High-Density Photoautotrophic Algal

7/30/2019 High-Density Photoautotrophic Algal

1/8

High-Density Photoautotrophic AlgalCultures: Design, Construction,and Operation of a NovelPhotobioreactor SystemMinoo Javanmardian and Bernhard 0. Palsson*Cellular Biotechno logy Laboratory, Departm ent of Chem ical Engineering,Universi ty of Michigan, An n A rbo r, Michig an 48109Received December 3, 1990lAccepted Apr i l 26, 1991

A photobioreactor system has been designed, constructed,and implemented to achieve high photosynthetic rates inhigh-density photoautotrophic algal cell suspensions. Thisunit is designed for efficient oxygen and biomass produc-tion rates, and it also can be used for the production ofsecreted products. A fiber-optic based optical transmissionsystem that is coupled to an internal light distributionsystem illuminates the culture volume uniformly, at lightintensities of 1.7 mW/cm2 over a specific surface area of3.2 cmz/cm3.Uniform light distribution is achieved through-out the reactor without interfering with the flow patternrequired to keep the cells in suspension. An on-line ultra-filtration unit exchanges spent with fresh medium, and itsuse results in very high cell densities, up to lo9cells/mL[3% (w/v)] for eukaryotic green alga Chlorella vulg aris. DNAhistograms obtained from flow cytometric analysis revealthat on-line ultrafiltration influences the growth pattern.Prior to ultrafiltration the cells seem to have halted at a par-ticular point in the cell cycle where they contain multiplechromosomal equivalents. Following ultrafiltration, thesecells divide, and the new cells are committed to division sothat cell growth resumes. The prototype photobioreactorsystem was operated both in batch and in continuous modefor over2 months. The measured oxygen production rate of4-6 mmol/L culture h under continuous operation is consis-tent with the predicted performance of the unit for the pro-vided light intensity.Key words: photobioreactor ultrafiltration photoauto-trophic - DNA histograms cell cycle flow cytometryChlorella vulgarisINTRODUCTIONOver the past decade we have seen the biotechnologicaluse of genetically-engineered bacteria, yeast, and mam-malian cells grow at a rapid rate and reach a relativelyhigh degree of sophistication. However, despite theirrecognized biotechnological potential, the developmentof photoautotrophic cell cultures has been slower.6 Thehistory of the commercial use of algal cultures spansabout50years with application to wastewater treatment,mass production of different strains such as ChlorellaandDunaliella, and production of some specialty chemi-cals., Large-scale open algal cultures, such as linedand unlined ponds, typically reach low cell densities of0.01-0.06% (w/v),,~esulting in expensive harvestingprocedures and unfavorable economic * ^ Even with* To whom all correspondence should be addressed.

Biotechnology and Bioengineering,Vol. 38, Pp. 1182-1189 (1991)0 1991 J ohn Wiley & Sons, Inc.

the development of closed tubular photobioreactors withcontrolled COzsupplementation, values of up to only0.13% have been reported.Currently significant interest is developing in the useof high-density algal cultures to produce high-valueproducts, such as pharmaceuticals and genetically-engineered product ^These include antibacterial, anti-viral, antitumor/anticancer, antihistamine, and manyother biologically valuable pr od~~ts. - ~ ~~ ~~~eneti-cally engineered cyanobacteria have the potential to beused for a variety of purposes, such as biological controlof insect larvae.26The genetic manipulation of cyano-bacteria and microalgae, however, has lagged behindparallel developments with bacteria and yeast.One of the impediments to the commercial use ofhigh-density photoautotrophic algal cultures for produc-tion of high-value products is the availability of suitablephotobioreactors. In this article we describe the design,construction, and implementation of a novel photobio-reactor system for the cultivation of photoautotrophiccells. This photobioreactor system is capable of cul-tivating Chlorella vulgaris at densities of 10 cells/mL[=3% (w/v)].ORDER-OF-MAGNITUDE CALCULATIONSBefore embarking on the detailed design of photobiore-actor systems, it is useful to estimate their maximalachievable performance and identify the design limit-ing factors. Simple calculations based on known physico-chemical and biological parameters enable us to set ourdesign targets (Table I).Volumetric ProductivityThe volumetric oxygen production rate for growing cellscan be used to specify reactor performance, and it canbe calculated as follows:Volumetric oxygen production rate = specific

oxygen yield (1)x steady-state cell concentration (2)

x growth rate (3)Numerical values for these three quantities are available.

CCC 0006-3592/91/01001182-08$04.00


2/8

Table I .bioreactor.Design targets for an optimally perf orming photo-

Parameter TargetL ight intensity ( 10)(usable light)Specifi c surface area (S /V)Cell densityCell growth rateVolumetric productivity

1-2 mw/cm5-10 cm/cm10 celI s/mL0.3-0.4 g/L h0.02mol Oz/L h

The specific oxygen yield (mol Oz/cell) can be esti-mated from the stoichiometry of photosynthesis. Thestoichiometries presented here are based on those ob-served by Pirt et aLZ2:0.89C02+O.61H2O +O.127NH3

0.89 C-mol biomass+O2(ammonia as nitrogen source)

O.71CO2+OS9H2O +0.101NO;*0.71 C-mol biomass+O2(nitrate as nitrogen source)

One C-mole of ash-free biomass (the amount that con-tains 1rnol carbon) is taken to be C I , o H ~. ~~0~. ~~~N ~. 1~3,which has the formula weight of 21.8 for nitrate as nitro-gen source, and Cl.oH1.800.432N0.143,ith the formulaweight of 22.7 for ammonia as nitrogen source.15~17~22Based on photosynthetic stoichiometry, the specificoxygen yield is estimated to be in the range of 0.044-0.046mol (=1L) of oxygen generated per gram dryweight generated biomass. This factor is an intrinsicproperty of algal biochemistry and for practical pur-poses must be considered as an inherent biologicalconstraint.The steady-state cell concentration is primarilya function of bioreactor design, with the illuminatedsurface-area-to-volume ratio and the rate of oxygenremoval being the most important design variables.Assuminga cell volume on the order 30 fL/cell (repre-sentative of cell volume of C. vulgarisat high cell densi-ties, >loscells/mL, in photoautotrophic cultures; thevalue was measured by a Coulter Channelyzer), thepacking density of cells will be about 3 X 10 cells/mL.Therefore, a density of about lo9cells/mL, or 3% (w/v),may be considered as a reasonable design goal for thesteady-state cell concentration. Above concentrationsof 10% (w/v) the culture medium will turn to a creamymaterials and will become difficult to process.The growth rate is assumed to be0.06h-, which cor-responds to 12h doubling time. The volumetric oxygenproductivity can then be estimated to be 0.02molOz/L h. An efficient photobioreactor must be able togrow the cells to high cell concentrations, sustain cellsat these high concentrations, and maintain high growthand oxygen production rates.

Light RequirementsA major theoretical and practical limitation on micro-algal productivity is the photosynthetic efficiency (bio-chemical energy stored in biomass per energy absorbedby the cells in the form of light). The minimum lightrequirement for production of 1 rnol oxygen has been asource of controversy since the Z-scheme of photosyn-thesis mechanism was put forth in 1960.14According tothe Z-scheme, eight photons of light (quanta) are re-quired to produce 1mol oxygen. However, reviews onthis to pi^^^ have suggested that less than eight photonsmay be required for the formation of one oxygen mole-cule, and environmental conditions such as temperaturehave a strong influence on the efficiency of photosyn-thesis. The minimum photon requirement for Chlorellacells was estimated to be in the range of 5-6 hv/Oz.21Based on this quantum requirement and assuming thatthe energy content (enthalpy) of Chlorella biomass isabout 25kJ/g dry ~ t , ~ ~ ~maximum photosyntheticefficiency in the range of 32-54%, depending on assimi-lation quotient, is predicted. This efficiency is 10-20%higher than the predicted photosynthetic efficiency bythe Z-scheme theory (23-30%).Wewill use the Z-scheme stoichiometry of 8hv/Oztospecify the photobioreactor design requirement. Thisvalue corresponds to 160pE/mL h, or in the range of8-10 mW/mL (550-680 nm average light frequency) forproduction of 0.02mol Oz/L h. Then, by assuming aphotosynthetic efficiency of 23-30%, the target oxygenproduction rate will correspond to a biomass accumula-tion rate of 0.3-0.4 g dry wt/L h.Specific Surface Area RequirementsFollowing the total light energy requirement, the nextquestion that needs to be addressed is the light penetra-tion distance at high cell densities. The light intensitymay be described in Beers law as an exponentially de-caying function of distance into the culture volume witha characteristic length constant that is proportional tocell density.* For Chlorella pyrenoidosa, the penetrationdistance at 680nm has been measured to be about1cmat cell concentrations of about 10 cells/mL.8 Thus atlo9cells/mL the penetration distance is estimated to beabout 1 mm.The light penetration distance can also be calculatedfrom the following empirical equation:

6000CCd = -

whereC, is the algal dry weight concentration in milli-grams per liter andd is the light penetration distance incentimeters. This equation predicts a light penetrationdistance of about2mm at 30g/L dry weight.This information indicates that for a target cell con-centration of lo9cells/mL, the required specific area

J AVANMARDIAN AND PALSSON: HIGH-DENSITY PHOTOAUTOTROPHIC ALGAL CULTURES 1183


3/8

for a reactor is about 5-10 cm2/cm3, and the desiredlight intensity will range between 1and 2 mW/cm ofusable light for the cells (in the red and blue regions ofthe spectrum). Table I summarizes the design targetsfor an optimal photobioreactor system.PHOTOBIOREACTOR DESIGNOur photobioreactor system consists of six major com-ponents: the light source, the optical transmission sys-tem, the photobioreactor, the gas-exchange unit, theultrafiltration unit, and the on-line sensors for monitor-ing the condition of the culture (Fig. 1).The culturemedium circulates in a closed loop between the reactor,the gas-exchange module, and the ultrafiltration unit.The circulation rate is calibrated so that the incomingstream to the photobioreactor is low in oxygen and theexiting stream is close to saturation. The gas suppliedto the gas-exchange module is a mixture of nitrogen,oxygen, and carbon dioxide, whose compositions can becontrolled. The effluent gases from the gas-exchangemodule flow through a condenser, trapping the watervapor prior to analysis of the gas composition. The pH,dissolved oxygen, and dissolved carbon dioxide concen-

trations in the inlet and outlet stream are measured con-tinuously and fed to a computer which stores the data.The on-line data acquisition system provides a directmeasurement of the carbon dioxide fixation and oxygenproduction rates. An on-line ultrafiltration unit is usedto dialyze the culture medium at a relatively high flowrate. This unit allows us to selectively separate thewaste and/or secreted products and exchange them withfresh medium. More detailed descriptions of the opticaltransmission system, the gas-exchange module, and theultrafiltration unit are found in the Appendix.As discussed above, one of the main challenges inthe design and construction of the photobioreactor is toprovide high illuminated surface-area-to-volume ratio.Figure 2 illustrates the geometric configuration of theprototype photobioreactor unit. The reactor has a cul-ture volume of 0.6 L and the photobioreactor systemhas a total volume of 1L . The light radiators inside thecylindrical bioreactor are arranged as concentric verticalcylinders to provide3.2 cm2 radiative surface per mLof algal suspension.

Light entering the reactor through the fiber-opticcables passes through +-in.-thick acrylic cylinders. Thesecylinders form the center of light radiators. They consist

Figure 1 F lowsheet for photobioreactor system used in study.

1184 BIOTECHNOLOGY A ND BIOENGINEERING, VOL. 38, NO . 10, DECEMBER 5, 1991


4/8

SIDE VIEW

Irtificial Illumination SystemI (Xenon lamp andfilters)Iibcr optic cables

Inlet culture

V

Figure 2. Details of prototype photobioreactor unit used in study.of two $-in.-thick annuli inserted into each other withwedge-shaped cuts of increasing depth in inner surfaceof the outer annulus, and outer surface of the innerannulus, allowing the light to escape evenly along thelength of the annuli to illuminate the algal suspension.This method of construction provides the most efficientlight radiation system, and since the rough surfaces arenot in contact with cells, the clogging of light openingswith microalgae will be prevented. The specificationsof the prototype photobioreactor are summarized inTable 11.MATERIALS AND METHODSChlorellu vulguris, Emerson strain, from Carolina Bio-logical Supply was cultured in the bioreactor system inN-8 medium at a pH of 5.6.27The medium consisted of(mg/L): Na2HP04 2H20, 260; KH2P04,740; CaC12,10; Fe EDTA, 10; MgS04 7H20,50; K N03, 1000; and1mL of trace element stock with composition of (g/L):A12(S04)3. 18H20, 3.58; MnC12. 4H20, 12.98;CuS04 5H20,1.83; and ZnS04.7H20, 3.2.

Table11. Specifications of prototype photobioreactor unit.Diameter 10 cmHeight 17 cmCulture volume 600mLI lluminated sur face/volumeof cultureL ight radiator thickness In.M aterial A crylicI rradiance (400-700 nm) 1.7 mW/cm'Usable irradiance

0.6 rnW/cm'

3.2 cm'/cm'I .

(red and blue part of spectrum)Number of fiber-optic bundles 100

Algal cell concentration was measured with a CoulterCounter Model ZM. This system also includes a CoulterChannelyzer which can measure particle size distribu-tions. The culture was cultivated in the system with areasonably constant temperature of 25 k1"C and a cir-culation flow rate of =2 L/min. The gas composition tothe cartridge was controlled by multitube flowmeters.The input gas composition to the hollow fiber was keptat20%02,% C02,and 75% N2with a total flow rateof 300 mL/min. The rate of ultrafiltration was about8mL/min, and the molecular cut-off of the membranewas 100 kD. The light intensity inside the reactor wasmeasured by an LI-COR light meter model LI-185 fromLAMBDA Instruments Corporation which indicatedan intensity of 1.7 mW/cm of light inside the reactor,and the total light energy provided to the reactor wasabout 3.2 W. The gas composition from hollow fibercartridges was examined by a Hewlett-Packard gaschromatograph Model 5890 with autosampler for di-rect measurement of inlet and outlet gas to and fromthe cartridge.The DNA content of the cells were quantified byusing a flow cytometer. Samples containing a total num-ber of 2 x lo6cells were centrifuged (280g, 10 rnin),suspended in a small volume (


5/8

scale were acquired to a total count of 30,000 events.The linear voltage was adjusted to allow the inclusion ofall DNA peaks in a single linear histogram. T he DN Ahistograms were bitmapped off two-parameter histo-grams to eliminate electronic noise and signals gener-ated from cellular debris. The light energy emitted fromthe PI -DNA complex was directed to a photomultipliertube by a 550-nm dichroic beam splitter followed by a630-nm long-pass fil ter. T he microscopic examinationof the cells comprising peaks in the DNA histogramsand sorted by flow cytometer revealed an insignificantnumber of clumps. Therefore, the DN A histograms areassumed to indicate the per-cell DNA content withina population.

RESULTSThe photobioreactor system can grow C. vulguris intohigh cell densities. A typical growth curve for batchcultivation is shown in Figure 3. C. vulguris inoculatedat lo7cells/mL was grown up to lo9cells/mL over 16-20 days. The use of the ultrafiltration unit was critical(Fig. 3) in order to achieve cell concentrations abovelo8cells/mL. As discussed below, the ultrafiltration unitserves to remove secreted autoinhibitory compoundsenabling growth into high densities. A repeated ultra-filtration at densities of lo9cells/mL at day 26 did notlead to further growth.Figure 4A shows the growth curve of an experimentof longer duration with the same initial conditions andmore frequent useof the ultrafil tration unit. This experi-ment lasted about 2 months. T he cells were cultivated ina batch mode until a concentration of lo9cells/mL wasachieved. Again the key factor in achieving high cellconcentration was the use of the on-line ultrafi ltrationunit. The pH and the specific oxygen production rate ofthe culture were monitored on-line through three peri-ods of ultrafi ltration (Figs. 4D, E).Following the onsetof ultrafi ltration the pH dropped rapidly to about 5.6,which is the pH of fresh medium, and then rose gradu-ally after termination of ultrafil tration. T he oxygenproduction rate increased following the period of ultra-filtration, and a concomitant expansion in the cell num-ber took place. This result indicates that prior to on-l ineultrafi ltration, the culture is limited by either lack ofnutrients or the accumulation of autoinhibitory com-pounds that are secreted by the cells.To examine the effectsof ultrafiltration on cell growthat high cell concentrations, flow cytometric DNA histo-grams were obtained prior to and following ultrafiltra-tion (Fig. 5). Before ultrafiltration, the culture was notactively growing, and the DN A histogram had two dis-tinct peaks showing the presence of two cell popula-tions, one with a single set of chromosomes and anotherwith multiple sets of chromosomes. This observationindicates that the growth is halted both at the commit-ment and at the division stages in the cell cycle. Fol-

loor------

10'0 8 16 24 32 40Timefdays)

Figure 3. Growth curve for C. vulgaris grown in a batch modein the photobioreactor system, shown in F ig. 1. Arrows indicatetiming of on-line ultrafiltration. Each ultrafiltration was carriedout for 8 h against fr esh medium.lowing ultrafiltration the multiple chromosome peakdisappears, and more cells enter the cell cycle from theGo/GIphase to the DNA synthesis phase (S phase), andthe culture resumes its normal growth condition.Based on this observation, it appears that at high celldensities the cells secrete autoinhibitory compoundsthat are responsible for the cell cycle alteration. To ver-ify this phenomena, an experiment was conducted toinvestigate the effect of the autoinhibitory compoundson low-density cultures. Exponentially growing cellswere inoculated at low density into fresh medium thatwas supplemented with spent medium from high den-sity (lo9 cells/mL) culture. The DNA histograms fromdifferent cultures clearly show the effect of these com-pounds on the cell cycle of C. vulgaris (Fig. 6). Com-parison between the control (0%conditioned medium)and 50% and 90% conditioned medium indicates thatthe broad exponential pattern of the DNA histogramsresolves into separate peaks with two or three chromo-somal equivalents. This observation verifies that thesecreted products have the same effect on low-densitycultures (-lo6 cells/mL) as they do on high-density cul-tures. This observation also rules out nutrient limita-tion as the cause for the growth inhibition, since thenumber of the cells at low-density cultures is three or-ders of magnitude less than high-density cultures andin both cases the same composition of fresh mediumwas used.The photobioreactor system was placed in continuousmode on day 52, with a dilution rate of 0.15 per day(150 mL/day). The oxygen production rate under theseconditions was in the range between 4 and 6 mmol/Lculture/h (Fig. 4F). T he steady-state concentration atthis dilution rate was measured to be 8 x 10' cells/mL(Fig. 4C). Assuming a cell volume of 30-femtol and dry-to-fresh-weight ratio of 0.25, 63 mg dry Chlorellu wasproduced per liter of reactor per hour.The total amount of light provided to the reactorfrom the xenon lamp was 3.2 W with 30-35% of thepower in the usable frequency range for photosyn-thesis (-1.2 W ). Based on the usable provided light(0.6mW/cm2), the oxygen production rate in this unit

1186 BIOTECHNOLOGY A ND BIOENGINEERING, VOL. 38, NO. 10, DECEMBER 5, 1991


6/8

lo9

0 ( C)0o 8 0

lo8: 5

uPtsY8is

7.0 ( D)

3 5 7 9 11 13 15Time (days)

0 C,.3

48 50 52 54 56 58Time(days)

3 5 7 9 11 13 15Time(days)

6

d.2 4Ei2

2

54 55 56 57 58 59Time(days)

Figure 4. Operation of photobioreactor system over a 2-month period. (A ) Growth curve of C. vulgaris in batch and continuous mode.Arrows indicate the timing of on-line ultrafi ltration. Arrows5and 6 indicate ultrafiltration against medium with five-fold elevated potas-sium nitrate concentration (5g/L) over normal medium(1g/L). These particular ultrafiltrations were donetoverify that the culture isnotnutrient limited. The concentration of potassium nitrate was increased in the medium because this compound is the most likely nutrient tobe depleted in the processof growth. As the growth curve shows, the additional availability of nitrogen source did not result in substantialstimulationof growth. After the reactor was operated over8days in continuous mode, i t had to be stopped due to lossof light. (B) Growthcurve during the first three ultrafi ltrations. Arrows indicate time of ultrafiltration. (C ) Growth curve during continuous mode. (D) On-linepH monitoring during ultrafiltration. Arrows indicate timeof ultrafiltration. (E)On-line dissolved oxygen concentration monitoring dur-ing ultrafi ltration. Arrows indicate time of ultrafiltration. (F )Oxygen production rate during continuous operation.

was very close to the predicted amount by the order-of-magnitude calculations. Assuming the Z-scheme forphotosynthetic mechanism and 23% efficiency for con-verting light energy tobiomass, it is expected to produce60-80 mg dry biomass/L h, which agrees with the ac-tual biomass production (63mg/L h) rate. These resultsdemonstrate that the reactor is performing as predicted.

D~SCUSS~ONA photobioreactor for mass cultivation of algal cells athigh cell concentrations has been designed, constructed,and implemented. Our present unit meets all the speci-fications, based on physicochemical and biological con-straints, except the surface light intensity. Based on the

JA VANMA RDIAN AN D PAL SSON: HIGH-DENSITY PHOTOAUTOTROPHIC A LGAL CULTURES 1187


7/8

18001500 '1200 .900 .600.3 300.

1 5 0 0 :3 0 -

7Hours afterUltrafiltrationUltrafiltrationstarted

27Hour s afterUltrafiltration 40Hour s afterUltrafiltra tion

Channel N umberFigure 5. T ime sequence of DNA histograms of C. vulgaris follow-ing ultrafiltration during second ultrafiltr ation peri od.

provided light and nutrients to the photobioreactor, theperformance of the system meets the estimated achiev-able growth rate and oxygen production rate. This per-

formance could be improved further by increasing thelight intensity at- the illuminating surfaces inside thereactor to reach the estimated performance of 0.3-0.4g/1 h biomass production and0.02mol/L h oxygenproduct on.Even with illumination that falls short of the desireddesign goal, we demonstrate that very high density algalcultures can be established and maintained. This tech-nological achievement has led to the discovery of un-usual cell cycle behavior at high cell densities. Theautoinhibitory compounds alter the cell cycle kineticsofChlorella.These factors may be analogous to the phero-monal mating factors known to be produced by othersingle cellular eukaryotes." These secreted compoundsalso may have bacteriocidal activity. There have beensome reports about the bacteriocidal activity of spentmedium fromChlorella culture ' but this phenomenonhas not been extensively studied.The attainment of active high-density algal cultureswill allow algal cultures tobe economically feasible forthe production of high-value pharmaceutical productsand perhaps recombinant proteins.

2500, T imedl(Day1)

0 64 128 192 :Channel No.50% Cond. M ediaControl(0%Cond. M edia)

6

90%Cond. M edia

Day 10 Day 1C1500l " - I L , ,;I,,b, ,

00 64 128 192 0 64 128 192 0 64 128 192 : 6Channel Number

Figure6. T ime sequence of DNA histograms of C. vulgaris cultivated in condi-tioned medium.

1188 BIOTECHNOLOGY A ND BIOENGINEERING, VOL. 38, NO. 10, DECEMBER 5, 1991


8/8

APPENDIXOptical Transmission Sys temThe main goal in designing the optical transmissionsystem was to provide uniform light with the requiredintensity at the illuminating surfaces of the reactor.This task was accomplished by conducting the requiredportion of the light spectrum via fiber-optic bundles tothe il luminating surfaces of the reactor. In this fashioninternal illumination is provided for the system, whilethe light source is kept out of the reactor, protectingthe reactor from the lamp waste heat, and preventingunnecessary design and performance complications.A xenon lamp is used as an artificial near-duplicatesource of solar spectrum for controlled illumination.However, the efficiency of arc lamps is about 2-3%.The efficiency can be signif icantly improved by usingother artificial light sources such as light emitting di-odes (LEDs) if the use of artificial light proves to benecessary. These light sources can have efficiencies upto 80% in converting electrical energy to light at specificwavelengths, such as 660 nm (red light).3Gas-Exchange Devic eHollow fiber cartridges with microporous hydrophobicpolypropylene fibers are used as eff icient devices to ex-change C 02and O2 n the suspension culture. The ad-vantages of these units are their high efficiency at lowflow rates (90%)and high mass transfer coefficients inthe range of 0.05-0.3 mmol/h mm gas at liquid flowrates of 0.5-7.0 l/min. The gas-exchange membranesare external to the reactor to avoid unnecessary compli-cations with respect to the light delivery to the culture.They can be used as a single unit or in serial arrange-ment. In serial arrangement oxygen can be stripped offthe culture in one unit under low pressure, and carbondioxide can be dissolved in the culture in another unitunder high pressure to increase the solubility.Ult raf i l t rat ion Uni tOn-l ine ultrafiltration units allow the system to achievehigh cell densities as well as the separation of autoin-hibitory compounds without removing the cells fromthe system. The significance of this unit is demon-strated in the preliminary results. In this particular sys-tem, we have used Minisette acrylic cell from FiltronTechnology Corporation, which is a high-performancetangential flow membrane unit and can be scaled upeasily. The membrane is an open-channel polyethersul-fone (PES)with a molecular cut-off of 100kD.The sur-face area is 0.75 ft2, the retentate flow rate is 2 L /min,and the filtrate flow rate is 8 mL/min, which can beincreased to 30 mL/min.

The authors wish to thank Uris Upatneiks from Environ-mental Research Institute of Michigan (E RIM ) for his help

in constructing the photobioreactor unit and designing thedetails of the internal illumination system and CharlesJ acobus (E R I M ) for many stimulating discussions. Wewould also li ke to thank Mak Cameron of The University ofMichigan for his help with the flow cytometry. This projectwas funded through an A nn Arbor based NASA center forCommercial Development of Space (CCDS).

References1.

2.3.4.

5.

6.

7.

8.9.10.11.12.

13.14.15.16.

17.18.

19.

20.21.22.23.24.25.26.27.

A motz, A . B ., A vron, M . 1989. p. 91 I n: R .C . C resswell,T. A.V. Rees, and H . Shah (eds.), Algal and cyanobacterial bio-technology. L ongman Scientif ic&Technical, Harlow, England.Avron, M . 1989. p. 385 In: Y. Cohen and E. Rosenberg (eds.),M icrobial mats. A merican Society of Microbiology.Barta, D. J ., Tibbits, W.T., Bula, R. J ., Morrow, R . C. 1991.A dvances in space research, Pergamon Press, Oxford, in press.Benemann, J . R., K oopman, B. L., Weissman, J . C., E isenberg,D. E., Goebel, R . P. 1980. p. 457 In: G. Shelef and C . Soeder(eds.), A lgal biomass. E lsevier Biomedical, A msterdam.Borowitzka, L. J ., Borowitzka, M. A . 1989. p. 294 In: R.C.Cresswell, T. A.V. Rees, and H . Shah (eds.), Algal and cyano-bacterial biotechnology. L ongman Scientif ic & Technical,Harlow, England.Chapman, D. J ., Gell enbeck, K .W. 1989. p. 1 I n: R . C .Cresswell, T.A.V. Rees, and H . Shah (eds.), Algal and cyano-bacterial biotechnology. L ongman Scientif ic & Technical,Harlow, England.Chaumont, D., T hepenier, C ., Gudin, C. 1988. p. 198 I n:T . Stadler, J . Mollion, M -C Verdus, Y. Karamanos, H. Morvan,and D. Chri stiaen (eds.), A lgal biotechnology. Elsevier Ap-plied Science, L ondon.Emerson, R. , Lewis, C. M. 1943. A m. Bot. 30: 165.Faulkner, D. J . 1977. Tetrahedron 33: 1421.Faulkner, D. J . 1978. Top. A ntibiot. Chem. 2: 9.Faulkner, D. . 1986. Nat. Prod. Rep., 3: 1.Glombitza, K .W., K och, M . 1989. p. 161 In: R . C. Cresswell,T.A.V. Rees, and H . Shah (eds.), A lgal and cyanobacterial bio-technology. L ongman Scientif ic &Technical, Harlow, England.Goldman, J . C. 1979. Wat. Res. 13: 119.Hill, R., Bendall, F. 1960. Nature 186: 136.K ok, B. 1960. p. 566 In: W. Ruhland (ed.), Encyclopedia ofplant physiology. Springer, Berlin.M anney,T . R ., Duntze, W., Betz, R. 1981.p. 21 In: D. H .ODayand P. A . Horgen (eds.), Sexual interactions in eukaryoti cmicrobes. A cademic, New Y ork.Myers, J . 1980. p. 1 In: P. G. Falkowski (ed.), Primary produc-tivi ty in the sea. Plenum, New Y ork.Myers, J . 1953. I n: J . S. Burlew (ed.), Algal culture from labo-ratory to pilot plant. Carnegie Institution of Washington,Washington, DC .Oswald, W. J . 1988. p. 357 In: M .A . Borowitzka and L . .Borowtizka (eds.), M icroalgal biotechnology. Cambridge Uni-versity Press, Cambridge.Paul, J . H . 1982. Biochem. J . 203: 109.Pirt, S. . 1986. New Phytol. 102: 3.Pirt, S. J ., L ee, Y -K . R ichmond, A ,, Watts-Pirt, M . 1980.J . Chem. T echno]. Biotechnol. 30: 25.Pratt, R ., Daniels, T. C., E iler, J . B., Stain, H. H . 1944. Science99: 351.Ri chmond, A . 1986. p. 199 In: A . R ichmond (ed.), Handbookof algal mass culture. CRC Press, Boca Raton, FI.Shinho, K . 1986. J pn. Kokai Tokyo K oho, J P 61 69.728(8669.728) 105: 30034t.Tandeau de Marsac, N., de la Torre, F ., Szulmajster, J . 1987.Mol. Gen. Gent. 209: 396.V onshak, A . 1986.p. 117 In: A . Richmond (ed.), Handbook ofmicroalgal mass culture. CR C Press, Boca Raton, FL .

J AVANMARDIAN AND PALSSON: HIGH-DENSITY PHOTOAUTOTROPHIC ALGAL CULTURES 1189

Documents

High-Density Photoautotrophic Algal