This article was downloaded by: [Aston University] On: 16 January 2014, At: 23:45 Publisher: Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Environmental Technology Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/tent20

Harvesting Chlorella vulgaris by natural increase in pH: effect of medium composition a

a

a

a

Thi Dong Phuong Nguyen , Matthieu Frappart , Pascal Jaouen , Jérémy Pruvost & Patrick ab

Bourseau a

LUNAM Université, Université de Nantes, CNRS, GEPEA, UMR 6144, CRTT, 37 bd de l'Université, BP 406, Saint-Nazaire Cedex 44602, France b

Univ. Bretagne-Sud, EA 4250, LIMATB, Lorient F-56100, France Published online: 08 Jan 2014.

To cite this article: Thi Dong Phuong Nguyen, Matthieu Frappart, Pascal Jaouen, Jérémy Pruvost & Patrick Bourseau , Environmental Technology (2014): Harvesting Chlorella vulgaris by natural increase in pH: effect of medium composition, Environmental Technology, DOI: 10.1080/09593330.2013.868531 To link to this article: http://dx.doi.org/10.1080/09593330.2013.868531

PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http:// www.tandfonline.com/page/terms-and-conditions

Environmental Technology, 2014 http://dx.doi.org/10.1080/09593330.2013.868531

Harvesting Chlorella vulgaris by natural increase in pH: effect of medium composition Thi Dong Phuong Nguyena , Matthieu Frapparta , Pascal Jaouena , Jérémy Pruvosta , and Patrick Bourseaua,b∗ a LUNAM

Université, Université de Nantes, CNRS, GEPEA, UMR 6144, CRTT, 37 bd de l’Université, BP 406, Saint-Nazaire Cedex 44602, France; b Univ. Bretagne-Sud, EA 4250, LIMATB, Lorient F-56100, France

Downloaded by [Aston University] at 23:45 16 January 2014

(Received 3 October 2013; accepted 18 November 2013 ) The freshwater microalga Chlorella vulgaris was harvested by autoflocculation resulting from the precipitation of magnesium or calcium compounds induced by a slow increase in pH in the absence of CO2 input. Autoflocculation was tested in two − culture media with, respectively, ammonium (NH+ 4 ) and nitrate (NO3 ) ions as nitrogen source. The culture pH increased + because of photosynthesis and CO2 stripping. pH rose to 11 after 8 h in the NO− 3 medium, but did not exceed 9 in the NH4 − medium. No flocculation took place in any of the media. Autoflocculation tests were repeated in the NO3 -based culture medium by progressively increasing the concentrations of Ca2+ and Mg2+ until inorganic compounds precipitated and flocculated microalgae. The minimal concentrations for flocculation were found to be 120 mg Ca2 L−1 and 1000 mg Mg2+ L−1 . These values were, respectively, 3.5 times and 20 times higher than those allowing flocculation by NaOH addition. Energy-dispersive X-ray spectroscopy, zeta potential measurement, and ionic chromatography suggest that the mechanisms involved are different. The rate of cell removal was close to 90% in both cases, but cells were more concentrated in the aggregates obtained by magnesium compound precipitation, with an estimated concentration close to 33 g (dry matter) L−1 , against 19 g L−1 for calcium phosphates. Keywords: dewatering; harvesting; Chlorella vulgaris; autoflocculation; pH-induced flocculation; medium composition

1. Introduction Microalgae contain valuable compounds such as pigments (e.g. astaxanthin, B-phycorerythrin, and C-phycocyanin), lipids, and polysaccharides. Until now microalgae have mainly been farmed for juveniles and shellfish larvae feeding in aquaculture or to produce high-value products for the cosmetic and nutraceutical industries. However, their exploitation is now developing in the food and livestock feed sectors, and more recently in the mass markets for energy (biofuel) and green chemistry.[1–5] Harvesting is a critical step in microalgae exploitation, as it can represent 20–30% of the overall process costs.[6,7] Centrifugation, the standard technique for niche markets, is ruled out for mass markets, and for biofuel in particular, in favour of cheaper harvesting methods such as flocculation or membrane processes, according to the dewatering rate needed.[8] Microalgae can be harvested and pre-concentrated by the flocculation methods used in wastewater treatment based on aluminium or iron salts, or polyelectrolyte addition.[9–11] Several other methods have also been investigated, including bioflocculation by microalgal polymers and/or bacterial exopolysaccharides,[12–14] electrocoagulation,[15] electroflocculation,[16] or combined flocculation and flotation.[17] Some conditions ∗ Corresponding

author. Email: [email protected]

© 2014 Taylor & Francis

can cause microalgae to autoflocculate with no reactant addition. Autoflocculation has been attributed to various causes including (1) the ability of some microalgal strains to autoflocculate at pH values that are weakly basic (< 8) or moderately basic (≈ 10 − 11),[18,19] (2) bridging between microalgal cells by extracellular organic matter excreted by the microalga itself [12] or by other microorganisms,[13] and (3) the precipitation of salts contained in the culture medium at high pH (≈ 8.5 − 10.5) reached by a slow, natural rise due to photosynthesis [20] and the stripping of dissolved CO2 by air bubbling. Only this last type of flocculation is considered here. For culture media with a suitable composition, the pH may reach the solubility limit of some salts such as calcium phosphates (pH ≈ 8.5 − 9.0) or carbonates (pH ≈ 9.1 − 9.5), or magnesium hydroxide (pH ≈ 10.2 − 11.5).[20–28] Flocculation by precipitation of inorganic compounds has been reported to result from mechanisms similar to those involved in the flocculation of colloidal particles, namely charge neutralization, sweep flocculation, or weighting effects.[20,21,23,25] It is noteworthy that the mechanisms of microalgae flocculation by salt precipitation have been discussed most often in the framework of studies on chemical flocculation (alkali addition), but rarely in research

Downloaded by [Aston University] at 23:45 16 January 2014

2

T.D.P. Nguyen et al.

on autoflocculation, with the notable exception of the pioneering work of Sukenik and Shelef.[20] Chlorella vulgaris, a unicellular non-flagellate, green fresh water microalga, is widely cultivated for nutrition or food complement purposes because it is easy to grow, and it contains proteins, pigments, and fatty acids with high nutritional value.[29] C. vulgaris is also recognized as a potential candidate for microalgae biofuel because of its ability to accumulate neutral lipids with high productivity.[5,30] In this study, the ability of C. vulgaris to autoflocculate was tested in two culture media with, respectively, ammonium (Sueoka’s medium) and nitrate ions (modified Bold’s basal (mBB) medium) as nitrogen source. The minimal ion concentrations for flocculation were first estimated by chemical flocculation, with a sharp increase in pH obtained by adding 1 N sodium hydroxide. The autoflocculation of C. vulgaris by the precipitation of calcium phosphate or magnesium compounds was then tested in the two media. The concentrations of Mg2+ or Ca2+ ions in the culture medium that allowed autoflocculation were established.

2.

Materials and methods

2.1. Strain and growth conditions Chlorella vulgaris is a eukaryotic unicellular green freshwater alga.[31] Its cells are spherical or ellipsoidal with a mean diameter of around 4–5 μ m. The strain used in the study was C. vulgaris 211–19 (SAG, Germany), chosen for its ability to assimilate both ammonium (NH+ 4 ) and nitrate ions (NO− ), with a preference for ammonium.[32,33] The 3 strain was, therefore, grown in two media. The first one, with NH+ 4 ions as nitrogen source, was adapted from the autotrophic Sueoka’s medium [34] described by Harris.[35] The second was a mBB medium with NO− 3 as the nitrogen source.[36] The nutrient concentrations in the media given in Table 1 were adjusted so as to reach a biomass concentration of 2 g DM L−1 (DM, dry matter) in a batch culture without mineral limitation. Nitrate-based media are widely used to grow algae, and the ammonium-based medium was tested because it is highly assimilated by the microalga and can, therefore, be recycled in the photobioreactor (PBR) after cell harvesting without causing accumulation of minerals.[37,38] Culture media were sterilized before feeding a 1 L flat airlift PBR operated in continuous mode. The medium flow rate was regulated at 720 mL day−1 . The PBR was illuminated by a panel of white cold light LEDs at a flux Table 1.

of 150 μ mol m−2 s−1 , and pH and temperature were controlled, respectively, at 7.5 by CO2 injection and 25◦ C. The protocol allows recovery of a microalgae suspension with a final biomass concentration of ca. 0.8 g DM L−1 .

2.2.

Flocculation experiments

The minimal concentrations of Ca2+ and Mg2+ required for flocculation were first estimated by sharply increasing the pH by NaOH addition. Chemical flocculation by base addition is often considered as a surrogate technique of autoflocculation.[20,25,26] Autoflocculation was then tested in the two media. As no flocculation was observed after 9 h in the initial media, experiments were repeated in one medium (mBB) doped with magnesium or calcium in order to determine the minimal concentrations of [Ca2+ ]min and [Mg2+ ]min for flocculation. Chemical flocculation and autoflocculation tests were run at T = 24 ± 1◦ C in test tubes containing 120 mL of a 0.4 g DM L−1 microalgal suspension obtained by twice diluting the 0.8 g DM L−1 harvest with fresh, cell-free culture medium. Test tubes were placed in front of a light panel similar to that used for the microalga culture in PBR. Ca2+ or Mg2+ concentrations in the medium were increased when required by adding small volumes of stock solutions of MgSO4 .7 H2 O or CaCl2 .2 H2 O at 50 g L−1 in increments of 20–30 mg L−1 . Efficient mixing was achieved by a magnetic stirrer set at 500 rpm for at least 10 min. Chemical flocculation tests to estimate the minimal concentration of Ca2+ or Mg2+ were performed by fast addition of 1 N NaOH with stirring under a light intensity of 150 μ mol m−2 s−1 until the pH reached 11.8. It was considered that the minimum concentration corresponded to a settling efficiency of 80%. As cells settled, the suspension clarified so that the fraction of the intensity emitted by the light panel that was transmitted through the culture increased. Thus the settling efficiency was evaluated with a photovoltaic cell placed 5 cm below the suspension surface. All experiments on a given medium were performed in triplicate on withdrawals taken from the same culture over a total period of 4 days (one experiment per day). In the autoflocculation tests, the test tube was exposed to a mean illumination of 500 μ mol m−2 s−1 with the CO2 regulation stopped. Gentle mixing was ensured by bubbling sterilized atmospheric air injected at the base of the test tube (Figure 1). The pH then rose slowly because of the CO2 depletion of the medium by the photosynthesis and

Ionic composition of the two culture media (mg L−1 ).

Sueoka mediuma mBBa a 0.5 mL L−1

Na+

K+

NH+ 4

Mg2+

Ca2+

Cl−

NO− 3

HCO− 3

SO2− 4

PO3− 4

230 541.5

28.7 28.7

243.9 –

13.8 13.8

6.8 6.8

493.1 28.7

– 840.4

610 610

54.6 54.6

69.8 69.8

of Hutner’s trace elements solution [39] is also added to the solution.

Environmental Technology

Downloaded by [Aston University] at 23:45 16 January 2014

Figure 1.

Schematic diagram of settling by NaOH addition (A) or by self-flocculation test (B).

the stripping of dissolved CO2 . Experiments on a given medium were performed in duplicate on withdrawals taken from the same culture (one experiment per day). Samples were taken regularly for analysis. 2.3.

3

Settling efficiency and cell concentration in the aggregate zone

The performance of cell recovery by flocculation was evaluated using two criteria: (1) settling efficiency and (2) estimated cell density in the aggregate zone. Settling efficiency E was taken as the percentage of cells flocculated, and was computed according to the Beer–Lambert law as E=

OD682 i − OD682 s , OD682 i

(1)

where OD682i and OD682s are, respectively, the optical densities of the processed suspension and supernatant measured at 682 nm with a Lambda 2S spectrophotometer (Perkin Elmer). OD682s was measured 20 min after the beginning of chemical flocculation tests or after the total decantation of cells for autoflocculation tests. Floc density was estimated by computing the cell concentration in the flocculated zone by mass balance as Vi .Ci .E (g DM L−1 ), (2) Vf where Vi is the initial volume of the suspension (mL), Vf the volume of the cell aggregate zone after settling (mL), and Ci the cell concentration in the processed suspension (g L−1 ). Cest = Ci +

2.4. Analysis Cell concentration is expressed on a DM basis. The stability of the PBR operation was monitored by checking the constancy of the cell concentration in the PBR, and by assay of the total chlorophylls and carotenoids. Total chlorophylls are roughly proportional to the cell content, but the measure is available more quickly. A change in the

carotenoid-to-chlorophyll ratio reveals some stress in the culture. Changes in the medium composition during autoflocculation were followed by ion chromatography. Cell superficial charge and its time course during the pH increase and salt precipitation were evaluated by measuring the zeta potential (ζ ) of cells: ζ gives qualitative information on the sign and order of magnitude of cell charge. Finally, the structure and the chemical composition of mineral deposits in the flocs were examined by scanning electron microscopy with energy-dispersive X-ray spectroscopy (SEM–EDX). 2.4.1. DM and pigments on cells Biomass dry weight was determined by gravimetry. The sample was filtered through a rinsed glass fibre filter (Whatman GF/F), pre-dried, and weighed. The filter was dried for 24 h at 105◦ C, cooled in a desiccator, and weighed again. Measurement was computed as the average of a triplicate, and the experimental error was estimated as the average absolute deviation of the experimental values. Pigments were extracted with pure methanol, incubated for 45 min at 45◦ C, and centrifuged. The total chlorophyll (Chl-t) and carotenoid contents were determined according to Ritchie [40] from the mesurement of absorbances at 652 and 665 nm. 2.4.2. Zeta potential of cells (ζ ) Cell ζ potential was determined by measuring their electrophoretic mobility (Zetasizer Nano-ZS, Malvern, USA). Each data item was computed as the average of 30 readings on 1 mL samples. At neutral or basic pH, the C. vulgaris strain had a negative zeta potential, due to the presence of charged anionic carboxyl, phosphoric, phosphodiester, or hydroxyl functional groups on the cell surface.[41] The ζ value was measured in triplicate in both the Sueoka and mBB media at pH 7.5. It was found to be around -25 mV with an average absolute deviation of 11%, with no significant difference between the two media. The deviation indicated by the apparatus was, however, much higher: approximately 12 mV. This high value was due to the low

4

T.D.P. Nguyen et al.

Downloaded by [Aston University] at 23:45 16 January 2014

conductivity of the culture media. In addition, the nature of microalgal suspensions and the presence of impurities can also be the cause of low reliability and reproducibility.[42] The ζ potential time course during the autoflocculation experiment was plotted here with an estimated experimental error of 11%. 2.4.3. Ionic chromatography Cations were analysed with a cationic chromatograph (Dionex ICS 1100) equipped with a CG16 guard column (5 × 50 mm) and a CS16 separation column (5 × 250 mm). The cationic eluent was an H2 SO4 solution at 18 mM running at 1.25 mL min−1 . Anions were analysed with an anionic chromatograph (Dionex ICS 900) equipped with an AG9-HC guard column (4 × 50 mm) and an AS9-HC separation column (4 × 250 mm). The anionic eluent was a solution of Na2 CO3 at 7.2 mM and NaHCO3 at 1 mM running at 0.9 mL min−1 . Salts were detected by a conductivity meter, and the acquired data were processed with a dedicated software program (Chromeleon). Drifts in measurement accuracy were detected by daily injection of a solution containing all the tested ions at a concentration lying within the calibration range, and the device was recalibrated if the deviation was too large. For each ion, the deviation recorded the day of the analyses was taken as an estimate of the experimental uncertainty.

2.5.

Scanning electron microscopy and energy-dispersive X-ray spectroscopy

Cell aggregates from flocculation were examined by SEM used with EDX. Observations were made just after flocculation experiments or after 12 h for long experiments (autoflocculation). Cell aggregates were filtered on a 0.45 μ m glass-fibre filter, and if necessary pre-treated with gold deposition at the sample surface. Measurements were made in hard vacuum mode (10−4 Pa).

3. Results and discussion 3.1. Estimation of minimal Mg2+ and Ca2+ concentrations for flocculation in Sueoka (NH+ 4) and mBB (NO− 3 ) media The minimal concentrations of calcium or magnesium ions, [Ca2+ ]min and [Mg2+ ]min , required to flocculate cells were estimated by sharply increasing the pH to 11.8 by adding a small volume of 1 N NaOH to a harvest sample. Minimal Ca2+ concentrations required to induce the precipitation of calcium phosphate were determined by keeping the phosphate concentration at its initial value in the culture media (69.8 mg DM L−1 ). Figure 2 reports the values of [Ca2+ ]min and [Mg2+ ]min in both media as a function of the cell concentration in the suspension, ci . [Mg2+ ]min increased linearly with ci , with a similar slope for both media, [Mg2+ ]min

being higher by ca. 16 mg L−1 in the nitrate-based medium whatever the cell concentration (Figure 2(A)). By contrast, [Ca2+ ]min varied non-linearly with ci (Figure 2(B)). [Ca2+ ]min increased more slowly in the nitrate than in the ammonium medium, and [Ca2+ ]min was the same for the two media at ci ≈ 0.4 g DM L−1 . [Mg2+ ]min and [Ca2+ ]min values agreed with those found, respectively, by Smith and Davis [25] and Vandamme et al., [26] and by Sukenik and Shelef.[20] The differences in behaviour in the two media may be interpreted as a sign that the mechanisms involved in the inorganic salt precipitation are different. As reported by Stumm and O’Melia,[43] there must be a linear relationship between the coagulant level and the concentration of algal cells when charge neutralization is the predominant destabilization mechanism in the flocculation processes. Such a linear relationship can no longer apply if the precipitate generates large aggregates by bridging with the cells.[24] The volume of the flocculated zone (Vf ), the flocculation efficiency (E), and the cell concentration in the aggregate zone (Cest ) are reported for the two media in Table 2. Significant differences in Cest values and in floc densities were observed. Flocs were much denser in the ammonium (Sueoka) than in the nitrate medium (mBB) for the precipitation of magnesium compounds, the estimated concentrations of cells in the aggregate zone being Cest = 10.3 ± 1.6 and 3.2 ± 0.1 g DM L−1 , respectively, for the Sueoka and the mBB media. The opposite was found for the precipitation of calcium compounds, with Cest = 7.2 ± 0.8 for the mBB medium and 5.5 ± 0.4 g DM L−1 for the Sueoka medium. These observations confirm that the medium composition has a strong impact on flocculation quality,[44] even when ionic strength, pH, and zeta potential are similar. 3.2. Autoflocculation tests In this part, cell suspensions at 0.4 g DM L−1 were left under a light intensity of 500 μmol m−2 s−1 with air bubbling, but with no CO2 input, to let the pH rise slowly by photosynthesis and the stripping of dissolved carbon dioxide. Both [Ca2+ ] and [Mg2+ ] in the media were below the minimal values estimated by NaOH precipitation. Cells were thus not expected to precipitate. Flocculation tests were, however, carried out first without supplementing the media with [Ca2+ ] or [Mg2+ ] in order to test the capacity of the pH to increase ‘naturally’ (i.e. with bubbling and illumination, but with no nutrient input) in the media. The minimal concentrations of [Ca2+ ] and [Mg2+ ] inducing cell flocculation and decantation were then determined. The various autoflocculation tests performed are summarized in Table 3. 3.2.1. Sueoka’s medium (NH+ 4) The capacity of the pH to increase ‘naturally’ (i.e. with bubbling and illumination, but with no nutrient input) in the Sueoka medium was tested (test No. 1 in Table 3).

Downloaded by [Aston University] at 23:45 16 January 2014

Environmental Technology

5

Figure 2. Minimal amounts of divalent cations required for cell flocculation by NaOH addition as a function of the initial cell concentration Ci for the two culture media: (A) magnesium ion and (B) calcium ion ([Mg2+ ]0 and [Ca2+ ]0 are the initial concentrations in Mg2+ and Ca2+ in both culture media). Table 2. Flocculation   efficiency (E) and cell concentration in the aggregate zone (Cest ) determined at the minimal requirements of 3− 2+ 2+ Mg or Ca ( PO4 = 69.8 mg L−1 , concentration of cells = 0.41 g DM L−1 , sharp increase in pH up to 11.8 by 1 N NaOH addition). SUEOKA medium (NH+ 4)

Vf (mL) E (%) Cest (g DM L−1 )

BB medium (NO− 3)

[Mg2+ ]min = 24.2 mg L−1 (1 mM)

[Ca2+ ]min = 34 mg L−1 (0.85 mM)

[Mg2+ ]min = 44.9 mg L−1 (1.85 mM)

[Ca2+ ]min = 34 mg L−1 (0.85 mM)

3.5 ± 0.5 84 10.3 ± 1.6

6 ± 0.5 74 5.5 ± 0.4

13 ± 0.5 87 3.2 ± 0.1

5 ± 0.5 83 7.2 ± 0.8

The pH increased moderately from 7 to 9, and the zeta potential from −30 to −23 mV during the first five hours (Figure 3), but no cell aggregation was observed, nor any mineral precipitation. The pH then fell again from 9 to

8.25 between 5 and 7 h, while the zeta potential of cells increased more strongly. The limited increase in pH (pH increased up to 11 in the mBB medium, see Section 3.2.2 and Figure 4(A)) can be attributed to the nitrogen source:

6

T.D.P. Nguyen et al. Table 3. Experimental design L = 159 μ mol m−2 s−1 ).

of

autoflocculation

investigation

(C.

vulgaris

ci = 0.40 g DM L−1 ;

Concentration (mg L−1 ) Section 3.2.1 3.2.2 3.2.3 3.2.4

Test no.

Mediuma

pH increase

Mg2+

Ca2+

PO3− 4

Flocculation occurrence

1 2 3 4 5 6 7 8

Sueoka Sueoka + Ca2+ id. mBB mBB + Ca2+ mBB ++ Ca2+ mBB + Mg2+ mBB ++ Mg2+

Natural id. Chemical Natural id. id. id. id.

13.8 id. id. 13.8 id. id. 44.9 1000

6.8 61 id. 6.8 34 120 6.8 id.

70 id. id. 69.8 id. id. id. id.

– – Yes – – Yes – Yes

Downloaded by [Aston University] at 23:45 16 January 2014

a ‘+Ca2+ /Mg2+ ’ means that medium was supplemented up to the minimal concentration of Ca2+ or Mg2+ required for flocculation by NaOH addition, and ‘++ Ca2+ /Mg2+ ’ that the medium was doped well above the minimal concentration required for flocculation by NaOH addition; id. (idem) means that the value is the same as on the previous one.

Figure 3. Natural pH increase in Sueoka (NH+ 4 -based) medium with no CO2 injection: time course of pH and zeta potential; no flocculation observed.

NH+ 4 assimilation by the microalgae comes with the release of hydrogen instead of hydroxyl ions for negative nutrients (carbonates, phosphates, etc.),[45] leading to a lower rate of pH increase by photosynthesis in the ammonium-based Sueoka medium than in the nitrate-based mBB medium. On the other hand, the decrease in pH after 5 h reveals that a process consumes hydroxyl ions and competes with the photosynthesis and the CO2 stripping that tend to raise the pH. A possible cause is the stripping of NH3 and the resulting shift in the NH+ 4 /NH3 equilibrium. Other phenomena must, however, be considered to explain the increase in zeta potential after 5 h, such as the deposition at the cell surface of algogenic organic matter (AOM), which would hide the surface charges: NH3 is known to have inhibitory effects [46] that can increase the release of AOM. As the pH did not increase beyond 9, the precipitation of magnesium compounds could not occur, but that of calcium phosphates was possible. Thus a second test was carried out

with a Ca2+ addition up to ca. 2 times the minimal value estimated by NaOH addition (that is 61 mg L−1 , 1.53 mM), but the pH did not increase beyond 8, and did not allow the aggregation of cells (test No. 2, results not shown). On the contrary, flocculation was observed when a sharp increase in pH was obtained by adding 1 N NaOH in the same medium (test No. 3). This confirms that the natural increase in pH in the Sueoka medium was too low to precipitate calcium phosphates and flocculate C. vulgaris cells. These results suggest that the autoflocculation of C. vulgaris cannot be obtained in an ammonium-based culture medium with low salinity. 3.2.2. mBB medium (NO− 3) The capacity of the pH to increase ‘naturally’ was then tested in the mBB medium (test No. 4 and Figure 4(A)). The pH increased more than in the Sueoka medium (up to 10.4

Downloaded by [Aston University] at 23:45 16 January 2014

Environmental Technology

7

Figure 4. Natural increase in pH in mBB (NO− 3 -based) medium with no CO2 injection: time course of pH and zeta potential (A) and ion concentrations (B) as a function of time; no flocculation observed.

at t = 5 h instead of 9 for BMM). However, no cell settling or even turbidity due to mineral precipitation was observed visually (maybe because of the high turbidity already due to the cells), as expected because of the low concentration of Ca2+ in the medium. The time course of the zeta potential during the natural increase in pH differed noticeably from that observed in the Sueoka medium: variations were slighter, and the initial absolute value was lower in the mBB medium. There was no clear evidence of any possible cause of these differences, but it can be hypothesized that some biological mechanisms may modify the chemistry of the cell surface or lead to the release of algogenic matter masking the surface charges. The slight decrease followed by a slight increase in ζ could point to competitive effects, such as the dissociation of negatively charged groups, and the precipitation of positive compounds at the cell surface

(Mg2+ for instance). However, the time course of the zeta potential of the cell surfaces should be interpreted with caution because of the high experimental deviation. The decrease in Ca2+ and PO3− in the first 5 h 4 (Figure 4(B)) indicates that some moderate mineral precipitation occurs, maybe at the cell surface. Simulation with the VMinteq software [47] predicts that the solubility limits of hydroxyapatite (HAP, Ca5 (PO4 )3 OH) and amorphous tricalcium diphosphate Ca3 (PO4 )2 are reached when pH rises in the mBB medium, and these compounds may precipitate. As shown in Figure 4(A), the pH increases naturally up to nearly 10.8, suggesting that greater mineral precipitation leading to the flocculation of cells can be obtained, provided that the medium is doped with Ca2+ or Mg2+ . Tests of this type are described in the following sections.

Downloaded by [Aston University] at 23:45 16 January 2014

8

T.D.P. Nguyen et al.

2+ 3.2.3. mBB medium (NO− addition 3 ) with Ca The natural drift in pH was followed in the mBB medium supplemented with [Ca2+ ] up to the minimal value estimated by NaOH addition, [Ca2+ ]min = 34 mg L−1 (0.85 mM, see Section 2.1 and Figure 2), but no flocculation occurred (test No. 5). [Ca2+ ] had to be increased up to 120 mg L−1 (3 mM, test No. 6) to obtain a rapid settling of cells. The flocculation was then efficient, with a removal rate E of 92%, and the cell concentration in the precipitate was rather high, with a Cest value of 19 ± 4 g L−1 . The minimal concentration of Ca2+ for flocculation was thus roughly 3.5 times higher than that estimated for NaOH flocculation, and the cell concentration in the precipitate was also higher, with a Cest value of 19 g DM L−1 instead of 7 g DM L−1 for chemical flocculation. However, the large difference is not necessarily significant, because the pH is forced in chemical flocculation up to 11.8, which is probably much above the optimal NaOH level for the minimization of the volume of aggregates generated. Even so, the results highlight the difference between chemical flocculation and autoflocculation. Cell flocculation starts instantaneously at pH 7.8. This is in line with the results obtained by Sukenik and Shelef,[20] who reported a pH of precipitation of calcium compounds 2+ between 7.2  and  8 for [Ca ] = 2.5 mM (instead of 3 3− here) and PO4 = 0.74 mM (the same as here). Ion chromatography measurement confirmed that PO3− 4 disappeared instantly after Ca2+ injection, and SEM–EDX on cell aggregates showed a deposit composed mainly of Ca and P both at the cell surface and between cells (Figure 5).

estimated by NaOH addition (44.9 mg L−1 , 1.85 mM, test No. 7). The pH increased up to 10.8 in 8 h and remained stable for 16 h, but no flocculation occurred as for Ca2+ addition. The concentration in Mg2+ had to be increased to as high as 1000 mg L−1 (41.15 mM), a value close to the Mg2+ concentration in seawater, to obtain autoflocculation (test No. 8). Flocculation started after 11 h, just before the pH levelled off at 10.4 after 14 h (Figure 6(A)) due to photosynthesis stopping and the precipitation of magnesium hydroxide. The rate of pH increase was lower than in the native mBB medium (Figure 4(A)). A possible explanation is that the high concentration of magnesium sulphate added here (nearly 10 g L−1 ) limited the metabolism of C. vulgaris, a fresh water strain. Alyabyev et al. [48] reported such a limitation of C. vulgaris in a culture medium with a high concentration of NaCl. Zeta potential remained around 10 mV with non-significant changes, and its time course is therefore not reported in Figure 6(A). Values of ζ were higher than in the mBB medium, with no magnesium addition, which can be explained by the stronger adsorption of Mg2+ ions or other positively charged magnesium compounds at the cell surface. It can be noted, however, that ζ remained negative even when there was flocculation. A possible explanation is that charge neutralization does not

2+ 3.2.4. mBB medium (NO− addition 3 ) with Mg Attempts at autoflocculation by the precipitation of magnesium compounds were performed in the same way as for calcium compounds. First, the mBB medium was supplemented with Mg2+ up to the minimal concentration

Figure 5. Cell aggregates obtained by flocculating C. vulgaris in the mBB medium doped with Ca2+ (SEM–EDX).

Figure 6. Natural increase in pH in mBB (NO− 3 -based) medium enhanced with Mg2+ with no CO2 injection: time course of pH and zeta potential (A) and ion concentrations (B) as a function of time.

Downloaded by [Aston University] at 23:45 16 January 2014

Environmental Technology play a major role in flocculation, because coagulation does not intervene, and flocculation is mainly due to a sweeping mechanism. Another explanation was given by Cacace et al. [49] for the flocculation of montmorillonite particles: the presence of divalent cations such as Ca2+ has a very marked effect on the surface hydrophobicity, and coagulation may occur due to hydrophobic attraction between cells, even when their zeta potential is still sufficient to ensure electrostatic repulsion. As already observed for autoflocculation with calcium compounds, the settling efficiency was close to 90%, and the aggregate zone was more compact than for the precipitation with sodium hydroxide, with an estimated concentration of cells [Cest ] ≈ 33 g DM L−1 against 3 g DM L−1 for NaOH precipitation (Table 2). The difference is higher than for calcium compounds, possibly because solutions rich in Mg2+ tend to become more gelatinous when large quantities of NaOH are added.[28] Contrary to the observations made for calcium- and phosphate-induced flocculation, there was no specific mineral structure visible between cells, but instead a fine, regular deposit at the cell surface (Figure 7(B)). This suggests that weighting effects are not involved in cell precipitation, which would thus be more likely due to sweeping by magnesium compounds, probably magnesium hydroxide. On the other hand, large deposits clearer than flocculated zones B are visible in Figure 7(A), for instance in zone C. The deposits are partially inserted into flocculated zones. EDX analysis performed on zone C indicates that the weight percentages of Mg and P in the deposit were, respectively, 26% and 22%, which are values very close to the theoretical percentages of these elements in tribasic magnesium phosphate Mg3 (PO4 )2 (respectively, 28% for Mg and 24% for P). It seems, therefore, that magnesium-induced flocs contain an appreciable amount of magnesium phosphate-based deposit. Precipitation of magnesium phosphate was corroborated by ionic chromatographic measurements: the low concentration of calcium ions in the medium (6.8 mg L−1 ) can explain only a limited part of the total disappearance of phosphate ions (ca. 63 mg L−1 ). In addition, the decrease in the magnesium ion concentration began before phosphate exhaustion and continued thereafter. This evolution can be explained by the successive precipitation of phosphate magnesium compounds and magnesium hydroxide. Simulation with the Vminteq software confirms that tribasic magnesium phosphate can precipitate at low pH (≈ 8) for the concentrations in magnesium and phosphate present in the medium (respectively, 1000 and 69.8 g L−1 ). However, whether or not magnesium phosphate plays a role in the flocculation of microalgae cells cannot be inferred from these observations. The minimal concentration of magnesium required for autoflocculation was more than 20 times higher than that obtained for a sharp rise in pH by NaOH addition (1000 against 45 mg L−1 ), and the ratio was much larger than for the minimal concentration of calcium. This again suggests

9

Figure 7. Cell aggregates obtained by flocculating C. vulgaris in mBB medium doped with Mg2+ (SEM–EDX; B: enlargement of the framed area on snapshot A; C: zone with a deposit rich in magnesium and phosphate).

that base addition cannot always be considered as a surrogate technique of autoflocculation as is currently claimed in the literature.[20,25,26] It is known that solution chemistry can have a marked influence on flocculation efficiency. Thus Cao et al. [50] report that hydroxyapatite precipitation is inhibited by the presence of several compounds, in particular Mg2+ , and to a lesser extent SO2− 4 and silica. Some anions (sulphate, bicarbonate, and chloride) may also have considerable effects on the coagulation by trivalent aluminium salts.[44] As calcium and magnesium also have hydrolyzing properties, similar effects for the flocculation of microalgae in culture media cannot be excluded. These facts suggest that kinetic aspects play an important role in flocculation, together with the large difference in the rates of pH increase between autoflocculation and chemical flocculation. Another possible cause of the difference in chemical flocculation and autoflocculation is the release during the natural increase in pH of AOM, which is known to be able to decrease the flocculation efficiencies considerably.[25,27,51]

Downloaded by [Aston University] at 23:45 16 January 2014

10

T.D.P. Nguyen et al.

4. Conclusions Autoflocculation of C. vulgaris was found to be possible in an mBB culture medium with nitrate as nitrogen source, but not in a Sueoka medium with ammonium in place of nitrate. However, the medium made up on the basis of cellular growth must be supplemented with either calcium or magnesium to allow the precipitation of inorganic compounds, and the flocculation of cells. The minimal concentrations of Ca or Mg required for autoflocculation were found to be much higher than those estimated by a sharp NaOH-induced flocculation. Therefore, it is suspected that base addition cannot be considered as a surrogate technique as it is currently claimed in the literature. In Bold’s basal (BB) culture medium enhanced with magnesium, EXD and ion chromatography analyses indicate that magnesium phosphate compounds are included in cell flocculates. Finally, our results confirm that autoflocculation by slow pH increase with no CO2 supply can be envisaged for freshwater strains only in media are supplemented with calcium or magnesium. Acknowledgements The authors thank Guillaume Roelens (assistant engineer, GEPEA, Université de Nantes) and Jérémie Castello (PhD research engineer, TIMR 4297, Université de Technologie de Compiègne) for their contribution to scanning electron microscopy.

Funding This work was supported by the French National Research Agency and is part of the ‘DIESALG’ project [ANR-12BIME-000] on biodiesel fuel production from microalgae (http://www.biosolis.org/). The Vietnamese government is also acknowledged for funding the PhD of T.D.P. Nguyen.

References [1] Olaizola M. Commercial development of microalgal biotechnology: from test tube to the marketplace. Biomol Eng. 2003;20:459–466. [2] Chisti Y. Biodiesel from microalgae. Biotechnol Adv. 2007;25:294–306. [3] Brennan L, Owende P. Biofuels from microalgae – a review of technologies for production, processing, and extractions of biofuels and co-products. Renew Sust Energ Rev. 2010;14:557–577. [4] Greenwell HC, Laurens LML, Shields RJ, Lovitt RW, Flynn KJ. Placing microalgae on the biofuel priority list: a review of the technological challenges. J R Soc Interface. 2010;7:703– 726. [5] Mata TM, Martins AA, Caetano NS. Microalgae for biodiesel production and other applications: a review. Renew Sust Ener Rev. 2010;14:217–232. [6] Gudin C, Therpenier C. Bioconversion of solar energy into organic chemicals by microalgae. Adv Biotechnol Process. 1986;6:73–110. [7] Molina Grima E, Belarbi E-H, Acién Fernández FG, Robles Medina A, Chisti Y. Recovery of microalgal biomass and metabolites: process options and economics. Biotechnol Adv. 2003;20:491–515.

[8] Uduman N, Qi Y, Danquah MK, Forde GM, Hoadley A. Dewatering of microalgal cultures: a major bottleneck to algal-based fuels. J Renew Sust Ener. 2010;2, 012701: 1–15. [9] Golueke CG, Oswald WJ. Harvesting and processing sewage grown planktonic algae. J Water Pollut Control Fed. 1965;37(4):471–498. [10] Divakaran R, Pillai VNS. Flocculation of algae using chitosan. J Appl Phycol. 2002;14:419–422. [11] Vandamme D, Foubert I, Meesschaert B, Muylaert K. Flocculation of microalgae using cationic starch. J Appl Phycol. 2010;22:525–530. [12] Tenney MW, Verhoff FH. Chemical and autoflocculation of microorganisms in biological wastewater treatment. Biotechnol Bioeng. 1973;XV:1045–1073. [13] Lee AK, Lewis DM, Ashman PJ. Microbial flocculation, a potentially low-cost, harvesting technique for marine microalgae for the production of biodiesel. J Appl Phycol. 2009;21:559–567. [14] Duraiarasan S, Mani V. Influence of bioflocculation parameters on harvesting Chlorella salina and its optimization using response surface methodology. J Environ Chem Eng. 2013;1(4):1051–1056. [15] Dassey AJ, Theegala CS. Reducing electrocoagulation harvesting costs for practical microalgal biodiesel production. Environ Technol. 2013. doi:10.1080/09593330.2013. 842602 [16] Poelman E, De Pauw N, Jeurissen B. Potential of electrolytic flocculation for recovery of micro-algae. Resour Conserv Recy. 1997;19:1–10. [17] Phoochinda W, White DA, Briscoe BJ. An algal removal using a combination of flocculation and flotation processes. Environ Technol. 2004;24:1385–1395. [18] Dupré C, Guary JC, Grizeau D. Culture of an autoflocculant microalga in a vertical tubular photobioreactor for phycoerythrin production. Biotechnol Tech. 1995;9:185–190. [19] Spilling K, Seppälä J, Tamminen T. Inducing autoflocculation in the diatom Phaeodactylum tricornutum through CO2 regulation. J Appl Phycol. 2011;23(6):959–966. [20] Sukenik A, Shelef G. Algal autoflocculation – verification and proposed mechanisms. Biotechnol Bioeng. 1984;XXVI:142–147. [21] Leentvar J, Rebhun ME. Effects of magnesium and calcium precipitation on coagulation-flocculation with lime. Water Res. 1982;16:655–662. [22] Ayoub GM, Lee S-I, Koopman B. Seawater induced algal flocculation. Water Res. 1986;20:1265–1271. [23] Semerjian L, Ayoub GM. High-pH–magnesium coagulation–flocculation in wastewater treatment. Adv Environ Res. 2003;7:389–403. [24] Schlesinger A, Eisenstadt D, Bar-Gil A, Carmely H, Einbinder S, Gressel J. Inexpensive non-toxic flocculation of microalgae contradicts theories; overcoming a major hurdle to bulk algal production. Biotechnol Adv. 2012;30: 1023–1030. [25] Smith BT, Davis RH. Sedimentation of algae flocculated using naturally-available, magnesium-based flocculants. Algal Res. 2012;1:32–39. [26] Vandamme D, Foubert I, Fraeye I, Meesschaert B, Muylaert K. Flocculation of Chlorella vulgaris induced by high pH: role of magnesium and calcium and practical implications. Bioresour Technol. 2012;105: 114–119. [27] Wu Z, Zhu Y, Huang W, Zhang C, Li T, Zhang Y, Li A. Evaluation of flocculation induced by Ph increase for harvesting microalgae and reuse of flocculated medium. Bioresour Technol. 2012;110:496–502.

Downloaded by [Aston University] at 23:45 16 January 2014

Environmental Technology [28] Besson A, Guiraud P. High-pH-induced flocculation– flotation of the hypersaline microalga Dunaliella salina. Bioresour Technol. 2013;147:464–470. [29] Shi X-M, Chen F, Yuan J-P, Chan H. Heterotrophic production of lutein by selected Chlorella strains. J Appl Phycol. 1997;9:445–450. [30] Pruvost J, Van Vooren G, Le Gouic B, Couzinet-Mossion A, Legrand J. Systematic investigation of biomass and lipid productivity by microalgae in photobioreactors for biodiesel application. Bioresour Technol. 2010;102:150–158. [31] Beyerinck MW. Culturversuche mit Zoochlorellen, Lichenengonidien und anderen niederen Algen. Botanische Zeitung. 1890;47:724–785. [32] Reisner GS, Gering RK, Thompson JF. The metabolism of nitrate and ammonia by Chlorella. Plant Physiol. 1960;35:48–52. [33] Schuler JF, Diller VM, Kersten HJ. Preferential assimilation of ammonium ion by Chlorella vulgaris. Plant Physiol. 1952;28:299–303. [34] Sueoka N. Mitotic replication of deoxyribonucleic acid in Chlamydomonas reinhardi. Proc Nat Acad Sci. 1960;46: 83–91. [35] Harris EH. The Chlamydomonas source book. Introduction to Chlamydomonas and its laboratory use. New York, NY: Academic Press Inc.; 2009. [36] Bischoff HW, Bold HC. Some soil algae from Enchanted Rock and related algal species. Phycol Stud. 1963; 4:1–95. [37] Hadj-Romdhane F, Jaouen P, Pruvost J, Van Vooren G, Bourseau P. Development and validation of a minimal growth medium for recycling Chlorella vulgaris culture. Bioresour Technol. 2012;123:366–374. [38] Hadj-Romdhane F, Zheng X, Jaouen P, Pruvost J, Grizeau D, Croue JP, Bourseau P. The culture of Chlorella vulgaris in a recycled supernatant: effects on biomass production and medium quality. Bioresour Technol. 2013;132:285–292. [39] Hutner SH, Provasoli L, Schatz A, Haskins CP. Some approaches to the study of the role of metals in the

[40] [41] [42]

[43] [44] [45] [46] [47] [48]

[49] [50] [51]

11

metabolism of microorganisms. Proc Am Philosophical Soc 1950;94:152–170. Ritchie R. Consistent sets of spectrophotometric chlorophyll equations for acetone, methanol and ethanol solvents. Photosynth Res. 2006;89:27–41. Hadjoudja S, Deluchat V, Baudu M. Cell surface characterisation of Microcystis aeruginosa and Chlorella vulgaris. J Colloid Interf Sci. 2010;342:293–299. Baya DT, Effebi KR, Tangou TT, Keffala C, Vasel JL. Effect of hydroxyapatite, octacalcium phosphate and calcium phosphate on the auto-flocculation of the microalgae in a high-rate algal pond. Environ Technol. 2013;4(16):2407–2414. Stumm W, O’Melia CR. Stoichiometry of coagulation. J Am Water Works Assoc. 1968;60(1968):514–539. Duan J, Gregory J. Coagulation by hydrolysing metal salts. Adv Colloid Interfac Sci. 2003;100–102:475–502. Goldman JC, Brewer PG. Effect of nitrogen source and growth rate on phytoplankton-mediated changes in alkalinity. Limnol Oceanogr. 1980;25:352–357. Abeliovich A, Azov Y. Toxicity of ammonia to algae in Sewage oxidation ponds. Appl Environ Microb. 1976;31(6): 801–806. Gustafsson JP. Visual MINTEQ ver. 3.0 website; 2011. Alyabyev AJ, Loseva NL, Gordon LK, Andreyeva IN, Rachimova GG, Tribunskih VI, Ponomareva AA, Kemp RB. The effect of changes in salinity on the energy yielding processes of Chlorella vulgaris and Dunaliella maritima cells. Thermochim Acta. 2007;458:65–70. Cacace MG, Landau EM, Ramsden JJ. The Hofmeister series: salt and solvent effects on interfacial phenomena. Q Rev Biophys. 1997;3:241–277. Cao X, Harris WG, Josan MS, Nair VD. Inhibition of calcium phosphate precipitation under environmentally-relevant conditions. Sci Total Environ. 2007;383:205–215. Vandamme D, Foubert I, Muylaert K. Flocculation as a low-cost method for harvesting microalgae for bulk biomass production. Trends Biotechnol. 2013;31:233–239.