Algal Research 35 (2018) 561–568

Contents lists available at ScienceDirect

Algal Research
journal homepage: www.elsevier.com/locate/algal

Magnetic–cationic cassava starch composite for harvesting Chlorella sp. T

TISTR8236
Kanthida Jangyubola, Kittiwut Kasemwongb, Theppanya Charoenrata, Supenya Chittapuna,
⁎

a
Department of Biotechnology, Faculty of Science and Technology, Thammasat University, Rangsit Centre, Pathum Thani 12120, Thailand
b
National Nanotechnology Center, National Science and Technology Development Agency, 111 Thailand Science Park, Phahonyothin Road, Khlong-Luang, Pathumthani
12120, Thailand

ARTICLE INFO ABSTRACT

Keywords: Microalgae are cultured for their useful chemicals, but their development is restricted by the costs and diffi-
Microalgae culties of harvesting them. New composites, magnetic-cationic cassava starch, were successfully developed to
Separation facilitate harvesting. Compared to naked magnetic particles, the composites exhibited higher potential for algal
Cationic cassava starch cell recovery. In addition, harvest capacity through the composites was positively related to the degree of ca-
Composite
tionic substitution (DS) in modified cassava starch. The application of 500 mg L−1 composites with DS
Magnetic harvesting
value ≥ 0.76 to recover 1 g L−1 Chlorella sp. TISTR8236 (1.67 ± 0.01 g-dry cell weight g-composites−1) from
medium broth at pH 9.5 was significantly (p ≤ 0.005) > 98% effective. The separation efficiency of this com-
posite for 1 g L−1 algal biomass was evaluated under different composite doses (200 to 800 mg L−1) and pH
levels (4 to 10). Ninety-five percent of the algal cell separation efficiency of the magnetic and 0.76 DS cationic
starch composite was recorded, when applied to the 300 mg L−1 composite to harvest 1 g L−1 Chlorella sp.
TISTR8236 at pH 10 within 2 min. The adsorption isotherm fit the Langmuir model. In addition, desorption of
algal cells was conducted by adjustment of the pH of the culture medium. The highest desorption (9.69%) was
recorded at pH 12. These composite magnetic-cationic cassava starch particles were shown to be a new ad-
sorption material for algal cell harvesting. This alternative composite for magnetic separation may represent an
effective method for saving time and energy in algal industrial processing.

1. Introduction
Microalgae are a valuable natural source of beneficial pigments,
protein, lipids and compounds that have wide applications including
nutraceuticals, pharmaceuticals, biofuel, food and feed additives [1–3].
However, large-scale production of microalgae requires a high invest-
ment cost for cultivation, harvesting, extraction and purification pro-
cesses. The harvesting cost of microalgae is 20–30% of the total pro-
duction cost, which is accounted for by the high energy and time
consumption in biomass separation because of their small size (3–30 μm
diameter), low density (< 1 g L−1) and colloidal stability [1,3–5]. A
major contribution to the use of microalgae would be improved har-
vesting.
Magnetic separation is a promising technique that has been used for
biomass collection for both marine and freshwater microalgae [6]. The
principle of magnetic separation is the application of magnetic particles
to attach to microalgae cells and subsequently separate the biomass
from the liquid culture medium via an external magnetic bar [7]. The
advantages of magnetic separation are that it is simple, fast, and in-
expensive with no energy and chemicals requirements, which makes it
an energy efficient and cost effective technique [5,8]. However, simple
naked magnetic particles have the problems of particle oxidation and
low electrostatic interaction between the negative charge of algal cell
surfaces and the neutral charge of the magnetic particle surface, re-
sulting in the requirement of a high dosage of magnetic particles [9].
Coating the particles with a polymer reduces contact with oxygen. In
addition, giving the coated polymer a positive charge was shown to
enhance the efficiency of algal biomass harvesting [10].
Coating naked magnetite surfaces with polymers to make magnetic
composites before the adhesion of algal cell is called the “immobilized-
on” strategy, which was shown to achieve higher efficiency than the
“attached-to” approach, where cells are coated with a polymer and are
subsequently attached by naked magnetics [4,5]. Electrostatic interac-
tion between polymers and algae was a main factor that affected the
performance of harvesting efficiency [5]. Synthetic polymers that have
been used include poly-(diallyldimethylammonium chloride) [11,12],

⁎
Corresponding author.
E-mail address: supenyac@tu.ac.th (S. Chittapun).

https://doi.org/10.1016/j.algal.2018.09.027
Received 3 April 2018; Received in revised form 10 September 2018; Accepted 26 September 2018
Available online 05 October 2018
2211-9264/ © 2018 Published by Elsevier B.V.
K. Jangyubol et al.
polyethylenimine [9,13], diethylaminoethyl [14] and cationic poly-
acrylamide [7]. Furthermore, natural polymers have been used for
coating with magnetic composites, including: chitosan [3,15], starch
[16] and silica [6]. Harvesting of Chlorella ellipsoidae cells, a green
microalga, by using 20 mg L−1 of magnetic nanocomposites coated with
polyethylenimine, achieved 97% of harvesting efficiency after 2 min,
which was a tenfold lower particle dosage requirement compared to
that of naked Fe3O4 magnetic nanoparticles of 250 mg L−1 [9,13]. The
application of silica-coated magnetic particles for Chlamydomonas re-
inhardtii, Chlorella vulgaris, Phaeodactylum tricornutum and Nanno-
chloropsis salina separation provided high separation efficiencies >
95% with maximum particle loads for C. reinhardtii, C. vulgaris, and P.
tricornutum of 30, 30 and 77 g g−1, respectively [6]. Moreover, Zhao
et al. [5] examined the hybrid strategy by mixing “immobilized-on” and
“attached-to” together with the composite (polyaluminum chloride
(PACl) + Fe3O4 or polyacrylamide (PAM) + Fe3O4 and flocculants
(PAM or PACl)), respectively. The results showed that 99% of algal
biomass could be harvested within 30 s by adding PACl+Fe3O4
(0.625 mmol Al L−1 + 10 g L−1) and PAM (3 mg L−1).
From the previous studies mentioned above, it is clear that coated
magnetic composites resulted in increased separation efficiency.
Therefore, the polymer used for coating and the dosage of magnetic
composites are the main factors influencing the harvesting efficiency,
but different types of polymer coatings on naked particles showed dif-
ferent efficiencies in algal biomass harvesting. Therefore, the objective
was to test natural polymers with cationic properties for coating naked
Fe3O4, that are ecologically sound and, could increase algal biomass
harvesting efficiency.
Starch is the second-most abundant natural polymer on earth.
Cassava starch has attracted increasing attention as a raw material for
various industrial uses because of its low price, renewability and bio-
degradability. However, the utilization of native starch was reported to
be limited by its physicochemical properties, such as water insolubility
and its tendency to form unstable pastes and gels [17]. Therefore, the
physical and chemical modification of native starch could improve its
functionality for other applications. Cationic starches and modified
starches, which are widely applied in the paper and textile industry and
in waste water treatment, have thermoplasticity and water dissolution
properties. In addition, two commercial cationic starches, which are
commonly used in waste water treatment and in the paper industry,
were reported to be efficient flocculants for freshwater algae, Para-
chlorella and Scenedesmus. Another advantage is that cationized poly-
saccharides are biodegradability, have a relatively low price and are
more effective flocculants for highly negatively charged colloidal par-
ticles, such as algae [18]. Therefore, this study aimed to synthesize
magnetic-cationic cassava starch composite material of different de-
grees of substitution (DS) with cations into the molecule of starch. In
addition, the potential of a novel composite for algal biomass separa-
tion was examined by harvesting the microalgae, Chlorella sp.
TISTR8236. Moreover, the pH is an important factor that affects the
protonation/deprotonation of the composite and the microalgal surface
and, consequently, affects the algal cell harvesting efficiency [13].
Thus, the present experiments were performed under different pH levels
and dosages of new composites. Moreover, the adsorption isotherm
between the composites and the algal cell was investigated.
2. Materials and methods
2.1. Synthesis of composites magnetic-cationic cassava starch
The magnetic particles were synthesized by the procedure pre-
viously reported by Mikhaylova and her group [19]. Cationic cassava
starch was synthesized following the approach of Wang et al. [17], and
the ratio of glycidyl trimethyl ammonium chloride:anhydroglucose unit
(GTMAC:AGU) was modified to vary its DS value. DS was calculated
according to the following Eq. (1) [20].
562
Algal Research 35 (2018) 561–568
DS =
162N%
1400 151.5N% (1)
where N% is nitrogen content determined by a carbon hydrogen ni-
trogen element analyzer (TruSpec CHN), and 162 and 151.5 are the
molecular weights of AGU and GTMAC, respectively.
The synthetic magnetic composite was modified according to Lee
et al. [3]. Two grams of modified cationic cassava starch was dissolved
in 10 mL of acetic acid at pH 4; next, 1 g of magnetic particles were
gradually added into the solution, which was mixed in a sonication bath
at 60 Hz for 3 h. For cross-linking between cationic cassava starch and
magnetic particles, 15 mL of 0.15% (w/v) sodium tripolyphosphate
solution in distillated water was added and incubated overnight. The
composites were collected using a small permanent magnet bar. The
supernatant was tested by the phenol‑sulfuric acid method [21] to ex-
amine the cationic starch to ensure that there was sufficient cationic
starch for synthesizing it with magnetic material and to confirm that all
magnetic particles were formed into the composites. The composite was
washed with distilled water several times and redispersed and pre-
served in distilled water for further use.
2.2. Microalgae source and cultivation
Chlorella sp. TISTR8236, was supplied by the Thailand Institute of
Scientific and Technological Research and cultured in BG-11 medium
[22] (pH 7.0) in a 500 mL Erlenmeyer flask that was incubated in a
shaker at 120 rpm at 25 ± 1 °C and was illuminated continuously 12 h
per day at 45 μmol photons m−2 s−1 for 24 days. The initial Chlorella sp.
biomass concentration was 0.075 g L−1, and after 24 days of culture,
the cell concentration was 1 g L−1, and the culture pH was 9.5.
2.3. Harvesting efficiency of the composites
Algal cells were harvested by centrifugation at 1008 ×g for 10 min
and washed twice with distilled water to remove the secondary meta-
bolites or organic matter, which may affect the electrostatic interaction
between algal cells and composites during an experiment. Chlorella sp.
was resuspended in BG-11 medium with a final concentration of 1 g L−1
of cells and was used for investigation of the harvesting efficiency.
Naked magnetic or the composite magnetic particles were added to the
10 mL of microalgal suspension, which was then mixed by a vortex
mixer for 1 min, and cell separation was performed by placing the vial
on a permanent magnetic bar for 2 min. The supernatant was sampled
and measured for optical density (OD) using a spectrophotometer at
750 nm to determine the microalgae concentration before harvesting
(ODi) and after harvesting (ODf).
The optimal magnetic dosage was preliminarily investigated at
pH 9.5 by varying the quantity of naked magnetic and composite
magnetic agents at 200, 300, 400, 500, 600, 700 and 800 mg L−1. The
dosage of adsorption material that provided moderate harvesting effi-
ciency was selected for the following study. The influence of DS values
of composites were studied using the composite with different DS va-
lues in the range of 0.22 to 0.91. The effect of pH on the algal har-
vesting efficiency was studied by varying the pH of microalgal sus-
pension in the range of pH 4 to 10, which had been reported as the
optimum pH for several microalgal species [8,23]. The optimum dosage
and DS values of the adsorption material and optimum pH were chosen
and were validated for harvesting efficiency and recovery capacity.
The harvesting efficiency and recovery capacity of the magnetic
composites-cationic cassava starch for the Chlorella sp. cells (g-DCW g-
composites−1) were calculated using Eqs. (2) and (3), respectively.
ODi ODf
Harvesting efficiency(%) = × 100
ODi (2)
where ODi is the initial OD before harvest and ODf is the OD of the
supernatant after harvest.
K. Jangyubol et al.
Recovery capacity =
(C 0 Ct ) V
m (3)
where C0 is the concentration of microalgae in the culture medium
before harvesting (g L−1), Ct is the concentration of supernatant after
harvesting (g L−1), V is the volume of microalgal cultures used for
harvesting (L), and m is the amount of composites used for harvesting
(g).
2.4. Adsorption isotherm
Among the mathematical models used for the description of ad-
sorption isotherms, the Langmuir and Freundlich models are the most
popular due to their reasonable accuracy and relative simplicity [8]. To
quantitatively evaluate the adsorption ability of magnetic composites
for Chlorella sp., the adsorption isotherms were tested at the particle
dosage of 500 mg L−1 with different biomass concentration (0.2 to
2 g L−1).
The equilibrium data were analyzed using Langmuir and Freundlich
models. The linear forms of the Langmuir isotherm and the Freundlich
isotherm are shown in Eqs. (4) and (5), respectively.
Ce 1 Ce
= +
Qe QmKL Qm (4)
1
logQe = logKF + logCe
n (5)
where Ce is equilibrium concentration of the microalgal cells in the
solution (g L−1), Qe is the amount of microalgal cells adsorbed per unit
weight of composites (g-DCW g-composites−1), Qm is the maximum
adsorption capacity (g-DCW g-composites−1), KL is the Langmuir ad-
sorption equilibrium constant (L g−1), KF is the physical constants of
the Freundlich adsorption isotherm indicating the adsorption capacity
(g-DCW g-composites−1) and n is Freundlich constants, respectively.
2.5. Desorption of algal cells
After harvesting microalgal cells at pH 10 by composites
(DS = 0.76), an aqueous solution (pH 2 to 12) was added for detach-
ment. Microalgae were separated from the magnetic composites
through electrostatic repulsion generated by pH and vortexed in a
mixer for 2 min. The composites were harvested by a permanent mag-
netic bar, and the optical density of supernatant containing drifting
microalgae was examined using a spectrophotometer at 750 nm.
The detaching efficiency was determined by the concentration of
microalgae both before and after detaching, which can be calculated by
using the following Eq. (6).
ODf
Detaching efficiency (%) = × 100%
ODi (6)
where ODi is the optical density (ODi) of initial culture and ODf is the
OD of supernatant liquid after the magnetic composites with micro-
algae were removed by the permanent magnet.
2.6. Analytical methods
The determination of nitrogen content in cationic cassava starch
was examined by the CHN element analyzer. A sample of 0.1 g cationic
starch was placed in a combustion tube and combusted at ~1000 °C
under helium gas and a high purity of oxygen environment. During the
combustion process, nitrogen was converted to oxides of nitrogen and
nitrogen gas. The gasses were then passed through copper and the
oxides of nitrogen were converted to N2. The N2 was separated in a
chromatographic column and was detected by a thermal conductivity
detector (TCD).
To determine the surface structure, the infrared spectrum was
measured. Naked magnetic particles and native and modified cationic
563
Algal Research 35 (2018) 561–568
cassava starch and composites were analyzed by using Fourier trans-
form infrared spectroscopy (FTIR) (Perkin Elmer SPECTRUM GX FTIR
System) via the KBr disc technique. The preparation of pellets was
carried out by mixing a few milligrams of the starch or the composite
sample with 200 mg of potassium bromide. The mixture was placed into
a pellet-forming die to make a transparent pellet and was placed in a
sample holder in FTIR spectroscopy. Spectra were obtained between the
wavelength ranging from 400 and 4000 cm−1, 64 times per sample
with a resolution of 4.
The size distribution of naked magnetic particles and composites
examined by using Malven Mastersizer 3000 (Malvern, UK) in the size
range of 0.01 to 3500 μm. A sample of 1 mL was dispersed into 20 mL
distilled water. To make the solution homogeneous, the stirring rate
was set at 1200 rpm. The particle size distribution was determined by
laser diffraction method. The emitted laser beam was scattered onto the
particles in the suspensions and the detectors measured the intensity of
the light scattered on the particles to categorize their size.
The zeta potential measurements of Chlorella sp., naked magnetic
particles and composites were measured in the model environments at
25 °C using the Zetasizer Nano-ZS (Malvern, UK) fitted by size and the
zeta potential folded capillary cell (DTS1070). This cell is used for
measurements of aqueous-based samples. A total of 0.05 g sample was
suspended in 30 mL distilled water and was placed in a test chamber.
The pH of sample suspension was adjusted from 2 to 10 using a 0.1 M
NaOH and 0.1 M HCl solution. The zeta potential measurement was
applied under a field of 12 v, and all measurements were converted into
zeta potentials using the Smoluchowski equation.
2.7. Data analysis
All experiments were performed in three replications and the results
were presented as the mean values ± standard deviation. One way
analysis of variance (ANOVA) was also carried out to evaluate differ-
ences among the treatments of algal harvesting efficiency, recovery
capacity and detachment efficiency at p level of 0.05. Tukey's HSD test
was applied for multiple comparisons of each treatment when sig-
nificant (p ≤ 0.05).
3. Results and discussion
3.1. Characteristics of naked magnetic particles and composites
The magnetic (Fe3O4) composites were generated by introducing
cationic cassava starch onto the surface of Fe3O4 particles by means of
electrostatic interaction. The adsorption bands of FTIR at 1654 cm−1
presented amine NeH stretching vibrations, which confirmed the pre-
sence of the cationic group in the composites. However, the bands of
FTIR at 1654 cm−1 were also observed in spectra (a) of Fig. 1, which
was the FTIR spectra of naked magnetic particles. This could have been
caused of the residual ammonium ion (NH4+), since the ammonium ion
is used in naked magnetic particle preparation. A similar result was
reported by Hu et al. [13].
The bands at 1154 cm−1 and 1028 cm−1 were attributed to the CeO
bond stretching vibration of the anhydroglucose units (Fig. 1), as pre-
viously reported by Wang et al. [17]. The absorption peak at
3400 cm−1 corresponded to OeH stretching as a result of hydrogen
bonding involving the hydroxyl groups on the starch molecules. Com-
pared to the spectra of naked magnetic particles, those of the compo-
sites were observed, as the band at approximately 3400 cm−1 was more
pronounced and broader. The strong adsorption band at 584 cm−1 was
assigned to FeeO adsorption, which was observed from magnetic
composites coated with polyethyleneimine [13]. The FTIR data con-
firmed the crosslink between modified cationic cassava starch and the
magnetic particles.
The size distribution of naked magnetic particles and composites is
shown in Fig. 2. Uncoated particles synthesized from this study varied
K. Jangyubol et al. Algal Research 35 (2018) 561–568

Fig. 3. Harvesting efficiency of Chlorella sp. from the algal culture medium by
Fig. 1. FTIR spectra of naked magnetic particles (a), composite-cationic starch varying naked magnetic particles from 200 to 800 mg L−1 at pH 9.5. (The let-
at DS = 0.22 (b), composite-cationic starch at DS = 0.76 (c), cationic cassava ters (a, b, c, d, e) are significantly different from each other (one way ANOVA
starch DS = 0.22 (d) and cationic cassava starch DS = 0.76) (e). and Turkey's HSD, p ≤ 0.05)).

3.2. Harvesting efficiency of adsorption materials

3.2.1. Harvesting efficiency of naked magnetic particles

The effect of the adsorption material dosage on harvesting efficiency
was studied at pH 9.5, which was the pH of the algal culture medium
after 24 days of cultivation. The experiment of increasing the naked
magnetic particles resulted in an improvement in the Chlorella sp.
harvesting efficiency. The correlation between the dosage of naked
particles and harvesting efficiency followed a sigmoidal pattern. At 200
to 400 g L−1 of particles the harvesting efficiency was approximately
46.09–47.67%. The increase of the dosage to 500 g L−1 gave a mod-
erate harvesting efficiency of 52.67 ± 5.34%. The highest harvesting
efficiency of 91.97–95.20% was derived from applying the magnetic
condition of 700 and 800 mg L−1 (Fig. 3).

3.2.2. Harvesting efficiency of the composites

An appropriate amount of adsorbent for harvesting the efficiency
investigation of the composites is very important. The application of an
Fig. 2. Size distribution of naked magnetic particles and composites-cationic overdose of adsorbent may result in saturated adsorption, and conse-
cassava starch with DS = 0.22 and DS = 0.76. quently, the different harvesting efficiencies between naked and com-
posite particles may not be detected. Therefore, the effect of the DS
values of the composite on the harvesting efficiency was studied by
from 0.2 to 40 μm and were mostly distributed in the range of 5.1 to
fixing the composite dosage at 500 mg L−1, which showed moderate
10.0 μm, which was similar to the commercial magnetite powder used
harvesting efficiency for the naked particles. With the same dosage of
for study magnetic flocculation of Chlorella vulgaris by Zhao et al. [5],
adsorption materials, the magnetic composite-cationic cassava starch
while the coated particles were widely distributed between 0.2 and
showed higher potential in Chlorella sp. recovery compared to naked
127 μm. The bonding composite with cationic cassava starch resulted in
magnetic particles. The application of the composite bonded with a
an increase in the number of small particles (< 5.0 μm). Additionally,
higher DS value exhibited improved separation efficiency (Table 1 and
modified cassava starch coated with higher DS values (DS = 0.76)
Fig. S1 in Supplementary data). The composites with a DS value ≥ 0.76
compared to that of DS = 0.22 increased the smaller particles of 1 to
were significantly (p ≤ 0.05) effective in algal cell harvesting efficiency
5 μm from 27% to 32%. This demonstrated that binding a magnetic
(≥98.09 ± 0.74%) at pH 9.5 (Table 1). This study demonstrated that
particle to a starch particle may reduce the magnetostatic and Van der
the composite showed a good performance as a separator agent and
Waals function among the naked particles, which results in dis-
enhanced the harvesting efficiency compared to naked magnetic par-
aggregation of a large naked particle. This effect was previously ob-
ticles, an effect previously reported from Hu et al. [9,13]. These re-
served by Wang et al. [4]. This mechanism improved a dispersibility of
searchers found that a lower dosage of coated particle was required to
composite particles, which can increase the chances of composites at-
achieve equal separation effectiveness. In addition, magnetic compo-
taching to algal cells and consequently enhance their harvesting effi-
sites-cationic cassava starch with higher DS values showed higher ef-
ciency.
ficiency in algal cell recovery. Vandamme et al. [24] showed that a

564
K. Jangyubol et al. Algal Research 35 (2018) 561–568

Table 1
Harvesting efficiency of Chlorella sp. by 500 mg L−1 composite with different DS value at pH 9.5.
Degree of substitution (DS) Zeta potential (mV) Harvesting Recovery capacity
efficiency (%)1 (g-DCW g-composites−1)1

Naked magnetic particles −2.06 ± 0.21c 51.50 ± 4.28d 0.88 ± 0.07g

0.22 −5.24 ± 0.07bc 63.18 ± 1.91c 1.07 ± 0.03f
0.54 −6.32 ± 0.17b 67.47 ± 0.92c 1.15 ± 0.02e
0.60 −6.27 ± 0.52b 69.24 ± 2.16bc 1.18 ± 0.04d
0.62 −6.40 ± 0.30b 75.31 ± 4.02b 1.28 ± 0.07c
0.76 −15.70 ± 1.67a 98.09 ± 0.74a 1.67 ± 0.01b
0.82 −15.90 ± 3.07a 98.62 ± 0.32a 1.68 ± 0.01a
0.91 −14.76 ± 0.27a 98.99 ± 0.31a 1.68 ± 0.01a

1
The values are represented in terms of the mean ± standard deviation (SD); the values in the same column with different superscript letters are
significantly different (one way ANOVA and Turkey's HSD, p ≤ 0.05)).

result that a high separation efficiency was obtained. The high algal cell
harvesting efficiency of this study were recorded from the composite
with DS ≥0.76, which was a higher, different value of the zeta po-
tential between composites and algal cells than in the composite with
DS < 0.76 (Fig. 4).

3.2.3. Effects of pH on composite harvesting efficiencies

Fig. 4. Zeta potential of Chlorella sp., from naked magnetic particles and
composites with different DS values of cationic cassava starch at pH 9.5.
commercial cationic starch with 0.15 DS had more efficient flocculation
in Parachlorella and Scenedesmus than in others with 0.11 DS.
At pH 9.5, Chlorella sp. exhibited positive zeta potential, while
naked particles and composites displayed negative zeta potential
(Fig. 4). Composites with DS values higher than 0.76 showed a stronger
negative charge at pH 9.5. Since alkali was added, the surface of the
cationic particles became less positive, and the negative potential of
zeta was detected. In addition, different zeta potential values between
algal cells and composites were observed, where higher separation ef-
ficiency was detected (Table 1 and Fig. 4). Prochazkova et al. [14]
reported that a stronger electrostatic interaction was formed with the
The effect of the pH value on the harvesting efficiency was studied
by adjusting the pH in the range of 4 to 10, using 1 M H3PO4 or 1 M
NaOH with a concentration of 1 g L−1, for the microalgal growth stage
of 24 days. The high harvesting efficiency was at pH 4, 9 and 10. The
maximum harvesting efficiency of 98.94 ± 0.35% was obtained at a
pH 10 (Table 2 and Fig. S2 in Supplementary data). The recovery ca-
pacity of the magnetic-cationic cassava starch composites for the mi-
croalgal cells was 1.98 ± 0.01 g-DCW g-composites−1. However, har-
vesting efficiency of Chlorella sp. showed a slight decrease in the pH
range of 5 to 8 and subsequently increased in the pH range of 9 to 10. In
comparison with other harvesting methods on Chlorella ellipsoidea, the
magnetic separation by Fe3O4–PEI nanocomposites provided higher
harvesting efficiency at pH 4 [13]. The decreased harvesting efficiency
at the pH range of 5 to 8 was due to the competition between the hy-
droxyl ions and the microalgal cells for the magnetic composite ad-
sorption sites. However, the high harvesting efficiency at pH 4, 9 and 10
would have given different surface ionization characteristics to the
magnetic composites and microalgae which was also observed in this
work (Fig. 5) and was reported by Hu et al. [9] to be at high pH values.
These researchers observed that the diameter of cell aggregates

Table 2
Harvesting efficiency of Chlorella sp. from the algal culture medium by varying
pH 4 to 10 using magnetic composites (DS = 0.76) at 500 mg L−1.
pH Harvesting efficiency (%)1 Recovery capacity (g-DCW g-composites−1)1

4 91.91 ± 0.35b 1.84 ± 0.01b

5 90.48 ± 1.20bc 1.81 ± 0.02bc
6 89.27 ± 0.35c 1.79 ± 0.01c
7 88.66 ± 0.82c 1.77 ± 0.02c
8 90.63 ± 1.14bc 1.81 ± 0.02bc
9 96.98 ± 0.57a 1.94 ± 0.01a
10 98.94 ± 0.35a 1.98 ± 0.01a

1
The values are represented in terms of the mean ± standard deviation
(SD); the values in the same column with different superscript letters are sig- Fig. 5. Zeta potential of magnetic composites and Chlorella sp. (● Chlorella sp.;
nificantly different (p ≤ 0.05). ○ magnetic particles; ▼ composites (DS = 0.76).

565
K. Jangyubol et al. Algal Research 35 (2018) 561–568

Table 3
Algal cell harvesting by magnetic separation.
Microalgae Agent Dosage (mg L−1) pH Reaction time Harvesting efficiency Reference
(min) (%)

B. braunii Fe3O4 nanoparticles 75 4 2–3 99.9 [8]

C. ellipsoidea Fe3O4 nanoparticles 300 7 2–3 98.9 [8]
C. vulgaris Iron oxide magnetic microparticles 900 4 10 95 [14]
N. maritima Fe3O4 nanoparticles 120 8 4 97.5 [9]
C. ellipsoidea Fe3O4-PEI nanocomposites 20 4 2 97 [13]
N. oceanica Iron oxide magnetic microparticles Mass ratio (2−10) 10 10 99 [27]
Chlorella sp. BaFe12O19 particles Not mention not mention 2–3 98.6–99.5 [28]
Chlorella sp. TISTR8236 Magnetic - cationic cassava starch composites 300 10 2 94.96 This study
(DS = 0.76) 800 10 2 99.15

increase, which enhanced the harvesting efficiency. A harvesting effi- DS = 0.76 under the optimum pH level. Moreover, the adsorption iso-
ciency > 95%, was observed at pH 9 and 10 (Table 2). therm between the composites and the algal cell should be determined.
The electrostatic attraction of algal cells and separator agent were
varied according to the algal species, agent types and pH. The high cell
3.2.4. Influence of the dosage of 0.76 DS composite under optimum pH
separation efficiencies of Chlorella ellipsoidea, Nannochloropsis maritima
The harvesting efficiency increased significantly (p ≤ 0.05) by in-
and N. oceanica were reported to achieve neutral and alkaline pH
creasing the dosage of magnetic composites, and it reached > 95%
(pH ≥ 7) (Table 3). Harvesting of C. ellipsoidea using Fe3O4 nano-
when the magnetic composites dosage rose to 400 mg L−1 at pH 10. The
particles accomplished the highest harvesting efficiency (> 98%) at
recovery capacity of the magnetic-cationic cassava starch composites at
pH 7 [8]. The increase in pH to alkaline condition (pH 8) led to a de-
pH 10 rose to 4.69 ± 0.01 g-DCW g-composites−1 of 200 mg L−1 do-
crease in the required dosage of magnetic particles for achieving a high
sage (Table 4). The harvesting efficiency increased as the dosage of
algal harvesting efficiency of N. maritima [9]. Alternatively, the high
magnetic composites increased until the highest harvesting efficiency
harvesting efficiency of Botryococcus braunii and Chlorella vulgaris can
value was reached. This relationship is mainly due to the ratio of the
be successful at lower pH levels. Harvesting of Chlorella ellipsoidea using
algal cells and the magnetic composites. At a low particle dosage, the
Fe3O4-PEI nanocomposites and C. vulgaris using iron oxide magnetic
magnetic particles cannot completely adsorb the algal cells, and the
microparticles achieved the highest harvesting efficiency of 97% [13]
algal cells cannot adsorb sufficient magnetic particles and cannot be
and 95% [14] at pH 4. There were increased or decreased pH values of
completely harvested by the external magnetic field due to the in-
solutions that affected the attraction force between algal cells and
sufficient magnetic force. At the optimal dosage, the quantity of the
composites [9,14]. Therefore, pH was an important factor that affected
magnetic composites was sufficient to interact with the algal cells and
the protonation/deprotonation of composite and microalgal surfaces
provided sufficient magnetic force to collect algal cells completely by
and consequently affected the algal cell harvesting efficiency [13].
an external magnetic field and the maximum harvesting efficiency was
Moreover, increasing the medium pH to 11 by adding 0.5 M NaOH can
achieved.
induce C. vulgaris flocculation and had been considered to be a poten-
tially useful method to preconcentrate freshwater microalgal biomass
during separation processes [23]. In contrast to Chlorococcum nivale, 3.3. Adsorption isotherm
Chlorococcum ellipsoideum and Scenedesmus sp., decreasing the pH of the
growth medium to 4 by adding 01. M HNO3 resulted in 90% floccula- The influence of the pH, over the range of 4 to 10 on the harvesting
tion efficiencies with a high cell concentration of 1 g L−1 [25]. In this efficiency was studied at a dosage of 500 mg L−1 magnetic composites
study, the flocculation of C. vulgaris could also be observed when 0.1 M under initial microalgal concentrations ranging from 0.2 to 2 g L−1 by
NaOH was added into a suspension and the pH reached 12. By in- concentrating or diluting the algal medium with BG-11 medium at the
creasing the pH, high pH induced the flocculation might have caused an growth stage of 24 days. The experimental data were fitted by the
increase in the negative surface charge [26]. Langmuir isotherm and Freundlich isotherm, and the fitted parameters
According to the results, the harvesting efficiency in the pH range of of two models along with the correlation coefficients (R2) were calcu-
4 to 10 was above 88%, which was due to the strong charge of the lated. The results indicated that pH played an important role in the
cationic composite. Therefore, pH 10 was chosen for next experiment to process of magnetic separation. The Langmuir model had a better fit
study the harvesting efficiency of the dosage of the composite with with the adsorption process compared to the Freundlich model
(Table 5). The Langmuir model demonstrated that the adsorption of the
magnetic composites (DS = 0.76) onto the algal cells gave monolayer
Table 4 coverage, and magnetic composites had limited adsorption to the
Harvesting efficiency of Chlorella sp. from algal culture medium by varying
magnetic composites (DS = 0.76) from 200 to 800 mg L−1 at pH 10. Table 5
Dosage Harvesting efficiency Recovery capacity
Model parameters for harvesting of Chlorella sp. by magnetic composites
(mg L−1) (%)1 (g-DCW g-composites−1)1 (DS = 0.76) at different pH levels.
pH Langmuir isotherm Freundlich isotherm
200 93.77 ± 0.26f 4.69 ± 0.01a
300 94.96 ± 0.43e 3.17 ± 0.01b
qm(g g−1) KL R2 KF n R2
400 96.38 ± 0.20d 2.41 ± 0.00c
500 97.74 ± 0.35c 1.95 ± 0.01d 4 1.33 22.51 0.998 1.32 4.94 0.840
600 98.36 ± 0.10bc 1.64 ± 0.00e 5 1.27 19.88 0.998 1.25 4.78 0.841
700 98.70 ± 0.10ab 1.41 ± 0.00f 6 1.26 16.23 0.992 1.23 4.05 0.808
800 99.15 ± 0.17a 1.24 ± 0.00g 7 1.35 11.09 0.989 1.26 4.05 0.734
1
8 1.32 16.29 0.998 1.28 4.35 0.799
The values are represented in term of mean ± standard deviation (SD); the 9 1.33 21.30 0.997 1.32 4.77 0.777
values in the same column with different superscript letters are significantly 10 1.37 25.43 0.998 1.37 5.09 0.787
different (p ≤ 0.05).

566
K. Jangyubol et al.
Table 6
Detachment efficiency of Chlorella sp. from magnetic
composites (DS = 0.76) at pH 2 to 12.
pH Detaching efficiency (%)1
2 3.73 ± 0.08bc
3 2.62 ± 0.32cd
4 0.01 ± 0.92e
5 3.14 ± 0.47bcd
6 4.05 ± 1.05bc
7 4.52 ± 0.71bc
8 2.68 ± 0.64cd
9 1.13 ± 0.53de
10 0.00 ± 0.00e
11 5.27 ± 1.38b
12 9.69 ± 1.28a
1
The values are represented in terms of the
mean ± standard deviation (SD); the values in the
same column with different superscript letters are sig-
nificantly different (p ≤ 0.05).
surface of the magnetic composite-coated algal cells due to the elec-
trostatic force avoiding multilayer adsorption. The electrostatic repul-
sion among the algal cells also prevented them from aggregating into
large particles. The maximum adsorption capacity, Qm, in this case was
calculated as 1.37 g g−1.
3.4. Desorption
The desorption of algal cells had been conducted by an adjustment
to the media pH over the range of 2 to 12. The highest detaching ef-
ficiency of 9.69% was recorded at pH 12 (Table 6). Similar results had
previously been reported by Seo et al. [28] who demonstrated that the
detachment between microalgae and BaFe12O19 was successfully
achieved at pH 12 and the large particle size on the detaching perfor-
mance was up to 85%. However, the desorption of algal cells in this
work has low efficiency. Two procedures of algal magnetic particle
aggregate processing reported by Xu et al. [8] consisted of the de-
magnetization of algal cells before the processing or extraction of ag-
gregated biomass without the detachment of magnetic particle.
Choosing any procedure is based upon the characteristic of algal species
and desired products [4]. Products applied in medical aspects must be
high-purity and uncontaminated by magnetic particles; therefore, de-
magnetization of algal cells before extraction and purification is a good
choice, whereas extraction of hydrocarbon from algal cells as B. braunii
is not influenced by the existence of magnetic particles [8]. Therefore,
aggregated biomass should be process. According to the results of this
study, with less desorption than 10%, algal biomass aggregated with
magnetic-cationic cassava starch can be further used as a biomass for
lipid extraction.
4. Conclusions
The modification of native cassava starch with different
GTMAC:AGU molar ratios produced cationic cassava starch with a de-
gree of substitution. The higher ratio of GTMAC: AGU promoted a
greater degree of substitution and stronger cationic charge. Compared
to naked magnetic particles, the successful magnetic composites-ca-
tionic cassava starch exhibited higher potential in Chlorella sp. re-
covery. The composites of cationic cassava starch at DS value ≥0.76
achieved ≥98.09% of Chlorella sp. cell harvesting efficiency at pH 9.5.
Using magnetic composite-cationic cassava starch of DS = 0.76 and a
dosage of 400 mg L−1 at pH 10, a relatively high harvesting efficiency
of 95% was achieved. Therefore, it can be concluded that the magnetic
composite-cationic cassava starch is a new potential composite
that performs as an effective agent for algal cell harvesting.
However, > 400 mg L−1 of composites may require to harvest 95%
567
Algal Research 35 (2018) 561–568
algal biomass from the algal culture system, because the organic sub-
stances secreted from the algal cell during cultivation may be compe-
titive with composites or alter pH condition. Moreover, the low de-
taching performance is a disadvantage, as it may limit an application of
the harvested algal cells. Nevertheless, the algal biomass with the
composite can be an alternatively raw material for biofuel production.
Authors' contributions
Kanthida Jangyubol: Literature review, Design of the research out-
line, Design of research tools, Analysis of data, Interpretation of the
data analysis, Drafting of the article (first draft), Revising the draft of
the article (second draft).
Kittiwut Kasemwong: Conception of the project, Design of the re-
search outline, Design of research tools, Decision on the data analysis
method, Interpretation of the data analysis.
Theppanya Charoenrat: Design of the research outline, Design of
research tools, Decision on the data analysis method, Interpretation of
the data analysis, Revising the draft of the article (second draft), Final
approval of the article (final draft).
Supenya Chittapun: Conception of the project, Literature review,
Design of the research outline, Design of research tools, Decision on the
data analysis method, Analysis of data, Interpretation of the data ana-
lysis, Revising the draft of the article (second draft), Final approval of
the article (final draft).
Conflict of interest
The authors declare that there are no conflicts of interest.
Statement of informed consent, human/animal rights
No conflicts, informed consent, or human or animal rights are ap-
plicable.
Acknowledgments
This research did not receive any specific grant from funding
agencies in the public, commercial, or not-for-profit sectors. The study
was conducted under the jurisdiction of the Department of
Biotechnology, Faculty of Science and Technology, Thammasat
University. The Department of Biotechnology and the Department of
Chemistry, the Faculty of Science and Technology, Thammasat
University and Nano Agro and Food Innovation (NAF) of National
Nanotechnology Center (NANOTEC) supported the facilitation of
equipment. We would like to thank Prof. Dr. Anthony Keith Thompson
for critically editing the manuscript.
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://
doi.org/10.1016/j.algal.2018.09.027.
References
[1] E. Molina Grima, E.H. Belarbi, F.G. Acién Fernández, A. Robles Medina, Y. Chisti,
Recovery of microalgal biomass and metabolites: process options and economics,
Biotechnol. Adv. 20 (2003) 491–515.
[2] C. Delattre, G. Pierre, C. Laroche, P. Michaud, Production, extraction and char-
acterization of microalgal and cyanobacterial exopolysaccharides, Biotechnol. Adv.
34 (2016) 1159–1179.
[3] K. Lee, S.Y. Lee, J.G. Na, S.G. Jeon, R. Praveenkumar, D.M. Kim, W.S. Chang,
Y.K. Oh, Magnetophoretic harvesting of oleaginous Chlorella sp. by using bio-
compatible chitosan/magnetic nanoparticle composites, Bioresour. Technol. 149
(2013) 575–578.
[4] S.K. Wang, A.R. Stiles, C. Guo, C. Liu, Harvesting microalgae by magnetic separa-
tion: a review, Algal Res. 9 (2015) 178–185.
[5] Y. Zhao, W.Y. Liang, L.J. Liu, F.Z. Li, Q.L. Fan, X.L. Sun, Harvesting Chlorella vulgaris
by magnetic flocculation using Fe3O4 coating with polyaluminium chloride and
K. Jangyubol et al.
polyacrylamide, Bioresour. Technol. 198 (2015) 789–796.
[6] M. Cerff, M. Morweiser, R. Dillschneider, A. Michel, K. Menzel, C. Posten,
Harvesting fresh water and marine algae by magnetic separation: screening of se-
paration parameters and high gradient magnetic filtration, Bioresour. Technol. 118
(2012) 289–295.
[7] S.K. Wang, F. Wang, Y.R. Hu, A.R. Stiles, C. Guo, C.Z. Liu, Magnetic flocculant for
high efficiency harvesting of microalgal cells, ACS Appl. Mater. Interfaces 6 (2014)
109–115.
[8] L. Xu, C. Guo, F. Wang, S. Zheng, C.Z. Liu, A simple and rapid harvesting method for
microalgae by in situ magnetic separation, Bioresour. Technol. 102 (2011)
10047–10051.
[9] Y.R. Hu, F. Wang, S.K. Wang, C.Z. Liu, C. Guo, Efficient harvesting of marine mi-
croalgae Nannochloropsis maritima using magnetic nanoparticles, Bioresour.
Technol. 138 (2013) 387–390.
[10] W. Wu, Q. He, C. Jiang, Magnetic iron oxide nanoparticles: synthesis and surface
functionalization strategies, Nanoscale Res. Lett. 3 (2008) 397–415.
[11] J.K. Lim, D.C. Chieh, S.A. Jalak, P.Y. Toh, N.H. Yasin, B.W. Ng, A.L. Ahmad, Rapid
magnetophoretic separation of microalgae, Small 8 (2012) 1683–1692.
[12] P.Y. Toh, S.P. Yeap, L.P. Kong, B.W. Ng, D.J.C. Chan, A.L. Ahmad, J.K. Lim,
Magnetophoretic removal of microalgae from fishpond water: feasibility of high
gradient and low gradient magnetic separation, Chem. Eng. J. 211–212 (2012)
22–30.
[13] Y.R. Hu, C. Guo, F. Wang, S.K. Wang, F. Pan, C.Z. Liu, Improvement of microalgae
harvesting by magnetic nanocomposites coated with polyethylenimine, Chem. Eng.
J. 242 (2014) 341–347.
[14] G. Prochazkova, N. Podolova, I. Safarik, V. Zachleder, T. Branyik, Physicochemical
approach to freshwater microalgae harvesting with magnetic particles, Colloids
Surf. B: Biointerfaces 112 (2013) 213–218.
[15] P.Y. Toh, B.W. Ng, C.H. Chong, A.L. Ahmad, J.W. Yang, C.J.C. Derek, J.K. Lim,
Magnetophoretic separation of microalgae: the role of nanoparticles and polymer
binder in harvesting biofuel, RSC Adv. 4 (2014) 4114–4121.
[16] T.T. Dung, T.M. Danh, L.T.M. Hoa, D.M. Chien, N.H. Duc, Structural and magnetic
properties of starch-coated magnetite nanoparticles, J. Exp. Nanosci. 4 (3) (2009)
259–267.
568
Algal Research 35 (2018) 561–568
[17] P.X. Wang, W. Xiu, D. Xue, K. Xu, Y. Tan, X. Du, W. Li, Preparation and char-
acterization of cationic corn starch with a high degree of substitution in
dioxane–THF–water media, Carbohydr. Res. 344 (2009) 851–855.
[18] S. Pal, D. Mal, R.P. Singh, Cationic starch: an effective flocculating agent,
Carbohydr. Polym. 59 (2005) 417–423.
[19] M. Mikhaylova, D.K. Kim, N. Bobrysheva, M. Osmolowsky, V. Semenov,
T. Tsakalakos, M. Muhammed, Superpaeamagnetism of magnetite nanoparticles:
dependence on surface modification, Langmuir 20 (6) (2004) 2472–2477.
[20] R. Kavalianuskaite, R. Klimaviciute, A. Zemaitaitis, Factors influencing production
of cationic starches, Carbohydr. Polym. 73 (2008) 665–675.
[21] M. Dubois, K.A. Gilles, J.K. Hamilton, P.A. Rebers, F. Smith, Colorimetric method
for determination of sugars and related substances, Anal. Chem. 28 (1956)
350–356.
[22] A. Ilavarasi, D. Mubarakali, R. Praveenkumar, E. Baldev, N. Thajuddin,
Optimization of various growth media to freshwater microalgae for biomass pro-
duction, Biotechnology 10 (6) (2011) 540–545.
[23] D. Vandamme, I. Foubert, I. Fraeye, B. Meesschaert, K. Muylaert, Flocculation of
Chlorella vulgaris induced by high pH: role of magnesium and calcium and practical
implications, Bioresour. Technol. 105 (2012) 114–119.
[24] D. Vandamme, I. Foubert, B. Meesschaert, K. Muylaertu, Flocculation of microalgae
using cationic starch, J. Appl. Phycol. 22 (2010) 525–530.
[25] J. Liu, Y. Zhu, Y. Tao, Y. Zhang, A. Li, T. Li, M. Sang, C. Zhang, Freshwater mi-
croalgae harvested via flocculation induced by pH decrease, Biotechnol. Biofuels 6
(2013) 98–108.
[26] S. Rakesh, S. Saxena, D.W. Dhar, R. Prasanna, A.K. Saxena, Comparative evaluation
of inorganic and organic amendments for their flocculation efficiency of selected
microalgae, J. Appl. Phycol. 26 (2014) 399–406.
[27] E. Boli, M. Savvidou, D. Logothetis, V. Louli, G. Pappa, E. Voutsas, F. Kolisis,
K. Magoulas, Magnetic harvesting of marine algae Nannochloropsis oceanica, Sep.
Sci. Technol. (2017) 1–8.
[28] J.Y. Seo, K. Lee, S.Y. Lee, S.G. Jeon, J.G. Na, Y.K. Oh, S.B. Park, Effect of barium
ferrite particle size on detachment efficiency in magnetophoretic harvesting of
oleaginous Chlorella sp, Bioresour. Technol. 152 (2014) 562–566.