Journal of Membrane Science 490 (2015) 92–102

Contents lists available at ScienceDirect

Journal of Membrane Science

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

Elaboration of novel tubular ceramic membrane from inexpensive raw

materials by extrusion method and its performance in microfiltration
of synthetic oily wastewater treatment
R. Vinoth Kumar, Aloke Kumar Ghoshal, G. Pugazhenthi n
Department of Chemical Engineering, Indian Institute of Technology Guwahati, Guwahati 781039, Assam, India

art ic l e i nf o a b s t r a c t

Article history: A tubular ceramic microfiltration membrane was prepared by an extrusion technique using inexpensive
Received 18 March 2015 clay mixtures namely, ball clay, kaolin, feldspar, quartz, pyrophyllite and calcium carbonate. The mixture
Received in revised form of clay powders extruded to form a porous tubular membrane without the addition of any organic
25 April 2015
additives. The dimensions, such as outer and inner diameters, wall thickness and length of the tube are
Accepted 27 April 2015
Available online 6 May 2015
11.5, 5.5, 3 and 100 mm, respectively. The sintered membrane possesses the porosity of 53%, water
permeability of 5.93
10  7 m/s kPa, an average pore size of 0.309 μm and mechanical strength of
Keywords: 12 MPa with very good corrosion resistance in acidic and basic conditions. The fabricated membrane is
Tubular membrane expected to have potential applications in the pretreatment and also can be used as support for
Microfiltration
ultrafiltration membranes. With this intention, the membrane is subjected to microfiltration of synthetic
Oily wastewater
oily wastewater emulsion experiments at various combinations of applied pressures (69–345 kPa), feed
Pore blocking model
Fouling mechanism concentrations (50–200 ppm) and cross flow rates (5.55
10  7–1.66
10  6 m3/s). An increase in the
applied pressure and flow rate of oily wastewater emulsion result a decreased oil rejection while, an
increase in the oil concentration results in enhanced rejection. The applied pressure of 69 kPa offers the
highest rejection of oily wastewater (99.98%) with permeate flux of 3.16
10  5 m/s. Additionally, the
membrane fouling mechanisms are investigated using diverse pore blocking models (complete,
standard, intermediate pore blocking and cake filtration model) with obtained experimental data. It is
found that the experimental results are well described by the cake filtration model. Finally, the rejection
potential of the membrane is compared with other membranes reported in the literature.
& 2015 Elsevier B.V. All rights reserved.

1. Introduction tasks for the upcoming development of ceramic membranes for

treatment of a large volume of waste matters contaminated liquids.
In recent years, the potential utilization of ceramic membranes Tubular ceramic membranes are especially suitable in applications
has greatly improved, especially in a broad range of industrial where the feed stream contains a relatively high proportion of large
processes and pollution treatment technologies. Ceramic mem- particles, and where the membranes are exposed to extreme pH
branes can be applied in extreme aggressive environments due to and temperature conditions. In general, the size of the feed flow
their distinct advantages, including good thermal stability, mechan- channels in tubular ceramic membranes is larger than other
ical strength, chemical resistance, long lifetime and defouling membrane modules. This type of open channel geometry reduces
properties. While organic polymer membranes cannot be applied the risk of the blockage of the feed channels and also the require-
owing to their restricted stability [1–3]. In general, most of the ment of costly pretreatment before microfiltration [6]. Generally,
commercialized ceramic membranes available in the market are ceramic membranes comprise of multi-layer formation with the
produced from alumina, silica, zirconia and titania materials. How- substrate (support), intermediate layer and skin layer (separating
ever, these ceramic membranes are severely restricted in large scale layer). The substrate of membranes can be produced by various
application due to the higher cost of starting materials and sintering techniques, including isostatic pressing, extrusion, slip casting, etc.,
processes [4–5]. Therefore, the preparation of ceramic membranes [7,8]. The fabrication of porous ceramic membrane with the tubular
with lower cost and excellent characteristics are the challenging configuration needs the expertise on choosing appropriate starting
materials and methods. Shaping by extrusion is achievable only if
the paste created from starting materials has rheological character-
n
Corresponding author. Tel.: þ 91 361 2582264; fax: þ 91 361 2582291. istics similar to those of clays [9]. As a consequence, currently many
E-mail address: [email protected] (G. Pugazhenthi). researches targeted on the development of ceramic membranes

http://dx.doi.org/10.1016/j.memsci.2015.04.066
0376-7388/& 2015 Elsevier B.V. All rights reserved.
 R. Vinoth Kumar et al. / Journal of Membrane Science 490 (2015) 92–102 93

with utilization of less expensive materials such as apatite powder, on various operating parameters, such as applied pressure, feed
dolomite, pyrophylite, Moroccan clay and kaolin, etc,. [10–13]. Talidi concentration and cross flow rate with respect to rejection and
et al. [10] prepared the tubular macroporous membrane using permeate flux. The fouling mechanisms are examined using
pyrophylite clay via extrusion and followed by a sintering process. obtained flux data during the microfiltration.
The properties of the porous pyrophyllite membrane were dis-
cussed as a function of sintering temperature in order to optimize
the preparation conditions. Saffaj et al. [11] prepared membrane 2. Materials and methods
support using Moroccan clay as a raw material by extrusion of the
clay paste. The structural and mechanical properties of fabricated The starting materials utilized for elaboration of the membrane
low cost membrane support were found suitable for membrane (kaolin, quartz, ball clay, pyrophyllite, and feldspar) were of
applications. Masmoudi et al. [12] produced tubular porous support mineral grade and obtained in the vicinity (Kanpur). Crude oil
using natural apatite powder and suggested for ultrafiltration (UF) was procured from Guwahati Refinery, IOCL (Assam). Calcium
and microfiltration (MF) applications. Bouzerara et al. [13] con- carbonate, hydrochloric acid and sodium hydroxide were supplied
structed membrane support from mixtures of doloma and kaolin. by Merck (I) Ltd, Mumbai.
They strongly recommended that prepared support can be utilized
for MF and UF processes. 2.1. Elaboration of tubular ceramic membrane
A huge volume of oily wastewater was created from different
process industries, including petrochemical, petroleum refineries, Tubular ceramic membrane was fabricated with length, inner
transportation and metallurgical industries [14]. Discharging of and outer diameters of 100, 5.5, and 11.5 mm, respectively. The
this effluent causes environmental pollution as well as decreases composition of clays taken for the fabrication of tubular mem-
the yield of oil. Therefore, the treatment of oily wastewater is brane is as follows: Ball clay – 18 wt%, Feldsper – 6 wt%, Kaolin –
compulsorily required before discharging. As per environment 15 wt%, Pyrophyllite – 15 wt%, Quartz – 28 wt%, Calcium carbonate
(protection) rules (1986, India), the central pollution control board – 18 wt%. Clay powders were accurately weighed according to the
framed the discharge limit of oil into surface water as 10 (mg/L), composition and mixed with the determined volume of Millipore
public sewers as 20 (mg/L), irrigation water as 10 (mg/L) and water to make the paste for extrusion. No organic additives were
coastal water as 20 (mg/L) from various industries [15]. Several used for the preparation of paste. The obtained paste was fed into
conventional techniques have been used for oily wastewater the extrusion cylinder. Then, evacuation piston forced the paste
treatment that includes electrostatic coalesce, heating treatment, through a die in a tabletop extruder to form a tubular shape
filter coalesce, centrifugal settling, gravity settling, pH adjustment membrane. The tubular membrane was extruded in the horizontal
and chemical emulsification. All of these methods have some direction with the forwarding velocity of  0.007 m/s at room
advantages and disadvantages [16,17]. Hence, membrane technol- temperature. When the tube reached the length of around
ogy comes into sight as a best competent technique for treatment 120 mm, the process extruder stopped and the tube was cut with
of oily wastewater amongst the various conventional techniques. sharp blades. Then, the obtained tubular membrane was subjected
In addition, it is very attractive due to the advantages, such as to natural drying at room temperature for 12 h. After which, the
compact design, lower energy requirement and higher separation membrane was dried at 100 1C for 12 h and 200 1C for 12 h in a hot
efficiency as compared to other existing treatment processes [18]. air oven. Subsequently, the membrane was taken to the sintering
However, the treatment cost can be further lowered by choosing process with a heating rate of 2 1C/min and sintered at 950 1C for
microfiltration (MF) membranes since they do not require so high 6 h in a box furnace. These restrained thermal treatment steps
trans-membrane pressures and have higher flux than ultrafiltra- were followed to avoid the formation of micro cracks and bends in
tion membranes [19]. Numerous authors have evaluated the the membrane. Finally, the elaborated membrane was washed
potential aspects of ceramic membranes in oily wastewater treat- with water and dried at 100 1C for further characterization.
ment. Mohammadi et al. [20] formulated a kaolin based tubular
ceramic membrane with a pore diameter of 10 mm for the treat- 2.2. Characterization of raw materials and tubular ceramic
ment of oily wastewater emulsions. The prepared membrane membrane
provided a good separation performance on various operating
conditions. Zhong et al. [21] examined the efficiency of ZrO2 The XRD study was carried out to identify the phase transforma-
ceramic membrane (average pore diameter of 0.2 μm) in oily tion behavior of the membrane before and after sintering. The
wastewater treatment. The prepared zirconia membrane displayed profiles measured in a Bruker AXS instrument using Cu Kα radiation
the oil removal of 99.4–99.9%. Yang et al. [22] attained 99.8% of oil source. The profiles were recorded in the 2θ range of 5–751 with a
rejection utilizing the commercial tubular (ZrO2/α-alumina) mem- scan rate of 0.051 s  1. Thermogravimetric (TG) and derivative
brane with an average pore diameter of 0.2 μm. In another study, thermogravimetric (DTG) analysis were performed to evaluate the
Cui et al. [17] obtained 99% of oil rejection employing the zeolite thermal stability and the minimum sintering temperature required
membrane having the pore size of 1.2 μm in microfiltration of oily for membrane fabrication. The thermal behaviors of mixed raw
wastewater. Amongst many research perceptions, the develop- materials were analyzed by Mettler Toledo TGA/SDTA 851s instru-
ment and employment of the ceramic membranes derived from ment in an air atmosphere from 30 to 1000 1C in 150 mL platinum
inexpensive raw materials for oily wastewater treatment are crucible with a heating rate of 10 1C/min. The tubular membrane
getting interest in recent days. was characterized using a field emission scanning electron micro-
Considering such research trends, the study presented here scope (FESEM, JEOL JSM-5600LV) to analyze the presence of
focuses on the preparation of the novel tubular ceramic membrane possible defects on the surfaces. A small size of the membrane
by an extrusion method and evaluation of separation performance sample was fixed on top of the stub and layered with gold using an
in treatment of oily wastewater emulsion. Contemplation on auto fine coating instrument (JEOL JFC-1300) preceding to morphol-
reducing the price of the membrane and to assess our natural ogy assessment. The porosity of the membrane was measured by
resources, the local clay materials: kaolin, quartz, ball clay, utilizing water as a soaking agent. The mechanical strength was
pyrophyllite, feldspar and calcium carbonate are utilized as raw measured using a standard three-point bending test by Computer-
materials. The performance of the membrane is investigated ized Universal Tester (DUTT-101, India). Four tubular ceramic
through microfiltration of synthetic oily wastewater emulsions membranes (length of 100 mm) were tested with the bending
94 R. Vinoth Kumar et al. / Journal of Membrane Science 490 (2015) 92–102

Flow Meter

Tubular Membrane

V2 Pressure
Gauge

V1

Permeate V3
Pump Tank
Electronic Balance

Retentate (or) Concentrate

Feed Tank

Fig. 1. Schematic representation of tangential microfiltration system (V1-by-pass valve, V2- inlet valve, V3- retentate valve).

strength instrument. The instrument support span adjusted to The pure water flux (JW) was measured by utilizing below
80 mm, before putting the sample on the support beam. Then the expression:
load was applied at a constant load rate of 10 N s  1 until failure in
Q
the breaking force value was observed. The mechanical strength of JW ¼ ð1Þ
At
the membrane was calculated from arithmetic mean of all values
obtained. The corrosion resistance of the membrane was evaluated where, Q is the permeated water in volume (m3); A is the effective
by means of loss of mass after treating in aggressive environments. membrane area (m2) and t is the measured time (s).
The acid and alkali solutions were prepared with extreme condi-
tions (HCl solution with pH 1 and NaOH solution with pH 14) and
soaked the membrane into the solutions for one week. The 2.5. Microfiltration of synthetic oily wastewater
corrosion resistance of the recovered membrane was evaluated by
weight decrement of the membrane. All the filtration tests were carried at an ambient temperature
( 25 1C) by using the same laboratory scale microfiltration system
illustrated in Fig. 1. The synthetic oily wastewater emulsions were
2.3. Microfiltration system made using crude oil with Millipore water as follows: a requisite
quantity of crude oil was added into 10 L of Millipore water to
A representational diagram of the overall tangential microfil- make the various concentrations (50, 75, 100, 150 and 200 ppm) of
tration system is presented in Fig. 1. The system was formulated in feed stocks. The suspension was subjected to ultrasonication (Elma
a series of feed tank, metering pump for circulation of the feed T460, India) for about 8–12 h to get stable oil–water emulsions.
solution, membrane module with tubular configuration, pressure The emulsification process was carried out without the addition
indicator, and flow meter to measure the cross flow rate along of any emulsifier agents. The oil droplet size in the feed was
with flow control valves (3 nos.). The prepared above tubular calculated by laser diffraction spectroscopy (Malvern Mastersizer,
ceramic membrane was incorporated into the module for tangen- United Kingdom). The size dispersion of oil droplet is presented in
tial cross flow microfiltration. The feed was pumped to the module the Fig. 2 and the mean droplet size varies from 0.71 to 0.74 μm for
from the feed tank and retentate employed with the recirculation 50, 75 and 100 ppm of oil concentrations and 1.18 to 1.21 μm for
arrangement. The permeate flow through the membrane was 150 and 200 ppm oil concentrations. The tangential microfiltration
determined using the electronic weighing balance at open atmo- study on synthetic oily wastewater was performed for a period of
spheric condition. The operating pressure was adjusted by hand 1 h at various operating conditions such as applied pressures of
operation using the flow control valves installed in by-pass (V1), 69–345 kPa, feed concentrations of 50–200 ppm and cross flow
inlet (V2) and retentate (V3) flow path. Cross flow rate was rates of 5.55
10  7–1.66
10  6 m3 s  1.
adjusted with a flow control valve (V3) equipped in retentate side The observed rejection was determined by below expression:
flow path. Cf  Cp
Rð%Þ ¼
100 ð2Þ
Cf

2.4. Measurement of water flux through membrane where Cf is the oil concentration in the feed stream, Cp is the oil
concentration in the permeate stream and R is the observed
In order to determine the water flux, the collection of permeate rejection (%).
was measured at different applied pressures (69–345 kPa) as a After performing the primary calibration studies, the accurate
function of time at a fixed cross flow rate using in-house made concentration of oily wastewater was determined by measuring
setup (see Fig. 1). The different pressures were regulated by a by- the absorbance of the oily wastewater emulsions using ultravio-
pass valve on the system starting from the lowest to the highest let–visible spectrophotometry (Thermo Scientific, UV-2300) at a
pressure. wavelength of 236 nm.
 R. Vinoth Kumar et al. / Journal of Membrane Science 490 (2015) 92–102 95

10 the following linearized equations of the membrane permeate flux

(J) and time (t) as:
50 ppm
75 ppm ðaÞ:Complete pore blocking : ln J  1 ¼ ln J 0 1 þ kb t ð3Þ
8
100 ppm
150 ppm ðbÞ:Standard pore blocking : J  0:5 ¼ J 0 0:5 þ ks t ð4Þ
200 ppm
6 ðcÞ:Intermediate pore blocking : J  1 ¼ J 0 1 þki t ð5Þ
Volume (%)

ðdÞ:Cake filtration : J  2 ¼ J 0 2 þkc t ð6Þ

4 1  0.5 1 2
Hence, the plot of ln (J ) vs. t, J vs. t, J vs. t and J vs. t
shall be a linear line with the slope of kb, ks, ki and kc, with
2
y-intercept of ln(J0 1), J0 0.5, J0 1 and J0 2 for complete, standard,
intermediate pore blocking and cake filtration model, respectively.
Calculating the correlation coefficient (R2) values acquired from
linear regression plots, the suitable fitness and competency of
0
various fouling models can be described. Ultimately, the model
0.01 0.1 1 10 100 that signifies experimental data with best fit R2 value (nearby 1) is
Droplet Size (μm) considered to refer the suitable fouling mechanism during micro-
filtration.
Fig. 2. Droplet size distribution of oil-water emulsions.

2.6. Membrane regeneration 4. Result and discussion

The membrane was thoroughly rinsed and regenerated after 4.1. Characteristics of tubular ceramic membrane
every experimental run. For cleaning the membrane, the following
steps were adopted sequentially: First, the membrane was washed The phase transformation activities of the membrane before
by passing the cleansing agent (commercially acquired surf-excel and after sintering were analyzed through XRD patterns as
powder solution of 1 g/L) to the filtration system for 1 h to depicted in Fig. 3. The sintering process generates various phase
eliminate the deposited oil on the surface of the membrane during transformations and reactions, which initiates the creation of new
the filtration. After which, the whole system was cleaned by phases. This gives the absence as well as transfer of the position of
passing Millipore water for 1 h. After conducting the cleaning the XRD peaks. The perfect crystalline phase is observed for the
process, the water flux of the membrane was evaluated to assure sintered membrane and the mainly identified phases for the
about negligible flux decline due to partial plugging. The full membrane (after sintering) are mullite (3Al2O3  2SiO2), quartz
recuperation of the membrane was inspected by calculating the (SiO2), Wollastonite (CaSiO3) and anorthite (CaO.Al2O3  2SiO2).
membrane hydraulic permeability equal to the actual hydraulic Kaolin to mullite through metakaolinite phase transformation is
permeability or within the limits of 7 2% of actual permeability. identified in the temperature range of 800 1C to 1000 1C [24]. This
takes place due to the degradation of kaolin structure. The thermal
investigations of mixed raw materials are performed to determine
3. Description of the pore blocking model for flux decline the sintering temperature to achieve a good solid membrane. TGA
analysis and DTG curves of mixed raw materials are presented in Fig. 4. In
TGA profile, the adsorbed water is gradually evolved at tempera-
To identify the presentable microfiltration flux decline profile tures up to 200 1C. The elimination of structural water/surface
of oily wastewater emulsion, four fouling models, such as hydroxyl group by dehyroxylation reaction takes place in the
(a) complete, (b) standard, (c) intermediate pore blocking and temperature range of 300–500 1C. A significant weight decrement
(d) cake filtration model, have been analyzed [23]. occurs owing to the thermal degradation (500–750 1C) of calcium
(a). Complete pore blocking takes place when the size of the carbonate to form calcium oxide and carbon dioxide. The porosity
solute particles is bigger than the pores of the membrane. It causes of the membrane mostly depends on the pathway of evolved CO2
the pore blocking on the surface of the membrane and does not gas. It is identified that above 850 1C, a negligible weight loss is
cause inside pores of the membrane. (b). Standard pore blocking
occurs due to the non-uniformity of pore paths. When the solute 900 2
particles are smaller than the pores of the membrane, it creates 800 (a) Before sintering
(b) After sintering
the pore blocking inside the pores of the membrane. Therefore, the
Intensity (Arbitrary units)

700
pore volumes of the membrane decrease proportionately to the
600
filtered volume of permeate. (c). Intermediate pore blocking
causes when the solute particle size and membrane pore size are 500

equivalent. It is considered in this model that the solute particles 400

1
are not significantly blocked the pores of the membrane. It may be 300 5
1 6
blocked by some solute particles settling over others. As a result, 1 3 4
200
the non blocked membrane surface area reduces with time. (d). 1 2
100
Cake filtration model is applicable where particles bigger than the (a)
(b)
average pore size accumulate on the surface of the membrane thus 0

forming a cake. Therefore, the cake grows up with time and gives 10 20 30 40 50 60 70 80

an extra porous barrier to the permeating liquid. The porous 2θ (Degrees)

barrier may also mainly impact membrane separation character- Fig. 3. XRD patterns of the membrane (a) before sintering and (b) after sintering
istics. The assorted pore blocking models can be delineated with (1- Kaolin, 2- Quartz, 3-Anorthite, 4- Calcium oxide, 5- Mullite, 6-Wollastonite).
96 R. Vinoth Kumar et al. / Journal of Membrane Science 490 (2015) 92–102

observed. It recommends that the membrane needs to be sintered microscope cross-sectional image of the membrane with their
at above 850 1C to obtain a good mechanical and thermal strength. dimensions. It gives satisfactory cross-sectional circularity. In addi-
DTG analysis shows an endothermic phenomenon from ambient tion, FESEM images were used to analyze both inner and outer
to below 100 1C is associated to the physisorbed water and another surface morphology of the fabricated membrane (see Fig. 5(c and d)).
endothermic phenomenon towards 480 1C is due to the dehydrox- These images provide information on consistency of prepared
ylation. A specific endothermic transformation at 650 1C–750 1C is membrane surfaces. One can see a homogeneous surface with no
due to the loss of structural hydroxyl groups of the kaolinite. cracks on the membrane surfaces. Also, the membrane is highly
The following possessions are most important for the develop- smooth and a flawless inner and outer surfaces. The overall surface
ment of a good quality membrane: excellent surface, including morphological analysis recommends that the nonappearance of
nonappearance of any big pores/defects/cracks, porosity, mechan- cracks/defects/ big pores is a key condition leading to a good quality
ical strength and corrosion resistance. membrane.
By following the optimized preparation procedure, the perfect The porosity of the membrane is measured using below
tubular shaped ceramic membrane was successfully prepared. The expression [25]:
tube obtained under this condition was straight with a constant W wet  W dry
thickness and diameter (see Fig. 5). Photograph of the sintered Porosity ð%Þ ¼
100 ð7Þ
ρwater
V membrane
membrane shown in Fig. 5 (a) confirms that the sintering process
was successful. Fig. 5 (b) displays the universal serial bus (USB) where, Wwet is the wet weight of the membrane (placed in
distilled water for overnight), Wdry is the dry weight of the
membrane (dried at 120 ◦C for 3 h), Vmembrane is the total volume
100 0.0
of the membrane and ρwater is the density of the water. The
average porosity of the tubular membrane is calculated to be 53%.
98 The standard three point bending test is performed to examine
-0.2
96 the mechanical strength of the membrane using the following
Derivate weight (% / C)

expression [1]:
ο

-0.4
94
Weight (%)

8FLdout
92 -0.6 σ¼ 4 4
ð8Þ
π ðdout  din Þ
90 -0.8 where, F is the applied force (N), L is the distance between the
88 sample support points (mm), din and dout are the inner and outer
TGA
-1.0 diameter of the tube (mm), respectively and σ is the mechanical
86
DTG strength of the membrane (MPa). The average mechanical stability
-1.2 of the membrane is calculated to be 12 MPa. This value is
84
comparable and even better than the pyrophyllite membrane
82 -1.4
(5.5 MPa) prepared by Talidi et al. [10].
100 200 300 400 500 600 700 800 900 1000
ο The corrosion resistance test was performed with respect to
Temperature ( C)
weight decrement of the membrane after keeping in harsh
Fig. 4. TGA and DTG curves of mixed raw materials. environments (acid and alkali). The weight decrement of the

Fig. 5. Photograph of the sintered membrane (a) Microscope image of cross sectional view of the membrane (b) FESEM images of inner (c) and outer surfaces of the
membrane (d).
 R. Vinoth Kumar et al. / Journal of Membrane Science 490 (2015) 92–102 97

Table 1
0.00030 Properties of the tubular ceramic membrane.
69 kPa
138 kPa Properties Membrane
0.00025 207 kPa Dimensions:
278 kPa Outer diameter (mm) 11.5
345 kPa
Channel diameter (mm) 5.5
Water flux (m /m s)

Thickness (mm) 3
2

0.00020
Length (mm) 100
3

Unit surface (m2) 1.72
10  3

Average pore size (μm) 0.309
0.00015 Porosity (%) 53
Water permeability (m/s kPa) 5.93
10  7
Mechanical strength (MPa) 12
0.00010 Chemical stability
Acid - weight loss (%) 9
Base - weight loss (%) 0
0.00005

0 100 200 300 400 500 600 700 800 900 pores are cylindrical in shape [27–29].
Time (s) εr 2 ΔP
JW ¼ ð10Þ
8μτl
0.00024 where, ε is the porosity of the membrane (0.53), r is the pore radius of
the membrane (m), ΔP is the applied pressures across the membrane
(0–345 kPa), l is the pore length (0.003 m) which is generally taken as
0.00020
thickness of the membrane, τ is the tortuosity factor (generally used
as 1), and μ is the viscosity of water (0.00089 Pa s).
Water flux (m /m s)

0.00016 By combining Eqs. 9 and 10, the corresponding pore radius of

2

the membrane can be evaluated from

3

 1=2
0.00012 L 8μτl
r¼ h ð11Þ
ε
0.00008 The average pore size of the membrane is calculated to be
0.309 μm (0.309
10  6 m). It is worth to mention that, this
0.00004 membrane can be directly utilized for making of ultrafiltration
membranes with exclusion of intermediate layers.
Table 1 summarizes the overall properties (characterization
0.00000 results) of the prepared tubular ceramic membrane. It can be
0 50 100 150 200 250 300 350 400 concluded that the prepared membrane could be applied for
Applied prressure (kPa) industrial utilization due to the excellent membrane characteris-
Fig. 6. (a) Water flux as a function of time for five applied pressures and (b) water tics such as quality surface, pore size, porosity, mechanical
flux as a function of applied pressure. strength, corrosion resistance and water permeability.

4.2. Performance in synthetic oily wastewater treatment

membrane is calculated to be 9% in acidic condition and 0% in
alkali condition. In acid and alkali conditions, the membrane The elaborated novel tubular ceramic membrane was applied
displays an excellent result in resisting corrosion. The obtained for the treatment of synthetic oily wastewater emulsions. Applied
results are more commensurable in comparison with the cordier- pressure, feed oil concentration and cross flow rates are the
ite membrane fabricated by Dong et al. [26]. important variables that affect the treatment process in terms of
Fig. 6(a) illustrates the water flux of the membrane as a function permeate flux and rejection. Hence, the effect of these parameters
of time for different applied pressures. The steady flux is attained was investigated as a function of time.
for the entire measured time. The variations of applied pressure on
water flux are also presented in Fig. 6(b). It can be noticed that the 4.2.1. Influence of applied pressure on permeate flux and rejection
water flux increases linearly with an increase of applied pressures Fig. 7(a) shows the permeate flux profiles with time for various
(69–345 kPa). This stipulates that the variation in pressure is the applied pressures (69, 138, 207, 278 and 345 kPa) at a cross flow
barely driving force for permeation. For transportation operation rate of 1.11
10  6 m3/s with feed concentration of 100 ppm. It can
exclusively by convection, the flow rate is proportionate to the be seen from the figure that the permeate flux declines sharply at
pressure, and is in accordance with Darcy's law. the initial period and becomes gradual thereafter in the micro-
The water permeability (Lh) of the membrane was determined filtration run. The reasons for the flux reductions are owing to
by the water flux at different applied pressures. It was evaluated concentration polarization on the surface of membrane and
from the slope of the pure water flux (JW) versus applied pressure blocking of pore in the porous ceramic structure. In addition, it
across the membrane (ΔP) (see Fig. 6 (b)) using the equation. is noticed that the permeate flux increases with increasing applied
pressures due to the higher driving forces acted on the membrane.
J W ¼ Lh ΔP ð9Þ However, the permeate flux profile on increasing applied pressure
does not follow the linear trend (see Fig. 7(a) (inset)). This is due to
The water permeability (Lh) of the membrane is found to be concentration polarization and adsorptive of oil on the surface that
5.93
10  7 m/s kPa. Hagen–Poiseuille equation was employed to creates further resistances to transport liquid through the mem-
estimate the average pore size of the membrane by assuming brane [30]. The rate of flux decline is higher at higher applied
98 R. Vinoth Kumar et al. / Journal of Membrane Science 490 (2015) 92–102

0.000126 0.00016

69 kPa 50 ppm
0.000112 138 kPa 0.00014 75 ppm
207 kPa 100 ppm
278 kPa 150 ppm
0.000098 345 kPa 0.00012

Permeate flux (m /m s)
200 ppm
Water flux (m /m s)

2
2

3
0.000084 0.00010
3

0.000070 0.00008

0.000056
0.00006

0.000042
0.00004
0.000028
0.00002

0 500 1000 1500 2000 2500 3000 3500 0 500 1000 1500 2000 2500 3000 3500
Time (s) Time (s)

100 100

98 98
Oil rejection (%)

Oil rejection (%)

96 96

200 ppm
69 kPa 150 ppm
94 138 kPa 94
100 ppm
207 kPa
75 ppm
278 kPa
50 ppm
92 345 kPa 92

90 90
10 20 30 40 50 60 10 20 30 40 50 60
Time (min) Time (min)

Fig. 7. Influence of applied pressure on (a) permeate flux and (b) oil rejection. Fig. 8. Influence of feed concentration on (a) permeate flux and (b) oil rejection.

pressure due to the immediate formation of oil layer on the surface resulting a declined premeate flux at higher feed concentration
of the membrane, which makes membrane fouling faster while (see Fig. 8(a) (inset)). Similar kind of trends is also reported in the
increasing the applied pressure. Fig. 7 (b) presents the rejection of literature [14]. From Fig. 8(b), it is apparent that the rejection
oil with time for various applied pressures. It is noticed that the oil enhances with an increase in feed concentration due to the
rejection declines with an increase in applied pressure. This occurs enhancement of oil droplet size and density of droplet at higher
due to the fact that higher pressures assist the enrichment of concentrations. At higher concentrations, the coalescence of the oil
coalescence and wetting of oil droplets, therefore imposing some droplets forms a larger droplet that offers a greater rejection. The
oil droplets to transfer through the membrane pores and attain the highest percentage of oil rejection (99.64%) is obtained for the feed
permeate side. Moreover, the percentage of oil rejection slightly concentration of 200 ppm with permeate flux of 1.74
10  5 m/s.
increases with duration of microfiltration. This is possibly due to
the reduction of membrane pore sizes, which happens by the
adsorption of oil droplets on the surface of the membrane. The 4.2.3. Influence of cross flow rate on permeate flux and rejection
fabricated membrane exhibits excellent rejection performance Fig. 9(a) represents permeate flux plots with respect to time for
according to the results depicted in Fig. 7(b). The membrane different cross flow rates (5.55
10  7, 1.11
10  6 and 1.66

shows the highest rejection of 99.98% at a lower applied pressure 10  6 m3/s) at applied pressure of 207 kPa with feed concentration
of 69 kPa with permeate flux of 3.16
10  5 m/s. of 100 ppm. It is noticed that the permeate flux of the membrane
enhances with increasing the cross flow rate (see Fig. 9(a) (inset)).
It can be explained that an increment of the cross flow rate
4.2.2. Influence of feed concentration on permeate flux and rejection reduces the concentration polarization. Additionally, increasing
Fig. 8(a) illustrates the permeate flux plots with time for the cross flow rate offers enhancement in the shear stress on the
various feed concentrations (50, 75, 100, 150 and 200 ppm) at membrane surface, which diminishes the thickness of the adhered
applied pressure of 207 kPa with the cross flow rate of 1.11
oil layer on the surface of the membrane. Fig. 9(a) signifies that an
10  6 m3/s. Fig. 8(a) clearly demonstrates that an increase in the oil increase in the cross flow rate decreases the rate of flux decline.
concentration results larger flux decline. The permeate flux greatly This is due to the fact that the increasing cross flow rate restricts
declines when the concentration of oily water emulsion augments. the cake layer occurrence on the surface of the membrane. Fig. 9
The coalesced oil sticks over the membrane, which creates fouling (b) points out that the percentage of oil rejection reduces slightly
 R. Vinoth Kumar et al. / Journal of Membrane Science 490 (2015) 92–102 99

model parameters of four filtration models for microfiltration of

0.00010 oily wastewater are summarized in the Table 2. Based on the R2
(correlation coefficient) values of various models, the best fit of the
model can be identified. According to the R2 value, the cake
0.00008 filtration model is regarded as best model to portray the flux
decline mechanism for microfiltration of oily wastewater. The
Permeate flux (m /m s)
2

identical results reported by Salahi et al., [31] also indicated that

3

the flux decline throughout the microfiltration of oily wastewater

0.00006
with the polymer membrane was well described by the cake
filtration model.
The values of model parameters (k) evaluated from different
0.00004 filtration models are also presented in the Table 2. As stated by the
physical meaning and description of Hermia's model parameters,
-7 3
5.55 × 10 m /s the value of ‘k’ designates the severity of fouling. It can be noticed
0.00002 -6 3
1.11 × 10 m /s that the initial permeate fluxes are calculated from intercept of
-6 3
1.66 × 10 m /s various filtration models are lower than the initial permeate flux
obtained from experimental results and thereby, all models have a
0 500 1000 1500 2000 2500 3000 3500 tendency to underrate the initial flux. In addition, a huge variation
Time (s) is detected except for the cake filtration model, and it offers a fine
conformity to the obtained experimental results as compared to
102 other models. The entire study confirms that the cake filtration
model best fits with the obtained experimental result. These
examinations throughout the microfiltration are an auspicious
100 act towards the life of the membrane, as cake filtration facilitates
exterior fouling and may not provide considerably towards the
irreversible fouling of the membrane.
Oil rejection (%)

98 The fitness of four pore blocking models for various feed

concentrations (50, 75, 100, 150, 200 ppm) and each cross flow
rates (5.55
10  7, 1.11
10  6 and 1.66
10  6 m3/s) are presented
-7 3
96 5.55 × 10 m /s as a supplementary data (see Figs. S1 and S2). Summary of
-6 3
1.11 × 10 m /s parameters associated with pore blocking models for different
feed concentrations and cross flow rates are given in the supple-
-6 3
1.66 × 10 m /s
94 mentary data as Tables S1 and S2.

4.4. Potential evaluation of the elaborated membrane with other

92
membranes
10 20 30 40 50 60
Time (min)
The performance of the present membrane in treatment of oily
Fig. 9. Influence of cross flow rate on (a) permeate flux and (b) oil rejection. wastewater emulsion is analyzed with other membranes reported
in the literature. It is evident from the Table 3 that the highest
rejection of 99.98% achieved in this exertion at lower applied
with increasing cross flow rate. This is due to the reduction of oil
pressure and cross flow rate is considerably equivalent and more
layer thickness on the surface of the membrane at a higher cross
preferable than those described in the literature. In the work
flow rate, as we discussed previously. At a lower cross flow rate,
investigated by Mueller et al., [32], α-alumina membrane exhibited
the formation of the cake layer is more, which acts as an additional
99.9% of oil rejection with permeate flux of 8.33
10  6 m3/m2 s
barrier on the surface of the membrane resulting improved oil
for oil concentration of 250 mg/L. It is noteworthy to mention that
rejection. The higher cross flow rates depreciates the formation of
the prepared membrane provides a tough competitive to replace
cake layer, consequently the oil rejection decreases. For that
α-alumina membrane. In particularly, the synthesized membrane
reason, the percentage of oil rejection declines with rising cross
demonstrates one order higher permeate flux with keeping the
flow rates. The highest percentage of oil rejection (99.86%) is
same rejection value (99.98%) even at lower feed concentration
obtained for lower cross flow rate of 5.55
10  7 m3/s with per-
(100 mg/L). From the assessment survey (Table 3), it can be stated
meate flux of 2.24
10  5 m/s.
that the elaborated membrane is better than that of other
membranes reported in the literature.
4.3. Fouling analysis

For microfiltration processes, it is essential to indentify the 4.5. Estimation of cost based on raw materials
pertinent flux decline mechanism. The permeate flux decline
mechanisms are investigated using various pore blocking models The cost of the prepared tubular ceramic membrane is eval-
during microfiltration of synthetic oily wastewater. These models uated on the basis of raw materials utilized in the present study
provide the platform to study the flux decline, which is influenced and it is estimated to be Rs. 1/membrane (see Table 4). The raw
by pore blocking or cake filtration according to fitness of obtained materials cost of the membrane per unit area is estimated to be Rs.
data from microfiltration experiments. Fig. 10 elucidates the 254/m2 (4 $/m2). In comparison, α-alumina ceramic symmetric
fitness of pore blocking models for various applied pressures (69, membranes cost approximately $500/m2 and symmetric stainless
138, 207, 278 and 345 kPa) with cross flow rate of 1.11
10  6 m3/s. steel membrane costs approximately $3000/m2 [36]. Therefore, it
From these plots, it can be described that cake filtration model can be concluded from the cost estimation that the prepared
offers the best fit with the obtained experimental data. The various membrane is inexpensive as compared to alumina membranes,
100 R. Vinoth Kumar et al. / Journal of Membrane Science 490 (2015) 92–102

11.0 240

220

10.5 200
3

1.5
180
ln (J ), m s/m

, ms /m
2

0.5
10.0 160
-1

-0.5
140

J
69 kPa
138 kPa 69 kPa
9.5 207 kPa 120 138 kPa
278 kPa 207 kPa
345 kPa 100 278 kPa
345 kPa
9.0 80
0 500 1000 1500 2000 2500 3000 3500 4000 0 500 1000 1500 2000 2500 3000 3500 4000
Time (s) Time (s)

50
210
45

40 180

35 150
6
3

J x 10 , m s /m
J x 10 , m s/m

30
4 2
2

120
25
-3

-7

90
20
-1

-2

69 kPa 60
15
138 kPa 69 kPa
10 207 kPa 30 138 kPa
278 kPa 207 kPa
5 345 kPa 278 kPa
0 345 kPa
0
0 500 1000 1500 2000 2500 3000 3500 4000 0 500 1000 1500 2000 2500 3000 3500 4000
Time (s) Time (s)

Fig. 10. Linearized plots of permeate flux versus time for various pore blocking models (applied pressures: 69–345 kPa; feed concentration: 100 ppm; cross flow rate:
1.11
10  6 m3/s) (a) Complete pore blocking model, (b) Standard pore blocking model, (c) Intermediate pore blocking model, (d) Cake filtration model.

Table 2
Summary of parameters associated with various pore blocking models for different applied pressures.

Pressure Complete pore blocking Standard pore blocking Intermediate pore blocking Cake filtration

1  0.5 1
(kPa) R 2
kb (s ) lnðJ o 1 Þ R2
ks (s 0.5
m ) J o 1
10  2 R 2
ki (m ) J 0 1
10  4 R2 kc (s m  2) J o 2
10  8

69 0.974 0.00011 10.34 0.982 0.01166 1.76 0.989 4.57046 3.07 0.997 353321.17 9.04
138 0.919 0.00028 9.75 0.954 0.02311 1.29 0.978 7.72818 1.60 0.999 450082.33 1.54
207 0.963 0.00029 9.59 0.986 0.02260 1.18 0.991 7.12208 1.30 0.997 368495.05 0.63
278 0.916 0.00026 9.46 0.950 0.01827 1.12 0.975 5.18478 1.21 0.998 216360.29 1.02
345 0.912 0.00026 9.39 0.948 0.01766 1.08 0.973 4.82586 1.12 0.998 186885.77 0.88

Table 3
Potential evaluation of other membranes with prepared membrane.

Membrane material Pressure (kPa) Oil concentration (mg/L) Cross flow rate (m3/s) Permeate flux (m3/m2 s) Rejection (%) Reference no.

α- alumina 69 250 1.94
10  3 8.33
10  6 99.9 [32]

α- alumina 125 141(TOC) 1.82
10  2 8.05
10  5 97.8(TOC) [33]
Zeolite/ α- alumina 100 100 8.05
10  5 1.66
10  5 98.8 [18]
Zirconia/ α- alumina 100 5500 2.43
10  2 7.75
10  6 94.3 [22]
Mullite 300 1000 8.10
10  3 1.11
10  5 94(TOC) [34]
Carbon 100 120 8.11
10  4 1.78
10  5 97.8 [35]
Mixed clays 69 100 1.11
10  6 3.16
10  5 99.98 Present work
 R. Vinoth Kumar et al. / Journal of Membrane Science 490 (2015) 92–102 101

Table 4
Cost analysis of the raw materials for making one tubular membrane.

Raw Unit price (Rs./ Materials used for preparation of one membrane Cost of raw materials used for the fabrication of one tubular membrane
materials kg) (g) (Rs.)

Kaolin 12 1.428 0.017

Quartz 20 2.628 0.052
CaCO3 360 1.694 0.609
Ball clay 6 1.737 0.010
Phyrophyllite 10 1.456 0.014
Feldspar 8 0.553 0.004
Total cost for raw materials used for the fabrication of one tubular membrane 0.706
(Round off value) Rs. 1.000

based on the sintering temperature and raw materials utilized in Vmembrane total volume of the membrane (m3)
this study [37]. F applied force (N)
L distance between the sample support points (mm)
din inner diameter of the tube (mm)
5. Conclusions dout outer diameter of the tube (mm)
r pore radius of the membrane (μm)
Low cost novel tubular ceramic membrane has been effectively ΔP applied pressure across the membrane (kPa)
fabricated using inexpensive clay mixtures by the extrusion l pore length of the membrane (m)
technique. The prepared membrane offers better porosity (53%), Lh water permeability of the membrane (m3 m  2 s  1
lower average pore size (0.309 mm), strong mechanical strength kPa)
(12 MPa) and excellent corrosion resistant. Moreover, the elabo- Jo initial permeate flux (m3 m  2 s)
rated membrane is highly fulfilled in oily wastewater treatment. kb complete pore blocking model constant (s  1)
Microfiltration studies demonstrate that the permeate flux kc cake filtration model constant (s m  2)
increases and the rejection decreases when the applied pressure ki intermediate pore blocking model constant (m  1)
and cross flow rate increase. The highest rejection of 99.98% is ks standard pore blocking model constant (m  0.5
attained with permeate flux of 3.16
10  5 m/s at applied pressure s  0.5)
of 69 kPa. The investigation on fouling mechanisms with various R2 square of correlation coefficient (dimension less)
pore blocking models elucidates that the cake filtration model Rs. Indian rupee
describes well the experimental data. The raw materials cost of the $ United States dollar
prepared membrane is estimated to be Rs. 254/m2, which is lower
as compared to other ceramic membranes. Overall, the presented
Greek symbols
work suggests the competency of the elaborated membrane in the
direction of utilization towards the treatment of oily wastewater
emulsion. In addition, the fabricated tubular membrane can be ρwater density of the water (kg m  3)
σ mechanical strength of the membrane (MPa)
used as a support for the manufacturing of ultrafiltration (UF) and
nanofiltration (NF) membranes. ε porosity of the membrane (%)
τ tortuosity factor (dimension less)
μ Viscosity of water (Pas)
Acknowledgments

We would like to thank the Central Instruments Facility at IIT

Guwahati for helping us to perform FESEM analysis. References

[1] O. David, Y. Gendel, M. Wessling, Tubular macro-porous titanium membranes,

Appendix A. Supporting information
J. Membr. Sci. 461 (2014) 139–145.
[2] G.C.C. Yang, C.M. Tsai, Effects of starch addition on characteristics of tubular
Supplementary data associated with this article can be found in porous ceramic membrane substrates, Desalination 233 (2008) 129–136.
[3] J. Fang, G. Qin, W. Wei, X. Zhao, Preparation and characterization of tubular
the online version at http://dx.doi.org/10.1016/j.memsci.2015.04.066.
supported ceramic microfiltration membranes from fly ash, Sep. Purif. Tech-
nol. 80 (2011) 585–591.
[4] P. Hristov, A. Yoleva, St. Djambazov, I. Chukovska, D. Dimitrov, Preparation and
characterization of porous ceramic membranes for micro-filtration from
natural zeolite, J. Chem. Technol. Metall. 47 (2012) 476–480.
Nomenclature [5] Y.C. Dong, J.E. Zhou, B. Lin, Y.Q. Wang, S.L. Wang, L.F. Miao, Y. Lang, X.Q. Liu, G.
Y. Meng, Reaction-sintered porous mineral-based mullite ceramic membrane
Jw pure water flux (m3 m  2 s  1) supports made from recycled materials, J. Hazard. Mater. 172 (2009) 180–186.
[6] M. Cheryan, Ultrafiltration and Microfiltration Handbook, second ed., Tech-
Q permeated water in volume (m3) nomic Publishing Company, Inc., Lancaster, 1998.
A effective membrane area (m2) [7] H.P. Hsieh, Inorganic Membranes for Separation and Reaction, Elsevier,
t measured time (s) Amsterdam, The Netherlands, 1996.
[8] J. Mulder, Basic Principle of Membrane Technology, 2nd edn., Kluwer Aca-
Cf oil concentration in the feed stream (mg L  1) demic Publishers, The Netherlands, 1997.
Cp oil concentration in the permeate stream (mg L  1) [9] A.J. Burggraaf, L. Cot, Fundamentals of Inorganic Membrane Science and
R observed rejection (%) Technology, Elsevier, Amsterdam, The Netherlands, 1996.
[10] A. Talidi, N. Saffaj, K.E. Kacemi, S.A. Younssi, A. Albizane, A. Chakir, Processing
Wwet wet weight of the membrane (g) and characterization of tubular ceramic support for microfiltration membrane
Wdry dry weight of the membrane (g) prepared from pyrophyllite clay, Sci. Study Res.: Chem. Chem. Eng., Biotechnol.
Food Indus. 12 (2011) 263–268.
102 R. Vinoth Kumar et al. / Journal of Membrane Science 490 (2015) 92–102

[11] N. Saffaj, M. Persin, S.A. Younsi, A. Albizane, M. Cretin, A. Larbot, Elaboration [24] P. Monash, G. Pugazhenthi, Development of ceramic supports derived from
and characterization of microfiltration and ultrafiltration membranes depos- low-cost raw materials for membrane applications and its optimization based
ited on raw support prepared from natural Moroccan clay: application to on sintering temperature, Int. J. Appl. Ceram. Technol. 8 (2011) 227–238.
filtration of solution containing dyes and salts, Appl. Clay Sci. 31 (2006) [25] R.V. Kumar, A.K. Basumatary, A.K. Ghoshal, G. Pugazhenthi, Performance
110–119. assessment of an analcime-C zeolite-ceramic composite membrane by
[12] M. Masmoudi, A. Larbot, H.E. Feki, R.B. Amar, Elaboration and characterisation removal of Cr(VI) from aqueous solution, RSC Adv. 5 (2015) 6246–6254.
of apatite based mineral supports for microfiltration and ultrafiltration [26] Y. Dong, X. Feng, D. Dong, S. Wang, J. Yang, J. Gao, X. Liu, G. Meng, Elaboration
membranes, Ceram. Int. 33 (2007) 337–344. and chemical corrosion resistance of tubular macro-porous cordierite ceramic
[13] F. Bouzerara, A. Harabi, S. Achour, A. Larbot, Porous ceramic supports for membrane supports, J. Membr. Sci. 304 (2007) 65–75.
membranes prepared from kaolin and doloma mixtures, J. Eur. Ceram. Soc. 26 [27] M.C. Almandoz, J. Marchese, P. Pradanos, L. Palacio, A. Hernandez, Preparation
(2006) 1663–1671. and characterization of non-supported microfiltration membranes from alu-
[14] P. Monash, G. Pugazhenthi, Effect of TiO2 addition on the fabrication of ceramic minosilicates, J. Membr. Sci. 241 (2004) 95–103.
membrane supports: a study on the separation of oil droplets and bovine [28] W.R. Bowen, A.W. Mohammad, N. Hilal, Characterisation of nanofiltrtation
membranes for predictive purposes-use of salts uncharged solutes and atomic
serum albumin (BSA) from its solution, Desalination 279 (2011) 104–114.
force microscopy, J. Membr. Sci. (1997) 91–105.
[15] Central Pollution Control Board (CPCB), India (2014). 〈http://www.cpcb.nic.in/
[29] A.K. Basumatary, R.V. Kumar, A.K. Ghoshal, G. Pugazhenthi, Synthesis and
Industry_Specific_Standards.php〉.
characterization of MCM-41-ceramic composite membrane for the separation
[16] T. Mohmmadi, M. Kazemimoghadam, M. Saadabadi, Modeling of membrane
of chromic acid from aqueous solution, J. Membr. Sci. 475 (2015) 521–532.
fouling and decline in reverse osmosis during separation of oil in water
[30] B. Chakrabarty, A.K. Ghoshal, M.K. Purkait, Ultrafiltration of stable oil-in-water
emulsions, Desalination 157 (2003) 369–375.
emulsion by polysulfone membrane, J. Membr. Sci. 325 (2008) 427–437.
[17] J. Cui, X. Zhang, H. Liu, S. Liu, K.L. Yeung, Preparation and application of zeolite/
[31] A. Salahi, A. Gheshlaghi, T. Mohammadi, S.S. Madaeni, Experimental perfor-
ceramic microfiltration membranes for treatment of oil contaminated water,
mance evaluation of polymeric membranes for treatment of an industrial oily
J. Membr. Sci. 325 (2008) 420–426. wastewater, Desalination 262 (2010) 235–242.
[18] M. Cheryan, N. Rajagopalan, Membrane processing of oily streams. Waste [32] J. Mueller, Y.W. Cen, R.H. Davis, Crossflow microfiltration of oily water,
water treatment and waste reduction, J. Membr. Sci. 151 (1998) 13–28. J. Membr.Sci. 129 (1997) 221–235.
[19] X.S. Yi, S.L. Yu, W.X. Shi, S. Wang, N. Sun, L.M. Jin, C. Ma, Estimation of fouling [33] S.R.H. Abadi, M.R. Sebzari, M. Hemati, F. Rekabdar, T. Mohammadi, Ceramic
stages in separation of oil/water emulsion using nano-particles Al2O3/TiO2 membrane performance in microfiltration of oily wastewater, Desalination
modified PVDFUF membranes, Desalination 319 (2013) 38–46. 265 (2011) 222–228.
[20] T. Mohammadi, A. Pak, M. Karbassian, M. Golshan, Effect of operating [34] M. Abbasi, M. Mirfendereski, M. Nikbakht, M. Golshenas, T. Mohammadi,
conditions on microfiltration of an oil-water emulsion by a kaolin membrane, Performance study of mullite and mullite-alumina ceramic MF membranes for
Desalination 168 (2004) 201–205. oily wastewaters treatment, Desalination 259 (2010) 169–178.
[21] J. Zhong, X. Sun, C. Wang, Treatment of oily wastewater produced from [35] C.W. Song, T.H. Wang, Y.Q. Pan, J.S. Qiu, Preparation of coal-based microfiltra-
refinery processes using flocculation and ceramic membrane filtration, Sep. tion carbon membrane and application in oily wastewater treatment, Sep.
Purif. Technol. 32 (2003) 93–98. Purif. Technol. 51 (2006) 80–84.
[22] C. Yang, G. Zhang, N. Xu, J. Shi, Preparation and application in oil-water [36] Mott Metallurgical Corporation, USA, (2007). 〈http://www.mottcorp.com〉.
separation of ZrO2/α-Al2O3 MF membrane, J. Membr. Sci. 142 (1998) 235–243. [37] V. Mavrov, H. Chmiel, J. Kluth, J. Meier, F. Heinrich, P. Ames, K. Backes, P. Usner,
[23] B.K. Nandi, B. Das, R. Uppaluri, M.K. Purkait, Microfiltration of mosambi juice Comparative study of different MF and UF membranes for drinking water
using low cost ceramic membrane, J. Food Eng. 95 (2009) 597–605. production, Desalination 117 (1998) 189–196.