Wastewater Purification: Aerobic Granulation in Sequencing Batch Reactors - Chapter 9 pdf

149
9
The Essential Role
of Cell Surface
Hydrophobicity in
Aerobic Granulation
Yu Liu and Zhi-Wu Wang
CONTENTS
9.1 Introduction 149
9.2 Cell Surface Hydrophobicity 150
9.2.1 What Is Hydrophobicity? 150
9.2.2 Cell Surface Property-Associated Hydrophobicity 151
9.2.2.1 Surface Properties of Amino Acids 151
9.2.2.2 Surface Properties of Proteins 151
9.2.2.3 Surface Properties of Polysaccharides 151
9.2.2.4 Surface Properties of Phospholipids 151
9.2.3 Determination of Cell Surface Hydrophobicity 152
9.3 The Role of Cell Surface Hydrophobicity in Aerobic Granulation 152
9.4 Factors Inuencing Cell Surface Hydrophobicity 156
9.5 Selection Pressure-Induced Cell Surface Hydrophobicity 160
9.6 Thermodynamic Interpretation of Cell Surface Hydrophobicity 161
9.7 Enhanced Aerobic Granulation by Highly Hydrophobic
Microbial Seed 170
9.8 Conclusions 176
References 176
9.1 INTRODUCTION
Aerobic granulation is a process of cell-to-cell self-immobilization that results
inaformofregularshape.Inviewofmasstransferandutilizationofsubstrate,
bacteriaindeedwouldpreferadispersedratherthanaggregatedstate.Thereshould
be triggering forces that can bring bacteria together and further make them aggre-
gate. It appears from the preceding chapters that cell hydrophobicity induced by

cultureconditionscanserveasatriggeringforceforaerobicgranulation.Infact,
it has been well known that the physicochemical properties of the cell surface
haveprofoundeffectsontheformationofbiolmsandbothanaerobicandaerobic
granules (Bossier and Verstraete 1996; Zita and Hermansson 1997; Kos et al. 2003;
53671_C009.indd 149 10/29/07 7:27:18 AM
© 2008 by Taylor & Francis Group, LLC
© 2008 by Taylor & Francis Group, LLC
150 Wastewater Purification
Liuetal.2004b).Whenbacteriabecamemorehydrophobic,increasedcell-to-cell
adhesionwasobserved,thatis,cellsurfacehydrophobicitymaycontributetothe
ability of cells to aggregate (Kjelleberg, Humphrey, and Marshall 1983; Del Re
etal.2000;Kosetal.2003;Liuetal.2004b).Thischapterlooksattheroleofcell
surfacehydrophobicityintheformationofaerobicgranularsludgeinasequencing
batch reactor (SBR).
9.2 CELL SURFACE HYDROPHOBICITY
9.2.1
WHAT IS HYDROPHOBICITY?
Hydrophobicity attraction is the strongest binding force occurring between parti-
cles or polymers immersed in water. The attraction between two apolar surfaces,
or between one apolar and one polar surface, in water, is traditionally called the
hydrophobiceffect.Hydrophobicsurfacesdonotrepelwaterbutinsteadattract
water (Hildebrand1979).Becauseofwaterhydrogenbonds,watermoleculesoften
presentintheformofwaterclusters(gure9.1),andthesizeoftheseclusterstends
to decrease with increase of temperature. The classical macroscopic scale inter-
actions between apolar and/or polar surfaces, immersed in a liquid, have been often
described by the well-known DLVO theory, which shows apolar Lifshitz–van der
Waals(LW)attractionandelectricaldoublelayer(EL)repulsionasafunctionof
distance. It can be shown that hydrophobic interaction becomes the main driving
forcewhichrepresentsnearlyallthemacro-scaleinteractionsinwaterintermsof
FIGURE 9.1 Illustrationofwatermoleculescluster.(FromChaplin,M.F.2000.Biophys

Chem 83:211–221.Withpermission.)
53671_C009.indd 150 10/29/07 7:27:19 AM
© 2008 by Taylor & Francis Group, LLC
© 2008 by Taylor & Francis Group, LLC
The Essential Role of Cell Surface Hydrophobicity in Aerobic Granulation 151
attraction or repulsion (Bergendahl et al 2002). Hydrophilic repulsion occurs only
whenpolarmolecules,particles,orcellsattractwatermoleculesmorestronglythan
the acid-base (AB) cohesive attraction between water molecules.
9.2.2 CELL SURFACE PROPERTY-ASSOCIATED HYDROPHOBICITY
Most biological surfaces have a low H
+
intheorderof0.1mJm
–2
.Thecellsurfaceiscom-
posed mainly of proteins, polysaccharides, and phospholipids. The combination charac
-
teristics of these substances in turn determine the overall cell surface hydrophobicity.
9.2.2.1 Surface Properties of Amino Acids
AccordingtoParker,Guo,andHodges(1986),theorderofaminoacidsidechains
beginning with the most hydrophobic can probably be summarized as follows: Trp,
Phe, Leu, Ile, Met, Val, Tyr, Cys, Ala, Pro, His, Arg, Thr, Lys, Gly, Glu, Ser, Asx,
Glu, Asp, where the amino acids to the right of Thr are more hydrophilic. It is evident
thatanaminoacidwithalargerhydrophobicsidechainismorehydrophobicthan
thosewithasmallhydrophobicsidechain.Thisseemstoindicatethatthesurface
propertyofaminoacidscansignicantlyinuencethecellsurfacehydrophobicity.
9.2.2.2 Surface Properties of Proteins
Proteins are made up of hydrophobic and/or hydrophilic amino acids. For water-
soluble protein, the majority of its hydrophilic amino acids presents at the water
interface, whereas the more hydrophobic amino acids are located inside the three-
dimensional framework of the macromolecule. However, once protein has made

contactwithahydrophobicsurface,itcanorientitsmosthydrophobicsitestothe
hydrophobicinterface(Leeetal.1973;vanOss1994a).Thisseemstoindicatethat
some proteins can shift between hydrophobicity and hydrophilicity, depending on
actual conditions.
9.2.2.3 Surface Properties of Polysaccharides
In contrast with protein that comprises hydrophilic and/or hydrophobic amino acids,
polysaccharides are made up of different sugars that are hydrophilic and soluble
inwater(vanOss1995).Obviously,thehighsolubilityofpolysaccharidesinwater
meansalowhydrophobicity.Aspolymericsubstances,thesesugarsmaybecome
more hydrophilic or more hydrophobic, depending on the structure of the polymer
molecule. It has been reported that the amount of extracellular polymers affects
the contribution of electrostatic interaction to cell attachment onto a solid surface
(Tsunedaetal.2003).Furthermore,inastudyofhydrophobicandhydrophilic
properties of activated sludge, it was found that a signicant portion of extracellular
polymersarehydrophobic(Jorandetal.1998).Likely,extracellularpolymer-induced
cell surface hydrophobic changes may be fundamental in microbial aggregation.
9.2.2.4 Surface Properties of Phospholipids
The general structure of biological membranes is a phospholipids bilayer. Phospho-
lipids containbothhighly hydrophobic(fatty acid) andrelatively hydrophilic (glycerol)
53671_C009.indd 151 10/29/07 7:27:19 AM
© 2008 by Taylor & Francis Group, LLC
© 2008 by Taylor & Francis Group, LLC
152 Wastewater Purification
moietiesandcanexistinmanydifferentchemicalformsasaresultofvariationin
the nature of the fatty acids or phosphate-containing groups attached to the glycerol
backbone (Madigan, Martinko, and Parker 2003). As phospholipids aggregate in an
aqueous solution, they tend to form a bilayer structure spontaneously with the fatty
acidsinahydrophobicenvironment,andthehydrophilicportionsremainexposedto
the aqueous external environment. Saturated alkyl chains of phospholipids can attract
eachotherstronglyinwater,withahydrophobicenergyofattractionof–102mJm

–2
in all cases (van Oss 1994b). The major proteins of the cell membrane generally have
veryhydrophobicexternalsurfacesintheregionsoftheproteinthatspanthemem
-
brane and have hydrophilic surfaces exposed on both the inside and the outside of
thecell.Theoverallstructureofthecytoplasmicmembraneisstabilizedbyhydrogen
bonds and hydrophobic interactions (Madigan, Martinko, and Parker 2003).
9.2.3 DETERMINATION OF CELL SURFACE HYDROPHOBICITY
Thereareanumberofmethodsavailabletocharacterizethecellsurfacehydropho-
bicity, including contact angle measurement, bacterial adherence to hydrocarbons,
hydrophobic interaction chromatography, salting out aggregation, adhesion to solid
surfaces,andbindingoffattyacidstobacterialcells(RosenbergandKjelleberg
1986;MozesandRouxhet1987).Contactanglemeasurementisthetraditionaland
widely used method, and it involves measuring the contact angle of a sessile drop
withaatbacteriaxedlteroralawnofthebacteriaonanagarplate(Mozesand
Rouxhet1987).Asthecellsurfacemoisturedecreaseswithevaporation,thecon
-
tact angle increases over time. The stationary-phase contact angle is often used to
characterizethecellsurfacehydrophobicity(Absolometal.1983).Accordingtothe
watercontactangle,cellsurfacehydrophobicitymayberoughlyclassiedintothree
categories:ahydrophobicsurfacewithacontactanglegreaterthan90°,amedium
hydrophobicsurfacewithacontactanglebetween50°and60°,andahydrophilic
surfacewithacontactanglebelow40°(MozesandRouxhet1987).
It should be noted that
H
LW
, H
+
, H
–

canbedeterminedbycontactanglemeasure-
mentswithatleastthreedifferentliquids(ofwhichtwomustbepolar)bytheYoung
equation (van Oss, Chaudhury, and Good 1987, 1988):
(1 cos ) 2
Lm
LW
L
LW
mL mL
 

 
QG G G GG GG
Becausee the contact angle method requires specic equipment, microbial
adhesion to solvents (MATS) has been developed to characterize microbial cell
surfaces (Bellon Fontaine, Rault, and van Oss 1996). This method is based on the
comparison between microbial cell afnity to a monopolar solvent and a polar solvent.
Acidic solvent serves as electron acceptor, and basic solvent as electron donor.
9.3 THE ROLE OF CELL SURFACE HYDROPHOBICITY IN
AEROBIC GRANULATION
An aerobic granule can form through cell-to-cell self-adhesion, and its formation is a
multiple-step process, and both physicochemical and biological forces are involved.
53671_C009.indd 152 10/29/07 7:27:20 AM
© 2008 by Taylor & Francis Group, LLC
© 2008 by Taylor & Francis Group, LLC
Table9.1liststhecommonlyusedthemonopolarsolventsintheMATSmethod.
The Essential Role of Cell Surface Hydrophobicity in Aerobic Granulation 153
It has been suggested that microbial adhesion can be dened in terms of the energy
involved in the formation of the adhesive junction. When a bacterium approaches
another bacterium, the hydrophobic interaction between them is a crucial force.

Wilschut and Hoekstra (1984) proposed a local dehydration model, and suggested
that under the physiological conditions, the strong repulsive hydration interaction
wasthemainforcetokeepthecellsapart.
So far, aerobic granules have been developed with various substrates, and
aerobic granulation by heterotrophic, nitrifying, denitrifying, and phosphorus-
accumulating bacteria has been reported. This implies that aerobic granulation is
notstrictlyrestrictedtosomespecicsubstrateandmicrobialspecies,anditcanbe
regardedasaprocessinwhichindividualcellsaggregatetogetherthroughcell-to-
cellhydrophobicinteractionandbinding.Itisbelievedthatcellhydrophobicityis
one of the most important afnity forces in microbial aggregation. Hydrophobicity
andhydrophilicityareusuallyusedtodescribeamoleculeorastructurehavingthe
featureofbeingrejectedfromanaqueousmedium(i.e.,hydrophobicity),orbeing
positively attracted (i.e., hydrophilicity). Hydration interaction becomes signicant
atsurfaceseparationsof2to5nmorless,dependingonthenatureofbacterial
surfaces. In terms of process thermodynamics, microbial aggregation is driven by
decreases of free energy, that is, increasing cell surface hydrophobicity results in a
corresponding decrease in the Gibbs energy of the surface, which in turn promotes
cell-to-cell interaction and further serves as an inducing force for cells to aggregate
out of hydrophilic liquid phase. Local dehydration of the surfaces that are a short
distance apart has been identied as the prerequisite for bacterial adhesion (Tay, Xu,
and Teo 2000).
Concrete evidence shows that the formation of aerobic granules under different cul
-
tureconditionsiscorrelatedverycloselytoanincreaseincellsurfacehydrophobicity,
as discussed in the preceding chapters. Li, Kuba, and Kusuda (2006) reported that the
surfacenegativechargeofbacteriadecreasedfrom0.203to0.023meqgVSS
−1
along
withaerobicgranulationinanSBR,whilesuchadecreaseinthedensity ofcellsurface
negativechargewasaccompaniedbyanincreaseintherelativecellsurfacehydro

-
cell surface charge results in weakened repulsive force of bacterium to bacterium, that
is, the decreased surface negative charge promotes cell to cell aggregation, ultimately
TABLE 9.1
Surface Tensions of Typical Organic Solvents
Liquid Formula
H
1
LW
(mJ m
–2
)
H
1
+
(mJ m
–2
)
H
1
–
(mJ m
–2
)
Decane C
10
H
22
23.9 0 0
Ethyl acetate C

4
H
10
O 23.9 0 19.4
Hexadecane C
16
H
34
27.7 0 0
Chloroform CHCl
3
27.2 3.8 0
Source: Data from Bellon Fontaine, M/ N., Rault, J., and van Oss, J.
1996. Colloid Surface B 7: 47–53.
53671_C009.indd 153 10/29/07 7:27:21 AM
© 2008 by Taylor & Francis Group, LLC
© 2008 by Taylor & Francis Group, LLC
phobicityfrom28.8%to60.3%(gure9.2).Asdiscussedearlier,reduceddensityof
154 Wastewater Purification
leading to aerobic granulation. In fact, an inverse proportional correlation between
cellsurfacehydrophobicityandsurfacenegativechargehasbeenestablishedforacti-
vatedsludgemicroorganisms(Liaoetal.2001).Thisseemstoimplythathighcell
surface hydrophobicity favors aerobic granulation.
Thecellsurfacehydrophobicityofacetate-fedaerobicgranuleswasfoundtobe
nearly two times higher than that of suspended seed sludge (Tay, Liu, and Liu 2002),
whileYang,Tay,andLiu(2004)reportedthatnitrifyingbacteriaexposedtohigh
freeammoniaconcentrationcouldnotformgranules,andalowcellsurfacehydro
-
phobicity of the nitrifying biomass was detected. As discussed in the preceding
chapters, cell surface hydrophobicity is very sensitive to the shear force and hydraulic

selectionpressurepresentinanSBR;however,theeffectoftheorganicloadingrate
intherangeof1.5to9.0kgCODm
–3
d
–1
on the cell surface hydrophobicity was
notsignicant.Zheng,Yu,andSheng(2005)alsofoundthattherewasasignicant
differenceincellsurfacehydrophobicitybeforeandaftertheformationofaerobic
granular sludge, for example, the mean contact angle values were 35.0° and 46.3°
for seed sludge and granular sludge, respectively. This suggests that the formation
ofaerobicgranularsludgeisassociatedwithanincreaseinthecellsurfacehydro-
phobicity, whereas the specic gravity of sludge increased with the increase of the
cellsurfacehydrophobicityalongwithaerobicgranulation.
Toh et al. (2003) investigated the cell surface hydrophobicity of aerobic granules
of various sizes, and in their study cell surface hydrophobicity was expressed as the
specic surface hydrophobicity determined by measuring phenanthrene adsorption
accordingtotheprocedureproposedbyKim,Stabnikova,andIvanov(2000).Inthis
method, a 2-ml sample was added to 4 ml of 60% (w/v) phenanthrene solution, while
inthecontroltest,2mlofdeionizedwaterisusedtoreplacethesample.Allmixtures
were incubated without shaking in the dark for 30 minutes. The incubated mixtures
were then ltered, and the ltrates were subsequently used for the determination of
0
0.05
0.1
0.15
0.2
0.25
0 5 10 15 20 25 30 35
Operation Time (days)
Surface Charge (–meq g

–1
VSS)
FIGURE 9.2 Change in the density of cell surface negative charge along with aerobic gran-
ulationinanSBR.(DatafromLi,Z.H.,Kuba,T.,andKusuda,T.2006.Enzyme Microb
Technol 38: 670–674.)
53671_C009.indd 154 10/29/07 7:27:22 AM
© 2008 by Taylor & Francis Group, LLC
© 2008 by Taylor & Francis Group, LLC
The Essential Role of Cell Surface Hydrophobicity in Aerobic Granulation 155
dry biomass, while the supernatants were assayed for phenanthrene concentration,
usingaluminescencespectrometer.Thespecicsurfacehydrophobicity(
H
s
)inmilli-
grams phenanthrene per gram volatile solids (VS) can be calculated as follows:
H
VF F
X
s
ce

()
in which F
c
and F
e
are the phenanthrene concentration in the control and the sample,
respectively,
V is the volume from which the concentration was measured, and X is the
total dry biomass used in the hydrophobicity test. It was found that the specic cell

surface hydrophobicity tended to increase from 2.46 to 5.92 mg phenanthrene g
–1
VS
asthegranulesizeincreasedfrom<1mmto>4mmindiameter(Tohetal.2003).
This was also conrmed by confocal laser scanning microscopy (CLSM) exami-
nationshowingthatwhenagranulegrewlarger,thebiomassdensityofaerobic
granulesincreasedinthesurfacelayerandthustheaccumulativehydrophobicityon
these bacterial cell surfaces could generate a higher hydrophobicity on the exterior
face of the granule (Toh et al. 2003).
Changes in cell surface hydrophobicity result from bacterial responses to certain
stressful culture conditions (Bossier and Verstraete 1996; Mattarelli et al. 1999). It is
most likely that the cell surface hydrophobicity induced by stressful conditions would
strengthen cell-to-cell interaction, leading to a stronger microbial self-attachment, which
inturnprovidesaprotectiveshellforcellsexposedtotheunfavorableenvironments.
To date, aerobic granulation phenomena have been observed only in SBRs,
whilenosuccessfulexampleofaerobicgranulationhasbeenreportedincontinu
-
ousculture.Comparedtoacontinuousculture,theuniquefeatureofanSBRisits
cycleoperation.Asaresultofthecycleoperation,microorganismsaresubjectto
aperiodicfastingandfeasting,thatis,thereisaperiodicstarvationphaseduring
thecycleoperationofanSBR(seechapter14).Ithasbeenshownthatthestarva
-
tionphasehasaprofoundimpactonthesurfacepropertiesofbacteria(Kjelleberg,
Humphrey, and Marshall 1983; Hantula and Bamford 1991; Bossier and Verstraete
1996). Some studies showed that starvation conditions could induce cell surface
hydrophobicity that in turn facilitates microbial adhesion and aggregation (Chesa,
Irvine, and Manning 1985; Bossier and Verstraete 1996). Through controlling feast
-
ingandfastingcyclesbyoperatingactivatedsludgesystemsinaplugoworby
feedingthesludgeintermittently,Chesa,Irvine,andManning(1985)foundthatsuch

an operation strategy yielded better-settling sludge with high cell surface hydro
-
phobicity. It is most likely that microorganisms can change their surface proper
-
ties when faced with starvation, and such changes can contribute to their ability to
aggregate. Therefore, the periodic starvation cycle would induce the cell surface
hydrophobicity, and then the induced cell surface hydrophobicity can help initiate
cell-to-cell self-aggregation.
Itshouldbepointedoutthattheeffectofstarvationoncellsurfacehydrophobicity
is still debatable, as discussed in chapter 14. The negative effect of starvation on
changesincellsurfacehydrophobicityhasbeenreported,forexample,upontransfer
fromarichgrowthmediumtostarvationconditions,cellsurfacehydrophobicity
droppedsharplybutrecovereditsinitialvaluewithin24to48hours(Castellanos,
53671_C009.indd 155 10/29/07 7:27:23 AM
© 2008 by Taylor & Francis Group, LLC
© 2008 by Taylor & Francis Group, LLC
156 Wastewater Purification
Ascencio,andBashan2000).Ontheotherhand,Sanin,Sanin,andBryers(2003)
reportedthatcellsurfacehydrophobicitiesstayedmoreorlessconstantduringcarbon
starvationconditions,whereastherewasasignicantdecreaseinhydrophobicity
when all three cultures were starved for nitrogen. Castellanos, Ascencio, and Bashan
(2000)alsonotedthatstarvationwasnotamajorfactorininducingchangesinthe
cell surface that led to the primary phase of attachment of
Azospirillum to surfaces.
9.4 FACTORS INFLUENCING CELL SURFACE HYDROPHOBICITY
Microbialcellsfavoradispersedratherthanaggregatedstateundernormalculture
conditions. Aerobic granulation is the result of cell response to stressful environ-
ments,whichleadtochangesinthesurfacecharacteristicsofbacteria(seechapter
2).Thehighhydrophobicityofmicroorganismsisusuallyassociatedwiththepres-
enceofspeciccellwallproteins(Singleton,Masuoka,andHazen2001;Kosetal.

2003). As discussed earlier, extracellular polymeric substances produced by bacteria
mainly consist of proteins and polysaccharides. Proteins are polymers of amino
acidscovalentlybondedbypeptidebonds.Aminoacidhasahydrophiliccarboxylic
acid group (-COOH) and a hydrophobic or hydrophilic side chain. The side chains of
amino acids have twenty different structures whose hydrophobicity varies markedly
(Parker,Guo,andHodges1986).Ifthecarboxylicacidgroup(-COOH)oftheamino
acidisconnectedwithanaminogroup(-NH
2
)ofanotheraminoacid,theconnected
polymerbecomesapolarmolecule,eithermonopolar(hydrophilic)oramphipathic
(one hydrophilic and one hydrophobic) (Parker, Guo, and Hodges 1986). Cell wall
proteinsmayworkintwoways:(1)exposedhydrophobicproteinsdirectlybindto
extracellular matrix proteins; or (2) alternatively, cell surface hydrophobicity may
mediate attachment by facilitating and maintaining specic receptor-ligand inter-
actions (Singleton, Masuoka, and Hazen 2001). Obviously, a sound understanding
of the factors that may inuence cell surface hydrophobicity is important for devel-
opingthestrategyforafastaerobicgranulation.
Extracellular polysaccharides have been considered to play an important role in
both the formation and stability of biolms and anaerobic and aerobic granules by
mediatingbothcohesionandadhesionofcells(SchmidtandAhring1994;Tay,Liu,
andLiu2001b,2001a;LiuandTay2002;Qin,Liu,andTay2004).Polymerscan
bridge physically or electrostatically to form a three-dimensional structure, which
favors attachment of bacterial cells (Ross 1984). In a pilot-scale upow anaerobic
sludgeblanket(UASB)reactor,QuarmbyandForster(1995)foundthatanaerobic
granules tended to become weaker as the surface negative charge of cells increased.
At the usual pH value, suspended bacteria are negatively charged and electrostatic
repulsionexistsbetweencells.Ithasbeensuggestedthatextracellularpolymerscan
change the surface negative charge of bacteria, and thereby bridge two neighbor
-
ing bacterial cells to each other as well as other inert particulate matters, and settle

outasoccusaggregates(Shen,Kosaric,andBlaszczyk1993;SchmidtandAhring
1994). This seems to indicate that the formation and stability of immobilized cell
communitieshaveastrongassociationwithextracellularpolysaccharides.
Thehighcellsurfacehydrophobicityisusuallyassociatedwiththepresenceof
brillar structures on the cell surface and specic cell wall proteins. Fibrils may
53671_C009.indd 156 10/29/07 7:27:24 AM
© 2008 by Taylor & Francis Group, LLC
© 2008 by Taylor & Francis Group, LLC
The Essential Role of Cell Surface Hydrophobicity in Aerobic Granulation 157
attach to the surface of receptors by piercing through the energy barrier between
cells into a strong Lewis acid-base (AB) force interaction distance (van Oss 2003).
Figure 9.3 schematically presents differences in accessibility of round spherical bod
-
iesandaatplate.Forcell-to-cellinteraction,theatplateshowningure9.3can
bedisplacedbyanothercell.Thesmoothhydrophilicsphericalcellcannotmake
contact with a smooth at hydrophilic surface because their mutual specic, macro-
scopic-scale repulsion prevents a closer approach. However, a similar spherical cell
with long thin spiky brils can easily penetrate the microscopic-scale repulsion eld,
leadingtoamacroscopic-scalespeciccontact.Ingure9.3,thedottedlineindi
-
catesthelimitofclosestapproachforasmoothhydrophiliccellwitharelatively
largeradiusofcurvature.
Starvation may induce changes in cell surface hydrophobicity. Kjelleberg and
Hermansson (1984) observed a large increase in surface roughness throughout the star
-
vation period for all studied strains that showed marked changes in physicochemical
characteristics. As discussed earlier, brillar surface structure can help to overcome
theintercellularenergybarrier.Itislikelythatincreasedcellsurfaceroughnessmight
havethesamefunctionascellsurfacehydrophobicityinmicrobialaggregation.
Culture temperature may also inuence cell surface hydrophobicity. Thermo-

dynamically, water will become a much stronger Lewis acid at higher temperature
(van Oss 1993, 1994b), for example, at 20°C,
GG
ww

25.5 mJ m
–2
, whereas at
38°C,
G
w

 32.4 and G
w

 18.5 mJ m
–2
.Usually, G
m

values of most biological
surfaces are extremely low, whereas the
G
m

value can be high or low, depending one
whetheritishydrophilicorhydrophobic.Thus,acellsurfacemaybedesignatedas
essentiallyelectrondonormonopolar(vanOss,Chaudhury,andGood1987;vanOss
1994b).Inthiscase,forelectrondonormonopolarcompounds,theincreaseof G
w


with temperature will markedly increase the value of the term
GG
mw

(a) (b)
FIGURE 9.3 Illustration of the different accessibilities of spheres with smooth (A) and rough
(B)surfacestoaatplatesurface.(FromvanOss,C.J.2003.J Mol Recognit 16: 177–190.
With permission.)
53671_C009.indd 157 10/29/07 7:27:28 AM
© 2008 by Taylor & Francis Group, LLC
© 2008 by Taylor & Francis Group, LLC
158 Wastewater Purification
ThisisinlinewiththendingbyBlancoetal.(1997)thatthemajorityofthe
forty-two strains of
Candida albicans studied were hydrophobic at 22°C, but hydro-
philicat37°C,andthehydrophobiccellsshowedaconsistentadherencecapacity
that was absent from the hydrophilic strains.
Thegrowthrateofmicroorganismsisanotherfactorthatcaninuencecellsur
-
facehydrophobicity.Inastudyoftheimpactofbrewingyeastcellageonfermentation
performance, Powell, Quain, and Smart (2003) reported that the occulation poten
-
tial of cells and cell surface hydrophobicity increased in conjunction with cell age,
whereasinaselectionofxenobiotic-degradingmicroorganisms,asimilartrendwas
foundbyAsconcabreaandLebeault(1995),thatis,thecellsurfacehydrophobicity
tendedtoincreasewiththegrowthrateintermsofdilutionrate.vanLoosdrechtetal.
(1987) also found that for a certain species, at high growth rates, bacterial cells would
become more hydrophobic. Meanwhile, Malmqvist (1983) observed an increase in cell
surfacehydrophobicityduringexponentialgrowth.Expressionofcellsurfacehydro

-
phobicityisinuencedbygrowthconditions,andisoftenexpressedaftergrowth
under nutrient-poor conditions, or starvation (Ljungh and Wadstrom 1995).
It has been shown that the change in cell surface hydrophobicity can result from
bacterial stress responses to certain culture conditions, such as low pH, high tem
-
perature, and hyperosmotic stress (Danniels, Hanson, and Phillips 1994; Bossier
and Verstraete 1996; Correa, Rivas, and Barneix 1999; Mattarelli et al. 1999).
Blanco et al. (1997) reported that the majority of the forty-two strains of
Candida
albicans studied were hydrophobic at 22°C, but hydrophilic at 37°C. As presented
in chapter 1, cell surface hydrophobicities of aerobic granules grown on glucose and
acetateshowednosignicantdifference,whereascellsurfacehydrophobicitywas
found to increase with increase in hydrodynamic shear force (c2). It appears from
chapter 1 that aerobic granules cultivated at different organic loading rates of 1.5 to
9.0 kg COD m
–3
d
–1
exhibited comparable cell surface hydrophobicity of 78% to
86%; however, cell surface hydrophobicity was signicantly improved as the cycle
time was shortened (chapter 6). In the preceding chapters, it can be seen that all
selection pressures may improve cell hydrophobicity, including settling time, dis
-
charge time, and exchange ratio of the SBR. This means that the cell surface hydro
-
phobicity induced by culture conditions strengthen cell-to-cell interaction, leading
toastrongermicrobialstructure.
Sun,Yang,andLi(2007)investigatedtheeffectofcarbonsourceonaerobic
granulation. It appears from gure 9.4 that during the period of the reactor start-up,

thetypeofcarbonsourceinuencesthesurfacepropertyofsludgeintermsofzeta
potential,whichisoftenusedtoquantifythedensityofsurfacecharges,forexample,
sludge grown on peptone shows the lowest surface charge density among all four
organic carbon sources studied. As a result, the peptone-fed aerobic granules had the
highestbiomassdensityovergranulesgrownonacetate,glucose,andfecula(Sun,
Yang,andLi2007).Asdiscussedearlier,thedensityofthesurfacechargeisinversely
relatedtothecellsurfacehydrophobicity,thatis,alowerchargedensitymeansahigher
cellsurfacehydrophobicity.Inaddition,gure9.4alsoimpliesthatthepropertyof
feed may inuence cell surface hydrophobicity during aerobic granulation.
When exposed to toxic or inhibitory substrates, microorganisms are able to regu
-
late their surface properties, especially cell surface hydrophobicity. In a study on
53671_C009.indd 158 10/29/07 7:27:29 AM
© 2008 by Taylor & Francis Group, LLC
© 2008 by Taylor & Francis Group, LLC
The Essential Role of Cell Surface Hydrophobicity in Aerobic Granulation 159
substrate-dependent autoaggregation of Pseudomonas putida CP1 during the degra-
dation of monochlorophenols and phenol, Farrell and Quilty (2002) found that cells
grown on the higher concentrations of monochlorophenol were more hydrophobic
than those grown on phenol and lower concentrations of monochlorophenol, whereas
when
Pseudomonas putida wasexposedtotoxicalcohols,thebacteriumchanged
degreesofcellsurfacehydrophobicityandadaptedtothealcohols(Tsubata,Tezuka,
andKurane1997).Jiang,Tay,andTay(2004)studiedthetoxiceffectofphenolon
aerobic granules, and found that the surface hydrophobicity of cultivated sludge sig
-
nicantlydecreasedfrom60%to40%asphenolloadingwasincreasedfrom1to
2.5 kg phenol m
–3
d

–1
. This may imply that increased hydrophobicity and the resultant
autoaggregationofbacteriaisamicrobialresponsetothetoxicityofsubstrates.
Bulkliquidsurfacetensionmayalsoaltercellsurfacehydrophobicity.Thaveesri
et al. (1994; 1995) reported that anaerobic granules grown in protein-rich media
exhibited lower hydrophobicity than carbohydrate, which in turn slowed down the
anaerobicgranulation.Onthecontrary,vanLoosdrechtetal.(1987)foundthatthe
effect of the growth substrate on cell surface hydrophobicity was not signicant.
Different observations could result from the fact that all substrates used by van
Loosdrecht(1987)werecarbohydratesandnoprotein-richmediawereincluded.
Thaveesrietal.(1995)proposedamodeltointerprettheinuenceofgrowthsub
-
strate on the surface hydrophobicity of anaerobic granules, and they thought that if
H
lv
is high, microorganisms with low surface energy (low H
mv
or hydrophobic bacteria)
canaggregateinordertoobtainminimalfreeenergy.Nevertheless,when
H
lv
is
low, bacteria with high surface energy (high
H
mv
or hydrophilic bacteria) can form
aggregates easily. Since there is no high-concentration chemical solvent present in
municipal wastewater, the rst case (high
H
lv

)iscommoninmunicipalwastewater
treatment practice, whereas the second case becomes true in treating chemical
Time (days)
0246810
Zeta Potential of Sludge Surface (mV)
–20
–18
–16
–14
–12
–10
–8
–6
FIGURE 9.4 Change in zeta potential of sludge in the course of aerobic granulation in
SBRs fed with acetate (D), glucose ($), peptone (O), and fecula (/).(DatafromSun,F.Y.,
Yang, C. Y., and Li, J. Y. 2007. Beijing Jiaotong Daxue Xuebao [J Beijing Jiaotong University]
31: 106–110.)
53671_C009.indd 159 10/29/07 7:27:30 AM
© 2008 by Taylor & Francis Group, LLC
© 2008 by Taylor & Francis Group, LLC
160 Wastewater Purification
solvent-containing industrial wastewater. In both cases, microbial aggregation seems
to be related to the relative hydrophobicity of microorganisms to the aqueous phase
in which they grow. Protein-rich substrate can lower the surface tension of reactor
liquid,andconsequentlyleadtotheformationofanaerobicgranuleswithlowhydro
-
phobicity because low hydrophobicity of anaerobic granules could decrease the free
energy at the interface between aggregates and liquid.
Cellsurfacehydrophobicityisalsorelatedtogeneexpressionofmicroorganisms.
The analytical study of biolm has demonstrated that adhesion launches the expres

-
sionofasetofgenesthatendswiththetypicalbiolmphenotype,particularlywith
an enhanced resistance to antimicrobial agents (Stickler 1999; Davey and O’Toole
2000;WatnickandKolter2000;Goldberg2002).Likewise,thegeneexpressionfor
enzyme is also different under free-living and immobilized conditions (Hata et al.
1998;Vallimetal.1998;Akaoetal.2002;Astheretal.2002).Forexample,Vallim
etal.(1998)foundthat
Phanerochaete chrysosporium showedadifferentialgene
expression for cellobiohydrolase when it was cultured under immobilized conditions.
Therefore, it is possible that the new gene expression gives rise to changes in the cell
surfacepropertiesthatinturnhelpsthecellsadapttothenewlivingenvironment.
Inconclusion,changesincellsurfacehydrophobicitycanresultfrombacterial
stressful responses to certain culture conditions. The cell surface hydrophobicity
inducedbycultureconditionsstrengthenscell-to-cellinteraction,leadingtoastronger
microbialstructurethatprovidesaprotectiveshellforcellsexposedtotheunfavor
-
able environments. Therefore, it is necessary to identify the key engineered parameter
that can induce cell surface hydrophobicity during biogranulation.
9.5 SELECTION PRESSURE-INDUCED CELL
SURFACE HYDROPHOBICITY
Aerobicgranulationisagradualprocessevolvingfromdispersedseedsludgeto
mature and stable aerobic granules with spherical outer shape. It is thought that cell
surface hydrophobicity may trigger and initiate aerobic granulation. As the seed
sludge used to inoculate bioreactors often has a very low cell surface hydrophobicity,
highcellsurfacehydrophobicityreportedinaerobicgranulationwouldnotbean
extantpropertyoftheseedsludge,anditwouldbeinducedduringaerobicgranula
-
tion. Many factors as discussed earlier may induce cell surface hydrophobicity, but
most of them cannot be manipulated in terms of process operation. Thus, it is neces
-

sary to identify the key engineered parameters that can induce cell hydrophobicity
during aerobic granulation.
Hydraulicselectionpressurehasbeenprovedtobeadecisiveparameterinthe
formationofbiogranules.AbsenceofanaerobicgranulationinUASBreactorswas
observed at very weak hydraulic selection pressure in terms of liquid upow velocity,
whileithasbeendemonstratedthataerobicgranulationisaprocessdrivenbyselec
-
tion pressure in SBRs. Under strong selection pressures, only the good-settling and
heavysludgeparticlesareretainedinthereactor,whilethelightandpoor-settling
sludge is washed out. Compared to seed sludge with a cell surface hydrophobicity
of20%,whenmicroorganismsaresubjecttohighselectionpressureintermsof
settling time, the cell surface hydrophobicity gradually increases to 70% during
53671_C009.indd 160 10/29/07 7:27:31 AM
© 2008 by Taylor & Francis Group, LLC
© 2008 by Taylor & Francis Group, LLC
The Essential Role of Cell Surface Hydrophobicity in Aerobic Granulation 161
aerobicgranulation.Thehydraulicselectionpressureseemstoinducechangesin
cell surface hydrophobicity, and a strong selection pressure results in a more hydro
-
phobic cell surface (Qin, Liu, and Tay 2004). Similarly, it was found that anaerobic
granular sludge in UASB reactors was more hydrophobic than the nongranular sludge
thatwashedout(Mahoneyetal.1987).Underthestrongselectionpressure,micro
-
organismshavetoadapttheirsurfacepropertiesinordertoavoidbeingwashedout
from the reactors through microbial self-aggregation. In this sense, biogranulation
wouldbeaneffectivedefensivestrategyofmicrobialcommunitiesagainstexternal
selectionpressure,andthusonemayexpecttomanipulatetheformationandcharac
-
teristicsofbiogranulesbycontrollingselectionpressure(chapter6).
9.6 THERMODYNAMIC INTERPRETATION OF

CELL SURFACE HYDROPHOBICITY
Similar to a chemical process, microbial aggregation is a function of changes in free
energy at interfaces. Bacterial attachment to a solid surface can be described by
the following:
Microorganism surface Microorganism surfacem  $G
adh
o'
(9.1)
in which microorganism–surface means microorganism attached onto a solid surface,
and $G
adh
o'
is the effective free energy change of microorganism-to-solid surface
attachment,whichchangeswiththeproceedingofmicrobialattachmentonthesur
-
face. For bacterial attachment onto a solid surface, $G
adh
o'
canbeexpressedas:
$G
adh
o
sm sl ml
'
GGG (9.2)
in which
H
sm
, H
sl

,andH
ml
are the solid-bacteria, solid-liquid, and bacteria-liquid inter-
facial free energies, respectively. The surface free energies are related to contact
angles (
R)accordingtotheYoungequation:
GQGG
lv sv sl
cos  (9.3)
in which
H
lv
and H
sv
are the liquid-vapor and solid-vapor interfacial free energies,
respectively.Itshouldberealizedthatequation9.3cannotbesolvedfortheinterfacial
free energies (
H
sv
and H
sl
)bymeasuringthecontactangleandliquidsurfacetension
(H
lv
) without additional assumptions (Bos, van der Mei, and Busscher 1999). In a
study of anaerobic granulation, Thaveesri et al. (1995) used the equation developed
byNeumannetal.(1974)asasupplementaryequationtoequation9.3:
G
G
GG

sl
sv lv
sv lv
r



()
.
2
1 0 015
(9.4)
53671_C009.indd 161 10/29/07 7:27:35 AM
© 2008 by Taylor & Francis Group, LLC
© 2008 by Taylor & Francis Group, LLC
162 Wastewater Purification
Compared to microbial attachment to a solid surface, microbial aggregation
isamicroorganism-to-microorganismself-immobilizationprocess,whichcanbe
depicted as follows:
Microorganism Microorganism Microorganism Mmiicroorganism $G
agg
o'
(9.5)
in which microorganism–microorganism represents microbial aggregation, and
$G
agg
o'
is the effective free energy change of microorganism-to-microorganism
aggregation, which changes with the proceeding of microbial aggregation. It has
been proposed that equation 9.2 can also be applied to the microbial aggregation

process by replacing the solid by an identical microorganism (Bos, van der Mei, and
Busscher 1999). Thus,
$G
agg
o'
of microbial aggregation can be written as follows:
$G
agg
o
mm ml
'
GG2
(9.6)
in which H
mm
is the microorganism-microorganism interfacial free energy. In a study
ofanaerobicgranulation,Thaveesrietal.(1995)postulatedthatwhenadhesionoftwo
identicalbacteriaisconsidered,H
mm
is equal to zero. Hence, equation 9.6 reduces to:
$G
agg ml
'
2G
(9.7)
Substituting equation 9.4 into equation 9.7 by replacing the subscript s by m yields:
$G
agg
mv lv
mv lv

'
()
.



2
1 0 015
2
GG
GG
(9.8)
According to equation 9.8, Thaveesri et al. (1995) proposed that if H
lv
is high,
low-energy surface types of microorganisms (low H
mv
or hydrophobic bacteria) can
aggregate in order to obtain minimal free energy, while when
H
lv
is low, high-energy
surface types of bacteria (high H
mv
or hydrophilic bacteria) can aggregate better. As
there is no high-concentration chemical solvent present in municipal wastewater, the
rst case (high H
lv
) is most commonly encountered in municipal wastewater treatment
practice, whereas the second case applies to treating chemical solvent-containing

industrial wastewater. No matter what the case is, microbial aggregation seems to be
relatedtorelativehydrophobicityofmicroorganismstotheaqueousphaseinwhich
theysurvive.Itshouldberealizedthatequation9.8isnotcorrelatedtoacharacter
-
isticparameterofmicrobialaggregation;itonlyoffersaqualitativeinterpretationof
surfacethermodynamicsofmicrobialaggregation.Moreover,intherealmicrobial
aggregation process, there are no completely identical bacteria, and the simplicity
assumedbyThaveesrietal.(1995)suchthat
H
mm
is zero is still debatable. Cells favor
adispersedratherthanaggregatedstateundernormalcultureconditions,hence
H
mm
should exist during cell-to-cell interaction.
As Bos, van der Mei, and Busscher (1999) noted, it is impossible to deter
-
mine experimentally the interfacial free energies in equation 9.2. Evidence shows
53671_C009.indd 162 10/29/07 7:27:39 AM
© 2008 by Taylor & Francis Group, LLC
© 2008 by Taylor & Francis Group, LLC
The Essential Role of Cell Surface Hydrophobicity in Aerobic Granulation 163
that the H
mm
and H
ml
termsinequation9.4arerelatedtobothcellsurfacecharge
andhydrophobicity(ZitaandHermansson1997;Kosetal.2003),whilecellsur
-
facehydrophobicityisinverselycorrelatedtothequantityofsurfacechargeof

microorganisms (Liao et al. 2001). There is strong evidence that the microbial
autoaggregation process is correlated to cell surface hydrophobicity (Del Re et al.
2000;Kosetal.2003;Liuetal.2003).AccordingtoextendedDLVOtheoryby
VanOss,GoodandChaudhury(1986),cellsurfacehydrophobicityrepresentsan
attractiveforce,whilecellsurfacehydrophilicityreectsrepulsionbetweencells.
This indicates that both
H
mm
and H
ml
in equation 9.4 are functions of the interaction
between cell surface hydrophobicity and cell surface hydrophilicity, that is, with the
increaseofcellsurfacehydrophobicityorthe decrease of cell surface hydrophilicity,
repulsiveforcesbetweencellsbecomesweakerandweaker.AccordingtoLiuetal.
(2004a),
$G
agg
o'
can be expressed as:
$$GGaRTH
agg
o
agg
o
ow
'
/
ln (9.9)
in which
$G

agg
o
is change of the standard free energy of the microbial aggregation
process,
a is a positive coefcient, and H
o/w
is relative cell hydrophobicity, which is
dened as follows:
H
ow/

Cell hydrophobicity
Cell hydrophilicty
(9.10)
It has been recognized that the formation of biolms and microbial aggregates
isamultiple-stepprocess,towhichphysicochemicalandbiologicalforcesmake
signicantcontributions(LiuandTay2002).Someparametersotherthancell
surface hydrophobicity, such as cell surface charge and extracellular polymers, may
also inuence or contribute to microbial attachment and aggregation. In general, the
surfaceofmicroorganismsisnegativelychargedundertheusualpHconditions,and
bacterial surface charge might affect bacterial attachment (Rouxhet and Mozes 1990).
There is evidence that microbial attachment on a solid surface can be improved or
enhanced by reducing electrical repulsion between cell and solid surfaces (Changui
etal.1987;RouxhetandMozes1990;Masui,Takata,andKominami2002),while
Strand, Varum, and Ostgaard (2003) reported that charge neutralization was not the
main microbial occulation mechanism. As pointed out earlier, microbial aggrega
-
tion is cell-to-cell interaction that is different from bacterial adhesion on a solid
surface. When bacteria that carry charges with the same sign approach each other,
thereshouldbeanelectricalrepulsiveforcebetweenbacterialsurfaces.Zitaand

Hermansson (1997) studied the correlations of cell surface hydrophobicity and charge
to adhesion of
Escherichia coli strainstoactivatedsludgeocs,andfoundthatthere
was a strong correlation between the cell surface hydrophobicity of the E. coli strains
andadhesiontothesludgeocs,whileforpositivecellsurfacechargesthecorrela
-
tionwasweakerthanforsurfacehydrophobicity,andnegativecellsurfacecharges
showed no correlation to adhesion. In order to enhance initial interaction between
microorganisms,Jiangetal.(2003)addedcalciumionstoneutralizethenega
-
tively charged bacterial surface, but the results were not as satisfactory as expected.
53671_C009.indd 163 10/29/07 7:27:41 AM
© 2008 by Taylor & Francis Group, LLC
© 2008 by Taylor & Francis Group, LLC
164 Wastewater Purification
Therefore, it seems that electrical interaction is not a triggering force of aggrega-
tionofmicroorganisms.Inaddition,extracellularpolymersplayaroleinmicrobial
attachment and aggregation. Tsuneda et al. (2003) reported that if the amount of
extracellular polymers is relatively small, cell adhesion onto solid surfaces is inhib
-
ited by electrostatic interaction, and if it is relatively large, cell adhesion is enhanced
by polymeric interaction. Jorand et al. (1998) studied hydrophobic and hydrophilic
properties of activated sludge extracellular polymers, and found that a signicant
proportion of extracellular polymers fraction was hydrophobic. Such results support
the hypothesis that hydrophobic extracellular polymers are involved in the organiza
-
tion of microbial ocs, biolm, and aggregates, that is, hydrophobic interaction may
befundamentalinmicrobialaggregation,asdiscussedearlier.
Similar to any chemical process, the overall change of free energy of the micro
-

bial aggregation process (
%G
agg
)inc reases with the increase of process resistance,
and decreases with the increase of the driving force of microbial aggregation process.
Liuetal.(2004a)putforwardahypothesisthattheoverallchangeoffreeenergyof
themicrobialaggregationprocessshouldbeformulatedasafunctionofthedriving
forceandresistanceoftheprocess,suchthat:
$$GGbRT
agg agg
o

'
ln
Resistance
Driving force
(9.11)
in which
b is a positive coefcient. Equation 9.11 is similar to those developed for
biological systems (Roels 1983). It has been demonstrated that aerobic granulation is
agradualprocessfromdispersedsludgetostablegranules,andbothgranulesizeand
density nally approach respective equilibrium value (Tay, Liu, and Liu 2001c). Thus,
the driving force of microbial aggregation is the potential of the process towards
aggregation that can be described by the difference between the current state of
microbial aggregates and the balanced state that microbial aggregates can thermo-
dynamicallyachieve.Itisobviousthatincreasingthecellsurfacehydrophobicity
cansimultaneouslycauseadecreaseintheexcessenergyofthemicrobialsurface,
whichinturnenhancescell-to-cellinteraction,leadingtoamorecompactstructure
of microbial aggregates (Liu et al. 2003). In the environmental engineering eld, the
density of microbial aggregates is often used to describe how compact and strong

themicrobialinteractionis.Theobserveddensityofmicrobialaggregatesisthe
result of balanced cell-to-cell interaction, which is a characteristic parameter repre
-
senting the state of microbial aggregates at time
t.Figure9.5showsanexampleof
the evolution of the density (
S)ofmicrobialaggregatesagainsttimeobservedinthe
SBRrunatasubstrateN/CODratioof5/100.Itappearsthat
S is gradually close to
its equilibrium.
Liu et al. (2004a) further proposed that the driving force of microbial aggrega
-
tioncanbedenedasthedifferencebetweenthedensityattime
t (S)andthedensity
of microbial aggregates at equilibrium state (
S
eq
). With the increase of S, the aggre-
gationprocesstendstoreachitsequilibrium.Asaresult,thedrivingforce(
S
eq
– S)
decreases and the resistance would increase accordingly. This shows that
S indeed
would reect the magnitude of aggregation resistance. Hence, equation 9.11 can be
translated to
53671_C009.indd 164 10/29/07 7:27:43 AM
© 2008 by Taylor & Francis Group, LLC
© 2008 by Taylor & Francis Group, LLC
The Essential Role of Cell Surface Hydrophobicity in Aerobic Granulation 165

$$GGbRT
agg agg
o


'
ln
R
RR
eq
(9.12)
When
S =0.5S
eq
, $G
agg
o'
isequalto∆G
agg
.Thisimpliesthat $G
agg
o'
canbedened
as the overall free energy change at S =0.5S
eq
,thatis,thedrivingforceofmicro-
bial aggregation is equal to the resistance force. As S approaches S
eq
,∆G
agg

goes to
innity and further aggregation becomes energetically impossible. In fact, this is in
agreementwithpreviousresearch(Bos,vanderMei,andBusscher1999).Substitu-
tionofequation9.9intoequation9.12yields:
$$G G aRT H bRT
agg agg
o
ow
 

ln ln
/
R
RR
eq
(9.13)
Solvingequation9.13forS gives:
RR
¥
§
¦
¦
´
¶
µ
µ


eq
ow

ab
GG
RT
b
H
e
agg
o
agg
()
/
/
()
/
$$
1
(()
/
/
H
ow
ab
(9.14)
Equation 9.14 can be rearranged as:
RR

eq
ow
n
agg o w

n
H
KH
()
()
/
/
(9.15)





 




FIGURE 9.5 The density (S)ofaggregatesagainstoperationtimeobservedintheSBR
operated at substrate N/COD ratio of 5/100. (From Liu, Y. et al. 2004a. Appl Microbiol
Biotechnol 64: 410–415. With permission.)
53671_C009.indd 165 10/29/07 7:27:47 AM
© 2008 by Taylor & Francis Group, LLC
© 2008 by Taylor & Francis Group, LLC
166 Wastewater Purification
in which
Ke
agg
GG
RT

m
agg
o
agg

¨
ª
©
©
·
¹
¸
¸

()
$$
(9.16)
and n
andmequala/band1/b, respectively. A simple least-square method can be
used to evaluate the constants involved in equation 9.15. If
S
eq
is known, K
agg
and n
canalsobeeasilyestimatedfromalinearplotofln[
S/(S
eq
– S)] versus lnH
o/w

.
ln ln ln
/
R
RR
eq
o w agg
nH K

 (9.17)
Values of K
a
gg
and n canbedeterminedfromtheslopeandinterceptof
equation 9.17.
Liuetal.(2004a)determinedcellsurfacehydrophobicityinthecourseofthe
formation of aerobic granules in SBRs operated at different substrate N/COD ratios of
5/100to30/100byweight.Thus,
H
o/w
canbecalculatedaccordingtoequation9.10:
H
ow/
%


Cell hydrophobicity (%)
cell hydro100 pphobicity (%)
(9.18)
The value of

H
o/w
is in between innite (absolutely hydrophobic) and 0
(absolutely hydrophilic). The density (
S)ofmicrobialaggregatesversusH
o/w
is
showningures9.6to9.9fordifferentsubstrateN/CODratios.Itcanbeseenthat
the prediction given by equation 9.15 is in good agreement with the experimental
data obtained, indicated by a correlation coefcient greater than 0.92. For microbial
aggregation at substrate N/COD ratios of 5/100 to 30/100, the respective value of
n






     





FIGURE 9.6 S versus H
o/w
observed in aerobic granulation at substrate N/COD ratio of
5/100.Thepredictiongivenbyequation9.15isshownbyasolidcurvewithacorrelation
coefcient of 0.97. S
eq

=13.3gL
–1
; K
agg
= 0.086; and n =4.0,(FromLiu,Y.etal.2004a.Appl
Microbiol Biotechnol 64: 410–415. With permission.)
53671_C009.indd 166 10/29/07 7:27:49 AM
© 2008 by Taylor & Francis Group, LLC
© 2008 by Taylor & Francis Group, LLC
The Essential Role of Cell Surface Hydrophobicity in Aerobic Granulation 167
estimated is 4.0, 3.2, 2.8, and 1.7. This implies that cell surface hydrophobicity has
amorepronouncedeffectonaerobicgranulationatthelowsubstrateN/CODratio,
which is further conrmed by the values of
K
agg
, which increases with the increase
ofthesubstrateN/CODratio.Ontheotherhand,thedensityofaerobicgranulesat
equilibrium was found to increase with the increase of the substrate N/COD ratio.
Thisimpliesthatamorecompactmicrobialstructureofaerobicgranulescanbe
obtainedathighsubstrateN/CODratio.Figure9.10showstherelationshipbetween
the relative activity of the nitrifying population over the heterotrophic population
andthespecicgrowthrate(µ
agg
) of aerobic granules by size, and gure 9.11 further
exhibitstheeffectofµ
agg
on n and S
eq
.













FIGURE 9.7 S versus H
o/w
observed in aerobic granulation at substrate N/COD ratio of
10/100.Thepredictiongivenbyequation9.15isshownbyasolidcurvewithacorrelation
coefcient of 0.97. S
eq
= 15.3 g L
–1
; K
agg
=0.39;andn =3.2.(FromLiu,Y.etal.2004a.Appl
Microbiol Biotechnol 64: 410–415. With permission.)















FIGURE 9.8 S versus H
o/w
observed in aerobic granulation at substrate N/COD ratio of
20/100.Thepredictiongivenbyequation9.15isshownbyasolidcurvewithacorrelation
coefcient of 0.93. S
eq
=17.9gL
–1
; K
agg
= 1.11; and n =2.8.(FromLiu,Y.etal.2004a.Appl
Microbiol Biotechnol 64: 410–415. With permission.)
53671_C009.indd 167 10/29/07 7:27:51 AM
© 2008 by Taylor & Francis Group, LLC
© 2008 by Taylor & Francis Group, LLC
168 Wastewater Purification
Intermsofchemistry,∆G
agg
issupposedtocorrelatetothemassquotientofthe
microbial aggregation process (Q
agg
) through the following equation:
$$GGRTQ

agg agg
o
agg
ln (9.19)
Substitution of equation 9.15 into equation 9.19 leads to:
K
Q
agg
agg
m

¥
§
¦
´
¶
µ
1
(9.20)
(SOUR)
N
/(SOUR)
H
0.20 0.39 0.50 0.64
μ
agg
(d
–1
)
0.04

0.06
0.08
0.10
0.12
FIGURE 9.10 Effect of (SOUR)
N
/(SOUR)
H
on µ
agg
ofaerobicgranules.(DatafromLiu,Y.
et al. 2004a. Appl Microbiol Biotechnol 64: 410–415. With permission.)












FIGURE 9.9 S versus H
o/w
observed in aerobic granulation at substrate N/COD ratio of
30/100.Thepredictiongivenbyequation9.15isshownbyasolidcurvewithacorrelation
coefcient of 0.92. S
eq

=19.8gL
–1
; K
agg
= 0.77; and n =1.7.(FromLiu,Y.etal.2004a.Appl
Microbiol Biotechnol 64: 410–415. With permission.)
53671_C009.indd 168 10/29/07 7:27:53 AM
© 2008 by Taylor & Francis Group, LLC
© 2008 by Taylor & Francis Group, LLC
The Essential Role of Cell Surface Hydrophobicity in Aerobic Granulation 169
Equation 9.20 reveals that K
agg
is inversely related to the process mass quotient.
When
K
agg
is small or Q
agg
is large, microorganisms proceed toward aggregation; on
the other hand, when
K
agg
is large or Q
agg
is small, microbial interaction tends away
fromtheaggregation.Thisprovidesaplausibleexplanationforwhy
S approaches
its maximum much more slowly when
K
agg

is large, whereas when K
agg
is small,
S increases steeply and breaks to the right sharply, as shown in gures 9.6 to 9.9.
Obviously, equation 9.20 offers a theoretical interpretation for the physical meaning
of
K
agg
.If∆G
agg
is close to zero, Q
agg
inequations9.19and9.20canbereplacedby
the equilibrium constant (
K
agg
eq
) of the microbial aggregation process. Consequently,
the magnitude of the
K
agg
value may represent the equilibrium position of a microbial
aggregation process.
Equation 9.15 can t the experimental data very well. The values of
n estimated
by equation 9.15 are closely related to the substrate N/COD ratios, that is, a low
substrate N/COD ratio results in a high
n value.Intermsofreactionkinetics,the
magnitude of
n describes how fast microorganism-to-microorganism hydrophobic

interaction is under given culture conditions. In fact, Chen and Strevett (2003)
reported that various substrate N/COD ratios inuence microbial surface thermo
-
dynamics reected in cell surface hydrophobicity, while evidence also indicates that
cellsurfacehydrophobicityandhydrophobicinteractionarecloselyrelatedtomicro
-
bialactivity(Liaoetal.2001;Tay,Liu,andLiu2001c;Liuetal.2003).Liuetal.
(2004a) determined the overall specic growth rate (µ
agg
) of microbial aggregates
by size and the activity distribution of heterotrophic and nitrifying populations in
stable aerobic granules in terms of (SOUR)
N
/(SOUR)
H
.Itwasfoundthatµ
agg
tended
to decease with the increase of substrate N/COD ratio. Such a changing trend of
µ
agg
with the substrate N/COD ratio indeed is within expectation because a high
substrate N/COD ratio results in a sustainable enrichment of slow-growing nitrify
-
further shows the relationships among µ
agg
, n,andS
eq
.Itappearsthatn is positively






        



















FIGURE 9.11 Effect of µ
agg
on n (D)andS
eq
($).(DatafromLiu,Y.etal.2004a.Appl Micro-
biol Biotechnol 64: 410–415.)

53671_C009.indd 169 10/29/07 7:27:55 AM
© 2008 by Taylor & Francis Group, LLC
© 2008 by Taylor & Francis Group, LLC
ingpopulationwithlowgrowthrateinaerobicgranules(gure9.10).Figure9.11
170 Wastewater Purification
correlated to µ
agg
, that is, the microbial aggregation process is faster at high µ
agg
than
that at low µ
agg
. Meanwhile, the lowered growth rate of aerobic granules results in
ahigher
S
eq
,thatis,amorecompactandstrongermicrobialstructure.Infact,high
growth rate encourages the outgrowth of aerobic granules, leading to a loose struc
-
tureaccompaniedwithalowdensity.Inastudyofbiolms,thestrengthofbiolms
was found to be negatively related to the growth rate of microorganisms (Tijhuis, van
Loosdrecht, and Heijnen 1995), while Kwok et al. (1998) reported that the biolm
density decreased as the growth rate increased. Similarly, in the anaerobic granula
-
tionprocess,ahighbiomassgrowthratealsoledtoareducedstrengthofanaerobic
granules (Quarmby and Forster 1995).
It seems that microorganism-to-microorganism interaction with a
H
o
/

w
value
greaterthan1favorscellself-aggregationtowardsatightandcompactmicrobial
structure. Hydrophobic cells may attach not only on the surface, but also can aggre
-
gatetoformmicrobialgranules,whereashydrophiliccellsdidnot(Olofsson,Zita,
andHermanson1998;SharmaandRao2003).RyooandChoi(1999)attemptedto
develop the fungal pellets from the surface thermodynamic balance between fungal
cellandliquidmedia.Aerobicgranulationisadynamicprocessevolvingfrom
dispersedsludgetomatureandstableaerobicgranules.Equation9.15shedslight
onasoundthermodynamicunderstandingofcellsurfacehydrophobicityinaerobic
granulation,anditwouldbeapplicableforbiolmandanaerobicgranulation.
9.7 ENHANCED AEROBIC GRANULATION BY
HIGHLY HYDROPHOBIC MICROBIAL SEED
Itisapparentthatcellsurfacehydrophobicitycanserveasanimportantinducing
forceforaerobicgranulation.Thisinturnimpliesthatenrichmentculturesofmicro
-
organisms with high cell surface hydrophobicity or self-aggregation ability can be
usedtoaccelerateorenhanceaerobicgranulation.Forthispurpose,afour-step
proceduretoselecthighlyhydrophobicbacteriahasbeendeveloped(Wang2004;
Ivanov et al. 2005).
Step1:precultivatedaerobicgranulesaredisintegratedbyabeaterfor5minutes
andarefurtherlteredthrougha35-µmcellstrainercapinordertoremove
particlesbiggerthan35µm.
Step 2: bacteria collected at the bottom of the test tube after 5 minutes of set
-
tlingtimearetransferredintoaliquidmedium.
Step3:thetransferredbacteriaarecultivatedintheliquidmediumfor48hours,
andthenmicrobialaggregatesformedareremovedafter5minutesofsedi
-

mentation time.
Step 4: those bacteria at the air–liquid interface are nally harvested as seed
forenhancedaerobicgranulation.
Theselectedbacteriahadanaverageaggregationindexof60%to80%comparedtothe
activatedsludgewithatypicalaggregationindexof5%to20%(Ivanovetal.2005).
Usingtheselectedhighlyhydrophobicbacteriaasaseed,Ivanovetal.(2005)
showedthataerobicgranulationwasshortenedfrom8daysinthecultureof
53671_C009.indd 170 10/29/07 7:27:56 AM
© 2008 by Taylor & Francis Group, LLC
© 2008 by Taylor & Francis Group, LLC
53671_C009.indd 171 10/29/07 7:27:58 AM
© 2008 by Taylor & Francis Group, LLC
© 2008 by Taylor & Francis Group, LLC
The Essential Role of Cell Surface Hydrophobicity in Aerobic Granulation 171
conventional activated sludge to 2 days in the enrichment culture (gure 9.12 and
gure 9.13). These offer a new operation strategy by which the time required for
aerobic granulation can be shortened markedly.
According to the interfacial chemistry, microorganisms situated at the water–air
interface are highly hydrophobic. Following the same procedure as presented earlier,
Time of Cultivation (days)
0 5 10 15 20
Mean Size of Bioparticles (mm)
0
1
2
3
4
1
2
Conventional diameter

of formed granules
fIgure 9.13 Evolution of bioparticle size during aerobic granulation with selected highly
hydrophobic bacteria (1) and conventional activated sludge ocs (2). (Data from Ivanov, V.
et al. 2005. Water Sci Technol 52: 13–19.)
(d)
(e)
(f)
(a)
(b)
(c)
10 mm
fIgure 9.12 Aerobic granulation after 5 days (a, d), 10 days (b, e), and 20 days (c, f)
observed in two SBRs seeded with conventional activated sludge ocs (a to c) and selected
highly hydrophobic bacteria (d to f), respectively. (From Ivanov, V. et al. 2005. Water Sci
Technol 52: 13–19. With permission.)
172 Wastewater Purification
Wang (2004) articially selected highly hydrophobic cells at the water–air interface
usinganinoculationloopasshowningure9.14.TwoSBRs,namelyR1andR2,
were then inoculated with conventional activated sludge with low cell surface hydro
-
phobicity and selected highly hydrophobic cells as seeds, respectively.
Table 9.2 summarizes the main characteristics of conventional activated sludge
and selected highly hydrophobic seed inoculated to R1 and R2. Obviously, the
selected seed was much more hydrophobic than the conventional activated sludge.
Figure9.15showschangesinthesludgevolumeindex(SVI)inthecourseofR1and
R2operations.TheSVIinR2remainedataverylowlevelof60mLg
–1
,whilea
sharpincreaseintheSVI,followedbyasignicantdecrease,wasobservedinR1.It
isclearthatafteraerobicgranulationfromday10onwards,theSVIobservedinR1

decreasedsignicantly,indicatinganimprovedsludgesettleability.
The evolution of sludge morphology in the course of operation of R1 and R2 was
trackedbyimageanalysistechnique.Ascanbeseeningure9.16,smooth,round
aerobicgranulesappearedinR2inoculatedbytheselectedhighlyhydrophobic
microbial seed after 10 days of operation, while the same phenomenon was observed
TABLE 9.2
Characteristics of Conventional Activated Sludge and
Selected Highly Hydrophobic Microbial Seed
Activated Sludge
Selected Highly
Hydrophobic Seed
Total soluble solids (g L
–1
) 2.43 0.33
Sludge volume index (mL g
–1
) 226 76.9
Particle size (μm) 42 102.9
Hydrophobicity (%) 32.6 80.1
Source: Wang, Z W. 2004. Unpublished report. Nanyang Technologiccal
University, Singapore.
FIGURE 9.14 Selectionofhighlyhydrophobiccellsatthewater–airinterface.(From
Wang,Z W.2004.Unpublishedreport.NanyangTechnologicalUniversity,Singapore.)
53671_C009.indd 172 10/29/07 7:27:59 AM
© 2008 by Taylor & Francis Group, LLC
© 2008 by Taylor & Francis Group, LLC
The Essential Role of Cell Surface Hydrophobicity in Aerobic Granulation 173
in R1 seeded by conventional activated sludge 20 days later. These observations
seem to indicate that aerobic granulation is shortened by 20 days if the selected
highly hydrophobic microbial seed is used. Meanwhile, gure 9.17 clearly shows

that aerobic granules developed from the selected highly hydrophobic microbial
seed look more compact and denser in structure than those cultivated from con
-
ventional activated sludge. Moreover, the mechanical shearing tests also reveal that
aerobic granules developed from the selected highly hydrophobic microbial seed are
muchmoreresistanttoexternalmechanicalstirringstrengththanthosecultivated
from conventional activated sludge (gure 9.18).
ChangesinbiomassconcentrationinR1andR2werefollowedthroughoutthe
granulation process (gure 9.19). Although the initial biomass concentration seeded
toR1wasmuchhigherthanthatinoculatedtoR2,afteronly1dayofoperationthe
biomassconcentrationinR1andR2becamecomparable.Thiswasduemainlyto
53671_C009.indd 173 10/29/07 7:28:02 AM
© 2008 by Taylor & Francis Group, LLC
© 2008 by Taylor & Francis Group, LLC
0
50
100
150
200
250
300
0 10 20 30 40
Time (days)
SVI (mL g
–1
)
fIgure 9.15 Changes in sludge volume index (SVI) in the course of operation observed
in R1 seeded by conventional activated sludge (⦁) and R2 (•) inoculated by selected highly
hydrophobic microbial seed. (Data from Wang, Z W. 2004. Unpublished report. Nanyang
Technological University, Singapore.)

R2
R1
5 days 10 days
15 days
30 days 40 days 0 days
fIgure 9.16 Evolution of sludge morphology in the course of aerobic granulation in R1 and
R2 seeded with conventional activated sludge and highly hydrophobic cells, respectively. (From
Wang, Z W. 2004. Unpublished report. Nanyang Technological University, Singapore.)