Resins that Reversibly Bind Algae for Harvesting and Concentration Jessica Jones, a Christine Beran, b James Beach, b and Martin Poenie b a Department of Biomedical Engineering, University of Texas at Austin, Austin, TX 78712 b Department of Molecular Cell and Developmental Biology, University of Texas at Austin, Austin, TX 78712; [email protected](for correspondence) Published online in Wiley Online Library (wileyonlinelibrary.com). DOI 10.1002/ep.11697 Algae have great potential to address a number of impor- tant needs such as water remediation and as a feedstock for biofuel and biochemicals. At present, harvesting or removing algae is expensive, especially in terms of the cost of making bio- fuel. The expense of harvesting algae arises from the cost of pumping and processing large amounts of water. In an effort to explore less expensive methods for harvesting, a series of weak anion exchange resins incorporating various weak bases with different pK a values were developed to bind and release algae as a function of pH. The best of these resins bind about 10% of their weight in algae, show 100% reversibility and reus- ability, and the ability to concentrate algae from dilute suspen- sion to 30 g L 21 . Furthermore, they have the potential for desalting algae grown in high salt media. These resins provide a low cost method for harvesting algae without pumping water or introducing large amounts of chemicals into the feedwater, harvested algae, or the environment. Ó 2012 American Institute of Chemical Engineers Environ Prog, 00: 000–000, 2012 Keywords: algae, biofuel, dewatering, harvest INTRODUCTION Algae have the potential to address current and future energy needs as feedstock for biofuels [1]. High oil content and biomass production rates [2], as well as higher photosyn- thetic efficiencies than terrestrial plants [3], make algae ideal for commercial-scale biofuel production. Algae consume car- bon dioxide which can reduce greenhouse gas emissions compared to petroleum and energy cost balances can be improved when using flue gas as a carbon source [3]. Fur- thermore, algae can be cultivated in brackish water unsuit- able for agricultural crops, and be used to remediate waste- water [4]. In addition to fuel, profitable coproducts can be generated, such as animal feed, biopolymers, and agricultural fertilizers [3]. Algae are also a sustainable source of high- value biochemicals [5]; for example, docosahexaenoic acid (DHA), an essential omega-3 fatty acid important to infant development and cardiovascular health [6], is synthesized and concentrated in algae [7]. One of the current barriers to using algae for biofuel on the commercial scale is the cost of growing and harvesting algae from dilute suspension. Most algae used for biofuel are small, 2–3 lm in diameter, and are difficult to separate from water by filtration or tank settling [8]. Furthermore, autotro- phic algal growth density is limited by light perfusion. While lab-scale studies report biomass density in the magnitude of grams per liter media [9], large-scale cultivation, such as in a raceway pond, often produce biomass concentrations of less than 0.5 g L 21 [1]. As a consequence, large volumes of water must be processed to generate the amount of biomass needed. For example, to recover just one kilogram of algal dry mass from an algal culture of 0.4 g L 21 density, at least 2500 liters (660 gallons) of water must be processed. Previ- ous studies have shown that biomass recovery contributes 20–30% of the total production costs [10]. Thus, the cost of water removal from algal biomass poses a significant obstacle to the expansion of algae-based bioenergy [8]. Current techniques used on the commercial production scale have proven inadequate, requiring considerable time or energy expenditure to harvest the algal biomass. Gravity sed- imentation is the most common separation method used in wastewater treatment, but new approaches must be devel- oped because of its characteristic long retention time and low removal rate [11]. Microfiltration is a physical separation technique that depends on the membrane pore size and flow characteristics [12], and thus may not be suitable for high throughput processing. Mechanical separation methods such as centrifugation produce the highest biomass yields, but also have high equipment and energy costs [10, 13]. Air flota- tion, in which hydraulic equipment generates small bubbles that adhere to algae and force them to float to the surface, can similarly incur high energy costs to aerate a large water volume [14]. Flocculation, the induction of algal cell aggregation, can be achieved by either charge neutralization or addition of charged polymers that coordinate the charged algal surfaces. Although operating cost can be relatively low depending on the flocculant, it involves continual addition of chemicals to large volumes of water. Some of these treatments introduce unwanted salts or metal ions whereas others may have long residence times and, in any case, can affect the output water [3]. Experimental harvesting techniques, such as ultrasound- induced aggregation, have been effectively demonstrated at the laboratory scale, but have been projected to be more ex- pensive than current centrifugation techniques [15]. Problems with the current methods have inspired the design, develop, and test of reusable resins for concentrating algae out of suspension. The strategy takes advantage of the negative surface charge on algae, which bind to weak anion exchange resins as a function of pH. While simple in princi- ple, the performance of commercial resins was disappointing due to nonspecific or non-pH dependent binding, as well as poor binding capacity. Here the binding capacity and algal Additional Supporting Information may be found in the online version of this article. Ó 2012 American Institute of Chemical Engineers Environmental Progress & Sustainable Energy (Vol.00, No.00) DOI 10.1002/ep Month 2012 1
Transcript
Resins that Reversibly Bind Algae for Harvesting
and ConcentrationJessica Jones,a Christine Beran,b James Beach,b and Martin Poenieba Department of Biomedical Engineering, University of Texas at Austin, Austin, TX 78712b Department of Molecular Cell and Developmental Biology, University of Texas at Austin, Austin, TX 78712; [email protected](for correspondence)
Published onli ne in Wiley Online Library (wileyonlinelibrary.com). DOI 10.1002/ep.1169 7
Algae have great potential to address a number of impor-tant needs such as water remediation and as a feedstock forbiofuel and biochemicals. At present, harvesting or removingalgae is expensive, especially in terms of the cost of making bio-fuel. The expense of harvesting algae arises from the cost ofpumping and processing large amounts of water. In an effortto explore less expensive methods for harvesting, a series ofweak anion exchange resins incorporating various weak baseswith different pKa values were developed to bind and releasealgae as a function of pH. The best of these resins bind about10% of their weight in algae, show 100% reversibility and reus-ability, and the ability to concentrate algae from dilute suspen-sion to 30 g L21. Furthermore, they have the potential fordesalting algae grown in high salt media. These resins providea low cost method for harvesting algae without pumping wateror introducing large amounts of chemicals into the feedwater,harvested algae, or the environment. � 2012 American Institute of
Chemical Engineers Environ Prog, 00: 000–000, 2012
Keywords: algae, biofuel, dewatering, harvest
INTRODUCTION
Algae have the potential to address current and futureenergy needs as feedstock for biofuels [1]. High oil contentand biomass production rates [2], as well as higher photosyn-thetic efficiencies than terrestrial plants [3], make algae idealfor commercial-scale biofuel production. Algae consume car-bon dioxide which can reduce greenhouse gas emissionscompared to petroleum and energy cost balances can beimproved when using flue gas as a carbon source [3]. Fur-thermore, algae can be cultivated in brackish water unsuit-able for agricultural crops, and be used to remediate waste-water [4]. In addition to fuel, profitable coproducts can begenerated, such as animal feed, biopolymers, and agriculturalfertilizers [3]. Algae are also a sustainable source of high-value biochemicals [5]; for example, docosahexaenoic acid(DHA), an essential omega-3 fatty acid important to infantdevelopment and cardiovascular health [6], is synthesizedand concentrated in algae [7].
One of the current barriers to using algae for biofuel onthe commercial scale is the cost of growing and harvestingalgae from dilute suspension. Most algae used for biofuel aresmall, 2–3 lm in diameter, and are difficult to separate fromwater by filtration or tank settling [8]. Furthermore, autotro-
phic algal growth density is limited by light perfusion. Whilelab-scale studies report biomass density in the magnitude ofgrams per liter media [9], large-scale cultivation, such as in araceway pond, often produce biomass concentrations of lessthan 0.5 g L21 [1]. As a consequence, large volumes of watermust be processed to generate the amount of biomassneeded. For example, to recover just one kilogram of algaldry mass from an algal culture of 0.4 g L21 density, at least2500 liters (660 gallons) of water must be processed. Previ-ous studies have shown that biomass recovery contributes20–30% of the total production costs [10]. Thus, the cost ofwater removal from algal biomass poses a significant obstacleto the expansion of algae-based bioenergy [8].
Current techniques used on the commercial productionscale have proven inadequate, requiring considerable time orenergy expenditure to harvest the algal biomass. Gravity sed-imentation is the most common separation method used inwastewater treatment, but new approaches must be devel-oped because of its characteristic long retention time andlow removal rate [11]. Microfiltration is a physical separationtechnique that depends on the membrane pore size and flowcharacteristics [12], and thus may not be suitable for highthroughput processing. Mechanical separation methods suchas centrifugation produce the highest biomass yields, butalso have high equipment and energy costs [10, 13]. Air flota-tion, in which hydraulic equipment generates small bubblesthat adhere to algae and force them to float to the surface, cansimilarly incur high energy costs to aerate a large water volume[14]. Flocculation, the induction of algal cell aggregation, canbe achieved by either charge neutralization or addition ofcharged polymers that coordinate the charged algal surfaces.Although operating cost can be relatively low depending onthe flocculant, it involves continual addition of chemicals tolarge volumes of water. Some of these treatments introduceunwanted salts or metal ions whereas others may have longresidence times and, in any case, can affect the output water[3]. Experimental harvesting techniques, such as ultrasound-induced aggregation, have been effectively demonstrated atthe laboratory scale, but have been projected to be more ex-pensive than current centrifugation techniques [15].
Problems with the current methods have inspired thedesign, develop, and test of reusable resins for concentratingalgae out of suspension. The strategy takes advantage of thenegative surface charge on algae, which bind to weak anionexchange resins as a function of pH. While simple in princi-ple, the performance of commercial resins was disappointingdue to nonspecific or non-pH dependent binding, as well aspoor binding capacity. Here the binding capacity and algal
Additional Supporting Information may be found in the online versionof this article.� 2012 American Institute of Chemical Engineers
Environmental Progress & Sustainable Energy (Vol.00, No.00) DOI 10.1002/ep Month 2012 1
release were studied as a function of resin chemistry. Throughseveral generations of synthesis and testing both bindingcapacity and release properties of the resins were progres-sively improved. The best of these resins can bind 10% of theirweight in algae with essentially 100% reversibility and the abil-ity to concentrate algae 100-fold from dilute suspension.
MATERIALS AND METHODS
Amberlite� IRA-67, Aminomethyl Chemmatrix�, Celite500, Dowex� Marathon2 WBA free base, Levatit� MP-62 freebase, poly(4-vinylpyridine hydrochloride), Sephadex2 G-25,Stratospheres2 PL-DETA, Stratospheres2 PL-DIPAM, Strato-spheres2 PL-PPZ, polymer-bound aniline (564761), andtriethylenetetramine resin (T1522) were obtained from Sigma(St. Louis, MO). Amberlite� CG-400 was obtained from Malli-nickrodt (St. Louis, MO). DEAE Sephadex2 was obtainedfrom Pharmacia Biotech (Sweden). Silica gel was obtainedfrom EMD Chemicals (Gibbstown, NJ). Di(ethylene glycol)vinyl ether (DEG), dimethylaminoethyl methacrylate (DMA),divinylbenzene (DVB), ethylene glycol dimethacrylate(EGDMA), methyl methacrylate (MMA), styrene (S), vinyl2-imidazole (IM), vinyl 4-pyridine (PYR), 2-hydroxyethylmethacrylate (HEMA), and azobisisobutyronitrile (AIBN)were obtained from Sigma.
Neochloris oleoabundans (UTEX LB 1185) was obtainedfrom The University of Texas at Austin culture collection ofalgae (Austin, TX). KAS 603, a saltwater species of Chlorella,was obtained from Kuehnle Agro Systems (HI).
Algal Cultivation and HarvestNeochloris oleoabundans was cultivated in freshwater (�5
psu) Bold 3N (B3N) medium [16], and KAS 603 was culti-vated in saltwater (�35 psu) f/2 medium [17]. Both cultureswere grown at room temperature (238C) under cool whitefluorescent lights on a 12 h:12 h, light:dark photoperiod inairlift photobioreactors aerated with ambient air using aquar-ium pumps. Immediately prior to testing, algae were concen-trated by centrifugation (Sorvall Legend XTR) at 8000 RPM(9700g) for 10 min. The supernatant was decanted and theconcentrate resuspended in freshly prepared B3N medium orf/2 medium prepared with low, medium, or high salinity sea-water (5, 7, 32 psu, respectively), all at pH 6.5.
Commercial Resin TestingCommercially available resins were evaluated for the abil-
ity to bind and release algae. For each assay, 100 mg ofwashed resin was loosely packed into a polyethylene col-umn. A 100 mL volume of algal suspension at 0.2 g L21 bio-
Table 1. Structures of pH-dependent reversible alga-binding resins
2 Month 2012 Environmental Progress & Sustainable Energy (Vol.0000, No.0000) DOI 10.1002/ep
mass density was loaded into the column until the resinreached binding saturation. Unbound algae were rinsed fromthe column with distilled water. The product of the volumeand optical density of the algal suspension at 680 nm(OD680), measured before and after it was passed throughthe resin bed, was used to determine the binding capacity ofthe resin. The resin was then eluted with 10 mM sodiumphosphate, dibasic, adjusted to pH 12.
Resin SynthesisMonomers were combined in ratios based on weight as
indicated in Table 1. EGDMA- and DVB-crosslinked resinswere synthesized with functional monomers DEG, DMA, IM,PYR, and HEMA. The remainder of resin weight was occu-pied by spacer monomers: MMA for EGDMA resins, or S forDVB resins. An equivalent volume of solvent was added tothe combined monomers, consisting of 50% toluene and 50%acetic acid:water (1:30 v/v). Porosity of the crosslinker net-work was held relatively constant by keeping solvent at thesame volume percentage for all resins, though the porosityof these materials could not be measured to confirm theirconsistency. Resin synthesis was carried out in round bottomflasks fitted with an argon bubbler and heated to 608C withconstant stirring. Polymerization was initiated by addition of1 mol % AIBN and polymerization continued until the mix-ture formed a brittle solid. The polymer was then dried in558C oven for 12 h, scraped from the flask and ground bymortar and pestle. The crushed resin was then sized between35 and 170 size stainless steel meshes to obtain particles of�100–500 lm diameter. Before use, the resin was washedwith distilled water.
Determination of Algal Dry Cell WeightRoutine determination of algal dry cell weight (DCW),
was obtained by measuring OD680 using a Shimadzu spectro-photometer. To convert OD680 to DCW, the OD680 wasrecorded for an algal dilution series and then the contents ofeach cuvette were collected onto preweighed cellulose ace-tate membranes (Pall, Port Washington, NY). The membraneswere then dried in a vacuum oven (15 in. Hg, 608C) for 12 hand then weighed to give the DCW for a given OD680. Toavoid optical filtering effects, algal suspensions were dilutedif needed, to keep the OD680 under 1.8. The calibration fitswere linear over the optical density range tested (R2
Neo 50.997, R2
KAS 5 0.999; Supporting Information Figure S1).DCW obtained by optical measurement was validated against
gravimetric measurement of original algal suspension andreleased algae.
Algal Binding and Release AssayTo test algal binding capacity and release characteristics of
newly synthesized resins, 1 g of resin was added to 100 mLof 0.55 6 0.04 g L21 algal suspension in a 250-mL flask, or 500mL of 0.40 6 0.01 g L21 algal suspension in a 1000-mL flask.The suspension was then gently agitated on an orbital shaker(VWR, 125 RPM) for 15 min. Subsequently, the suspension wasfiltered through #170 stainless steel mesh to isolate the resinand the OD680 of unbound algae was measured. The differ-ence in OD680 obtained before and after resin binding wasused to determine how much algae was bound to the resin.The resin was then transferred to a second flask containing anequivalent volume of 10 mM sodium phosphate, dibasic, at pH12 as the initial algal suspension (100 or 500 mL) and the mix-ture was again agitated for 1 h. The resin was then removed byfiltration through the steel mesh and OD680 of desorbed algaewas recorded. To determine the pH dependence of algaeunbinding, the same procedures were carried out except thatalgae were eluted using 10 mM sodium phosphate bufferadjusted to have a pH from 8 to 12.
To determine how well resins could concentrate algae, 1g of resin was added to 500 mL of 0.27 6 0.04 g L21 algalsuspension in a 1000 mL flask. Algae were then eluted offthe resin with different volumes (500, 50, 25, 10, or 5 mL) of10 mM sodium phosphate buffer at pH 12. Experiments werebased on samples run in triplicate. Mean DCW bound andreleased per gram resin was reported, with brackets indicat-ing standard error of the mean.
RESULTS
A number of commercially available resins were initiallytested for their ability to bind and release algae as a functionof pH. The results fell into three categories: resins that boundalgae irreversibly, those that bound reversibly but with verylow binding capacities, and some that showed no binding(Table 2). For example, Amberlite CG-400 and IRA 67, DEAESephadex, poly(4-vinylpyridine hydrochloride), and amino-methyl Chemmatrix all bound �30 mg of Neochloris or 10mg of KAS 603 per g of resin. However, algae could not beremoved by base. On the other hand, a number of resinsand chromatography materials including Dowex MarathonWBA, Levatit MP-62, Celite, silica gel, and Sephadex did notbind to algae even though some of them are anion exchange
Table 2. Commercial resins evaluated for algal binding and release.
Environmental Progress & Sustainable Energy (Vol.0000, No.0000) DOI 10.1002/ep Month 2012 3
resins. A few of the resins such as triethylenetetramine resin,Stratospheres resins PL-PPZ, PL-DETA, and PL-DIPAM, andpolymer-bound aniline, did show some ability to bind andrelease algae, binding �10 mg of Neochloris per g of resin buteven less KAS 603, a potentially high oil-producing species [18].
Based on preliminary results with commercial resins, itwas concluded that weak anion exchange resins showedpromise but the composition of the resin matrix and/ormethod of synthesis was also an important factor. It was alsosuspected that resin-bound weak bases with pKa values near7 would be easier to elute than those with high pKa values.With that in mind a series of resins were generated incorpo-rating one of three different functional groups: dimethyl-amine (DMA), pKa �10.2; imidazole (IM), pKa �7.7; and pyri-dine (PYR), pKa �5.5. Nonfunctionalized EGDMA (EGD-MA:MMA, 80:20 w/w) and DVB resins (DVB:S, 80:20 w/w)that were synthesized as controls showed no binding.
The binding capacity and adsorption reversibility forEGDMA resins containing 10% DMA, IM, or PYR were testedfor algae binding capacity and reversibility using Neochloris(Figure 1). Higher binding capacities were observed forEGDMA-DMA and EGDMA-IM, with 34.2 and 25.3 mg ofalgae bound per gram of resin, respectively, while EGDMA-PYR was less at 13.6 mg. Upon elution with pH 12 phos-phate solution, EGDMA-DMA released 23.2 mg (68% ofbound algae released), followed by the EGDMA-IM thatrelease 23.3 mg (92% release), and finally the EGDMA-PYRresin released 16.0 mg (100% release). Given the relativelypoor binding capacity of the EGDMA-PYR at pH 6.5, furtherstudies focused on the DMA and IM functional groups.
The initial binding and release data showed that bindingcapacity correlated positively with increasing pKa of theweak base component while reversibility correlated inverselywith pKa. Binding reversibility as a function of pH was thendetermined for EGDMA-DMA and EGDMA-IM. Fresh samplesof resins that were preloaded with KAS 603 were eluted with10 mM sodium phosphate solution adjusted to a range of pHvalues from 8 to 12. The proportion of algae released at eachpH is shown in Figure 2. At lower pH (8 and 9), algaeremain largely fixed to both resins. For EGDMA-DMA, 13.1and 28.1% of bound algae were released at pH 8 and 9,respectively, while for EGDMA-IM, 22.5 and 32.1% were
released. At pH 10, EGDMA-DMA released 33.8% of algaeback into suspension, while EGDMA-IM released 70.5%. At apH of 11 or 12, algal binding was mostly reversed, withEGDMA-DMA releasing 86.5 and 100% of algae from theresin while EGDMA-IM released 77.2 and 92%, respectively.
EGDMA-IM and EGDMA-DMA were both effective at bothbinding and releasing algae, but since binding was more eas-ily reversed with IM functionalized resin, this group wasincorporated when examining the effect of resin backbone.Initial results showed that IM incorporated into DVB resin(60:10:30 w/w/w ratio of DVB:IM:S) bound algae poorly.However when the ratio of IM was increased to 60:30:10 (w/w/w DVB:IM:S; Table 1), binding was comparable to that ofthe EGDMA resin containing 10% IM. These two resins werecompared using Neochloris in freshwater B3N medium andKAS 603 in freshwater, brackish, and saltwater f/2 media (7,15, and 32 psu). The results show that EGDMA-IM (Figure3a) had the highest binding capacity for Neochloris (25.3 mgalgae per g resin) whereas binding for KAS 603 was slightlylower in fresh water (22.0 mg g21) and decreased furtherwith increasing salt (brackish, 10.9 mg g21; seawater, 5.1 mgg21). DVB-IM (Figure 3b) had a higher binding capacity forKAS 603 in freshwater medium (41.2 mg g21) than Neochloris(23.5 mg g21). As with EGDMA-IM, KAS 603 binding capacityto DVB-IM was decreased with increasing salt (brackish, 14.6mg g21; seawater, 12.5 mg g21). Even though binding wasreduced at elevated salt concentration, the results wereactually encouraging because the commercial resins we hadtested showed no binding in salt water.
After exploring algal binding behavior at lower monomercomposition, a resin with higher binding capacity was devel-oped. Having shown that increasing the percentage of IMfunctional group improved the binding capacity of DVBresin, similarly the amount of IM was increased to 30 wt %for EGDMA resin. While this also increased binding capacityfor both algal species, a larger fraction of the algae nowbound irreversibly. In addition, previous results showed thatNeochloris was often difficult to completely remove from the
Figure 1. Algal binding capacity of methacrylate resinsincorporating dimethylamine (EGDMA-DMA), imidazole(EGDMA-IM), or pyridine (EGDMA-PYR) as the weak basefunctional group. Each formulation was then tested for itsability to accumulate Neochloris from freshwater suspension(0.4 g L-).
Figure 2. Elution of algae from resins as a function of pH.Sets of EGDMA-DMA and EGDMA-IM resins were preparedand loaded with KAS 603 in freshwater. Corresponding setsof elution solutions were prepared consisting of 10 mMsodium phosphate adjusted to different pH values rangingfrom 8 to 12. Each resin sample was then eluted with a dif-ferent pH buffer and the amount of algae released wasmeasured and compared with the amount originally bound.
4 Month 2012 Environmental Progress & Sustainable Energy (Vol.0000, No.0000) DOI 10.1002/ep
resin, even at 10 wt % IM. However, through testing a num-ber of other functional groups, it was found that inclusion ofeither DEG or HEMA in EGDMA resin with 30 wt % IMresulted in a high capacity, reversibly binding resin. ForEGDMA:IM:DEG (60:30:10 w/w/w), binding to Neochlorisincreased to 109.6 mg g21 resin while for EGDMA:IM:HEMA(60:30:10 w/w/w), binding increased to 134.4 mg g21 resin.Raising the pH to 12 released 90 and 80% of the boundalgae, respectively (Figure 4a). Binding capacity for KAS 603also increased for both EGDMA:IM:DEG (116.7 mg g21) andfor EGDMA:IM:HEMA (120.7 mg g21). Furthermore, after rais-ing the pH to 12 essentially 100% of the bound algae werereleased from both resins.
In the previous experiments, binding reversibility wasmeasured using large volumes of media such that entrapmentof algae in the resin bed, as opposed to binding, was not anissue. However, to be useful for harvesting it is important toshow how well resins can actually concentrate algae. To testthis, a set of EGDMA:IM:HEMA resin samples (Table 1) wasprepared and each sample was mixed with 500 mL of 0.27 60.04 g L21 KAS 603 algae. Once the algae were bound, resinsamples were collected on a filter and eluted with differentvolumes of pH 12 sodium phosphate solution (500, 50, 25,10, or 5 mL). The results (Figure 4b) show that with succes-sively smaller volumes of solution, the concentration of algae
in the eluate increased. As expected, as the volume of eluatedecreased, increasing amounts of algae were trapped in theresin bed. To show that this was merely entrapment ratherthan binding, once samples of the initial eluate were takenfor measurement, washing the resin bed with distilled water(�10–20 mL) in each case removed all the remaining algae.
DISCUSSION
This study was motivated by the need to develop lessexpensive methods to harvest algae from dilute suspension.The basic principle sought here is the ability to bind and con-centrate algae onto the resin in a lower pH medium such aspond water and then release the algae into a small volume ofsolution at a higher pH. Although some commercially availableresins showed promise in this regard, the resins that weredeveloped as a part of this study showed more than 10 timesthe binding capacity of any of the commercial resins tested.However, it is difficult to directly compare the binding capacityof our resin systems with commercial resins since they differ in
Figure 4. (a) Improved binding properties of resins contain-ing EGDMA:IM:HEMA (60:30:10 w/w/w) or EGDMA:IM:DEG(60:30:10 w/w/w). Resins were prepared and tested for bind-ing capacity and algae release at pH 12 for Neochloris in B3Nmedium, and for KAS 603 in f/2 medium at 5 psu. (b) Con-centration and recovery of KAS 603 as a function of elutionvolume for EGDMA:IM:HEMA (60:30:10 w/w/w). Each resinsample was loaded with algae from 500 mL of dilute suspen-sion (0.2 g L21), eluted with progressively smaller volumesof pH 12 sodium phosphate solution (50, 25, 10, or 5 mL),and measured for the amount of algae released and retainedin the resin bed for each elution volume.
Figure 3. Algal binding and release as a function of IM resinbackbone and media composition. Algal binding capacity of(a) EGDMA-IM and (b) DVB-IM resins incorporating vinylimidazole and release at pH 12 were measured for Neochlorisin B3N medium, and for KAS 603 in f/2 media with salinitymatched to fresh, brackish, and saltwater.
Environmental Progress & Sustainable Energy (Vol.0000, No.0000) DOI 10.1002/ep Month 2012 5
bead size and thus surface area/gram, crosslinker density, andmoles of specific binding substituent/gram. In addition thecommercial resins tested were based on DVB polymers, whichshowed a much lower binding capacity for a given percentageof IM as compared to EGDMA resins.
Although resin binding capacity is a significant factor inresin performance, complete reversibility is equally important.Any tendency of the resin to bind algae irreversibly will likelylead to fouling of the resin over time. The goal of this effortwas to develop resins where binding of algae was solely dueto the pH-dependent charge borne by its weak-base constitu-ents. It was apparent from these studies that other factors canalso be involved. IM resins showed a small amount of irreversi-ble binding, which increased with higher percentages of theIM monomer. One possible explanation for this is the bindingof heavy metals by imidazoles [19]. Heavy metal cations couldfunction as a bridge that link algae to the resin. It should benoted that this pH independent binding was reduced whenalcohol groups (DEG or HEMA) were incorporated into theresin. This allowed the generation of resins with higher bind-ing capacities yet with good reversibility. One might speculatethat the incorporation of alcohol ��OH groups into the resinmay have interfered with binding of metals but this will needto be resolved in future studies.
To use resins for harvesting algae, the pKa of the resin weakbase constituents must be higher than the pH of the algalgrowth media. At pH 8 or less, the IM resins show good algaebinding characteristics and the algae can be eluted by solutionsat pH 10 or above. With the DMA resins algae will bind at pHvalues >8 but the elution pH is also elevated to pH 11 orhigher. While this study assumed that algae would be grown ata pH around 8, there is the flexibility in resin design that inprinciple allows one to work over wide range of different pHvalues. However, an alternative mode of use can be envisionedusing a resin whose pKa is less than or equal to the pH of thenormal algal growth medium. This would be possible if the pHgrowth media were temporarily lowered, perhaps by bubblingCO2 into the media, to facilitate binding. Once the bound algaewere removed from the media, loss of CO2 would elevate thepH leading to release of algae from the resin.
It is important to note that polymers synthesized in this studywere convenient for lab-based comparisons but would not beideal for actual harvesting applications. The test resins were syn-thesized with a high crosslinker density, resulting in nonswel-ling, hard resins that could easily be pulverized and sized usingmesh sieves. This proved to be a rapid and simple method fortesting an array of resin formulations. Although the particle sizesare not exact, it provided a basis for comparison between resins.However, for algae harvesting, resin particles would not be thebest format or modality due to problems with entrainment asseen in Figure 4b. Furthermore, these resins, along with mostcommercial resins, are porous with the majority of the ionicbinding sites being internal and therefore unavailable for bind-ing algae. With porous beads, elution of algae by raising the pHwould require neutralizing both internal and external charges.Clearly, it would be better to use nonporous particles or thinfilms of resin rather than porous resin particles.
One proposed alternative to the use of resin beads wouldbe a belt harvester where the belt contained flexible bristles, toincrease surface area, that are coated with a thin film of resin.Assuming that the bristles are relatively far apart there wouldbe better drainage and less opportunity for entrainment ofalgae compared to a bed of small resin beads. A schematic forsuch an implementation is shown in Figure 5. Because thehighly cross-linked resin beads are not porous to algae, it isassumed that algae bind only at the surface of the resin par-ticles. The surface area per gram of resin can be roughly calcu-lated by assuming spherical beads of an average radius. Basedon these assumptions it was determined that 5 g algae wouldbe bound per square meter of surface area.
From the above calculations, a resin-coated conveyor beltwas modeled, 1 m in width and 7.5 m in length, covered onone face with bristles that are 1 mm in diameter, 10 cm inlength, and spaced 1 mm apart. This gives a surface area of592.5 m2 on the bristles (Figure 5). A belt of this surface areacould, if saturated, bind 3 kg of algae. If one assumed thatalgae were harvested from a source pond containing 0.4 g ofalgae per liter, this would correspond to pumping and proc-essing 7500 L of water (1981 gal) with each turn of the belt.
It is important to note that the amount of algae bound to thebelt would depend on a number of factors including pond algaeconcentration, pond volume, belt residence time and rate of cul-ture flow over the belt. However, even if the belt was not satu-rated, it should not greatly affect the final concentration of theharvest. This is because the eluting solution can be reused, oralternatively, a belt could be cycled through the solution until amaximum concentration (at least 50 g L21) is achieved.
In terms of economy of use, it is worthwhile to comparethe use of resins with pH-dependent autoflocculation, amethod of harvesting that also depends on a pH change.Both methods are, in some sense, charge neutralizationmethods for harvesting algae. A pH of �11 is typicallyneeded to flocculate algae efficiently [20–26] and one of themost economical ways to elevate the pH is to add lime (cal-cium oxide) or quicklime (calcium hydroxide or hydratedlime) to the algal suspension. Raising the pH above 10 con-verts magnesium salts in the media to magnesium hydroxidewhich is thought to be the main flocculant [27]. It should benoted that raising the pH causes other salts such as calciumcarbonate to form which coprecipitate with the algae.
Resins can be eluted using lime or calcium hydroxide as thebase (data not shown) and for the DMA resin, the pH neededis similar to that required to flocculate algae (�11). Further-more, once algae are eluted with a pure calcium hydroxide so-lution, algae will flocculate upon standing. Thus one can drawa number of parallels between resin-based harvesting andautoflocculation using calcium hydroxide as the base.
Despite these similarities, there are important differences.With autoflocculation, supersaturated solutions are often gen-erated to speed the flocculation and settling such that the re-sultant sludge contains a large amount of inorganic precipi-tate (mainly calcium carbonate) [28]. Thus relatively largeamounts of lime are needed to maintain the pH while drivingthe formation of calcium carbonate. Furthermore, when usedfor harvesting, large volumes of water must be treated withlime. In eluting resins with calcium hydroxide, only smallvolumes of water are needed and algae flocculate withoutforming an inorganic precipitate.
Interestingly, elution of resin-bound algae with pure cal-cium hydroxide solutions runs counter to expectations sincemagnesium hydroxide is thought to be the active flocculant.
Figure 5. Commercial application of harvesting resins. Sche-matic for algal harvest by reversibly binding resins in indus-trial application. The resin is applied as a high surface areasubstrate over a rotating platform, such as a conveyor belt.
6 Month 2012 Environmental Progress & Sustainable Energy (Vol.0000, No.0000) DOI 10.1002/ep
However, the resin results are consistent with a recent reportfrom Schlesinger et al., who reported that calcium ions alonecould flocculate algae and that the efficiency of flocculationgrew exponentially with algal concentration [29]. Resinsgreatly concentrate the algae which may promote floccula-tion as seen when algae were eluted with calcium hydroxide.
As a final note, although the resins in this study are lessefficient in seawater they still will bind algae under high saltconditions, while all commercial resins tested were ineffec-tive at binding in salt water. Given that algae can be elutedinto a low salt medium, depending on how effectively salt-water is drained from the algae, some degree of desaltingcould be achieved. This would be improved even more ifthe resin-bound algae were washed with fresh water prior toelution. This may be an important factor when consideringthe use of algae grown in saltwater for feedstock or fertilizer.
This work was supported in part by a grant from OpenAlgae.
LITERATURE CITED
1. Chisti, Y. (2007). Biodiesel from microalgae, Biotechnol-ogy Advances 25, 294–306.
2. Li, Q., Du, W., & Liu, D. (2008). Perspectives of micro-bial oils for biodiesel production, Applied Microbiologyand Biotechnology 80, 749–756.
3. Li, Y., Horsman, M., Wu, N., Lan, C.Q., & Dubois-Calero,N. (2008). Biofuels from microalgae, Biotechnology Pro-gress 24, 815–820.
4. Munoz, R. & Guieysse, B. (2006). Algal-bacterial proc-esses for the treatment of hazardous contaminants: Areview, Water Research 40, 2799–2815.
5. Gavrilescu, M. & Chisti, Y. (2005). Biotechnology—Asustainable alternative for chemical industry, Biotechnol-ogy Advances 23, 471–499.
6. Pyle, D.J., Garcia, R.A., & Wen, Z. (2008). Producingdocosahexaenoic acid (DHA)-rich algae from biodiesel-derived crude glycerol: Effects of impurities on DHAproduction and algal biomass composition, Journal ofAgricultural and Food Chemistry 56, 3933–3939.
7. Harwood, J. & Guschina, I. (2009). The versatility of algaeand their lipid metabolism, Biochimie 91, 679–684.
8. Uduman, N., Qi, Y., Danquah, M.K., Forde, G.M., &Hoadley, A. (2010). Dewatering of micralgal cultures: Amajor bottleneck to algae-based fuels, Journal of Renew-able and Sustainable Energy 2, 1–15.
9. Banerjee, A., Sharma, R., Chisti, Y., & Banerjee, U.C.(2002). Botryococcus braunii: A renewable source of
hydrocarbons and other chemicals, Critical Reviews inBiotechnology 22, 245–279.
10. Grima, E.M., Belarbi, E.H., Fernandez, F.C.A., Medina,A.R., & Chisti, Y. (2003). Recovery of microalgal biomassand metabolites: Process options and economics, Bio-technology Advances 20, 491–515.
11. Park, J.B.K., Craggs, R.J., & Shilton, A.N. (2011). Recy-cling algae to improve species control and harvest effi-ciency from a high rate algal pond, Water Research 45,6637–6649.
12. Hung, M.T. & Liu, J.C. (2006). Microfiltration for separa-tion of green algae from water, Colloids and Surfaces B:Biointerfaces 51, 157–164.
13. Olaizola, M. (2003). Commercial development of micro-algal biotechnology: From the test tube to the market-place, Biomolecular Engineering 20, 459–466.
14. Wiley, P.E., Brenneman, K.J., & Jacobson, A.E. (2009).Improved algal harvesting using suspended air flotation,Water Environment Research 81, 702–708.
15. Bosma, R., Spronsen, W.A.V., Tramper, J., & Wijffels, R.H.(2003). Ultrasound, a new separation technique to harvestmicroalgae, Journal of Applied Phycology 15, 143–153.
16. Brown, M.R. & Bold, H.C. (1964). Comparative studiesof the algal genera Tetracystis and Chlorococcum (pp. 213),Austin, Texas, USA: University of Texas Publication.
18. Jones, J., Manning, S., Montoya, M., Keller, K., & Poenie,M. (2012). Extraction of algal lipids and their analysis byHPLC and mass spectrometry, Journal of the American OilChemists’ Society 89, 1371–1381.
19. Kara, A., Uzun, L., Besirli, N., & Denizli, A. (2004). Poly(ethylene glycol dimethacrylate- n-vinyl imidazole)beads for heavy metal removal, Journal of HazardousMaterials 106B: 93–99.
20. Ayoub, G.M., Lee, S.I., & Koopman, B. (1986). Seawaterinduced algal flocculation, Water Research 20, 1265–1271.
21. Dziubek, A.M. & Kowal, A.L. (1984). Effect of magne-sium hydroxide on chemical treatment of secondaryeffluent under alkaline conditions, Proceedings of theWater Reuse Symposium III (pp. 1428–1436), San Diego:AWWA Research Foundation.
22. Dziubek, A.M. & Kowal, A.L. (1989). High-pH coagula-tion-adsorption: A new technology for water treatmentand reuse, Water Science and Technology 21, 511–517.
23. Elmaleh, S., Yahi, H., & Coma, J. (1996). Suspended sol-ids abatement by pH increase—Upgrading of an oxidan-tion pond effluent, Water Research 30, 2357–2362.
24. Lavoie, A. & Noue, J.D.L. (1986). Harvesting of Scene-desmus obliquus in wastewaters: Auto- or biofloccula-tion? Biotechnology and Bioengineering 30, 852–859.
25. Shin, H.S. & Lee, S.M. (1997). Removal of nutrients inwastewater by using magnesium salts, EnvironmentalTechnology 19, 283–290.
26. Yahi, H., Elmaleh, S., & Coma, J. (1994). Algal floccula-tion-sedimentation by pH increase in a continuous reac-tor, Water Science and Technology 30, 259–267.
27. Folkman, Y. & Wachs, A.M. (1973). Removal of algaefrom stabilization pond effluents by lime treatment,Water Research 7, 419–435.
28. Shelef, G., Sukenik, A., & Green, M. (1984). Microalgaeharvesting and processing: A literature review. Spring-field: National Technical Information Service.
29. Schlesinger, A., Eisenstadt, D., Hilla, A., Carmely, H., Ein-binder, S., & Gressel, J. (2012). Inexpensive non-toxicflocculation of microalgae contradicts theories: Overcom-ing a major hurdle to bulk algal production. Biotechnol-ogy Advances 30, 1023–1030.
Environmental Progress & Sustainable Energy (Vol.0000, No.0000) DOI 10.1002/ep Month 2012 7
