Synthesis and characterization of fluorescent dyes and their applications for the enhancement of growth rate of Chlorella vulgaris

Article abstract of DOI:10.1016/j.dyepig.2018.05.003

Spectral conversion of solar radiation from green to red light can enhance the photosynthesis efficiency of microalgae. This can lead to significant increase in microalgal growth. Four perylene-based fluorescent dyes were synthesized to convert wavelengths of light from green (500–570 nm) to orange-red (580–650 nm). These dyes were applied in microalgal cultivation system using two methodologies: (1) light-converting layer, and (2) light-converting medium. The dyes were dissolved in either ethanol or water to form light-converting layers between microalgae culture and the light source. Under converted light in both set-ups, Chlorella vulgaris cultures showed maximum increase in lipid productivity by 13.00% and 17.15%, and biomass productivity by 31.65% and 37.66% compared to the control. Overall, the cultivation results confirmed beneficial effects of modified light on microalgae cultures both on their lipid and biomass productivities. Note that both growth parameters achieved higher performance in the light-converting medium thereby the effectiveness of the cultivation method was validated.

Full text of DOI:10.1016/j.dyepig.2018.05.003

Accepted Manuscript
Synthesis and characterization of fluorescent dyes and their applications for the
enhancement of growth rate of Chlorella vulgaris
Hyeyoun Jang, Jinwoong Namgoong, Min-Gyu Sung, Yongkeun Chang, Jaepil Kim
PII:
S0143-7208(18)30623-5
10.1016/j.dyepig.2018.05.003
DYPI 6732
DOI:
Reference:
To appear in: Dyes and Pigments
Received Date: 20 March 2018
Revised Date: 3 May 2018
Accepted Date: 3 May 2018
Please cite this article as: Jang H, Namgoong J, Sung M-G, Chang Y, Kim J, Synthesis and
characterization of fluorescent dyes and their applications for the enhancement of growth rate of
Chlorella vulgaris, Dyes and Pigments (2018), doi: 10.1016/j.dyepig.2018.05.003.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT
Synthesis and characterization of fluorescent dyes and their applications for the
enhancement of growth rate of Chlorella vulgaris
§1
§1
2,3
2,4
Hyeyoun Jang , Jinwoong Namgoong , Min-Gyu Sung , Yongkeun Chang , Jaepil
1
*
Kim
1
Department of Materials Science and Engineering, Seoul National University, Seoul
1
51-744, Republic of Korea
2
Department of Chemical and Biomolecular Engineering, KAIST, 291 Daehak-ro,
Yuseong-gu, Daejeon 305-701, Republic of Korea
3
Division of Policy Research, Green Technology Center, 17th fl. Namsan Square Bldg.,
1
73, Toegye-ro, Jung-gu, Seoul 04554, Republic of Koera
4
Advanced Biomass R&D Center, #2502 Building W1-3, 291 Daehak-ro, Yuseong-gu,
Daejeon 305-701, Republic of Korea
§
Both authors contributed equally to the research.
Email address: jaepil@snu.ac.kr
Abstract: Spectral conversion of solar radiation from green to red light can enhance the
photosynthesis efficiency of microalgae. This can lead to significant increase in microalgal
growth. Four perylene-based fluorescent dyes were synthesized to convert wavelengths of
light from green (500 – 570nm) to orange-red (580– 650nm). These dyes were applied in
microalgal cultivation system using two methodologies: (1) light-converting layer, and (2)
light-converting medium. The dyes were dissolved in either ethanol or water to form light-
converting layers between microalgae culture and the light source. Under converted light in
both set-ups, Chlorella vulgaris cultures showed maximum increase in lipid productivity by

ACCEPTED MANUSCRIPT
1
3.00% and 17.15%, and biomass productivity by 31.65% and 37.66% compared to the
control. Overall, the cultivation results confirmed beneficial effects of modified light on
microalgae cultures both on their lipid and biomass productivities. Note that both growth
parameters achieved higher performance in the light-converting medium thereby the
effectiveness of the cultivation method was validated.
Keywords: Organic dyes, light-conversion, growth, microalgae, cultivation
1
. INTRODUCTION
Biomass is considered an alternative energy source for its renewable, biodegradable, and
rd
environmentally-friendly properties. Microalgae, known as the 3 generation biomass is
spotlighted as one of the most appropriate solution as it offers numerous advantages over
preceding generations. Unlike food crops, microalgae do not cause competition for food over
energy, nor demands as expensive chemical processing as lignocellulosic feedstock.
Henceforth, various approaches to stimulate rapid reproduction and boost lipid productivity
of microalgae were made. Assorted microalgae species were genetically engineered to bear
more growth determining components including photosynthetic pigments, fatty acids, and
other high value products.[1-2] The effect of temperature, mixing technique, gas transfer
control, and pH changes in bioreactors were also investigated through outdoor
cultivations.[3-7] Furthermore, new methods for efficient extraction and refinery have been
intensely scrutinized. Out of all, light-related parameters such as light intensity, illumination
time, day and night cycle, and diverse light sources appeared most critical for cell growth. In
particular, the intensity and wavelength of light.[8]
Light acts as a driving force in photoautotrophic cultivation. It makes logical sense to
provide utmost light energy to microalgae which photosynthesize to produce carbon
derivatives for growth and lipid production. However, microalgae are unable to absorb the

ACCEPTED MANUSCRIPT
entire spectrum of light. In fact, microalgae absorb most efficiently in photosynthetically
active radiation (PAR). This is primarily due to the photosynthetic pigments in the cells,
chlorophyll a and chlorophyll b, that absorb orange-red (600 – 700 nm) and reflect off green
light. Inferred from this light absorption range, extra orange-red light would contribute to
accelerate microalgal growth. A new challenge was to utilize an expendable spectrum of
energy through light conversion, where a portion of light less in need for growth were made
useful for photosynthetic activity.[9] Theoretically, the overall productivity of biomass and
lipid should be noticeably improved. Not only can algal cultivation overcome various
topographic limitations for set-up – even under suboptimal light supply – but also hint high
potential in economic feasibility on the road to commercialization.
In a previous study, Mohesenpour et al.[3] reported 20 – 70% increase in dry weight of
microalgae cultured under red light.. Wondraczek et al.[10] also carried out a similar
2
+
cultivation using fluorescence phosphorus powder derived from Eu as a light converter and
obtained 35% increase in cell weight. These promising results assured beneficial effect of
extra red light for microalgae cultivation.
Light-converting dyes can manipulate spectral properties of the light source to provide
more energy in the PAR. Four fluorescent dyes were synthesized for this purpose to absorb
light in the green wavelength range (500 – 580 nm) and emit in the orange-to-red wavelength
range (580 – 650 nm). Their emission spectra differed by approximately 10 – 20 nm to test
various shades of orange and red lights.
Perylene moiety was adopted in this study for its high quantum yield and high durability
which satisfy essential requirements for later application during cultivation. Also, it is rather
easy to tweak chemical and physical properties of perylene by the attachment of substituents
at peripheral and bay positions. Additionally, the chemicals used in the synthesis process are
relatively cheap and less toxic compared to reported studies.[10]

ACCEPTED MANUSCRIPT
Light-converting dyes were applied in cultivation system by two different methodologies
to create light-converting effect. Two dyes were dissolved in separate layers, namely light-
converting layers, adjacent to microalgae cultures. The water-soluble dyes were dissolved in
culture medium together with microalgae to act as a light-converting medium.[11]
The efficacy of both methodologies was examined in terms of lipid and biomass
productivities of microalgae while all other aspects of the experiment were designed to fulfill
optimal cultivation conditions.[12-14] Simultaneously, the effect of different colours on the
growth of microalgae is also discussed.
Note that Chlorella vulgaris species was selected as a local representative of eukaryotic
microalgae. Chlorella vulgaris contains lots of lipid, and enclose chlorophyll a and
chlorophyll b in their cell bodies to absorb light most efficiently in 400 – 500 nm and 600 –
7
00 nm ranges to photosynthesize, in other words, grow faster under red light. [9, 10, 15]
Experiments were aimed to examine if Chlorella vulgaris grew more efficiently when extra
light was provided in this particular wavelength range.
2
2
2
. EXPERIMENTAL
.1. SYNTHESIS
.1.1. CHEMICALS
Perylene-3,4,9,10-tetracarboxylic dianhydride (97%), 2,6-diisopropylaniline (97%),
propionic acid (≥99.5%, ACS reagent grade), sulfuric acid (95.0-98.0%, ACS reagent grade),
bromine (≥99.5%, ACS reagent grade), acetic acid (≥99.7%, ACS reagent grade), N-Methyl-
2
-pyrrolidone (NMP) (≥99.0%, ACS reagent grade), anhydrous potassium carbonate (≥99.0%,
ACS reagent grade), and phenol (≥99.0%, reagent plus) was purchased from Sigma-Aldrich,
and dichloromethane (>99.0%) was purchased from TCI Chemicals. Other reagents and
solvents were obtained from various commercial suppliers including Alfa Aesar, Duksan

ACCEPTED MANUSCRIPT
Chemicals, and Samchun Chemicals at reagent-grade quality.
Figure 1 Synthetic scheme of dyes 1, 2, 1', and 2'
2
.1.2. SYNTHESIS OF 1,7-DIBROMO-PERYLENE-3,4,9,10-TETRACARBOXYLIC
DIANHYDRIDE
Perylene-3,4,9,10-tetracarboxylic dianhydride (39.2 g, 0.10 mmol), iodine (0.474 g, 3.73
mmol), and 450 mL of 98% sulfuric acid were mixed and stirred for 2 hours at room
temperature. The reaction temperature was increased to 80°C and bromine (8.87 mL, 172
mmol) was added dropwisely. The mixture was heated for 16 hours then cooled to room
temperature while displacing excess Br (g) by flowing N (g) into the reaction flask. The
2
2
mixture was slowly transferred to a litre of ice-water and red precipitate was obtained. The
precipitate was collected by suction filtration and was washed with distilled water several
times until the aqueous layer became neutral to yield crude product. The product was dried in
a vacuum oven at 80°C and used in the next step without further purification.
2
.1.3. SYNTHESIS OF N,N’-BIS(2,6-DIISOPROPYLPHENYL)-1,7-DIBROMO-
PERYLENE-3,4,9,10-TETRACARBOXYDIIMIDE
1
,7-dibromoperylene-3,4,9,10-tetracarboxydiimide (5.50 g, 0.01 mol) was suspended in

ACCEPTED MANUSCRIPT
3
20 mL of propionic acid in a 500 mL round-bottom flask at 70°C for 1 hour. 2,6-
diisopropylaniline (13.2 mL, 0.07 mol) was added to the mixture after increasing the reaction
temperature to 140°C and heated overnight. The mixture cooled to room temperature and was
slowly transferred to 500 mL of methanol. Filtrate was collected after suction filtration;
washed thoroughly with methanol. Crude product was obtained after evaporating the filtrate
in rotatory evaporator. The product was purified by column chromatography on silica gel
using CH Cl as the eluent. A band containing tribrominated diimide was separated first,
2
2
1
followed by dibrominated and monobrominted. Yield: 53%; H NMR (DMSO-d , ppm): 9.56
6
(d, 2H), 8.83 (d, 2H), 8.79 (s, 2H), 7.48 (t, 2H), 7.38 (d, 2H), 3.00(m, 4H), 1.13 (d, 24H);
+
MALDI-TOF MS: m/z 896.94, 100% [M + H ]; Elemental analysis: Calcd for
C H Br N O : C, 66.37; H, 4.64; N, 3.22%. Found: C, 66.19; H, 4.72; N, 3.26.
48
40
2
2
2
2
.1.4. SYNTHESIS OF N,N’-BIS(2,6-DIISOPROPYLPHENYL)-1,7-DIPHENOXY-
PERYLENE-3,4,9,10-TETRACARBOXYDIIMIDE (1)
N,N’-Bis(2,6-diisopropylphenyl)-1,7-dibromoperylene-3,4,9,10-tetracarboxydiimide (4.0
g, 4.62 mmol) was suspended in 280 mL of N-Methyl-2-pyrrolidone (NMP) with phenol
(
0.94 g, 10.0 mmol), and anhydrous K CO (2.8 g, 20.3 mmol). The mixture was heated at
2 3
1
20°C for 24 hours then cooled to room temperature. The mixture was slowly poured into
00 mL of 5% HCl solution. A precipitate was then collected by suction filtration, repeatedly
3
washed with distilled water. Crude product was dried in vacuum oven at 80°C overnight. The
product was purified by column chromatography on silica gel using CH Cl as the eluent.
2
2
1
Yield: 82.8%; H NMR (DMSO-d , ppm): 8.69 (d, 2H), 8.13 (d, 2H), 7.33 (m, 10H), 7.15 (m,
6
6
H), 7.38 (d, 2H), 2.69(m, 4H), 1.04 (d, 24H); MALDI-TOF MS: m/z 896.25, 100% [M +
+
H ]; Elemental analysis: Calcd for C H N O : C, 80.51; H, 5.63; N, 3.13%. Found: C,
6
0
50
2
6
7
9.98; H, 5.59; N, 3.10.

ACCEPTED MANUSCRIPT
2
.1.5. SYNTHESIS OF N,N′-BIS(2,6-DIISOPROPYLPHENYL)-1,7-DI-[4
(SULFURIC ACID)PHENOXY]PERYLENE-3,4,9,10-
TETRACARBOXYDIIMIDE (1’)
N,N′-Bis(2,6-diisopropylphenyl)-1,7-diphenoxy-perylene-3,4,9,10-tetracarboxydiimide
(1 g, 1.12 mmol) was added to 2.5 mL of concentrated sulfuric acid and the flask was sealed.
The mixture was stirred at room temperature for 15 hours. 5 mL of distilled water was added
dropwisely to form a precipitate, then filtered under low pressure. The solid product was
washed several times with 30 mL of CH Cl . Red product was dried in a vacuum oven at
2
2
8
1
2
0°C. Synthetic scheme of dyes is shown in Fig. 1. Yield: 75%; MALDI-TOF MS: m/z
+
057.15, 100% [M + H ]. Elemental analysis: Calcd for C H N O S : C, 68.30; H, 4.78; N,
6
0
50
2
12 2
.65; S, 6.08%. Found: C, 69.56; H, 4.73; N, 2.68; S, 6.16.
2
.1.6. SYNTHESIS OF 1,6,7,12-TETRACHLOROPERYLENE-3,4,9,10-
TETRACARBOXYLIC DIANHYDRIDE
Perylene-3,4,9,10-tetracarboxylic dianhydride (39.2 g, 0.10 mmol), iodine (0.474 g, 3.73
mmol), and 450 mL of chlorosulfuric acid were mixed and stirred for 2 hours at room
temperature. The reaction temperature was increased to 70°C and the mixture was heated for
1
6 hours. The mixture was slowly transferred to a litre of ice-water and red precipitate was
obtained. The precipitate was collected by suction filtration and was washed with distilled
water several times until the aqueous layer became neutral to yield crude product. The
product was dried in a vacuum oven at 80°C and used in the next step without further
purification.
2
.1.7. SYNTHESIS OF N,N’-BIS(2,6-DIISOPROPYLPHENYL)-1,6,7,12-
TETRACHLOROPERYLENE-3,4,9,10-TETRACARBOXYDIIMIDE

ACCEPTED MANUSCRIPT
1
,6,7,12-tetrachloroperylene-3,4,9,10-tetracarboxydiimide (5.30 g, 0.01 mol) was
suspended in 320 mL of propionic acid in a 500 mL round-bottom flask at 70°C for 1 hour.
,6-diisopropylaniline (13.2 mL, 0.07 mol) was added to the mixture after increasing the
2
reaction temperature to 140°C and heated overnight. The mixture was cooled to room
temperature and was slowly transferred to 500 mL of methanol. Filtrate was collected after
suction filtration; washed thoroughly with methanol. Crude product was obtained after
evaporating the filtrate in rotatory evaporator. The product was purified by column
chromatography on silica gel using CH Cl as the eluent. A band containing product was
2
2
separated first, followed by tri-chlorinated, di-chlorinated and then mono-chlorinated. Yield:
1
7
6%; H NMR (DMSO-d , ppm): 8.69 (s, 4H), 7.50 (t, 2H), 7.38 (d, 4H), 3.01(m, 4H), 1.14
6
+
(
d, 24H); MALDI-TOF MS: m/z 848.52, 100% [M ]; Elemental analysis: Calcd for
C H Cl N O : C, 67.94; H, 4.51; N, 3.30%. Found: C, 67.79; H, 4.45; N, 3.62.
48
38
4
2
4
2
.1.8. SYNTHESIS OF N,N’-BIS(2,6-DIISOPROPYLPHENYL)-1,6,7,12-
TETRAPHENOXY-PERYLENE-3,4,9,10-TETRACARBOXYDIIMIDE (2)
N,N’-Bis(2,6-diisopropylphenyl)-1,6,7,12-tetrachloroperylene-3,4,9,10-
tetracarboxydiimide (3.91 g, 4.62 mmol) was suspended in 280 mL of N-Methyl-2-
pyrrolidone (NMP) with phenol (0.94 g, 10.0 mmol), and anhydrous K CO (2.8 g, 20.3
2
3
mmol). The mixture was heated at 120°C for 24 hours then cooled to room temperature. The
mixture was slowly poured into 300 mL of 5% HCl solution. A precipitate was then collected
by suction filtration, repeatedly washed with distilled water. Crude product was dried in
vacuum oven at 80°C overnight. The product was purified by column chromatography on
1
silica gel using CH Cl as the eluent. Yield: 86%; H NMR (DMSO-d , ppm): 7.97 (s, 4H),
2
2
6
7
.54 (t, 2H), 7.39 (m, 12H), 7.29 (m, 12H), 2.69(m, 4H), 1.01 (d, 24H); MALDI-TOF MS:
+
m/z 1080.29, 100% [M + H ]. Elemental analysis: Calcd for C H N O : C, 80.13; H, 5.42;
7
2
58
2
8

ACCEPTED MANUSCRIPT
N, 2.60%. Found: C, 79.95; H, 5.49; N, 2.56.
.1.9. SYNTHESIS OF N,N’-BIS(2,6-DIISOPROPYLPHENYL)-1,6,7,12-TETRA-[4-
SULFURIC ACID)PHENOXY]PERYLENE-3,4,9,10-
2
(
TETRACARBOXYDIIMIDE (2’)
N,N’-Bis(2,6-diisopropylphenyl)-1,6,7,12-tetraphenoxy-perylene-3,4,9,10-
tetracarboxydiimide (1.21 g, 1.12 mmol) was added to 2.5 mL of concentrated sulfuric acid in
a flask then sealed. The mixture was stirred at room temperature for 15 hours. 5 mL of
distilled water was added dropwisely to form a precipitate, then filtrated under low pressure.
The solid product was washed thoroughly with 30 mL of CH Cl . Red product was dried in a
2
2
+
vacuum oven at 80°C. Yield: 84%; MALDI-TOF MS: m/z 1401.06, 100% [M + H ].
Elemental analysis: Calcd for C H N O S : C, 61.79; H, 4.18; N, 2.00; S, 9.16%. Found: C,
7
2
58
2
20 4
6
1.64; H, 4.13; N, 1.98; S, 9.25.
2
.1.10. CHARACTERIZATION
1
H NMR spectra were measured on a Bruker Avance 500 spectrometer (National Center
for Inter-University Research Facilities at Seoul National University) at 500MHz using
dimethyl sulfoxide-d6 and chloroform-d as the solvent. Tetramethylsilane (TMS) was used as
1
an internal standard. Synthesized dyes were examined for chemical analysis via H NMR,
matrix-assisted laser desorption / ionization time of flight (MALDI-TOF) mass spectrometry,
1
and elemental analysis. H NMR and MALDO-TOF MS spectra were measured using a
Bruker Avance 500 spectrometer at 500MHz and Voyager-DE STR Biospectrometry
Workstation with α-cyano-4-hydroxy-cynamic acid (CHCA) matrix respectively. Elemental
analysis was conducted with a Thermo Scientific Flash EA 1112 elemental analyzer. All
measurements were carried out at National Center for Inter-University Research Facilities
(NCIRF) at Seoul National University

ACCEPTED MANUSCRIPT
Spectral characteristics were analyzed using PerkinElmer UV/Vis spectrometer Lambda
5 and PerkinElmer fluorescence spectrometer LS 55, and absorption and emission spectra
2
were scanned in 400 – 800 nm range. Quantum yield of dyes were measured using Jasco
International FP-8500ST spectrofluorometer.
2
2
.2.CULTIVATION
.2.1. STRAIN AND SEED CULTURE
Chlorella species have short and regular life cycle and are relatively easy to culture in a
lab environment. Most importantly, cells absorb red light efficiently owing to their pigment
content. Thereby, Chlorella vulgaris, known to contain rich fatty acid,was selected as the
working organism. Chlorella vulgaris obtained from Algal Biomass Centre, Korea Advanced
Institute of Science and Technology (KAIST) was cultivated in a sterilized BG-11
medium[16] including NaNO (1.5 g/L), K HPO (0.058 g/L), Na CO (0.020 g/L), and other
3
2
4
2
3
essential components.
2
.2.2. OPTIMIZATION OF PHOTOAUTOTROPHIC CONDITIONS IN BATCH
OPERATION
Apart from variable lighting conditions, cells were cultivated in optimal conditions for
photosynthesis.[12, 17] Medium selection was important as it signifies a living environment
microalgae are directly exposed to. TAP medium was first considered as it would stimulate
fast growth of microalgae. Though, the option was ruled out due to high carbon content in the
form of acetic acid. This is because microalgae should be solely dependent on self-produced
carbon compounds, and to test the effect of modified light. Alternative choice was BG-11
medium as it supplied no excessive carbon nutrients and the cells are dependent on
photosynthesis to produce carbon derivatives either to replicate or accumulate lipid in their
bodies.

ACCEPTED MANUSCRIPT
When seeding, Chlorella vulgaris was transferred from agar plate and inoculated in 200
mL culture medium. The medium and flasks were sterilized in an autoclave for 20 minutes at
1
21°C to prevent any contamination. Microalgae were incubated at 24°C under cool white
2
LED lamp which illuminated consistently at an intensity of 150 µmol/m /s. Flasks were
clipped onto Vision Scientific Co. LTD shaker set at 120 rpm, and CO enriched air (1%, v/v)
2
was bubbled into mixture through silicon tubing (Korea Ace Scientific) via a 20 µm filter.
The aeration was monitored using Dwyer’s flowmetre, set at 1.5 mL/min air flow rate into
the working volume. Gas input had a significant effect in mixing.
Inoculants in their lag phase were cultivated until the midst of exponential phase over
approximately 96 hours. A small portion of production medium was centrifuged at 3,000rpm
for 15 minutes to harvest cells. Then, the cells were washed twice with distilled water and re-
suspended in 400 mL of BG-11 medium. Determined by optical density (OD) measurement at
6
60 nm, equal amount of cells was transferred to a larger cultivation flask (500 mL with
vented cap) for full scale-testing. Initial optical density of fresh microalgae samples was 0.1.
Cells were cultured until late-exponential phase where growth curve reached a plateau.
2
.2.3. APPLICATION OF LIGHT-CONVERTING DYES IN MICROALGAE
CULTIVATION
Dyes were added to microalgae samples by two different methodologies according to
their solubility in various solvents (Fig. 2).

ACCEPTED MANUSCRIPT
Figure 2 Front and top views of experimental setup for two cultivation methodologies: (left) dyes 1 and 2
dissolved in ethanol-base light-converting layer, and (right) dyes 1' and 2' dissolved in light-converting
medium
2
.2.4. METHODOLOGY 1: LIGHT-CONVERTING LAYER
-6
Dyes 1 and 2 soluble in organic solvents were each dissolved at 10 M in 400 mL of
ethanol to act as a light-converting layer. Microalgae were grown separately in neighboring
flasks.The coupled flasks were placed next to the light source so that light had to penetrate
through the dye solution prior to reaching the microalgae layer. The meniscus level in
adjacent flasks were aligned to ensure the dye solution completely covered the surface of
culture flasks. This safeguarded the cells from exposure to unmodified light from the source
and the cells only received modified light through the light-converting layer. The dye-
containing flasks were tightly sealed to prevent evaporation and alteration in dye
concentration throughout cultivation.

ACCEPTED MANUSCRIPT
Similarly to the seeding step, the flasks were placed on a shaker set at 120 rpm to mix
content thoroughly without touching the vented caps to avoid contamination. Carbon dioxide
was supplied at the same rate as for seed culture via silicon tubing for gas supply and mixing.
2
.2.5. METHODOLOGY 2: LIGHT-CONVERTING MEDIUM
A preliminary cultivation was carried out to ensure microalgae cells can survive in the
light-converting medium setup (Fig. 3). Chlorella vulgaris were cultured in two different
conditions: with and without water-soluble dye in medium, over 72 hours. Both samples were
measured on dry cell weight and cell count. The cells grown in water-soluble dye medium
showed higher count in both, hence it was clear that the dye had a positive effect on the
growth of microalgae. Furthermore, the cells were tested at different concentrations of dye-
-7
-6
-5
-4
media (4×10 M, 4×10 M, 4×10 M, and 4×10 M). Later cultivations were carried out at
-
6
dye concentration of 10 M throughout, because the cells grew fastest in it during the trial
phase.
-6
Later, culture flasks containing BG-11 medium, 10 M of dyes 1’ or 2’, and microalgal
cells were tightly sealed and placed on a shaker set at 120rpm. The mixture was to mixed
thoroughly without touching the vented caps to avoid contamination. Gas was supplied in the
same manner as in light-converting layer setup for carbon supply and mixing.

ACCEPTED MANUSCRIPT
Figure 3 (Top) toxicity test of applying light-converting medium in microalgae cultivation, and (bottom)
water-soluble dyes 1' and 2' tested at various concentrations
2
.2.6. SCHEME OF EXPERIMENTS
Both methodologies followed the same cultivation and sampling protocols after initial
setup. Cells were cultured in a light room at 24°C for 120 hours under cool white LED with a
2
photon flux of 100 µmol/m /s, alight strength estimated to not cause photolysis of dyes nor
photo-inhibition of cells. These conditions were monitored on a 12-hour basis throughout
1
20-hour cultivation for consistency.
Cell density was measured every 12 hours using Nexcelom Bioscience’s Cellometer K2
Image Cytometer and Leica’s optical microscope. Photosynthetic activity and lipid
productivity were calculated from the raw data obtained from quantum yield and lipid
content measured post harvesting. All culture samples were experimented in duplicates, and
sampled in triplicates.
2
2
.3.GROWTH ANALYSIS
.3.1. CELL DENSITY AND BIOMASS PRODUCTIVITY
Cell density was determined from data obtained by Nexcelom Bioscience’s Cellometer K2

ACCEPTED MANUSCRIPT
Image Cytometer and Leica’s optical microscope where thecellometer deduced the value
from 20 µL of cells mounted on each slides. Cell density was plotted on a graph over time to
show a growth curve that eventually levels off. Biomass productivity was calculated using
Wang et al.’s analytical methods.[6]
2
.3.2. LIPID CONTENT AND LIPID PRODUCTIVITY
Once cells reached stationary growth phase, a small portion was collected for quantum
yield measurement and lipid extraction. The method was adapted from Bligh and Dyer’s
work[18] and was performed gravimetrically. Then lipid content and productivity were
calculated.
Sartorius balance was used to precisely measure the weight of dry biomass and extracted
lipid to four decimal places.
2
.3.3. PHOTOSYNTHETIC RATE
Quantum yield of cells were measured once at the end of 120-hour cultivation. A small
portion of cells were depleted of light for 30 minutes prior to testing in an Aquapen-C AP-C
1
00 Photon System Instrument. Blue and red excitation light at 455 nm and 620 nm were
2
emitted at the intensity of up to 3,000 µmol/m /s to estimate photon absorption of microalgae
suspension. The amount of photon absorbed with respect to light emitted computed to
quantum yield of cells which is pertinent to their photosynthetic rate.
3
. RESULTS AND DISCUSSION
3
.1. SYNTHESIS AND CHARACTERIZATION
Dyes were designed to absorb and emit in the green (500 – 580 nm) and the orange-to-red
580 – 650 nm) wavelength ranges in order to provide additional light in need for microalgal
growth.. 2,6-diisopropylaniline was attached at both peripheral positions of the halogenated
(

ACCEPTED MANUSCRIPT
perylene, and these substituents suppressed pi–pi interactions between the molecules, thereby
increasing the solubility in organic solvents. Next, halogen atoms attached at bay positions
were substituted with phenoxy group. The phenoxy group extended the pi-conjugation of
electrons from the perylene body and as an effect shifted the absorption and emission band
bathochromically by up to 20 nm. In this intermediary step of synthesis, a considerable
difference in structure of dyes 1 and 2 can be found at the bay positions where one has two
phenoxy groups attached to it, and the other has four phenoxy groups. Dyes 1 and 2 each
showed maximum absorbance at 540.0 nm and 574.0 nm, and maximum emission at 583.0
nm and 613.5 nm, respectively.
Dyes 1 and 2 were each divided into two aliquot parts – one for use in light-converting
layer, and another for use in light-converting medium after further synthetic reactions. Dyes 1’
and 2’ derived from dyes 1 and 2 via sulfonation under mild conditions. They showed
maximum absorbance at 550.0 nm and 564.0 nm, and maximum emission at 599.0 nm and
6
19.5 nm. Note that the maximum absorbance of dyes 1’ and 2’ lie between those of dyes 1
and 2. However, maximum emission of dyes 1’ and 2’ both exceed those of dyes 1 and 2.
All dyes showed spectral characteristics as predicted in the design process. They
absorbed most in the green wavelength range (mid-500 nm) in which chlorophyll-containing
organisms would mostly reflect off. Instead, they emitted in orange and red wavelength range
(
from late 500 nm to mid-600 nm) which coincides with photosynthetically active region for
microalgae.
Spectral characteristics of the light-converting dyes are presented in Fig. 4 and Table 1.

ACCEPTED MANUSCRIPT
Figure 4 Normalized absorption and emission spectra of synthesized dyes
Table 1 Absorption and emission maxima, stoke shift, and quantum yield of dyes
Stoke shift Quantum
Dye Solvent λabs (nm)
λemi (nm)
(nm)
yield
0.66
0.69
0.29
0.35
1
2
1
2
540.0
574.0
550.0
564.0
583.0
613.5
599.0
619.5
43.0
39.5
49.0
55.5
Ethanol
Water
’
’
Stoke shift for dyes 1, 2, 1’, and 2’ were 43 nm, 39.5 nm, 49 nm, and 55.5 nm,
respectively. Each pair of dyes showed similar stoke shifts, though, the sulfonated dyes 1’
and 2’ show broader spectra. Evidently, their average absorbance falls between the phenol-
substituted dyes, and both photoluminescence peaks exceed those of dyes 1 and 2.
The emission spectra were spread across from orange to red. While distinguishing

ACCEPTED MANUSCRIPT
colours is easy, understanding the chemical mechanism behind the colour deviation is rather
complicated. The complexity is due to the variety of inter- and intra-molecular interactions
that cause spectral shifts. At the simplest level, solvatochromism explains that light emission
from fluorophores is hugely dependent on solvent polarity. As explained by Jablonski
diagram for fluorescence with solvent relaxation [19], a fluorophore is typically excited to the
first singlet state (S ) by energy intake. Energy level shifts up and down via internal
1
conversion and vibrational relaxation, but returns to excited state µ almost instantly within
E
1
0 – 100 ps. Intersystem crossing is strongly dependent on solvent polarity where polar
solvent molecules take part in the fluorescence mechanism by reorienting around and
adjusting the excited energy level. As photons release energy from varied energy levels,
different colours of fluorescent light are emitted.
This effect depends on the polarity of solvent molecules where strongly polar solvent
lands at a lower energy level compared to weakly polar solvent. Emission occurs at a lower
energy level with increasing solvent polarity, in other words, in a more redshifted colour.
Since water is slightly more polar than ethanol, dyes 1’ and 2’ are bathochromically shifted in
both absorbance and emission by up to 20 nm in comparison to dyes 1 and 2. Not to mention,
polar fluorophores display stronger sensitivity to solvent polarity. Accordingly, the tetra-
substituted dyes (red) are undoubtedly more responsive to this phenomenon than the di-
substituted dyes (orange).
An important criterion in determining the performance of dyes as light-converting agents
is quantum yield. It measures the amount of photons emitted with respect to the amount
absorbed. In this case, quantum yield represents colour conversion rate from green to either
orange or red. Conversion efficiencies were measured using FP-8500ST spectrofluorometre
at UNIST.
Dyes 1 and 2 dissolved in ethanol demonstrated fairly high quantum yield at 66% and

ACCEPTED MANUSCRIPT
6
9%, respectively. Distinct orange and red colours were apparent. Quantum yield of dyes 1’
and 2’ reached 29% and 35% which meantfluorescence was not as strong. But glowing
colours were equally apparent to human eyes.
Significantly lower quantum yield of dyes 1’ and 2’ in polar solvent attributes to the
presence of photo-induced electron-transfer (PET) processes. Since PET process heavily
depends on the distance between hydrophilic substituent and chromophore, the distance
between sulfonyl groups and perylene body should have led to significant PET activity.[20]
The process causes a broadening of the emission spectra as well as a low quantum yield.
Regardless, the emission wavelengths seem to cast stronger impact on the growth rate of
microalgae compared to quantum yield as the results showed better performance under
bathochromically shifted light rather than under higher quantum yield dyes.
Dyes 1 and 2 and dyes 1’ and 2’ differ in structure at para-position of bay substituents.
Highly polar and bulky sulfonyl groups attached in these positions induce disproportionate
electron distribution, and the ionic nature of the functional group significantly increased
solubility of dyes 1’ and 2’ in water more than dyes 1 and 2. The four dyes were then applied
in two cultivation methodologies of either light-converting layer or light-converting medium,
depending on the solubility of dyes in either ethanol or water.
As explained in the previous section, each cultivation methodologies were carried out
using a pair of dyes with same substituents either di-substituted or tetra-substituted at bay
positions. Since red dyes (2 and 2’) are more bathochromically shifted compared to the
orange dyes (1 and 1’) in each cases, faster growth in the former samples were expected.
Inevitably, the light converting medium containing dyes 2 and 2’ somewhat showed higher
performance in increasing microalgal growth than light converting layers containing dyes 1
and 1’.
3
.2.CULTIVATION

ACCEPTED MANUSCRIPT
3
.2.1. METHODOLOGY 1: LIGHT-CONVERTING LAYER
In this setup, thelag phase of Chlorella vulgaris lasted approximately 24 hours and
continued onto exponential phase until stationary phase was reached at 96 hours. The result
was shown in Fig. 5 and Table 2.
Figure 5 Lipid productivity and cell density of Chlorella vulgaris cultured with light-converting layer
Table 2 Lipid and biomass productivities of Chlorella vulgaris cultured with light-converting layer shown
in percentage increases
Lipid
Lipid
% increase Cell count % increase
(cells / (compared
Sample
content productivity (compared

ACCEPTED MANUSCRIPT
(
%)
(g / L / day) to control)
mL)
to control)
67.07
8
1
2
11.09
11.74
0.11105
0.11779
0.09996
11.10
17.83
0.00
2
1
1
.080 · 108
.740 · 108
.245 · 10
39.76
Control 10.00
0.00
As hypothesized, lipid productivity increased in the order of red dye 2, orange dye 1,
then control. Red and orange dye samples each showed lipid productivity of 0.11105 g/L/day
and 0.11779 g/L/day which were equivalent to 11.10% and 17.83% escalation compared to
the control at 0.09996 g/L/day.
Cell density also showed improvement in a contrasting trend to lipid productivity where
the orange dye stimulated cells to replicate faster than the red dye. Dyes 1 and 2 samples
8
8
reached a cell density plateau at 2.080×10 cells/mL and 1.740×10 cells/mL, respectively.
8
These represent 67.07% and 39.76% increase compared to the control at 1.245×10 cells/mL.
3
.2.2. METHODOLOGY 2: LIGHT-CONVERTING MEDIUM
For this part, preliminary tests were carried out to specify appropriate concentration of
dye-medium for microalgae cells grow and reproduce. Cell growth was measured by
microscopic cell count and dry cell weight (DCW) after 120 hours. Cell count showed largest
7
-5
population size of 2.32×10 cells/mL in 4×10 wt% sample, and DCW weighed most heavily
-6
at 0.82 g/L in 4×10 wt% sample. This test proved that cells can grow faster in the medium.
Overall, the toxicity test turned out safe and acceptable to employ dyes in multi-functional
medium, co-acting as a production medium for cells as well as a light-converting layer.
Dyes 1’ and 2’ along with cells were dissolved in the medium during cultivation and as a
whole functioned as a light-converting layer. Lag phase lasted just as long as for light-
converting layers, for approximately 24 hours. Though, the cells reached the stationary phase
about half a day earlier, meaning that the cell count expanded at a slightly faster rate.
st
The results (Fig. 6 and Table 3) showed resemblance with the 1 batch of cultivation in

ACCEPTED MANUSCRIPT
rising trends of lipid content and cell density. Lipid productivity increased in the order of red
dye 2’, orange dye 1’, followed by control. Dyes 1’ and 2’ samples showed lipid productivity
of 0.10936 g/L/day and 0.12212 g/L/day which were equivalent to 4.91% and 17.15% rise
compared to the control at 0.10424 g/L/day.
Figure 6 Lipid productivity and cell density of Chlorella vulgaris cultured in light-converting medium
Table 3 Lipid and biomass productivities of Chlorella vulgaris cultured with light-converting medium
shown in percentage increases
Sample
Lipid
Lipid
% increase Cell count
% increase

ACCEPTED MANUSCRIPT
content
%)
productivity (compared (cells / mL)
(compared
to control)
(
(g / L / day) to control)
8
1
2
’
’
10.93
12.20
0.10936
0.12212
0.10424
4.91
17.15
0.00
37.66
17.09
0.00
2
1
1
.175 · 108
.850 · 108
.580 · 10
Control 10.40
Biomass productivity increased in a reverse order where orange dye sample overtook the
8
red dye sample, and control came last. Cell density curve reached a plateau at 2.175×10
8
cells/mL and 1.850×10 cells/mL, representing 37.66% and 17.09% increase from the control
8
at 1.580×10 cells/mL.
Data obtained from the experiments using light-converting layers or light-converting
medium showed similar trends. Orange dye samples in both cases significantly increased
biomass productivity by more than 30% whereas red dye samples showed higher lipid
productivity by around 17%. Chlorophylls in Chlorella vulgaris make no use of green light
for photosynthesis nor growth. Hence, orange and red dyes were used to convert light in this
range to other useful colours. As shown in Figure 4, both orange and red dyes absorb light in
the green wavelength range. However, the absorption of orange dye lies closer to the green
range compared to the red dye’s. This means that the orange dye absorbed more green light to
convert to orange light, inducing more active growth compared to red dye. Note that higher
biomass does not directly correlate with lipid content in cells.
The result from cultivation suggests that a new approach of applying dyes in light-
converting media was comparably effective as using light-converting layers. Furthermore,
maximum lipid and biomass productivities out of all dye sample were observed in light-
converting medium samples. In comparison to light-converting layer method, less portion of
incident light was lost by surface reflection, light absorption by host material, light emission
from the layer edges. Given that the working volume is recycled for several cultivation cycles,

ACCEPTED MANUSCRIPT
light-converting medium method suggests a possibility in constructing a more space efficient,
cost effective, and eco-friendly bioreactor.
3
.2.3. QUANTUM YIELD
All microalgae samples measured at the end of cultivation cycles exceeded 70% in
quantum yield. As quantum yield measures the ratio of light absorbed by the cells to light
emitted, the results can be interpreted as light usage efficiency of cells. It is also possible to
conclude that the absorption efficiency directly correlates to the rate of photosynthesis.
Supplementary tests using increasing concentration of dyes would have strengthened
data set as the results can define optimal dye concentration requiredto maximize the growth
rate of microalgae. It is also advisable to explore at what concentration the dyes begin to
impede beneficial effect on cell growth, predominantly caused by screening.
Similar trials on different light intensities controlled by light power, or multi-colours
such as blue, yellow, or green, are also recommended for future study.
4
. CONCLUSION
It is logical to conclude that spectral conversion can improve lipid and biomass
productivities of Chlorella vulgaris. Light alteration caused by the synthesized dyes offers
environmentally friendly, economically feasible cultivation technique for prosperous
production of lipid or biomass, ultimately for efficient biofuel production.
Four fluorescent dyes that absorb mostly in green wavelength range and emit in orange
and red regions allow more light energy available for photosynthesis. Cultivation results
showed that bathochromically-shifted dyes improved lipid productivity more extensively.
Specifically, red dye samples increased in lipid productivity more than orange dyes samples
in both light-converting layers and light-converting media. Though, orange dye samples
outperformed in biomass productivity instead, where population size expanded more in

ACCEPTED MANUSCRIPT
orange dye samples more than red samples. This can be explained by the absorption of
orange dyes that overlays green wavelength range more than that of red dyes. Technically,
orange dyes can absorb more light to convert to orange, inducing higher growth rate
compared to the red dye. This can be interpreted that red light induced fatty cells with less
cell count, and vice versa for orange light. Conversely, the effect of light on lipid and biomass
productivities are colour specific.
In general, light-converting layer and light-converting medium methodologies both
promoted growth rate of Chlorella vulgaris. Both methodologies used dyes in solution stateto
assure higher quantum yield, lower cost, and less toxicity over solid state in which
fluorescence of dyes can be severely hindered, especially when coated on polymethyl
methacrylate (PMMA) filter.[3] The implications of the second method are significant in the
area of bioreactor design as they overcome many critical limitations found in conventional
layered structures built for spectral conversion. Unlike conventional method of creating
reactor layers, light-converting medium can potentially add multifunctionality to a plain
medium. Effectiveness and viability of the new cultivation method proved numerous
advantages over light-converting layers. This method is simple, space-efficient, and
economically feasible. No additional layers are needed for light conversion, nor requires
larger area to build more layers. Despite environmental concerns on coloured water, the
recyclability of medium prevents drainage into open water, and when necessary, easy
separation of dyes is also possible by salt formation.
5
. ACKNOWLEDGEMENTS
This work is supported by Industrial Technology Innovation Program. No.10052798
funded by the Ministry of Trade, Industry & Energy (MOTIE, Korea), and based on a
master’s thesis by Hyeyoun Jang at Seoul National University.

ACCEPTED MANUSCRIPT
6
. REFERENCES
[
1] Sriram N, Shahidehpour M, Renewable biomass energy. IEEE Power Eng. Soc. Gen.
Meet. 2005: 1–6. doi:10.1109/PES.2005.1489459
2] Roessler PG, et al. Genetic Engineering Approaches for Enhanced Production of
Biodiesel Fuel from Microalgae. Enzymatic Conversion of Biomass for Fuels Production
994: 255–270. doi:http://dx.doi.org/10.1021/bk-1994-0566.ch013
3] Mohsenpour SF, Richards B, Willoughby N, Spectral conversion of light for enhanced
microalgae growth rates and photosynthetic pigment production. Bioresour. Technol.
012;125:75–81.
4] Seo YH, Cho C, Lee J-Y, Han J-I, Enhancement of growth and lipid production from
microalgae using fluorescent paint under the solar radiation. Bioresour. Technol.
014;173:193–7.
5] Cheirsilp B, Torpee S, Enhanced growth and lipid production of microalgae under
[
1
[
2
[
2
[
mixotrophic culture condition: effect of light intensity, glucose concentration and fed-batch
cultivation. Bioresour. Technol. 2012;110:510–6.
[
6] Wang X, et al. Two-stage photoautotrophic cultivation to improve carbohydrate
production in Chlamydomonas reinhardtii. Biomass and Bioenergy 2015;74:280–7.
7] Hamedi, S., Mahdavi, M. A. & Gheshlaghi, R. Improved lipid and biomass productivities
[
in Chlorella vulgaris by differing the inoculation medium from the production medium.
Biofuel Res. J. 2016;10(2):410–6.
[
8] Wang CY, Fu CC, Liu YC, Effects of using light-emitting diodes on the cultivation of
Spirulina platensis. Biochem. Eng. J. 2007;37(1):21–5.
9] Seo YH, Lee Y, Jeon DY, Han J-I, Enhancing the light utilization efficiency of microalgae
using organic dyes. Bioresour. Technol. 2015;181:355–9.
10] Wondraczek L, et al. Solar spectral conversion for improving the photosynthetic activity
[
[

ACCEPTED MANUSCRIPT
in algae reactors. Nat. Commun. 2013;4:2047-2052.
11] Hyeyoun Jang, Synthesis and characterization of fluorescent dyes and their applications
[
for the enhancement of growth rate of Chlorella vulgaris. Master Diss. Seoul National
University, 2017.
[
12] Scarsella M, Belotti G, De Filippis P, Bravi M, Study on the optimal growing conditions
of Chlorella vulgaris in bubble column photobioreactors. Chem. Eng. Trans. 2010;20:85–90.
13] Sostaric M, Golob J, Bricelj M. Klinar D, Pivec A, Studies on the Growth of Chlorella
vulgaris in Culture Media with Different Carbon Sources. Chem. Biochem. Eng. Q.
009;23(4):471–7.
14] Liang Y, Sarkany N, Cui Y, Biomass and lipid productivities of Chlorella vulgaris under
autotrophic, heterotrophic and mixotrophic growth conditions. Biotechnol. Lett.
009;31(7):1043–9.
15] Ghosh AK, Crespi HL, Katz JJ, Fluorescence studies on deuterated Chlorella vulgaris.
Biochim. Biophys. Acta - Biophys. Incl. Photosynth. 1966;120(1):19–22.
16] Rippka R, Deruelles J, Waterbury JB, Herdman M, Stanier RY, Generic Assignments,
Strain Histories and Properties of Pure Cultures of Cyanobacteria. J. Gen. Microbiol.
979;111:1–61.
17] Bruno LB, Udhaya R, Sandhya S, Biomass and Lipid Productivity by Four Fresh Water
Microalgae in Photoreactor. J. Mod. Biotechnol. 2013;2(3):82–8.
18] Bligh EG, Dyer WJ, A Rapid method of total lipid extraction and purification. Canadian
Journal of Biochemistry and Physiology 1959;37(1):911-7.
19] Lakowicz JR, Principles of fluorescence spectroscopy, 3rd Edition, Joseph R. Lakowicz,
[
2
[
2
[
[
1
[
[
[
editor. Principles of fluorescence spectroscopy, Springer, New York, USA, 3rd edn, 2006.
doi:10.1007/978-0-387-46312-4
[
20] Lor M, et al. Photoinduced electron transfer in a rigid first generation triphenylamine

ACCEPTED MANUSCRIPT
core dendrimer substituted with a peryleneimide acceptor. J. Am. Chem. Soc.
2
002;124(33):9918–25.