Microalgae triacylglycerols content by FTIR spectroscopy
Roberta Miglio, Stefano Palmery, Mario Salvalaggio, Lino Carnelli, Federico Capuano & Raffaella Borrelli Journal of Applied Phycology ISSN 0921-8971 J Appl Phycol DOI 10.1007/s10811-013-0007-6
1 23
Your article is protected by copyright and all rights are held exclusively by Springer Science +Business Media Dordrecht. This e-offprint is for personal use only and shall not be selfarchived in electronic repositories. If you wish to self-archive your work, please use the accepted author’s version for posting to your own website or your institution’s repository. You may further deposit the accepted author’s version on a funder’s repository at a funder’s request, provided it is not made publicly available until 12 months after publication.
1 23
Author's personal copy J Appl Phycol DOI 10.1007/s10811-013-0007-6
Microalgae triacylglycerols content by FT-IR spectroscopy Roberta Miglio & Stefano Palmery & Mario Salvalaggio & Lino Carnelli & Federico Capuano & Raffaella Borrelli
Received: 30 August 2012 / Revised and accepted: 13 February 2013 # Springer Science+Business Media Dordrecht 2013
Abstract The present study aims to develop a methodology via Fourier transform infrared (FT-IR) spectroscopy for the semiquantitative determination of triacylglycerols (TAGs) in microalgal consortia, consistent with the use of the technique as process control. FT-IR spectroscopy has proved to be a powerful analytical tool for the identification of macromolecular pools (e.g., proteins, lipids, and carbohydrates) and in monitoring biochemical changes (including lipids) in response to nutrient stress or environmental modifications. In the Oocystis-based consortium under examination, the synthesis of neutral lipid in the form of TAGs can be induced, applying stress condition, and these lipids are suitable as biodiesel precursors. In the exponential growing phase, the consortium shows a low TAGs content, in the order of 5 %w, that can be increased till around 22 %w on ash free dry matter, after nitrogen starvation. Keywords Microalga . Starvation . FT-IR . Triacylglycerol . Lipid
Introduction Recent reports on next-generation biofuels produced via biomass are showing significant improvements on greenhouse gases emission savings, carbon footprint, and environmental sustainability with respect to first-generation biofuels. In the group of biomasses, phototrophic algae are R. Miglio (*) : S. Palmery : M. Salvalaggio : L. Carnelli : R. Borrelli ENI–Research Center for Non Conventional Energies, via Fauser, 4 28100 Novara, Italy e-mail:
[email protected] F. Capuano ENI–Refining & Marketing Division, via Laurentina 441, Rome, Italy
extremely promising, also for their ability to mitigate aquatic pollution and to sequester CO2 from power plants. In the photosynthetic pathway, algae are able to produce lipids, in the order of 10–60 % by weight, and other compounds like proteins and polysaccharides, which are reconverted among each others during different growing phases (Richardson et al. 1969). An important challenge for algal-derived biofuels is the utilization of open ponds (OPs) for large-scale cultivation. The environmental in and around OPs cannot be completely controlled and a number of factors can interfere with the growing phases and can influence productivity and biomass quality (consortium population). A key factor for microalgae in the energy market is the lipid content (Griffiths and Harrison 2009), which must be maximized and preferentially routed to high content of triacylglycerols (TAGs). An accurate method for lipid quantification in algal biomass is necessary for the purpose of identifying optimum species and growth conditions. Moreover, the availability of efficient and rapid analytical techniques, providing biomass and lipid screening, is essential as process control on large scale and could be a tool to take immediate actions without wasting time and thus money. Traditional lipid analyses require relatively large amounts of biomass (>1 g of dry biomass), are time- and chemicalconsuming, and are not particularly effective for the analysis of a large number of algal samples. The conventional method involves solvent extraction and gravimetric determination (Bligh and Dyer 1959); the neutral lipid quantification requires crude extract separation and fractions chromatographic analysis (Eltgroth et al. 2005). Fourier transform infrared (FT-IR) spectroscopy has proven to be a powerful analytical tool for the identification of macromolecular pools (e.g., proteins, lipids, and carbohydrates) and in monitoring biochemical changes (including lipids) in response to nutrient stress or environmental modifications (Murdock and Wetzel 2009 and references therein;
Author's personal copy J Appl Phycol
Stehfest et al. 2005; Dean et al. 2010; Mi-Kyung and Kyung-Hee 2009; Kiefer et al. 1997). However, most of the studies reported in the literature concern the relative lipid content evolution on band ratios basis as indeces of nutrient stress-related changes (Stehfest et al. 2005; Dean et al. 2010). Recently, NREL researchers (Laurens and Wolfrum 2011) have demonstrated the application of FTIR attenuated total reflectance spectroscopy for algal lipid quantification in dried biomass. Using multivariate calibration models, they have estimated lipid levels in algae strains. All the experiments were based on exogenously added lipids to algal biomass, with a maximum content of 5 %w. The present study aims to develop a fast and highthroughput FT-IR method for the semiquantitative determination of TAGs in algal consortia, whose calibration is based on the method of exogenous lipid standard addition. Even if in principle the calibration is species-specific, this methodology is reliable as screening technique to monitor lipid content (a basic factor for the production of algal biomass-based biofuels) and is consistent with the use of the same as process control analysis. A similar approach could be used with carbohydrates and/or phospholipids.
Materials and methods Apparatus setup for algae cultivation A photobioreactor (PBR) in Pyrex with base dimensions 11×5.5 cm and height 18.5 cm was used for algae cultivation. It was open on the top (non-sterile condition). The pH was measured and maintained at the desired set point by bubbling CO2. Constant temperature was achieved with a thermostatic bath and an immersed coil. The light, necessary for growth, was provided by fluorescent lamps (lamps OSRAM Dulux D/E, 26 W/840, Lumilux cool white, T=4,000 K, G24q-3) placed on both sides of the Pyrex PBR. Photosynthetic active radiation (PAR) irradiance (400–700 nm) was measured with QSL-2201 supplied by Biospherical Instruments Inc., equipped with a scalar irradiance sensor. An image of the PBR is shown in Fig 1. The growth medium was prepared in accordance with Table 1; the water conductivity was 14 mS cm−1.
(1)
(2) (3) (5) (5)
(4)
Fig. 1 Photobioreactor for microalgae cultivation. (1)=CO2 sparger inlet; (2) inlet and outlet of thermo-stating coil; (3)=pH meter; (4)= irradiance sensor; (5) fluorescent lamps
was affected by vegetative cell phase. Dry weight was determined by filtering algae suspension with 0.45-μm porosity membranes and heating in flowing air at 105 °C for 2 h. The dry weight was equivalent to the difference between the final weight and the membrane weight over the algae suspension volume. The concentration of nitrate as nitrogen (N) in the culture medium was followed with pre-measured kits and a HANNA 83099 series photometer. The kit used was: HI 93728–0, suitable for a concentration range between 0 and 30.0 mg L−1 with a resolution of 0.1 mg L−1. FT-IR algal biomass was analyzed in transmission mode by drop-casting on CaF2 optical windows. On the basis of preliminary trials performed to determine the optimal cell aqueous suspension density and to assess the feasibility and operational modalities of this technique, the following procedure was developed for cell deposition: & &
Analytical methods & Algal growth was monitored by optical density (OD), after calibration of OD signal with dry weight measurements. The OD measurements were performed with a Hanna Instruments HI 83099 photometer at 610 nm (close to one of chlorophyll absorption maxima). The OD reading was taken within the range 0.25–1.2 by diluting the suspension with distilled H2O, the resolution was 0.01. OD signal calibration (linear correlation)
& &
About 40 mL of algal suspension (at about 200 g m−3 or equivalent concentrations) was collected in a Falcon tube; Cells were harvested by centrifugation at 2,400×g for 10 min; The concentrated phase was re-suspended in 40 mL of an isotonic 0.2 M NaCl solution; This suspension was centrifuged again at 2,400×g for 10 min, and the supernatant was removed; the concentration was 5–10 %w; A few milligrams of this dense phase were cast on a CaF2 optical window and oven-dried at 40 °C under vacuum or at 105 °C for 15 min;
Author's personal copy J Appl Phycol Table 1 Growth medium
Thermogravimetric analysis was performed under air atmosphere at the heating rate of 10 °C min−1 from room temperature up to 930 °C using a Perkin Elmer TGA7 thermobalance, Pt crucibles and sample masses of 3–6 mg. Thermogravimetric analysis was used for ash determination in algae samples. They were also performed to correlate thermal decomposition profiles with TGAs content, confirming lipid accumulation by a different technique. An optical microscope was used to check contamination by grazers, pathogens and other fast-growing heterotrophs during the cultivation. A Water-Pam (Walz GmbH) fluorometer was used for fluorescence analysis.
(g m−3) 500 45 6.3 500 25 5,500 150 3,100 2.85 1.00 0.50 0.50 0.50
NaNO3 KH2PO4.H2O FeCl3 CaCl2 FeSO4.7H2O NaCl KCl MgSO4.7H2O H3BO3 MnCl2.4 H2O CuSO4.5 H2O ZnSO4.7 H2O Na2MoO4
Cell growth and starvation To determine growth performances of the microalgae consortium (Oocystis-based consortium), several growth tests were run at different irradiances, in the range 100– 1,600 μmol photons m−2 s−1, in the apparatus described in the “Apparatus setup for algae cultivation” section. Tests conditions were: microalgae initial concentration about 50 g m−3 to avoid self-shading, no nutrients and CO2 limitations, temperature 23±0.5 °C, and pH=7±0.5 for about 48 h. After the necessary period of acclimatization, the growth test started at time t°, then samples were taken for determining the algae concentration as a function of time C(t), up to a final concentration of about 300 g m−3. Beyond that threshold, self-shading effects heavily modified light intensity. The specific growth rate (μ), associated to test light and temperature conditions, was calculated using Eq. (1),
The so prepared deposition was then analyzed in transmission mode at room temperature. FT-IR spectra were collected with a Thermo Nicolet Nexus 8700 FT-IR spectrometer equipped with a DTGS detector in the 4,000–1,000 cm−1 range, with 64 scans and a resolution of 2 cm−1. A soxhlet apparatus was used to obtain extracts from dried (105 °C for 2 h) algae samples (solvent ratio isopropanol/ hexane=1:3 w/w, ratio solvent/solid=60 w/w, temperature= 66 °C, cycles number=30). The solvent was later evaporated from the extracted lipids. Extracted lipids after transesterification with methanol in H2SO4 were analyzed by highresolution gas chromatography with flame ionization detector. Algae elementary analysis was performed. A Thermo Fisher Flash 2000 analyzer was used for C, H, N, S determination and a Thermo Fisher EA1110 analyzer for O2 determination. This analysis was to check if N content of algae was constant during growing and starvation phases. Protein N amidic bonds, if constant, could be considered as an internal standard in FT-IR analysis for TAGs semiquantitative determination. Fig. 2 Oocystis-based consortium inhibition curve at 23 °C and pH=7
CðtÞ ¼ Cðt-Þ expð μ t Þ
ð1Þ
All these data are summarized in the consortium inhibition curve (Fig. 2). PBR laboratory cultivations were conducted to produce growth and starvation of the aforementioned algal consortium and to monitor the TAG content
0.07 0.06
K-4
K-5
K-3
K-6 K-7
0.05
K-8
K-2
µ (h-1 )
0.04 K-1
0.03 0.02 0.01 0.00
K-0
-0.01 0
200
400
600
800
1000
P.A.R. Irradiance ( µ mol photons /(m2 *s) )
1200
1400
1600
Author's personal copy J Appl Phycol 2.0
1.0
0.5
1.5
Absorbance
Fig. 3 FT-IR spectra; in the insert bands area for semiquantitative determination of TAGs: triolein (dotteddashed line), algae sample (solid line), and extract from algae (dotted line). In the insert: A1: νC=O ester group (shaded area), A2: νC=O Amide I plus δ N-H Amide II (horizontal lines)
0.0
1800
1600
1400
1.0
0.5
0.0 3500
3000
2500
2000
1500
1000
-1
Wavenumber (cm )
Wave number (cm−1)
Assignment
Functional groups
if the lipid accumulation rate during starvation was influenced by the growing phase duration. The P1 test was carried out in batch mode from an initial algal concentration of ~400 g m−3 and was kept in batch mode for a total duration of 14 days. It was in N absence from day 7 of cultivation (starvation phase). The P2 test was carried out in semi-continuous way for about 17 days with a dilution factor suitable to maintain an algal concentration of ~500 g m−3. Starting from day 18, it was maintained in batch mode. It was in N absence from day 21 (starvation phase). Later, the culture was diluted again, N was added, and the culture came back to the exponential growing phase. Samples from P1 and P2 tests were harvested over time to follow the parameters of interest. The harvested samples were centrifuged at 2,400×g for 10 min.
~2,960
νas CH3
CH3 methyl group
~2,930 ~2,850
νas CH2 ν CH2, CH3
Procedure for the semiquantitative determination of TAGs via FT-IR
via FT-IR. For the growth-starvation tests (cultivation code: P1, P2), the culture medium was inoculated with an exponentially growing inoculum. The cultures were grown under PAR constant photon flux of 450–500 μmol photons m−2 s−1, close to the inhibition curve maximum specific growth rate. The cultures were kept 24/24 h at constant temperature of 23 °C. The pH was kept constant at pH 7 dosing (CO2) in nitrogen (N2); the growth medium, described in Table 1, had an initial N concentration of 100 g m−3. Two experiments P1 and P2 were carried out with different growing phase periods with non-limiting N to understand
Table 2 Main absorption for algae in the IR window
~1,745 ~1,655 ~1,545–1,540 ~1,455 ~1,398–1,380
CH2 methylene group CH2, CH3 methyl & methlyene group ν C=O Ester of lipids and fatty acids ν C=O (Amide I) Protein δ Ν−Η (Amide Protein II) δas CH3 & δas CH3, CH2 methyl & CH2 methlyene group δs CH2 & δs CH3 CH2, CH3, COO¯ νs C-O of COO¯
~1,240
νas P=O
~1,150
ν C-O-C
Phosphodiesters e phospholipids Polysaccharides carbohydrates
Infrared radiation can be absorbed by molecules promoting vibrational transitions. On the basis of the IR absorption bands, functional groups can be identified and, upon calibration curves, corresponding macromolecules (e.g., proteins, lipids, carbohydrates, and nucleic acid) can be quantified. In Fig. 3, for example, an FT-IR spectrum of dried microalgae cast on CaF2 was is reported in comparison to those of an algae extract and triolein, the latter two spectra having been collected in transmission simply by sandwiching the liquid between two optical windows. The microalgae spectrum clearly showed the green algae characteristic features (Murdock and Wetzel 2009); in particular, the relative
Author's personal copy J Appl Phycol 50 45 Base
40
P1-6 Absorbance
35 conc. (%ps)
30 25
0.5
20 15
0.0 1800
10
1600
1400 -1
Wavenumber (cm )
5 0
f atty acids
Fig. 4 Fatty acids distribution (GC analysis) of starved sample (P1-6) and initial sample (Base). In the insert: FT-IR spectra acquired during cultivation, showing lipid accumulation (1,745 cm−1 band increase)
high intensity bands at 1,200–1,000 cm−1 due to C-O-C stretch of polysaccharides as starch and cellulose (cell wall and energy storage products). Protein content was related to amide I (C=O stretch) and amide II (N-H bending) bands around 1,655 and 1,545 cm−1, respectively (Stehfest et al. 2005; Dean et al. 2010). The intense and broad absorption centered at 3,350 cm−1 was ascribable to the stretching of hydrogen bonded O-H and N-H groups (carbohydrates, proteins, etc.), while the 1,240 cm−1 absorption was mainly related to P=O stretches (nucleic acids, phosphoralated proteins, and polyphosphate storage products). Concerning the TAGs content, an analysis of the FT-IR spectra revealed the presence of the band characteristic of C=O stretching absorption at ~1,745 cm−1, which corresponded to the ester groups of TAGs (whose corresponding C-O-C stretch was observed at ~1,160 cm−1) and the occurrence of bands characteristic of C-H bonds vibrations at ~2,960–2,850 cm−1, primarily of
aliphatic chains. The main bands of interest on dried algae samples are summarized in Table 2. Apart the greater quantity of water in the extract with respect to triolein (water hydrogen bonded stretching centred at 3,300 cm-1 and the corresponding deformation at 1,630 cm−1), the similarity between the spectra was straightforward and clearly indicated that triolein was suitable as reference for standard addition method. As a further confirmation, the fatty acid analysis obtained from the hydrolysis of algae lipids (Fig. 4) showed fatty acids centered on C18. The overlapping of the triolein bands with those of algae (Fig. 3) confirmed that TAGs were an important algae component. To obtain semiquantitative data of the samples TAGs content, a calibration of FT-IR spectra was performed with a set of references, prepared by addition of triolein to a sample (base) of the same powdered algal consortium. The
Table 3 FT-IR analysis calibration for TAGs, -CH3, and -CH2Standard
Triolein (%wt/wa)
CH3 (%wCH3/wa)
CH2 (%wCH2/wa)
IR determinations A1/A2
B/A2
C/A2
1 2 3
4.7 8.2 14.1
0.3 0.4 0.8
3.5 6.1 10.6
0.029 0.044 0.081
0.0016 0.0018 0.0023
0.0029 0.0039 0.0048
4 5 6 7
16.1 20.1 29.7 33.6
0.9 1.1 1.6 1.8
12.1 15.1 22.3 25.2
0.105 0.138 0.184 0.167
0.0030
0.0068
0.0036
0.0078
Triolein (wt/wa =triolein weight/algae dry weight); CH3 (wCH3/wa =CH3weight/ algae dry weight); CH2 (wCH2/wa =CH2weight/algae dry weight) A1 absorbance at 1,745 cm−1 , assigned to ν C=O of triolein ester group; A2 absorbance between 1,780 and 1,480 cm−1 , assigned to Amide I and Amide II groups; B absorbance at 2,960 cm−1 , assigned to the νas (CH3) of triolein methyl groups; C absorbance at 2,930 cm−1 , assigned to the νas (CH2) of triolein methylene
Author's personal copy J Appl Phycol 30
40
30
y = 172,7x - 0,38 R² = 0,93
-CH3 -CH2- (%wCH/wa)
TAGs (%w t/wa)
35
25
20 15 10
5 0 0.00
0.05
0.10 (A1/A2)
0.15
added CH3 y = 342,45b - 0,34 R² = 0,94
25 20
15 10
added CH2 y = 2593,c - 4,24 R² = 0,90
5 0 0.000
0.20
0.002
0.004 B/(A2)
0.006 C/(A2)
0.008
0.010
Fig. 5 a, b Equations for FT-IR spectra calibration. a TAGs (%w/w) correlation with CO ester Area (A1) and Amide I+II bands area (A2) ratio. b –CHn (%w/w) correlation with -CHn bands high (B or C) and Amide I+II bands Area (A2) ratio
1,780 and 1,480 cm−1 (due to the Amide I and Amide II bands of the microalga), using linear baselines (Fig. 3 insert). The prepared standards and area estimations are listed in Table 3. The assumption of the Amide group absorptions as internal reference bands was based on the fact that only minor changes in the protein concentration (N determination) were observed under non limiting nutrients conditions (growth medium containing N; Dean et al. 2010). The Amide I band was also used by others as thickness band to compare the progressive increase of lipids (band at 1,745 cm−1) taking into account differences in the thickness of the analyzed samples. Triolein added contents, y (%wt/wa) as a function of (A1/A2), are reported in a graph (see Fig. 5a) and, as expected from the Beer– Lambert Law, the correlation obtained was linear and described by Eq. (2).
initial TAGs content of the base sample was determined by extraction with solvent and HR-GC analysis. The extraction was performed in a soxhlet apparatus. The solvent was evaporated, the extract was weighted and resulted 4.2 %w of the initial dry weight. After transesterification with methanol, the extract was analyzed by HR-GC. The quantity that eluted from GC was 23.7 % (by area), and ~74.8 % of the total peak amount corresponded to fatty acids. The base sample TAGs amount resulted, as a consequence, 0.74 %w (=4.2×23.7×74.8 %). The base sample was dried and finely ground using a cryo-grinder. Reference samples were then prepared by milling in a mortar about 100 mg of the powder (wa) with a known quantity of triolein (wt) in the range 5–50 mg, obtaining a smooth paste of known composition (see Table. 3). The so obtained triolein mull was then analyzed in transmission mode by deposition on CaF2 optical window. Each reference sample spectrum was elaborated with Omnic 8.0 spectroscopic software and used for the construction of calibration curve. For all references, the triolein ester band area (A1) was measured at 1,745 cm−1 (C=O stretching of the ester group) and the “thickness” area (A2) between
ð2Þ
The value of 0.38 % wt/wa represented the base sample TAGs initial content.
4000
45
N
P1-1
pH
3000
40
P1-2
35
T 2500
TAGs
P1-3 P1-0
2000
P1-5
30 25
P1-4
20
1500
15 1000 10 500
5
0
0
0
1
2
3
4
5
6
7 t (day)
8
9
10
11
12
13
N*10 (ppm) - pH - T(°C) - TAGs (%w)
50 algae
3500
algae (g/m3 )
Fig. 6 Cultivation test P1 profile as a function of time with indications of samples taken for FT-IR analysis, algae concentration (grams per cubic meter), temperature (°C), pH, and N from –NO3 concentration (parts per million) in the growing medium
y½%wt =wa ¼ 172:7 ðA1=A2Þ 0:38
Author's personal copy J Appl Phycol Fig. 7 Cultivation test P1 and P2: TAGs content (percentage weight) as a function of starvation days
30 P2
TAGs (%w)
25
P1
P1-4
P1-3
P1-5
20 15
P1-1
P1-2 P2-3
P2-4
P2-2
10
P2-1
5 0 0
1
To calibrate the analysis for the total amount of TAGs (added triolein plus unknown initial value of the base sample used for analysis calibration), a translation was operated obtaining Eq. (3) y½%wt =wa ¼ 172:7 ðA1=A2Þ
2
3 4 t (day of starvation)
5
6
7
For total amount –CH3 or –CH2 estimation, calibration Eqs. (7) and (8) were derived from Eqs. (5) and (6): CH3 ½%wCHn =wa ¼ 342:45 ðB=A2Þ
ð7Þ
CH2 ½%wCHn =wa ¼ 2; 593 ðC=A2Þ
ð8Þ
ð3Þ
TAGs content in term of %w was obtained with Eq. (4): y½%w ¼ 1=ð 1 þ 1=y½%wt =wa Þ
ð4Þ
The TAGs screening of unknown samples was based on calibration curve (3). In addition, for each reference standard, the absorbance (B) at 2,960 cm−1 (tangent baseline from 2,780–3,030 cm−1) was measured, band assigned to triolein methyl groups νas (CH3) and the absorbance (C) at 2,930 cm−1 (tangent baseline from 2,780– 3,030 cm−1), assigned to triolein methylene units νas (CH2; see Table 2). The –CH2- triolein content was 0.750 of triolein MW, – CH3 was 0.0548. The Amide band (A2, see Table 3) was used again as an internal reference to compare the increase of –CH2- (band at 2,930 cm−1) or –CH3 (band at 2,960 cm−1). The –CH3 or –CH2 [%wCHn/wa] correlations with (B/A2) or (C/A2) were linear (see Fig. 5b) and were described by Eqs. (5) and (6). CH3 ½%wCHn =wa ¼ 342:45 ðB=A2Þ 0:34
ð5Þ
CH2 ½%wCHn =wa ¼ 2; 593 ðC=A2Þ 4:24
ð6Þ
The observed linear correlation confirmed the goodness of the method.
Results The inhibition curve: specific growth rate (μ) vs. irradiance (I° PAR.) of the algal consortium under examination is shown in Fig. 2. Each specific growth rate value was calculated for an observation period of 21–24 h (t−t°), during which the cultivation was kept at constant temperature, pH, and incidence irradiance (I°). The concentration doubling time (t2) was extimated by Eq. (9) and resulted in the range 10.7–20 h. t2 ðhÞ ¼ ln 2=μ h1
ð9Þ
The specific growth rate data were interpolated using Haldane Eq. (10) (Haldane 1930):
Table 4 Test P1—elementary analysis of dry samples Sample
t (h)
N (%w)
C (%w)
H (%w)
S (%w)
O (%w)
Δ (%w)
N in H2O (g m−3)
P1-0 P1-6
96 240
3.4 3.4
45.0 47.2
6.8 7.0
0.6 0.3
30.4 25.9
13.8 16.1
37 0
t time; Δ balance difference to 100 %, N in H2O nitrogen concentration in the growth medium
Author's personal copy J Appl Phycol Table 5 Comparison of TAGs determination by GC and FT-IR Sample
a b c d * Ash (%w) Extract (%w) GC recovery (%w) TAGs in GC analysis (%w) TAGs extraction (%w) TAGs FT-IR (%w)
P1-0 P1-6 P2-6 *=(b)(c)(d)/(100−a)/ 100
13.8 16.1 15.0
16.4 26.0 19.0
65.0 77.5 79.5
85.0 92.3 97.3
10.5 22.2 17.3
10.8 22.3 19.3
Ash ash content of dry algae sample, extract solvent extractable content of dry algae sample, GC recovery fraction of extractable content recovered from GC analysis, TAGs in GC analysis TAGs fraction in GC analysis, TAGs extraction TAGs content of dry algae sample determined by solvent extraction and GC analysis normalized to ash-free base, TAGs FT-IR TAGs content of dry algae sample determined by FT-IR analysis
μ ¼ μ* max
I 2 Me K1 þ I þ KI i
ð10Þ
obtaining the following parameters: μmax (maximum theoretical specific growth rate)=0.103 h−1; KI (onset of saturation constant) =215 μmol photons m−2 s−1; Ki (onset of inhibition constant)=2,340 μmol photons m−2 s−1; Me (metabolism constant)=0.001 h−1. The cultivation for growth and starvation was performed at an irradiance value (I°=455 μmol photons m−2 s−1) close to the flat profile of the inhibition curve. Figure 6 shows the P1 cultivation test. The algae concentration was very high, and cell settling was observed in the bottom part of the reactor. The autoflocculation phenomenon started, more or less, with the starvation phase. The reported algal concentrations correspond to the fraction in suspension during the test due to CO2 and N2 bubbling. The FT-IR spectrum of harvested samples, during both P1 and P2 tests, showed the triolein group bands (2,960 νasCH3; 2,930 νasCH2, 2,850 νCH2 +CH3, 1,745 ν R-COOR) and confirmed the suitability of triolein choice in the calibration analysis. An estimation of sample TAGs content was made using the calibration Eq. (3). As shown in Fig. 7, TAGs increased, starting from the day in which the culture medium was Ndeprived (t=7 days) and then stabilized after ~5 days. Fig. 8 Yield (Y) as a function of irradiance (I) in PAM fluorometry for samples harvested during cultivation
Algae protein concentration was assumed constant to use the FT-IR method as semiquantitative TAGs analysis. Protein in algal biomass may be estimated (%w) by the N elementary analysis, as a first approximation, by multiplying N (percentage weight)×6.25. Nitrogen elementary analysis of initial and final samples was the same (see Table 4) and verified the previous assumption. The result further confirmed the possibility of using the bandwidth measured in FT-IR Amide I+II as a constant reference to proportion TAGs content of microalgae. To validate the FT-IR semiquantitative procedure as an estimation of algal biomass TAGs, a comparison between FT-IR technique and the more conventional extraction/GC technique was made on a few samples. The results are shown in Table 5. The FT-IR determinations were base on ash-free dry matter, so for comparison, the results of extraction/GC technique were recalculated on the same base (ash-free). GC analysis of fatty acids obtained from sample after starvation phase (P1-6) and sample before starvation phase (base sample) is shown in Fig. 4. The comparison showed that during starvation, the studied consortium accumulated neutral lipid (see FT-IR spectra insert in Fig. 4), mainly C18. FT-IR semiquantitative data of –CH2- and –CH3 content of algae samples further confirmed that during starvation, the studied consortium substantially moved from C16 to
Y (-) 0.7 P1-0
0.6
P1-1
0.5
P1-2 P1-3
0.4
P1-4
0.3
P1-5
0.2 0.1 0 0
200
400
600 800 1000 I (µmol photons/(m2 *s))
1200
1400
1600
Author's personal copy J Appl Phycol Table 6 Test P1—FT-IR details (band height ratios) Sample Ratio (TAGs/Amid) Ratio (PO/Amid) Ratio (TAGs/Carb) P1-0 P1-1 P1-2 P1-3 P1-4 P1-5
0.58 0.53 0.63 0.99 1.09 1.10
0.25 0.23 0.24 0.27 0.27 0.26
0.94 0.59 0.71 1.24 1.41 1.49
TAGs band height at 1,745 cm−1 , assigned to ν C = O of triolein ester group; Amid band height at 1,655 cm−1 , assigned to ν C = O of proteins (Amid I); PO band height at 1,240 cm−1 , assigned to νas P = O of phospholipid group; Carb band height at 1150 cm−1 , assigned to ν C-O-C of polysaccharides
carbohydrates and a likely carbohydrate conversion into TAGs. The thermogravimetric curves analysis of base sample (TAGs
