You are on page 1of 6

Separation and Purification Technology xxx (2016) xxx–xxx

Contents lists available at ScienceDirect

Separation and Purification Technology


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

Microfiltration of microalgae in the presence of rigid particles


M.T. Hung, J.C. Liu ⇑
Department of Chemical Engineering, National Taiwan University of Science and Technology, 43 Keelung Road, Section 4, Taipei 106, Taiwan

a r t i c l e i n f o a b s t r a c t

Article history: This study investigated the microfiltration of microalgae, Chlorella sp., in the presence of polymethyl-
Received 12 July 2016 methacrylate (PMMA) under dead-end and cross-flow modes. The concentration of microalgae was
Received in revised form 21 October 2016 10 mg/L, while PMMA concentration varied from 10 to 40 mg/L. Flux increased with increasing concen-
Accepted 31 October 2016
tration of PMMA in dead-end microfiltration, and the specific cake resistance decreased as PMMA con-
Available online xxxx
centration increased under three different trans-membrane pressure (TMP) of 20, 40, and 60 kPa. The
cake compressibility decreased in the presence of PMMA. In cross-flow microfiltration of microalgae in
Keywords:
the presence of PMMA, the flux increased as cross-flow velocity increased, and the effect became more
Compression
Harvesting
pronounced as PMMA concentration increased. It was because the rigid PMMA particles rendered the dis-
Microalgae tribution of local specific resistance more uniform along the cake thickness. In addition, the cake growth
Microfiltration became more containable by increasing cross-flow velocity when more PMMA particles were present,
Water treatment and resulted in higher steady flux under both laminar and turbulent flow conditions.
Ó 2016 Elsevier B.V. All rights reserved.

1. Introduction in severe fouling in cross-filtration of Chlorella sorokiniana [6]. In


microfiltration of algal suspension, EOM significantly increased
Microalgae in excess amount in surface water are known to the resistance [8]. Hung and Liu [9] pointed out that the dissolved
cause trouble in sedimentation basin and shorten filter run in rapid polysaccharide-like substances derived from EOM cause remark-
sand filtration unit in water treatment plants [1]. Cyanotoxins, able increase in fouling resistance of the hydrophobic membrane
such as microcystin, pose risks to human health and its removal via adsorption on membrane surfaces. Characteristics of AOM as
is challenging although various technologies have been examined affected by different nutrient conditions and subsequent influences
[2]. Furthermore, microalgae also lead to taste, odors, formation on microfiltration membrane fouling have been investigated [10].
of disinfection by-products, and higher coagulant demand by the It is noted that microalgae may not the only fine particles in
presence of algogenic organic matter (AOM) [3]. Unfortunately, nature water. There are plenty of inorganic particles present in
the conventional water treatment processes are not adequate for water, such as clay. Contrary to microalgae that are compressible,
the removal of algal cells, while additional unit may be costly clays are rigid and incompressible in terms of their structures.
[4]. The use of membrane-based technologies in water treatment The filtration of microalgae in the presence of incompressible par-
have been shown to be reliable and capable of producing high- ticles needs to be investigated from water treatment viewpoint.
quality water. Understanding the influence of rigid particles on the microfiltra-
Microalgae have gained enormous attention due to their poten- tion of microalgae will be beneficial for developments in water
tial for various applications, such as biofuel and wastewater treat- treatment technology and microalgae harvesting method. In the
ment [5]. And membrane filtration is considered as an alternative current study, the filtration of microalgae in the presence of rigid
method for microalgae harvesting [6]. However, membrane fouling polymethylmethacrylate (PMMA) particles at mass ratio of
is an important issue encountered in the membrane processes for microalgae to PMMA at 1:0, 1:1, 1:2, and 1:4, were examined,
water treatment as well as for algae harvesting, and extensive respectively. The main reason for choosing PMMA as surrogate of
research has been carried out to elucidate the fouling mechanisms. clay is because, similar to clay, it is negatively charged in a wide
Both algae size and extracellular organic matter (EOM) are critical pH range. Its possible electrostatic interaction with microalgae will
factors in membrane fouling [7]. Deposition of microalgae started thus be similar to that of clay. Both dead-end and cross-flow micro-
to occur even at a very low permeate and would eventually result filtration were conducted at different trans-membrane pressure
(TMP) and cross-flow velocity. The specific cake resistance and
⇑ Corresponding author. cake compressibility were assessed. Furthermore, the resistances-
E-mail address: liu1958@mail.ntust.edu.tw (J.C. Liu).

http://dx.doi.org/10.1016/j.seppur.2016.10.063
1383-5866/Ó 2016 Elsevier B.V. All rights reserved.

Please cite this article in press as: M.T. Hung, J.C. Liu, Microfiltration of microalgae in the presence of rigid particles, Separ. Purif. Technol. (2016), http://dx.
doi.org/10.1016/j.seppur.2016.10.063
2 M.T. Hung, J.C. Liu / Separation and Purification Technology xxx (2016) xxx–xxx

in-series model was used to assess the fouling mechanisms during measured with a small-angle light scattering instrument (Malvern,
microfiltration. Mastersizer 2000) as shown in Fig. 1. The hydrophilic mixed cellu-
lose ester membrane (Corning, Membra-Fil) with a nominal pore
size of 0.22 lm was utilized in this study.
2. Materials and methods

2.1. Materials 2.2. Apparatus and procedures

The microalgae species, Chlorella sp., was cultivated in the lab- 2.2.1. Dead-end microfiltration
oratory. The procedure for microalgae cultivation and the subse- The experimental apparatus for dead-end microfiltration is
quent cleaning are described in our previous work [9]. The mean shown in Fig. 2. The 2-L microalgae suspension was poured into
diameter of Chlorella sp. is 3.8 lm. The commercial PMMA particles the tank, which was kept at 25 °C throughout the process. The
(Soken, MX-300) with a mean diameter of 3.68 lm were used to membrane module made of acrylic with the effective filtration area
mimic incompressible particulates in natural water. The stock sus- of 4 cm2 was submerged into suspension. As filtrate began to flow
pension of PMMA (1000 mg/L) was first prepared using 1-L ultra- into the receiver while the vacuum pump (Gast, DOA-P104-AA)
pure water. After stirring for 24 h, the suspension was filtered was switched on to create negative pressure. Trans-membrane
with a 0.45 lm GF/C membrane. The retained PMMA particles pressure (TMP) for dead-end microfiltration was set at 20, 40,
were again suspended with 1-L ultrapure water. Aforementioned and 60 kPa, respectively. The weight of filtrate was recorded as a
procedures were repeated twice to clean PMMA particles. The con- function of operation time by the personal computer that acquired
centration of microalgae was controlled at 10 mg/L for all experi- data from the electronic balance (Ohaus, adventurer). The proce-
ments with PMMA concentration of 0, 10, 20, 40 mg/L, dure for the classification of hydrodynamic resistances, such as
respectively. The size distribution of microalgae with PMMA was membrane resistance (Rm,b), fouling resistance (Rm,f), cake resis-
tance (Rc), and total resistance (Rt) is illustrated in the previous
work [9].

2.2.2. Cross-flow microfiltration


The equipment of cross-flow microfiltration used in the current
study was described in our earlier work [9]. It was a customer-
made plate and sheet type module which measures 4 cm in length,
1 cm in width, and 2 mm in height. The 2-L microalgae suspension
was first placed in the thermostatic tank. The peristaltic pump
(Cole-Parmer, Masterflex) not only accounted for transportation
of suspension but also provided trans-membrane pressure (TMP)
as the driving force for filtration. The rotameter (Aalborg, P11A2)
prior to filter module was used to monitor cross flow velocity,
which was controlled at 0.43 and 1.11 m/s, respectively, in order
to assess laminar and turbulent flow conditions. Two pressure
gauges were used to examine the TMP, which was kept at 20, 40
and 60 kPa, respectively. The technique used for recording perme-
ate weight as a function of time was the same as that for dead-end
microfiltration. The desired level of the overall particle concentra-
tion was monitored using turbidimeter (Merek, Turbiquant-1500T)
Fig. 1. Size distribution of microalgae with various concentration of PMMA. and kept via supplementing ultrapure water, so that the solid

Fig. 2. Equipment of dead-end microfiltration system.

Please cite this article in press as: M.T. Hung, J.C. Liu, Microfiltration of microalgae in the presence of rigid particles, Separ. Purif. Technol. (2016), http://dx.
doi.org/10.1016/j.seppur.2016.10.063
M.T. Hung, J.C. Liu / Separation and Purification Technology xxx (2016) xxx–xxx 3

loading of suspension was constant throughout the filtration


experiment.

2.3. Determination of the specific cake resistance and the cake


compressibility

The specific cake resistance, a, is a parameter associated with


the cake porosity in membrane filtration. Cake resistance is deter-
mined by the following equation:
Rc ¼ a  m ð1Þ
where Rc is the cake resistance (1/m), a the specific cake resistance
(m/kg), and m the cake mass (kg/m2). The method for determining a
is as follows [11]:
t lRm alcv
¼ þ ð2Þ
v DP 2 DP
where t is the operating time (s), v the permeate volume (m3/m2), l
the viscosity of permeate (kg s1 m), DP the TMP (N m2), and c the
particle concentration (kg/m3). Therefore, a plot of t/v versus v
should yield a straight line for the duration under cake formation.
From the slope of the line, the specific cake resistance (a) can be cal-
culated. The relationship between the specific cake filtration resis-
tance, a, and the TMP can be expressed as the following empirical
equation [12]:

a ¼ A 1  DP n ð3Þ
where A1 is the empirical coefficient (m/kg) and n the compressibil-
ity coefficient of cake.

3. Results and discussion

3.1. Dead-end microfiltration

The effect of TMP on flux of dead-end microfiltration was first


examined with the algal concentration of 10 mg/L. The initial flux
was 7.25  104, 15.70  104, and 21.00  104 m3/m2 s, respec-
tively, under TMP of 20, 40, and 60 kPa. The flux at the end of
experiment dropped to 1.78  104, 2.28  104, and
4 3 2
2.36  10 m /m s, respectively. It was found that the initial flux
increased with increasing TMP; however, more drastic flux decline
was observed when under higher TMP. The flux data can be ana-
lyzed by plotting dt/dv versus v, which are shown in Fig. 3. Three

Fig. 4. Filtration curves of the plot of dt/dv versus v of 10 mg/L of microalgae with
various concentration of PMMA under TMP of (a) 20 kPa; (b) 40 kPa; and (c) 60 kPa.

filtration curves all show a concave upward profile, implying the


transition of fouling mechanism from pore blocking to cake forma-
tion during microfiltration [13]. It could also result from the defor-
mation of the particles and the compression of the filter cake,
which exhibits viscoelastic behavior and its properties vary with
Fig. 3. Filtration curves of the plot of dt/dv versus v of 10 mg/L of microalgae under the solid compressive pressure and the compression time [14,15].
three TMPs.

Please cite this article in press as: M.T. Hung, J.C. Liu, Microfiltration of microalgae in the presence of rigid particles, Separ. Purif. Technol. (2016), http://dx.
doi.org/10.1016/j.seppur.2016.10.063
4 M.T. Hung, J.C. Liu / Separation and Purification Technology xxx (2016) xxx–xxx

Table 1
The classification of hydrodynamic resistance as affected by PMMA concentration under three TMPs.

PMMA (mg/L) Rm,b (1/m) Rm,f (1/m) Rc (1/m) Rt (1/m)


TMP: 20 kPa
0 3.17E+10 (28.56%) 2.50E+09 (2.25%) 7.68E+10 (69.19%) 1.11E+11 (100%)
10 2.97E+10 (35.48%) 4.10E+09 (4.9%) 4.99E+10 (59.62%) 8.37E+10 (100%)
20 3.13E+10 (36.06%) 3.50E+09 (4.03%) 5.20E+10 (59.91%) 8.68E+10 (100%)
40 3.16E+10 (30.38%) 2.40E+09 (2.31%) 7.00E+10 (67.31%) 1.04E+11 (100%)
TMP: 40 kPa
0 2.63E+10 (14.86%) 4.10E+09 (2.32%) 1.47E+11 (82.82%) 1.77E+11 (100%)
10 2.90E+10 (17.90%) 4.80E+09 (2.96%) 1.28E+11 (79.14%) 1.62E+11 (100%)
20 3.06E+10 (24.68%) 2.10E+09 (1.96%) 9.13E+10 (73.63%) 1.24E+11 (100%)
40 2.86E+10 (25.09%) 1.80E+09 (1.58%) 8.36E+10 (73.33%) 1.14E+11 (100%)
TMP: 60 kPa
0 2.99E+10 (11.77%) 3.00E+09 (1.18%) 2.21E+11 (87.05%) 2.54E+11 (100%)
10 3.04E+10 (15.75%) 5.70E+09 (2.95%) 1.57E+11 (81.30%) 1.93E+11 (100%)
20 3.17E+10 (17.51%) 2.60E+09 (1.44%) 1.47E+11 (81.05%) 1.81E+11 (100%)
40 3.24E+10 (19.29%) 4.10E+09 (2.44%) 1.36E+11 (80.71%) 1.68E+11 (100%)

A greater curvature of filtration curve represents that cake is sub- Table 2


jected to higher compression [16]. Specific cake resistance, cake mass, and cake resistance as affected by PMMA
concentration under three TMPs.
The filtration of microalgae of 10 mg/L in the presence of vari-
ous concentration of PMMA under TMP of 20 kPa was examined. PMMA Specific cake Cake mass Cake resistance
It could clearly be seen that flux significantly increased by the (mg/L) resistance (m/kg) (g/m2) (1/m)
addition of PMMA particles into the microalgae. Fig. 4a shows TMP: 20 kPa
the curves of dt/dv versus v. The curvature of filtration continu- 0 5.63E+12 13.6 7.68E+10
ously decreased with increasing PMMA concentration and eventu- 10 2.29E+12 21.8 4.99E+10
20 1.33E+12 39.1 5.20E+10
ally reached zero. In other words, the filtration profile became a 40 9.89E+11 70.8 7.00E+10
straight line. It implied that the presence of PMMA particles in
TMP: 40 kPa
microalgae effectively suppressed cake compression in microfiltra- 0 1.06E+13 14.4 1.47E+11
tion. Notably, a straight line in the plot of dt/dv versus v does not 10 4.21E+12 30.4 1.28E+11
necessarily represent that the corresponding cake is totally incom- 20 1.68E+12 54.3 9.13E+10
pressible [16]. It is because of the deformation of a solid network 40 9.09E+11 92.0 8.36E+10
(i.e. the rearrangement of particle packing) makes the cake slightly TMP: 60 kPa
compressible. Somehow, the filtration curve of PMMA microfiltra- 0 1.37E+13 16.1 2.21E+11

tion in the plot of dt/dv versus v shows a prefect linear profile (data
10 3.62E+12 43.3 1.57E+11
20 1.91E+12 76.7 1.47E+11
not shown). Unlike the experiments under TMP of 20 kPa, flux 40 1.11E+12 118.6 1.36E+11
increased with increasing PMMA concentration when under TMP
of 40 kPa (Fig. 4b). The curvature of filtration curve decreased at
PMMA concentration of 10 mg/L. Furthermore, the filtration curve helped to counteract cake compression. However, it also led to
exhibited a linear profile when the PMMA concentration increased the increase in cake mass. Basically, cake resistance is determined
to 20 and 40 mg/L. Fig. 4c shows the results under TMP of 60 kPa. by the product of cake mass and specific cake resistance. Table 2
The effects of PMMA particles on flux and curvature of filtration shows how cake resistance was simultaneously affected by the
curve under TMP of 60 kPa were identical to those observed under specific cake resistance and the cake mass. When under TMP of
TMP of 40 kPa. 20 kPa, cake resistance decreased in the presence of 10 mg/L of
Table 1 shows the classification of hydrodynamic resistances PMMA. It was because the decrease in specific cake resistance
(Rm,b, Rm,f, Rc) as affected by PMMA concentration under three was more significant than the increase of the cake mass in con-
TMPs. It is obvious that cake resistance (Rc) accounted for the tributing to cake resistance. Nevertheless, the increase of the cake
majority of total resistance among all dead-end microfiltration. mass became more dominant in determining cake resistance when
The fouling resistance (Rm,f) was not affected by the concentration 20 mg/L or more PMMA were involved. Therefore, the cake resis-
of PMMA, while the Rm,f always contributed the least to total resis- tance began to increase as long as the concentration of PMMA
tance. The presence of PMMA particles did not inflict additional was higher than 10 mg/L when under TMP of 20 kPa. On the con-
fouling resistance on membrane even though some of PMMA par- trary, the flux gradually increased as PMMA concentration
ticles were just slightly greater than membrane pores. This result increased from 10 to 40 mg/L when under TMP of 40 and 60 kPa,
further verified that all concave upward profiles in the plot of dt/ respectively. It is probably because the specific cake resistance
dv versus v were mainly due to cake compression instead of pore played a dominant role in determining the cake resistance rather
blockage. Table 2 shows that the specific cake resistance decreased than the cake mass when under higher TMP. Table 3 shows the
as PMMA concentration increased under three TMPs. It implied
that the packing of PMMA particles gained more space for filtrate
to pass through the cake layer, which rendered the specific cake Table 3
resistance lower. It implied that the presence of clay particles, rigid Cake compressibility index (n) as affected by mass ratio of microalgae to PMMA.

and negatively charged just like PMMA, would benefit filtration Mass ratio of microalgae to PMMA Compressibility index (–)
separation of microalgae. However, the PMMA used in the current 1:0 0.82
study was mono-dispersed with same size of microalgae, while 1:1 0.47
clay particles in natural water are not. Further study is needed to 1:2 0.33
verify this. In general, the presence of PMMA in the microalgae 1:4 0.08

Please cite this article in press as: M.T. Hung, J.C. Liu, Microfiltration of microalgae in the presence of rigid particles, Separ. Purif. Technol. (2016), http://dx.
doi.org/10.1016/j.seppur.2016.10.063
M.T. Hung, J.C. Liu / Separation and Purification Technology xxx (2016) xxx–xxx 5

cake compressibility as affected by the mass ratio of algae to


PMMA. The cake compressibility dramatically decreased with
increasing PMMA concentration. The role of PMMA in microalgae
filtration is similar to body feed filtration, in which cellulose or
diatomaceous earth were added to enhance separation of microbial
suspension [17].

3.2. Cross-flow microfiltration

Cross-flow microfiltration of the microalgae affected by TMP


under laminar and turbulent flow condition was examined. The
cross-flow velocity was controlled at 0.43 and 1.11 m/s, so that
the corresponding Reynolds number was 1583 and 4115, respec-
tively. When under laminar flow condition, the initial flux was
4.29  104, 7.96  104, and 13.20  104 m3/m2 s, respectively,
under TMP of 20, 40, and 60 kPa. The flux at end of experiment
dropped to 1.15  104, 1.25  104, and 1.25  104 m3/m2 s,
respectively. When under turbulent flow condition, the initial flux
was 8.85  104, 12.63  104, and 17.60  104 m3/m2 s, respec- Fig. 5. Filtration curves of the plot of dt/dv versus v of 10 mg/L of microalgae at
tively, under TMP of 20, 40, and 60 kPa. The flux at end of experi- various concentration of PMMA under laminar flow condition.
ment decreased to 1.09  104, 1.22  104, and 1.30  104 m3/
m2 s, respectively. It can be seen that the initial flux increased with
increasing cross-flow velocity, which is due to the increased drag 10 mg/L of PMMA, and it showed a linear profile as PMMA concen-
force exerted on the particles [18]. However, the flux at the end tration increased to 20 mg/L. The filtration curve became concave
of experiment under turbulent flow condition was lower than downward and gradually leveled off at the late period at 40 mg/L
those under laminar flow condition under identical TMP, probably of PMMA. Table 4 shows that the fouling resistance (Rm,f) was
because the algal cake was subjected to more significant compres- not significantly influenced by PMMA. For cross-flow microfiltra-
sion due to higher drag force when under turbulent flow condition. tion, cake formation still contributed the dominant resistance. It
In our pervious study, increasing cross-flow velocity from 0.43 to is known that the distribution of specific filtration resistance in a
0.84 m/s resulted in insignificant improvement of flux in the cake formed by deformable particles is extremely non-uniform
microfiltration of algal suspension [9]. along cake height [14,16,18]. In contrast, the cake composed of
The cake compressibility as affected by the presence PMMA was PMMA particles had more uniform distribution of specific filtration
evaluated at TMP of 40 kPa. When under laminar flow condition, resistance. The more PMMA particles present, the more uniform
the initial flux was 7.88  104, 12.50  104, 11.85  104, and the distribution of a in the algal cake would be. As a result, the
12.00  104 m3/m2 s, respectively at PMMA concentration of 0, restraint to cake growth could prevent flux from further decay at
10, 20, and 40 mg/L. Table 4 shows the classification of hydrody- the late period of filtration. The initial flux under turbulent flow
namic resistances as affected by PMMA concentration for cross- condition was 14.30  104, 15.15  104, 13.85  104, and
flow microfiltration under laminar and turbulent flow conditions. 13.20  104 m3/m2 s, respectively, at PMMA concentration of 0,
It is clearly seen that the membrane resistance (Rm,b) of the exper- 10, 20, and 40 mg/L. The flux at the end of experiment was
iment without PMMA under laminar flow was abnormally high, 1.21  104, 1.85  104, 3.04  104, and 3.45  104 m3/m2 s,
which resulted in the decrease in the initial flux. The flux at the respectively, at PMMA concentration of 0, 10, 20, and 40 mg/L.
end of experiment was 1.25  104, 1.85  104, 1.83  104, and The increase in PMMA concentration did not lead to the decrease
2.26  104 m3/m2 s, respectively at PMMA concentration of 0, in the initial flux. Reducing cake compression by adding PMMA
10, 20, and 40 mg/L. Obviously, the presence of PMMA particles particles was considered as the compensation to the increase of
enhanced cross-flow microfiltration. Although the difference cake resistance by increased cake mass. Therefore, the initial flux
between fluxes at the end of experiments with PMMA concentra- was not affected by the overall solid loading of feed concentration.
tion of 10 and 20 mg/L was insignificant, the flux trajectory showed Limited by the length allowed, the role of EOM was not covered in
less decay at late period at 20 mg/L of PMMA. Furthermore, the flux the current work. However, judging from the insignificant fraction
profile at late period of the experiment at PMMA of 40 mg/L grad- of fouling resistance (Rm,f), it was probable that EOM did not con-
ually leveled off, which could be approximated as the steady-state tribute much to total resistance. The role of EOM in microfiltration
flux. Fig. 5 shows the curves of dt/dv versus v. Similar to dead-end of microalgae in the presence of rigid particles certainly deserved
microfiltration, the curvature of filtration curve decreased at further study. Fig. 6 depicts the filtration curve in the plot of

Table 4
The classification of hydrodynamic resistance as affected by PMMA concentration under laminar and turbulent flow conditions.

PMMA (mg/L) Rm,b (1/m) Rm,f (1/m) Rc (1/m) Rt (1/m)


Cross-flow velocity: 0.43 m/s (laminar flow)
0 4.56E+10 (14.3%) 2.70E+09 (0.8%) 2.71E+11 (84.7%) 3.20E+11 (100%)
10 2.56E+10 (11.7%) 1.30E+09 (0.6%) 1.89E+11 (87.5%) 2.16E+11 (100%)
20 2.99E+10 (13.7%) 1.90E+09 (0.8%) 1.87E+11 (85.5%) 2.19E+11 (100%)
40 2.59E+10 (14.6%) 2.00E+09 (1.1%) 1.49E+11 (84.2%) 1.77E+11 (100%)
Cross-flow velocity: 1.11 m/s (turbulent flow)
0 2.78E+10 (8.4%) 2.20E+09 (0.7%) 3.01E+11 (90.9%) 3.31E+11 (100%)
10 2.46E+10 (11.4%) 3.50E+09 (1.6%) 1.88E+11 (87.0%) 2.16E+11 (100%)
20 2.84E+10 (21.5%) 2.70E+09 (2.0%) 1.01E+11 (76.5%) 1.32E+11 (100%)
40 2.60E+10 (22.4%) 3.50E+09 (3.0%) 8.65E+10 (74.6%) 1.16E+11 (100%)

Please cite this article in press as: M.T. Hung, J.C. Liu, Microfiltration of microalgae in the presence of rigid particles, Separ. Purif. Technol. (2016), http://dx.
doi.org/10.1016/j.seppur.2016.10.063
6 M.T. Hung, J.C. Liu / Separation and Purification Technology xxx (2016) xxx–xxx

faces. The presence of PMMA particles in microalgae suspension


significantly enhanced the microfiltration performance. It
increased the cake porosity and reduced the cake compressibility.
For dead-end microfiltration, more PMMA particles made cake
more resistant to compression, thus higher flux could be attained
by increasing TMP. For cross-flow microfiltration, the distribution
of local specific cake resistance along cake height became more
uniform when PMMA particles were present. Therefore, the flux
became higher when increasing cross-flow, which effectively
reduced the cake thickness. Furthermore, steady state flux was
obtained at low concentration of PMMA particles when under tur-
bulent flow condition. As long as PMMA particles were involved,
the cake development at the late period of experiment was more
readily limited by applying higher shear stress to cake surface.

References

[1] G. Joh, Y.S. Choi, J.K. Shin, J. Lee, Problematic algae in the sedimentation and
filtration process of water treatment plants, J. Water SRT-Aqua 2011 (60)
(2011) 219–230.
Fig. 6. Filtration curves of the plot of dt/dv versus v of 10 mg/L of microalgae at [2] J.A. Westrick, D.C. Szlag, B.J. Southwell, J. Sinclair, A review of cyanobacteria
various concentration of PMMA under turbulent flow condition. and cyanotoxins removal/inactivation in drinking water treatment, Anal.
Bioanal. Chem. 397 (2010) 1705–1712.
[3] M. Pivokonsky, J. Naceradska, I. Kopecka, M. Baresova, B. Jefferson, X. Li, R.K.
dt/dv versus v. The filtration curve was approximately linear at Henderson, The impact of algogenic organic matter on water treatment plant
operation and water quality: a review, Crit. Rev. Environ. Sci. Technol. 2016
10 mg/L of PMMA. It then became concave downward as PMMA (46) (2016) 291–335.
concentration increased to 20 and 40 mg/L. It could be reasoned [4] C.R. Dunlap, K.S. Sklenar, L. Blake, A costly endeavor: addressing algae
that steady state flux was more readily obtained when under problems in a water supply, J. Am. Water Works Assoc. 107 (2015) 255–262.
[5] M.R. Bilad, H.A. Arafat, I.F.J. Vankelecom, Membrane technology in microalgae
higher cross-flow velocity and in the presence of PMMA. cultivation and harvesting: a review, Biotechnol. Adv. 32 (2014) 1283–1300.
The reduction in cake compressibility when in the presence of [6] F. Wicaksana, A.G. Fane, P. Pongpairoj, R. Field, Microfiltration of algae
10 mg/L PMMA particles seemed more significant under turbulent (Chlorella sorokiniana): critical flux, fouling and transmission, J. Membr. Sci.
387–388 (2012) 83–92.
flow than that under laminar flow, judging by the variation in the
[7] K.J. Hwang, S.J. Lin, Filtration flux-shear stress-cake mass relationships in
curvature of filtration curves. The particle size distribution, shown microalgae rotating-disk dynamic microfiltration, Chem. Eng. J. 244 (2014)
in Fig. 1, indicated that a certain amount of PMMA particles were 429–437.
submicron, ranging between 0.28 and 0.99 lm. Hwang et al. [19] [8] S. Babel, S. Takizawa, Microfiltration membrane fouling and cake behavior
during algal filtration, Desalination 261 (2010) 46–51.
reveal that the deposition probability decreases rapidly as particle [9] M.T. Hung, J.C. Liu, Microfiltration for separation of green algae from water,
size increases since the lift force causes particles to depart from the Colloids Surf. B: Biointerfaces 51 (2006) 157–164.
contact points. This means that more submicron PMMA particles [10] W. Huang, H. Chu, B. Dong, Characteristics of algogenic organic matter
generated under different nutrient conditions and subsequent impact on
would deposit on membrane surface when increasing cross-flow microfiltration membrane fouling, Desalination 293 (2012) 104–111.
velocity. Consequently, the cake compression was more effectively [11] R. Rautenbach, R. Albrecht, On the behavior of asymmetric membranes in
suppressed since more PMMA particles were present in cake depo- pervaporation, J. Membr. Sci. 19 (1984) 1–22.
[12] F.M. Tiller, O.V. Anantharamakrishnan, Filtration with variable slurry
sition when increasing cross-flow velocity. concentration, J. Chem. Eng. Jpn. 13 (1980) 380–385.
In summary, the presence of rigid PMMA was beneficial to [13] K.J. Hwang, C.Y. Liao, K.L. Tung, Analysis of particle fouling during
microfiltration of microalgae in both dead-end and cross-flow microfiltration by use of blocking models, J. Membr. Sci. 287 (2007) 287–293.
[14] K.J. Hwang, C.L. Hsueh, Dynamic analysis of cake properties in microfiltration
mode. Specific cake resistance and cake compressibility decreased of soft colloids, J. Membr. Sci. 214 (2003) 259–273.
significantly in the presence of PMMA in dead-end microfiltration. [15] K.J. Hwang, Y.T. Wang, E. Iritani, N. Katagiri, Effects of porous gel particle
Cake resistance was the dominant one in cross-flow microfiltra- compression properties on microfiltration characteristics, J. Membr. Sci. 341
(2009) 286–293.
tion. Increasing cross-flow velocity resulted in more pronounced
[16] W.M. Lu, K.L. Tung, S.M. Hung, J.S. Shiau, K.J. Hwang, Constant pressure
improvement in flux, since distribution of local specific resistance filtration of mono-dispersed deformable particle slurry, Sep. Sci. Technol. 36
became more uniform. (2001) 2355–2383.
[17] K. O’Mahony, R. Freitag, B. Dhote, P. Muller, I. Schumacher, Capture of bacteria
from fermentation broth by body feed filtration: a solved problem?, Biotech
4. Conclusions Prog. 2 (2006) 471–483.
[18] W.M. Lu, K.L. Tung, C.H. Pan, K.J. Hwang, Cross-flow microfiltration of mono-
dispersed deformable particle suspension, J. Membr. Sci. 198 (2002) 225–243.
The microfiltration of microalgae yielded very low flux in both [19] K.J. Hwang, Y.L. Hsu, K.L. Tung, Effect of particle size on the performance of
dead-end and cross-flow mode due to serious cake compression, cross-flow microfiltration, Adv. Powder Technol. 17 (2006) 189–206.
and the formation of a very thin and dense layer on membrane sur-

Please cite this article in press as: M.T. Hung, J.C. Liu, Microfiltration of microalgae in the presence of rigid particles, Separ. Purif. Technol. (2016), http://dx.
doi.org/10.1016/j.seppur.2016.10.063

You might also like