Jou nal o Wa e P ocess Enginee ing 55 (2023) 104086
A ailable online 4 Augus 2023
2214-7144/© 2023 The Au ho s. Published by Else ie L d. This is an open access a icle unde he CC BY license (h p://c ea i ecommons.o g/licenses/by/4.0/).
Memb ane ac ioning o p e- ea ed was e ac i a ed sludge o he
eco e y o aluable biocompounds
Daniel Nú˜
nez
a
, Paula Oulego
a
, Mahdi Nikbakh Fini
b
, Jens Mu
b
, Se gio Collado
a
,
F ancisco A. Rie a
a
, Ma io Díaz
a
,
*
a
Depa men o Chemical and En i onmen al Enginee ing, Uni e si y o O iedo, c/Juli´
an Cla e ía 8, 33006 O iedo, Spain
b
Depa men o Chemis y and Bioscience, Aalbo g Uni e si y Esbje g, Niels Boh s Vej 8, Esbje g 6700, Denma k
ARTICLE INFO
Keywo ds:
Biomolecules
Modelisa ion
Ul a il a ion
Was e alo isa ion
We oxida ion
ABSTRACT
The eco e y o added- alue biocompounds om was e ac i a ed sludge (WAS) is a p omising al e na i e o i s
cu en managemen . In his s udy, WAS was pa ially we oxidised p oducing a highly complex s eam mainly
composed o humic acids, p o eins and ca bohyd a es. This s eam was ul a il e ed o assess he in luence o
memb ane ma e ial and molecula weigh cu -o (MWCO) on i s ac ioning based on he di e en sizes o he
biomolecules con ained in he oxidised WAS.
Fla -shee polye he sul one (PES), pe manen ly hyd ophilic polye he sul one (PESH), and polyac yloni ile
(PAN) memb anes wi h a MWCO o 50 KDa we e e alua ed. The bes pe o mance was ob ained wi h PES
memb ane, achie ing high e en ion alues (a ound 70 % o p o eins and ca bohyd a es, and 47 % o humic
acids) and high di e ences in selec i i y be ween p o eins and ca bohyd a es wi h humic acids (a ound 23 %).
Fo he size ac ioning expe imen s, PES memb anes o 10 and 3 kDa (PES10 and PES3) we e used, ob aining he
bes esul s when il e ing he PES50 memb ane pe mea e wi h he PES3 memb ane, e aining 83 % o ca bo-
hyd a es, 87 % o p o eins and 69 % o humic acids. These esul s open he possibili y o sepa a ing ca bohy-
d a es, p o eins and humic acids h ough an in eg a ed memb ane p ocess.
Besides, memb anes we e cha ac e ized by a omic o ce mic oscopy, in a ed spec oscopy, and con ac angle
measu emen s.
Mul iple ouling models we e assessed, and he main ouling in PAN memb ane, and o a lesse ex en in PES,
was e e sible. Con e sely, PESH memb ane ouling had a s ongly i e e sible cha ac e . Cake il a ion can be
conside ed he main ouling mechanism in all expe imen s.
1. In oduc ion
The bio e ining o biowas es has been p oposed as a sus ainable
means o was e alo isa ion, ob aining ene gy and biochemical e-
sou ces while he olume o inal was e is educed [1]. Al hough i is ye
a i s concep ual phase [2], he s eady ising end o he bio e ine y
ma ke alue, expec ed o show an annual g ow h o 2.2 % un il
eaching a o al alue o USD 52680 million by 2027 [3], is a eliable
ma ke o i s po en ial. Was e ac i a ed sludge (WAS) is a p omising
aw ma e o bio e ine y, as i is a sou ce o biomolecules (p o eins,
lipids, ca bohyd a es, humic acids and enzymes), phospho us, bio-
plas ics, bio-pes icides and also ene gy [4–10]. The cu en global
ma ke size o jus he biomolecules p esen in WAS is o almos USD 200
billion, u ning i s eco e y economically appealing [11].
WAS mainly consis s o locs o bac e ial cells, which a e o med by
weak physical chemical in e ac ions wi h ex acellula polyme ic sub-
s ances [12] (a complex mix u e o polyme s gene a ed by bac e ial
exc e ion, eleased a e cell lysis o p esen in he incoming was ewa e
[13]). Thus, in o de o eco e he abo e-men ioned biocompounds, a
p io solubilisa ion o he sludge is needed in o de o b eak bo h he
s uc u e o med by he ex acellula polyme ic subs ances and he cell
walls, hus eleasing hei in acellula con en o he liquid medium.
Se e al solubilisa ion me hods ha e been es ed o WAS, including
ul a-sonica ion [14,15], ca i a ion [16], alkali ea men [15,17],
ozonolysis [18], we oxida ion (WO) and he mal hyd olysis [19].
Among hese echniques, WO is pa icula ly in e es ing, due o i s
easibili y o be used a an indus ial scale o sludge s abilisa ion [20].
I should be no ed ha his solubilisa ion esul s in a complex ma ix,
* Co esponding au ho .
E-mail add ess: [email p o ec ed] (M. Díaz).
Con en s lis s a ailable a ScienceDi ec
Jou nal o Wa e P ocess Enginee ing
jou nal homepage: www.else ie .com/loca e/jwpe
h ps://doi.o g/10.1016/j.jwpe.2023.104086
Recei ed 19 Decembe 2022; Recei ed in e ised o m 3 July 2023; Accep ed 22 July 2023
Jou nal o Wa e P ocess Enginee ing 55 (2023) 104086
2
whe e in e ac ions be ween he di e en eleased biomolecules can
di icul hei pu i ica ion. Fo ins ance, elec os a ic in e ac ions be-
ween p o eins and humic acids occu while binding o hea y me als
such as Cu
2+
, Zn
2+
, and Cd
2+
[21]; and e en agg ega es a e o med
when complexed wi h Cu
2+
[22], hinde ing a sui able sepa a ion o
hese molecules by immobilised me al a ini y ch oma og aphy. Besides,
a selec i e p ecipi a ion is also no possible om his complex ma ix, as
p o eins, ca bohyd a es and humic acids co-p ecipi a e wi h se e al
p ecipi a ion me hods [19]. These molecules ha e impo an indus ial
applica ions sepa a ely: p o eins a e used in cosme ics, ood indus y,
pha maceu icals and animal eed; humic acids can be applied in ag i-
cul u e, pha maceu icals o ecological emedia ion, among o he uses;
and ca bohyd a es a e o en used in he ood indus y [11]. Thus, hei
sepa a ion is o g ea in e es o he inco po a ion o WAS as a aluable
aw ma e ial in a con ex o ci cula economy. Besides, lipid eco e y o
i s use as bio uel om WAS aces di icul ies du ing i s pu i ica ion wi h
sol en ex ac ion, as o he lipidic con aminan s such as wax es e s,
e penoids and polycyclic a oma ic hyd oca bons a e ex ac ed oge he
wi h he desi ed lipids [4].
Fo hese easons, ac ioning he solubilised WAS would imp o e
he e iciency o u he sepa a ion and pu i ica ion s eps. To ha end,
memb ane il a ion is a sui able echnology o his pu pose due o i s
ad an ages, such as high selec i i y, low ene gy consump ion, low cos ,
and mild ope a ing condi ions [23]. Ne e heless, he pe o mance o
he memb ane il a ion is a ec ed by se e al ac o s, especially he
choice o memb ane ma e ial and i s molecula weigh cu -o (MWCO)
[24–26]. Mos memb anes a e polyme ic, and he choice o his polyme
is c i ical o he e iciency o he ope a ion, since i a ec s he pe me-
abili y a e, he sepa a ion abili y o he ouling p ocess, key pa ame e s
in he il a ion p ocess [23].
Despi e hei u e impo ance, polyme ic ma e ials ha e no ye
been s udied o he il a ion o solubilised WAS, and he e ec o
MWCO has been sca cely s udied. Hence, only Li e al. ha e es ed he
use o 1, 10, 30, and 50 kDa polysulphone memb anes o concen a e
humic acids [27,28]. The e o e, he objec i e o his wo k was o s udy
he in luence o he memb ane ma e ial and MWCO on he ac iona ion
o hyd o he mally solubilised WAS in o de o sepa a e ca bohyd a es,
p o eins, and humic acids based on hei size di e ences aiming o
ob ain pa ially pu i ied s eams o hese compounds. In his sense, 3
di e en polyme ic ma e ials: polye he sulphone, hyd ophylic poly-
e he sulphone and polyac yloni ile, wi h a MWCO om 3 KDa o 50
KDa, we e e alua ed, paying special a en ion o ouling modelling.
Hyd ophilic polye he sul one and polyac yloni ile ha e a hyd ophilic
cha ac e , unlike polye he sul one, which is a hyd ophobic ma e ial.
Besides, an in eg a ed memb ane p ocess o he eco e y o bio-
molecules om WAS was also p oposed.
2. Expe imen al
2.1. Oxidised was e ac i a ed sludge
Was e ac i a ed sludge was collec ed om he hickening uni o a
was ewa e ea men plan loca ed in no he n Spain (Baí˜
na, As u ias).
The collec ion was pe o med by ained plan pe sonnel o ensu e he
ep esen a i eness o he samples. WAS was immedia ely solubilised by
a pa ial WO a 160 ◦C and 40 ba o 80 min. These oxida ion condi ions
we e selec ed o maximise he p oduc ion o he a ge molecules: i he
in ensi y o he ea men is oo high, he a ge molecules ge oxidised
o mine alised; i he in ensi y is oo low, he sludge does no comple ely
solubilise [29]. Addi ionally, as he oxida ion in ensi y inc eases, he
pa icle size dec eases, educing he e en ion capabili ies o he mem-
b anes. A cons an low o 1200 mL/min o O
2
sa u a ed wi h s eam was
main ained du ing he en i e eac ion. The con en o he eac o was
kep s i ing a 150 pm. A mo e de ailed desc ip ion o he eac o can
be ound in [30]. A e he eac ion, he oxidised WAS was cen i uged in
o de o wo k wi h he liquid phase. Sodium azide 0.1 % (w/ ) was
added o he oxidised WAS in o de o p e en mic obiological g ow h.
The oxidised sludge was s o ed a 4 ◦C o 15 days and hen eplaced
wi h esh oxidised sludge.
2.2. Memb ane il a ion
2.2.1. Memb anes
MQ (Synde Fil a ion) polye he sul one (PES), UH050 (Mic odyn
Nadi ) hyd ophylic polye he sul one (PESH), and MW (Suez) poly-
ac yloni ile (PAN) la -shee memb anes wi h MWCO o 50 kDa (named
as PES50, PESH50, and PAN50, espec i ely) we e employed o pe o m
he ma e ial sc eening expe imen s. Addi ionally, ST (Synde Fil a ion)
and VT (Synde Fil a ion) PES la -shee memb anes wi h MWCO o 3
and 10 kDa, espec i ely (named as PES3 and PES10), we e used o he
cu -o size sc eening expe imen s. All memb anes we e cu o a ci cula
shape o 9 cm o diame e , and a il a ion a ea o 63.62 cm
2
.
Memb ane hyd ophilici y was cha ac e ized by con ac angle mea-
su emen s. The images we e ob ained wi h a CAM 200 op ical con ac
angle me e (KSV Ins umen s L d., Finland). Sessile wa e d ople s we e
d opped on he clean and ouled memb ane su aces using a sy inge and
le o sp ead eely. Images o he d ople s we e aken by a high-
esolu ion CCD came a a 40 ms in e als o he i s 0.36 s, and a 1
s in e als o he subsequen 19 s. Equilib ium sessile d op con ac
angles we e de e mined om he s eady-s a e angles using he KSV CAM
200 so wa e by measu ing he angle be ween he baseline o a liquid
d op and he angen a he solid–liquid bounda y. All con ac angle
measu emen s we e pe o med in iplica e using h ee di e en mem-
b ane samples.
2.2.2. Equipmen and il a ion condi ions
Fil a ion expe imen s we e ca ied ou in duplica e using an FT17
C oss- low Fil a ion Uni (A m ield L d., Uni ed Kingdom), which al-
lows o pe o m angen ial low il a ions wi h la shee memb anes.
P io o conduc ing he expe imen s, all polyme ic memb anes we e p e-
condi ioned by unning he equipmen wi h no p essu e o 30 min using
dis illed wa e . Subsequen ly, wa e was il e ed unde he ope a ional
condi ions (indica ed below) o an addi ional 30 min. The lux ob ained
in his s ed was conside ed as he lux a =0. The pe mea e lux o he
clean memb anes was measu ed du ing his s ep o u he ouling
modelling. All he ma e ial sc eening il a ion expe imen s we e pe -
o med unde he ollowing condi ions: empe a u e o 50.0 ±0.4 ◦C,
ansmemb ane p essu e (TMP) o 4.0 ±0.2 ba and c oss low eloci y
(CFV) o 3.00 m/s. The oxidised WAS was il a ed wi hou pe mea e
eci cula ion un il a olume concen a ion a e (VCR) o 2.5 was
eached.
Du ing he MWCO sc eening expe imen s, he oxidised WAS was
il e ed wi h he PES50, PES10 and PES3 memb anes. Besides, in o de
o assess he iabili y o he ac ioning o he oxidised WAS, he
pe mea e ob ained a e he il a ion wi h PES50 was subsequen ly
il e ed wi h he PES10 o wi h he PES3 memb anes, naming hese
pe mea es as PES50-10 and PES50-3, espec i ely. The expe imen s
we e pe o med unde he same condi ions han hose used in he ma-
e ial sc eening ones. Only o ob aining he PES50-3 pe mea e, p essu e
was se a 30 ba and a VCR alue o 1.25 was achie ed.
Pe mea e low was de e mined by g a ime ic measu emen s o he
pe mea e, which we e collec ed au oma ically by he FT17 C oss- low
Fil a ion Uni so wa e. Flux was calcula ed by he ollowing equa ion
(Eq. (1)):
J=QP
AM
(1)
Whe e J is he pe mea e lux (m⋅s
−1
), Q
P
is he pe mea e low (m
3
⋅s
−1
),
and A
M
is he memb ane su ace a ea (m
2
).
In addi ion, samples o he pe mea e and e en a e we e collec ed
pe iodically and kep a 4 ◦C o u he analysis.
D. Nú˜
nez e al.
Jou nal o Wa e P ocess Enginee ing 55 (2023) 104086
3
A e he il a ions, he ouled memb ane was insed wi h dis illed
wa e un il a cons an lux was ob ained, and i s pe meabili y was
measu ed o u he ouling modelling.
2.2.3. Fouling modelling
Resis ance-in-se ies, He mia’s, and Men ha’s ouling models we e
employed o cha ac e ise bo h he e e sibili y and main mechanism o
memb ane ouling occu ed du ing he di e en il a ion expe imen s.
Resis ance-in-se ies model exp esses he o al hyd aulic esis ance o
he memb ane (R
T
, m
−1
) as he sum o di e en esis ances caused by
e e sible ouling (R
e
, m
−1
), i e e sible ouling (R
i e
, m
−1
), o by he
memb ane i sel (R
m
, m
−1
) (Eq. (2)). Hyd aulic esis ance can be
calcula ed as shown in Eq. (3):
RT=Rm+R e +Ri e (2)
R=TMP
μ
J(3)
Whe e
μ
is he dynamic iscosi y o he WAS a 50 ◦C (kg⋅m⋅s
−1
). By
adding o sub ac ing he esis ances ob ained wi h he clean, ouled, o
insed memb ane luxes, R
m
, R
e
, and R
i e
can be easily calcula ed. A
mo e de ailed explana ion o hese calcula ions can be ound in he
Appendix o [31]).
The main ouling mechanism occu ed on each memb ane du ing
ul a il a ion was de e mined h ough He mia’s model [32] (Eq. (4)):
dJ
d = − Kj⋅(J−J0)⋅J2−n(4)
Whe e is ime (min), K
j
is he model cons an ha depends on he
ouling phenomenon, J
0
is he limi ing lux (m⋅s
−1
), and n is a cons an
ha a ies o he ouling mechanism: comple e po e blocking (CPB) (n
=2, K
b
in min
−1
), whe e he ac i e memb ane a ea is blocked by pa -
icles la ge han he po e size; in e nal po e blocking (IPB) (n =1.5, K
i
in m
−1
), whe e memb ane po es a e blinded by ei he adso p ion o
deposi ion o pa icles smalle han he po e size; pa icle po e blocking
(PPB) (n =1, K
p
in m
−1
), whe e pa icles migh seal a po e o e ime, o
b idge i and no block i comple ely; and cake il a ion (CF) (n =0, K
c
in min⋅m
−2
), whe e a cake o pa icles ha does no en e he po es is
o med on he memb ane su ace [33].
The K
j
o he ou models we e calcula ed by minimizing he di -
e ence be ween he p edic ed alues and he expe imen al da a,
calcula ed as he sum o squa ed esiduals (SSR). The model wi h he
lowes SSR was chosen as he mos sui able one o each se o expe i-
men al da a.
In addi ion, lux was also modelled using he Meh a’s model [34],
which akes in o accoun he wo lux decline domains ha ake place
du ing memb ane il a ion: domain 1, whe e a apid lux decline occu s
du ing he ea ly s age o il a ion; and domain 2, whe e he lux decline
dec eases un il he lux emains quasi-s able [35]. I can be exp essed as
ollows (Eq. (5)):
J=J0−J∞1⋅exp−
α
+ (J∞1 −J∞2)⋅exp−β +J∞2 (5)
Whe e J∞1 is he lux a he end o domain 1 (m⋅s
−1
); J∞2 is he lux a
he end o domain 2 (i.e., a he end o he expe imen ) (m⋅s
−1
); and
α
(min −1)and β(min −1)a e wo cons an s de e mined expe imen ally
ha desc ibe he a e o lux decline associa ed wi h he memb ane
ouling and he concen a ion pola iza ion and gel laye o ma ion,
espec i ely.
2.3. A omic o ce mic oscopy
The oughness o he ouled and clean memb anes was analysed by
a omic o ce mic oscopy (AFM). All AFM measu emen s we e pe o med
a oom empe a u e (20 ◦C) using a Nanoscale scanning unnelling
mic oscope (Nano ec Ce an es FullMode SPM), wo king in con ac
mode in ai medium wi h gold coa ed silicon ni ide ips. Memb ane
samples we e ixed o he sample holde o he mic oscope wi h high-
acuum silicone g ease. Roughness pa ame e s we e de e mined om
he collec ed da a using he WSxM 5.0 so wa e [36]. Memb ane
oughness was compa ed in e ms o mean oughness (R
a
[nm/
μ
m]),
oo mean squa e oughness ( ms), peak- o-peak dis ance (nm/
μ
m), and
su ace skewness and ku osis. R
a
is he mean alue o he su ace
ela i e o he cen e plane; ms is he s anda d de ia ion o he heigh s
o all he pixels in he image om he a i hme ic mean [37]; and
skewness and ku osis desc ibe he shape o a p obabili y dis ibu ion,
e lec ing he obliqui y and he la ness o he cu e, espec i ely [38].
2.4. In a ed spec oscopy
In a ed spec a (FTIR) o he clean and ouled memb anes we e
aken in he ange om 600 o 4000 cm
−1
using Va ian 670-IR FTIR
spec ome e equipped wi h a Golden Ga e ho izon al a enua ed o al
e lec ance (ATR) accesso y. Expe imen al condi ions we e 32 scans, 4
cm
−1
esolu ion and ape u e open.
2.5. Analy ical me hods
P o eins, humic acids, ca bohyd a es, colou numbe (CN) and
chemical oxygen demand (COD) we e measu ed by colo ime ic
me hods. P o eins and humic acids we e measu ed ollowing he
modi ied Low y me hod desc ibed by F ølund e al. [39], using bo ine
se um albumin and comme cial humic acid as s anda ds. Ca bohyd a es
we e measu ed acco ding o he Dubois me hod [40] using D-glucose as
s anda d. Spec al abso bance coe icien s (SAC [cm
−1
]) we e measu ed
a 436, 525 and 620 nm and used o calcula e he CN alue (cm-1) ac-
co ding o Eq. (6):
CN =SAC2
436 +SAC2
525 +SAC2
620
SAC436 +SAC525 +SAC620
(6)
The abso bances o p o eins, humic acids, ca bohyd a es and SAC
we e measu ed wi h a Helios Alpha UV–Vis spec opho ome e (The mo
Scien i ic, USA).
Densi y was measu ed a 50 ◦C and 1 a m wi h a pycnome e . Ki-
nema ic iscosi y was measu ed a 50 ◦C and 1 a m wi h a Cannon-
Fenske in e sed- low iscome e (P o on, UK). Dynamic iscosi y was
calcula ed by mul iplying he kinema ic iscosi y by he densi y. pH was
measu ed wi h a Basic 20 pH me e (C ison, Spain). COD alues we e
de e mined by he po assium dich oma e me hod [41], and he abso -
bance a 600 nm was measu ed wi h a HACH DR/2500 spec opho-
ome e (Hach Company, USA). To al o ganic ca bon (TOC) was
de e mined wi h a Shimadzu TOC-V
CSH
TOC analyse (Shimadzu,
Japan).
Rejec ion coe icien s (RC
i
) we e calcula ed acco ding o he Eq. (7):
RCi=1−CPm,i
CR ,i
(7)
Whe e CPm,i and CR ,i he concen a ion o he compound “i” in he
pe mea e and he e en a e (g⋅L
−1
), espec i ely. All analy ical mea-
su emen s we e conduc ed a leas h ee imes.
3. Resul s and discussion
3.1. Oxidised was e ac i a ed sludge
The oxidised WAS was sligh ly acid and p esen ed a deep b own
colou . I s main physical-chemical cha ac e is ics a e shown in Table 1.
D. Nú˜
nez e al.
Jou nal o Wa e P ocess Enginee ing 55 (2023) 104086
4
3.2. Memb ane ma e ial sc eening
3.2.1. Con ac angle measu emen s
Measu ed con ac angles o wa e on he polyme ic memb anes used
in he ul a il a ion o he oxidised WAS a e shown in Table 2. A se-
lec ion o he pic u es o he sessile d ops, om which he con ac angles
we e calcula ed, can be ound in Fig. A1.
A su ace is conside ed hyd ophilic i he con ac angle is lowe han
90◦[42]. Thus, PES50 could be conside ed hyd ophobic, while PESH50
and PAN50 we e ound o be hyd ophilic. This was in acco dance wi h
he esul s ob ained by o he au ho s ela ed o ouling esis ance o
ul a il a ion memb anes [43]. A e il e ing he oxidised WAS, he
ouled PES50 and PESH50 u ned mo e hyd ophilic han he p is ine
ones, while PAN50 became less hyd ophilic a e being ouled, which
showed he di e en na u e o he oulan -memb ane in e ac ions
depending on he memb ane ma e ial: i seems ha PES50 and PESH50
we e coa ed wi h mo e hyd ophilic oulan s, while PAN50 in e ac ed
wi h oulan s less hyd ophilic han i sel .
3.2.2. Pe meabili y es s
The luxes ob ained wi h PES50, PESH50 and PAN50 a e shown in
Fig. 1.
Final luxes app oxima ely 3 imes highe we e ob ained wi h
PESH50 (67.2 ±0.9 LMH) and PAN50 (74 ±3 LMH) compa ed o he
one achie ed by PES50 (25 ±2 LMH), due o hei hyd ophilic cha -
ac e . In his sense, memb ane hyd ophilici y p e en ed ouling, in
acco dance o wha was epo ed by o he au ho s [44,45]. The
esis ance-in-se ies models (Fig. 2) con i med he a o emen ioned abou
hyd ophobici y and i s highe endency o ouling. In his sense, all he
esis ances (memb ane, e e sible and i e e sible) measu ed o
PES50, which added up a o al hyd aulic esis ance o 1.07⋅10
14
m
−1
,
we e highe han hose o PESH50 (13.3, 2.5 and 1.1 imes highe ,
espec i ely) and PAN50 (3.4, 2.4 and 6.3 imes highe , espec i ely).
On he o he hand, he di e en beha iou be ween luxes in PESH50
and PAN50 can be explained based on he alues o i e e sible ouling
o each memb ane. Thus, he esis ance-in-se ies modelling showed
ha he highe endency o ouling obse ed in PESH50 is due o i e-
e sible ouling, since i s R
i e
accoun ed o he 49.6 % o i s o al
hyd aulic esis ance, while R
i e
obse ed in PAN50 only ep esen ed
12.4 % o he o al hyd aulic esis ance. The o al R
i e
also seemed o be
co ela ed wi h he hyd ophobici y o he memb ane, as PES50 showed
he highes R
i e
([2.8 ±0.4]⋅10
13
m
−1
), ollowed by PESH50 ([2.45 ±
0.01]⋅10
13
m
−1
) and PAN50 ([4.3 ±0.6]⋅10
12
m
−1
).
Besides, e e sible ouling was ound o be he main ouling in PES50
and PAN50 memb anes, co esponding o a 53.4 % and o a 68.9 % o
he o al ouling o PES50 and PAN50, espec i ely.
The luxes ob ained in his s udy we e in he same o de o magni-
ude han hose ob ained by o he au ho s when PES memb anes we e
used du ing he il a ion o milk [46], e ine y and pe ochemical
was ewa e [47] and oil-in-wa e emulsion [48,49]; and wi h PAN
memb anes when ap wa e [50] and oil-in-wa e emulsion [49] we e
il e ed. Besides, lowe ini ial luxes we e ob ained when PESH mem-
b anes we e employed o he il a ion o molasses [51], while simila
luxes we e a ained o oli e oil washing was ewa e [52].
I should be no ed ha he highe luxes ob ained wi h PESH50 and
PAN50 come along wi h lowe ejec ion coe icien s and lowe selec-
i i ies be ween p o eins and humic acids (Table 3). In pa icula , he
lowes ejec ion coe icien s o CN, TOC, and he h ee measu ed bio-
compounds we e hose co esponding o PESH50. On he o he hand,
he highes ejec ions we e ob ained wi h PES50, also achie ing he
highes ejec ion di e ences be ween p o eins and humic acids (22 %
di e ence s 7 % di e ence ob ained wi h PAN50, and 10 % di e ence
ob ained wi h PESH50); and be ween ca bohyd a es and humic acids
(23 % di e ence s 13 % ob ained wi h PAN50 and 17 % di e ence
ob ained wi h PESH50).
These highe e en ions may be due o he o ma ion o a hicke cake
laye on op o he memb ane, which would ac as a seconda y il a ion
mesh, inc easing he selec i i y o he memb ane [53]. The o ma ion o
his hicke ouling cake could be obse ed h ough he esis ance-in-
se ies model (Fig. 2), whe e he R
e
, mainly associa ed wi h he o -
ma ion o he ouling cake [54], was mo e han wo- old highe a e
il e ing wi h PES50 ([5.7 ±0.8]⋅10
13
m
−1
) han wi h PESH50 ([2.3 ±
0.2]⋅10
13
m
−1
) o PAN50 ([2.41 ±0.07]⋅10
13
m
−1
). Rejec ion di e -
ences be ween PESH50 and PAN50 a e cohe en wi h his explana ion,
as he R
e
o PESH50 is sligh ly lowe han ha o PAN50.
As bo h highe ejec ion coe icien s and highe ejec ion di e ences
be ween humic acids and he o he biomolecules (p o eins and ca bo-
hyd a es) we e achie ed wi h PES50, he ac iona ion es s wi h
memb anes o di e en MWCO (MWCO sc eening expe imen s) was
ca ied ou wi h PES memb anes.
3.2.3. Flux modelling
As i can be seen in Fig. 3 and Table 4 (SSR), CF was he bes - i ing
He mia’s model in all h ee cases, al hough he i ings indica ed ha
none o he He mia’s models ully explain he ouling mechanism, hus
indica ing se e al ouling mechanisms may ha e occu ed h oughou
he il a ion. Indeed, he ac ha di e en ouling mechanisms occu a
di e en s ages o he il a ion is well documen ed in he li e a u e
[55,56] and i was aken in o accoun by Meh a’s model [34]. Thus, CF
was he main ouling mechanism o e all, al hough i e e sible po e
blocking also occu ed; especially du ing he il a ion wi h PESH50,
whe e he IPB model showed be e i ing han in he il a ions wi h
PES50 and PAN50, e lec ing he abo e-men ioned mo e i e e sible
na u e o he PESH50 ouling. Besides, CF has also been desc ibed by
o he au ho s as he main ouling mechanism o na u al o ganic ma e
du ing ul a il a ion wi h PES memb anes [57,58].
Howe e , i should be no ed ha he bes i ing o he expe imen al
da a was achie ed wi h Meh a’s model, which p o ides in o ma ion
abou he e ec o memb ane ouling (pa ame e
α
) and concen a ion
pola iza ion and gel laye o ma ion (pa ame e β) on he lux decline.
The op imised alues o hese wo cons an s a e shown in Table 4.
Highe
α
and β alues ep esen as e ini ial memb ane ouling and
as e s abilisa ion o he lux by he es ablishmen o he concen a ion
pola iza ion g adien and gel laye o ma ion, espec i ely. The
α
alues
ob ained o he il a ions wi h PESH50 and PAN50 we e 10- old highe
han hose o PES50 (3.88), while he β alues we e 4- old lowe han
hose ob ained o PES50 (5.20⋅10
−2
). This indica es ha a s ong ini ial
memb ane ouling occu ed a e s a ing he il a ion o he oxidised
Table 1
Main physical-chemical cha ac e is ics o he oxidised was e
ac i a ed sludge.
Pa ame e Value
pH 5.04 ±0.03
COD
a
(g O
2
L
−1
) 20.5 ±0.5
TOC
a
(g L
−1
) 8.00 ±0.01
CN
a
(cm
−1
) 3.9 ±0.2
P o eins (g L
−1
) 3.4 ±0.3
Humic acids (g L
−1
) 8.4 ±0.2
Ca bohyd a es (g L
−1
) 2.75 ±0.03
a
COD: chemical oxygen demand; TOC: o al o ganic ca -
bon; CN: colou numbe .
Table 2
Wa e su ace con ac angles on he s udied memb anes.
Memb ane size Memb ane ma e ial Memb ane s a e Con ac angle
50 kDa PES Clean 90 ±3
Fouled 70 ±2
PESH Clean 72 ±1
Fouled 54 ±11
PAN Clean 39 ±4
Fouled 57 ±10
D. Nú˜
nez e al.
Jou nal o Wa e P ocess Enginee ing 55 (2023) 104086
5
WAS wi h PESH50 and PAN50, causing a apid dec ease in he lux.
A e his ini ial d op, he s abilisa ion o he lux by concen a ion po-
la iza ion occu ed mo e slowly. Rega ding he beha iou du ing he
il a ion wi h PES50, i was opposi e o ha o he o he polyme ic
memb anes: he ini ial d op caused by memb ane ouling was less
d as ic, and he equilib ium in concen a ion pola iza ion was eached
as e .
This esul is in acco dance wi h he li e a u e, and can be explained
conside ing he pola iza ion sie ing model [59], based on he di e -
ences o hyd ophilici y be ween he memb anes:
Thus, in hyd ophilic memb anes, an ini ial i e e sible adso p ion
laye is o med, ega dless o he solu e concen a ion, and subsequen
0
200
400
600
800
1000
1200
1400
1600
0
0.2
0.4
0.6
0.8
1
1 1.2 1.4 1.6 1.8 2 2.2 2.4 2.6
J (LMH)
J/J0
VCR
b)
0
20
40
60
80
100
0
0.2
0.4
0.6
0.8
1
1 1.2 1.4 1.6 1.8 2 2.2 2.4 2.6
J (LMH)
J/J0
VCR
0
60
120
180
240
300
360
420
0
0.2
0.4
0.6
0.8
1
1 1.2 1.4 1.6 1.8 2 2.2 2.4 2.6
J (LMH)
J/J0
VCR
c)
a)
Fig. 1. No malised lux a ia ion o e VCR o he oxidised WAS il a ion wi h (a) PES50 (J
0
=102.4 ±0.2 L/m
2
h), (b) PESH50 (J
0
=1590 ±40 L/m
2
h) and (c)
PAN50 (J
0
=420 ±40 L/m
2
h).
Fig. 2. R
m
(
), R
e
(
) and R
i e
(
) a e il a ion wi h PES50, PESH50 and PAN50.
D. Nú˜
nez e al.

Jou nal o Wa e P ocess Enginee ing 55 (2023) 104086
6
ouling will appea in he o m o a gel-pola iza ion laye . On he o he
hand, in hyd ophobic memb anes, he size o he i e e sible adso p ion
laye inc eases un il i s hickness p o ec s he hyd ophobic su ace om
he adso bed molecules, which gene a es highe alues o i e e sible
ouling han in hyd ophilic su aces, and only a e his limi is eached,
he gel-pola iza ion laye s a s o o m [59]. Acco ding o his, he
beha iou o PES50 can be explained by i s hyd ophobic cha ac e , since
mo e ime was equi ed o he ini ial memb ane ouling o be ully
occu , and he inal concen a ion pola iza ion laye needed less ime o
s abilise, hus s a ing o o m a highe eed concen a ions.
3.2.4. A omic o ce mic oscopy
AFM images we e aken om clean and ouled memb anes in o de
o analyse he su ace mo phology (Fig. 4). The e ical p o ile o he
memb ane su ace was ep esen ed by he colou in ensi y, wi h ligh e
colou s indica ing highe egions, and da ke colou s indica ing
dep essions.
The su ace o all he h ee clean memb anes was clea ly a anged in
a “c es and alley” o “nodule and alley” pa e n, o igina ed by he
andom o ien a ion and o e lapping o he ib e s uc u e [60]. This
su ace a angemen has also been obse ed by o he au ho s when PES,
PAN and polyamide memb anes wi h la -shee , ubula and hollow-
ib e geome ies we e used in ul a il a ion and nano il a ion p o-
cesses [60–63]. The p esence o alley-like o ma ions is highly ela ed
o ouling, as oulan pa icles end o deposi in hese o ma ions [64].
This ac was suppo ed by he images o ouled memb anes, whe e
c es -like o ma ions could no longe be seen, indica ing ha he alley-
like egions had been clogged by oulan s.
In addi ion o AFM imaging, memb ane oughness was compa ed in
e ms o mean oughness (R
a
[nm/
μ
m]), oo mean squa e oughness
( ms), peak- o-peak dis ance (nm/
μ
m), and su ace skewness and
Table 3
Rejec ion coe icien s (RC) ob ained wi h he 50 kDa polyme ic memb anes.
PES50* PESH50* PAN50*
RC
CN*
0.80 ±0.02 0.69 ±0.03 0.73 ±0.04
RC
TOC*
0.47 ±0.03 0.34 ±0.04 0.38 ±0.05
RC
COD*
0.48 ±0.08 0.41 ±0.03 0.4 ±0.1
RC
CH*
0.70 ±0.02 0.57 ±0.07 0.60 ±0.06
RC
P o *
0.69 ±0.06 0.50 ±0.05 0.54 ±0.09
RC
HA*
0.47 ±0.04 0.40 ±0.03 0.47 ±0.03
*CN: colou numbe ; TOC: o al o ganic ca bon; COD: chemical oxygen demand;
CH: ca bohyd a es; PROT: p o eins; HA: humic acids; PES50: poly-
e he sulphone, 50 kDa; PESH50: pe manen ly hyd ophilic polye he sulphone,
50 kDa; PAN: polyac yloni ile, 50 kDa.
0
20
40
60
80
100
120
140
0 5 10 15 20 25 30 35 40
J (LMH)
(min)
0
200
400
600
800
1000
1200
1400
1600
1800
0 5 10 15 20 25 30 35 40
J (LMH)
(min)
0
50
100
150
200
250
300
350
400
450
0 5 10 15 20 25 30 35 40
J (LMH)
(min)
a)
b)
c)
Fig. 3. He mia’s (comple e po e blocking [ ], in e media e po e blocking [ ], pa ial po e blocking [ ] and cake o ma ion
[]) and Meh a’s ( ) lux models o PES50 (a), PESH50 (b) and PAN50 (c) expe imen al luxes (●).
Table 4
Fi ing pa ame e s o he adjus ed models.
PES50 PESH50 PAN50
He mia’s models
CPB K
b
(min
−1
) 3.31⋅10
−2
5.29 9.43⋅10
−2
SSR 172,041.83 617,253.55 628,747.63
IPB K
i
(m
−1
) 4.58⋅10
−3
1.37⋅10
−1
7.31⋅10
−3
SSR 116,241.52 590,551.29 431,129.86
PPB K
p
(m
−1
) 6.54⋅10
−4
1.40⋅10
−3
5.97⋅10
−4
SSR 77,545.23 852,391.39 286,513.46
CF K
c
(min⋅m
−2
) 1.45⋅10
−5
4.42⋅10
−6
3.86⋅10
−6
SSR 32,776.51 255,336.20 129,005.00
Meh a’s model
J∞1 (LMH) 46.05 114.16 116.08
α
(min −1)3.88 30.70 30.70
β(min −1)5.20⋅10
−2
1.33⋅10
−2
1.26⋅10
−2
SSR 700.38 1401.89 783.73
D. Nú˜
nez e al.
Jou nal o Wa e P ocess Enginee ing 55 (2023) 104086
7
ku osis in o de o analyse i s ela ionship wi h ouling (Table 5).
The loss o no malised lux was in e sely ela ed o R
a
and ms
alues. In his sense, PES50, whose no malised lux dec eased he leas
(77 %) du ing he ul a il a ion o he oxidised WAS, also showed he
lowes R
a
, ms and peak- o-peak dis ance alues. Fu he mo e, he
highes alues we e ob ained wi h PESH50, which los he highes
p opo ion o no malised lux (96 %). The di e ence be ween no mal-
ised luxes o PES50 and PAN50 was less ma ked han he oughness
alues may sugges , which could be explained based on hei di e en
hyd ophilic p ope ies ( ouling allegedly inc eases wi h hyd ophobici y
[61]). The ela ionship be ween loss o no malised lux and R
a
and ms
alues is in acco dance wi h he exis ing li e a u e, indica ing ha su -
ace oughness plays a key ole in lux loss [61,64–66]. As i was p e-
iously commen ed, in he ini ial s ages o il a ion, he pa icles end o
deposi in he “ alley-like” o ma ions o he memb anes, clogging hese
dep essed egions. Memb anes wi h lowe su ace oughness p esen
ewe “ alley-like” o ma ions on hei su ace, so he a achmen o
solu e molecules is es ic ed [64]. The e ec o his ewe p esence o
alley-like egions can be a ended con as ing he oughness alues
om Table 5 wi h he images in Fig. 4: alley o ma ions we e mo e
e iden in PAN50 and PESH50 han in PES50, and he dep h o he
alleys was lowe in he la e memb ane, as i can be con i med by he
peak- o-peak dis ance.
As p o ed by he pe meabili y es s and he AFM measu emen s, lux
loss is a complex phenomenon, which depends on se e al ac o s, among
which, he na u e o he memb ane su ace (hyd ophilic su aces p e-
en ouling) and i s ugosi y ( he highe he ugosi y, he mo e space
he solu e molecules ha e o deposi ing) can be conside ed
de e minan .
3.3. Memb ane molecula weigh cu -o sc eening
3.3.1. Pe meabili y es s
As PES was selec ed as he mos sui able ma e ial o WAS ac-
ioning (Sec ion 3.2.2), he MWCO sc eening expe imen s we e ca ied
ou wi h memb anes made o his polyme ic ma e ial (PES10 and PES3).
The lux ob ained wi h PES3 was ex emely low (a ound 0.5 LMH on
a e age), making his il a ion un easible, and hus no da a om his
expe imen a e depic ed. The luxes ob ained wi h PES10, PES50-10 and
PES50-3 a e shown in Fig. 5.
The highes lux ob ained du ing he MWCO sc eening was achie ed
wi h PES50-10. Howe e , he low ejec ion coe icien s obse ed made
his op ion un easible (Table 6). The highes ejec ions in e ms o CN,
TOC and COD we e achie ed wi h PES50-3. Ne e heless, in he case o
he biomolecules, sligh ly di e ences we e obse ed in ejec ions and
ejec ion di e ences be ween PES10 and PES50-3.
Based on he ejec ion coe icien s and he obse ed dec ease in lux
du ing he ul a il a ion o he oxidised WAS using PES10, as well as he
pe mea es ob ained om PES50 wi h PES10 and PES3, i can be
concluded ha he majo i y o molecules e ained by he PES10 mem-
b ane can also be e ained by he PES50 memb ane. The e o e, coupling
hese wo memb anes would be edundan and unnecessa y. This
beha iou was in acco dance wi h he wo k by U ea e al. [67], whe e
he e ec o WO on he di e en molecula weigh ac ions o WAS was
s udied. They epo ed ha , a e a WO ea men a 190 ◦C and 90 min,
he molecula weigh o he majo i y o he p esen molecules was
comp ised in he anges be ween 0 and 35 kDa ( e e ed by U ea e al.
as low molecula weigh molecules) and 35–150 kDa (medium molec-
ula weigh molecules). Mo eo e , hyd ophobic molecules we e also
p esen due o he in e ac ions wi h size-exclusion column. Acco ding o
he esul s a ained in his s udy, he sizes o he majo i y o he low
molecula weigh molecules we e comp ised be ween 0 and 10 kDa, and
Fig. 4. AFM images o clean and ouled PES50 (A and B), PESH50 (C and D) and PAN50 (E and F) o memb ane su ace mo phology analysis.
Table 5
Memb ane su ace oughness pa ame e s.
Clean Fouled
PES50 PESH50 PAN50 PES50 PESH50 PAN50
R
a
[nm/
μ
m] 23.0 ±
0.6
74 ±5 67 ±8 56 ±
15
164 ±
37
92 ±6
ms 30.4 ±
0.9
94 ±7 89 ±5 41 ±
12
212 ±
47
119 ±7
peak- o-
peak
dis ance
[nm/
μ
m]
142 ±
5
391 ±
32
365 ±
16
255 ±
59
918 ±
208
487 ±
26
Skewness −0.23
±0.04
– −0.08
±0.03
– – −0.05
±0.03
Ku osis 3.35 ±
0.03
2.72 ±
0.05
2.8 ±
0.3
3.4 ±
0.4
2.89 ±
0.01
2.7 ±
0.1
D. Nú˜
nez e al.
Jou nal o Wa e P ocess Enginee ing 55 (2023) 104086
8
he ones o he medium molecula weigh molecules we e abo e 50 kDa.
The esis ance-in-se ies model (Fig. 6) showed ha app oxima ely
hal o he memb ane ouling du ing he il a ion wi h bo h PES10 and
PES50-3 was i e e sible, which con as o he esul s ob ained wi h
PES50, whe e only 33 % o he ouling esis ance was a ibu able o
i e e sible ouling. Mo eo e , a compa ison be ween he alues o he
R
e
and he R
i e
ob ained when il e ing wi h PES50 and PES10 shows
ha no signi ican di e ences could be ound be ween he alues o
e e sible ouling, whe eas he i e e sible one o he PES10 memb ane
was a ound 2.5 imes highe han ha o he PES50 memb ane. Thus,
he addi ional lux decay obse ed be ween he il a ion wi h PES10
and PES50 was exclusi ely due o an inc ease in i e e sible ouling. I
has been epo ed ha i e e sible ouling du ing he il a ion o na -
u al o ganic ma e is p ima ily caused by he hyd ophilic ac ion o
his o ganic ma e [68–70]. The e o e, he addi ional i e e sible
ouling obse ed in his s udy is likely a ibu ed o hyd ophilic sub-
s ances, p esumably oxidised HA, since hei ejec ion coe icien s we e
lowe han hose o he o he biomolecules, which sugges s a lowe
a e age size. In his sense, se e al au ho s ha e epo ed ha he
oxida ion o HA inc eased hei hyd ophilici y [71–74] by oxidizing
benzene g oups in o di e en aldehydes and ca boxylic acids [72,74].
A e a subsequen educ ion o he memb ane MWCO o 3 kDa, he
ouling p o ile emained simila (sligh ly mo e i e e sible han
e e sible ouling), p o ing ha he molecules e ained by he PES50
memb ane mainly caused e e sible ouling. In his sense, Taniguchi
e al. [57] compa ed he ouling o PES memb anes wi h MWCOs om
10 o 1000 kDa du ing he UF o na u al o ganic ma e , and epo ed
ha memb anes wi h lowe MWCO (10 and 30 kDa) showed highe
i e e sible ouling, al hough ouling was mos ly e e sible in all cases.
3.3.2. Fouling modelling
The i ing o he s udied ouling models o he expe imen al da a
co esponding o he il a ions wi h PES10, PES50-10 and PES50-3 is
shown in Fig. 7.
Simila ly o he il a ion wi h he 50 kDa memb anes, he bes i ing
He mia’s model was CF in he h ee il a ions. This esul s we e in
acco dance o hose epo ed by Pee a e al. [75] ela ed o he ul a-
il a ion o humic acid solu ions wi h PES memb anes. Howe e , as i
was discussed in he p e ious sec ion, he i ing showed ha his model
alone could no adequa ely explain he obse ed ouling endency. In
Fig. 5. Flux a ia ion o e VCR o he oxidised WAS il a ion wi h PES10 (a) (J
0
=101.5 ±0.9 L/m
2
h), PES50-10 (b) (J
0
=102 ±3 L/m
2
h) and PES50-3 (c) (J
0
=
13.72 ±0.01 L/m
2
h).
Table 6
Rejec ion coe icien s ob ained wi h PES10, PES50-10 and PES50-3.
PES10 PES50-10 PES50-3
RC
CN
0.90 ±0.01 0.35 ±0.01 0.9942 ±0.0009
RC
TOC
0.70 ±0.16 0.17 ±0.09 0.88 ±0.01
RC
COD
0.60 ±0.04 0.09 ±0.04 0.66 ±0.04
RC
CH*
0.79 ±0.05 0.14 ±0.05 0.83 ±0.05
RC
P o *
0.84 ±0.09 0.21 ±0.08 0.87 ±0.06
RC
HA*
0.63 ±0.05 0.15 ±0.06 0.67 ±0.06
*CH: ca bohyd a es; PROT: p o eins; HA: humic acids.
D. Nú˜
nez e al.
Jou nal o Wa e P ocess Enginee ing 55 (2023) 104086
9
his case, he ul a il a ion o bo h he oxidised WAS wi h PES10 and
he pe mea es ob ained om PES50 wi h PES10 and PES3 su e ed a
p opo iona ely highe i e e sible ouling compa ed wi h he ul a il-
a ion o oxidised WAS wi h he PES50 memb ane, which in all cases
could be mainly a ibu ed o PPB, as i was he second-bes i ing
He mia’s model (Table 7). The ac ha PPB emained he main i e-
e sible ouling mechanism in memb anes wi h lowe MWCO is in
acco dance wi h he li e a u e. Thus, i has been epo ed ha , when
Fig. 6. R
m
(
), R
e
(
) and R
i e
(
) a e il a ion wi h PES10 and PES50-3 memb anes.
Fig. 7. He mia’s (comple e po e blocking [ ], in e media e po e blocking [ ], pa ial po e blocking [ ] and cake o ma ion
[]) and Meh a’s ( ) lux models o PES10 (a), PES50-10 (b) and PES50-3 (c) expe imen al luxes (●).
D. Nú˜
nez e al.