You are on page 1of 37

Accepted Manuscript

Comparison of flocculation methods for harvesting Dunaliella

Kristin Pirwitz, Liisa Rihko-Struckmann, Kai Sundmacher

PII: S0960-8524(15)00985-2
DOI: http://dx.doi.org/10.1016/j.biortech.2015.07.032
Reference: BITE 15265

To appear in: Bioresource Technology

Please cite this article as: Pirwitz, K., Rihko-Struckmann, L., Sundmacher, K., Comparison of flocculation methods
for harvesting Dunaliella, Bioresource Technology (2015), doi: http://dx.doi.org/10.1016/j.biortech.2015.07.032

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.
Comparison of flocculation methods for harvesting
Dunaliella
Kristin Pirwitza , Liisa Rihko-Struckmanna,∗, Kai Sundmachera,b
a
Max Planck Institute for Dynamics of Complex Technical Systems, Process Systems
Engineering, Sandtorstr. 1, 39106 Magdeburg, Germany.
b
Otto von Guericke University Magdeburg, Process Systems Engineering,
Universitätsplatz 2, 39106 Magdeburg, Germany.

Abstract

Low cell concentrations of Dunaliella salina in production scale cultivations


require high energy input for biomass harvesting. Flocculation is a potential
preconcentration method to lower the dewatering costs for the β-carotene
production. In the present study, optimal flocculant dosages were deter-
mined for several metal salts, NaOH, Ca(OH)2 and Al-electrolysis. Beside
harvesting efficiency ηH and concentration factor CF , also the recyclability
of the separated medium as well as the influence of the cell physiology on
the harvesting performance were analyzed for selected flocculants. To as-
sess the possible recycle of non-sedimented cells for the inoculation of new
cultivations, cell vitality and the photosynthetic activity of D. salina were
analyzed after the flocculation. As a result, the flocculation with NaOH led
to a clear inhibition of both, the algal growth on recycled medium and the
algal photosynthetic activity. The addition of FeCl3 seems most promising
to flocculate D. salina.


Corresponding author. E-mail address: rihko@mpi-magdeburg.mpg.de, Tel. +49 391
6110 318, Fax: +49 391 6110 353.

Preprint submitted to Bioresource Technology July 10, 2015


Keywords: Dunaliella, Harvesting methods, Flocculation

1 1. Introduction

2 The use of microorganism to produce valuable products represents an


3 attractive alternative to chemical production processes. Hereby, microalgae
4 offer the possibility to biochemically convert the natural resources CO2 , wa-
5 ter and sunlight into high-valuable products. Furthermore, microalgae can
6 be cultivated in outdoor plants without competing with arable cropland.
7 Dunaliella salina is an industrial exploited representative of potential mi-
8 croalgal production organisms. The halotolerant microalga is one of the
9 most important producer of natural β-carotene for the nutrition and cos-
10 metic sector. However, under production conditions D. salina reaches rela-
11 tively low cell concentrations which cause high production cost and energy
12 demand for biomass dewatering. Mostly, centrifugation is the harvesting
13 method of choice for microalgal production processes, since it works highly
14 effective (Pittman et al., 2011). Also in the case of D. salina, centrifuga-
15 tion is successfully applied for biomass dewatering in large scale operation
16 (Ben-Amotz, 2008). However, centrifugation is extremely energy consuming
17 (Molina Grima et al., 2003). A possibility to lower the energy demand is
18 the preconcentration of biomass by flocculation prior to centrifugation (Van-
19 damme et al., 2013). Flocculation is a process to agglomerate single algal
20 particles into bigger flocs, which leads to an increased sedimentation veloc-
21 ity. This effect can be approximately described by the Stokes’s law (Liss
22 et al., 2004). Without the influence of a flocculant, coulomb repulsion pre-
23 vents the single algae to agglomerate. In natural environments single algal

2
24 particles are stable in suspension due to their negative charge caused by the
25 presence of deprotonated functional groups on the algal like phosphates, car-
26 boxylates and hydroxides (Brady et al., 2014). The destabilization of the
27 algae suspension can be induced by the addition of flocculants as a result
28 of neutralization, bridging or sweeping flocculation. With that the coulomb
29 repulsion of the particles decreases and the flocculation of microalgal cells is
30 facilitated. The use of metal salts leads to charge neutralization of the nega-
31 tively charged microalgae surfaces by multivalent metal cations like Al+3 or
32 Fe+3 (Wan et al., 2014). Cationic polymers like chitosan or modified starch
33 are able to bind especially freshwater algae by the mechanism of bridging
34 (Vandamme et al., 2013). Furthermore, microalgal biomass can be concen-
35 trated by sweeping flocculation as a consequence of the precipitation of Mg+2
36 ions in the medium after pH increase (Besson and Guiraud, 2013; Vandamme
37 et al., 2012). Recently, the former method was successfully investigated in
38 combination with air flotation for D. salina by Besson and Guiraud (2013).
39 Physical flocculation can be induced by electrolysis, ultrasound or magnetic
40 forces (Wan et al., 2014). In the case of electrolysis, the solved metal ions
41 from the sacrificial electrode react with water to positively charged metal
42 hydroxide species which can adhere the negatively charged algae surfaces.
43 The resulting charge neutralization facilitates the agglomeration into flocs
44 (Vandamme et al., 2011). Furthermore, also sweeping flocculation occurs, if
45 insoluble aluminium hydroxide precipitates were generated during electroly-
46 sis (Gao et al., 2010). Thus, a contamination of the biomass with metal ions
47 is unavoidable, which also holds true for the previously mentioned types of
48 chemical flocculation. This can be problematic for the downstream process-

3
49 ing of the biomass as well as for the quality of the product.
50 Although there are numerous studies on microalgal flocculation, less informa-
51 tion is available about the flocculation behavior of D. salina. Furthermore,
52 most flocculation studies focus only on the harvesting efficiency or the dewa-
53 tering properties of the investigated methods. The present work aims for the
54 experimental comparison of several flocculation methods for the model or-
55 ganism D. salina under consideration of various assessment criteria. Initially,
56 potential flocculation methods and their optimal flocculant doses are iden-
57 tified, namely flocculation by the addition of different aluminium and ferric
58 salts, by pH increase with either NaOH or Ca(OH)2 or by electrolysis with
59 aluminium electrodes. After this, the influence of the harvesting method on
60 cell vitality and photosynthetic activity, the reusability of culture medium
61 after flocculation as well as the biomass contamination with the flocculant
62 is investigated. Furthermore, we also compare the flocculation performance
63 of nitrogen deprived, stressed cells and cells grown under control condition.
64 Finally, a reliable evaluation of the different flocculation types is presented.

65 2. Materials and Methods

66 2.1. Strain and cultivation conditions

67 D. salina (CCAP 19/18, delivered 2011 by the Culture Collection of Algae


68 and Protozoa, Scotland) was grown photoautotrophically in a rotary shaking
69 incubator (Multitron, Infors AG, Switzerland) at 3.5 % CO2 in air, a temper-
70 ature of 26 ◦ C and a shaking frequency of 100 rpm. Therefore, 2 L shaking
71 flasks were filled with 1 L of the growth medium (pH 7.5) previously described
72 by Lamers et al. (2010). The medium was prepared without NaFeEDTA,

4
73 Na2 EDTA and HEPES to avoid inhibitory effects of these medium ingredi-
74 ents on flocculation performance. Since iron is a crucial trace element for
75 algal growth, 5 µM FeCl3 was added instead into the medium. To simulate a
76 day/night cycle, 16 h light at 75 µE m−2 s−1 and 8 h darkness were applied
77 every day. Carotenogenesis of D. salina was introduced by applying abiotic
78 stress in form of continuous high light conditions (1500 µE m−2 s−1 ) and
79 nitrogen depletion. For this purpose 1 L nitrate free culture medium was
80 inoculated to reach approximately 1×106 cells mL−1 with a culture grown
81 under normal conditions as described above. This cultivation was carried out
82 in a flat panel photobioreactor (FMT 150, PSI, Czech Republic) at a pH of
83 7.5, aerated with 400 mL min−1 of a synthetic air mixture containing 3.5 %
84 CO2 .
85 For the determination of algal growth and the harvesting efficiency ηH the
86 optical density of the samples was measured at 735 nm using an UV/Vis spec-
87 trophotometer (Specord S600, Analytik Jena AG, Germany). Cell concen-
88 trations were determined by a cell counter (Cellometer Auto T4, Nexcelom
89 Bioscience LLC., USA).

90 2.2. Flocculation of D. salina

91 Unless otherwise noted, all harvesting experiments were carried out un-
92 der lab scale conditions in duplicates of beaker glass vessels or electrolysis
93 chambers and 30 min sedimentation. Therefore, D. salina cultures in mid-
94 logarithmic growth phase were adjusted to a pH of 7.5 and a culture density
95 of approximately OD735nm = 1 prior to flocculation. This optical density cor-
96 responds to approximately 0.3 gdw L−1 biomass which is comparable with
97 the harvesting concentration of D. salina in production scale (Curtain and

5
98 Snook, 1983). The flocculation procedure was induced by different floccu-
99 lants, namely the addition of metal ions via Al2 (SO4 )3 × 16H2 O (alum),
100 AlCl3 , FeCl3 × 6H2 O, FeSO4 × 7H2 O, Fe2 (SO4 )3 × 6H2 O (0.1 M stock so-
101 lutions) or by aluminium electrode electrolysis as well as the pH increase via
102 NaOH (5 M stock solution) or Ca(OH)2 (0.5 M stock solution). To identify
103 the optimal flocculant doses the concentrations of the chemicals were varied
104 between 0-1.2 mM for the metal salts and between 0-25 mM for Ca(OH)2 and
105 0-50 mM for NaOH. Therefore, beaker glass vessels were filled with 100 mL
106 culture suspension. The respective flocculants were added to the cultures
107 under mixing at 250 rpm. After the flocculant addition, the cultures were
108 mixed for 10 min at 250 rpm and thereafter allowed to settle down for an-
109 other 30 min. Samples were taken from the middle of the culture height to
110 determine the harvesting efficiency ηh according to following equation:
OD735nm (t0 ) − OD735nm (tend )
ηh = · 100 (1)
OD735nm (t0 )
111 with OD735nm (t0 ) as the absorbance of the algal suspension at 735 nm before
112 and OD735nm (tend ) after 30 min of flocculant addition. The concentration
113 factor CF was determined as follows:
H1
CF = (2)
H2
114 where H1 and H2 are the optically determined algal suspension height at the
115 beginning and the end of the harvesting experiment, respectively.
116 For the electrolysis experiment the flocculant dose was varied by apply-
117 ing different electrolysis times from 5-20 min or by applying current densities
118 between 3.4-17 mA cm−2 via a power supplier (peqPOWER 250, Peqlab, Ger-
119 many). Therefore, glass chambers (5.5×10.8×12 cm3 ) were used, equipped

6
120 with two Al electrodes with an electrode distance of 4 cm, respectively. The
121 glass chambers were filled with 300 mL of the culture prepared as described
122 above. In the experiments an active area of 44 cm2 sacrificial electrode was
123 applied to the culture under continuous mixing at 250 rpm. After electroly-
124 sis the electrodes were removed and the culture was mixed a further 10 min.
125 Samples were taken as described above to determine ηh and CF according
126 to Eqs. 1-2. The energy consumption Eh and the mass of aluminium mAl
127 dissolved during the electrolysis were calculated according to Faraday’s law.

128 2.3. Measurements of cell vitality

129 2.3.1. Pulse amplitude modulation fluorometry


130 The photochemical activity of photosystem II (PSII) of D. salina was de-
131 termined with pulse amplitude modulation (PAM) fluorometry (Dual-PAM-
132 100, Walz, Germany) in order to investigate the cell stress induced by the
133 harvesting procedure. For this purpose, 1.5 mL culture samples adjusted to
134 5x106 cells mL−1 were placed in a glass cuvette with 1 cm path length. Floc-
135 culated cells were decollated via resuspension in the medium. The samples
136 were dark adapted for 5 min, at 26 ◦ C and mixed at 150 rpm. After the
137 dark adaptation, the initial (F0 ) and maximum (Fmax ) fluorescence levels
138 were measured with a measuring radiation of 460 nm at 5 µE m−2 s−1 PAR,
139 an actinic radiation of 635 nm at 166 µE m−2 s−1 and a saturating actinic
140 excitation pulse of 635 nm at 2000 µE m−2 s−1 PAR for 0.5 s. The Fv /Fmax
141 value (with Fv = Fmax -F0 ) was calculated to quantify the maximum quantum
142 yield of PSII (biochemical efficiency of PSII).

7
143 2.3.2. Flow cytometry
144 Cell vitality was analyzed by flow cytometry applying the dyer fluores-
145 cein diacetate (FDA) (Hejazi et al., 2004). Metabolically active cells convert
146 the non-fluorescent FDA into green fluorescent fluorescein with the help of
147 esterase enzymes. For the cell staining 20 µL of a FDA stock solution (2 mg
148 mL−1 in aceton) was added to 1 mL of the cell suspension, mixed and in-
149 cubated in the dark for 5 min at room temperature. The analysis of the
150 stained cells was done by a flow cytometer (Cyflow Space, Partec, Germany)
151 with a blue argon solid state excitation laser (488 nm). To get reliable re-
152 sults at least 300000 particles were analyzed using a flow rate of 1 µL s−1 and
153 1.5 M NaCl solution as sheath fluid. The background signals were excluded
154 in the analysis by considering only the particles with a red chlorophyll fluo-
155 rescence emission signal. Data were acquired with FloMax software (Version
156 2.70, Partec, Germany). As negative control 1 mL of D. salina culture was
157 heated for 5 min at 70 ◦ C, stained and measured as described above. All
158 results were normalized to the mean received signal strength of the control
159 (non-flocculated) cells.

160 2.4. Reuse of separated culture medium

161 The separated supernatant of the harvested cells was collected to investi-
162 gate the recyclability of cultivation medium after the harvesting procedure.
163 The supernatant was readjusted to a pH of 7.5. MgCl2 was added to one part
164 the samples harvested via pH increase to regain the typical Mg concentration
165 in the medium. In order to remove contamination from the medium, the re-
166 cycled supernatants were filtered (0.22 mM) prior to the second cultivation.
167 To test on reusability of the separated culture media, duplicates of 200 mL

8
168 shaking flasks were filled with 100 mL medium, respectively. Fresh medium
169 was used for the control cultivation. All flasks were inoculated with D. salina
170 cells to reach an initial optical density of approximately OD735nm =0.1 and
171 cultivated under normal conditions as described above.

172 2.5. Elemental analysis of the separated culture medium

173 The contamination of the supernatant and the biomass with Al, Fe, Mg, P
174 or Ca was estimated by analyzing the separated medium after the harvesting
175 procedure with inductively coupled plasma optical emission spectrometry
176 (ICP-OES, Currenta, Germany). The remaining flocculant load mF in the
177 biomass was estimated as follows:

mF (t0 ) − mF (tend )
mF = · 100 (3)
mdw

178 with mF (t0 ) as the mass of the receptive flocculant added prior to floccu-
179 lation, mF (tend ) as the mass of the respective flocculant measured in the
180 supernatant of the flocculated algae suspension and mdw as dry weight of the
181 harvested biomass.

182 3. Results and discussion

183 3.1. Flocculation by the addition of metal salts

184 Cationic metal ions lead to charge neutralization at the negatively charged
185 algae surface. With increasing pH or increasing flocculant concentration also
186 sweeping flocculation occurs due to the decreasing solubility of metal hydrox-
187 ides (Duan and Gregory, 2003). Fig. 1 illustrates ηh and CF for D. salina cells
188 reached by the addition of Fe2 (SO4 )3 , FeCl3 , FeSO4 , Al2 (SO4 )3 and AlCl3 .

9
189 In all cases, except FeSO4 , flocculant concentrations of 0.5 mM resulted in
190 a ηh of approximately 80 % which further increased to 90-97 % by applying
191 flocculant concentration up to 1 mM (see Fig. 1a). Comparable tendencies
192 were observed for Chlorella minutissima and Scenedesmus sp. using the sim-
193 ilar cation concentration Fe3+ and Al3+ (Papazi et al., 2010; Chen et al.,
194 2013). In contrast to the work of Papazi et al. (2010), all trivalent metal ions
195 in form of Fe2 (SO4 )3 , FeCl3 , Al2 (SO4 )3 and AlCl3 used in the present study,
196 reached comparable results regarding ηh , irrespective of whether chloride or
197 sulfate were used as anion. The addition of divalent iron ions in form of
198 FeSO4 was not applicable for D. salina since it did not cause flocculation for
199 all applied concentration (see Fig. 1a). By considering the CF values reached
200 with the different flocculants (Fig. 1b), it became obvious that the addition
201 of aluminum salts was most effective to get a high dewatering effect of the
202 biomass. In the case of ferric salts, FeCl3 caused the best values regarding
203 CF . For that reason, the addition of 0.3 mM Al2 (SO4 )3 and 1 mM FeCl3
204 were chosen as candidates for subsequent analyzes (see Sections 3.4-3.7).

205 3.2. Flocculation via electrolysis

206 In principle, electrolysis is more applicable for marine microalgae than


207 for fresh water species (Vandamme et al., 2013) since highly saline medium
208 has a high electric conductivity and thereby reduces the energy demand for
209 electrolysis (Uduman et al., 2011). Thus, flocculation of D. salina by elec-
210 trolysis seems to be promising as the used culture medium contains 1.5 M
211 NaCl leading to a high conductivity. Another effect to potential promote
212 electro-flocculation is the reduced thickness of the diffuse part of the electric
213 double layer on the particle surface due to the high ionic strength of the sur-

10
214 rounding medium (Duan and Gregory, 2003). Fig. 2 shows the flocculation
215 efficiencies ηH and concentration factors CF for D. salina cells. Current den-
216 sities of 3.4-17 mA cm−2 were applied with aluminium electrodes for 5, 10,
217 15 and 20 min, respectively. The higher the flocculant load, the higher was
218 ηH (Fig. 2a). The current density of 17 mA cm−2 resulted in the best algae
219 recovery when applying electrolysis for 5 or 10 min. ηH values of more than
220 95 % were reached (see Fig. 2a), being comparable to literature data (Gao
221 et al., 2010; Uduman et al., 2011). However, the current densities needed
222 to flocculated D. salina are higher compared to the 0.6-3 mA cm−2 used for
223 the marine microalga Phaeodactylum tricornutum (Vandamme et al., 2011).
224 Other studies dealing with electrolysis of Dunaliella or other marine microal-
225 gae applied comparable (Zenouzi et al., 2013) or even 2-3 fold higher current
226 densities (Uduman et al., 2011) to reach similar values for ηH .
227 The best dewatering effect was achieved using 3.4 mA cm−2 for 5 min. How-
228 ever, an increase of the flocculant concentration directly caused a decrease
229 in the CF values (Fig. 2b). It is reported in literature that the optimal pH
230 for electrolysis is below pH 7 (Gao et al., 2010; Vandamme et al., 2011),
231 where charge neutralization by monomeric or cationic hydrolyzed forms of
232 Al ions efficiently occurs. At higher pH, the solubility of aluminium hydrox-
233 ides decreases (Duan and Gregory, 2003), leading to its precipitation and the
234 embedding of algal particle. This mechanism seems to be less efficient for
235 electrolysis (Gao et al., 2010; Vandamme et al., 2011). The Dunaliella sus-
236 pension used in the present work has a pH of 7.5, similar to the production
237 pond. Here, sweeping flocculation is the predominant mechanism. This was
238 visible by the formation of white insoluble aluminium hydroxide and by the

11
239 increase of the pH value during electrolysis (see Table 1). High concentra-
240 tion of insoluble aluminium hydroxide precipitates in the algal suspension
241 increase ηH (Vandamme et al., 2011), but enlarge the volume of flocculated
242 algae sludge as well, which is observable as a decreased CF value.
243 With respect to the economic feasibility of the harvesting procedure, we de-
244 cided to use an electrolysis time of 5 min with a current density of 3.4 mA cm−2
245 in the following studies discussed in Sections 3.4-3.7. With this parametric
246 setup the lowest flocculant load and energy demand per kg harvested biomass
247 were required (see Table 1) going along with the best harvesting performance
248
dw the theoretical flocculant de-
in terms of the CF value. With 84 gAl kg−1
249
dw needed for
mand of the electrolysis is higher compared to the 68.5 gAl kg−1
250 the flocculation with 0.3 mM Al2 (SO4 )3 .

251 3.3. Flocculation by pH increase

252 Fig. 3 illustrates the harvesting efficiency ηH , the concentration factor CF


253 and the pH values of a D. salina suspension after flocculation by NaOH or
254 Ca(OH)2 . In both cases, increased base concentration resulted as expected in
255 an increase of the pH and thereby in an increase of ηH . The highest ηH val-
256 ues were reached with a pH above 12. Similar results were recently reported
257 by Chen et al. (2013) and Garcı́a-Pérez et al. (2014). In the latter work
258 Chlorella vulgaris was flocculated most efficient by adjusting the pH to 12.
259 However, ηH is not only a function of the pH but also of the cell concentra-
260 tion as well as the concentration of Mg in the culture medium (Garcı́a-Pérez
261 et al., 2014; Wu et al., 2012). In contrary, the work of Besson and Guiraud
262 (2013) illustrated that pH is less important than base dose for optimal NaOH
263 flocculation-flotation of D. salina using similar NaOH concentration to the

12
264 present work.
265 The mechanism behind high pH flocculation can be considered as indirect,
266 since the added base is not the flocculant itself. In fact, the pH increase
267 destabilize the algae suspension by sweeping flocculation due to Mg or Ca
268 precipitation or by charge neutralization (Vandamme et al., 2013). Regard-
269 ing the ηH values, the addition of Ca(OH)2 seems preferable compared to
270 that of NaOH. 10 mM of Ca(OH)2 reached a ηH value of nearly 98 % for D.
271 salina, whereas the highest applied NaOH concentration (50 mM) reached
272 only a maximum efficiency of 82 % after 30 min settling (see Fig. 3a, b). Sim-
273 ilar findings were reported in literature for various other microalgal species
274 (Castrillo et al., 2013; Schlesinger et al., 2012). Accordingly, Ca(OH)2 bears
275 a higher flocculation strength than NaOH. Prepared from limestone as raw
276 material, Ca(OH)2 represents an attractive harvesting agent as it is inexpen-
277 sive and widely available. Furthermore, the biomass flocculated by Ca(OH)2
278 is accepted to be used for animal feed (Schlesinger et al., 2012).
279 However, by comparing the CF values of both approaches, it becomes evi-
280 dent that the dewatering effect of NaOH addition clearly exceeded that of
281 Ca(OH)2 (Fig. 3a, b). The flocs generated by NaOH addition are morpho-
282 logically more dense than that of Ca(OH)2 (see Supplement Fig. 1g, h),
283 resulting in a lower algae sludge volume. High pH flocculation obviously
284 tend to decreased CF values for pH values above 11 (Besson and Guiraud,
285 2013; Horiuchi et al., 2003; Garcı́a-Pérez et al., 2014). The higher the pH, the
286 higher is the amount of precipitated Mg(OH)2 in the biomass. Brady et al.
287 (2014) calculated for pH above 10.5 an exponential increase of Mg(OH)2 in
288 the medium accompanied with an decrease of functional algal surface groups

13
289 bound via Mg+ sorption. Furthermore, co-precipitation of other medium in-
290 gredients is possible. If the pH rises above 11, the solved Ca in the medium
291 can precipitates into insoluble CaCO3 (Şirin et al., 2013), which increases the
292 algal sludge volume as well. Since the addition of 20 mM NaOH resulted in
293 an extraordinary CF value of 50, this concentration was used for the more
294 detailed investigation later on this work.

295 3.4. Reuse of the separated culture medium

296 Recycling of nutrients and water from medium after the dewatering step
297 would be cost efficient and beneficial for biomass production and product
298 yields (Farooq et al., 2015). Therefore, this issue was investigated in de-
299 tail in the present work. Fig. 4 depicts the growth of D. salina on recycled
300 medium after centrifugation or flocculation induced by 0.3 mM Al2 (SO4 )3 ,
301 1 mM FeCl3 , 3.4 mA cm−2 electrolysis for 5 min and 20 mM NaOH. Fresh
302 medium was used for control cultivation. D. salina was able to grow in
303 the recycled medium after harvesting with Al2 (SO4 )3 , FeCl3 or electrolysis.
304 Here, the alga reached a growth rate comparable to that of fresh medium
305 or culture medium separated by centrifugation (see Fig. 4a, b). This ob-
306 servation was expected, since the separated media contained only traces of
307 the possibly growth inhibitory flocculants Fe or Al (see Table 2). Mg is an
308 essential trace element for marine microalgae growth (Ho et al., 2003). It
309 is, therefore, not astonishing that the medium separated from the floccula-
310 tion via NaOH was not applicable for recycling without readjusting the Mg
311 concentration (Fig. 4a). Induced by NaOH flocculation almost all Mg ions
312 precipitated in the medium (see Table 2). However, the used medium con-
313 tains a relatively low Mg concentration (7.3 mg L−1 ) compared to natural

14
314 salt water. Therefore, it is not expectable that the addition of NaOH would
315 cause Mg limitation in natural sea water. For that reason, the concentration
316 of Mg in the separated medium was readjusted to its typical level in a further
317 experiment. When cultivating D. salina in this readjusted medium, growth
318 was detected, but strongly inhibited in comparison to the control culture
319 (Fig. 4b). These observations arose the presumption, that there are further
320 effects along the precipitation of Mg and Ca influencing algal growth on the
321 recycled medium. Triggered by the high pH values above 12 (see Fig. 3a), it
322 is conceivable that co-precipitation of further medium ingredients occurred.
323 Castrillo et al. (2013) demonstrated that by readjusting of NO−
3 and PO4
−3

324 ion to the original level, algal biomass yields were comparable for cultures
325 grown on recycled medium after NaOH flocculation and control cultures. In
326 addition, the work of Sukenik and Shelef (1984) indicated that high pH floc-
327 culation also induces orthophosphate and calcium phosphate precipitation.
328 However, the analysis of the initial and residual Ca and P concentration in
329 the here used medium after NaOH addition indicated no significant decreases
330 in the concentration (see Table 2). Nevertheless, the reusability of NaOH for
331 D. salina cultivation is possible only to a limited extent.

332 3.5. Biomass contamination

333 The Mg (P, Ca), Fe and Al concentrations of the separated culture


334 medium after flocculation D. salina were analyzed to indirectly estimate
335 the biomass contamination with the respective flocculant. In all cases the
336 flocculant was concentrated in the biomass nearly completely (see Table 2).
337
dw , the highest flocculant concentration was found in the
With 205.84 mgF e g−1
338 biomass harvested via FeCl3 . The lowest contamination of the biomass with

15
339
dw resulted from the flocculation with NaOH. In that case, an
26.96 mgM g g−1
340 additional contamination with co-precipitates of P or Ca is negligible (see
341 Sections 3.3). The residual flocculants in the concentrated biomass can in-
342 terfere with subsequent downstream processing steps (Uduman et al., 2010)
343 and possibly lower the product quality. However, remaining flocculants in
344 the biomass can also promote the downstream processing. Rwehumbiza et al.
345 (2012) illustrated for polyaluminium flocculated Nannochloropsis salina cells,
346 that the concentration of flocculant decreases with every downstream pro-
347 cessing step and do not affect the final product composition. Furthermore, it
348 is possible to remove the flocculant by integrating it as a reactant in down-
349 stream processes like extraction or transesterification (Seo et al., 2015), by
350 pH decrease of the biomass (Vandamme et al., 2014; Farooq et al., 2015)
351 or by conducting washing steps of the product (Vandamme et al., 2014; Fa-
352 rooq et al., 2015). Thus, the biomass contamination caused by flocculation
353 is not necessarily a disadvantage. For Dunaliella biomass it was recently
354 shown, that a pH increase of the carotenogenic culture sludge leads to higher
355 concentration of β-carotene in the dried algae powder (Ben-Amotz, 2014).
356 Therefore, the application of high pH flocculation could even be favorable in
357 the β-carotene production process.

358 3.6. Photosynthetic activity and cell vitality after flocculation

359 Viable unharvested cells in the separated medium could be used for the
360 inoculation of a new cultivation process. In order to investigate the abiotic
361 stress level induced by the flocculation procedure, concentrated D. salina cells
362 were analyzed via PAM fluorometry and vitality staining with flow cytome-
363 try (see Table 3). In all cases, except the NaOH flocculation, the harvested

16
364 cells showed photosynthetic activity with a comparable extent to the control
365 culture. Accordingly, no loss of photochemical cell productivity occurred.
366 However, at high pH over 12 induced by NaOH, a complete stagnation of
367 photosynthetic activity in D. salina occurred. This pH value is far away
368 from the pH optimum of the alga and obviously caused cell stress. Neverthe-
369 less, the cell vitality remained at a level of 64 % indicating that cells could
370 be reactivated by readjusting the pH to the normal level. Compared to the
371 control cells, the vitality of D. salina was 30-40 % reduced after flocculation
372 with Al2 (SO4 )3 or FeCL3 , respectively. Electrolysis flocculation caused in the
373 highest rate of cell lysis by losing 56 % of the cell vitality. The large precipi-
374 tated aluminium hydroxide particle might physically damage the fragile algae
375 surface. Nevertheless, in all cases the residual cell activity was higher than
376 44 %. Consequently, non-sedimented cells can be recycled within the culture
377 medium to inoculate new cultivations which is favorable for commercial algal
378 biomass production (Besson and Guiraud, 2013; Horiuchi et al., 2003).

379 3.7. Influence of cell physiology on flocculation performance

380 Finally, we carried out a comparison of the flocculation performance for


381 normal and carotenogenic algae cells (Fig. 5). If D. salina is used for β-
382 carotene production, abiotic stress in form of high light intensity and ni-
383 trogen depletion have to be induced. Under this condition the harvesting
384 behavior of D. salina might be different from that reached under normal,
385 unstressed growth condition. The first differences in harvesting behavior was
386 observed for Al2 (SO4 )3 flocculation. The flocculant concentration had to be
387 increased from 0.3 mM to 0.6 mM in order to induce flocculation of the
388 stressed cells. With regard to the flocculation efficiency ηH it is visible that

17
389 the cell physiology mainly impacts the mechanism of charge neutralization
390 caused by the addition of metal salts (Fig. 5a). Here, a higher concentration
391 of flocculant is necessary to flocculate stressed cells with the same efficiency
392 than the unstressed cells. When using sweeping flocculation caused by NaOH
393 or electrolysis, no differences in the ηH values were observed. More remark-
394 able differences were detected for the CF values of stressed and unstressed
395 D. salina cells (Fig. 5a). In detail, the FeCl3 and electrolysis led to higher
396 concentration factors for the stressed D. salina culture, whereas the floccu-
397 lation with NaOH or Al2 (SO4 )3 was more appropriate for unstressed cells.
398 In consequence, the physiology of D. salina is crucial for the identification of
399 the optimal preconcentration strategy for D. salina. For cells under normal
400 and carotenogenic conditions, NaOH flocculation performed best.

401 3.8. Final assessment of the different flocculation methods

402 To reliable compare the applied flocculation methods for D. salina, a


403 summary of the present examinations is given in Table 4. Regarding the
404 flocculation performance, indicated by ηH and CF , flocculation of D. salina
405 with NaOH seems most promising for normal and carotenogenic cells. Fur-
406 thermore, high pH values of the biomass induced by NaOH can be ben-
407 eficial for the product yield (Ben-Amotz, 2014). However, the reusability
408 of separated medium for cultivation is limited. Flocculation via electroly-
409 sis turned out as most inappropriate method in the present study, since it
410 showed a weak harvesting performance as well as a limited possibility for
411 downstream integration. Similar conclusion was taken for the flocculation
412 with Al2 (SO4 )3 . In spite of a promising harvesting performance, the con-
413 tamination of the biomass with flocculant can not be completely removed

18
414 in subsequent downstream steps (Rwehumbiza et al., 2012). Under the con-
415 sideration of all applied assessment criteria, flocculation of D. salina with
416 FeCl3 seems most favorable. Besides good results in terms of the ηH and
417 CF values, also the recyclability of the medium and the integration of the
418 flocculant into the downstream processing is possible.

419 4. Conclusion

420 In the present work, flocculation of D. salina induced by metal salts, elec-
421 trolysis and pH increase were compared based on several assessment criteria.
422 Generally, the physiology of the alga strongly influences the applicability of
423 the flocculation method. Expect NaOH, the investigated flocculation tech-
424 niques preserved the photosynthesis activity of the alga. All methods led to
425 biomass contamination and reduced vitality of D. salina. Regarding harvest-
426 ing results, NaOH flocculation performed best for normal and carotenogenic
427 biomass. However, this method has weakness in terms of medium recycle.
428 Under consideration of all criteria, FeCl3 flocculation turned out as the most
429 promising one to harvest D. salina.

430 Acknowledgments

431 This research work was partly supported by the Center for Dynamic Sys-
432 tems (CDS) funded by the Federal State Saxony-Anhalt (Germany).The au-
433 thors would like to thank S. Nickel, M. Ikert and A. Reichelt for technical
434 assistance in performing the experiments.

19
435 Ben-Amotz, A., 2008. Bio-fuel and CO2 capture by marine microalgae.

436 Ben-Amotz, A., 2014. Method for producing β-carotene rich Dunaliella pow-
437 der, US8722057 B25.

438 Besson, A., Guiraud, P., 2013. High-pH-induced flocculation – flotation of the
439 hypersaline microalga Dunaliella salina. Bioresour Technol 147, 464–470.

440 Brady, P.V., Pohl, P.I., Hewson, J.C., 2014. A coordination chemistry model
441 of algal autoflocculation. Algal Res 5, 226–230.

442 Castrillo, M., Lucas-Salas, L.M., Rodrı́guez-Gil, C., Martı́nez, D., 2013. High
443 pH-induced flocculation-sedimentation and effect of supernatant reuse on
444 growth rate and lipid productivity of Scenedesmus obliquus and Chlorella
445 vulgaris. Bioresour Technol 128, 324–329.

446 Chen, L., Wang, C., Wang, W., Wei, J., 2013. Optimal conditions of different
447 flocculation methods for harvesting Scenedesmus sp. cultivated in an open-
448 pond system. Bioresour Technol 133, 9–15.

449 Şirin, S., Clavero, E., Salvadò, J., 2013. Potential pre-concentration methods
450 for Nannochloropsis gaditana and a comparative study of pre-concentrated
451 sample properties. Bioresour Technol 132, 293–304.

452 Curtain, C.C., Snook, H., 1983. Method for harvesting algae, US6/511135.

453 Duan, J., Gregory, J., 2003. Coagulation by hydrolysing metal salts. Adv
454 Colloid Interfac 100-102, 475–502.

455 Farooq, W., Moon, M., Ryu, B.G., Suh, W.I., Shrivastav, A., Park, M.S.,
456 Mishra, S.K., Yang, J.W., 2015. Effect of harvesting methods on the

20
457 reusability of water for cultivation of Chlorella vulgaris, its lipid produc-
458 tivity and biodiesel quality. Algal Res 8, 1–7.

459 Gao, S., Yang, J., Tian, J., Ma, F., Tu, G., Du, M., 2010. Electro-
460 coagulation-flotation process for algae removal. J Hazard Mater 177, 336–
461 343.

462 Garcı́a-Pérez, J.S., Beuckels, A., Vandamme, D., Depraetere, O., Foubert,
463 I., Parra, R., Muylaert, K., 2014. Influence of magnesium concentration,
464 biomass concentration and pH on flocculation of Chlorella vulgaris. Algal
465 Res 3, 24–29.

466 Hejazi, M.A., Holwerda, E., Wijffels, R.H., 2004. Milking microalga
467 Dunaliella salina for beta-carotene production in two-phase bioreactors.
468 Biotechnol Bioeng 85, 475–481.

469 Ho, T.Y., Quigg, A., Finkel, Z.V., Milligan, A.J., Wyman, K., Falkowski,
470 P.G., Morel, F.M.M., 2003. The elemental composition of phytoplankton.
471 J Phycol 39, 1145–1159.

472 Horiuchi, J.I., Ohba, I., Tada, K., Kobayashi, M., Kanno, T., Kishimoto,
473 M., 2003. Effective cell harvesting of the halotolerant microalga Dunaliella
474 tertiolecta with pH control. Biosci Bioeng 95, 412–415.

475 Lamers, P.P., van de Laak, C.C.W., Kaasenbrood, P.S., Lorier, J., Janssen,
476 M., De Vos, R.C.H., Bino, R.J., Wijffels, R.H., 2010. Carotenoid and fatty
477 acid metabolism in light-stressed Dunaliella salina. Biotechnol Bioeng 106,
478 638–648.

21
479 Liss, S.N., Milligan, T.G., Droppo, I.G., Leppard, G.G., 2004. Methods of
480 Analyzing Floc Properties. CRC Press.

481 Molina Grima, E., Belarbi, E.H., Acién Fernández, F.G., Robles Medina, A.,
482 Chisti, Y., 2003. Recovery of microalgal biomass and metabolites: process
483 options and economics. Biotechnol Adv 20, 491–515.

484 Papazi, A., Makridis, P., Divanach, P., 2010. Harvesting Chlorella minutis-
485 sima using cell coagulants. J Appl Phycol 22, 349–355.

486 Pittman, J.K., Dean, A.P., Osundeko, O., 2011. The potential of sustainable
487 algal biofuel production using wastewater resources. Bioresour Technol
488 102, 17–25.

489 Rwehumbiza, V.M., Harrison, R., Thomsen, L., 2012. Alum-induced floc-
490 culation of preconcentrated Nannochloropsis salina: Residual aluminium
491 in the biomass, fames and its effects on microalgae growth upon media
492 recycling. Chem Eng J 200, 168–175.

493 Schlesinger, A., Eisenstadt, D., Bar-Gil, A., Carmely, H., Einbinder, S., Gres-
494 sel, J., 2012. Inexpensive non-toxic flocculation of microalgae contradicts
495 theories; overcoming a major hurdle to bulk algal production. Biotechnol
496 Adv 30, 1023–1030.

497 Seo, Y.H., Sung, M., Kim, B., Oh, Y.K., Kim, D.Y., Han, J.I., 2015. Ferric
498 chloride based downstream process for microalgae based biodiesel produc-
499 tion. Bioresour Technol 181, 143–147.

500 Sukenik, A., Shelef, G., 1984. Algal autoflocculation-verification and pro-
501 posed mechanism. Biotechnol Bioeng 26, 142–147.

22
502 Uduman, N., Bourniquel, V., Danquah, M.K., Hoadley, A.F.A., 2011. A
503 parametric study of electrocoagulation as a recovery process of marine
504 microalgae for biodiesel production. Chem Eng J 174, 249–257.

505 Uduman, N., Qi, Y., Danquah, M.K., Forde, G.M., Hoadley, A., 2010. De-
506 watering of microalgal cultures: A major bottleneck to algae-based fuels.
507 J Renew Sustain Ener 2, 012701.

508 Vandamme, D., Beuckels, A., Markou, G., Foubert, I., Muylaert, K., 2014.
509 Reversible flocculation of microalgae using magnesium hydroxide. Bioenerg
510 Res , 1–10.

511 Vandamme, D., Foubert, I., Fraeye, I., Meesschaert, B., Muylaert, K., 2012.
512 Flocculation of Chlorella vulgaris induced by high pH: Role of magnesium
513 and calcium and practical implications. Bioresour Technol 105, 114–119.

514 Vandamme, D., Foubert, I., Muylaert, K., 2013. Flocculation as a low-cost
515 method for harvesting microalgae for bulk biomass production. Trends
516 Biotechnol 31, 233–239.

517 Vandamme, D., Pontes, S.C.V., Goiris, K., Foubert, I., Pinoy, L.J.J., Muy-
518 laert, K., 2011. Evaluation of electro-coagulation-flocculation for harvest-
519 ing marine and freshwater microalgae. Biotechnol Bioeng 108, 2320–2329.

520 Wan, C., Alam, M.A., Zhao, X.Q., Zhang, X.Y., Guo, S.L., Ho, S.H., Chang,
521 J.S., Bai, F.W., 2014. Current progress and future prospect of microalgal
522 biomass harvest using various flocculation technologies. Bioresour Technol
523 184, 251–257.

23
524 Wu, Z., Zhu, Y., Huang, W., Zhang, C., Li, T., Zhang, Y., Li, A., 2012.
525 Evaluation of flocculation induced by pH increase for harvesting microalgae
526 and reuse of flocculated medium. Bioresour Technol 110, 496–502.

527 Zenouzi, A., Ghobadian, B., Hejazi, M., Rahnemoon, P., 2013. Harvesting
528 of microalgae Dunaliella salina using electroflocculation. J Agr Sci Tech
529 15, 879–887.

24
530 Caption of figures

531 Figure 1: Flocculation a) efficiency ηH and b) concentration factor CF of


532 D. salina determined 30 min after addition of different concentration of alu-
533 minium and iron salts. Experiments were conducted in duplicates.
534

535 Figure 2: Flocculation a) efficiency ηH and b) concentration factor CF of


536 D. salina determined 30 min after electrolysis using aluminium electrodes.
537 Flocculant load was varied by applying current densities of 3.4 mA cm−2
538 (squares), 10.2 mA cm−2 (circles) and 17 mA cm−2 (triangles) for 5, 10, 15
539 and 20 min, respectively.
540

541 Figure 3: Flocculation efficiency ηH , concentration factor CF and pH of


542 a D. salina suspension determined 30 min after addition of a) NaOH or b)
543 Ca(OH)2 . Experiments were conducted in duplicates.
544

545 Figure 4: Growth of D. salina on separated medium after flocculation, in-


546 dicated as OD735nm . Cultivation was carried out in recycled medium after
547 a) centrifugation for 30 min and 500 × g (circles), electrolysis for 5 min at
548 3.4 mA cm−2 (triangles), the addition of 0.3 mM Al2 (SO4 )3 (alum, hexagons),
549 the addition of 20 mM NaOH (open, reversed triangles) or b) the addition of
550 20 mM NaOH with readjusted MgCl3 concentration (open, reversed triangles)
551 or the addition of 1 mM FeCl3 (stars). Control cultivation was conducted in
552 fresh medium (squares). The experiments were conducted in duplicates.
553

554 Figure 5: Flocculation a) efficiency ηH and b) concentration factor CF of

25
555 D. salina determined 120 min after electrolysis for 5 min at 3.4 mA cm−2 ,
556 the addition of 20 mM NaOH, the addition of 0.3 mM Al2 (SO4 )3 (alum) (or
557 0.6 mM Al2 (SO4 )3 for stressed cells) and the addition of 1 mM FeCl3 for un-
558 stressed and stressed cells (high light, nitrogen depletion). All experiments
559 were conducted in duplicates.

26
a)
100

80
FeCl3
60
ηH (%)

FeSO4
40 Fe2(SO4)3
AlCl3
20 Al2(SO4)3

0.0 0.4 0.8 1.2


Flocculant (mM)

b) 50

40

30
CF (-)

20

10

0
0.0 0.4 0.8 1.2
Flocculant (mM)

Figure 1

27
a)
100

90

80
ηH (%)

70

60
3.4 mA cm-2
10.2 mA cm-2
50
17 mA cm-2
40
5 10 15 20
t (min)

b)

50

40

30
CF (-)

20

10

0
5 10 15 20
t (min)

Figure 2

28
a)
100 ηH
CF
80 pH 12
ηH (%), CF
60
10

pH
40

20 8

0
0 10 20 30 40 50
NaOH (mM)

b)
100

80 12
ηH (%), CF

60

pH
10
40

20
8
0
0 5 10 15 20 25
Ca(OH)2 (mM)

Figure 3

29
a)
3.5 Control
Centrifuge
3.0
Electrolysis
2.5 NaOH
Alum
OD735nm (-)
2.0

1.5

1.0

0.5

0.0
0 50 100 150 200 250
t (h)

b)
2.5 Control
NaOH + MgCl3
2.0 FeCl3
OD735nm (-)

1.5

1.0

0.5

0.0
0 50 100 150 200 250 300
t (h)

Figure 4

30
a) Unstressed cells Stressed cells
100

80

60
ηH (%)

40

20

0
Settling Electrolysis NaOH Alum Ferric chloride

Procedure

b)
50
Concentration factor (-)

40

30

20

10

0
Settling Electrolysis NaOH Alum Ferric chloride

Procedure

Figure 5

31
Table 1: Final pH, calculated mass of solved aluminium mAl3+ and energy demand EH
of electrolysis in dependence of the applied electrolysis time using current densities of
3.4 mA cm−2 , 10.2 mA cm−2 and 17 mA cm−2 , respectively.

pHend (-) mAl3+ (g kg−1


dw ) E (kWh kg−1
dw )

mA cm−2 3.4 10.2 17 3.4 10.2 17 3.4 10.2 17


5 min 7.61 7.61 7.80 84.0 277.86 275.71 0.2 0.68 1.15
10 min 7.66 7.64 7.92 144.72 500.06 535.84 0.36 1.49 2.07
15 min 7.75 7.72 8.12 188.33 564.27 808.46 0.45 1.68 3.25
20 min 7.76 7.95 8.22 264.09 819.87 845.7 0.65 2.44 4.25

32
Table 2: Initial and residual flocculant in the culture medium and biomass contamination
after flocculation of D. salina via electrolysis for 5 min at 3.4 mA cm−2 , the addition of
20 mM NaOH, the addition of 0.3 mM Al2 (SO4 )3 or the addition of 1 mM FeCl3 .

Method Flocculant Initial Flocculant Residual Flocculant Contamination


(mg/L (%)) (mg/L (%a )) (mg/gdw )
Electrolysis Al 11 0.13 (1.18) 56.17
NaOH Mg 7.3 0.25 (3.42) 26.96
P 27 27 (100) –
Ca 0.72 0.65 (90.27) 0.27
Al2 (SO4 )3 Al 16.19 0.1 (0.62) 62.19
FeCl3 Fe 55.85 0.05 (0.27) 205.84
a
percentage of the initial applied flocculant in the separated culture medium after
flocculation

33
Table 3: Photosynthetic activity of photosystem II (PSII) and cells vitality of D. salina
cells flocculated via electrolysis for 5 min at 3.4 mA cm−2 , the addition of 20 mM NaOH,
the addition of 0.3 mM Al2 (SO4 )3 or the addition of 1 mM FeCl3 .

Method PSII activitya Cell vitalityb


(-) %
Control (positiv) 0.679 ± 0.003 100 ± 7.56
Control (negativ) - 0.44
Electrolysis 0.657 ± 0.003 44.30 ± 4.08
NaOH n.d. 64.31 ± 2.44
Al2 (SO4 )3 0.693 ± 0.007 70.56 ± 0.23
FeCl3 0.674 ± 0.005 60.30 ± 2.67
a
measured by PAM fluorometry
b
FDA cell staining detected by flow cytometry

34
Table 4: Comparison of the harvesting performance after flocculation of D. salina via
electrolysis for 5 min at 3.4 mA cm−2 , the addition of 20 mM NaOH, the addition of
0.3 mM Al2 (SO4 )3 or the addition of 1 mM FeCl3 . Assessment criteriaa are the harvesting
performances given by the flocculation efficiency ηH and the concentration factor CF, the
reusability of medium and non-sedimented cells as well as the flocculant load in the biomass
after flocculation and the possibility to integrate the flocculant in downstream processes.

Method Electrolysis NaOH Al2 (SO4 )3 FeCl3


ηH + ++ ++ ++
CF + ++ + +
Medium recycle ++ - ++ ++
Cell recycle - - + +
Biomass load - + - --
b
Downstream integration - ++c - b
++d
Rank 4 2 3 1
a
indicated as ++ excellent, ++ good, - insufficient, - - poor
b
flocculant concentration reduced during downstream processing (Rwehumbiza et al.,
2012)
c
flocculant integration in downstream processes (Ben-Amotz, 2014) or removal by pH
decrease (Vandamme et al., 2014)
d
flocculant integration in downstream processes (Seo et al., 2015) or removal by pH
decrease (Farooq et al., 2015)

35
,ŝŐŚůŝŐŚƚƐŽĨƚŚĞŵĂŶƵƐĐƌŝƉƚ͗

ͣŽŵƉĂƌŝƐŽŶŽĨĨůŽĐĐƵůĂƚŝŽŶŵĞƚŚŽĚƐĨŽƌŚĂƌǀĞƐƚŝŶŐƵŶĂůŝĞůůĂ͟

LJ

<ƌŝƐƚŝŶWŝƌǁŝƚnj͕>ŝŝƐĂZŝŚŬŽͲ^ƚƌƵĐŬŵĂŶŶĂŶĚ<Ăŝ^ƵŶĚŵĂĐŚĞƌ

• ŽŶƐŝĚĞƌŝŶŐĂůůĐƌŝƚĞƌŝĂĨůŽĐĐƵůĂƚŝŽŶŽĨƵŶĂůŝĞůůĂǀŝĂ&ĞůϯƉĞƌĨŽƌŵĞĚďĞƐƚ͘

• džƉĞĐƚEĂK,ĨůŽĐĐƵůĂƚŝŽŶ͕ŵĞĚŝƵŵƌĞĐLJĐůĞǁĂƐƉŽƐƐŝďůĞĨŽƌĂůůŝŶǀĞƐƚŝŐĂƚĞĚ

ŵĞƚŚŽĚƐ͘

• ůĞĐƚƌŽůLJƐŝƐĂŶĚEĂK,ƐƚƌŽŶŐůLJŝŵƉĂŝƌĞĚĐĞůůǀŝƚĂůŝƚLJĂŶĚƉŚŽƚŽƐLJŶƚŚĞƐŝƐ͕

ƌĞƐƉĞĐƚŝǀĞůLJ͘

• ĞůůƉŚLJƐŝŽůŽŐLJŽĨƵŶĂůŝĞůůĂŝŵƉĂĐƚĞĚƚŚĞŚĂƌǀĞƐƚŝŶŐƌĞƐƵůƚƐ͘


You might also like