Bioresource Technology 112 (2012) 212220

Contents lists available at SciVerse ScienceDirect

Bioresource Technology
journal homepage: www.elsevier.com/locate/biortech

Harvesting of microalgae by occulation with poly (c-glutamic acid)

Hongli Zheng, Zhen Gao, Jilong Yin, Xiaohong Tang, Xiaojun Ji, He Huang
State Key Laboratory of Materials-Oriented Chemical Engineering, College of Biotechnology and Pharmaceutical Engineering, Nanjing University of Technology,
No. 5, Xinmofan Road, Nanjing 210009, Peoples Republic of China

a r t i c l e

i n f o

Article history:
Received 7 October 2011
Received in revised form 7 February 2012
Accepted 13 February 2012
Available online 27 February 2012
Keywords:
Microalgae
Microbial occulant
Response surface methodology
Biomass harvest

a b s t r a c t
In an effort to search for an efcient and environmentally friendly harvesting method, a commercially
available microbial occulant poly (c-glutamic acid) (c-PGA) was used to harvest oleaginous microalgae.
Conditions for occulation of marine Chlorella vulgaris and freshwater Chlorella protothecoides were optimized by response surface methodology (RSM) and determined to be 22.03 mg L1 c-PGA, 0.57 g L1 biomass, and 11.56 g L1 salinity, and 19.82 mg L1 c-PGA and 0.60 g L1 biomass, respectively. Application
of the two optimized occulation methods to Nannochloropsis oculata LICME 002, Phaeodactylum tricornutum, C. vulgaris LICME 001, and Botryococcus braunii LICME 003 gave no less than 90% occulation efciency and a concentration factor greater than 20. Micrographs of the harvested microalgal cells
showed no damage to cell integrity, and hence no lipid loss during the process. The results show that occulation with c-PGA is feasible for harvesting microalgae for biodiesel production.
2012 Elsevier Ltd. All rights reserved.

1. Introduction
Two of the challenging global problems are the exhaustion of
fossil fuels and climate change. Microalgae are among the most
primitive forms of plant life able to capture CO2. In addition, some
microalgae can produce lipids suitable for biodiesel (Chiu et al.,
2009; Sialve et al., 2009). Compared with other energy crops, the
advantages of deriving biodiesel from microalgae include rapid
growth rates and a high per-acre yield. In addition, biodiesel has
low toxicity, is highly biodegradable and contains no sulfur (Hsieh
and Wu, 2009; Fu et al., 2009). Considering all the steps involved in
the biodiesel production from microalgae, harvest is a particularly
important step. Harvesting of microalgae is challenging because of
low cell concentrations (<0.5 kg m3 dry mass) in the medium and
their small sizes (330 lm in diameter) (Grima et al., 2003), the
small density differences between microalgae and growth media,
the ability of microalgae to remain in suspension, and the high ionic strength of seawater (Lee et al., 2009). Efcient and environmentally friendly harvesting would be important for commercial
biodiesel production from microalgae.
Flocculation is one of the most convenient methods for harvesting microalgae (Oh et al., 2001). Inorganic occulants such as ferric
chloride (FeCl3) and aluminum sulfate (Al2(SO4)3) and organic synthetic high-polymer occulants such as polyacrylamide derivatives
and polyethylene imine have been used for recovering microalgal
biomass (Papazi et al., 2009; Pushparaj et al., 1993). However,
Corresponding author. Tel./fax: +86 25 83172094.
E-mail address: biotech@njut.edu.cn (H. Huang).
0960-8524/$ - see front matter 2012 Elsevier Ltd. All rights reserved.
doi:10.1016/j.biortech.2012.02.086

these occulants are required in high doses (the dosages of FeCl3

and Al2(SO4)3 are both 0.75 g L1, and that of polyacrylamide is
0.35 mmol L1) (Papazi et al., 2009; Grima et al., 2003), and thus
may result in contamination of the microalgal lipids for biodiesel
production. Because the occulants are not readily degradable, a
major environmental concern, water pollution, may result. Moreover, the feedstock for biodiesel from microalgae is intracellular
lipids. Cell integrity may be affected by some chemical occulants,
causing lipid loss during occulation (Papazi et al., 2009), leading
to a decrease in the efciency of lipid recovery in the downstream
microalgal lipid extraction process.
In recent years, naturally occurring microbial occulants have
been applied to harvest microalgae for aquaculture and biodiesel
production because of their high harvesting efciency, innocuity
and biodegradability (Lee et al., 2009). Poly (c-glutamic acid)
(c-PGA) has been produced as an extracellular product of Bacillus
subtilis at large scales (Yokoi et al., 1996) and used commercially
as a microbial occulant in wastewater treatment (Taniguchi
et al., 2005). The purpose of this study was to evaluate c-PGA as
a occulant for harvesting lipid-rich microalgae by response surface methodology (RSM) and to investigate the effect of c-PGA
on harvested cell integrity using microscopy.
2. Methods
2.1. Materials

c-PGA produced by B. subtilis was purchased from Rundo

Biotech Japan Co. Ltd, Nanjing, China.

213

H. Zheng et al. / Bioresource Technology 112 (2012) 212220

2.2. Microalgal strains and cultivation conditions

Species of microalgae were obtained from the Culture Collection
of Algae at the University of Texas at Austin (Chlorella protothecoides UTEX 255 and Phaeodactylum tricornutum UTEX 640), the
China Center for Type Culture Collection at Wuhan in China (marine Chlorella vulgaris, strain CCTCC M 209256), and our laboratory
isolations (freshwater C. vulgaris LICME 001, Nannochloropsis
oculata LICME 002, and Botryococcus braunii LICME 003). Marine
C. vulgaris and N. oculata were grown in medium composed of (in
mg L1): KNO3, 100; KH2PO4, 10; Na2EDTA, 10; FeSO47H2O, 2.5;
MnSO4, 0.25; Vitamin B1, 0.006; Vitamin B12, 0.00005; instant
ocean synthetic sea salt (Aquarium Systems, Inc., USA), 26,000.
Freshwater C. vulgaris LICME 001, C. protothecoides, and B. braunii
LICME 003 were grown in BG-11 medium. P. tricornutum was
grown in F/2 medium. The media were autoclaved at 121 C for
20 min without pH adjustment. A 10 L bubble column photobioreactor (50.0 cm in height, 16.0 cm in diameter, a closed system) culture system with a working volume of 8 L was used. The culture
temperature of 25 C was regulated by water recycled in the outer
layer of the photobioreactor. Ten uorescent lamps were arranged
around the photobioreactor to supply continuous illumination of
80 lmol photons m2 s1 with a 12/12 h light/dark cycle. At the
bottom of the reactor, there was a gas sparger. CO2 of 3.0% was
prepared with a combination of room air and pure CO2 from a compressor and an aeration rate of 200 mL min1 was carried out. The
cultivation cycle was 15 days.
2.3. Analytical methods
The biomass concentrations (dry mass) of microalgae (BC, g L1)
were calculated from measurements of the optical density (OD) of
cultures at 680 nm according to the following equations: marine or
freshwater C. vulgaris BC = 0.560  OD680 (R2 = 0.986); N. oculata
BC = 0.580  OD680 (R2 = 0.995); P. tricornutum BC = 0.652  OD680
(R2 = 0.988); C. protothecoides BC = 0.558  OD680 (R2 = 0.994); B.
braunii BC = 0.885  OD680 (R2 = 0.991).
The microalgal suspension of 150 ml was placed into each of the
250 mL glass beakers, and the salinity of the media was adjusted
by addition of instant ocean synthetic sea salt or distilled water
to 10, 20, 30, 40, and 50 g L1 for the salinity effect experiment
(Figs. 1C and 2C) or according to Table 1. pH values were adjusted
to 6.5, 7.0, 7.5, 8.0, and 8.5 with 0.5 M HCl or 0.5 M NaOH for the
pH effect experiment, otherwise the pH was kept at 7.5. The initial
optical density of the microalgal suspension in the beakers was
measured at 680 nm. The c-PGA powder was added under magnetic stirring (HJ-3, Jiangsu Tianyou Co. Ltd, Jiangsu Province,
China) at a stirring rate of 500 rpm for 5 min. The microalgal suspension was left to settle for 2 h without agitation. Subsequently,
the optical density of the supernatant from half the height of the
claried layer and the sludge was measured. The occulation efciency was dened as the ratio of the mass of cells recovered to the
total mass of cells and the concentration factor was the ratio of the
nal product concentration to the initial concentration (Bosma
et al., 2003). The occulation efciency and concentration factor
were calculated as:

Flocculation efficiency %

Concentration factor

A2
A0

A0 V 0  A1 V 1
 100
A0 V 0

where A1 is OD680 of the supernatant from half the height of the

claried layer after occulation, A2 is OD680 of the sludge after

occulation, and A0 is OD680 of the microalgal suspension before

occulation. V0 is the volume of microalgal suspension before occulation, and V1 is the volume of microalgal supernatant after
occulation.
2.4. Experiment design
2.4.1. Evaluation of occulation parameters for C. vulgaris and C.
protothecoides
In order to optimize the occulation of microalgae with c-PGA,
C. vulgaris and C. protothecoides were used as model systems for
marine and freshwater microalgae, respectively. The effects of
the occulation parameters such as c-PGA dosage, biomass
concentration, pH and salinity on C. vulgaris and c-PGA dosage, biomass concentration and pH on C. protothecoides were individually
investigated by analyzing occulation efciency and concentration
factor.
The zeta potentials of the microalgal suspensions before occulation (1.2 g L1 biomass, pH 7.5 and 30 g L1 salinity for C. vulgaris
and 1.2 g L1 biomass and pH 7.5 for C. protothecoides) and those
of the occulated suspensions (obtained from the above c-PGA
dosage experiment containing 20 and 30 mg L1 c-PGA, respectively) were measured with a Zeta Potential Analyzer utilizing
phase analysis light scattering (Brookhaven Instruments Corporation, USA).
2.4.2. Optimization of occulation of C. vulgaris and C. protothecoides
with c-PGA
To improve occulation efciencies and concentration factors,
the interaction between the three most signicant factors (c-PGA
dosage, biomass concentration and salinity) for C. vulgaris and that
between the two most signicant factors (c-PGA dosage and biomass concentration) for C. protothecoides identied by preliminary
evaluation experiments were studied. Since it is known that RSM
can evaluate the interaction between the signicant factors of an
experiment and optimize them (Ghosh and Hallenbeck, 2010; Ji
et al., 2009), RSM using central composite design was applied to
determine the optimal levels of the three selected variables for C.
vulgaris and the two selected variables for C. protothecoides, which
signicantly affected the occulation efciency and concentration
factor. The three independent factors with ve different levels
(1.682, 1, 0,+1,+1.682) of C. vulgaris and the two independent
factors with ve different levels (1.414, 1, 0,+1,+1.414) of C.
protothecoides were investigated and the experimental designs
are shown in Tables 1 and 2. The factors were coded according to
the following equation:

xi

Xi  X0
;
DX

i 1; 2; 3; . . . ; k

where xi is the coded independent factor, Xi is the real independent

factor, X0 is the value of Xi at the center point and DX is the step
change value.
The occulation efciencies and concentration factors of c-PGA
were tted using a polynomial equation and four multiple regressions of the data were carried out to obtain four empirical models
related to the three and two most signicant factors in the case of
C. vulgaris and C. protothecoides, respectively. The general form of
the polynomial equation is:

Y b0

biXi

bii X 2i

bij X i X j ; . . . i; j 1; 2; 3; . . . ; k

where Y is the predicted response, Xi and Xj are independent factors,

b0 is the intercept, bi is the linear coefcient, bii is the quadratic
coefcient, and bij is the interaction coefcient.

214

H. Zheng et al. / Bioresource Technology 112 (2012) 212220

100

Flocculation efficiency (%)

a
80

ab

100
a

Flocculation efficiency (%)

d
60

40

20

80

d
e

60

40

20

0
10

15

20

25

0.4

30

0.8

-PGA dosage (mg L )

1.2

1.6

2.0

-1

-1

Biomass concentration (g L )

100

a
b

Flocculation efficiency (%)

80

c
60

40

20

0
10

20

30

40

50

CK1

CK2

-1

Salinity (g L )
Fig. 1. Effects of c-PGA dosage, biomass concentration and salinity on the occulation efciency of marine Chlorella vulgaris. The different letters in the graphs indicate a
signicant difference at P < 0.05. (A biomass concentration: 1.2 g L1, pH: 7.5 and salinity: 30 g L1; B c-PGA: 20 mg L1, pH: 7.5 and salinity: 30 g L1; C c-PGA: 20 mg L1,
biomass concentration: 1.2 g L1 and pH: 7.5; CK1 10 g L1 sea salt with 1.2 g L1 biomass concentration and pH of 7.5 without the addition of c-PGA; CK2 50 g L1 sea salt
with 1.2 g L1 biomass concentration and pH of 7.5 without the addition of c-PGA.).

To maximize the two response variables occulation efciency

and concentration factor simultaneously (m = 2), an optimization
using the global desirability function (D) was performed for
C. vulgaris and C. protothecoides, respectively, which consisted in
converting each response into a single desirability function (di)
ranging from 0 to 1 (0 6 di 6 1) (Derringer and Suich, 1980). The
individual desirabilitys were then combined using the geometric
mean, which gives the overall desirability D:

m
Y

!1=m
di

i1

Microalgal biomass samples harvested at optimal occulation

parameters and cells harvested before occulation (control), were
examined microscopically using a scanning ber-optic microscope
(Quanta 200, FEI Company, USA) and a Leica microscope (Leica DM
1000, Leica Microsystems, Germany). The powder forms of microalgae were sputter-coated with gold by the JFC-1600 auto ne
coater (JEOL Ltd., Tokyo, Japan) before observation using the scanning ber-optic microscope.
2.5. Data analysis and software
Statistical software Statistica 6.0 (StatSoft Inc., Oklahoma, USA)
was applied to the experimental design and statistical analysis of
the experimental data. The experiment was designed and carried

out at random. All the treatments were repeated three times and
data are reported as the mean SD values.
3. Results and discussion
3.1. Evaluation of occulation parameters
3.1.1. Effect of c-PGA dosage on occulation of C. vulgaris and C.
protothecoides
Figs. 1A and 2A show the effect of c-PGA dosage on the occulation efciency and the concentration factor for C. vulgaris. The optimal c-PGA dosage was 20 mg L1 with a occulation efciency of
82% and a concentration factor of 15.1. Both the occulation efciency and concentration factor increased signicantly (P < 0.05)
with increasing c-PGA dosage up to a concentration of 20 mg L1.
However, both occulation efciency and concentration factor decreased when the c-PGA dosage was increased above 20 mg L1.
Similar results were found for C. protothecoides occulation
(Fig. 3A and C) at an optimal c-PGA dosage of 20 mg L1 with a occulation efciency of 90% and a concentration factor of 23.7. Godos
et al. (2011) reported similar results for above and below optimum
dosages of ve polymeric occulants including chitosan. Our result
indicated that overdosing of c-PGA resulted in dispersion restabilization. Similar results were obtained by Vandamme et al. (2009).
The zeta potentials of the microalgal suspensions before occulation
were 19.08 and 13.62 mV for C. vulgaris and C. protothecoides,

215

H. Zheng et al. / Bioresource Technology 112 (2012) 212220

18

16

b
c

16

Concentration factor

Concentration factor

18

14
12

b
10

8
6

20

d
e

14
12

10
8
6

2
0

0
10

15

20

25

0.4

30

-PGA dosage (mg L )

0.8

1.2

1.6

2.0

-1

-1

Biomass concentration (g L )

20
18

a
b

16

Concentration factor

14
12

10
8
6

0
10

20

30

50

40

CK1

CK2

-1

Salinity (g L )
Fig. 2. Effects of c-PGA dosage, biomass concentration and salinity on the concentration factor of marine Chlorella vulgaris. (Same legends as in Fig. 1).

Table 1
The central composite design of RSM for optimization of the occulation parameters of marine Chlorella vulgaris with c-PGA.
Run

Factors

c-PGA dosage
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16

Biomass concentration

X1

P (mg L1)

X2

B (g L1)

X3

S (g L1)

1
1
1
1
1
1
1
1
1.682
1.682
0
0
0
0
0
0

15
15
15
15
25
25
25
25
11.59
28.41
20
20
20
20
20
20

1
1
1
1
1
1
1
1
0
0
1.682
1.682
0
0
0
0

0.5
0.5
1.5
1.5
0.5
0.5
1.5
1.5
1.0
1.0
0.16
1.84
1.0
1.0
1.0
1.0

1
1
1
1
1
1
1
1
0
0
0
0
1.682
1.682
0
0

10
30
10
30
10
30
10
30
20
20
20
20
3.18
36.82
20
20

respectively, and those of the corresponding occulated suspensions with optimal 20 and overdosing 30 mg L1 c-PGA were
+0.83 and +21.50 mV,+2.04 and +22.37 mV, respectively. These

Flocculation efciency (%)

Concentration factor

86 2
79 3
82 4
70 2
88 1
85 4
87 3
75 2
74 3
86 2
90 2
83 4
90 3
78 2
87 2
87 3

18.6 0.6
11.2 0.4
11.9 0.5
5.4 0.8
18.8 0.6
13.6 0.7
8.8 0.3
8.3 0.4
8.1 0.5
9.3 0.3
20.4 0.6
11.8 0.5
18.1 0.4
13.3 0.3
16.2 0.4
16.4 0.4

Salinity

results indicate that c-PGA could adsorb at the surface of the microalgae and such adsorption caused a reduction of surface potential by
charge neutralization and a resulting destabilization of the

216

H. Zheng et al. / Bioresource Technology 112 (2012) 212220

Table 2
The central composite design of RSM for optimization of the occulation parameters of freshwater Chlorella protothecoides with c-PGA.
Run

Factors

c-PGA dosage
1
2
3
4
5
6
7
8
9
10

P (mg L1)

X2

B (g L1)

1
1
1
1
1.414
1.414
0
0
0
0

15
15
25
25
12.93
27.07
20
20
20
20

1
1
1
1
0
0
1.414
1.414
0
0

0.5
1.5
0.5
1.5
1.0
1.0
0.29
1.71
1.0
1.0

100

60

40

20

100

b
d

80

60

40

20

10

15

20

25

0.4

30

D
a

2.0

b
24

c
Concentration factor

16

1.6

32
28

20

1.2

Biomass concentration (g L )

28

24

0.8

-1

-1

-PGA dosage (mg L )

Concentration factor

27.5 0.8
8.9 0.6
24.8 0.9
20.5 0.5
10.7 0.6
19.6 0.7
32.9 0.8
14.8 0.4
25.4 0.5
25.5 0.6

12

92 2
80 2
93 1
87 2
87 3
90 1
96 2
82 1
94 1
94 2

Flocculation efficiency (%)

Flocculation efficiency (%)

Concentration factor

Biomass concentration

X1

80

Flocculation efciency (%)

d
e

20
16

d
12
8
4
0

10

15

20

25

30

-1

-PGA dosage (mg L )

0.4

0.8

1.2

1.6

2.0

-1

Biomass concentration (g L )

Fig. 3. Effects of c-PGA dosage and biomass concentration on the occulation efciency and concentration factor of freshwater Chlorella protothecoides. The different letters in
the graphs indicate a signicant difference at P < 0.05. (A and C biomass concentration: 1.2 g L1 and pH: 7.5; B and D c-PGA: 20 mg L1 and pH: 7.5).

microalgae. Continuous adsorption beyond the point of charge neutralization by overdosing c-PGA caused charge reversal and restabilization occured.
3.1.2. Effect of biomass concentration on occulation of C. vulgaris and
C. protothecoides
Flocculation efciency and concentration factor of c-PGA as a
function of biomass concentration are shown in Figs. 1B and 2B.
Biomass concentration was strongly correlated with occulation
efciency and concentration factor as both decreased signicantly

(P < 0.05) with increasing biomass concentration. When the biomass concentration increased from 0.4 to 2.0 g L1, the occulation
efciency decreased from 89% to 65% and the concentration factor
decreased from 17.1 to 9.8. Similar results were found for
C. protothecoides occulation (Fig. 3B and D). The occulation
mechanisms of microbial occulants were not well established
(Esser and Kues, 1983), but a series of occulation mechanisms
of microbial occulants, like charge neutralization, bridging,
sweep-out and precipitation enmeshment (Divakaran and Pillai,
2002; Salehizadeh and Shojaosadati, 2001; Strand et al., 2002),

217

H. Zheng et al. / Bioresource Technology 112 (2012) 212220

have been proposed. Individual microalgal cells were visible in the

microalgal suspensions (Supplementary Fig. S1A and S1B) and the
cells were interlaced with c-PGA in ocs, indicating inter-cell
bridging between microalgal cells (Supplementary Fig. S1C and
S1D). Based on the observations of zeta potentials and SEM images
of the suspensions, the occulation mechanisms were likely
mainly cell aggregation by charge neutralization and bridging with
c-PGA, but more detailed investigations are needed to further validate this hypothesis.
3.1.3. Effect of pH on occulation of C. vulgaris and C. protothecoides
The surface electric property of the particles for occulation in
the suspension changed with pH, which inuences occulation
with microbial occulants (Chaiwong and Nuntiya, 2008). c-PGA
is a homopolymer of D- and L-glutamic acid units produced by B.
subtilis (Shih and Van, 2001), and the dissolution of c-PGA in
microalgal suspension may be inuenced by pH. Microalgae were
harvested at the late logarithmic phase of growth and the pH
values of the culture media for C. vulgaris and C. protothecoides were
approximately 8.4 and 7.8, respectively. In order to investigate the
effect of pH on occulation efciency and concentration factor
with c-PGA, pHs of 6.5, 7.0, 7.5, 8.0 and 8.5 were evaluated. Flocculation efciencies of C. vulgaris were approximately 81% and
concentration factors were approximately 15.2 with pH values
ranging from 6.5 to 8.5. Flocculation efciencies (89%) and
concentration factors (23.6) of C. protothecoides varied little within the same pH range. This demonstrates that pH had little effect
on occulation efciency and concentration factor. Yokoi et al.
(1996) also reported high occulation activity for a kaolin suspension with c-PGA and only small changes were observed when the
pH changed from 6.0 to 8.0.
3.1.4. Effect of salinity on occulation of C. vulgaris
High salinity is an important feature of culture media for marine microalgae. Sukenik et al. (1988) reported that microalgal occulation with cationic polymers was inhibited by the high ionic
strength of sea water. Figs. 1C and 2C show the occulation efciency and concentration factor of c-PGA with salinity levels of
10, 20, 30, 40 and 50 g L1 for C. vulgaris. Both the occulation efciency and concentration factor of c-PGA decreased signicantly
(P < 0.05) with increasing salinity and a maximum efciency of
88% and a maximum concentration factor of 17.4 were obtained
at a salinity of 10 g L1, which was the lowest salinity tested. In order to study the effect of salinity on C. vulgaris occulation without
addition of c-PGA, microalgal suspensions with a salinity of 10
(CK1) and 50 g L1 (CK2) were designed as controls. The results
showed that salinity had little effect on C. vulgaris occulation
without c-PGA (Figs. 1C, 2C and Fig. 4). Increasing salinity inhibited occulation with c-PGA thus salinity was one of the most
important occulation parameters for C. vulgaris. This result might
be explained by increasing salinity affecting the conformation of cPGA and higher sea salt concentration (ionic strength) causing the
chain of c-PGA to adopt a random coil arrangement (He et al.,
2000; Shih and Van, 2001), which induces a loose structure of
the ocs, resulting in a decrease in occulation efciency (Bajaj
and Singhal, 2011).
3.2. Optimization of occulation of C. vulgaris and C. protothecoides
with c-PGA
Since c-PGA dosage, biomass concentration and salinity had
highly signicant effects (P < 0.01) on occulation of C. vulgaris
and c-PGA dosage and biomass concentration had highly signicant effects (P < 0.01) on occulation of C. protothecoides with
c-PGA, it was desirable to investigate the interaction between
the three most signicant factors for C. vulgaris and the two most

signicant factors for C. protothecoides and optimize them in an attempt to obtain higher occulation efciencies and concentration
factors.
The results from the optimization experiments were analyzed
by standard ANOVA and the central composite design was tted
with the polynomial equations:
C. vulgaris

Flocculation efficiency 0:3362 0:0489x1  0:0012X 21

0:0451x2  0:0237X 22 0:0034x3
 0:0001X 23 0:0010x1 x2 0:0001X 1 X 3
 0:0035X 2 X 3  100%

Concentration factor 17:5716 4:5617x1  0:1192X 21

 3:2846x2  1:4524X 22  0:5508x3
 0:0050X 23  0:1400x1 x2
0:0205x1 x3 0:1400x2 x3

C. protothecoides

Flocculation efficiency 0:5158 0:0441x1  0:0012X 21

0:0005x2  0:1075X 22
0:0060x1 x2  100%

Concentration factor 25:5439 6:9772x1  0:1967X 21

 36:1743x2  2:2750X 22 1:4300x1 x2

where X1, X2 and X3 are c-PGA dosage, biomass concentration and

salinity (all for real values), respectively.
The t of the models was checked by the coefcients of determination R2, which were calculated to be 0.96, 0.99, 0.94 and
0.99, implying that 96%, 99%, 94% and 99% of the variability in
the response could be explained by Eqs. (6)(9) (Table 3). The statistical signicance of the model equations was evaluated by the Ftest for ANOVA. The model F-values were more than 13.00 and
their very low P-values (P < 0.05) indicated that all the models
were signicant. There was less than 5% chance that every model
with an F-value this large could result from noise. The lack of t
F-values of less than 200.10 implied that there was no less than
5% chance that every lack of t F-value could occur due to noise.
These results indicated that the models were suitable to describe
the relationships between occulation efciency and the signicant factors and between concentration factor and the signicant
factors. The regression models developed can be represented in
3-D response surface plots to gain a better understanding of the
interaction between the variables and to determine the optimum
level of each variable for maximum response (Supplementary
Fig. S2S4).
In this study, with the aim of achieving high values of occulation efciency and concentration factor, a contradiction in parameter settings is evident between the models. In practice, high
efciency is more important than a high concentration factor,
otherwise biomass is lost (Bosma et al., 2003). Based on the results
of RSM, occulation was further optimized by the application of
the global desirability function. The combinations predicted by
the application of the global desirability function were,
22.03 mg L1 c-PGA, 0.57 g L1 biomass, and 11.56 g L1 salinity
for C. vulgaris and 19.82 mg L1 c-PGA and 0.60 g L1 biomass for
C. protothecoides. The values predicted for the responses were
occulation efciencies of 91 and 97% and concentration factors
of 20.7 and 29.5 for C. vulgaris and C. protothecoides, respectively.
In order to conrm the optimization results, occulation was
studied using the optimal occulation parameters (c-PGA dosage

218

H. Zheng et al. / Bioresource Technology 112 (2012) 212220

Table 3
ANOVA for the response surface models.
Source
Marine Chlorella vulgaris
Flocculation efciencya
Model
Residual
Lack of t
Pure error
Total
Concentration factorb
Model
Residual
Lack of t
Pure error
Total
Freshwater Chlorella protothecoides
Flocculation efciencyc
Model
Residual
Lack of t
Pure error
Total
Concentration factord
Model
Residual
Lack of t
Pure error
Total
a
b
c
d

The
The
The
The

coefcient
coefcient
coefcient
coefcient

of
of
of
of

determination
determination
determination
determination

(R2)
(R2)
(R2)
(R2)

of
of
of
of

the
the
the
the

model
model
model
model

Sum of squares

DF

500.62
20.32
19.82
0.50
520.94

9
6
5
1
15

325.55
2.05
2.03
0.02
327.60

F-value

p-value

55.62
3.39
3.96
0.50

16.43

0.0016

7.93

0.2630

9
6
5
1
15

36.17
0.34
0.41
0.02

105.94

<0.0001

20.29

0.1669

194.46
11.94
11.44
0.50
206.40

5
4
3
1
9

38.89
2.99
3.81
0.50

13.03

0.0138

7.63

0.2587

635.21
3.01
3.00
0.01
638.22

5
4
3
1
9

127.04
0.75
1.00
0.01

169.03

<0.0001

200.09

0.0519

was
was
was
was

0.96.
0.99.
0.94.
0.99.

22.03 mg L1, biomass concentration 0.57 g L1, and salinity

11.56 g L1 for C. vulgaris and 19.82 mg L1 c-PGA and 0.60 g L1
biomass for C. protothecoides). Maximum occulation efciencies
under optimal occulation parameters were observed at 2 h of
91 and 98% for C. vulgaris and C. protothecoides (Fig. 4), respectively.
Their corresponding maximum concentration factors were 20.5
and 29.8, respectively. These results were in good agreement with
the predicted values. Both occulation efciencies greater than
90% and concentration factors exceeding 20.0 demonstrated the

100
90
80

Flocculation efficiency (%)

Mean square

70
60

feasibility of c-PGA as a promising microbial occulant for harvesting microalgae.

3.3. Application of c-PGA as a microbial occulant to other microalgal
species
In an attempt to verify that the optimal occulation parameters
with c-PGA were applicable to other microalgae, two marine microalgal species (P. tricornutum and N. oculata LICME 002) and two
freshwater species ( C. vulgaris LICME 001 and B. braunii LICME
003) were occulated using the optimal occulation parameters
of marine C. vulgaris and freshwater C. protothecoides with c-PGA,
respectively. The occulation efciencies and concentration factors
for C. vulgaris LICME 001, B. braunii LICME 003, P. tricornutum and
N. oculata LICME 002 were 90% and 20.1, 92% and 21.4, 97% and
28.2, and 96% and 27.6, respectively, indicating effectiveness of
occulation with c-PGA for harvesting microalgae.
3.4. Effect of c-PGA on cell integrity

50
40
30
20
10
0
0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

2.2

Flocculation time (h)

Fig. 4. Time course of the occulation under the optimized parameters. (j
occulation of marine Chlorella vulgaris with c-PGA; d occulation of freshwater
Chlorella protothecoides with c-PGA; N Control: marine Chlorella vulgaris without
the addition of c-PGA;  Control: freshwater Chlorella protothecoides without the
addition of c-PGA).

The harvesting process may cause cell disruption and affect

downstream processing and lipid recovery. In order to assess the
impact of c-PGA on the microalgal biomass harvest, the direct effects of c-PGA on the cell wall of marine C. vulgaris (Supplementary
Fig. S1AD) and the other ve microalgae (data not shown) were
observed using scanning electron and light microscopes. Supplementary Fig. S1A and S1B show the state of microalgal cells before
the addition of c-PGA, and Supplementary Fig. S1C and S1D show
the state of microalgal cells after occulation with c-PGA. Comparing Supplementary Fig. S1A and S1C, it can be easily demonstrated
that c-PGA occulates microalgal cells with very little visual
change in their morphology. In a previous study (Zheng et al.,
2011), the structure of disrupted cells of marine C. vulgaris showed
signicant deformation (Supplementary Fig. S1E) compared with
intact cells. In addition, c-PGA had very little effect on the

219

H. Zheng et al. / Bioresource Technology 112 (2012) 212220

Table 4
Comparison of harvesting efciencies of different methods.
Methods

Microlgal species

Marine/freshwater microalgae

Harvesting
efciencies (%)

References

Flocculation with c-PGA

Chlorella protothecoides, Chlorella vulgaris

LICME 001, and Botryococcus braunii LICME 003
Chlorella vulgaris, Nannochloropsis oculata LICME 002,
and Phaeodactylum tricornutum
Thalassiosira pseudonana
Chlorella minutissima
Phaeodactylum tricornutum

Freshwater microalgae

>95

Current study

Marine microalgae

>90

Current study

Marine microalga
Freshwater microalga
Marine microalga

90
>90
94

Heasman et al. (2000)

Papazi et al. (2009)
Heasman et al. (2000)

Flocculation with c-PGA

Flocculation with chitosan
Flocculation with AlCl3
Centrifugation

morphology of the other microalgal cells. These results indicate

that occulated microalgal cells with c-PGA did not show lysis.
Similar results were obtained by Divakaran and Pillai (2002) using
chitosan as a microbial occulant for harvesting Spirulina, Oscillatoria, Chlorella and Synechocystis. The lipids of the cells in our study
would not be lost during the occulation process.
3.5. Comparison of microalgae harvesting efciency with c-PGA and
conventional harvesting methods
The optimal occulation method with c-PGA evaluated in this
work was compared with some conventional harvesting methods
(Table 4). The harvesting efciencies in this study showed no signicant difference compared with those of the conventional harvesting methods (P > 0.05). c-PGA was able to occulate marine
and freshwater microalgae. Moreover, the microalgal cells were intact and no metallic occulants were used. The price of c-PGA applied in this work is approximately 5 US dollars per kg, which is
sufcient to treat up to 45,000 L of microalgal suspensions. However, it is also noteworthy to point out that the products of cPGA from different bacterial species may have different harvesting
performances for different microalgae, an area requiring further
research.

4. Conclusion
The work focused on optimizing occulation parameters of
marine C. vulgaris and freshwater C. protothecoides with c-PGA. A
maximum occulation efciency and concentration factor of 91%
and 20.5 of C. vulgaris and 98% and 29.8 of C. protothecoides, respectively, were obtained. The optimal occulation parameters of cPGA dosage, biomass concentration and salinity for C. vulgaris
and c-PGA dosage and biomass concentration for C. protothecoides
were successfully applied to harvest other microalgae. c-PGA had
little effect on microalgal cell integrity. Our results demonstrate
that c-PGA has potential as an efcient and sustainable microbial
occulant for harvesting microalgae in biodiesel production.

Acknowledgements
This work was supported by the Major State Basic Research
Development Program of China (973 Project) (Grant Nos.
2011CB200904 and 2011CB200906), and our sincere thanks to
Dr. Ailish OHalloran from Institute of Technology Tallaght, Ireland
for her language assistance.

Appendix A. Supplementary data

Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.biortech.2012.02.086.

References
Bajaj, I.B., Singhal, R.S., 2011. Flocculation properties of poly-(c-glutamic acid)
produced from Bacillus subtilis isolate. Food and Bioprocess Technology 4, 745
752.
Bosma, R., Spronsen, W., Tramper, J., Wiffels, R., 2003. Ultrasound, a new separation
technique to harvest microalgae. Journal of Applied Phycology 15, 143
153.
Chaiwong, N., Nuntiya, A., 2008. Inuence of pH, electrolytes and polymers on
occulation of kaolin particle. Chiang Mai J. Sci. 35 (1), 1116.
Chiu, S.Y., Kao, C.Y., Tsai, M.T., Ong, S.C., Chen, C.H., Lin, C.S., 2009. Lipid
accumulation and CO2 utilization of Nannochloropsis oculata in response to
CO2 aeration. Bioresource Technology 100, 833838.
Derringer, G.C., Suich, R., 1980. Simultaneous optimization of several response
variables. J. Qual. Technol. 12, 214219.
Divakaran, R., Pillai, V.N.S., 2002. Flocculation of algae using chitosan. Journal of
Applied Phycology 14, 419422.
Esser, K., Kues, U., 1983. Flocculation and its implication for biotechnology. Process
Biochemistry 18, 2123.
Fu, C.C., Su, C.H., Hung, T.C., Hsieh, C.H., Suryani, D., Wu, W.T., 2009. Effects of
biomass weight and light intensity on the performance of photosynthetic
microbial fuel cells with Spirulina platensis. Bioresource Technology 100, 4183
4186.
Ghosh, D., Hallenbeck, P.C., 2010. Response surface methodology for process
parameter optimization of hydrogen yield by the metabolically engineered
strain Escherichia coli DJT135. Bioresource Technology 101, 18201825.
Godos, I., Guzman, H.O., Soto, R., Garca-Encina, P.A., Becares, E., Mu, R., Vargas, V.A.,
2011. Coagulation/occulation-based removal of algalbacterial biomass from
piggery wastewater treatment. Bioresource Technology 102, 923927.
Grima, E.M., Belarbi, E.H., Acin Fernndez, F.G., Medinaa, A.R., Chisti, Y., 2003.
Recovery of microalgal biomass and metabolites: process options and
economics. Biotechnology Advances 20, 491515.
He, L.M., Neu, M.P., Vanderberg, L.A., 2000. Bacillus lichenformis c-Glutamyl
Exopolymer: physicochemical characterization and U (VI) interaction.
Environmental Science and Technology 34, 16941701.
Heasman, M., Diemar, J., O Connor, W., Sushames, T., Foulkes, L., Nell, J.A., 2000.
Development of extended shelf-life microalgae concentrate diets harvested by
centrifugation for bivalve molluscsa summary. Aquaculture Res. 31 (89),
637659.
Hsieh, C.H., Wu, W.T., 2009. Cultivation of microalgae for oil production with
cultivation strategy of urea limitation. Bioresource Technology 100, 3921
3926.
Ji, X.J., Huang, H., Du, J., Zhu, J.G., Ren, L.J., Li, S., Nie, Z.K., 2009. Development of an
industrial medium for economical 2,3-butanediol production through
cofermentation of glucose and xylose by Klebsiella oxytoca. Bioresource
Technology 100, 52145218.
Lee, A.K., Lewis, D.M., Ashman, P.J., 2009. Microbial occulation, a potentially lowcost harvesting technique for marine microalgae for the production of biodiesel.
Journal of Applied Phycology 21 (5), 559567.
Oh, H.M., Lee, S.J., Park, M.H., Kim, H.S., Kim, H.C., Yoon, J.H., Kwon, G.S., Yoon, B.D.,
2001. Harvesting of Chlorella vulgaris using a bioocculant from Paenibacillus sp.
AM49. Biotechnology Letters 23, 12291234.
Papazi, A., Makridis, P., Divanach, P., 2009. Harvesting Chlorella minutissima using
cell coagulants. Journal of Applied Phycology 22 (3), 349355.
Pushparaj, B., Pelosi, E., Torzillo, G., Materassi, R., 1993. Microbial biomass recovery
using a synthetic cationic polymer. Bioresource Technology 43, 5962.
Salehizadeh, H., Shojaosadati, S.A., 2001. Extracellular biopolymeric occulants
recent trends and biotechnological importance. Biotechnology Advances 19,
371385.
Shih, I.L., Van, Y.T., 2001. The production of poly-(c-glutamic acid) from
microorganisms and its various applications. Bioresource Technology 79,
207225.
Sialve, B., Bernet, N., Bernard, O., 2009. Anaerobic digestion of microalgae as a
necessary step to make microalgal biodiesel sustainable. Biotechnology
Advances 27, 409416.
Strand, S.P., Nordengen, N., tgaard, K., 2002. Efciency of chitosans applied for
occulation of different bacteria. Water Research 36 (19), 47454752.
Sukenik, A., Bilanovic, D., Shelef, G., 1988. Flocculation of microalgae in brackish and
sea waters. Biomass 15, 187199.

220

H. Zheng et al. / Bioresource Technology 112 (2012) 212220

Taniguchi, M., Kato, K., Matsui, O., Xu, P., Nakayama, H., Usuki, Y., Ichimura, A.,
Fujita, K., Tanaka, T., Tarui, Y., Hirasawa, E., 2005. Flocculating activity of crosslinked poly-c-glutamic acid against Bentonite and Escherichia coli suspension
pretreated with FeCl3 and its interaction with Fe3+. Journal of Bioscience and
Bioengineering 100 (2), 207215.
Vandamme, D., Foubert, I., Meesschaert, B., Muylaert, K., 2009. Flocculation of
microalgae using cationic starch. Journal of Applied Phycology 22 (4), 525530.

Yokoi, H., Arima, T., Hirose, J., Hayashi, S., Takasaki, Y., 1996. Flocculation properties
of poly (-glutamic acid) produced by Bacillus subtilis. Journal of Fermentation
and Bioengineering 82 (1), 8487.
Zheng, H.L., Yin, J.L., Gao, Z., Huang, H., Ji, X.J., Dou, C., 2011. Disruption of Chlorella
vulgaris cells for the release of biodiesel-producing lipids: a comparison of
grinding, ultrasonication, bead milling, enzymatic lysis and microwaves.
Applied Biochemistry and Biotechnology 164, 12151224.