Microalgal Downstream Processing: Harvesting, Drying, Extraction, Separation, and Purification 99
diatoms were floated to the surface of the water column by a rising steam of air bubbles. Harvesting
efficiencies exceeding 90% were achieved using optimized parameters regarding the bubble size between
0.1–0.5 mm outer diameter, air flow rate of 2.25 L/min, pH 7.5 and run time of 30 min.
This method of separation of algae can only be applied to algae which naturally secrete surfactants or
by amphiphilic additives. These additives will then have to be removed from the algal cake, thus making
this method less attractive.
Filtration
Filtration of microalgae cultures with filter presses operated under pressure or vacuum were shown
to produce a cake with 27% solids when filtering relatively large microalgae such as Coelastrum
proboscideum, but proved to be inefficient with smaller microalgae such as Scenedesmus and Dunaliella
(Mohn 1980). Harvesting microalgal biomass using membrane filtration technology has also been
investigated. Application of submerged microfiltration that applies lower pressure in absence of any
cross-flow velocity is commonly used in submerged membrane bioreactors (MBRs) for wastewater
treatment. Bilad et al. (2012) investigated the harvesting efficiency of such submerged filtrations fitted
with various types of membranes and using the improved flux-step method and batch up-concentration
filtrations. The membranes were homemade from 9, 12, and 15% w/w [Polyvinylidene fluoride (PVDF),
Mw ~ 534,000/N,N-Dimethylformamide (DMF)] to achieve various pore sizes and casted on a
polypropylene non-woven support. For all three polymer concentrations the membrane pore size was
far below the size of the diatom P. tricornutum. Almost complete algae retention was achieved for all
membranes, associated with lower degrees of fouling. Operational costs of such technology shows an
energy consumption of 0.91 kW h/m
3
and a final concentration of 22% w/v. This energy consumption is
reported as the lowest compared to other harvesting techniques.
Tangential flow filtration (TFF) was evaluated for concentrating T. suecica (Danquah et al. 2009a).
The unit consisted of a Millipore Pellicon cassette system fitted with a 0.22 mm Pellicon 2 filter.
The optimal transmembrane pressure was determined to be 2 bar corresponding to a flow rate of ca.
20 L/m
2
h. Under these optimal conditions, the final retentate concentrations were 4.66 and 8.88% w/v for
45 and 150 minutes run time, respectively. TFF can also be influenced by the microalgal growth phase as
demonstrated by Danquah et al. (2009b). Results from TFF (2 bar over 25 min) of low growth rate phase
and high growth rate phase of T. suecica algal broth showed that the final retentate concentrations were
5.57 and 2.47% w/v for an energy consumption of 0.38 and 0.51 kWh/m
3
, respectively.
Two ultrafiltration systems were compared on microalgal suspensions of C. fusiformis and
Skeletonema costatum (Frappart et al. 2011). The first was a cross flow filtration unit equipped with a flat
sheet membrane (Rayflow system) and the second was a rotating disk module placed close to the stationary
membrane. In both cases, the membrane was an IRIS 3038 membrane made of polyacrylonitrile. This
membrane is hydrophilic, with a molecular weight cut-off of 40,000 Da. This study demonstrated a
better performance of the rotating disk module which had fluxes twice higher than the Rayflow system.
Rossignol et al. (1999) compared the efficiency of cross-flow microfiltration and ultrafiltration using
polyvinylidilene difluoride (PVDF), polyethersulfone (PES), and polyacrylnitrile (PAN) membranes
applied to cultures of Haslea ostrearia and S. costatum. For long-term operation and hence reduced
running costs, ultrafiltration with PAN membrane turned out to be the most efficient technique when
operating at low pressure and low tangential velocity.
Harvesting microalgae was also performed by high gradient magnetic filtration (HGMF) (Cerff
et al. 2012). For that purpose, precipitated magnetite (Fe 3 O 4 ) and hydrophilic silica-coated magnetic
particles MagSilica 50–85, referred to as precipitated magnetite and MagSilica, respectively, were used in
combination with the microalgal cultures of P. tricornutum and N. salina. For P. tricornutum, MagSilica
was found to be adsorbed to the cell at pH 8, while at pH 12, the precipitated magnetite was incorporated
into large flocules of the algae, probably formed by secreted extracellular polysaccharides. For N. salina,
low adsorption rates of MagSilica and precipitated silica were observed, but was drastically increased
at pH 12, with enhanced flocculation to follow. Separation efficiencies of 90% were obtained for
diatoms were floated to the surface of the water column by a rising steam of air bubbles. Harvesting
efficiencies exceeding 90% were achieved using optimized parameters regarding the bubble size between
0.1–0.5 mm outer diameter, air flow rate of 2.25 L/min, pH 7.5 and run time of 30 min.
This method of separation of algae can only be applied to algae which naturally secrete surfactants or
by amphiphilic additives. These additives will then have to be removed from the algal cake, thus making
this method less attractive.
Filtration
Filtration of microalgae cultures with filter presses operated under pressure or vacuum were shown
to produce a cake with 27% solids when filtering relatively large microalgae such as Coelastrum
proboscideum, but proved to be inefficient with smaller microalgae such as Scenedesmus and Dunaliella
(Mohn 1980). Harvesting microalgal biomass using membrane filtration technology has also been
investigated. Application of submerged microfiltration that applies lower pressure in absence of any
cross-flow velocity is commonly used in submerged membrane bioreactors (MBRs) for wastewater
treatment. Bilad et al. (2012) investigated the harvesting efficiency of such submerged filtrations fitted
with various types of membranes and using the improved flux-step method and batch up-concentration
filtrations. The membranes were homemade from 9, 12, and 15% w/w [Polyvinylidene fluoride (PVDF),
Mw ~ 534,000/N,N-Dimethylformamide (DMF)] to achieve various pore sizes and casted on a
polypropylene non-woven support. For all three polymer concentrations the membrane pore size was
far below the size of the diatom P. tricornutum. Almost complete algae retention was achieved for all
membranes, associated with lower degrees of fouling. Operational costs of such technology shows an
energy consumption of 0.91 kW h/m
3
and a final concentration of 22% w/v. This energy consumption is
reported as the lowest compared to other harvesting techniques.
Tangential flow filtration (TFF) was evaluated for concentrating T. suecica (Danquah et al. 2009a).
The unit consisted of a Millipore Pellicon cassette system fitted with a 0.22 mm Pellicon 2 filter.
The optimal transmembrane pressure was determined to be 2 bar corresponding to a flow rate of ca.
20 L/m
2
h. Under these optimal conditions, the final retentate concentrations were 4.66 and 8.88% w/v for
45 and 150 minutes run time, respectively. TFF can also be influenced by the microalgal growth phase as
demonstrated by Danquah et al. (2009b). Results from TFF (2 bar over 25 min) of low growth rate phase
and high growth rate phase of T. suecica algal broth showed that the final retentate concentrations were
5.57 and 2.47% w/v for an energy consumption of 0.38 and 0.51 kWh/m
3
, respectively.
Two ultrafiltration systems were compared on microalgal suspensions of C. fusiformis and
Skeletonema costatum (Frappart et al. 2011). The first was a cross flow filtration unit equipped with a flat
sheet membrane (Rayflow system) and the second was a rotating disk module placed close to the stationary
membrane. In both cases, the membrane was an IRIS 3038 membrane made of polyacrylonitrile. This
membrane is hydrophilic, with a molecular weight cut-off of 40,000 Da. This study demonstrated a
better performance of the rotating disk module which had fluxes twice higher than the Rayflow system.
Rossignol et al. (1999) compared the efficiency of cross-flow microfiltration and ultrafiltration using
polyvinylidilene difluoride (PVDF), polyethersulfone (PES), and polyacrylnitrile (PAN) membranes
applied to cultures of Haslea ostrearia and S. costatum. For long-term operation and hence reduced
running costs, ultrafiltration with PAN membrane turned out to be the most efficient technique when
operating at low pressure and low tangential velocity.
Harvesting microalgae was also performed by high gradient magnetic filtration (HGMF) (Cerff
et al. 2012). For that purpose, precipitated magnetite (Fe 3 O 4 ) and hydrophilic silica-coated magnetic
particles MagSilica 50–85, referred to as precipitated magnetite and MagSilica, respectively, were used in
combination with the microalgal cultures of P. tricornutum and N. salina. For P. tricornutum, MagSilica
was found to be adsorbed to the cell at pH 8, while at pH 12, the precipitated magnetite was incorporated
into large flocules of the algae, probably formed by secreted extracellular polysaccharides. For N. salina,
low adsorption rates of MagSilica and precipitated silica were observed, but was drastically increased
at pH 12, with enhanced flocculation to follow. Separation efficiencies of 90% were obtained for