Separation and Purication Technology 118 (2013) 179187

Contents lists available at SciVerse ScienceDirect

Separation and Purication Technology

journal homepage: www.elsevier.com/locate/seppur

Development of an asymmetric carbon microltration membrane:

Application to the treatment of industrial textile wastewater
Nouha Tahri a, Ilyes Jedidi a, Sophie Cerneaux b, Marc Cretin b, Raja Ben Amar a,
a
b

Laboratoire Sciences des Matriaux et Environnement, Facult des Sciences de Sfax, Route de Soukra Km 4, 3000 Sfax, Tunisia
Institute Europen des Membranes, UMR 5635 (CNRS, ENSCM, UM II), 1919 Route de Mende, 34293 Montpellier Cedex 5, France

a r t i c l e

i n f o

Article history:
Received 27 February 2013
Received in revised form 24 June 2013
Accepted 25 June 2013
Available online 4 July 2013
Keywords:
Carbon/carbon microltration membrane
Phenolic resin
Carbon powder
Carbonization

a b s t r a c t
Tubular carbon microltration membranes have been prepared using mineral coal powder (100 lm average particle size) mixed with phenolic resin solution and organic additives. The porosity of the support
can be sharply controlled by xing the organic additives percentages, the particles size and the homogeneity of the carbon powder. A mesoporous active layer was deposited on the inner face of the carbon support by slip casting using a deocculated slip made of mineral coal powder (1.76 lm) suspended in a
phenolic resin solution. An active layer with an average pore diameter of 0.60 lm is obtained after carbonization under nitrogen atmosphere. The obtained membrane has very interesting characteristics
regarding mechanical and chemical resistances. The application to the treatment of industrial textile
wastewater shows good performances in term of permeate ux and efciency (retention of COD and
salinity of 50% and 30% respectively and almost a total retention of turbidity and color).
2013 Elsevier B.V. All rights reserved.

1. Introduction
Many efforts to achieve economical and efcient membranes
for various uses have resulted in a selection of new materials, an
improvement in membrane preparation techniques and an increase in the number of applications [1,2]. At present, the interest
is derived toward inorganic membranes due to their superior permeabilityselectivity combination and suitable performance for
high temperature or corrosive environment compared to polymeric membranes [3].
Different studies on carbon membranes have shown that they
can successfully compete with polymeric membranes and other
porous inorganic membranes, such as silica- and zeolite-based
membranes, in separation processes of industrial interest. Carbon
material possess a very good chemical and mechanical resistance
whatever the temperature used.
Carbon membrane is generally prepared by carbonization of
various carbonaceous materials [2] such as polyimide [46],
poly-acrylonitrile [7], phenolic resins [810] and polyfurfuryl
alcohol [1113]. Recent studies [14] showed that coal is a good
carbon material that can be used as source of carbon to prepare

Abbreviations: COD, chemical oxygen demand; MF, microltration; R, retention;

TMP, transmembrane pressure.
Corresponding author. Address: Facult des Sciences de Sfax, Laboratoire
Sciences des Matriaux et Environnement, Route de Soukra Km 4 BP 1171, Sfax,
Tunisia. Tel.: +216 74 276 400, +216 74 276 763; fax: +216 74 274 437.
E-mail address: [email protected] (R. Ben Amar).
1383-5866/$ - see front matter 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.seppur.2013.06.042

asymmetric microltration carbon membranes with high porosity,

controllable pore size and narrow pore size distribution.
Due to their low cost and high carbon yield, phenolic resins are
widely used as precursors for the preparation of carbon membranes with molecular sieving properties [15]. Phenolformaldehyde resins are thermosetting polymers. They do not ow at high
temperatures which maintain the initial shape unchanged during
the heat treatment. In addition, when pyrolyzed under inert atmosphere, phenolic resins have a high carbon yield which is crucial for
the mechanical properties of the nal carbon material. Their use is
made easier by the large variety of presentation (in solution or
powdered) and of types (Novolac and Resole thermosetting respectively in the presence or absence of a cross-linking agent).
Nowadays, a lot of researchers have demonstrated that extrusion is proven to be the best technique to use for preparing tubular
carbon membranes. During the last decade, tubular carbon supports have been produced by extrusion using a novolac type phenolic resin [1015]. The resin was semi-cured then ground in ne
powder which was mixed (or not) with different additives and
sometimes another source of carbon is added to better control
the nal porosity characteristics of the nal material [10]. A carbonization step under nitrogen atmosphere takes care about the
pores opening and the porosity is then created.
Much works were done concerning the preparation of carbon
molecular sieves for gas separation by carbonizing under inert
atmosphere a phenolic resin layer deposited by slip casting
process.

180

N. Tahri et al. / Separation and Purication Technology 118 (2013) 179187

Centeno et al. and Fang et al. [16,9] prepared carbon membrane

using thermosetting phenolic resin as polymeric precursor. The
membrane was formed by a thin microporous carbon layer obtained by pyrolysis of a phenolic resin lm supported on the inner
face of a porous alumina tube. The coatingcarbonization cycle was
repeated several times to obtain carbon membranes, which have
good separation properties for H2/N2, H2/CH4 and O2/N2. Kishore
et al. [17] used phenolformaldehyde resin to prepare carbon
membrane for nanoltration.
Little work was done to deposit carbon active layers on a carbon macroporous support. Katsaros et al. [10] prepared a carbon
membrane for gas separation purpose by coating a phenolic resin
solution on the outer face of a carbon macroporous support. To
do so, a paste made of partially cured phenolic resin particles
mixed to water and porosity agents, was extruded and then
carbonized under inert atmosphere at 800 C. Song et al. [14]
prepared coal-based symmetric microltration carbon membrane
for the treatment of oilwater mixtures. The result showed that
the oil rejection coefcients of oily wastewater were up to 97%,
and the oil concentrations of the permeate were less than
10 mg/L.
Based on these considerations and the body of previous
research, this work describes the preparation of new tubular carbon/carbon microltration membranes by slip-casting method
using mineral coal powder and thermosetting resin. The effects
of particle size, casting time, and concentration of organic additives
are studied. The performances of the application of this membrane
on the treatment of industrial textile wastewater are then
discussed.

2. Experimental
2.1. Material
2.1.1. Carbon material
Mineral Coal powder was used as the main carbon material
source. To determine the effect of particle size, carbon powders
with different average particle size were prepared. A 200 lm average particles size powder was obtained by crushingsieving of a
raw mineral coal powder. This late powder was ground in a planetary crusher at 300 rpm respectively during 30 min, 2 h and
15 h to obtain three different powders having average particle size
respectively equal to 100 lm, 20 lm and 1.76 lm.
The particle sizing system CILAS 920 was used to determine the
particle size distribution. Thermogravimetric analysis (TGA) and
differential scanning calorimetry (DSC) were performed with
simultaneous DSCTGA 2960 TA instrument to estimate the materials weight loss in a temperature interval varying from 20 to
1000 C with a heating rate of 10 C/min, under nitrogen ow of
1 ml/min.
2.1.2. Thermosetting resin
Phenolic resin marketed by the company Irons Resins S.A, Spain
(Sumitomo Bakelite Co.) was used as carbon precursor, binder and
porosity agent. It is sold in powder form soluble in ethanol containing 7% of hexamine as curing agent.
2.1.3. Organic additives
Organic additives were dry mixed to the carbon powder. Some
of them, as starch powder, were used to enhance the total porosity
of the nal material (since it sublimate during carbonization),
other additives, such as ethylene glycol, amijel and methyl-cellulose are used respectively as lubricants and plasticizers of the paste
to be extruded in tubular form.

To study the inuence of starch percentage on the nal material

structure, carbon supports differing only by starch content, were
prepared (Table 1).
2.2. Preparation of carbon membrane
Membrane synthesis consisted on two main steps: preparation
of macroporous support and active microltration layer deposition. The process of microltration membrane preparation is
described in Fig. 1.
Tubular support (OD/ID = 10 mm/8 mm) were obtained by
paste extrusion under a pressure of 3040 bars. The paste was prepared following the composition described on previous work
[18,19].
The extruded green tubular bodies were heat treated in a tubular electrical furnace (VTF7 VECSTAR) connected to inert gas purging during the carbonization and following the temperaturetime
schedule given in Fig. 2.
The active microltration layer was prepared by slip-casting
process which process consists essentially on four steps (Fig. 3):
 Preparation of the suspension: the elaborated suspension was
made of 20% of carbon powder (1.76 lm average particle size)
suspended in alcoholic solution of phenolic resin (Resin/Ethanol = 15/85 wt%).
 Homogenization of the suspension by magnetic stirring followed by ultrasound exposure.
 Slip cast the suspension on the inner face of the porous carbon
support for few minutes at room temperature. In the case of the
tubular membranes, the tube was closed at one end and lled
with the solution.
 The deposited layer was treated following the same temperature program used to prepare the macroporous carbon
supports.
2.3. Material physicochemical characterization
The porosity and the average pore diameter of the nal material
were determined using Hg-porosimetry on a Micrometrics Autopore II9220 V3.05. The presence of possible defects in the membrane was checked using scanning electron microscopy (SEM).
Mechanical resistance of the material was measured by the three
point bending method (using LLOYD Instrument) on test bars having the following dimensions: 32 mm/10 mm/2 mm. The distance
separating the two base points was 28 mm. The Shrinkage degree
was evaluated by comparing the dimensions of 10 tubes (using a
caliper rule) before and after carbonization step. Contact angle
technique was used to evaluate the hydrophobic character of the
elaborated membrane material.

Table 1
Materials used to prepare the paste: quantity and function.
Component

Function

Quantity

Powder carbon
Ethylene glycol
Starch (RG 03408, Cerestar)a
Amijel (Cplus 12076,
Cerestar)
Methocel (The Dow Chemical
Company)
Alcoholic solution of phenolic
resin

Mineral charge
Lubricant
Porosity agent
Plasticizer,
binder
Plasticizer

150 g
25.5 g
0%, 4%, 8%, 10%, 12%, 15%
5.75 g

Binder, porosity
agent

120130 ml

5.75 g

a
The different starch quantities are calculated based on the total weight of the
dry components only.

N. Tahri et al. / Separation and Purication Technology 118 (2013) 179187

181

Fig. 1. Main steps of the processing route for microltration membrane preparation.

60 C) and basic (NaOH 2% at 80 C) solutions were alternatively

circulated for 20 min. The membrane was in between rinsed with
deionized water until neutral pH. The efciency of the used cleaning protocol was veried by measuring water ux after the cleaning cycle. The membrane had been applied to treat wastewater
coming from Tunisian textile industry.
2.5. Efuent characterization
Physico-chemical parameters of real efuent and of permeate
were determined according to the standard methods suggested

Fig. 2. Temperaturetime schedule.

2.4. Microltration treatment

Single channel tubular carbon membrane was employed with 8/
10 mm in inner/outer diameter and 150 mm in length. The crossow mode of microltration was adopted in this experiment. Schematic diagram of the experimental apparatus is shown in Fig. 4.
Microltration (MF) operations were carried out at a temperature
of 25 C, under the transmembrane pressure (TMP) between 1
and 5 bars. The transmembrane pressure was controlled by an
adjustable valve at the concentrate side. Before the tests, the membrane had been conditioned by immersion in pure deionized water
for at least 24 h. The water ux through the membrane was measured as a function of time at different TMP values.
The regeneration of the membrane was carried out by rstly,
back-ushing procedures for 15 min, then acidic (nitric acid 2% at

Fig. 3. Scheme of slip-casting process.

182

N. Tahri et al. / Separation and Purication Technology 118 (2013) 179187

Fig. 4. Scheme of the pilot plant.

Fig. 5. DSCTGA data under air of the mineral coal powder.

by American Public Health Association (Standard Methods for the

Examination of Water and Wastewater 1998). Conductivity and
pH determinations were done by means of a conductimeter, Tacussel model 123 and pH-meter, Metrohm 744 respectively. Turbidity
was measured using a turbidity meter (Hach RATIO 2100A) in
accordance with standard method 2130B; a calibration standard
is needed.
Color measurement was determined with standard dilution
multiple method [20] and by comparing absorbance to a calibration curve [21]. Dye decolorization was determined by monitoring
the decrease in the absorbance peak at the maximum wavelength
for the global efuent. For the wastewater used in this study only
one pick was observed at 420 nm. UVVisible spectrophotometer
(Perkin Elmer Lambda 20 UV/VIS Spectrophotometer) was used
in all experiments.
COD was estimated by open reux method. The protocol present a method derived from the Standard AFNOR T90-101. The sample was reuxed in an acidic medium with a known excess of
potassium dichromate and the remaining dichromate was titrated
with ferrous ammonium sulphate. The COD values were obtained
using a Fisher Bioblock Scientic reactor COD 10119 type COD
meter.
3. Result and discussion
3.1. Thermogravimetric analyses of raw materials
3.1.1. Carbon material
The DSCTGA (from 0 to 1000 C) under air of a coal powder
sample, shows a weight loss of 8.23% at 730 C which corresponds
to y ash (Fig. 5).
This result shows that y ash represents a minor phase in the
coal rock. Knowing that the y ash is in the molecular state and
scattered through the mineral coal rock [22,23], thus, it can be considered that the membrane is made totally of coal carbon.

Fig. 6. DSCTGA data under nitrogen atmosphere of paste prepared with Novolac
Resin and carbon powder.

The pore structure is formed after all volatile components present in the green body (the additives part and the noncarbonaceous
part of the resin) are sublimated.

3.1.3. Starch
Starch is a complex Glucide composed of molecular chains of DGlucose having a brute formula as: (C6H10O5)n.
The result of DSCTGA analyzes of a sample of starch (Fig. 7)
shows that the mass loss took place in the temperature range of
250550 C and reached the maximum of 61%,37% at 300 C.
It can be clearly seen that when the nal temperature of 750 C
was reached, all starch quantity was sublimated and was then
responsible in part of the material porosity.

3.2. Optimization of carbon support composition

3.1.2. Thermosetting resin
The weight loss during the carbonization of a paste prepared
from resin and carbon powder without any additives was evaluated by thermogravimetric analysis. Fig. 6 shows a maximum
weight loss of about 25% reached at 700 C. Since the carbon content in the resin represents 25% of its total molecular weight, it can
be concluded that coal carbon material does not participate in the
total porosity of the nal material. Porosity results principally from
resin weight loss. Therefore, besides its role as binder and source of
carbon, resin acts also as porosity agent.
In addition, the sublimation of the other organic additives
(especially starch) contributes also to the nal porosity of the
coal-based carbon membrane.

As known, powder particle size and organic porous agent percentages (starch in our case), in addition to sintering temperature,
are the most important parameters affecting the structure and the
porosity of porous materials [19]. In the case of carbon membranes
prepared following the process described in this paper, the consolidation of the material is due to the phenolic resin crosslinking. In
addition, resin participate with a constant value to the nal porosity, but the main parameters to control both the pore size and the
porous volume of the nal material are respectively coal particle
size and starch percentage the powder particle size and the presence of starch should also contribute to the structure and porosity
of the membrane.

N. Tahri et al. / Separation and Purication Technology 118 (2013) 179187

183

Fig. 7. The DSCTGA data under nitrogen atmosphere of the starch.

3.2.1. Effect of carbon particle size

Different tubular supports were prepared using four carbon
powders with different average particle sizes: 200 lm, 100 lm,
20 lm and 1.76 lm. Fig. 8 shows that total porous volume remains
unchanged at a value of 38% for all prepared supports. However,
average pore size increases when particle size increases. Average
pore diameters were equal to 0,066 lm, 0.1 lm, 8 lm and 12 lm
respectively for powder particle size of 1.76 lm, 20 lm, 100 lm
and 200 lm. Similar results were obtained by Song et al. [14]
whose found that bigger is the coal particles, higher is the average
pore sizes.
3.2.2. Effect of starch percentage
In order to determine the effect of starch percentage on the
physical characteristics of the nal material, carbon supports were
prepared using various starch percentages: 0%, 2%, 4%, 8%, 10%, 12%
and 15%.
Fig. 9 shows that all samples have very good mechanical
strength since they resisted to a mean maximum force before
break varying from 29N for 0% starch content to 18N for 15% starch
content. The materials mechanical resistance decreased with the
increase of starch content due to the porous volume increase.
On the other hand, the porous volume increased with starch
content from 25% (0% starch) to 38% (12% starch). Beyond 12%
starch content, the porous volume remained constant. 25% of
porosity obtained at 0% is due to resin weight loss. The increase

Fig. 9. Evolution of porous volume and mechanical resistance with the starch
content.

of the porosity until 12% added starch (limit value) can be attributed to the starch sublimation.
SEM pictures (Fig. 10) shows a drastic change in the nal material texture with the added percentage of starch. 12% starch content show the greatest porosity.
As we can see, the porous texture of the obtained tubular support is not the result of a sintering phenomenon as observed during
the formation of conventional ceramic porous materials [19,24,25].
In fact, the elaborated paste is no more than a composite polymer
in which the resin represents the polymer matrix and the carbon
particles are the mineral charge. During the carbonization step organic additives sublimate and the resin lose weight, thus, only the
carbon skeleton of the resin and the carbon coal particles remained
(Fig. 11).
12% of starch content was found to be the most appropriate to
obtain a material with a good mechanical strength, porosity and
texture.
3.3. Active layer characterization

Fig. 8. Variation of pore size distribution with particle size of carbon powder.

A microltration active layer was deposited on the optimized

support by slip casting. The same suspension (described earlier)
was used to cast the layer. Nevertheless, different casting times
(2, 4, 6 and 8 min) were tested. Fig. 12 presents the evolution pore

184

N. Tahri et al. / Separation and Purication Technology 118 (2013) 179187

Fig. 10. SEM images of two samples of carbon supports using 0% starch content and 12% starch content.

Fig. 11. SEM pictures of the nal carbon support.

size distribution of the carbon membranes with casting time. It can

be observed that the average pore diameter of the active layer
(0.60 lm) remained unchanged for all casting times.
Surface and cross-section morphologies for the different samples obtained with the different casting times were characterized
by SEM micrographs. All samples showed a typical asymmetric
structure but the thicknesses of the active layer increased with

casting time respectively from 6 lm for 2 min, to 30 lm for

8 min casting time (Fig. 13).
When analyzing the surface texture of the different samples, it
can be concluded that the sample obtained with 6 min casting time
do not contain defects (cracks and pinholes) at the opposite of all
other samples.
So the best layer is obtained with 6 min casting time with thickness around 22 lm which is a good thickness for a microltration
layer [18,19,24,26,27].
3.4. Determination of microltration membrane permeability

Fig. 12. Mercury porosimetry analysis of the microltration membrane for the
different casting time.

The permeability of the support and MF membrane was determined using pure distilled water. Flux values of distilled water at
different operating pressures were measured and were plotted
against transmembrane pressure. Samples were conditioned in
the pure water during the 24 h which precede the ltration, this allows obtaining a fast stabilization of the permeate ux. Permeability was obtained according to the Darcys law (J = DP/l.TMP)
(Fig. 14).
The average values for the elaborated support and MF membrane permeability were respectively 280 L/h m2 bar and 100 L/
h m2 bar.
Contact angle technique is used to estimate the hydrophobicity
character of the material. The measure was repeated 4 times using
a water drop volume varying between 10 and 15 lL. The measures

185

N. Tahri et al. / Separation and Purication Technology 118 (2013) 179187

Fig. 13. SEM micrographs for carbon microltration membrane obtained with different casting times. Surface texture is shown in the inserts.

1000

cal substances such as dyes, detergents, salts, auxiliaries (e.g. surfactants, emulsiers) and caustic soda are used. Generally, the
wastewater was coming from the dyeing, washing and bleaching
baths (Table 2).

900
800

600
500
400
300
200
100

MF

Support

0
0

0.5

1.5

2.5

3.5

4.5

TMP (Bar)
Fig. 14. Determination of the support and MF membrane permeability.

of the angle were taken after 12 s. Macroporous supports contact

angle value was equal to 80 showing a slight hydrophobic character while it was found 96.1 for the microltration layer corresponding to a stronger hydrophobic character. This difference in
wettability, even if both support and layer had a similar chemical
composition, can be related to the difference in pore size [28,29].
When using membranes, a cleaning cycle is needed for regeneration according to a protocol using an alternating rinsing of basic
and acid solutions. Tests of corrosion resistance were performed
on carbon supports by immersion in basic and acid solutions. For
this, membrane sample was soaked at 80 C for 30 days in bath
solution at 2 wt% nitric acid (pH not exceeding 2) and NaOH
(pH = 14). Every 24 h, a sample was removed, washed, dried and
weighed to estimate the weight loss. It was concluded that no
weight loss was recorded throughout the testing period. Therefore,
it appears that the elaborated carbon membrane has a very high
chemical resistance.

3.5.2. Microltration treatment

The ltration performances were determined at a TMP of 3 bar,
a velocity of 5.6 m/s and a temperature of 25 C. The evolution of
the ux with time is given by Fig. 15. The ux decreased during
the rst 10 min from 42 L/h m2 to 35 L/h m2 and then stabilized
at 32 L/h m2. The decline of permeate ux with time is a typical
behavior of membrane processes, and can be interpreted mostly
in terms of concentration polarization and fouling due to the interaction between membrane material and solution [30].

Table 2
Principle physico-chemical characteristics of textile wastewater.

Parameters

pH

Salinity (g/L)

COD (mg/L)

Turbidity (NTU)

Colorb

Values

8.5

22.8

960

210

52

Integral of the absorbance curve in the whole visible range (400800 nm).

50
45
40
35

J (L/h.m)

J (L/h.m)

700

30
25
20
15
10

3.5. Application to textile wastewater treatment

5
0

3.5.1. Wastewater characterization

The study was conducted with a textile wastewater sample
supplied from a Tunisian textile factory in which, different chemi-

10

20

30

40

50

Time (min)
Fig. 15. Evolution of the permeate ux with time.

60

70

186

N. Tahri et al. / Separation and Purication Technology 118 (2013) 179187

100%

R (%)

80%
60%
40%
20%

Carbon
Membrane

0%

Salinity

Alumina
Membrane

COD
Turbidity
Color

Fig. 16. Evolution of retention of the different parameters by carbon and alumina
microltration membrane.

carbon precursor using a new process. The porosity and the texture
of the nal material were controlled by adjusting coal particle size
and percentage of organic additives. Tubular support (OD/
ID = 10 mm/8 mm) elaborated by paste extrusion had a porous volume of 38% and a good mechanical and chemical resistances.
The active microltration layer was prepared by slip-casting
process. The casting time was optimized to 6 min which produced
a layer with thickness around 22 lm.
Performances of this membrane toward the treatment of industrial textile wastewater have been studied in term of permeate ux
and efciency. The retention of polluting substances regarding
COD, Turbidity, salinity and color was respectively 57%, 90%, 30%
and 80%.
By comparison with alumina membranes of similar porous
characteristics, carbon microltration membrane showed the best
performances.

Acknowledgements
In order to explain permeate ux decline, the resistance series
model can be used:

RT Rm Rrev Rirrev
RT is the total ltration resistance which represents the distribution of the different resistances. Rm is the inherent hydraulic resistance of clean membrane caused by the membrane itself, it was
given by the determination of pure water permeability. The Rrev
resistance is due to concentration polarization and solids (cake
layer) on the membrane surface, which can be removed by rinsing
with water after the ltration experiment. At the contrary, the Rirrev
resistance is due to pore blocking and absorption of materials onto
the membrane surface and pores, which cannot be removed by a
simple water rinsing but needs a chemical cleaning. After each
run, the membrane was cleaned with pure water and then the permeate water ux was determined, given the Rirrev. The Rrev value
was calculated following:

Rrev RT  Rm Rirrev
In our case, the different resistance values are: RT = 3.37  1011 m1, Rirrev = 2  1011 m1 et Rrev = 1.29  1011 m1. The fouling is
rather irreversible (Rrev > Rirrev).
Regarding the permeate quality; a total retention of color and
turbidity was achieved. By an other hand the retention of polluting
substances was higher than 50% except for salinity which was of
only 30%. Indeed, the main mechanism of microltration is usually
sieving but adsorption of organic and mineral pollution towards
membrane surface can also occurred due to ionic and hydrophobic
interaction. So, at high concentration of salt, surface charge effect
can occurred which explains the retention of 30% of salt with initial
concentration of 22 g/L. The difference of charge between the
membrane surface (negative since pH is of 6.9) and the solution
having a pH of 8.5 contributed to this retention. By comparison
with alumina microltration membrane, carbon membrane shows
similar retention of turbidity and COD and higher removal of color
(Fig. 16).
Removal of color is due mainly to the adsorption of dyes onto
the membrane material. Retention of salinity by carbon membrane
is almost twice higher than that obtained by alumina membrane.
Therefore, the carbon microltration membrane seems to be suitable to textile efuent treatment.
4. Conclusion
New carbon/carbon asymmetric microltration membranes
have been prepared based on mineral coal and phenolic resin as

Authors would like to thank the Ministry of higher education

and scientic research, Tunisia for funding this work. Authors are
also very grateful to Sumitomo Bakelite Europe (Barcelona) for
their help by given samples of resin.

Appendix A. Supplementary material

Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.seppur.2013.06.042.

References
[1] W.J. Koros, R. Mahajan, Pushing the limits on possibilities for large scale gas
separation: which strategies, J. Membr. Sci. 175 (2000) 181191.
[2] A.F. Ismail, L.B. David, A review on the latest development of carbon
membranes for gas separation, J. Membr. Sci. 193 (2001) 118.
[3] H.P. Hsieh, Inorganic Membranes for Separation and Reaction, Elsevier, New
York, 1996.
[4] A.B. Fuertes, T.A. Centeno, Preparation of supported asymmetric carbon
molecular sieve membranes, J. Membr. Sci. 144 (1998) 105111.
[5] M. Yamamoto, K. Kusakabe, J. Hayashi, S. Morooka, Carbon molecular sieve
membrane formed by oxidative carbonization of a copolyimide lm coated on
a porous support tube, J. Membr. Sci. 133 (1997) 195205.
[6] H. Hatori, T. Kobayashi, Y. Hanzawa, Y. Yamada, T. Iimura, M. Kimura, Nanospace structure of carbon from polyimide containing nickel nitrate and their
function on catalytic reaction, Synth. Metal. 125 (2002) 183188.
[7] L.B. David, A.F. Ismail, Inuence of the thermastabilization process and soak
time during pyrolysis process on the polyacrylonitrile carbon membrane for
O2/N2 separation, J. Membr. Sci. 213 (2003) 285291.
[8] A.B. Fuertes, D.M. Nevskaia, T.A. Centeno, Carbon composite membranes from
Matrimid and Kapton polyimides for gas separation, Micropor. Mesopor.
Mater. 33 (1999) 115125.
[9] J. Fang, G. Qin, W. Wei, X. Zhao, Preparation and characterization of tubular
supported ceramic microltration membranes from y ash, Sep. Purif. Technol.
80 (2011) 585591.
[10] F.K. Katsaros, T.A. Steriotis, A.K. Stubos, A. Mitropoulos, N.K. Kanellopoulos, S.
Tennison, High pressure gaz permeability of microporous carbon membrane,
Micropor. Mater. 8 (1997) 171.
[11] M.B. Shiett, H.C. Foley, On the preparation of supported nanoporous carbon
membranes, J. Memb. Sci. 285 (2000) 275282.
[12] H. Wang, L. Zhang, G. Gavalas, Preparation of supported carbon membranes
from furfuryl alcohol by vapor deposition polymerization, J. Memb. Sci. 177
(2000) 2531.
[13] Y.D. Chen, R.T. Yang, Preparation of carbon molecular sieve membrane and
diffusion of binary mixtures in the membrane, Ind. Eng. Chem. Res. 33 (1994)
3146.
[14] C. Song, T. Wang, Y. Pan, T. Qiu, Preparation of coal-based microltration
carbon membrane and application in oily wastewater treatment, Sep. Purif.
Technol. 51 (2006) 8084.
[15] X. Zhang, H. Hu, Y. Zhu, S. Zhu, Effect of carbon molecular sieve on phenol
formaldehyde novolac resin based carbon membranes, Sep. Purif. Technol. 52
(2006) 261265.
[16] T.A. Centeno, J.L. Vilas, A.B. Fuertes, Effects of phenolic resin pyrolysis
conditions on carbon membrane performance for gas separation, J. Membr.
Sci. 228 (2004) 4554.

N. Tahri et al. / Separation and Purication Technology 118 (2013) 179187

[17] N. Kishore, S. Sachan, K.N. Rai, A. Kumar, Synthesis and characterization of a
nanoltration carbon membrane derived from phenolformaldehyde resin,
Carbon 41 (2003) 29612972.
[18] I. Jedidi, S. Khemakhem, A. Larbot, R. Ben Amar, Elaboration and
characterization of y ash based mineral supports for microltration and
ultraltration membranes, Ceram. Int. 35 (2009) 27472753.
[19] I. Jedidi, Elaboration de membranes inorganiques de ltration: application la
purication de solutions chimiques, PhD thesis, University of Sfax, Tunisia,
2005.
[20] I. Khouni, B. Marrot, R. Ben Amar, Decolourization of the reconstituted dye
bath efuent by commercial laccase treatment: optimization through
response surface methodology, Chem. Eng. J. 15 (2010) 121133.
[21] J. Weili, Y. Zengwei, B. Jun, S. Li, Conserving water by optimizing production
schedules in the dyeing industry, J. Clean. Prod. 18 (2010) 16961702.
[22] M. Ilic, C. Cheeseman, C. Sollars, J. Knight, Mineralogy and microstructure of
sintered lignite coal y ash, Fuel 82 (2003) 331336.
[23] D.A. Spears, Role of clay minerals in UK coal combustion, Appl. Clay. Sci. 16
(2000) 8795.

187

[24] S. Masmoudi, A. Larbot, H. El Feki, R. Ben Amar, Elaboration and properties of

new ceramic microltration membranes from natural and synthesized apatite,
Desalination 190 (2006) 89103.
[25] A. Larbot, S. Alami-Younssi, M. Persin, J. Sarrazin, L. Cot, Preparation of a calumina nanoltration membrane, J. Membr. Sci. 97 (1994) 167173.
[26] S. Khemakhem, A. Larbot, R. Ben Amar, New ceramic microltration
membranes from Tunisian natural materials: application for the cuttlesh
efuents treatment original, Ceram. Int. 35 (2009) 5561.
[27] A.J. Burggraaf, L. Cot, Fundamentals of inorganic membrane science and
technology, Membrane Science and Technology, vol. 4, Elsevier, 1996.
[28] K. Lum, D. Chandler, J.D. Weeks, Hydrophobicity at small and large length
scales, J. Phys. Chem. B. 103 (1999) 45704577.
[29] D.J. Ridder, J.Q. Verberk, S.G.J. Heijman, G.L. Amya, J.C. VanDijk, Zeolites for
nitrosamine and pharmaceutical removal from demineralised and surface
water: mechanisms and efcacy, Sep. Purif. Technol. 89 (2012) 7177.
[30] V. Chen, A.G. Fane, S. Maedani, I.G. Wenton, Particle deposition during
membrane ltration of colloids: transition between concentration polarization
and cake formation, J. Membr. Sci. 125 (1997) 109122.