionhave332

alreadybeenhighlightedin severalrecentstudies(Alturki et al.,2013; Codayet al., 2014).333

With the exceptionof clofibric acid, the rejection of thesehydrophilic and biologically334

persistentTrOCsby theFO membraneincreasedtheir permeationthroughtheMF membrane335

andthusreducedtheremovalby theMF-MBR channel(Fig. 6). Theremovalof clofibric acid336

via the MF-MBR channel gradually increasedwith time, althoughsomefluctuationswere337

observed. The reasonfor this phenomenonis not clear, possibly due to an enhanced338

biodegradationwith the increasedMLSS concentrationin the bioreactor(Cirja et al., 2008).339

Of six biologically persistentcompoundsnotedabove,theremovalof atrazineby the OMBR340

channel was also observed to decreasegradually with time. Atrazine has moderate341

hydrophobicity(Log DpH 7 = 2.6), andthusthe observedlow andreducedremovalcould be342

13

attributed to its adsorptionand partitioning into the membranesurface followed by a343

diffusion throughthemembrane(Nghiemetal., 2004).344

4. Conclusion345

This studycomparedthe waterquality of the FO andMF permeatesin an integratedO/MF-346

MBR systemregardingtheconcentrationof TOC,TN, PO43--P andTrOCs.TheFO permeate347

had a higher water quality than the MF permeatedue to the effective rejectionof the FO348

membrane.Theconcentrationof hydrophobicTrOCsandhydrophiliccompoundscontaining349

strongelectrondonatingfunctional groupswas low in both MF and FO permeatesas they350

could be well removedby the activatedsludge.However,the concentrationof hydrophilic351

and biologically persistentTrOCs which containedstrongelectronwithdrawing functional352

groupsin theFO permeatewasmuchlower thanthat in theMF permeate. In addition,dueto353

the high rejectionof the FO membrane,thesehydrophilic andbiologically persistentTrOCs354

couldaccumulatein the bioreactorandbe transferredinto theMF permeate.Thus,thewater355

flux ratio betweenMF andFO canbeoptimisedto reducesalinity build-up in thebioreactor356

while ensuringadequateMF permeatequality.357

5. Acknowledgement358

This researchwassupportedunderAustralianResearchCouncil’sDiscoveryProjectfunding359

scheme(project DP140103864).The authorswould like to thank the ChineseScholarship360

CouncilandtheUniversityof Wollongongfor thePhDscholarshipsupportto WenhaiLuo.361

6. References362

Achilli, A., Cath, T.Y., Marchand, E.A., Childress, A.E. 2009. The forward osmosis363

membranebioreactor:A low fouling alternativeto MBR processes.Desalination,239 (1-364

3), 10-21.365

Alturki, A.A., McDonald, J., Khan, S.J., Hai, F.I., Price, W.E., Nghiem, L.D. 2012.366

Performanceof a novelosmoticmembranebioreactor(OMBR) system:Flux stabilityand367

removalof traceorganics.Bioresour.Technol.,113,201-206.368

Alturki, A.A., McDonald,J.A., Khan,S.J.,Price,W.E., Nghiem,L.D., Elimelech,M. 2013.369

Removal of trace organic contaminantsby the forward osmosisprocess.Sep. Purif.370

Technol.,103,258-266.371

Boo, C., Lee, S., Elimelech,M., Meng, Z., Hong, S. 2012. Colloidal fouling in forward372

osmosis:Roleof reversesaltdiffusion.J.Membr.Sci.,390-391,277-284.373

14

Choi,J.H.,Lee,S.H.,Fukushi,K., Yamamoto,K. 2007.Comparisonof sludgecharacteristics374

andPCR-DGGEbasedmicrobialdiversityof nanofiltrationandmicrofiltration membrane375

bioreactors.Chemosphere,67 (8), 1543-1550.376

Cirja, M., Ivashechkin,P., Schäffer, A., Corvini, P.F.X.2008.Factorsaffectingthe removal377

of organicmicropollutantsfrom wastewaterin conventionaltreatmentplants(CTP) and378

membranebioreactors(MBR). Rev.Environ.Sci.Biotechnol.,7 (1), 61-78.379

Coday,B.D., Yaffe, B.G.M., Xu, P., Cath,T.Y. 2014.Rejectionof traceorganiccompounds380

by forwardosmosismembranes:A literaturereview.Environ.Sci.Technol.,48 (7), 3612-381

3624.382

Cornelissen,E.R., Harmsen,D., Beerendonk,E.F., Qin, J.J., Oo, H., de Korte, K.F.,383

Kappelhof,J.W.M.N. 2011. The innovativeosmoticmembranebioreactor(OMBR) for384

reuseof wastewater.WaterSci.Technol.,63 (8), 1557-1565.385

Drewes,J.E.,Anderson,P., Denslow,N., Olivieri, A., Schlenk,D., Snyder,S.A., Maruya,386

K.A. 2013.Designingmonitoringprogramsfor chemicalsof emergingconcernin potable387

reuse: Whatto includeandwhatnot to include?WateSci.andTech.,67 (2), 433-439.388

Grelier, P., Rosenberger,S., Tazi-Pain, A. 2006. Influence of sludge retention time on389

membranebioreactorhydraulicperformance. Desalination,192(1-3), 10-17.390

Hai, F.I., Tessmer,K., Nguyen,L.N., Kang,J.,Price,W.E., Nghiem,L.D. 2011.Removalof391

micropollutantsby membranebioreactorundertemperaturevariation.J. Membr.Sci.,383392

(1), 144-151.393

Hai, F.I., Yamamoto,K., Lee,C.H. 2014.MembraneBiological Reactors:Theory,Modeling,394

Design,ManagementandApplicationsto WastewaterReuse.IWA Publishing,London.395

Han, S.S.,Bae, T.H., Jang,G.G., Tak, T.M. 2005. Influence of sludgeretentiontime on396

membranefouling andbioactivities in membranebioreactorsystem.ProcessBiochem.,40397

(7), 2393-2400.398

Hancock,N.T., Xu, P., Roby, M.J., Gomez,J.D., Cath,T.Y. 2013.Towardsdirect potable399

reusewith forward osmosis:Technicalassessmentof long-term processperformanceat400

thepilot scale.J.Membr.Sci.,445,34-46.401

Hochstrat,R., Wintgens,T., Kazner,C., Melin, T., Gebel,J. 2010.Optionsfor waterscarcity402

anddroughtmanagement-theroleof desalination.Desalin.WaterTreat.,18 (1-3), 96-102.403

Hoek, E.M.V., Elimelech, M. 2003. Cake-enhancedconcentrationpolarization: A new404

fouling mechanismfor salt-rejectingmembranes.Environ. Sci. Technol.,37 (24), 5581-405

5588.406

15

Holloway, R.W., Childress,A.E., Dennett, K.E., Cath, T.Y. 2007. Forward osmosisfor407

concentrationof anaerobicdigestercentrate.WaterRes.,41 (17),4005-4014.408

Holloway, R.W., Regnery,J., Nghiem, L.D., Cath, T.Y. 2014. Removalof trace organic409

chemicals and performance of a novel hybrid ultrafiltration-osmotic membrane410

bioreactor.Environ.Sci.Technol.,48 (18), 10859-10868.411

Holloway, R.W., Wait, A.S., Fernandesda Silva, A., Herron,J., Schutter,M.D., Lampi, K.,412

Cath, T.Y. 2015. Long-term pilot scale investigationof novel hybrid ultrafiltration-413

osmoticmembranebioreactors.Desalination,363,64-74.414

Irvine, G.J.,Rajesh,S.,Georgiadis,M., Phillip, W.A. 2013.Ion selectivepermeationthrough415

celluloseacetatemembranesin forwardosmosis.Environ.Sci. Technol.,47 (23), 13745-416

13753.417

Kanazawa,H., Okada, A., Higaki, M., Yokota, H., Mashige, F., Nakahara,K. 2003.418

Stereospecificanalysis of omeprazolein human plasma as a probe for CYP2C19419

phenotype.J.Pharm.Biomed.Anal., 30 (6), 1817-1824.420

Lampard,J., Leusch,F.D.L., Roiko, A., Chapman,H.F. 2010.Contaminantsof concernin421

recycledwater.Water,37 (8), 54-60.422

Lay, W.C.L., Liu, Y., Fane,A.G. 2010. Impactsof salinity on the performanceof high423

retentionmembranebioreactorsfor waterreclamation:A review.WaterRes.,44 (1), 21-424

40.425

Lay, W.C.L., Zhang,Q., Zhang,J.,McDougald,D., Tang,C., Wang,R., Liu, Y., Fane,A.G.426

2012. Effect of pharmaceuticalson the performanceof a novel osmotic membrane427

bioreactor(OMBR). Sep.Purif. Technol.,47 (4), 543-554.428

Luo, W., Hai, F.I.,Price,W.E.,Nghiem,L.D. 2015.Waterextractionfrom mixedliquor of an429

aerobicbioreactorby forward osmosis:Membranefouling and biomasscharacteristics430

assessment.Sep.Purif. Technol.,145,56-62.431

Luo, Y., Guo,W.S.,Ngo, H.H., Nghiem,L.D., Hai, F.I., Zhang,J., Liang, S.,Wang,X.C.C.432

2014.A reviewon theoccurrenceof micropollutantsin theaquaticenvironmentandtheir433

fateandremovalduringwastewatertreatment.Sci.TotalEnviron.,473,619-641.434

McCutcheon,J.R.,Elimelech,M. 2008.Influenceof membranesupportlayerhydrophobicity435

on water flux in osmoticallydriven membraneprocesses.J. Membr. Sci., 318 (1), 458-436

466.437

Nawaz,M.S.,Gadelha,G., Khan,S.J.,Hankins,N. 2013.Microbial toxicity effectsof reverse438

transporteddraw solute in the forward osmosismembranebioreactor (FO-MBR). J.439

Membr.Sci.,429,323-329.440

16

Nghiem, L.D., Coleman,P.J.2008.NF/ROfiltration of thehydrophobicionogeniccompound441

triclosan: Transportmechanismsand the influence of membranefouling. Sep. Purif.442

Technol.,62 (3), 709-716.443

Nghiem,L.D., Schäfer,A.I., Elimelech,M. 2005.Nanofiltrationof hormonemimicking trace444

organiccontaminants.Sep.Purif. Technol.,40 (13),2633-2649.445

Nghiem, L.D., Schäfer, A.I., Elimelech, M. 2004. Removal of natural hormonesby446

nanofiltration membranes:Measurement,modeling, and mechanisms.Environ. Sci.447

Technol., 38 (6), 1888-1896.448

Nguyen,L.N., Hai, F.I., Kang,J.,Price,W.E., Nghiem,L.D. 2012.Removalof traceorganic449

contaminantsby a membranebioreactor-granularactivatedcarbon(MBR-GAC) system.450

Bioresour.Technol.,113,169-173.451

Qiu, G.L., Ting, Y.P. 2014. Direct phosphorusrecovery from municipal wastewatervia452

osmotic membranebioreactor(OMBR) for wastewatertreatment.Bioresour.Technol.,453

170,221-229.454

Qiu, G.L., Ting, Y.P. 2013.Osmoticmembranebioreactorfor wastewatertreatmentandthe455

effect of salt accumulationon systemperformanceandmicrobial communitydynamics.456

Bioresour.Technol.,150,287-297.457

Radjenovi�, J., Petrovi�, M., Barceló,D. 2009.Fateanddistribution of pharmaceuticalsin458

wastewaterandsewagesludgeof the conventionalactivatedsludge(CAS) andadvanced459

membranebioreactor(MBR) treatment.WaterRes.,43 (3), 831-841.460

Tadkaew,N., Hai, F.I., McDonald,J.A., Khan, S.J.,Nghiem,L.D. 2011.Removalof trace461

organicsby MBR treatment:The role of molecularproperties.WaterRes., 45 (8), 2439-462

2451.463

Tanoue,R., Sato,Y., Motoyama,M., Nakagawa,S., Shinohara,R., Nomiyama,K. 2012.464

Plant uptake of pharmaceuticalchemicalsdetectedin recycled organic manure and465

reclaimedwastewater.J.Agric. Food.Chem.,60 (41),10203-10211.466

Van der Bruggen,B., Patricia, L. 2015. Forward osmosis:Understandingthe hype. Rev.467

Chem.Eng.,31 (1), 1-12.468

Vogel, D., Simon,A., Alturki, A.A., Bilitewski, B., Price,W.E., Nghiem,L.D. 2010.Effects469

of fouling andscalingon the retentionof traceorganic contaminantsby a nanofiltration470

membrane:Therole of cake-enhancedconcentrationpolarisation.Sep.Purif. Technol.,73471

(2), 256-263.472

17

Wang,X., Chen,Y., Yuan, B., Li, X., Ren,Y. 2014a.Impactsof sludgeretentiontime on473

sludge characteristicsand membrane fouling in a submergedosmotic membrane474

bioreactor.Bioresour.Technol.,161,340-347.475

Wang, X., Yuan, B., Chen,Y., Li, X., Ren, Y. 2014b. Integrationof micro-filtration into476

osmotic membranebioreactorsto prevent salinity build-up. Bioresour.Technol., 167,477

116-123.478

Wintgens,T., Salehi,F., Hochstrat,R., Melin, T. 2008.Emergingcontaminantsandtreatment479

optionsin waterrecyclingfor indirectpotableuse.WaterSci.Technol.,57 (1), 99-107.480

Wu, C., Spongberg,A.L., Witter, J.D., Sridhar, B.B.M. 2012. Transfer of wastewater481

associatedpharmaceuticalsand personalcare products to crop plants from biosolids482

treatedsoil. Ecotoxicol.Environ.Safety,85,104-109.483

18

LIST OF CAPTIONS

Fig. 1: Schematicdiagramof a laboratory-scaleintegratedO/MF-MBR hybrid system.Draw

solution was replaceddaily to avoid overflow and contaminantaccumulationin the draw

solution reservoir.High concentrateddraw solution was addedmanuallyon a daily basis.

Sampleswere taken from feed, bioreactor,MF permeate,and draw solution reservoirfor

analysis.

Fig. 2: Variation of mixed liquor conductivity and FO water flux with time. Experimental

conditions:HRT = 24 h; DO concentration= 5 ± 1 mg/L; draw solution = 58.5 g/L NaCl;

cross-flow rate= 1 L/min (i.e. cross-flow velocity = 9 cm/s);FO mode;temperature= 22 ± 1

°C. Waterflux of MF wasadjustedfrom 1.6 to 2.6 L/m2h to compensatethe flux declineof

FOto keepaconstantbioreactorworkingvolumeandHRT.

Fig. 3: Removalof (a) TOC, (b) TN, and(c) PO43--P by OMBR andMF-MBR channelsof

theintegratedO/MF-MBR system.

Fig. 4: Variationof biomassconcentrationin thebioreactorwith time.

Fig. 5: MeasuredTrOC concentrationsin the feed,MF andFO permeates,andtheir removal

by MF-MBR andOMBR channelsof an integratedO/MF-MBR system.Error barsrepresent

thestandarddeviationof eightmeasurements(onceaweek).

Fig. 6: Time-dependentremovalof six hydrophilic and biologically persistentTrOCs (i.e.

diclofenac,atrazine,carbamazepine,propoxur,fenopropandclofibric acid) via MF-MBR and

OMBR channelsof theintegratedO/MF-MBR system.

19

Peristaltic pump

Computer

High Conc.tank

Biological reactor

Circulating pump

Flowmeter

Flowmeter

Balance

FO membrane cell

Feed Drawsolution

MF

MF permeate

Mixed liquor

Cond. controller

Peristaltic pump

Peristaltic pump

Figure 1

20

0 10 20 30 40 50 600

200

400

600

800

1000

1200

Mixedliquorconductivity(�S/cm)

Time (d)

Mixed liquor conductivity

0

1

2

3

4

5

6

7

FO water flux

FOwaterflux(L/m

2 h)

Figure 2

21

0

2

4

6

20

40

60

80

100

0

5

10

15

20

25

30

35

40

0 10 20 30 40 50 600

2

4

6

8

10

12

14

(a)

(b)

(c)

Concentration in: Feed Mixed liquor supernatantMF permeate FO permeate

Removal by: MF-MBR channel OMBR channel

TOC(mg/L)

0

20

40

60

80

100

Removal(%)

TN(mg/L)

0

20

40

60

80

100

Removal(%)

PO

43--Pconcentration(mg/L)

Time (d)

0

20

40

60

80

100

Removal(%)

Figure 3

22

0 10 20 30 40 50 600.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

Biomassconcentration(g/L)

Time (d)

MLSSMLVSS

0.0

0.2

0.4

0.6

0.8

1.0

MLVSS/MLSS ratioMLVSS/MLSSratio

Figure 4

23

Salicy

licac

id

Clofibr

icac

id

Metro

nidaz

ole

Fenop

rop

Ketopr

ofen

Napro

xen

Primido

ne

Ibupr

ofen

Propo

xur

Diclof

enac

Enter

olacto

ne

Carba

mazep

ine

Gemfib

rozil

Amitri

ptyli

ne

DEET

Estriol

Atrazin

e

Pentac

hloro

phen

ol

Ametryn

Benzo

phen

one

4-ter

t-But

ylphe

nol

Estron

e

Bisphe

nolA

Oxybe

nzon

e

17-e

thyny

lestra

diol

17-e

strad

iol

-Estr

adiol

-17-

acet

ate

4-ter

t-Octy

lphen

ol

Triclos

an

Octocr

ylene

0

1000

2000

3000

4000

5000Hydrophobic (Log D > 3)Hydrophilic (Log D < 3)

�

Concentration(ng/L)

Concentration in: Feed MF permeate FO permeate

0

20

40

60

80

100

�

�

Removal by: MF-MBR channel OMBR channel

Removal(%)

-1

0

1

2

3

4

5

6

7Log D

LogDatpH=7

Figure 5

24

0

20

40

60

80

100

0

20

40

60

80

100

0 10 20 30 40 50 600

20

40

60

80

100

0 10 20 30 40 50 60

Diclofenac MF-MBR channel OMBR channel

Removal(%)

Atrazine

Carbamazepine

Removal(%)

Propoxur

Fenoprop

Removal(%)

Time (d)

Clofibric acid

Time (d)

Figure 6

The role of forward osmosisand microfiltration in an integrated

osmotic-microfiltration membranebioreactor system

SupplementaryData

Revisedmanuscript submitted to Chemosphere

April 2015

WenhaiLuo a, FaisalI. Hai a, JinguoKangb, William E. Priceb, LongD. Nghiema*,

MenachemElimelechc

a StrategicWaterInfrastructureLaboratory, Schoolof Civil Mining andEnvironmental

Engineering,Universityof Wollongong,Wollongong,NSW2522,Australia

b StrategicWaterInfrastructureLaboratory,Schoolof Chemistry,Universityof Wollongong,

Wollongong,NSW2522,Australia

c Departmentof ChemicalandEnvironmentalEngineering,YaleUniversity,NewHaven,CT06520-8286,UnitedStates

* Correspondingauthor:longn@uow.edu.au; Ph:+61(2) 42214590.

TableS1:Physicochemicalpropertiesof theselectedtraceorganicchemicals.

CompoundsChemicalformula

Log Dat pH = 7

MW(g/mol)

Chemicalstructure

Salicylicacid C7H6O3 -1.13 138.1

Clofibric acid C10H11ClO3 -1.06 214.6

Metronidazole C6H9N3O3 -0.14 171.2

Fenoprop C9H7Cl3O3 -0.13 269.5

Ketoprofen C16H14O3 0.19 254.3

Naproxen C14H14O3 0.73 230.3

Primidone C12H14N2O2 0.83 218.3

Ibuprofen C13H18O2 0.94 206.3

Propoxur C11H15NO3 1.54 209.2

Diclofenac C14H11Cl2NO2 1.77 296.2

Enterolactone C18H18O4 1.89 298.33

Carbamazepine C15H12N2O 1.89 236.3

Gemfibrozil C15H22O3 2.07 250.3

Amitriptyline C20H23N 2.28 277.4

DEET C12H17NO 2.42 191.3

Estriol C18H24O3 2.53 288.4

Atrazine C8H14ClN5 2.64 215.7

PentachlorophenolC6HCl5O 2.85 266.4

Ametryn C9H17N5S 2.97 227.3

Benzophenone C13H10O 3.21 182.2

4-tert-Butylphenol C10H14O 3.4 150.2

Estrone C18H22O2 3.62 270.4

BisphenolA C15H16O2 3.64 228.3

Oxybenzone C13H10O 3.89 228.2

17� -ethynylestradiol

C20H24O2 4.11 296.4

17� -estradiol C18H24O2 4.15 272.4

� -Estradiol 17-acetate C20H26O3 5.11 314.4

4-tert-Octylphenol C14H22O 5.18 206.3

Triclosan C12H7Cl3O2 5.28 289.5

Octocrylene C24H27N 6.89 361.5

Source:SciFinderScholar(ACS)database.

TableS2: Basicwaterquality in differentunitsof O/MF-MBR (average± standard

deviation*)

WaterparametersContaminantconcentration(mg/L)

Feed Mixed liquor supernatant MF permeate FOpermeateTOC 71.4± 9.6 2.7± 1.2 2.2± 1.0 1.7± 0.8TN 18.3± 4.9 14.3± 4.3 12.2± 4.1 5.3± 3.5

NH4+-N 10.5± 1.7 2.5± 1.4 1.0± 0.4 0.6± 0.3

PO43--P 10.9± 1.1 9.1± 1.5 8.9± 1.6 0.0± 0.0

*Standarddeviationwascalculatedfrom 20 measurements(onceevery3 days).

Fig. S1: Photographof the FO membranesurfaceat the conclusionof the experiment.

Membranecleaningwas not conducted.Experimentalcondition: CTA-FO membrane;FO

mode;draw solution = 1 M NaCl; cross-flow rate = 1 L/min (i.e. cross-flow velocity = 9

cm/s);HRT = 24 h; DO concentration = 5 ± 1 mg/L; temperature= 22 ± 1 °C; MF waterflux

= 1.6– 2.6mL/min.

0 10 20 30 40 50 600

1

2

3

4

5

6

TMPprofileofMF(kPa)

Time (d)

Fig. S2: TheTMP profile of theMF membranewith time.

0 10 20 30 40 50 600

1

2

3

4

5

SOUR(mgO

2/gMLVSSh)

Time (d)

Fig. S3: SOURof theactivatedsludgewith time.