Arabian Journal of Chemistry (2021) 14, 103231

King Saud University

Arabian Journal of Chemistry

www.ksu.edu.sa
www.sciencedirect.com

ORIGINAL ARTICLE

Invent of a simultaneous adsorption and separation

process based on dynamic membrane for treatment
Zn(II), Ni(II) and, Co(II) industrial wastewater
Ali Aghababai Beni a,*, Akbar Esmaeili b, Yaghoub Behjat c

a
Department of Chemical Engineering, Shahrekord Branch, Islamic Azad University, Shahrekord, Iran
b
Department of Chemical Engineering, North Tehran Branch, Islamic Azad University, PO Box 1651153311, Tehran, Iran
c
Process Development and Equipment Technology Division, Research Institute of Petroleum Industry, Tehran, Iran

Received 9 March 2021; accepted 17 May 2021

Available online 23 May 2021

KEYWORDS Abstract Heavy metal ions wastewater threatens the environment. Removal of heavy metal ions at
Adsorption; low concentrations using bio-sorbents is a new solution, but separating the adsorbent from the fluid
Cs/PVA nanofiber; is also difficult and costly. In this work, a plug reactor equipped with the dynamic membrane (DM)
Sargassum glaucescens was used for the adsorption of Zn (II), Co(II), and Ni(II) ions from industrial wastewater. DM
nanoparticles; module equipped with vibrators for decreased membrane fouling. Central composite design
Hydroxyapatite powder; (CCD) was applied to the experiments design. Brunauer–Emmett–Teller (BET) theory, scanning
Dynamic membrane electron micro-scope (FE-SEM), and Fourier transform infrared spectroscopy (FT-IR) were
employed to characterize and analyze. The Sargassum glaucescens nanoparticles (SGN) and
hydroxyapatite with 50–150 nm and 10–20 lm of particle size and 12.99–172.46 m2 g
1 of specific
surface area were tested as adsorbent. DM support prepared with electrospinning of chitosan
(Cs) and polyvinyl alcohol (PVA) on the polyester web. Cs/PVA nanofibers were cross-linked with
the heat at 110 °C for 10 h. Membrane resistance and index fouling were calculated at different
transmembrane pressure (TMP) and corresponding flux. In this study, the removal efficiencies of
heavy metal ions and turbidity were obtained more than 90%.
Ó 2021 The Authors. Published by Elsevier B.V. on behalf of King Saud University. This is an open access
article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

1. Introduction

In the present century, population growth and meeting various

* Corresponding author. human needs have increased water consumption. Therefore the
E-mail address: [email protected] (A. Aghababai Beni). volume of effluent from various urban and industrial sources
Peer review under responsibility of King Saud University. has increased. One of the most dangerous effluents is heavy
metal wastewater, which has a high potential for contamina-
tion of food sources and can lead to a variety of cancers in
the human body (Xu et al., 2020). There are many reports
Production and hosting by Elsevier

https://doi.org/10.1016/j.arabjc.2021.103231
1878-5352 Ó 2021 The Authors. Published by Elsevier B.V. on behalf of King Saud University.
This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
2 A. Aghababai Beni et al.

on different methods of treatment of heavy metal wastewaters, the use of DM in activated sludge systems in municipal
such as: Coagulation and flocculation, Ion exchange, photo- wastewater treatment, but the DM process for heavy metal
catalysis, reverse osmosis, nanofiltration, microfiltration, ultra- wastewater treatment in a non-living system has not been
filtration, distillation technologies and adsorption (Benjelloun reported. Separation of nanoparticle size adsorbents from
et al., 2021); but traditional methods to treat wastewater with the effluent is a major problem. In this study, SGN and
low concentrations of heavy metal ions are inefficient given the hydroxyapatite powder was used for adsorption and then sep-
cost and waste of operations (Leong and Chang, 2020). In arated with Cs/PVA nanofibers as DM support. This system is
recent years, biosorption has attracted the attention of scien- an innovation because the mixing of adsorbent and effluent is
tists as biosorbent sources are free and have high adsorption done without the use of a mechanical stirrer. In this system,
capacity (Beni and Esmaeili, 2019). Different types of biosor- adsorbent particles are used to form a dynamic membrane.
bents are used for this purpose, for example, Sargasom algae Therefore, separation of adsorbent nanoparticles from the
showed a high capacity to absorb heavy metal ions (Qin treated wastewater is performed at a low cost. A creative idea
et al., 2020) or hydroxyapatite from bone (Al-Ahmed et al., was used in the design of the membrane module, small vibra-
2020). The chemical modifies and geometrical deformation of tors were used around the membrane support to reduce mem-
adsorbents are attractive topics to increase yield and reduce brane fouling and keep the process stable.
operating costs. Preparation of adsorbent at the nano size
can increase adsorption kinetics and adsorption capacity 2. Experiment
(Zhou et al., 2017). For example, the electrospinning technique
provides the nanofibers with high flexibility, stability and selec- 2.1. Material
tivity (Esmaeili and Beni, 2014); such as cellulose acetate/hy-
droxyapatite nanofiber (Hamad et al., 2020) or, b-
Chitosan (average Mw200,000), polyvinyl alcohol (average
cyclodextrin/chitosan/polyvinyl alcohol nanofiber (Fan et al.,
Mw72,000) from Sigma–Aldrich and Acetic acid (glacial)
2019). Despite the development of the nano-sized adsorbents
100% was obtained from Merck. Sargassum glaucescens was
synthesis, the separation of nano-sized adsorbents from the
collected from the Oman Sea on the coast of Chabahar, Iran.
aqueous phase after the adsorption process is difficult and
Nonwoven polyester fabric was purchased from Baftineh, Iran.
costly (Chandrashekhar Nayak et al., 2020). So, dynamic
Bovine bones were prepared and cleaned of meat. Industrial
membranes are an attractive alternative to solid–liquid separa-
wastewater was collected from shahrekord industrial park,
tion (Alibardi et al., 2016a).
Iran. The wastewater details are given in Table 1.
Dynamic membrane (DM) is a new technique in industrial
wastewater treatment (Guan et al., 2018b). In the DM process,
2.2. Preparation of dynamic particle and DM support
a layer of wastewater material (cake layer) is formed on the
support membrane (fabric, filter, stainless steel mesh or non-
woven fibers) (Ersahin et al., 2014). The DM consists of a cake According to Fig. 1a SGN was washed twice with water. It was
layer and a gel layer on top of the cake layer, the structure of dried at 80 °C for 20 h in the oven and crushed with a high-
the gel layer is similar to a normal membrane, which plays an energy planetary ball mill (Fara Pajouhesh, Isfahan FP2, Iran)
important role in DM permeability. The DM process consists at 600 rpm for 2 h. The bovine bones were boiled in water four
of three steps: i) The cake layer formation, ii) Stable filtration times at 120 °C for 3 h, 4 times and all the fat of bovine bones
and, iii) DM cleaning for regeneration. Liu et al. (Liu et al., was removed. The bones were washed and crushed in the range
2009) defined four stages for cake layer formation: substrate of 2–5 mm. Bovine bone was calcined to 850 °Cfor 6 hin the
formation, separation layer formation, fouling layer formation furnace, and then hydroxyapatite was crushed to the particle
and filtration cake formation. Cake layer formation and DM size range of 10–20 lm using a kitchen mill (Pars Khazar,
stability depend on the membrane support and operating GR-1.2.3P, Iran) for 20 min.
conditions. The Cs (5 wt:%) and PVA (10 wt%) solution were dissolved
Membrane support with 100lm of mesh size has higher in acetic acid (2%, v\v) and were stirred for 18 h at 45 °C.

2
outflow flux (up to 1000Lh
1 m ); but, this mesh size increases
the time of the cake layer formation, considerably and the
solid particles leave the membrane system. The stable mode Table 1 Wastewater details at each stage of the process (at
is dependent on the deposition of suspended solids in the fluid. 17.7 °C).
In a stable mode, the flux output is approximately 120– Characteristic Untreated Treated

2
50Lh
1 m and the filtration time could vary from several wastewater wastewater
hours to several days (Alibardi et al., 2016b). One of the obsta- pH 6.362 6
cles to the development of DM is the mass accumulation on Turbidity (NTU) 154.46 13.90
the cake layer and reduction in permeability (Guan et al., Electrical Conductivity 2750 2712
2018a). Membrane fouling is relieved using air and water flow (lscm
1 )
across the membrane support or brush without chemicals Settleable Solids (mgL
1 ) 105 8
(Alibardi et al., 2018). Chemical Oxygen Demand 96 14.5
Low cost, low power consumption and easy cleaning are (mgL
1 )
the advantages of DM because the filtration resistance is less Co(II) (mgL
1 ) 8.41 0.160
than the micro and ultrafiltration, transmembrane pressure Zn(II) (mgL
1 ) 14.72 0.618
(TMP) is low and the DM process can be achieved using the Ni(II) (mgL
1 ) 23.81 0.762
force of gravity (Li et al., 2018). Previous studies have reported
Invent of a simultaneous adsorption and separation process 3

Fig. 1 Schematic of DM preparation steps (a), Schematic of the DM filtration process and DM cake after 60 min (b), the perspective
drawing of membrane module (c) and, strategy of experiments and methodology (d).

Then, Cs and PVA were blended at a ratio of 7:3 on the mag- 2.3. Design process
netic heater stirrer for 24 h at 45 °C. The polyester web was
installed on the collector at 50 °C The Cs/PVA solution was A plug reactor was designed from a plastic pipe with a length
electrospanned on the polyester web at a high DC voltage of of 16 m and an internal diameter of 12 mm. This long pipe was
20 KV, with a tip-collector distance of 17 cm and a flow rate wrapped around a cylinder with a diameter of 100 mm. The
of 0.6 mlh
1 . In order to heat-cross-linking, the Cs/PVA nano- input of the plug reactor was connected to a wastewater tank
fiber membrane was placed in the oven at 110 °C for 10 h. with a pump (Water Pump, 12 Volts). According to Fig. 1b, the
4 A. Aghababai Beni et al.

Fig. 1 (continued)

effluent was mixed with the adsorbent (SGN and hydroxyap- where Wd is the mass of the dried Cs/PVA nanofiber mem-
atite powder) in a circular flow inside the plug reactor. brane and Ww is the wet mass for the swollen Cs/PVA nanofi-
According to Fig. 1c the membrane module made of two ber membrane. The piece of membrane was removed from the
polyethylene pieces. The DM support (Cs/PVA nanofiber @ water and was kept in the laboratory environment for 2 min to
polyester) was installed between polyethylene pieces (DM sup- remove the extra droplets, then; Ww was weighed.
port diameter = 20 mm). Four small vibrators (vibration DC,
5 Volt, 1500 round per minute, Taiwan) were installed in the 2.5. Adsorption and separation experiments
piece. This system was equipped with an air compressor to
fluid flowing through the DM (Positive pressure TMP). A 2000 ml of wastewater including Co(II), Zn(II) and Ni(II) ions
valve was installed at the down of the polyethylene tube for was pumped in the plug reactor. The wastewater and adsor-
rapid evacuation. bents contacted during a circular flow in the plug reactor;
Co(II), Zn(II) and Ni(II) ions were adsorbed onto the hydrox-
2.4. Measurement and methods yapatite powder and SGN. This mixture (wastewater and
adsorbents) was introduced into the tube of DM module, in
The heavy metal ions concentration was determined with atomic several cycles (Each charge = 400 ml). The hydroxyapatite
absorption spectrometry (Varian Atomic Absorption 240, USA). powder and SGN were remained on the DM support and
Turbidity was measured with a turbidity meter (ET266020, Lovi- the DM cake was formed.
bond Corporation, Germany). Nanofiber and nanoparticles mor- Fig. 1d shows the strategy of experiments and methodol-
phology were studied using scanning electron microscope (FEI ogy. Experiments were done in three Sections: i) Adsorption:
Quanta 200 SEM, USA). Functional groups of DM elements optimizing parameters affecting in the adsorption process in
were determined with FT-IR spectra in the range of 400–4000 the plug reactor: The effect of pH 4–9 at SGN = 6 g, hydrox-
cm
1 (Nicolet 8700, Thermo Fisher Scientific, Japan). The struc- yapatite powder = 4 g and, Cs/PVA nanofiber@ polyester
ture of hydroxyapatite was analyzed using X-ray diffraction web = 10 mg with retention time of 40 min. The effect of
(Philips PW3040). The specific surface area of dynamic mem- adsorbent mass SGN = 2–10 g, hydroxyapatite powder = 1–
brane was determined using the Brunauer–Emmert–Teller 5 g (Fig. S4 c) at optimized pH and retention time = 40 min.
method (Costech International Sorptometer 1042, Italy). In order The effect of retention time 10–120 min at optimized adsorbent
to determine the stability of Cs/PVA nanofiber membranes, des- mass and pH. Experiments were performed at constant room
welling experiments were done on the platform shaker. The temperature 28 °C to make the process cost-effective. This sec-
pieces of Cs/PVA nanofiber membrane (10mm  10mm) were tion was independent of the next two section.
immersed in 50 mL of deionized water at 25–45 °C for 24 h ii) Separation: optimizing mass of SGN and hydroxyapatite
Degree of deswelling was calculated from Eq. (1): powder and, interval between two vibration shocks at constant
  TMP (0.2 bar) with standard response surface methodology
Ww
Wd
Deswelling% ¼  100 ð1Þ (RSM) design called CCD at the Design-Expert 11 software.
Wd
Invent of a simultaneous adsorption and separation process 5

Table S1 shows the ranges as well as coded and un-coded levels 2.6. DM resistance and fouling index
of the variables considered in this study.
iii) Optimization of TMP: experiment of filtration at differ- In the cake filtration model, membrane fouling led to the
ent TMP (Fig. 3a) at optimized conditions. deposit of foulant material on the external surface of the mem-
In the output of plug reactor, the removal efficiency of brane, which increased resistance (Choi et al., 2009) DM resis-
heavy metal ions (R%) and adsorption capacity (qe ) were cal- tance was calculated using Eq. (5) (Tong et al., 2011):
culated using Eq. (2) and Eq. (3), respectively:
TMP
R¼  3600 ð5Þ
C0
Ct lJ
R% ¼  100 ð2Þ
C0 where R is the total hydraulic resistances of DM (m
1 ), TMP is
the trans-membrane pressure (Pa), l is water viscosity (Pas),
Jis the instantaneous membrane fluxðm3 m
2 h
1 Þ and, 3600 is
ðC0
Ct ÞV
qe ¼ ð3Þ a conversion factor for converting seconds to hours.
M
Harouna et al. (Mahamadou Harouna et al., 2019) devel-
oped the first order derivative equation by finite difference
where C0 is the initial concentration (mgL
1 ) of heavy metal methods and conducted a direct comparison of the membrane
ions, Ct is any of the heavy-metal ion concentrations fouling under different conditions. The fouling index (m
1 ) was
(mgL
1 ) at arbitrary time t and, M is the adsorbent mass (g) calculated using Eq. (6) (Xiong et al., 2016):
added to V volume of the solution (L). In the DM permeate,
1 dJ
the removal efficiency of turbidity (RT %) was calculated using Fouling index ¼
ð6Þ
Eq. (4): J0 dðVAÞ

where Jand J0 are the filtrate flux and initial flux, respectively
T0
Tt (m3 m
2 h
1 ), VA is the specific throughput (m). Almost all the
RT % ¼  100 ð4Þ
T0 fouling index are being investigated for reverse osmosis, ultra-
filtration and nanofiltration process, but in this research, for
 
where T0 is the effluent turbidity (NTU) of plug reactor out-
DM, Jvs. AV was plotted and slope d dJV was calculated, there-
put, Tt is effluent turbidity (NTU) at arbitrary time t, output ðAÞ
of DM. after, fouling index was obtained.

Fig. 2 Characterization of DM: The SEM image and diameter distribution of Cs/PVA nanofiber membranes (a), SGN (b) and
hydroxyapatite powder (c), FT-IR spectra of Cs/PVA nanofiber membranes, SGN and hydroxyapatite (d) and deswelling test of nanofiber
membranes (e), X-ray diffraction profile of hydroxyapatite (f).
6 A. Aghababai Beni et al.

Fig. 3 TMP profiles (a), Instantaneous membrane flux profiles (b), TMP effect on the removal efficiency of turbidity (e) and DM
resistance profiles (f).

3. Results for 2 h; the specific surface area of SGN with BET analysis
was obtained as 12.99 m2 g
1 (Fig. S1 b) and according to
3.1. Characterization of DM Fig. 2b, this nanoparticle average diameter was observed
95.6 nm. Also, the SEM of hydroxyapatite powder in Fig. 2c
shown, the particle diameter is about 10–20 lm. There are
The SEM of Cs/PVA in Fig. 2a showed that electrospanning
two reasons that SGN are nano-size but hydroxyapatite parti-
produced a homogeneous and microporous nanofiber. The
cles are not nano-size: SGN precipitate in the effluent after a
nanofibers have an average diameter and surface area of
short time; but nano-hydroxyapatite is very stable in the aque-
95.5nmand 172.46 m2 g
1 , respectively (Fig. S1 a). The Cs/
ous phase and does not precipitate easily; also, to create proper
PVA nanofiber membrane put in the oven at 110 °C for 10 h
porosity in the dynamic membrane, the presence of a
to cross-linking and nanofiber membrane stability in the aque-
micrometer-sized particle is necessary to prevent membrane
ous solution was increased. Cross-linking leads to a decrease in
fouling.
polar O
H groups (Fig. S2), and water resistance increases
According to FT-IR results in Fig. 2d, on the hydroxyap-
(Tian et al., 2019). Cross-linked and non-cross-linked mem-
atite spectrum, free O
Hbands are indicative at 3580 cm
1 .
brane stability was compared in Fig. 2e. For non-cross-
linked membranes in the range of 10 –20 h at 25 –45 °C the The absorption bands from 947 to 1080 cm
1 were assigned
degree of deswelling reached 60%, while under this condition, to the phosphate esters stretching R
P with two strong bands
deswelling degree of cross-linked membrane decreased by only at 965 and 1048 cm
1 , while Googerdchian et al.
20%. According to previous results (Esmaeili and Beni, 2014), (Googerdchian et al., 2018) reported that the phosphate is vis-
the high stability degree of the Cs/PVA nanofiber membranes ible from 460 to 1100cm
1 . In the Cs/PVA, SGN and hydrox-
in the aqueous medium with 350% of the average deswelling yapatite spectrums, the vibrational band of from 2900 cm
1
degree indicates that the heat cross-linked is a more suitable was attributed to C
H stretching of the alkyl groups. The
method than cross linking by glutaraldehyde vapor with 93% broad and strong vibration around 3200 to 3400 cm
1 is
of average deswelling degree (Aliabadi et al., 2014). indicative of O
H stretching in H-bonded type of vibration
The SGN was obtained with average size of 50–120 nm and the strong peak at 1390 cm
1 was assigned to the CH2 and
(Fig. 2b). In previous work (Akbar Esmaeili and Aghababai CH3 stretching vibrations (Fan et al., 2019). The SGN for
Beni, 2015), SGN were produced within 1 hin the planetary broad band at 3400 cm
1 was assigned to OAH stretching
ball mill with range of 30–300 nm and a specific surface area vibrations and other at 2900 cm
1 was CAH stretching. The
of 11.25 m2 g
1 , but in this work, planetary ball mill worked stretching vibrations of carboxylate OACAO related to
Invent of a simultaneous adsorption and separation process 7

2889 cm
1. The C‚O signal related to 1566 cm
1 SGN and As shown in Fig. S3d, the adsorption process was very fast.
hydroxyapatite spectrums, respectively. The absorption At 10 min of retention time, the removal efficiency was about
415 cm
1 related to CAOH vibration and OACAO (carboxy- 60%, at 30 and 90 min the removal efficiency was 94% and
late group) stretching vibration for SGN. The bands measured 98%, respectively. The system was received to pseudo-
at 1311, 1080 and 1029 cm
1 related to CACAH and equilibrium after 40 min.
OACAH, to CAO stretching vibrations. The bands CAO
and CAC vibrations of rings. The band 946 cm
1 in SGN indi- 3.3. Separation process in the DM module
cated of bond presence by the CAO stretching vibration
hydroxyapatite spectrums. XRD pattern of hydroxyapatite 3.3.1. Optimization of DM effective factors
was shown in Fig. 2f. It can be seen that the prepared hydrox- The 3D surface response plot was used to investigate the
yapatite belongs to the hexagonal structural type of hydroxya- effects of interaction between the SGN mass, hydroxyapatite
patite with space group P63 =m. According to Fig. S1 c, the powder mass and interval between two vibration shocks on
water contact angle test of Cs/PVA nanofiber @ polyester the removal efficiency of turbidity. The quadratic effects such
web and DM (Cake layer) after separation process shown as the interaction effects betweenX1 , X2 and X3 were consid-
DM is hydrophilic. ered moderate. Fig. S6 a shows the effect of SGN and hydrox-
yapatite mass on the yield of turbidity removal ðY1 Þ at 5 s of
3.2. Adsorption studies in the plug reactor interval between two vibration shocks. DM performance
depends on the external surface of particles, shape and density
Adsorption of heavy metal ions onto adsorbents involves a of particles, solution concentration and quality of the agitator
combination of various mechanisms such as electrostatic (Li et al., 2019). Table 2 compares different DM materials and
adsorption, complexation, ion exchange, covalent forces, yield of DM process. According to the experiment results, the
Van der Waals forces and surface adsorption (Esmaeili and turbidity removal efficiency was 91% at these conditions: 10 g
Aghababai Beni, 2018). The pH is very important in the of SGN and 5 g of hydroxyapatite powder with 5 s of interval
adsorption process with effects on the adsorbents-ion reactions between two vibration shocks. According to Fig. S6 b (at
(Beni, 2021). According to Fig. S3a, at pH = 6 adsorption effi- X1 = 10 g and X2 = 5 g) with increasing the interval between
ciency was more than 95%, while at pH 3 and 9 the adsorption two vibration shocks ðX3 Þ, a yield of turbidity removal ðY1 Þ
efficiency decreases. When the pH of the neutral metal ion was improved because the local shock reduces and DM has
solution was alkaline, the ions were converted to insoluble the opportunity of becoming more stable. Optimization was
hydroxide (Googerdchian et al., 2012). evaluated using software optimization strategy. The software
The stability of the membrane support layer was evaluated suggested 100 optimization strategy, as shown ramp in
at acid and alkaline conditions. The pieces of Cs/PVA nanofi- Fig. S5, first optimization strategy with 0.957 of desirability
ber@ polyester web (DM support) were contacted with a was selected.
200 ml wastewater sample for 40 min at pH 4, 6 and 8.
According to Fig. S3b, at pH = 6 the stability of Cs/PVA 3.3.2. TMP profile and filtration resistance of DM
nanofiber@ polyester was highest; these results showed that The instantaneous membrane flux has a direct relationship
most heavy metal ions are adsorbed by SGN and hydroxyap- with TMP, experiments started using a constant TMP 0.2
atite (DM cake) compared to CS/PVA nanofibers. In the pre- bar, after 4 min, instantaneous membrane flux was 2870
vious study, the optimum pH for the removal of Ni(II) and Co
Lm
2 h
1 and after 8 min; it decreased by 20% of the initial
(II) by SGN and Cs/PVA nanofiber (Esmaeili and Aghababai
flux. According to Fig. 3b, the pattern of instantaneous mem-
Beni, 2018) at pH = 6 was optimized. Googerdchian et al.
brane flux was reduced with quick slope after 8, 12 and 36 min,
(Googerdchian et al., 2018) reported pH = 3 is suitable for
because thickness of the cake layer increased significantly. At
the Pb(II) adsorption with nano-hydroxyapatite. At low pH
constant pressure and with the passage of the time, membrane
the concentration of Hþ ions is high and causes protons to
flux decreases. Almost in all membrane bioreactor was the DM
bind to functional groups in the adsorbent active sites and to
cleaned when flux decreased by 10% relative to the initial flux
interfere with the binding of metal ions and adsorbents (Beni
(Alibardi et al., 2016b). The increase in TMP was done to com-
and Esmaeili, 2020). The adsorbent mass and composition
pensate membrane flux (Tong et al., 2011) in the second series
ratio were tested according to the pattern in Fig. S3c. The
experiments. Increasing the TMP or the DM cleaning affects
results showed that with increasing the adsorbent mass: the
the membrane stability. In line with our results, Saleem et al.
removal efficiency of heavy metal ions increased but the adsor-
(Saleem et al., 2017) showed that almost all NTU values
bent capacity decreased. Previous results showed the SGN
became stable after the first few minutes (7 min); while, in
capacity for Ni(II) and Co(II) adsorption is about 7–9 and
other systems DM can became stable after 1–2 days. Alibardi
2–3 mgg
1 , respectively (Esmaeili and Aghababai Beni, et al. (Alibardi et al., 2016a) showed that the anaerobic
2015). Googerdchian et al. (Googerdchian et al., 2018) synthe- dynamic membrane bioreactor reached a stable mode after
sized nanohydroxyapatite from bovine bone for Pb(II) adsorp- approximately three months; they reported the use of support
tion with 180 mgL
1 of concentration and adsorption capacity mesh with large pore size of 200 lm and stable fluxes at TMP
reported 200mgg
1 . Therefore, by using 6 g of SGN and 5 g values lower than 0.1 bar were made. In this work, the DM
hydroxyapatite powder, more than 90% of the heavy metal reconstruction and membrane development to stable mode
ions were removed. The SGN and hydroxyapatite are widely was rapid. The DM resistance is a function of physical and
used in the wastewater treatment due to their low cost, high chemical properties of the fluid, membrane flux and TMP. In
adsorption capacity and high removal efficiency (Beni and general, increased DM resistance results in increased permeate
Esmaeili, 2020). quality (Saleem et al., 2018).
8 A. Aghababai Beni et al.

Table 2 Comparison of DM Properties.

Ref. Removal efficiency Particles size specific surface area Material DM material
ðlmÞ ðm2 g
1 Þ doseðgL
1 Þ
(Wu et al., 2017) 80% COD 43 1126.06 15 Powder activated
95% Turbidity carbon
(Hu et al., 2017) 88% COD 29.4 1100 3 Powder activated
98% Ammonia carbon
37% Total phosphorus
(Hu et al., 2016) 91% COD 0.4–2000 – 3.9 Activated sludge
83% Turbidity
99% Ammonia
40% Total phosphorus
(Chu et al., 95% COD 11.54–29.76 – 5.56 Activated sludge
2014) 98% Ammonia
100% Total phosphorus
85% Total nitrogen
(Kim et al., 42.6% Algal organic 0.1 37.5 12.5 Iron oxide
2017) matter 1 Goethite
In this work 98.1% Co (II) 95.6nm 12.99 5 SGN
96.8% Ni (II) 10-20lm – 2.5 Hydroxyapatite
95.8% Zn (II) 95.5nm 172.46 – Cs/PVA nanofiber
91% Turbidity

According to Fig. 3c turbidity removal efficiency was mea- Lm
2 h
1 ) (Hu et al., 2016). In this filtration process DM for-
sured at different TMP profile. By appropriate support can be mation was rapid because this DM system was made of non-
improved performance of DM such as DM formation time. living algae nanoparticle without biological metabolisms; also
Time in the DM process is affected to support pore size; for DM support coating with Cs/PVA nanofiber and polyester
example, if pore size of mesh was 10 lm, DM formation time web pore size was improved. So, the pore size of DM supports
would be rapid, but membrane fouling will increase, therefore mesh affected the DM formation, operation stability and
was selected mesh with 25 lm of pore size; DM formed at TMP.
lower than 5 min with membrane flux was acceptable (46–85

Fig. 4 Schematic diagrams of DM filtration in vibrator ‘‘on” and ‘‘off” mode (a), the effect of turning off the vibrator on DM resistance
(b) and removal efficiency of turbidity (c).
Invent of a simultaneous adsorption and separation process 9

In Fig. 3d, four stages are visible based on variations in activating the vibrator makes it impossible to form a stable
membrane resistance: in stage 1, the resistance value is con- DM and turbidity increases in output flux. The adsorption
stant. In stage 2, the resistance increases at low rates. In stage was dependent on the pH. Also, the retention time, dynamic
3, the rate of resistance increases quickly and then at stage 4, particles and vibration shocks optimized.
the resistance to its maximum value remains almost constant.
By increasing the process, time resistance increases because the Declaration of Competing Interest
thickness of the cake layer increases and the membrane flux
decreased. Comparing the TMP profiles in Fig. 3a and mem- The authors declare that they have no known competing
brane resistance profiles in Fig. 3d, it was observed that the financial interests or personal relationships that could have
effect of increase in TMP on the resistance was very high than appeared to influence the work reported in this paper.
passing time. Therefore, after 28 min, a drain valve was rapidly
opened and closed at the bottom of the membrane module; Appendix A. Supplementary material
adsorbent concentrated fluid 300 ml was drained (The DM
cake layer was removed) and the cycle continued, DM resis-
Supplementary data to this article can be found online at
tance decreased from1:097  1011 to 4:696  1010 m
1 : Also, https://doi.org/10.1016/j.arabjc.2021.103231.
an increase in TMP was not necessary with this strategy.
References
3.4. Effect of vibration on DM performance
Al-Ahmed, Z.A., Al-Radadi, N.S., Ahmed, M.K., Shoueir, K., El-
Experimental condition was set in optimal state, DM filtration Kemary, M., 2020. Dye removal, antibacterial properties, and
process was performed for 28 min without vibration and DM morphological behavior of hydroxyapatite doped with Pd ions.
resistance was calculated. When the vibrator turned off, Arab. J. Chem. 13, 8626–8637. https://doi.org/10.1016/j.
according to Fig. 4a-b, the instantaneous membrane flux was arabjc.2020.09.049.
decreased, resistance increased and removal efficiency of tur- Aliabadi, M., Irani, M., Ismaeili, J., Najafzadeh, S., 2014. Design and
bidity was improved because DM was made with dense layers. evaluation of chitosan/hydroxyapatite composite nanofiber mem-
brane for the removal of heavy metal ions from aqueous solution. J.
Roebuck and Tremblay (Roebuck and Tremblay, 2016) synthe-
Taiwan Inst. Chem. Eng. 45, 518–526. https://doi.org/10.1016/j.
sized highly porous 3D network of mono-dispersed particles jtice.2013.04.016.
without compressed; so, the membrane resistance was not Alibardi, L., Bernava, N., Cossu, R., Spagni, A., 2016a. Anaerobic
increased. Turn on vibrator prevented DM layer condensation dynamic membrane bioreactor for wastewater treatment at ambi-
and reduced fouling index from 1.040 to 0.798 m
1 . The cross- ent temperature. Chem. Eng. J. 284, 130–138. https://doi.org/
flow can be used to remove the attached sludge from the sur- 10.1016/j.cej.2015.08.111.
face of DM support (Liu et al., 2009), whereas, in this research, Alibardi, L., Lavagnolo, M.C., Saleem, M., Spagni, A., Bertucco, A.,
the DM backwashed after 56 min filtration with 500 ml water 2018. Assessment of dynamic membrane filtration for biological
at pressure of 12 bar for 30 s (Gasemloo et al., 2019). The treatment of old landfill leachate. J. Environ. Manage. 213, 27–35.
https://doi.org/10.1016/j.jenvman.2018.02.057.
vibrator placed DM particles in the best orientation; the large
Alibardi, L., Spagni, A., Lavagnolo, M.C., Cossu, R., Saleem, M.,
DM particles were placed close to the support layer (Fig. 4c).
2016b. Analysis of fouling development under dynamic membrane
In vibrator ‘‘off” state, DM support layer lost its quality, after filtration operation. Chem. Eng. J. 312, 136–143. https://doi.org/
backwashing because nanoparticles were imprisoned between 10.1016/j.cej.2016.11.123.
polyester base and PVA/Cs nanofibers. It should be noted that Beni, A.A., 2021. Design of a solar reactor for the removal of uranium
a continuous activation of the vibrator was inappropriate from simulated nuclear wastewater with oil-apatite ELM system. Arab.
because it prolongs the formation of DM layer, also the J. Chem. 14,. https://doi.org/10.1016/j.arabjc.2020.102959 102959.
removal efficiency of turbidity reduced severely, so according Beni, A.A., Esmaeili, A., 2020. Biosorption, an efficient method for
to the optimization results, Fig. S5 selected 5 s ‘‘on” and 5 s removing heavy metals from industrial effluents: a review. Environ.
‘‘off” pattern by programmable switch for the vibrator. Technol. Innov. 17,. https://doi.org/10.1016/j.eti.2019.100503
100503.
Beni, A.A., Esmaeili, A., 2019. Design and optimization of a new
4. Conclusion reactor based on biofilm-ceramic for industrial wastewater treat-
ment. Environ. Pollut. 255,. https://doi.org/10.1016/j.en-
Treatment of effluents contaminated with heavy metal ions is vpol.2019.113298 113298.
very important before entering the environment. SGN and Benjelloun, M., Miyah, Y., Akdemir Evrendilek, G., Zerrouq, F.,
hydroxyapatite powder had a high adsorption capacity to Lairini, S., 2021. Recent advances in adsorption kinetic models:
their application to dye types. Arab. J. Chem. 14,. https://doi.org/
adsorb Ni(II), Co(II) and Zn(II) ions from industrial wastew-
10.1016/j.arabjc.2021.103031 103031.
ater. In this study, a new design of a plug reactor was devel-
Chandrashekhar Nayak, M., Isloor, A.M., Inamuddin, Lakshmi, B.,
oped. In this design, the adsorbent particles were mixed with Marwani, H.M., Khan, I., 2020. Polyphenylsulfone/multiwalled
the wastewater effectively and completely. The adsorbent carbon nanotubes mixed ultrafiltration membranes: Fabrication,
was separated from the treated wastewater using the DM characterization and removal of heavy metals Pb2+, Hg2+, and
method. The Cs/PVA nanofiber @polyester web was cross- Cd2+ from aqueous solutions. Arab. J. Chem. 13, 4661–4672.
linked by the thermal method as DM supports. The cake layer https://doi.org/10.1016/j.arabjc.2019.10.007.
formed rapidly on the DM support with adsorbent dynamic Choi, J.S., Hwang, T.M., Lee, S., Hong, S., 2009. A systematic
particles. The results shown, the instantaneous membrane flux approach to determine the fouling index for a RO/NF membrane
is inversely related to the thickness of the cake layer. The back- process. Desalination 238, 117–127. https://doi.org/10.1016/j.
desal.2008.01.042.
washing DM prevented the output flux reducing. Continuously
10 A. Aghababai Beni et al.

Chu, H., Zhang, Y., Zhou, X., Zhao, Y., Dong, B., Zhang, H., 2014. Kim, B.C., Nam, J.W., Kang, K.H., 2017. Dynamic membrane
Dynamic membrane bioreactor for wastewater treatment: Opera- filtration using powdered iron oxide for SWRO pre-treatment
tion, critical flux, and dynamic membrane structure. J. Memb. Sci. during red tide event. J. Memb. Sci. 524, 604–611. https://doi.org/
450, 265–271. https://doi.org/10.1016/j.memsci.2013.08.045. 10.1016/j.memsci.2016.11.081.
Ersahin, M.E., Ozgun, H., Tao, Y., van Lier, J.B., 2014. Applicability Leong, Y.K., Chang, J.-S., 2020. Bioremediation of heavy metals using
of dynamic membrane technology in anaerobic membrane biore- microalgae: recent advances and mechanisms. Bioresour. Technol.
actors. Water Res. 48, 420–429. https://doi.org/10.1016/j. 303,. https://doi.org/10.1016/j.biortech.2020.122886 122886.
watres.2013.09.054. Li, L., Xu, G., Xing, J., 2019. Diatomite enhanced dynamic membrane
Esmaeili, A., Aghababai Beni, A., 2018. Optimization and design of a technology for simultaneous backwash sludge pre-dewatering and
continuous biosorption process using brown algae and chitosan/ backwash wastewater recycling. J. Clean. Prod. 211, 1420–1426.
PVA nano-fiber membrane for removal of nickel by a new https://doi.org/10.1016/j.jclepro.2018.11.283.
biosorbent. Int. J. Environ. Sci. Technol. https://doi.org/10.1007/ Li, L., Xu, G., Yu, H., Xing, J., 2018. Dynamic membrane for micro-
s13762-017-1409-9. particle removal in wastewater treatment: performance and influ-
Esmaeili, Akbar, Aghababai Beni, A., 2015. Novel membrane reactor encing factors. Sci. Total Environ. 627, 332–340. https://doi.org/
design for heavy-metal removal by alginate nanoparticles. J. Ind. 10.1016/j.scitotenv.2018.01.239.
Eng. Chem. https://doi.org/10.1016/j.jiec.2014.11.023. Liu, H., Yang, C., Pu, W., Zhang, J., 2009. Formation mechanism and
Esmaeili, A., Aghababai Beni, A., 2015. Biosorption of nickel and structure of dynamic membrane in the dynamic membrane biore-
cobalt from plant effluent by Sargassum glaucescens nanoparticles actor. Chem. Eng. J. 148, 290–295. https://doi.org/10.1016/j.
at new membrane reactor. Int. J. Environ. Sci. Technol. https://doi. cej.2008.08.043.
org/10.1007/s13762-014-0744-3. Mahamadou Harouna, B., Benkortbi, O., Hanini, S., Amrane, A.,
Esmaeili, A., Beni, A.A., 2014. A novel fixed-bed reactor design 2019. Modeling of transitional pore blockage to cake filtration and
incorporating an electrospun PVA/chitosan nanofiber membrane. modified fouling index – Dynamical surface phenomena in mem-
J. Hazard. Mater. https://doi.org/10.1016/j.jhazmat.2014.08.048. brane filtration. Chem. Eng. Sci. 193, 298–311. https://doi.org/
Fan, J.P., Luo, J.J., Zhang, X.H., Zhen, B., Dong, C.Y., Li, Y.C., 10.1016/j.ces.2018.07.054.
Shen, J., Cheng, Y.T., Chen, H.P., 2019. A novel electrospun Β- Qin, H., Hu, T., Zhai, Y., Lu, N., Aliyeva, J., 2020. The improved
CD/CS/PVA nanofiber membrane for simultaneous and rapid methods of heavy metals removal by biosorbents: a review.
removal of organic micropollutants and heavy metal ions from Environ. Pollut. 258. https://doi.org/10.1016/j.envpol.2019.113777.
water. Chem. Eng. J. 378. https://doi.org/10.1016/j. Roebuck, K., Tremblay, A.Y., 2016. Optimal aggregate size distribu-
cej.2019.122232. tion for the formation of highly efficient nanosheet dynamic
Gasemloo, S., Khosravi, M., Sohrabi, M.R., Dastmalchi, S., Ghar- membranes. J. Memb. Sci. 514, 143–154. https://doi.org/10.1016/j.
bani, P., 2019. Response surface methodology (RSM) modeling to memsci.2016.04.047.
improve removal of Cr (VI) ions from tannery wastewater using Saleem, M., Alibardi, L., Cossu, R., Lavagnolo, M.C., Spagni, A.,
sulfated carboxymethyl cellulose nanofilter. J. Clean. Prod. 208, 2017. Analysis of fouling development under dynamic membrane
736–742. https://doi.org/10.1016/j.jclepro.2018.10.177. filtration operation. Chem. Eng. J. 312, 136–143. https://doi.org/
Googerdchian, F., Moheb, A., Emadi, R., 2012. Lead sorption 10.1016/j.cej.2016.11.123.
properties of nanohydroxyapatite-alginate composite adsorbents. Saleem, M., Lavagnolo, M.C., Concheri, G., Stevanato, P., Squartini,
Chem. Eng. J. 200–202, 471–479. https://doi.org/10.1016/j. A., Spagni, A., 2018. Application of anaerobic dynamic membrane
cej.2012.06.084. bioreactor (AnDMBR) for the successful enrichment of Anammox
Googerdchian, F., Moheb, A., Emadi, R., Asgari, M., 2018. Opti- bacteria using mixed anaerobic and aerobic seed sludge. Bioresour.
mization of Pb(II) ions adsorption on nanohydroxyapatite adsor- Technol. 266, 532–540. https://doi.org/10.1016/j.
bents by applying Taguchi method. J. Hazard. Mater. Elsevier B.V. biortech.2018.06.100.
https://doi.org/10.1016/j.jhazmat.2018.01.056. Tian, H., Yuan, L., Wang, J., Wu, H., Wang, H., Xiang, A., Ashok,
Guan, D., Dai, J., Ahmar Siddiqui, M., Chen, G., 2018a. Comparison B., Rajulu, A.V., 2019. Electrospinning of polyvinyl alcohol into
of different chemical cleaning reagents on fouling recovery in a crosslinked nanofibers: an approach to fabricate functional adsor-
Self-Forming dynamic membrane bioreactor (SFDMBR). Sep. bent for heavy metals. J. Hazard. Mater. 378. https://doi.org/
Purif. Technol. 206, 158–165. https://doi.org/10.1016/j. 10.1016/j.jhazmat.2019.120751.
seppur.2018.05.059. Tong, J., Mai, S., Wang, Z., Wu, Z., Zhang, X., Lu, F., Wei, T., 2011.
Guan, D., Dai, J., Watanabe, Y., Chen, G., 2018b. Changes in the Membrane fouling in an anaerobic dynamic membrane bioreactor
physical properties of the dynamic layer and its correlation with (AnDMBR) for municipal wastewater treatment: characteristics of
permeate quality in a self-forming dynamic membrane bioreactor. membrane foulants and bulk sludge. Process Biochem. 46, 1538–
Water Res. 140, 67–76. https://doi.org/10.1016/j. 1544. https://doi.org/10.1016/j.procbio.2011.04.002.
watres.2018.04.041. Wu, S.E., Hwang, K.J., Cheng, T.W., Lin, Y.C., Tung, K.L., 2017.
Hamad, A.A., Hassouna, M.S., Shalaby, T.I., Elkady, M.F., Abd Dynamic membranes of powder-activated carbon for removing
Elkawi, M.A., Hamad, H.A., 2020. Electrospun cellulose acetate microbes and organic matter from seawater. J. Memb. Sci. 541,
nanofiber incorporated with hydroxyapatite for removal of heavy 189–197. https://doi.org/10.1016/j.memsci.2017.07.006.
metals. Int. J. Biol. Macromol. 151, 1299–1313. https://doi.org/ Xiong, B., Zydney, A.L., Kumar, M., 2016. Fouling of microfiltration
10.1016/j.ijbiomac.2019.10.176. membranes by flowback and produced waters from the Marcellus
Hu, Y., Wang, X.C., Tian, W., Ngo, H.H., Chen, R., 2016. Towards shale gas play. Water Res. 99, 162–170. https://doi.org/10.1016/j.
stable operation of a dynamic membrane bioreactor (DMBR): watres.2016.04.049.
operational process, behavior and retention effect of dynamic Xu, Y., Wei, Y., Long, T., Wang, R., Li, Z., Yu, C., Wu, T., He, M.,
membrane. J. Memb. Sci. 498, 20–29. https://doi.org/10.1016/j. 2020. Association between urinary metals levels and metabolic
memsci.2015.10.009. phenotypes in overweight and obese individuals. Chemosphere 254.
Hu, Y., Yang, Y., Wang, X.C., Hao Ngo, H., Sun, Q., Li, S., Tang, J., https://doi.org/10.1016/j.chemosphere.2020.126763.
Yu, Z., 2017. Effects of powdered activated carbon addition on Zhou, K., Liu, Y., Yang, Z., Liu, H., 2017. Biosorption of U(VI) by
filtration performance and dynamic membrane layer properties in a modified Hottentot Fern: kinetics and equilibrium studies. J.
hybrid DMBR process. Chem. Eng. J. 327, 39–50. https://doi.org/ Environ. Radioact. 167, 13–19. https://doi.org/10.1016/
10.1016/j.cej.2017.06.072. j.jenvrad.2016.11.012.