Desalination 297 (2012) 8796

Contents lists available at SciVerse ScienceDirect

Desalination
journal homepage: www.elsevier.com/locate/desal

Reactive dyes rejection and textile efuent treatment study using ultraltration and
nanoltration processes
A. Aouni a, 1, C. Fersi a, c,, B. Cuartas-Uribe b, 2, A. Bes-Pa b, 2, M.I. Alcaina-Miranda b, 2, M. Dhahbi a, 1
a
b
c

Water and Membrane Technologies Laboratory, CERTE, BP 273, Soliman 8020 Tunisia
Department of Chemical and Nuclear Engineering, Polytechnic University of Valencia, Camino de Vera s/n, 46022 Valencia, Spain
Method and Technical Analysis Laboratory, INRAP, Sidi Thabet 2020, Tunisia

a r t i c l e

i n f o

Article history:
Received 30 November 2011
Received in revised form 18 April 2012
Accepted 22 April 2012
Available online 14 May 2012
Keywords:
Ultraltration
Nanoltration
Reactive dyes
Rinsing wastewater

a b s t r a c t
Ultraltration and nanoltration processes were used to treat synthetic reactive dyes aqueous solutions and a
raw textile efuent supplied from rinsing baths of Spanish textile industry. The inuence of the reactive dyes
molecular weights and the effect of the used membranes types and cut-offs were investigated with regard to
the permeate ux at different transmembrane pressures (27 bar for UF and 415 bar for NF). The extent of
colour retention, COD and conductivity was determined in order to monitor the membrane's separation efciencies aiming at waste water treatment, water reuse and chemical usage minimisation. High COD retentions (80100%) were achieved using UF and NF processes. Good conductivity rates (80%) and high COD
and colour retention rates (> 90%) were obtained for both NF 200 and NF 270 membranes for all studied
dyes solutions. An improvement of the rinsing wastewater quality was obtained using UF and NF processes.
2012 Elsevier B.V. All rights reserved.

1. Introduction
The characteristics of wastewater from textile processing operations are comprehensively reviewed. The categorisation of wastewaters proceeds through a consideration of the nature of the various
industrial processes employed by the industry and the chemicals associated with these operations. Chemical pollutants arise both from
the raw material itself and a broad range of additives used to produce
the nished product. The industrial categories considered include
sizing and weaving, scouring, bleaching, mercerising, carbonising, dyeing and nishing. Pollutants of concern range from non-biodegradable
highly-coloured organic dyes to pesticides from special nishes such
as insect-proong [1].
Textile industries consume large volumes of water and chemicals
for wet processing of textiles. The chemical reagents used are very diverse in chemical composition, ranging from inorganic compounds to
polymers and organic products [2,3]. Chitin and chitosan were used in
lobster shell wastes for colour removal from aqueous solutions [4].
The presence of very low concentrations of dyes in efuent is highly
visible and undesirable [5]. Due to their chemical structure, dyes are
resistant to fading on exposure to light, water and many chemicals
[6,7].

Corresponding author at: Method and Technical Analysis Laboratory, INRAP, Sidi
Thabet 2020, Tunisia. Tel.: + 216 71 537 666; fax: + 216 71 537 767.
E-mail address: cheimafersi@yahoo.fr (C. Fersi).
1
Tel.: + 216 79 325 798; fax: + 216 79 325 802.
2
Tel.: + 34 96 3879633; fax: + 34 96 3877639.
0011-9164/$ see front matter 2012 Elsevier B.V. All rights reserved.
doi:10.1016/j.desal.2012.04.022

Many dyes are difcult to decolourise due to their complex structure and synthetic origin.
There are many structural varieties, such as, azo, diazo, anthroquinone,
triphenylmethane, phetalocyanine, stilbene, and metal complex dyes.
Decolouration of textile dye efuent does not occur when treated aerobically by municipal sewerage systems [8].
The conventional method of textile wastewater treatment consists
of chemical coagulation (using ferrous, lime and polyelectrolytes), biological treatment followed by activated carbon adsorption. The conventional coagulation process generates huge volume of hazardous
sludge and poses a problem of sludge disposal. In order to meet the
legal requirements for the discharge of textile wastewater, researchers
are attempting a combination of two or more treatment methods for
the complete and successful removal. Combination of electrochemical
treatment and chemical coagulation [9], combined chemical coagulation, electrochemical oxidation, and activated sludge process [10],
and combination of electrochemical method, chemical coagulation,
and ion exchange [11] were reported for textile efuent treatment
to comply with the legal requirements or for reuse standards. Each
treatment method has its own advantages and disadvantages, and
the selection of the method mainly depends on the treatment target
to be achieved.
The uctuating compound concentrations and ow rates make the
conventional processes quite insufcient for the treatment of textile
wastewaters, especially for colour removal [12,13].
The discharge regulations are becoming more stringent, and there is
a growing tendency and interest in the advanced treatment methods
like ozonation, photo catalysis, and membrane processes for a better
treatment of the textile wastewaters [14].

88

A. Aouni et al. / Desalination 297 (2012) 8796

NaO3SOCH2CH2O2S

SO2CH2CH2OSO3Na
OH
N

NaO3S

H2N
N

SO3Na

Fig. 1. Chemical structure of Everzol Black (MBlack = 991 g mol 1).

The recovery and reuse of textile wastewaters have been studied

extensively by membrane researchers [1524]. Membrane techniques hold great promise in this eld, as they have the potential to
either remove the dyes stuffs or allow reuse of the auxiliary chemicals
used for dyeing or to concentrate the dyestuffs and auxiliaries and
produce puried water.
In fact, membrane processes can be used for the purication of these
complex wastewater streams [2125]. There have been many investigations on the treatment of the dye house efuents using microltration
[26,27], ultraltration [26,2832], nanoltration [26,3339], and reverse
osmosis membranes [26,4044].
Several applications of the ultraltration process (UF) have been
reported for the separation of certain dyes, such as indigo, direct,
disperse and reactive dyes [2530]. A special attention was given to
the application of membrane processes; especially the nanoltration
(NF), on the rejection of reactive dyes, whose xation rates on the
fabric is one of the lowest (6090%) [3133].
In this work, different UF and NF membranes were used to treat
three synthetic reactive dyes aqueous solutions (using black, blue and
red EVERZOL dyes). The effect the of dye molecular weight (MW), the
cut-off and the type of used membranes were investigated with regard
to the permeate ux and separation efciency.
These processes were also used to treat a rinsing textile efuent
supplied from rinsing baths of Spanish textile industry. The extent
of colour retention, COD and conductivity were determined in order
to monitor the membrane's separation efciencies aiming at waste
treatment, water reuse and chemical usage minimisation.
2. Materials and methods
2.1. Synthetic model solutions
Experiments were conducted using three reactive azo dyes (Everzol
Black, Everzol Blue and Everzol Red) which are known to contain anionic sulphonate groups to various degrees.
Three synthetic model solutions were prepared using reactive azo
dyes and chloride sodium. The three synthetic solutions of azo reactive dyes were prepared by dissolving 15 g of Everzol dye powder in
a ask containing two liters of ultrapure water. This mixture was
heated for 30 min at a temperature of 80 C to ensure complete dissolution of the dye. After cooling, the contents of the ask were poured

into a tank of 25 L. Then, an amount of 12.5 g of NaCl was added and

supplemented with ultrapure water up to the mark.
Just after, pH was adjusted to 10 with 5 M sodium hydroxide solution and the solutions prepared were homogenised by simple magnetic stirrer. The dye concentration and NaCl concentration are
equal to 600 mg L 1 and 500 mg L 1 respectively.
The choice of pH value is in relation with the simulation of dies
solutions to a real textile wastewater.
In fact, the real textile wastewater is highly alkaline (pH 10.012.5)
and is characterised by strong colour and high pH. The textile industry
plant consists mainly of printing and dying processes. In these processes, the use of substances such as NaCl and Na2CO3 is required to improve
the aggregation of dye ions on the ber to x the dyestuffs.
The molecular structures and weights of the three dyes are presented in Figs. 1, 2 and 3.
The main characteristics of these solutions are shown in Table 1.
2.2. Textile wastewater characteristics
The study was conducted with a textile wastewater sample supplied from the rinsing baths of COLORTEX textile industry, Spain.
The main characteristics of this sample are given in Table 2.
2.3. Analytical methods
pH values were determined by means of a GLP 22 pH-meter
(Crison). The conductivities were measured by a Crison (GLP 32) instrument type conductivitymeter. Both conductivity and pH sensors
used for these analyses allowed automatic and continuous correction
of the values by taking into account the sample temperature. Turbidity was measured with a DINKO D-112 turbidimeter according to the
ISO 7027:1999. Turbidity and conductivity were measured with accuracies of 2 NTU and 1.0%, respectively. The COD concentrations
were obtained using a Spectroquant Nova 60 from MERCK (Germany)
type COD-meter, whereas SAC (spectral absorption coefcient) values
were obtained by UVvis absorption using a HP 8453 spectrophotometer (1 cm cell width) after samples ltration with a 0.45 m lter,
according to the ISO 7887:1994 method [45].
COD analyses were carried out in a cell tests from MERCK (denominated kits). These tubes contain the required reagents for the
oxidation (potassium dichromate, sulphuric acid and silver sulphate).

NH2
SO3Na

Cl
N
N

SO3Na

SO3Na

N
Ph

HO

N
O

HN

N
Me
NaO3S
Fig. 2. Chemical structure of Everzol Red (MRed = 788 g mol 1).

SO2CH2CH2OSO3Na
Fig. 3. Chemical structure of Everzol Blue (MBlue = 626 g mol 1).

A. Aouni et al. / Desalination 297 (2012) 8796

Table 1
Synthetic model solutions characteristics.
Parameter
(g mol 1)

MW of dye
pH
Conductivity
Colour
COD
[Dye]
[NaCl]

(S cm 1)
(mg L 1)
(mg L 1)
(mg L 1)

Everzol Black
solution

Everzol Blue
solution

Everzol Red
solution

991
9.89
2490
2.05
566
600
500

626
10.46
1407
1.92
450
600
500

788
10.27
2240
1.76
470
600
500

For the COD measurements, accuracies were depending on the COD

range measured: 1.2 for 10150 mgCOD L 1 and 2.0 for 50
500 mgCOD L 1. Colour based on the spectral absorption coefcient
(SAC) values at three different standard wavelengths in the visible
range (436, 525, and 620 nm) were measured. The colour intensity
of dyes solutions and textile efuent were analysed by HEWLETT
PACKARD HP-84-53 type spectrophotometer, according to the standard method EN ISO 7887:1994 (Section 3) [45] which measures
the sample absorbance at three wavelengths in the visible range,
and determined using Eq. (1).
Colour

A2436nm A2525nm A2620nm

A436nm A525nm A620nm

where, A represents the absorbance values measured at the different

standard wavelengths (436 nm, 525 nm and 620 nm) [46].
Colour retention and salt retention were calculated by the following equation:
R 100 

Cp
1
Cf

!
2

Where Cp and Cf are respectively permeate and feed concentrations.

The main ions (sodium, potassium, magnesium, calcium, sulphate,
and chloride) and the total dissolved solids were determined according
to the Standard Methods [45]. The particle size distribution (PSD) was
measured using Mastersizer 2000 from Malvern Instruments according
to the ISO 133201 [47].
2.4. Membranes
Two nanoltration membranes (NF200 and NF270) provided by
Dow/Filmtec (USA) and two ultraltration membranes (10 kDa and
1 kDa) purchased respectively from Rhodia Orelis and Synder Filtration
manufacturers were used in this work. Polyethersulfone membranes
(PES) are widely used for UF applications because they have wide temperature limits, wide pH tolerances, fairly good chlorine resistance and
wide range of pore sizes.

Table 2
Textile wastewater characteristics.
Parameter
pH
Conductivity
Turbidity
COD
Particle size distribution
TDS
Colour
Cl
SO42
Ca2+
Mg2+
Na+

Value
(S cm 1)
(NTU)
(mg L 1)
(m)
(mg L 1)
1

(mg L )
(mg L 1)
(mg L 1)
(mg L 1)
(mg L 1)

7.11
3840
4.02
708
0.768
3137
0.12
>125
28
42
18.6
1008

89

Table 3 provides the specication of these membranes as given by

the manufacturers. The effective surface for all these membranes is
equal to 90 cm 2. The water permeability was established for all used
membranes. Values are shown in Table 3.
There are great differences between NF 200 and NF 270, especially in
terms of pore diameter size and thickness. In fact, UF membranes have
the greater values (1100 nm as a pore size diameter and 150 10 3 nm
as a thickness) comparing with NF membranes (0.93 and 0.84 nm as a
pore size diameter and 1.81 103 nm and 1.87 103 nm as a thickness).
These values conrm that UF membrane retains the larger molecules
that NF membrane cannot reject.
2.5. Membrane ltration units
Membrane experiments were performed in a pilot plant designed
in the Department of Chemical and Nuclear Engineering, at the Polytechnic University of Valencia. This pilot plant is equipped with a
plane and at module.
The feed rectangular tank (25 L) included a mixer model RZR-1
(Heidolph) to homogenise the feed solution, and a level switch that
enables continuous operation of the pilot plant. Permeate and concentrate streams were recycled and returned to the feed tank. Then,
it can be considered that the concentration of the feed solution was
kept constant for each experiment.
As temperature affects permeate ux, this variable was kept constant by a cooling system at 22 C. A 10 m pre-lter was placed
upstream of a variable speed pump motor model 270 (cat pumps),
in order to reduce the pass of particles into the system. Next to the
feed pump exit there was a retentate security valve situated just
before the owmeter model R-005 (Comaquinsa S.A.) which was
used to regulate the transmembrane pressure and to limit the maximum operating pressure in the pilot plant. Two manometers at each
side of the membrane module and a pressure tube of stainless steel
were used to measure the transmembrane pressure. A owmeter
was located in the retentate line to measure the retentate volumetric
ow rate. The operating feed ow rate was xed to 400 L h 1 for all
experiments. The experimental set-up is illustrated in Fig. 4.
The operating time was established according to the steady state
conditions. It was about 120 min. Permeate uxes (JP) and retention
rates (R) were measured every 120 min. Thus, a total recirculation
mode was used at the beginning of the experiment and a constant
permeate volume was extracted from the system when the steady
state was reached. The operating transmembrane pressure (P) was
varied. In each series of experiments, three pressures were applied
(357 for UF membranes and 51015 for NF membranes).
After thoroughly cleaning the system several times using chemical
process and pure water, the membrane permeability was re-evaluated.
The rst step of the chemical cleaning consists on circulating distilled
water in the system for 40 min and then applying basic cleaning using a
sodium hydroxide solution (5 mol L 1) at pH=11during 20 min. The
second step of the procedure consists in rinsing again with distilled
water for 40 min and then applying acidic cleaning using a chlorhydric
acid solution at pH =3 for 20 min. Finally, a rinsing step with distilled
water was applied for 40 min.
The steps of membrane cleaning procedure were followed without applying any pressure and increasing the feed ow rate in order
to eliminate the organic and inorganic substances that clogged the
membrane. After that, the membrane permeability was re-evaluated
in order to check the membrane fouling state.
3. Results and discussions
3.1. Reactive dyes rejection using ultraltration
Synthetic model solutions were prepared at xed dye and NaCl
salt concentrations which were respectively 600 and 500 mg L 1 as

90

A. Aouni et al. / Desalination 297 (2012) 8796

Table 3
Membrane specications.
Membrane

UF 10 kDa

UF 1 kDa

NF 200

NF 270

Manufacturer
Material
Surface charge (pH 7)
Maximum operating pressure (bar)
Maximum operating temperature (C)
pH range
Cut-off (Da)
Contact angle ()
Isoelectric point
Roughness (nm)
Pore diameter size (nm)
Thickness (nm)

Rhodia orelis
Polyethersulfone
Neutral
7
50
314
10,000
56 38 [52]
N.D.
12.7 1.3A [52]
1100 [58]
150 103 [58]

Synder ltration
Polyethersulfone
Neutral
8
50
211
1000
56 38 [52]
N.D.
12.7 1.3A [52]
1100 [58]
150 103 [58]

Dow/Filmtec
Polyamide
Negative [51]
41
45
310
200300
30.3 [53]
4.3
5.19 [53]
0.93a
1.81 103a

Dow/Filmtec
Polyamide
Negative [51]
41
45
310
200300
51.4 [56]
3.3 [57]
19.80 [57]
0.84 [55]
1.87 103 [59]

N.D: Not Dened.

a
By authors.

given in Table 1. The transmembrane pressure was varied from 3 to

7 bar during each experiment. Permeate uxes of Everzol dyes solutions versus transmembrane pressure using UF 10 kDa and UF 1 kDa
membranes are presented in Fig. 2. The equations of linear parts of
obtained plots are shown in corresponding gures. It is evident that
membrane uxes were lower than for the corresponding pure water
ux. This is a phenomenon that can be related to the presence of organic and inorganic solutes, which can be retained at the membrane
surface causing concentration polarisation [34].
However, fouling due to dissolved inorganic and organic compounds specically is more problematic to nanoltration [48]. Membrane fouling can lower the permeate ux and might inuence the
retention of organic compounds [49]. The factors that affect membrane
fouling can be classied into three categories: membrane properties,
operating parameters and rinsing wastewater itself [50].
Membrane fouling is a complicated phenomenon of which causes
of fouling are multiple and poorly understood. The core of the issue is
interactions between fouling materials and membrane, and between
fouling materials themselves. It has been recognised that electrostatic
interaction and hydrophobic/hydrophilic interaction between membranes and fouling materials have a signicant bearing on membrane
fouling. This is particularly true to more difcult fouling problems caused by adsorption of natural organic matters and biopolymers on the
membrane. The balance between the forces of electrostatic repulsion

and hydrophobic adhesion determines the outcomes of membrane

fouling, as well as the efciency of chemical cleaning.
Dye molecules can be adsorbed on membrane surface, which affects
the permeate ux due to physicochemical interactions, i.e. hydrophobic
interactions (dispersion forces), polar interactions (dipole forces) and
charge transfer (hydrogen bonding). The nature of the membrane material, the type of solute, the solute concentration, crossow velocity
and pH are the most important parameters that can determine the
extent of dye adsorption [34].
Organic adsorption was reported to cause the greatest synergistic
effect due to the ability to reduce repulsive interaction between
foulants (organic and inorganic colloidal foulants) and the membrane
as well as that among foulants. After the adsorption taken place, the
dye aggregates will build multiple sublayer and nally the bulk concentration of dye will increase after the compacted sublayer had
been developed.
During the permeation of dye solutions, the adsorption of dye
molecules at the membrane surface and inside the pores contributes
to the decrease of the membrane pore sizes and as a consequence
to the alteration of the pure water permeability coefcient, Lp.
Dye molecules have amino, sulphonic and hydroxyl groups as substituents bound to the aromatic rings. Those functional groups could
interact with the membrane that has also functional groups as carboxyl, amide and amino.

Fig. 4. Membrane ltration experimental set-up.

A. Aouni et al. / Desalination 297 (2012) 8796

100
90
80

RBlack (%)

70
60
50

COD(Black)

40

Colour(Black)

30

Conductivity(Black)

20
10
0
0

J0 (Water) = 62.8 P

Jp (L h-1 m-2)

390

Jp (Black) = 46.4 P - 1.6

360
330
300
270
240
210
180
150

90
80
70
60
50
40
COD(Black)

20

Colour(Black)

10

Conductivity(Black)

P (bar)

Jp (Red) = 49.2 P - 0.03

Fig. 6. Retention rates as a function of transmembrane pressure after Everzol Black solution treatment using UF membranes ([Dye] = 600 mg L 1; [NaCl] = 500 mg L 1).
A) UF 10 kDa; B) UF 1 kDa.

P (bar)
30
J0 (Water) = 3.7 P
Jp (Black) = 1.5 P - 0.2

25

Jp (L h-1 m-2)

Jp (Blue) = 55.8 P - 0.1

120
90
60
30
0

100

30

P (bar)

RBlack (%)

The nature of the membrane material, the type and concentration

of solute, cross-ow velocity and pH are the parameters that can affect the dye adsorption in dye and NaCl mixtures. In each case, different electrostatic interactions take place between dye, NaCl and the
membrane.
Fig. 5A. shows that permeate uxes plots were not linear for a
transmembrane pressure higher than 3 bar. An important decline of
permeate uxes were observed for all Everzol dyes solutions at high
transmembrane pressures using UF 10 kDa membrane. However, in
the case of UF 1 kDa membrane, all plots were linear for the studied
transmembrane range as shown in Fig. 5B. This result demonstrates
that an important concentration polarisation occurs using the rst
UF membrane for transmembrane pressure higher than 3 bar.
Fig. 5 illustrates that Everzol Blue solution presented the higher
permeate uxes and Everzol Black solution presented the lowest permeate uxes for the whole transmembrane range. This is related to
the difference between molecular weights of studied dyes as given
in Table 1.
Fig. 6 shows the evolution of colour, COD and conductivity retention rates with transmembrane pressure in the case of Everzol Black
solution using respectively UF 10 kDa and UF 1 kDa membranes. The
same experiment conditions were used for Everzol Blue and Everzol
Red solutions as shown in Figs. 7 and 8. High COD retention rates
(>80%) were obtained for all dyes solutions using both UF 10 kDa
and UF 1 kDa membranes.

91

Jp (Blue) = 1.9 P - 0.4

Jp (Red) = 1.5 P - 0.3

20
15
10
5
0
0

P (bar)
Fig. 5. Permeate ux of Everzol dyes solutions versus transmembrane pressure using UF
membranes ([Dye] = 600 mg L 1; [NaCl]= 500 mg L 1). A) UF 10 kDa; B) UF 1 kDa.

It can be observed that transmembrane pressure affects especially

salt retention. In fact, an increase of the conductivity rate was observed by increasing transmembrane pressure from 3 to 7 bar for all
dyes solutions.
When UF membrane was placed in contact with an ionic solution
(Everzol solution or textile efuent), the association and dissociation
of functional group on the surface of UF membrane leads to the formation of charges. Thus, the charge generated can be related to the
ionisation of functional group existing on the surface of the membrane, which strongly depends on pH, thereby resulting in acid and
base dissociation. At pH = 10, if the charge generated is negative, a repulsion strength with negative ions appears and leads to a decrease of
the conductivity value.
Other possible interpretation can also be taken into consideration:
Na + has a strong afnity for sulfonate groups to complex with anionic
dye molecules and partially neutralises their negative charge. Due to
the opposite charge of solutes (cationic salt particles and anionic
dye), the electrostatic interactions prevent the formation of highmolecular-weight complexes between dye molecules and salt particles. This can affect the solution conductivity.
Due to its higher molecular weight cutoff, UF 10 kDa membrane
presented retention rate values lower than those obtained with UF
1 kDa membrane in terms of conductivity, colour and COD. As an example, Fig. 7 shows that conductivity rates for Everzol Blue solution
didn't exceed 40% using UF 10 kDa membrane. However, these rates
increased to 80% in the case of UF 1 kDa membrane as shown in
Fig. 7B. This result can be related to the size of the complex formed
and the pore sizes of the membranes.

92

A. Aouni et al. / Desalination 297 (2012) 8796

100
90
80

RBlue (%)

70
60
COD(Blue)

50

Colour(Blue)

40

Conductivity(Blue)

30
20
10
0
0

P (bar)

100
90
80

RBlue (%)

70
60
50
40

pressure using NF 200 and NF 270 membranes are presented in Fig. 9.

The equations of obtained plots are shown in the corresponding gure
for pure water and studied dyes solutions.
We notice that permeate ux increases proportionally with the
pressure drop within the pressure range studied in all cases, which
indicates that the operation is in the pressure- controlled region.
Permeate uxes using dye solutions were lower than the
corresponding pure water uxes for both NF 200 and NF 270. In the
case of nanoltration processes, this phenomenon can be related,
not only, to the concentration polarisation but also to the presence
of an important osmotic pressure near the membrane-efuent interface which leads to a decrease in the effective driving force [32]. In
fact, the retention of the ionic species results in an osmotic pressure
build-up across the membrane. The osmotic pressure causes ux to
decline, but this is due to reduced driving force rather than an increase in the resistance to mass transfer.
It can be also observed that the initial deviation of the permeate
ux depends on the molecular weight of used dyes. In fact, Everzol
Blue which characterised by the smaller molecular weight presented
the lower intercept as shown in Fig. 9 (1.5 L h 1 m 2 for NF 200
membrane and 1.1 L h 1 m 2 for NF 270 membrane). This result
proves that osmotic pressure occurs progressively in the case of
Everzol Blue solution. Nevertheless, this solution presented the lower
permeability (the lower slope which was about 10.3 L h 1 m 2 bar 1
in the case of NF 200 and 9.3 L h 1 m 2 bar 1 in the case of NF 270).
This observation is related to pore plugging with Everzol Blue molecules
which can be increased at high transmembrane pressure.

COD(Blue)

30

Colour(Blue)

20

Conductivity(Blue)

10

100
90

0
0

80

P (bar)

Although Everzol Black dye presents the higher molecular weight

comparing to Everzol Blue and Everzol Red, its corresponding colour
retention rates were the lowest ones. In fact, Fig. 6 shows that black
retention rates didn't exceed 15% in the case of UF 10 kDa membrane
and 80% in the case of UF 1 kDa membrane even at high transmembrane pressure. A total decolourisation was observed for blue and
red dyes using UF 1 kDa membrane. This result could be related to
the spatial conguration of the Everzol black dye.
In fact, the chemical structure (shape, number of active groups,
stability of a negative charge) of the dye could give the difference in
the afnity of these reactive dyes toward surface membrane.
Comparing the performances of the used ultraltration membranes (UF 10 kDa and UF 1 kDa) in the treatment of the studied reactive dyes solutions, we can conclude that ultraltration seems to
be an appropriate technology for the COD decrease. In addition, ultraltration using UF 1 kDa membrane could be used as pretreatment
step especially for highly charged efuent. This membrane process
could be followed by nanoltration or reverse osmosis processes
depending on the efuent characteristics and the treatment objective.
3.2. Reactive dyes rejection using nanoltration

RRed (%)

COD(Red)

60

Colour(Red)

50

Conductivity(Red)

40
30
20
10
0
0

P (bar)

100
90
80
70

RRed (%)

Fig. 7. Retention rates as a function of transmembrane pressure after Everzol Blue solution treatment using UF membranes ([Dye] = 600 mg L 1; [NaCl] = 500 mg L 1).
A) UF 10 kDa; B) UF 1 kDa.

70

60
50
40
30

COD(Red)

20

Colour(Red)
Conductivity(Red)

10
0

The same reactive dyes solutions using Everzol Black, Blue and Red
dyes and NaCl were treated using two nanoltration membranes (NF
200 and NF 270) which are described in Table 3. The transmembrane
pressure was varied from 3 to 12 bar during each experiment. The variation of permeate uxes of Everzol dyes solutions with transmembrane

P (bar)
Fig. 8. Retention rates as a function of transmembrane pressure after Everzol Red solution treatment using UF membranes ([Dye] = 600 mg L 1; [NaCl] = 500 mg L 1).
A) UF 10 kDa; B) UF 1 kDa.

A. Aouni et al. / Desalination 297 (2012) 8796

A
240

J0 (Water) = 12.6 P
Jp (Black) = 12.3 P - 16.8

210

Jp (Blue) = 10.3 P - 1.5

Jp (L h-1 m-2)

180

Jp (Red) = 12.6 P - 3.1

150
120
90
60
30
0
0

10

12

14

16

18

P (bar)

B
J0 (Water) = 15.6 P

240

Jp (Black) = 11.3 P - 10.8

210

Jp (Blue) = 9.3 P - 1.1

Jp (Red) = 11.4 P - 7.4

Jp (L h-1 m-2)

180
150

93

can be concluded that permeate ux, contact angle and roughness

values conrmed the fouling tendency showed by the NF 270 membrane in the laboratory tests. This conrms that the NF 270 membrane is more sensitive to fouling phenomenon.
Fig. 10 shows the retention rates for colour, COD and conductivity
in the case of Everzol Black solution using respectively NF 200 and NF
270 membranes. The some experiment conditions were used for
Everzol Blue and Everzol Red solutions as shown in Figs. 11 and 12.
Comparing nanoltration membrane performances, similar results
were obtained: high COD and colour retention rates (>90%) were
obtained for both NF 200 and NF 270 membranes even using Everzol
Black dye as shown in Fig. 10. Conductivity retention rate exceeded
85%.
In fact, negatively charged membranes increase electrostatic repulsion of anionic compounds. The studied reactive dyes are characterised
by sulfonate groups which highly interact with the negative charge
membrane. This phenomenon justies the high colour retention and
the improvement of chloride retention which affects the conductivity
retention rates.
These results are very interesting in terms of environment, economy and industrial management using nanoltration process. This preliminary study using synthetic dye and salt mixtures is interesting for
predicting nanoltration performances in the treatment of industrial
efuents using similar experimental conditions.

120
90
60

30

100

0
0

10

12

14

16

18

It can be also observed that all dyes solutions presented a decrease

of the permeate ux comparing to the pure water ux for both NF 200
and NF 270 as shown in Fig. 9. This decrease was more important in
the case of NF 270 (from 27 to 40%) than NF 200 (from 0 to 18%)
for all used reactive dyes solutions and for all applied transmembrane
pressure. This demonstrates that NF 270 membrane is more sensitive
to fouling phenomena.
Contact angle is a semi-quantitative index associated to the hydrophilicity/hydrophobicity of the membrane surface. The lower contact
angle, the higher hydrophilicity and fouling tendency decreases [54].
In addition, the contact angle of the membrane 270 NF (51.4) is
higher than that of NF 200 (30.3), so the surface hydrophobicity of
the NF 270 membrane is greater, generating a hooking of particles
more accentuated on the surface of the membrane.
In this way, it seems that NF 200 membrane presented a larger
resistance to fouling because of its low contact angle value (30.3).
On the contrary, NF 270 membrane has the highest value (51.4)
and therefore the major ux ratio reduction and maximum fouling.
Other parameter to be considered is the surface roughness. This
parameter is a physical property that denes the surface morphology
which is also connected with the fouling tendency. In fact, the roughness of the NF 270 membrane is equal to 19.80 nm and it is greater
than that of the NF 200 membrane which is equal to 5.9 nm.
Particularly, the NF 270 membrane presented the maximum ux
decrease and the highest roughness value (19.80 nm). On the contrary, NF200 membrane showed the minimum ux decrease and
the smoothest surface (5.9 nm). This small roughness value allows
conrming the lack and the difculty for linking foulants. Thus, it

90

80
COD(Black)
Colour(Black)

10

Conductivity(Black)

0
0

10

12

14

16

P (bar)

B
100

RBlack (%)

Fig. 9. Permeate ux of Everzol dyes solutions versus transmembrane pressure using

NF membranes ([Dye] = 600 mg L 1; [NaCl] = 500 mg L 1). A) NF 200; B) NF 270.

RBlack (%)

P (bar)

90

COD(Black)

80

Colour(Black)
Conductivity(Black)

10
0
0

10

12

14

16

P (bar)
Fig. 10. Retention rates as a function of transmembrane pressure after Everzol Black solution treatment using NF membranes ([Dye] = 600 mg L 1; [NaCl] = 500 mg L 1).
A) NF 200; B) NF 270.

94

A. Aouni et al. / Desalination 297 (2012) 8796

100

90

90

RBlue (%)

RRed (%)

100

80

80

COD(Red)

COD(Blue)

Colour(Red)

Colour(Black)

10

Conductivity(Red)

10

Conductivity(Black)

0
0

10

12

14

16

10

12

14

16

P (bar)

P (bar)

100

90

90

RBlue (%)

RRed (%)

100

80

80

COD(Red)

COD(Blue)

Colour(Red)

Colour(Blue)
Conductivity(Blue)

10

Conductivity(Red)

10
0

0
0

10

12

14

16

10

12

14

16

P (bar)

P (bar)
Fig. 11. Retention rates as a function of transmembrane pressure after Everzol Blue
solution treatment using NF membranes ([Dye] = 600 mg L 1; [NaCl] = 500 mg L 1).
A) NF 200; B) NF 270.

Fig. 12. Retention rates as a function of transmembrane pressure after Everzol Red
solution treatment using NF membranes ([Dye] = 600 mg L 1; [NaCl] = 500 mg L 1).
A) NF 200; B) NF 270.

3.3. Application of membrane processes in textile efuent treatment

water produced is called dye bath water and after washing the waste
water generated is called wash water. Dye Bath contains higher solids
in the range 45% whereas wash water contains only 0.51% solids.
The main characteristics of this sample were given previously in
Table 2. It can be observed that this efuent presents particularly high
conductivity (l4 mS cm 1) and high COD (708 mg L 1) which
causes environmental problems and makes impossible its reuse.
30
25

Jp (L h-1 m-2)

3.3.1. Application of ultraltration in textile efuent treatment

Ultraltration process using UF 1 kDa membrane was used in the
treatment of an industrial wastewater supplied from the rinsing
baths of a Spanish textile industry.
The textile industry plant of the study mainly consists of printing
and dying processes. In these processes, the use of substances such
as NaCl and Na2CO3 is required to favour the aggregation of dye
ions on the bre to x the dyestuffs.
About 10% of the total wastewater comes from the dying and printing residual baths whereas 85% comes from rinsing. Hence, water
recycling in rinse process is the key point for saving fresh water in this
industry.
Rinsing efuents were considered for on-site treatment and reuse
in the same or in other production processes, being characterised by a
large volume but low concentration of contaminants. Their direct
reuse is not suggested for the wide quality variability in con contrast
to the standard quality required in the processes, which cannot be assured without a proper treatment of the raw efuents.
Textile dyeing industries need huge quantity of water for textile
dyeing, which they normally pump out repeatedly from the ground
or natural water sources resulting in depletion of ground water level.
In the dyeing process textile industries generate huge quantity of
toxic efuent containing colours, sodium sulphate, sodium chloride,
sodium hydroxide and traces of other salts. These are generated after
dyeing and after washing of garments/fabrics. After dyeing the waste

J0 (Water) = 3.7 P
Jp (Rinsing) = 2.0 P - 0.5

20
15
10
5
0
0

P (bar)
Fig. 13. Rinsing wastewater permeate ux versus transmembrane pressure using UF
1 kDa membrane.

100

100

90

90

80

80

70

70

60

RRinsing (%)

RRinsing (%)

A. Aouni et al. / Desalination 297 (2012) 8796

COD
Colour
Conductivity

50
40

95

60
50
40

30

30

20

20

10

10

COD
Colour
Conductivity

0
0

9 10 11 12 13 14 15 16

P (bar)

P (bar)
Fig. 14. Retention rates as a function of transmembrane pressure after rinsing wastewater treatment using UF 1 kDa membrane.

Fig. 16. Retention rates as a function of transmembrane pressure after rinsing wastewater treatment using NF 270 membrane.

The variation of water and permeate uxes with the transmembrane pressure is presented in Fig. 13. Similar results were observed
comparing to Everzol Blue solution using the same ultraltration
membrane.
Fig. 14 illustrates COD, colour and conductivity retention rates of
the rinsing wastewater after ultraltration process. More than 90%
of colour was retained by UF 1 kDa membrane for a transmembrane
pressure varied from 3 to 7 bar. Good COD retention was also
obtained (70% at 3 bar and more than 80% at 7 bar).
The same gure demonstrates that ultraltration didn't highly improve the quality of the studied textile efuent in terms of conductivity. In fact, conductivity retention rates didn't exceed 30% even at high
transmembrane pressure.
These results conrm the efciency of ultraltration process in
COD reduction and offer the possibility of using this process as pretreatment to improve the waste water quality.

COD, colour and conductivity retention rates of the rinsing wastewater after nanoltration process are shown in Fig. 16. More than 95%
of COD retention and more than 98% of colour retention were observed for all transmembrane pressure range using NF 270 membrane. It can be also observed that an important improvement was
occurred for conductivity retention rates. In fact, this parameter increases with the transmembrane pressure and reaches 71% at
15 bar. These results demonstrate that nanoltration highly improves
the rinsing wastewater quality.

3.3.2. Application of nanoltration in textile efuent treatment

The same industrial efuent was treated using NF 270 membrane.
Fig. 15 illustrates the variation of water and permeate uxes with the
transmembrane pressure. Comparing to results obtained previously in
Fig. 9B, it can be observed that the initial deviation of the permeate
ux was more important in the case of the industrial efuent sample
than using the Everzol dyes solutions. This result conrms the complexity of the wastewater matrix which leads to a rapid pore plugging.
200
180

J0 (Water) = 9.7 P

160

Jp (Rinsing) = 6.0 P - 15.8

Jp (L h-1 m-2)

140

4. Conclusions
In this study, membrane separation efciencies were evaluated by
determining three parameters retentions (colour, COD and conductivity) for treating textile model dyes solutions and real rinsing efuents. Addition of chemicals was not required, to favorise better
permeate quality. A step forward in water treatment is to obtain an
efuent quality suitable to produce reusable water.
The following conclusions are made based on this experimental
study:
High COD retention rates (>80%) were obtained for all dyes solutions using both UF 10 kDa and UF 1 kDa membranes.
A total decolourisation was observed for blue and red dyes using UF
1 kDa membrane.
Good conductivity rates (80%) and high COD and colour retention
rates (>90%) were obtained for both NF 200 and NF 270 membranes for all studied dyes solutions.
An improvement of the rinsing wastewater quality was obtained
using UF 1 kDa in terms of COD and colour.
NF process was chosen as an appropriate and an efcient process for
reducing conductivity, COD and colour in industrial efuent, to assure
constant good water quality demanded by textile nishing processes.

120

References

100
80
60
40
20
0
0

9 10 11 12 13 14 15

P (bar)

Fig. 15. Rinsing wastewater permeate ux versus transmembrane pressure using NF

270 membrane.

[1] V.M. Correia, T. Stephenson, S.J. Judd, Characterisation of textile watsewatersa

review, Environ. Technol. 15 (1994) 917929.
[2] G. Mishra, M. Tripathy, A critical review of the treatments for decolourization of
textile efuent, Colourage 40 (1993) 3538.
[3] I.M. Banat, P. Nigam, D. Singh, R. Marchant, Microbial decolorization of textiledyecontaining efuents: a review, Bioresour. Technol. 58 (1996) 217227.
[4] R.S. Juang, R.L. Tseng, F.C. Wu, S.J. Lin, Use of chitin and chitosan in lobster shell
wastes for colour removal from aqueous solutions, J. Environ. Sci. Health 31
(1996) 325338.
[5] P. Nigam, G. Armour, I.M. Banat, D. Singh, R. Marchant, Physical removal of textile
dyes and solid state fermentation of dye-adsorbed agricultural residues, Bioresour.
Technol. 72 (2000) 219226.
[6] V.J.P. Poots, J.J. Mckay, The removal of acid dye from efuent using natural adsorbents, Water Res. 10 (1976) 10611066.

96

A. Aouni et al. / Desalination 297 (2012) 8796

[7] G. Mckay, Waste colour removal from textile efuents, Am. Dyest Rep. 68 (1979)
2936.
[8] N. Willmott, J. Guthrie, G. Nelson, The biotechnology approach to colour removal
from textile efuent, JSDC 14 (1998) 3841.
[9] S.H. Lin, C.F. Peng, Treatment of textile wastewater by electrochemical method,
Water Res. 28 (1994) 277282.
[10] S.H. Lin, C.F. Peng, Continuous treatment of textile wastewater by combined coagulation, electrochemical oxidation and activated sludge, Water Res. 30 (1996)
587592.
[11] S.H. Lin, M.L. Chen, Treatment of textile wastewater by chemical methods for
reuse, Water Res. 31 (1997) 868876.
[12] H. Shu, C. Huang, Degradation of commercial azo dyes in water using ozonation
and UV enhanced ozonation process, Chemosphere 31 (1995) 38133825.
[13] H. Abdul Aziz, S. Alias, M.N. Adlan, A.H. Faridah, M. Asaari, Shahrir Zahari, Colour removal from landll leachate by coagulation and occulation processes, Bioresour.
Technol. 98 (2007) 218220.
[14] C. Tang, V. Chen, Nanoltration of textile wastewater for water reuse, Des 143
(2002) 1120.
[15] A. Aouni, C. Fersi, M. Ben Sik Ali, M. Dhahbi, Treatment of textile wastewater by a
hybrid electrocoagulation/nanoltration process, J. Hazard. Mater. 168 (2009)
868874.
[16] C. Fersi, L. Gzara, M. Dhahbi, Flux decline study for textile wastewater treatment
by membrane processes, Des 244 (2009) 321332.
[17] C. Fersi, M. Dhahbi, Treatment of textile plant efuent by ultraltration and/or
nanoltration for water reuse, Des 222 (2008) 263271.
[18] C. Fersi, L. Gzara, M. Dhahbi, Treatment of textile efuents by membrane technologies, Des 185 (2005) 18251835.
[19] A. Bes-Pi, A. Iborra-Clar, C. Garca-Figueruelo, S. Barredo-Damas, M.I. AlcainaMiranda, J.A. Mendoza-Roca, M.I. Iborra-Clar, Comparison of three NF membranes
for the reuse of secondary textile efuents, Des 241 (2009) 17.
[20] A. Bes-Pi, M.I. Iborra-Clar, A. Iborra-Clar, J.A. Mendoza-Roca, B. Cuartas-Uribe,
M.I. Alcaina-Miranda, Nanoltration of textile industry wastewater using a physicochemical process as a pre-treatment, Des 178 (2005) 343349.
[21] J.M. Gozlvez-Zafrilla, D. Sanz-Escribano, J. Lora-Garca, M.C. Len Hidalgo,
Nanoltration of secondary efuent for wastewater reuse in the textile industry,
Des 222 (2008) 272279.
[22] M. Brik, P. Schoeberl, B. Chamam, R. Braun, W. Fuchs, Advanced treatment of textile wastewater towards reuse using a membrane bioreactor, Process. Biochem.
41 (2006) 17511757.
[23] H. Yen, S. Kang, Mu-Hoe, Separation of textile efuents by polyamide membrane
for reuse, Polym. Test. 21 (2002) 539543.
[24] C. Suksaroj, M. Hran, C. Allgre, F. Persin, Treatment of textile plant efuent by
nanoltration and/or reverse osmosis for water reuse, Des 178 (2005) 333341.
[25] N. Zaghbani, A. Haane, M. Dhahbi, Removal of Eriochrome Blue Black R from
wastewater using micellar-enhanced ultraltration, J. Hazard. Mater. 168
(2009) 14171421.
[26] N. Zaghbani, A. Haane, M. Dhahbi, Removal of Safranin T from wastewater using
micellar enhanced ultraltration, Des 222 (2008) 348356.
[27] N. Zaghbani, A. Haane, M. Dhahbi, Separation of methylene blue from aqueous
solution by micellar enhanced ultraltration, Sep. Purif. Technol. 55 (2007)
117124.
[28] K. Majewska-Nowak, The effect of a polyelectrolyte on the efciency of dyesurfactant solution treatment by ultraltration, Des 221 (2008) 395404.
[29] J.J. Porter, R.S. Porter, Filtration studies of selected anionic dyes using asymmetric
titanium dioxide membranes on porous stainless-steel tubes, J. Membr. Sci. 101
(1995) 6781.
[30] A.L. Ahmad, S.W. Puasa, Reactive dyes decolourization from an aqueous solution
by combined coagulation/micellar-enhanced ultraltration process, Chem. Eng.
J. 132 (2007) 257265.
[31] Y. He, G. Li, H. Wang, Z. Jiang, J. Zhao, H. Su, Q. Huang, Experimental study on the
rejection of salt and dye with cellulose acetate nanoltration membrane, J. Taiwan
Inst. Chem. Eng. 40 (2009) 289295.
[32] I. Petrini, N.P.R. Andersen, S. ostar-Turk, A.M. Le Marechal, The removal of reactive
dye printing compounds using nanoltration, Dyes Pigm. 74 (2007) 512518.
[33] A. Akbari, S. Desclaux, J.C. Rouch, J.C. Remigy, Application of nanoltration hollow
bre membranes, developed by photografting, to treatment of anionic dye solutions, J. Membr. Sci. 297 (2007) 243252.
[34] B. Van der Bruggen, L. Braeken, C. Vandecasteele, Flux decline in nanoltration
due to adsorption of organic compounds, Sep. Purif. Technol. 29 (2002) 2331.

[35] I. Koyuncu, Inuence of dyes, salts and auxiliary chemicals on nanoltration of

reactive dye baths: experimental observations and model verication, Desalination
154 (1) (2003) 7988.
[36] A.D. Dhale, V.V. Mahajani, Studies in treatment of disperse dye waste:
membrane-wet oxidation process, Waste Manage. 20 (2000) 8592.
[37] K.M. Pastagia, S. Chakraborty, S. DasGupta, J.K. Basu, S. De, Prediction of permeate
ux and concentration of two-component dye mixture in batch nanoltration,
J. Membr. Sci. 218 (2003) 195210.
[38] A. Akbari, S. Desclaux, J.C. Rouch, P. Aptel, J.C. Remigy, New UV-photografted
nanoltration membranes for the treatment of colored textile dye efuents,
J. Membr. Sci. 286 (2006) 342350.
[39] I. Koyuncu, D. Topacik, M.R. Wiesner, Factors inuencing ux decline during
nanoltration of solutions containing dyes and salts, Water Res. 38 (2004) 432440.
[40] S.K. Nataraj, K.M. Hosamani, T.M. Aminabhavi, Nanoltration and reverse osmosis
thin lm composite membrane module for the removal of dye and salts from the
simulated mixtures, Desalination 249 (2009) 1217.
[41] C. Allegre, P. Moulin, M. Maisseu, F. Charbit, Savings and re-use of salts and water
present in dye house efuents, Desalination 162 (2004) 1322.
[42] A. Bdalo-Santoyo, J.L. Gmez-Carrasco, E. Gmez-Gmez, F. Mximo-Martn, A.M.
Hidalgo-Montesinos, Application of reverse osmosis to reduce pollutants present
in industrial wastewater, Desalination 155 (2003) 101108.
[43] I. Kim, K. Lee, Dyeing process wastewater treatment using fouling resistant
nanoltration and reverse osmosis membranes, Desalination 192 (2006) 246251.
[44] A. Aouni, C. Fersi, B. Cuartas-Uribe, A. Bes-pi, M. Isabel Alcaina-Miranda, M. Dhahbi,
Study of membrane fouling using synthetic model solutions in UF and NF processes,
Chem. Eng. J. 175 (2011) 192200.
[45] American Public Health Association, American WaterWorksAssociation, Water
Pollution Control Federation, in: S.A. Daz Santos (Ed.), Standard Methods of Examination of Water and Wastewater, APHA/AWWA/WPCF, Madrid, Espana,
1992.
[46] B. Cuartas-Uribe, M.C. Vincent-Vela, S. Alvarez-Blanco, M.I. Alcaina-Miranda,
E. Soriano-Costa, Nanoltration of sweet whey and prediction of lactose retention as
a function of permeate ux using the KedemSpiegler and Donnan Steric Partioning
models, Sep. Purif. Technol. 56 (2007) 3846.
[47] International Organization for Standardization, Particle Size AnalysisLaser
Diffraction MethodsPart 1: General Principles, ISO 133201:1999, 2009.
[48] B. Van der Bruggen, M. Manttari, M. Nystrom, Drawbacks of applying nanoltration
and how to avoid them: a review, Sep. Purif. Technol. 63 (2008) 251263.
[49] L. Braeken, B. Van der Bruggen, C. Vandecasteele, Flux decline in nanoriltration
due to adsorption of dissolved organic compounds: model prediction of time
dependency, J. Phys. Chem. B 110 (2006) 29572962.
[50] M. Cheryan, Ultraltration and Microltration Handbook, Technomic Publishing
Company, Inc., Lancaster-Basel, 1998.
[51] Yi-Li Lin, Pen-Chi Chiang, E.-E. Chang, Removal of small trihalomethane precursors from aqueous solution by nanoltration, J. Hazard. Mater. 146 (2007) 2029.
[52] J. Pieracci, J.V. Crivello, G. Belfort, Photochemical modication of 10 kDa polyethersulfone ultraltration membranes for reduction of biofouling, J. Membr.
Sci. 156 (1999) 223240.
[53] C. Bellona, J.E. Drewes, The role of membrane surface charge and solute physicochemical properties in the rejection of organic acids by NF, J. Membr. Sci. 249
(2005) 227234.
[54] D. Norberg, S. Hong, J. Taylor, Y. Zhao, Surface characterization and performance
evaluation of commercial fouling resistant low-pressure RO membranes, Desalination 202 (2007) 4552.
[55] L.D. Nghiem, P.J. Coleman, NF/RO ltration of the hydrophobic ionogenic compound triclosan: transport mechanisms and the inuence of membrane fouling,
Sep. Purif. Technol. 62 (2008) 711718.
[56] J. Tanninen, S. Platt, A. Weis, M. Nystrom, Long-term acid resistance and selectivity of NF membranes in very acidic conditions, J. Membr. Sci. 240 (2006) 1118.
[57] N. Park, B. Kwon, I.S. Kim, J. Cho, Biofouling potential of various NF membranes
with respect to bacteria and their soluble microbial products (SMP): characterizations, ux decline, and transport parameters, J. Membr. Sci. 258 (2005)
4354.
[58] M. Takht Ravanchi, T. Kaghazchi, A. Kargari, Application of membrane separation
processes in petrochemical industry: a review, Desalination 235 (2009) 199244.
[59] Amparo Bes-Pi, Beatriz Cuartas-Uribe, Jose-Antonio Mendoza-Roca, Maria Isabel
Alcaina-Miranda, Study of the behaviour of different NF membranes for the reclamation of a secondary textile efuent in rinsing processes, J. Hazard. Mater. 178 (2010)
341348.