COMMUNICATION TO THE EDITOR
Investigation of Photobioreactor Design
for Enhancing the Photosynthetic
Productivity of Microalgae
Masahiko Morita, Yoshitomo Watanabe, Hiroshi Saiki
Bio-Science Department, Abiko Research Laboratory, Central Research
Institute of Electric Power Industry (CRIEPI), 1646 Abiko, Abiko-shi, Chiba
270-1194, Japan; telephone: +81-471-82-1181; fax: +81-471-83-3347; e-mail:
y-wata@criepi.denken.or.jp
Received 4 December 1999; accepted 23 April 2000
Abstract: As photosynthetic efficiencies are relatively
high at irradiation levels of <500 µmol m−2 s−1, photo-
synthetic productivity could be increased by redistribut-
ing strong light over a larger photo-receiving area using
conical, helical, tubular photobioreactors (HTP). When
Chlorella were exposed to light irradiation of 980 µmol
m−2 s−1, the ratio of productivities was 1.00:1.13:1.23:1.66
for conical HTPs with cone angles of 180° (flat type), 120°,
90°, and 60°, respectively. This suggests that photo-
redistribution technology is a highly effective and conve-
nient approach for increasing the photosynthetic produc-
tivity of microalgae. © 2000 John Wiley & Sons, Inc. Biotech-
nol Bioeng 69: 693–698, 2000.
Keywords: microalgae; CO2; photosynthesis; photobio-
reactor; photo-redistribution
INTRODUCTION
Microalgae have a greater capacity to assimilate CO2
through photosynthesis than higher plants and are capable
of synthesizing a variety of valuable substances (Benemann
et al., 1987). Microalgal CO2 fixation/conversion technol-
ogy is one approach for effectively utilizing CO2 in indus-
trial waste gases such as stack-emitted gases from thermal
power plants. The fixation product, algal biomass, could
then be used as livestock fodder, replacing forage crops,
thereby reducing greenhouse gases (N2O and CH4) indi-
rectly from crop fields, while helping to ameliorate the
greenhouse effect (Watanabe et al., 1992).
The photosynthetic productivity of microalgae is depen-
dent on the efficiency with which it utilizes light energy.
Photosynthetic efficiency is calculated from the energy ex-
pended synthesizing hexose: under standard conditions
(600-nm illumination), 3 mol of adenosine triphosphate
(ATP) and 2 mol of nicotinamide adenine dinucleotide
phosphate (NADPH) are used in synthesizing 1 mol of hex-
ose from CO2, assuming a theoretical efficiency of photo-
synthesis of 30% (Stryer, 1988). Unfortunately, this level of
efficiency can only be achieved under low irradiation; pho-
Correspondence to: Y. Wantanabe
© 2000 John Wiley & Sons, Inc.
tosynthetic productivity is usually diminished under strong
irradiation as a consequence of light saturation.
It would be useful if the genes controlling photosynthetic
reactions could be manipulated so as to improve photosyn-
thetic efficiency under strong irradiation, but this technol-
ogy is not presently available. As an alternative, redistrib-
uting light energy by introducing light-diffusing optical fi-
bers into the light path has proven to be effective (Mori,
1985). A somewhat different approach to photo-redis-
tribution is spatial redistribution, in which input light energy
is redistributed to a larger photo-receiving area. Using this
approach to increase the photosynthetic efficiency of mi-
croalgae, we designed a conical HTP (Watanabe and Hall,
1996) that increases photosynthetic productivity by spread-
ing light energy over a larger photo-receiving area. In this
study, we used both theoretical and experimental ap-
proaches to investigate the effect of photobioreactor design
on photo-redistribution and photosynthetic efficiency.
THEORY
The conical HTP is a spatial photo-redistribution photobio-
reactor that should, in principle, increase photosynthetic
productivity by redistributing high levels of light energy
over a larger photo-receiving area. Figure 1a shows the
theoretical relationship between light intensity and photo-
synthetic productivity (Benemann and Oswald, 1994): At
low light intensities, photosynthetic productivity increases
with light intensity until it eventually saturates and then
declines as light intensity continues to increase (solid line).
The expected increase in productivity achievable through
the use of a photo-redistribution photobioreactor is exem-
plified in Figure 1b. If the input light intensity is ␣, the
intensity of the light received by a flat reactor with an area,
S, will be the same ␣. On the other hand, in a conical HTP
with a 60° cone angle, the photo-receiving area is 2S, and
the light intensity received is only ␣/2; that is, the light
received has been diluted by a factor of 2 over the receiving
larger area. Thus, the expected photosynthetic productivities
of the flat and conical HTP are X and 2Y, respectively.

Figure 1. Concept underlying increased photosynthetic productivity
through photo-redistribution and schematic depiction of photo-
redistribution photobioreactors. (a) General relationship between light in-
tensity and photosynthetic productivity. (b) Comparison of the expected
products obtained using photo-redistribution photobioreactors employing
photostages of 180° (flat) or 60° (conical). (c) Schematic depiction of our
conical HTP system, which is comprised of ten parts: (1) a conical pho-
tostage made of transparent polyvinylchloride (PVC) tubing (internal di-
ameter 16 mm; wall thickness 2 mm; cone angles 180°, 120°, 90°, and
60°); (2) a passage (PVC tubing) for circulating liquid medium; (3) a heat
exchanger consisting of 11.6 m of PVC tubing set in a water bath; (4) a
degasser; (5) three metal halide lamps; (6) three timers; (7) an air pump; (8)
a CO2 gas cylinder; (9) a gas flowmeter; and (10) a cooling device. Arrows
indicate the direction of air or fluid flow.
MATERIALS AND METHODS
Organisms and Culture Medium
Chlorella sp. strain HA-1 (Watanabe et al., 1992) was
grown in M4N medium with the following composition
(milligrams per liter): KNO 3 , 5000; KH 2 PO 4 , 1250;
MgSO4 ⭈ 7H2O, 2500; FeSO4 ⭈ 7H2O, 3; H3BO3, 2.86;
MnCl2 ⭈ 4H2O, 1.81; ZnSO4 ⭈ 7H2O, 0.22; CuSO4 ⭈ 5H2O,
0.08; Na2MoO4, 0.021. All nutrients were dissolved in dis-
tilled water; the initial pH was 6.0.
Culture Systems and Operations
A culture system involving a single circular photostage
(0.01 m2) was used to investigate the relationship between
light intensity and photosynthetic productivity. The photo-
694 BIOTECHNOLOGY AND BIOENGINEERING, VOL. 69, NO. 6, SEPTEMBER 20, 2000
stage tube length was 0.5 m and the internal diameter of the
photostage was 0.16 m. Ten-day batch cultures were carried
out in triplicate experiments. The photostage was illumi-
nated from above for 12 h day−1; light intensities of 200,
500, 750, 1000, 1200, 1600, and 2000 ␮mol m−2 s−1 were
achieved by varying the distance between a metal halide
lamp and the photostage. The cell concentration of the in-
oculum was 0.2 g dry biomass L−1. A total of 0.45 L of
medium was used, and the air flow rate was 0.2 L min−1.
The batch cultures were carried out under an atmosphere of
10% CO2-enriched air at 25°C.
Four conical HTPs (Fig. 1c) with cone angles of 180°,
120°, 90°, and 60°; with corresponding photo-receiving ar-
eas of 0.50, 0.58, 0.71 and 1.0 m2; and with photostage tube
lengths of 25.9, 27.7, 34.4, and 49.0 m, were used to inves-
tigate the effect of photo-redistribution. All the reactors had
the same installation area (0.50 m2), and the cultures were
illuminated for 12 h day−1. The outer surface of the photo-
stage was covered with aluminum foil to prevent the ab-
sorption of outside light. The average light intensity at the
level of the top of the photostage was found to be 980 ␮mol
m−2 s−1 using a quantum meter (LI-189; Li-Cor Corp.). This
light intensity corresponds to the average photosynthetic
photon flux density outdoors between April and September
in central Japan. By measuring light intensities at a number
of selected points on the photostages, the average light in-
tensity at the receiving areas was determined. Batch cultures
were carried out at 28 ± 3°C for 8 days using 10% CO2-
enriched air supplied at various flow rates. The cell concen-
tration of the inoculum added to the conical HTPs was 0.3
g dry biomass L−1. A total of 14 L of medium was used.
Analytical Methods
Medium from each culture was sampled every 12 h for
analysis. The dry weights of samples of Chlorella biomass
were measured in suspension as a function of the absor-
bance at 750 nm using a spectrophotometer (Model DU650;
Beckman Instruments, Inc.). The relationship between dry
weight and absorbance was as follows: dry weight (g L−1)
⳱; 0.270 × OD750 (OD750 < 0.3). To determine carbon
content, batch cultures were centrifuged, and the harvested
cells were washed, dried at 105°C for 24 h, and crushed into
a powder. The carbon content in the powdered biomass was
determined using an elemental analyzer (Sumigraph NC-
800; Sumika Analysis Service). The pH of the culture me-
dium was measured with a pH sensor.
RESULTS AND DISCUSSION
Relationship between Light Intensity and
Photosynthetic Productivity
Initially, the relationship between light intensity and photo-
synthetic productivity was determined so that an equation
describing the effect of photo-redistribution could be em-
pirically derived. This was accomplished using Chlorella
sp. strain HA-1 cultured on a circular photostage (Fig. 2).
The data were taken from the latter logarithmic phase, and
this value was maximal during 10-day batch culture. The
culture period seemed to be enough time given for light
adaptation because the growth was recovered under strong
light after a decrease in the earlier period. Photosynthetic
productivity was found to increase with increasing light
intensity up to 500 ␮mol m−2 s−1 (Fig. 2, area A), at which
level light energy was the limiting factor. Photosynthetic
productivity remained almost constant between 500 ␮mol
m−2 s−1 and 1200 ␮mol m−2 s−1 (Fig. 2, area B); beyond that
level, productivity declined with increasing light intensity
(Fig. 2, area C).
Theoretically, for the linear ascending and plateau phases
of the curve (Fig. 2, areas A and B), the relationship be-
tween light intensity and photosynthetic productivity can be
expressed as:
P = PMI Ⲑ 共KS + I兲 − KM 共I ⱕ IC兲 (1)
−1 −1
where P is the photosynthetic productivity (g L day ), PM
is a constant (g L−1 day−1), I is the light intensity (␮mol m−2
s−1), KS is the light-saturation constant (␮mol m−2 s−1), KM
is a constant related to maintenance metabolism (g L−1
day−1), and IC is the light intensity at which inhibition
emerges (␮mol m−2 s−1). Because the degree of light inhi-
bition would be related to the amount of surplus light energy
(I − IC), in the descending phase of the curve (Fig. 2, area
C), the relationship between light intensity and photosyn-
thetic productivity can be expressed as:
P = PMI 兵1 − C共I − IC兲其 Ⲑ 共KS + I兲 − KM 共I ⱖ IC兲 (2)
where C is a constant (m s ␮mol ).
2 −1
The empirical equations describing the relationship be-
tween light intensity and photosynthetic productivity were
obtained as follows. IC was estimated to be 1200 ␮mol m−2
Figure 2. Relationship between light intensity and photosynthetic pro-
ductivity. (䊉) Experimental results where each data point represents the
mean of triplicate experiments; (—) outcome predicted by our theoretical
calculation.
s−1 based on the experimental results, whereas KM was es-
timated to be 0.044 g L−1 day−1 from the average decrease
in biomass during a period of darkness. Using the IC and KM
values obtained, a curve fitting Eq. (1) to the experimental
data was calculated for light intensities below 1200 ␮mol
m−2 s−1. From the fitted curve, PM and KS were estimated to
be 0.248 g L−1 day−1 and 108 ␮mol m−2 s−1, respectively.
Using these values, a curve fitting Eq. (2) to the experimen-
tal data was generated for light intensities beyond 1200
␮mol m−2 s−1, and C was estimated to be 0.00089 m2 s
␮mol−1, yielding the following equations:
P = 0.248 I Ⲑ 共108 + I兲 − 0.044 共I ⱕ 1200兲 (3)
P = 0.248 I 兵1 − 0.00089 共I − 1200兲其 Ⲑ 共108 + I兲 (4)
− 0.044 共I ⱖ 1200兲
The fitted curves are shown as a solid line in Figure 2; note
the good agreement between the calculated curve and the
experimental results.
Theoretically Derived Photo-Redistribution Effect
Photosynthetic productivity is primarily affected by the
strength of solar radiation. During summer, light saturation
reduces the efficiency of light-energy utilization, even at
median latitudes. In the case of Chlorella sp. strain HA-1,
light saturation was observed at irradiation levels >500
␮mol m−2 s−1. To increase photosynthetic productivity,
therefore, bioreactors must be configured such that strong
light input is in some way diluted. For example, at an input
light intensity of 1000 ␮mol m−2 s−1, the intensity of the
light received by a flat reactor is unchanged, and the pho-
tosynthetic productivity is 0.18 g L−1 day−1 (Fig. 2). By
contrast, at the same input intensity, light received by a 60°
conical HTP is only 500 ␮mol m−2 s−1 (a twofold redistri-
bution), because the photo-receiving area is twice as large.
As a result, the photosynthetic productivity in the conical
HTP is 0.32 (0.16 × 2) g L−1 day−1, a 1.78-fold increase
over the flat reactor.
The photo-redistribution described in the aforementioned
example was accomplished using the cone angle to expand
the photo-receiving area. Figure 3a shows the effect of cone
angle on photo-receiving area and photostage tube length.
Each increases as cone angle decreases, which means that,
as cone angle becomes smaller, the photo-redistribution ef-
fect becomes larger. Using Eq. (3) and the curve shown in
Figure 3a, the relationship between cone angle and photo-
synthetic productivity at a light irradiation of 980 ␮mol m−2
s−1 was calculated and normalized to the photosynthetic
productivity of a flat (180°) reactor (Fig. 3b). The relative
photosynthetic productivity per reactor was then estimated
by multiplying the normalized photosynthetic productivity
per unit photo-receiving area, which increased with increas-
ing cone angle until saturation, by the photo-receiving area.
Reactor productivity was thus found to increase dramati-
cally near 0°, then gradually decline with increasing cone
angle. The theoretical ratio of the photosynthetic produc-
tivities was 1.00:1.13:1.34:1.78 for the 180°, 120°, 90°, and
60° reactors, respectively. In general terms, for cone angles
COMMUNICATION TO THE EDITOR 695
 comitant changes in the pH of the culture medium are
shown in Figure 4. Maximum photosynthetic productivity
was obtained at air flow rates of 0.6, 0.3, 0.6, and 1.8 L
min−1 in the 180°, 120°, 90°, and 60° reactors, respectively
(data not shown). The highest productivity was achieved
with the 60° reactor, in which biomass dry weight reached
4.36 g L−1. This was followed by the 90°, 120°, and 180°
reactors, in decreasing order of productivity. The effect on
medium pH was similar in all four bioreactors: pH was
increased during the light cycle and decreased during the
dark cycle. The uptakes of nitrate and CO2 were considered
to be a cause of the increase in pH during the growth be-
cause Chlorella sp. is a freshwater microalga, and the nu-
trient concentration is at a relatively high level. The con-
tinued injection of 10% CO2-enriched air seemed to be a
cause of the decrease in pH during the dark cycle. On the
whole, the pH of the medium increased over time.
The results of our evaluation of the photosynthetic per-
formance in the conical HTPs are summarized in Table I.
The light energy inputs were virtually identical in all four
conical HTPs. As an example, the photosynthetically active
radiation (PAR; 400 to 700 nm) input to the 180° photo-
bioreactor was calculated as: 980 ␮mol m−2 s−1 (mean of
the photo-receiving area) × 0.5 m2 (photo-receiving area) ÷

Figure 3. (a) Effect of a cone angle on photo-receiving area and photo-

stage tube length. Inset shows the relation for the indicated cone angles on
an expanded scale. (b) Estimated relationship between cone angle and
photosynthetic productivity when light irradiation is 980 ␮mol m−2 s−1.
Data are normalized to the photosynthetic productivity of a 180° reactor,
which was arbitrarily set to 1. Relative photosynthetic productivity per unit
photo-receiving area (---) and per reactor (—) are shown.

of >0°, the smaller the cone angle, the larger the photosyn-
thetic productivity.

Experimentally Derived
Photo-Redistribution Effect
Although smaller cone angles are theoretically better for
photosynthetic productivity, the increasing photostage tube
length that accompanies a decreasing cone angle must also
be taken into consideration (Fig. 3a). For example, photo-
stage tube lengths in the 60° and 50° reactors were 49.0 and
58.0 m, respectively. When tube lengths were >60 m, cir-
culation of medium via an air-lift system was difficult. After
consideration of the stability of a medium flow, and making
allowances for practical technology, cone angles of ⱖ60°
were deemed suitable.
To assess the effect of photo-redistribution, the photosyn-
thetic productivities of four conical HTPs with cone angles
Figure 4. Time courses of the changes in the dry weights of Chlorella
of 180°, 120°, 90°, and 60°, photo-receiving areas of 0.50,
biomass and the pH of the medium in four conical HTPs obtained under
0.58, 0.71, and 1.0 m 2 , respectively, and a photo- conditions yielding maximum photosynthetic productivity. Depicted are
redistribution ratio of 1.00:1.16:1.42:2.00, were evaluated. reactor types and corresponding air flow rates: 180°, 0.6 L min−1 (䊏);
Photosynthetic production of Chlorella biomass and con- 120°, 0.3 L min−1 (⽧); 90°, 0.6 L min−1 (䊉); and 60°, 1.8 L min−1 (䉱).

696 BIOTECHNOLOGY AND BIOENGINEERING, VOL. 69, NO. 6, SEPTEMBER 20, 2000
Table I. Maximum photosynthetic productivity and efficiency in the four tested conical HTPs.
Maximum photosynthetic
Received light energy productivity
(kJ reactor−1 day−1, (g L−1 day−1,
Type of reactor 12-h light condition) 12-h light condition)
(A) 180° reactor 4601 0.61
(B) 120° reactor 4294 0.69
(C) 90° reactor 4441 0.75
(D) 60° reactor 4545 1.01
a
Calculated from experimental data of carbon contents and calorific value for 1 g carbon of microalgal biomass (Platt and Irwin, 1973).
4.6 ⳱; 106.5 (J s−1 or W), where 4.6 is the conversion factor
from W m−2 to ␮mol m−2 s−1 in the metal halide lamps (Hall
and Scurlock, 1993). The total energy input for an entire day
was calculated as: 106.5 (J s−1) × 60 × 60 × 12 ⳱; 4601 kJ
reactor−1 day−1 (12-h light/12-h dark cycle). Under an av-
erage photosynthetic photon flux density, simulating the
outdoors in central Japan, the maximum photosynthetic pro-
ductivity per installation area was 28.3 g dry biomass m−2
installation area per day (12-h light cycle) for the 60° reac-
tor. Photosynthetic efficiency, defined as the ratio of the
light energy recovered as biomass to the total light energy
received, was then assessed to evaluate the utilization effi-
ciency of the light energy received (Table I). The carbon
content in the powdered biomass was found to be 46.1%. As
the chemical energy of the carbon was 47.7 kJ g carbon−1
(Platt and Irwin, 1973), the energy of the total biomass was
estimated to be 22.0 kJ g dry biomass−1. The maximum
photosynthetic efficiency was thus 6.84% (PAR) for the 60°
reactor and 4.09% (PAR) for the 180° reactor.
Experimentally, the ratio of the photosynthetic produc-
tivities was 1.00:1.13:1.23:1.66 (Table I). Thus, the experi-
mental findings showed the same tendency as the theoretical
calculation for the four conical HTPs.
Productivity of the Conical Helical
Tubular Photobioreactor
In the present study, the photo-redistribution effect of our
conical HTP was confirmed both theoretically and experi-
mentally. We assessed photosynthetic performance of the
conical HTP in terms of the absolute quantity of biomass
produced. Differences in bioreactor design make it impos-
sible to compare directly the photosynthetic efficiency on
that basis; instead, photosynthetic productivity per unit area
was used for comparison. Using Spirulina platensis during
summer in Italy, Tredici and Materassi (1992) reported pro-
ductivities of 18 and 24 g m−2 day−1 with vertical- and
sun-oriented panel reactors, respectively, whereas Vonshak
and Guy (1992) reported that, in Israel, the productivity of
Spirulina was 20.8 g m−2 day−1 in an outdoor pond
equipped with a paddle-wheel. In addition, a photosynthetic
productivity of 10.5 g carbon m−2 day−1 was achieved with
the marine chlorophyte, Tetraselmis suecica, in shallow out-
door flumes under full sunlight in Hawaii (Laws and Bern-
ing, 1991). This value was translated into 29.2 g m−2 day−1,
Maximum photosynthetic Maximum
productivity per installation Recovered light energya photosynthetic
area (g m−2 day−1, (kJ reactor−1 day−1, efficiency
12-h light condition) 12-h light condition) (%) [PAR]
17.1 188 4.09
19.3 213 4.96
21.0 231 5.20
28.3 311 6.84
assuming that carbon accounts for about 36% of the dry
weight of Tetraselmis (Parsons et al., 1961) used. By com-
parison, the photosynthetic productivity of the 60° conical
HTP was 28.3 g dry biomass m−2 installation area per day
(12-h light condition). This figure was equivalent to or a
little higher than that obtained with other culture systems.
Tredici and Zittelli (1998) recently summarized the pho-
tosynthetic efficiencies obtained with various outdoor Spi-
rulina platensis cultivation systems in Florence, Italy during
the summer. They reported photosynthetic efficiencies of
6.6% (PAR) in a coiled tubular reactor, 4.6% (PAR) in a
sun-oriented plate, and 6.0% (PAR) in a vertical plate — the
highest photosynthetic efficiency being achieved with a
three-dimensional rather than a plate reactor. We similarly
obtained higher efficiencies with cone reactors than with a
flat reactor. In particular, the 60° HTP yielded a photosyn-
thetic efficiency of 6.84% (PAR). With respect to develop-
ment of photo-redistribution technology, the high photosyn-
thetic efficiency of our conical HTP represents an important
advantage in the laboratory condition. However, it may be
difficult to keep the extremely high photosynthetic produc-
tivity and efficiency under field conditions. An outdoor cul-
ture experiment is further required to prove and establish the
effectiveness of photo-redistribution technology in an out-
door environment.
The authors dedicate this article to the late Prof. D.O. Hall,
King’s College London, University of London.
References
Benemann JR, Oswald WJ. 1994. Systems and economic analysis of mi-
croalgae ponds for conversion of CO2 to biomass. Third quarterly
technical progress report no. DE-FG22-93PC93204. Berkeley, CA:
University of California. p 25.
Benemann JR, Tillett DM, Weissman JC. 1987. Microalgal biotechnology.
Trends Biotechnol 5:47–53.
Hall DO, Scurlock JMO. 1993. Biomass production and data. In: Hall DO,
Scurlock JMO, Bolhar-Nordenkampf HR, Leegood RC, Long SP, edi-
tors. Photosynthesis and production in a changing environment. Lon-
don: Chapman & Hall. p 425–444.
Laws EA, Berning JL. 1991. A study of the energetics and economics of
microalgal mass culture with the marine chlorophyte Tetraselmis sue-
cica: implications for use of power plant stack gases. Biotechnol Bio-
eng 37:936–947.
Mori K. 1985. Photoautotrophic bioreactor using visible solar rays con-
COMMUNICATION TO THE EDITOR 697
 densed by Fresnel lenses and transmitted through optical fibers. Bio-
technol Bioeng Symp 15:331–346.
Parsons TR, Stephens K, Strickland JDH. 1961. On the chemical compo-
sition of eleven species of marine phytoplankters. J Fish Res Bd Can
18:1001–1016.
Platt T, Irwin B. 1973. Caloric content of phytoplankton. Limnol Oceanogr
18:306–310.
Stryer L. 1988. Photosynthesis. In: Biochemistry. New York: WH Free-
man. p 517–543.
Tredici MR, Materassi R. 1992. From open ponds to vertical alveolar
panels: the Italian experience in the development of reactors for the mass
cultivation of phototrophic microorganisms. J Appl Phycol 4:221–231.
698 BIOTECHNOLOGY AND BIOENGINEERING, VOL. 69, NO. 6, SEPTEMBER 20, 2000
Tredici MR, Zittelli GC. 1998. Efficiency of sunlight utilization: tubular
versus flat photobioreactors. Biotechnol Bioeng 57:187–197.
Vonshak A, Guy R. 1992. Photoadaptation, photoinhibition and produc-
tivity in the blue-green alga, Spirulina platensis grown outdoors. Plant
Cell Environ 15:613–616.
Watanabe Y, Hall DO. 1996. Photosynthetic production of the filamentous
cyanobacterium Spirulina platensis in a cone-shaped helical tubular
photobioreactor. Appl Microbiol Biotechnol 44:693–698.
Watanabe Y, Ohmura N, Saiki H. 1992. Isolation and determination of
cultural characteristics of microalgae which functions under CO2 en-
riched atmosphere. Energy Convers Mgmt 33:545–552.