You may notice some images loading slow across the Open Collections website. Thank you for your patience as we rebuild the cache to make images load faster.

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Growth optimization of Synechococcus elongatus PCC 7942 in lab flask and 2D photobioreactor Kuan, David 2013

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata

Download

Media
24-ubc_2013_fall_kuan_david.pdf [ 2.64MB ]
Metadata
JSON: 24-1.0074247.json
JSON-LD: 24-1.0074247-ld.json
RDF/XML (Pretty): 24-1.0074247-rdf.xml
RDF/JSON: 24-1.0074247-rdf.json
Turtle: 24-1.0074247-turtle.txt
N-Triples: 24-1.0074247-rdf-ntriples.txt
Original Record: 24-1.0074247-source.json
Full Text
24-1.0074247-fulltext.txt
Citation
24-1.0074247.ris

Full Text

     GROWTH OPTIMIZATION OF Synechococcus elongatus PCC 7942 IN LAB FLASK AND 2D PHOTOBIOREACTOR  by DAVID KUAN B.A.Sc., The University of British Columbia, 2010  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  MASTER OF APPLIED SCIENCE  in  The Faculty of Graduate and Postdoctoral Studies (Chemical and Biological Engineering)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  August 2013    ? David Kuan, 2013  ii  Abstract One of the most promising mechanisms for the production of high value biologically active products is through the cultivation of microalgae. In addition to serving as a carbon capture system, this photosynthetic microorganism has demonstrated potential for recombinant protein expression as an approach towards sustainable development in biotechnology. Extensive studies on cyanobacterium Synechococcus elongatus PCC 7942 have assessed the dual function of carbon capture with product generation such as biodiesel and recombinant protein. In order to maximize CO2 fixation and production rates of valuable product, a high cell growth rate needs to be achieved. Consequently, challenges in photobioreactor operation and cultivation need to be addressed, such as CO2 mass transfer limitations, light availability, and minimizing energy consumption. Thus, the effects of the major growth factors need to be studied. In this research, the objectives were to optimize the specific growth rate and biomass concentration of S. elongatus by investigating the effects of medium composition, light intensity, temperature, and CO2 concentration. Preliminary studies at the shake flask scale revealed that an optimization of components in BG-11 medium resulted in no significant improvements for the specific growth rate and biomass concentration. However, a maximum specific growth rate of 0.0519 h-1 and a maximum biomass concentration of 0.496 g/L were achieved at 33?C and 120 ?E/m2/s.  A 1 L airlift photobioreactor was used to investigate the effects of light intensity, CO2 concentration, and gas flow rate on the specific growth rate and biomass concentration. Additional experiments carried out in this photobioreactor revealed that air enriched with 5% CO2 at 1 L/min, 33?C, and 120 ?E/m2/s achieved a maximum biomass concentration of 1.006 g/L at a reduced specific growth rate of 0.0234 h-1. Further increases in CO2 % and light intensity, as well as light/dark cycles, reduced the growth rate and biomass concentration. Mass transfer experiments also revealed that 5% CO2 provided the best growth conditions, as growth was significantly limited by CO2 when supplied with air, whereas 10% CO2 reduced the pH and consequently reduced the specific growth rate. iii  Preface This research is part of a collaboration with Dr. S. Yewalkar of the University of British Columbia and Dr. F. Nano and his research team of the University of Victoria. The microalgae strains were isolated by Dr. Nano. The experiments and data analysis in Chapter 4 were performed by Dr. Yewalkar and me. The experiments and data analysis in Chapters 5 and 6 are my original work.   iv  Table of Contents Abstract ........................................................................................................................................... ii Preface............................................................................................................................................ iii List of Tables ................................................................................................................................ vii List of Figures ................................................................................................................................ ix List of Symbols and Abbreviations................................................................................................ xi Acknowledgements ...................................................................................................................... xiii Dedication .................................................................................................................................... xiv 1    Introduction ............................................................................................................................... 1 1.1    The need for economical CO2 capture ............................................................................... 1 1.2    Microalgae as a direct CO2 mitigation technology ............................................................ 2 1.2.1    Microalgal productivity ............................................................................................... 4 1.2.2    Current industrial applications using microalgae ........................................................ 6 1.2.3    Economics of microalgae cultivation .......................................................................... 8 1.3    Microalgae as a platform for production of high value bioproducts .................................. 9 1.3.1    Advantages of microalgal protein production ............................................................. 9 1.3.2    Potential applications of genetically-modified microalgae ....................................... 11 1.3.3    Challenges of using genetically modified algae ........................................................ 12 1.4    Synechococcus elongatus PCC 7942 ............................................................................... 12 1.5    Closed photobioreactor design ......................................................................................... 13 1.5.1    Design of airlift reactors ............................................................................................ 15 1.6    Optimizing growth of microalgae .................................................................................... 17 1.6.1    Overview of media for microalgae ............................................................................ 17 1.6.2    Medium and pH buffer selection ............................................................................... 17 1.6.3    Effects of temperature on growth .............................................................................. 18 1.6.4    Gas-liquid mass transfer ............................................................................................ 18 1.6.5    Effects of light intensity on growth ........................................................................... 20 1.6.5.1    Photosynthesis ........................................................................................................ 20 1.6.5.2    Light Distribution ................................................................................................... 21 1.6.5.3    Light/dark cycles .................................................................................................... 24 1.6.6    Effects of O2 and CO2 on growth .............................................................................. 24 v  1.6.7    Effects of agitation on growth ................................................................................... 25 1.7    Previous Research on Optimization of Growth of Microalgae ........................................ 26 1.8    Modelling microalgae growth .......................................................................................... 26 1.8.1    Cell growth models .................................................................................................... 26 1.8.2    Logistic growth model ............................................................................................... 29 2    Research Objectives ................................................................................................................ 30 3    Materials and Methods ............................................................................................................ 31 3.1    Synechococcus elongatus PCC 7942................................................................................ 31 3.1.1    Subculturing of S. elongatus in BG-11 liquid medium ............................................. 31 3.1.2    Preparation of BG-11 liquid medium ........................................................................ 32 3.1.3    Preparation of BG-11 agar plates .............................................................................. 32 3.1.4    Sterilization ................................................................................................................ 33 3.2    Measurement of biomass concentration ........................................................................... 34 3.3    Logistic growth model and carbon uptake rate ................................................................ 35 3.3    Optimization of medium composition.............................................................................. 36 3.3.1    Defining conditions for screening medium composition .......................................... 36 3.3.2    Defining conditions for optimization of medium composition ................................. 37 3.4    Experimental setup of photobioreactor ............................................................................ 37 3.4.1    Calculation of the concentrations for each carbon species ........................................ 39 3.4.2    Specifications of photobioreactor .............................................................................. 40 3.5    Hydrodynamic parameters of the photobioreactor ........................................................... 42 3.5.1    Procedure for determining mixing time ..................................................................... 42 3.5.2    Procedure for determining the Bodenstein number ................................................... 43 3.5.3    Procedure for calculating gas hold-up ....................................................................... 44 3.5.4    Procedure for determining volumetric mass transfer coefficient .............................. 44 3.6    Procedure for optimization of ?max and Xmax in shake flasks ............................................ 46 3.7    Procedure for optimization of ?max and Xmax in photobioreactor ...................................... 46 4    Results for Shake Flask Scale Experiments ............................................................................ 48 4.1.1    Screening of BG-11 medium components ................................................................. 48 4.1.2    Optimization of BG-11 component concentrations ................................................... 52 5    Mass Transfer Studies in 2D Airlift Photobioreactor ............................................................. 62 vi  5.1    Determination of Bodenstein number and gas hold-up in photobioreactor ..................... 62 5.2    Aeration by CO2-enriched air for mass transfer characterization in 2D photobioreactor 63 6    Optimization of Xmax and ?max in photobioreactor .................................................................. 71 6.1    Effects of light intensity on growth .................................................................................. 75 6.2    Effects of inlet CO2 concentration on growth .................................................................. 76 6.3    Effect of 12:12 h light/dark cycle on S. elongatus growth ............................................... 79 7    Conclusion .............................................................................................................................. 82 8    Future Work ............................................................................................................................ 85 Appendices .................................................................................................................................... 96 Appendix A:    Sample calculations .......................................................................................... 96 Appendix B:     Bubble size in photobioreactor (33?C) ............................................................. 98 Appendix C:    S. elongatus growth at various conditions in 2D photobioreactor .................... 99 Appendix D:    Polymath program for calculating carbon content ......................................... 100   vii  List of Tables Table 1: General characteristics of open pond and photobioreactor systems (National Research Council, 2012)................................................................................................................................. 6 Table 2: A selection of microalgal species and their products and application areas (Pulz & Gross, 2004) .................................................................................................................................... 7 Table 3: Comparison of features of recombinant protein production from various bioreactor systems (Walker et al., 2005) ........................................................................................................ 10 Table 4: Potential applications of genetically modified microalgae (Enzing & Nooijen, 2012) .. 11 Table 5: Previous studies of exogenous protein expression in S. elongatus PCC 7942 ............... 13 Table 6: BG-11 composition ......................................................................................................... 17 Table 7: BG-11 medium stock solutions (UTEX, 2009) .............................................................. 32 Table 8: Experimental design for screening of medium components........................................... 36 Table 9: Central composite design for optimization of BG-11 medium components .................. 37 Table 10: Central composite design for optimization of ?max and Xmax using two factors ............ 46 Table 11: Experimental design for optimization of ?max and Xmax using two factors in photobioreactor ............................................................................................................................. 47 Table 12: Full factorial design of BG-11 components.................................................................. 48 Table 13: Results from the BG-11 medium factorial design ........................................................ 50 Table 14: Effects and coefficients of variables estimated using a screening design for response ?max ................................................................................................................................................ 51 Table 15: Effects and coefficients of variables estimated using a screening design for response Xmax ................................................................................................................................................ 51 Table 16: Values of independent variables in different levels of the optimization design ........... 52 Table 17: Experimental results from optimization of BG-11 component concentrations ............ 53 Table 18: Optimization of BG-11 component concentrations for response Xmax (R2=0.91) ........ 54 Table 19: Optimization of BG-11 component concentrations for response ?max (R2=0.89) ......... 55 Table 20: Components of the original and modified media ......................................................... 56 Table 21: Summary of Xmax and ?max ............................................................................................ 57 Table 22: Results from the optimization of light intensity and temperature ................................ 58 Table 23: Optimization of light intensity and temperature for response ?max ............................... 58 viii  Table 24: Optimization of light intensity and temperature for response Xmax .............................. 60 Table 25: Bo and tm ....................................................................................................................... 62 Table 26: Saturation concentration for CO2 at 23?C and 33?C and a constant gas flow rate of 0.5 L/min ............................................................................................................................................. 67 Table 27: Comparison of kLa values at different inlet gas flow rates and CO2 %. T=33?C ......... 69 Table 28: Factorial design of reactor conditions........................................................................... 71 Table 29: Results of full factorial design in airlift photobioreactor .............................................. 72 Table 30: Optimization of light intensity and CO2 % for response ?max in photobioreactor (R2=0.85) ....................................................................................................................................... 72 Table 31: Optimization of light intensity and CO2 % for response Xmax in photobioreactor (R2=0.44) ....................................................................................................................................... 74 Table 32: Reproducibility of growth rate data from the photobioreactor ..................................... 76 Table 33: Results for growth under continuous light and 12:12 light/dark cycle ......................... 80 Table 34: Energy input required for growth at continuous light and 12:12 light/dark cycle ........ 81 ix  List of Figures Figure 1: GHG emissions by economic sector (Mt CO2e) (Environment Canada, 2012a) ............ 2 Figure 2: Pond Biofuels CO2 capture process (Pond Biofuels, 2011) ............................................ 3 Figure 3: Open pond cultivation (Seambiotic Ltd., 2010) .............................................................. 5 Figure 4: Airlift reactors (a) split-cylinder internal loop; (b) draft tube internal loop; (c) external loop (Kilonzo & Margaritis, 2004) ............................................................................................... 14 Figure 5: Flat-plate airlift photobioreactor ................................................................................... 16 Figure 6: Photosynthesis (Mayer, 2008) ....................................................................................... 21 Figure 7: Light intensity changes with distance in absorbing medium ........................................ 22 Figure 8: Light penetration depth of Chlorella kessleri as a function of light intensity (W/cm2) and cell concentration (cell number/mL) ...................................................................................... 23 Figure 9: Typical growth curve for batch cultivation ................................................................... 27 Figure 10: Incubation of subcultures ............................................................................................ 31 Figure 11: Petri dish of S. elongatus ............................................................................................. 33 Figure 12: Determined relationship between dry cell concentration and optical density ............. 34 Figure 13: Example of a logistic growth curve ............................................................................. 35 Figure 14: Photobioreactor setup .................................................................................................. 38 Figure 15: Dissolved CO2 electrode calibration ........................................................................... 38 Figure 16: CO2 species and pH (Haynes et al., 2012) .................................................................. 39 Figure 17: Reactor dimensions ..................................................................................................... 41 Figure 18: Front view of photobioreactor ..................................................................................... 42 Figure 19: Measurement of bubble diameter with ImageJ ........................................................... 45 Figure 20: Photobioreactor and light panel setup ......................................................................... 47 Figure 21: Growth curve and fitted logistic model for estimating ?max ........................................ 49 Figure 22: Growth in optimized media. Lines added to visualize data. ....................................... 56 Figure 23: Response curve for ?max as a function of temperature and light intensity (R2=0.83) .. 59 Figure 24: Response curve for Xmax as a function of temperature and light intensity (R2=0.85).. 61 Figure 25: Relationship between gas holdup and gas flow rate .................................................... 63 Figure 26: DCO2 and DO2 at 33?C, constant gas flow rate of 0.5 L/min, and at 5% CO2 .............. 65 Figure 27: CO2 kLa at 23?C and 33?C for a constant gas flow rate of 0.5 L/min ......................... 65 x  Figure 28: CO2 kLa at various gas flow rates and a constant temperature of 33?C ...................... 66 Figure 29: Oxygen kLa at various gas flow rates and a constant temperature of 33?C ................. 67 Figure 30: Saturation concentration of CO2 at various gas flow rates and a constant temperature of 33?C .......................................................................................................................................... 68 Figure 31: Saturation concentration of O2 at various gas flow rates and a constant temperature of 33?C .............................................................................................................................................. 68 Figure 32: Superficial gas velocities and gas hold-ups at various flow rates ............................... 69 Figure 33: Response curve for ?max as a function of light intensity and CO2 % (R2=0.85) .......... 73 Figure 34: Culture pH at various conditions in the photobioreactor at 33?C and 1 L/min gas flow rate................................................................................................................................................. 74 Figure 35: Xmax for 0.04, 5 and 10% CO2, and 60 and 120 ?E/m2/s light intensity ...................... 75 Figure 36: Specific growth rate ?max for 0.04, 5 and 10% CO2, and 60 and 120 ?E/m2/s light intensity ......................................................................................................................................... 75 Figure 37: Carbon balance from S. elongatus growth in air at 1 L/min, 33?C, 60 ?E/m2/s ......... 77 Figure 38: Carbon balance from S. elongatus growth in 5% CO2 at 1 L/min, 33?C, 60 ?E/m2/s 78 Figure 39: Carbon balance from S. elongatus growth in 10% CO2 at 1 L/min, 33?C, 60 ?E/m2/s....................................................................................................................................................... 79 Figure 40: Comparison of growth in continuous light and 12:12 light/dark cycle at 33?C, 60 ?E/m2/s, 5% @1 L/min ................................................................................................................. 80    xi  List of Symbols and Abbreviations Symbol Units Description  Bo  Bodenstein number CS g/L Saturation concentration DCO2 mg/L Dissolved CO2 DO2 mg/L Dissolved O2 kLa h-1 Volumetric mass transfer coefficient ? h-1 Specific growth rate ? g/m3 Density Re  Reynolds number Sh  Sherwood number Sc  Schmidt number  v m/s Flow velocity V L Volume X g/L Biomass concentration   xii  Abbreviation Description  CTR  Carbon dioxide transfer rate CUR  Carbon uptake rate GHG  Greenhouse gas emissions ILI  Incident light intensity PBR  Photobioreactor vvm  Gas volume flow per unit of liquid volume per minute        xiii  Acknowledgements First, a thank you for the funding received from the National Science and Engineering Research Council (NSERC), without which I wouldn?t have had this opportunity. Next, I would like to give my most sincere thank you to my supervisors Dr. Sheldon Duff, Dr. Dusko Posarac, and Dr. Xiaotao Bi for their continued guidance, encouragement, and mentorship. You?ve helped me to really grow as an engineer and an individual and for that I am truly grateful. I would also like to thank Dr. Swati Yewalkar for providing her expertise and assistance in conducting this research, as well as Dr. Francis Nano and his research team for providing the microalgae strains. Also thanks to my lab mates for helping to create such an enjoyable environment to work in. Finally, I would like to thank my friends, Mom and Dad, and sister for their unwavering support they provided throughout my journey.   xiv  Dedication                 To my family1  1    Introduction 1.1    The need for economical CO2 capture Increasing concerns of the effects of climate change has elicited a response for better solutions to mitigate greenhouse gas (GHG) emissions. Humans are accelerating the accumulation of GHGs through combustion, waste gas emissions, and poor land use practices. Efforts to reduce the impacts of climate change are making significant progress towards the goal of reducing Canada?s greenhouse gas emissions by 17% from 2005 levels by 2020. As of August 2012, 2020 emissions are projected to be one half of the way to the target (Environment Canada, 2012a). In the past decade, the Canadian government has taken a robust approach in reducing these emissions to tackle climate change (Environment Canada, 2012b). The most prevalent of GHGs is CO2, which accounted for 79% of total Canadian GHG emissions (Environment Canada, 2012a). GHG mitigation generally involves reductions in anthropogenic emissions, increasing the capacity of carbon sinks (ex. plants and algae), and the use of renewable energy (e.g. wind power). Commercial viability of carbon capture technologies has been a challenge as these methods have not been shown to be economical. This can be attributed to high operating and disposal costs. The 2009 federal budget has allocated $1 billion over 5 years towards the development of new technologies and is expected to generate at least $2.5 billion in clean energy investment (Flaherty, 2009). Much of this funding has been dedicated toward large-scale carbon capture and storage facilities primarily concerning the electricity and oil/gas sectors. As shown in Figure 1, both these sectors combined to produce 253 million tons of carbon dioxide equivalents, or 36.6% of Canada?s total GHG emissions (Environment Canada, 2012c). The need to reduce industrial GHG emissions is imperative.  2   Figure 1: GHG emissions by economic sector (Mt CO2e) (Environment Canada, 2012a) 1.2    Microalgae as a direct CO2 mitigation technology Recent developments on green technologies have had success in mitigating climate change. Three conceptually different modes of atmospheric CO2 reduction include reducing fossil fuel consumption, removing CO2 from the atmosphere, and capturing CO2 from emissions before it is released into the atmosphere (Benemann, 1997). An example of the latter mode is the cultivation of microalgae in open ponds to deal with point source emissions. This has been attractive for its potential to fix CO2 directly from the atmosphere or flue gases using photosynthesis, and convert it to biomass. This biomass can then be converted into bioproducts or biofuels. What sets microalgae apart from other microorganisms is that they are capable of taking in CO2 and light energy. They are abundant in almost every habitat because of their tolerance to extreme environmental stresses. However, to date, direct mitigation methods with microalgae are generally associated with higher capital costs and the use of greater amount of resources. It is estimated to cost between $50-$250/ton of carbon to remove CO2 from a conventional power plant using algal technology, when economically feasible sequestration costs should be about $10/ton of carbon (PowerplantCCS, 2010). Estimates only consider the carbon fixed in biomass. The cost would be higher if carbon losses and costs from electricity, transportation, etc. were 22.3%14.3%24.0%10.8%11.4%10.0%7.2%Oil and GasElectricityTransportationEmissions Intensive & Trade Exposed IndustriesBuildingsAgricultureWaste and Others3  included. Open ponds require land, water, and appropriate climatic conditions, making microalgae a less favourable option (J. Lee & Lee, 2003). Toxic gases present in flue gas such as SOx and NOx pose a problem so a pretreatment step is required to minimize their inhibitory effects before introducing CO2 to the reactors (J. N. Lee et al., 2000). In other words, cultivating algae strictly for commercialization of biomass/biofuel to offset carbon capture costs is not economical. Rather, Benemann (1997) argued that microalgae are better used for specific biological technologies to produce high value products while simultaneously capturing CO2, which requires the use of photobioreactors. If utilization of CO2 into high-value compounds is possible, then CO2 removal costs can be minimized. This has been the driving force in the research on the conversion of CO2 into useful algae-based products. It has been shown that the application of microalgae for carbon capture can be economical by genetically modifying them to produce therapeutic proteins and other biological products (Specht et al., 2010).  One development that has come to fruition is the application of microalgae to capture CO2 from a cement plant in Bowmanville, Ontario (Pond Biofuels, 2011). Pond Biofuels has become the first company to successfully use CO2 emissions to produce biomass by using microalgae as a carbon sink, as shown in Figure 2.  Figure 2: Pond Biofuels CO2 capture process (Pond Biofuels, 2011) The biomass can then be dried from industrial waste heat and used as fuel, or they can be used as animal feed. Emissions can be reduced by as much as 40% (Hodge, 2009). Of course, 4  combustion of algal biomass will release the captured CO2; however, useful energy is generated as a result of the process. Although this only demonstrated the application of microalgae for production of low-value products, it opens the door for potential commercialization of high-value microalgal products from carbon capture applications. Besides the additional ability to sequester carbon dioxide from the atmosphere, microalgae can produce a wide range of primary and secondary metabolites, such as proteins, lipids, vitamins, and other biologically active compounds (J. Kim & Lee, 2005). The prospect of using recombinant microalgae to develop high value products has only been a recent development; nonetheless pharmaceutical companies and biochemical producers alike have made strides in its research. The potential of using algae as a source of novel compounds has been reviewed by several researchers (Franklin & Mayfield, 2005; Walker et al., 2005). Of particular concern is determining the best conditions for microalgae growth to make the process economical. Optimizing algal growth rate can improve the production rate of the target metabolite. Cell growth rate depends on light intensity, nutrient type and concentration, temperature, and dissolved CO2 concentration within the reactor (Sasi, 2009). With this in perspective, the focus of my research is to cultivate an engineered Synechococcus elongatus PCC 7942 strain as a vehicle for recombinant protein production in a bubble column and assess economic feasibility for industrial scale production. This provides the groundwork for further expansion of therapeutic and industrial products in the microalgae field. 1.2.1    Microalgal productivity Biomass productivity is defined as the rate of production of algae biomass per unit volume, and it is highly dependent on the culture technique. Selection of a suitable cultivation method is crucial because it has a profound influence on productivity and purity and it depends on the purpose of the production facility. Biomass productivity is the production of algal biomass concentration per unit time for a given volume. Purity can indicate the contamination level or the percentage of recovered product. As a commonly used configuration for cultivation, open pond cultures are often large scale systems with nutrients provided by runoff water or effluent from wastewater treatment plants. Figure 3 illustrates a schematic of an open pond system (Seambiotic Ltd., 2010). Water is agitated by paddle wheels or similar rotating structures. The culture often consists of a mix of algal strains. This technology typically yields productivities of 14-50 g?L-5  1?day-1, making this a popular design for commercial production of biofuels. However, environmental conditions are difficult to control as these ponds are located outdoors, and this makes them susceptible to contamination and therefore lower purity.  Figure 3: Open pond cultivation (Seambiotic Ltd., 2010) Small scale production, such as in closed photobioreactors (PBR), allows for better control of growth conditions and sterility. Currently, closed PBRs are the most popular system for conducting research in optimizing algae growth and to make high-value products such as enzymes and food supplements. Closed PBRs provide a sterile environment that enables the growth of a single or several specific species to be optimized for producing a desired product. Water, nutrients, and CO2 are supplied in a controlled manner. Internal PBR conditions such as pH, light intensity and temperature are controlled and monitored. They can be illuminated by artificial light or natural sunlight. Much work has been carried out to optimize these systems (Chisti & Moo-Young, 1993; Kajiwara et al., 1997; J. Kim & Lee, 2005). Although the potential for microalgae to achieve high productivities was based on projections of limited data from short term and laboratory experiments, they still possess features that can make them a valuable commodity in the biotechnology industry. In the laboratory, photosynthetic efficiency, which is the fraction of light energy that is converted into chemical energy, is usually at 20-24% (Benemann, 1997). Commercial microalgae cultivations are mostly done in open ponds where cells are exposed to sunlight. However, cell productivity does not increase in proportion to the 6  photosynthetic photon flux density. Table 1 compares the operating requirements for the two methods of algae cultivation. Table 1: General characteristics of open pond and photobioreactor systems (National Research Council, 2012) Open pond system Photobioreactor Energy source Natural sunlight Focused sunlight/artificial light CO2 source Atmosphere or introduced CO2 CO2 from carbon capture system Nutrients From waste water or introduced Accurate dosing of nutrients Capital cost Low cost for simple engineering High cost for sophisticated system Economics Generic analysis for energy-only schemes is not promising High value co-products could give viability to niche markets Electricity consumption Modest power demand High power demand Algal product quality Variable quality algae product with risk of contamination High quality monoculture and secreted co-products possible Potential products Only suitable as a fuel precursor Potential for high value products Water use and loss High water demand and high potential loss of evaporation High water demand for solution management. Low evaporation loss Land area Large Small 1.2.2    Current industrial applications using microalgae For the past few decades, mass culturing microalgae has been a widely-used process for wastewater and water pollution treatments, biofuels production, pharmaceutical bioproducts and carbon sequestration (Benemann & Oswald, 1996). From a commercial perspective, microalgae have been used to produce both biofuels and high-value compounds (Table 2). In waste treatment ponds, microalgae provide oxygen to bacteria, which, in turn, break down biodegradable waste sludge (Alabi et al., 2009). Consequently, useful nutrients like CO2 and phosphates are released and this promotes algae growth, turning waste into algal biomass. It also addresses the problem of eutrophication in watersheds. Kuehnle AgroSystems in Hawaii has implemented such a process at an oil refinery by removing CO2 from wastewater (Higuchi, 2012). Kuehnle?s system produced Chlorella by piping CO2 and wastewater from Chevron?s Hawaii refinery into tubular photobioreactors to accelerate algae growth. The algae were then converted into biofuel, which was used to reduce the refinery?s greenhouse gas emissions. In November 2011, the project was the first successful algae production process using industrial CO2 from an oil refinery. 7  As discussed previously, microalgae can also be used to fix CO2 from flue gas by directing waste emissions into photobioreactors and the biomass generated can be harvested for biofuel production. In Israel, flue gas from a coal-fired power station was directed to a series of open algae ponds to produce dry algae biomass for biofuels (Ben-Amotz, 2008). The 1000 m2 pilot project cultivated Nannochloropsis sp, and it had a maximum algae biomass yield of 25 kg/day. The most successful application of microalgae to produce high-value compounds in a commercial scale has been the production of astaxanthin by Cyanotech using Haematococcus microalgae. This health supplement has been cultivated in shallow ponds installed with a paddlewheel to provide circulation i.e. raceway ponds (Guerin et al., 2003). The 40 hectare raceway ponds are located in Hawaii, where the low rainfall and year-round sunlight are most suitable for microalgae.  The commercial exploitation of non-transgenic microalgae as a nutritional supplement is well-developed. Other high-value products such as food additives, pigments, enzymes and anti-oxidants derived from algae exist but are extremely limited in the market. Polyhydroxyalkanoate (PHA), a biodegradable polyester produced by diatoms, is being commercialized for the plastics industry (Loo & Sudesh, 2007). The versatility of microalgae is advantageous across many applications. More examples of current algae-related applications are listed in Table 2. Table 2: A selection of microalgal species and their products and application areas (Pulz & Gross, 2004) Species/group Product Application areas Basins/reactors Spirulina platensis/Cyanobacteria Phycocyanin, biomass Health food, cosmetics Open ponds, natural lakes Chlorella vulgaris/Chlorophyta Biomass Health food, food supplement, feed surrogates Open ponds, basins, glass-tube PBR Dunaliella salina/Chlorophyta Carotenoids, ?-carotene Health food, food supplement, feed Open ponds, lagoons Haematococcus pluvialis/Chlorophyta Carotenoids, astaxanthin Health food, pharmaceuticals, feed additives Open ponds, PBR Odontella aurita/Bacillariophyta Fatty acids Pharmaceuticals, cosmetics, baby food Open ponds Porphyridium cruentum/Rhodophyta Polysaccharides Pharmaceuticals, cosmetics, nutrition Tubular PBR Isochrysis galbana/Chlorophyta Fatty acids Animal nutrition Open ponds Phaedactylum tricornutum/Bacillariohyta Lipids, fatty acids Nutrition, fuel production Open ponds, basins 8  1.2.3    Economics of microalgae cultivation In the current market, the use of algae to mitigate CO2 with biofuel production as the sole income stream is not economically feasible due to competitive crude oil prices of $93/barrel (as of January 2013). As a comparison, algal biofuels have been estimated to cost $240/barrel (includes CO2 mitigation credits) (Lundquist et al, 2010). The biggest economic hurdle in commercialization of a closed microalgae cultivation system is the high capital cost (Alabi et al., 2009). Benemann (1997) reported costs of over $100/tonne for PBRs in Spain, Germany and Israel, as opposed to costs of around $10/tonne for open ponds. In PBRs, gas and liquid handling systems will be present, with spargers, valves and pumps. Biomass yields determine the frequency of harvesting and extraction. Many microalgal strains can be grown in defined media for the purpose of maintaining high growth rates and the purity of cultures (Andersen, 2005). Some strains such as Dunaliella flourish in undefined media like wastewater, which is a cheap source of inorganic nutrients. However, restrictions may apply in a PBR because pre-treatment may be necessary to sterilize wastewater and meet other downstream process conditions. Medium costs are strain-dependent. Carbon can be supplied by industrial-grade CO2 in PBRs or flue gas for any production scale. But other gases (sulphur oxides, nitrogen oxides, oxygen, etc.) must be considered as they have significant negative effects on the growth rate (Doucha et al., 2005). SO2 can lower the pH to 4 if the concentration reaches 400 ppm, but this is commonly mitigated by adding NaOH (Oilgae, 2009). Nitrogen oxides can also affect the pH of the medium but to a lesser degree. Soot dust and heavy metals were found to have very little impact on algae growth. Separation of biomass from the product stream in PBRs costs less than in open ponds due to higher biomass densities. Harvesting costs come from the electricity used for centrifugation and can make up 20-30% of the total cost of algal biomass (Molina et al., 2003).  Energy is also required for agitation of medium, delivery of gas, illumination, and heating. In some cases, cooling might be necessary for warmer climates. Over time, cells may adhere to the surface of the PBR, so this requires periodic cleaning and shutdown. Replacement of culture is also mandatory to maintain high optimal yields (Andersen, 2005). Other costs include production 9  of inoculum, and algae subculturing. It appears that, at the current economic state, production of high-value products is most profitable amongst all microalgae applications. 1.3    Microalgae as a platform for production of high value bioproducts Select strains of microalgae can naturally produce high-value compounds; they can produce complex molecules which are otherwise difficult to produce by chemical synthesis (Borowitzka, 1995). These include antibodies (Schillberg et al., 2003), polysaccharides (Walker et al., 2005), and carotenoids (Pulz & Gross, 2004). They are an important source of chemical products that are applied in the feed, food, nutrition, and pharmaceutical industries. What makes microalgae even more attractive in addition to their ability to capture CO2 is that they can be genetically modified to produce recombinant products that have important value in the chemical and pharmaceutical applications. This opens up the possibility of producing a greater variety of valuable molecules that was not technologically feasible before. In the past three decades there has been significant focus on selection of algae strains for pharmacologically-active compounds. These microorganisms are now being looked upon as genetically-modified hosts for production of biopharmaceuticals, polymers, vaccines, enzymes, and antibodies.  Perhaps most importantly, algae are now recognized for their potential to reduce material costs compared to conventional biochemical production plants, and for the potential for scale-up. The desirable solution to carbon capture is to combine the benefits of carbon fixation in algae with the production of high-value compounds. 1.3.1    Advantages of microalgal protein production Microalgae strains are generally simpler to cultivate commercially compared to macroalgae. High-value products are defined as any product when multiplied by its fraction in biomass has a value of over US$ 10/kg (Carlsson et al., 2007). In the algae commercialization industry, the products classified under this definition have been limited because most have been in the early stages of development. Producing therapeutic proteins in microalgae has many economic and qualitative benefits, such as reduced health risks from pathogen contamination and relatively high yields. Furthermore, existing infrastructure in production facilities can be used to cultivate, harvest, store, and process transgenic organisms and require relatively little capital investment. It 10  has been estimated that the cost of producing recombinant proteins in algae could be 10- to 50-fold lower than producing the same protein by E. coli fermentation, depending on the host transgenic plant or microalga (Kusnadi et al., 1997). Microalgae-derived products were estimated to have a global market value of US$ 5-6.5 billion in 2004 (Carlsson et al., 2007). US$ 1.25-2.5 billion were generated by the health food sector, US$ 1.5 billion from the production of docosahexanoic acid (DHA) and US$ 700 million from aquaculture. The United States is responsible for 50% of the global algae production. Algae production potential is greatest in regions where there is an abundance of land, water and sunlight; however, regions with these resources must also have the infrastructure to develop these production systems, limiting the number of locations suitable for cultivation. Although in its infancy, the use of transgenic microalgal systems as bioreactors for production of therapeutic and industrial organic compounds is promising. It is advantageous to use unicellular algae compared to higher plants as they have faster growth rates and can be grown at high densities in controlled reactors. A comparison of the various bioreactor systems currently available are listed in Table 3.  Table 3: Comparison of features of recombinant protein production from various bioreactor systems (Walker et al., 2005) Features Bacteria Yeast Mammalian cell culture Transgenic plants Transgenic microalgae Production time Short Medium Long Long Short Production cost Medium Medium High Low Very low Reactor scale Large Large Medium Small Small Scale up cost High High High Low Very low Cost Cheap Cheap Expensive Cheap Cheap Production scale Limited Limited Limited Worldwide Worldwide Propagation Easy Easy Hard Easy Very easy Distribution Feasible Feasible Difficult Easy Easy Delivery vehicle No No No Possible Possible Gene size Unknown Unknown Limited Not limited Not limited Glycosylation Absent Incorrect Correct Correct Correct Protein yield Medium High Medium-high High Unknown Risk Yes Unknown Yes Unknown Unknown Safety to human Low Unknown Medium Medium High  11  Theoretically, recombinant microalgae can potentially produce proteins at a fraction of the cost of conventional cell culture systems and higher plants. Production of algal biomass is estimated to cost $3/kg at commercial scale (Chisti, 2007). Approximately 25% of biomass is soluble protein, so if recombinant protein makes up 2% of soluble protein, facilities can presently produce recombinant protein at about $600/kg prior to purification. This is close to cost estimates for the least expensive protein expression systems in the current market (Rasala et al., 2010). Plant-based proteins are less likely to be contaminated with pathogens than those derived from animal cells because plants are not susceptible to human pathogens (Specht et al., 2010). Microalgae cells can also fold complex human proteins that bacteria and yeast may not fold well.  1.3.2    Potential applications of genetically-modified microalgae  There have been a few recent developments in the establishment of genetically-modified microalgae for production of high-value products. The most notable breakthrough for expression of human genes is the unicellular algae Chlamydomonas reinhardtii. It has been widely used in biology laboratories because its genome has been sequenced and has been cultured for many years. It is able to grow photosynthetically and heterotrophically, and can be cultured at a scale of up to 500,000 litres. From a genetic engineering perspective, the DNA of its chloroplasts, nucleus and other organelles can be easily altered. In studies by Tran (2012), C. reinhardtii was successfully modified to produce mammalian serum amyloid protein in 2007 and human antibody in 2010. In the most recent advancement, a malarial vaccine was developed (Dauvill?e & Delhaye, 2010). These achievements open up the opportunity to make complex drugs economically and in larger quantities. More examples of potential products are listed in Table 4. Table 4: Potential applications of genetically modified microalgae (Enzing & Nooijen, 2012) Species/group Product Application areas Reactor Cyanobacteria Antimicrobials, antivirals, antifungals Pharmaceuticals PBR Spirulina, Ulva conblogbata Neuroprotective agents Pharmaceuticals PBR Chlamydomonas reinhardtii Human therapeutic proteins Pharmaceuticals PBR Nostoc Cyanobionts Anti-cancer drugs Pharmaceuticals PBR    12  1.3.3    Challenges of using genetically modified algae Programs for genetic modification of algae strains have lagged compared to research efforts for plants and microbes. Few companies have had algal-based products reach or approach the stage of regulatory approval. Because of the uncertainty of the regulatory process involving novel technology, there is an inherent high risk that slows down the product development stage. Every compound goes through vigorous screening, testing, and approval stages before it reaches the market. Statistically, in the pharmaceutical industry, about 3 compounds out of 1000 make it past the first round of clinical trials and only one would be commercialised (Watkins, 2002). For example, Mera Pharmaceuticals submitted an application to begin large-scale production of modified alga C. reinhardtii, which produced human immunoglobulin-A protein and other pharmaceuticals (Dunwell & Ford, 2005). However, because of a lack of experience with engineered algae, a turn of events led to a judicial decision requiring an environmental assessment before pilot trials could be conducted. These trials have never taken place. There have also been increasing concerns of the modified algae being released into the environment, which can threaten native organisms. Secondly, the highly scrutinized stages of drug development can take 12 to 15 years, which is attributed to the long discovery stage and preclinical testing, trials involving thousands of patients, and the FDA review. Thirdly, the cost of developing a new drug using conventional methods is estimated to be US $800 million to $1.7 billion. Both the cost of development and FDA approval time have been rising over the past years, indicating the challenge of having genetically-modified algae within the purview of biotechnology firms.  1.4    Synechococcus elongatus PCC 7942 Cyanobacterium S. elongatus PCC 7942 was chosen as the species for this research because it has been extensively studied by the international research community, most notably its acclimation to nutrient stresses and adaptation to variations in temperature and light intensity (Joint Genome Institute, 2013). This photoautotrophic organism is found in freshwater habitats. Like land-based plants, it is photosynthetic, meaning that it can convert light energy into useful chemical energy using a complex biological process.  It grows best in BG-11 nutrient composition (I. S. Suh et al., 1998). In addition to physiological features, the genetic code of S. elongatus has been established and makes it an effective host for molecular genetics techniques 13  (Kajiwara et al., 1997). Numerous studies have been conducted on exogenous protein expression of this strain, as listed in Table 5 (Yamada et al., 1997). Table 5: Previous studies of exogenous protein expression in S. elongatus PCC 7942 Exogenous protein Source Carbonic anhydrase Human ?-galactosidase Escherichia coli Manganese superoxide dismutase Escherichia coli lacIQ repressor protein Escherichia coli ?-amylase Bacillus amyloliquefaciens A50 Entomocidal toxin Bacillus shaericus 1593M cIts repressor protein Bacteriophage lambda Fatty acid desaturase (desA) Synechocystis PCC6803 Plastocyanin Anabaena PCC7937 Phytoene desaturase Erwinia uredovora Ethylene-forming enzyme Pseudomonas syringae  1.5    Closed photobioreactor design Closed PBRs are ideal for the production of primary and secondary metabolites because they provide greater control of cell growth and metabolite production (I. S. Suh & Lee, 2001). Because closed PBRs are significantly smaller in size, it is necessary to maximize both productivity and photosynthetic efficiency. Various reactor configurations have been developed to address these factors. The three most common configurations of closed PBRs are tubular reactors, airlift reactors and bubble columns. A common feature among all closed photobioreactors is the high surface area to volume ratio for the purpose of increasing light exposure and biomass yields. Airlift reactors (Figure 4) are appealing because they require only a simple construction and can achieve circulation without moving parts (Chisti & Moo-Young, 1993). Gas-induced circulation of liquid is a key aspect of reactor design because it controls important parameters such as gas-liquid mass transfer, heat transfer and mixing. This setup also offers better O2/CO2 gas exchange compared to other configurations (I. S. Suh & Lee, 2001). Aeration rates must be chosen to balance a variety of factors. It has been shown that mixing of algal cultures has multiple effects on cell growth, namely shear forces on cell membranes and light penetration (Mir?n et al., 2000). As liquid circulation increases, cells are exposed to greater 14  mechanical forces from bubbles and reactor surfaces, causing cell lysis and foaming. Mixing also controls mass transfer of nutrients, and removal of potential harmful metabolites.  Figure 4: Airlift reactors (a) split-cylinder internal loop; (b) draft tube internal loop; (c) external loop (Kilonzo & Margaritis, 2004) At the same time, aeration rates cannot drop below a certain lower limit to ensure a well-mixed culture. Efficient mass transfer of CO2 is essential. Although aeration rate has an important influence on cell productivity, further consideration must be made on culture depth because bubbles can also affect light penetration (Mir?n et al., 1999). The presence of bubbles reduces internal irradiance because they can refract and reflect some light away from the reactor. Kobayashi has shown using Chlorella sp. cultures that growth rate and biomass concentration decreased with increasing diameter of vertical tubes over the range 0.016-0.050 m (Kobayashi & Fujita, 1997). At a certain depth, the light intensity fell to growth-limiting levels and a dark zone was encountered. This depth was dependent on reactor geometry, light source and intensity, biomass concentration, and other factors.  A flat-plate photobioreactor, comprising of transparent flat panels and a relatively high surface area to volume ratio, has been developed to allow for more light energy available for cells (Sierra et al., 2008). This reduces the effects of shading within the reactor. However, exposure of cells to excess light will result in photoinhibition (Krupa et al., 1991). Studies of PBRs are thorough and advanced; however, there is a lack of research in the optimization of growth of certain microalgal species in PBRs, including the freshwater cyanobacterium S. elongatus PCC 7942. Due to these 15  advantages, experimental work in this study was conducted with a flat-plate airlift photobioreactor. 1.5.1    Design of airlift reactors Airlift reactors have no focal points of energy dissipation and shear distribution is homogeneous throughout (Merchuk & Gluz, 2002). A typical flat-plate airlift photobioreactor is shown in Figure 5. The riser is located above the sparger, where gas bubbles are introduced into the liquid phase. Circulation of the liquid is mainly driven by the density gradient between the riser and downcomer (Chisti & Moo-Young, 1993).  The riser is sparged with gas, lowering the bulk density in this section. The downcomer is almost completely free of bubbles so the bulk density there is relatively higher. This induces a liquid circulation rate that is dependent on the geometry of the reactor, gas flow rate, and liquid and gas physical properties. It is a major characteristic of airlift reactors because this controls the circulation time, mixing, and mass transfer. The gas-liquid separation chamber is designed such that most bubbles from the riser are released from the liquid, leaving the downcomer with a negligible amount of bubbles. The gas entrainment chamber allows for bubbles to escape because expansion of the cross-sectional area reduces the liquid velocity at the entrance of the downcomer, giving more time for the bubbles to rise. To ensure that cultures are always well-mixed, a sufficient aeration rate should be utilized. Conversely, high aeration rates are not recommended because this would introduce unnecessary shear forces and larger bubbles that can decrease internal light intensities. 16    Figure 5: Flat-plate airlift photobioreactor The key hydrodynamic parameters in a reactor are gas hold-up, liquid circulation velocity and the mixing time. For airlift reactors, gas hold-up is calculated using the following equation:              DRRDRR VVV ?=? 1 where VDR is the dispersed riser volume and VR is the ungassed riser volume plus the gas entrainment volume. A useful characterization of flow pattern in a reactor is the mixing time, which can be quantitatively measured. Mixing time is the time required to reach the desired degree of homogeneity (for example 95%) after tracer injection. Axial mixing in loop reactors is commonly characterized by the Bodenstein number Bo, which is the ratio between the mixing effects of the axial dispersion coefficient and bulk movement of medium (Onken & Tr?ger, 1990):  axLDLwBo ?=  2 Where wL is liquid velocity, L is length of flow path, Dax is effective axial dispersion coefficient. A reactor is considered well mixed when Bo<0.1 and acting as plug flow at Bo>20 (Mir?n et al., 2000). 17  1.6    Optimizing growth of microalgae 1.6.1    Overview of media for microalgae Media are important because they provide the nutrients for microorganisms? metabolism and growth. They are generally composed of three components: macronutrients, trace elements, and vitamins. Macronutrients like nitrogen and phosphorous provide the nutrients essential for growth of all algae (Oilgae, 2009). However, trace elements can vary with different algae strains. Each trace element has an important function; for example, iron is required for electron transfer, and magnesium is needed for chlorophyll production. Lack of sufficient trace metals will limit the magnitude of growth. Each nutrient is usually prepared as a stock solution. Stock solutions are useful because they provide convenience and consistency; repeated individual weighing of chemicals would otherwise be time-consuming and prone to error. Microalgae are commonly grown in chemically defined media to ensure a pure and consistent cell culture environment. Undefined media contain complex ingredients and a mixture of numerous components in unknown concentrations. Undefined media are suitable for applications that require growing greater variety of microalgae or for economical reasons. 1.6.2    Medium and pH buffer selection Some medium recipes like BG-11 (Table 6) are formulated for multiple freshwater algae species; others are derived from detailed studies of a particular organism. For the purpose of optimizing the specific growth rate for a chosen species, it is necessary to establish specific nutrient requirements. BG-11 medium has shown to be successful in supporting growth of S. elongatus, thus this was chosen as the base case medium for the optimization experiments.  Table 6: BG-11 composition Component mM NaNO3 17.6 K2HPO4 0.22 MgSO4?7H2O 0.03 CaCl2?2H2O 0.2 C6H8O7?H2O 0.03 C6H11FeNO7 0.02 Na2EDTA?2H2O 0.002 Na2CO3 0.18 Trace metals 18  Each component in the medium plays an important role in sustaining cell growth. NaNO3 serves as the main source of nitrogen, which is converted by the bacteria to produce organic compounds like amino acids (Srivastava et al., 2013). The phosphorus in K2HPO4 is involved in a variety of physiochemical reactions and is important for the metabolism of cells. MgSO4 provides sulphur, which is also an essential nutrient that is found in amino acids, peptides, vitamins, and secondary metabolites (Jain et al., 2006). It is also the magnesium source that induces chlorophyll production. CaCl2 plays an important role in regulating osmotic pressures in cells (Srivastava et al., 2013). C6H8O7 is a central part of the citric acid cycle, which is a series of chemical reactions used to generate energy (Siegesmund, 2011). C6H11FeNO7 provides the iron required for photosynthetic apparatus (Shcolnick et al., 2009). Na2EDTA is a chelating agent that removes harmful heavy metal ions (De Philippis et al., 2011). Na2CO3 provides an additional source of inorganic carbon (Schwarz et al., 2011) and adds a buffering capacity to the medium. Trace metals act as cofactors for enzymatic reactions in the cell (Todar, 2012). The pH of the medium is also buffered to prevent any abrupt changes in pH that may potentially hinder cell growth.  1.6.3    Effects of temperature on growth In general, algae growth rates increase with rising temperatures until an optimum temperature is reached, after which growth declines. For every microalgal species, temperature greatly influences cellular chemical composition, uptake of nutrients, and carbon dioxide fixation. In most research involving S. elongatus PCC 7942, the subcultures of the strain were cultivated at a constant temperature of 30?C (Kajiwara et al., 1997). The effect of temperature on the rates of biological processes is well understood; however, each species responds differently to these changes (Konopka & Brock, 1978). For example, Microcystis isolates possess a thermal growth optimum between 28.8 and 30.5?C (Kr?ger & Eloff, 1978), while Chlorella sorokiniana has a thermal growth optimum at 35?C (Wageningen UR, 2011). In addition, there has been a lack of research focusing on the effects of temperature on the growth of S. elongatus PCC 7942. PBRs provide the ability to control temperature. Unlike outdoor reactors, fluctuations in ambient temperature will not affect the internal temperature of the PBR.  1.6.4    Gas-liquid mass transfer Monitoring dissolved CO2 concentrations and CO2 transfer rates is a standard procedure in microalgae cultivation to ensure adequate delivery of CO2 for algae. In this experiment, the on-19  line dissolved CO2 data were used to determine kLa, the volumetric mass-transfer coefficient for the CO2-medium system. CO2 is supplied to the culture medium by sparging CO2-enriched air through the medium. Thus, CO2 enters the reactor in air bubbles of fairly narrow size distribution. Empirical data and theory suggest that the major resistance to CO2 and O2 transfer is the liquid film surrounding the gas bubbles (Doran, 2012). Rates of mass transfer are often described by the classic two-film theory, where the main resistance to mass transfer lies in the liquid film.  For the estimation of theoretical kLa values at conditions in which CO2 concentrations are too small to be measured (ie. CO2 in air), the Rand and Marshall equation (Eq. 3) is useful (Bird et al., 2007). Assuming forced convection around the gas bubbles during operation of the photobioreactor, the Sherwood (Sh) equation for a spherical bubble is: 3/12/160.02 ScReSh += 3 where Re is the Reynolds number and Sc is the Schmidt number . Re is a dimensionless group that indicates the ratio of inertial and viscous forces in a fluid system (Eq. 4).  ??vdRe =  4 where ? is the density of the bulk liquid, v is the bubble velocity relative to the liquid, d is the diameter of the bubble and ? is the viscosity of the liquid. Sc is another dimensionless group that indicates the ratio of viscosity and mass diffusivity (Eq. 5). 2CODSc??=  5 where DCO2 is the diffusivity of CO2 in the liquid phase. Equation 5 is valid for constant physical properties and small net mass transfer rates. kLa can then be estimated using Equation 6: ShLDkL =  6 20  where D is the diffusivity of CO2 in water and L is the characteristic length. Similar equations can be used to determine kLa for oxygen (using DO2 instead of DCO2). Gas holdup is also important because it controls the residence time of gas in the liquid. In addition to bubble size, it controls the gas-liquid interfacial area available for mass transfer (Chisti & Moo-Young, 1988). 1.6.5    Effects of light intensity on growth  1.6.5.1    Photosynthesis Photosynthesis is the biological process of using light energy to produce chemical energy and stored in the molecular bonds of organic molecules (e.g. sugars) (Lai et al., 2005). This process occurs in plants and photoautotrophic algae like cyanobacteria. In cyanobacteria (ie. blue-green algae), photosynthetic reactions take place in light-harvesting organelles called phycocyanin (Mazel et al., 1988). CO2 is reduced to form carbohydrates, with water as an electron donor and O2 as a waste product. The overall chemical reaction is: nCO2 + nH2O (+ light energy)  (CH2O)n + nO2 7 The process of photosynthesis is dependent on the light intensity, which can also be expressed as the irradiance or the quantity of incident light on a surface. The unit for irradiance is ?mol photons/m2/s or ?E/m2/s (i.e. microeinsteins/m2/s). Photosynthesis is a combination of two reactions: a light phase and a dark phase (Prezelin, 1981). Light phase reactions only occur when cells are exposed to light while dark phase reactions occur in the absence of light. The purpose of the light-dependent reactions is to convert available light energy into a form that can be used readily in metabolic processes. Energy of the sunlight is converted to a biochemical reductant NADPH2 (nicotinamide-adenine dinucleotide phosphate) and a high energy compound ATP (adenosine triphosphate) (Masojidek et al., 2004). The dark phase reactions involve the utilization of the NADPH2 and ATP molecules produced by the light dependent reactions. The mechanisms of carbon fixation, called the Calvin cycle, occur here. The basic process of photosynthesis is illustrated in Figure 6. 21   Figure 6: Photosynthesis (Mayer, 2008) 1.6.5.2    Light Distribution Because light is the main source of energy for phototrophic algae, the distribution and intensity of light is one of the important factors for growth. Light attenuation is the main limitation in algae growth in cultivation systems due to decreases in light intensity and mutual shading (C. Lee, 1999). According to Lambert?s Law of Absorption, light intensity decreases along the path length in the photobioreactor as illustrated in Figure 7: lCainoutaceII lg?= 8 where Iout = light intensity exiting (?E/m2/s)            Iin = light intensity at interface where light enters (?E/m2/s)            ac = spectral-averaged absorption coefficient (m2/g)            Calg = algae concentration (g/m3)            l = path length (m) 22   Figure 7: Light intensity changes with distance in absorbing medium This law allows for determination of the light gradient inside the photobioreactor. Initially, the light strikes the reactor with constant intensity but past the surface, it is converted to photonic energy in algae cells. As a result, less light exits the reactor than entered. Integration of the above equation over the path length gives the average light intensity Iav.  [ ])exp(1 lglglCalCaII acacinav ??=  9 At very low light intensities, the rate of photosynthesis is equal to the rate of respiration (compensation point), resulting in no cell growth. As the intensity increases, photosynthesis increases until it reaches a light saturation point, where growth rate is at a maximum (Krupa et al., 1991). Higher light intensities will lead to a reversible damage of the photosynthetic apparatus of algae and decrease the photosynthetic rate (L?nneborg et al., 1988). This phenomenon, called photoinhibition, can be also triggered by a combination of light and other environmental factors such as low temperatures or drought. This effect can be reversed if the algae cells are exposed to favourable growth conditions. It is essential to determine the light intensity in which S. elongatus reached the light saturation point and to operate as close as possible to it. Algae cells absorb almost all the light projected on them but not all can be used for photosynthesis (N. Kim et al., 2002). At high cell densities, almost all light is absorbed by a thin top layer of cells, leaving the rest with less available light (self-shading) and decreases the Light source Iin Iout 23  photosynthetic efficiency. Figure 8 shows that light intensity decreases as penetration depth or cell concentration increases (C. Lee, 1999). The growth rate of microalgae depends on the light energy available to the cells. In the early stage of growth, microalgal cells adjust to conditions and start to grow exponentially. The organisms reach a maximum growth rate, but at the same time, mutual shading effects become more apparent. On average, less light is available for each cell, so the specific growth rate starts to decline. The biomass concentration continues to increase but the growth rate begins to decrease. Graphically, the biomass concentration increases linearly with time. This linear growth phase in batch cultures is a common observation for microalgae during photoautotrophic growth.  Figure 8: Light penetration depth of Chlorella kessleri as a function of light intensity (W/cm2) and cell concentration (cell number/mL)  Minimizing the effect of mutual shading requires reducing the light path length of algae culture for a more homogeneous light distribution. This is facilitated by proper mixing to ensure that individual cells are not stationed entirely in the dark or light zones of the photobioreactor. The intensity of light is sometimes expressed as the average light irradiance per algal cell.  Since S. elongatus has a preferable absorption spectrum of 631-684 nm (Espinosa et al., 2007), light used for all experiments has been selected to provide energy within this spectrum. The best light source for this spectrum and application is a compact fluorescent light bulb (CFL) that provides 6500K color temperature. The color temperature of a light source is the temperature of Cell concentration (cell/m2) Penetration Depth (cm) 24  an ideal black body that radiates light identical to that of the light source (Wyszecki & Stiles, 1982).  CFLs are cost effective, energy efficient and stable over the extended operating time (Geider & Osborne, 1992). 1.6.5.3    Light/dark cycles The presence of light and dark cycles also has been suggested to have a beneficial effect on the growth of S. elongatus. Photosynthesis is comprised of light dependent and light independent reactions so it is preferable to utilize a light/dark pattern for optimal growth. Literature involving photobioreactors has suggested that mixing rates create an artificial light/dark effect (Phillips & Myers, 1954). Turbulence causes individual cells to move between the high intensity of the front surface and the low intensity of the back surface. Some studies have focused on the effect of different light/dark cycles (Jacob-Lopes et al., 2009). Jacob-Lopes? results indicated that duration of light periods was a significant factor in the performance of photobioreactors. Biomass production and CO2 fixation was observed to decrease as the duration of the light period decreased, with the exception of 12:12 (light hours:dark hours) cycles. This can be explained by the circadian rhythm of microalgae, in which cells have adapted to grow best under 12:12, like natural sunlight patterns (Suzuki & Johnson, 2001). Specifically for S. elongatus, it is not known if culturing in photobioreactors with diurnal cycles will save energy and improves cell growth in comparison to continuous lighting. This is to be investigated. 1.6.6    Effects of O2 and CO2 on growth Microalgae are naturally adapted to atmospheric oxygen so exposure to air during growth is not a concern (Pope, 1975). At atmospheric conditions, the solubility of oxygen in water (7 ppm at 35 ?C and 1 atm air) is extremely low. However, high levels of O2 and CO2 can inhibit growth. There were some known cases in which high concentrations of O2 led to growth inhibition. For Chlorella pyrenoidosa, oxygen inhibition was only evident at 50% vvm O2 or higher, resulting in a dissolved O2 concentration of at least 15.7 ppm (Shelp & Canvin, 1980). Other microalgae respond differently at high O2 concentrations; however, there have been no published data on the effects of oxygen inhibition for S. elongatus (Pope, 1975). For S. elongatus, inhibition from CO2 was observed at concentrations greater than 5% (Kajiwara et al., 1997). There was also evidence of growth inhibition at inlet CO2 concentrations above 25  10% (Kajiwara et al., 1997). Due to the equilibrium reactions of CO2 in water, carbonic acid is also produced (Poling et al., 2001).  3(l)2(l)22(l) COHOHCO ?+  10 At equilibrium, only a small fraction (0.2-1%) of dissolved CO2 is converted to H2CO3. Nevertheless, high inlet CO2 concentrations will produce a greater concentration of H2CO3, causing pH to drop and adversely affect cell growth. For microalgae growth, CO2 and HCO3- are the most important carbon sources. S. elongatus PCC 7942 has been shown to be capable of growing at over 0.9 g/L in culture aerated with CO2-enriched air. Kajiwara determined growth characteristics of S. elongatus for CO2 fixation by cultivating under various inlet CO2 concentrations in a photobioreactor and comparing final biomass yields and growth rates to determine which conditions gave the best yields. His external loop airlift column achieved the highest specific growth rate of 0.048 h-1 at 5% CO2. At higher CO2 concentrations, the maximum biomass concentration and specific growth rate have been observed to be lower than 0.048 h-1, signifying that elevated CO2 concentrations above 5% CO2 did not correlate to higher biomass productivities. In this case, excess CO2 acidified the medium at levels that inhibited growth. Yet productivities grown with CO2-enriched air were at least 30% higher than for culture grown with air. It is paramount to understand the effects of inlet CO2 concentration and pH changes on growth to minimize the inhibitory effects while achieving high biomass productivity. Nonetheless, the findings suggested that S. elongatus is capable of growing at elevated CO2 concentrations, paving the way for the application of industrial waste emissions as a carbon source.  1.6.7    Effects of agitation on growth As discussed previously, at high algae concentrations, almost all the available light is absorbed by the top cell layer. Thus, mixing must be sufficient to maintain cell suspension and light homogeneity. Improper mixing could also lead to sedimentation, uneven thermal distribution, and inefficient gas exchange. Mixing also decreases the boundary layer surrounding each cell, promoting the uptake of nutrients and release of metabolic products. Sources have shown that both growth rate and maximum biomass concentration increased when the aeration rate increased, with CO2 concentration constant at all times (Zhang et al., 2002; Q. Hu et al., 1996). Furthermore, it was concluded that if a gas containing a low CO2 concentration was used, a high 26  critical kLa value would be required to meet the CO2 requirement for algae growth. This indicated that there was room for optimizing the vertical airlift photobioreactor by varying the aeration rate.  1.7    Previous Research on Optimization of Growth of Microalgae Studies on growth optimization of numerous microalgal species have been performed over the past few decades, but in most cases, high growth rates have been elusive. One study demonstrated that ?max of S. elongatus decreased from 0.033 h-1 to 0.0290 h-1 as the light intensity increased by a factor of 16 in an external loop photobioreactor (Sasi, 2009). In terms of the growth response from varying the inlet CO2 concentration, cultivation with air produced the lowest ?max of 0.015 h-1, while 5% CO2 concentration resulted in the highest ?max of 0.043 h-1, but further increases in CO2 concentration did not generate an increase in ?max.  Information on the growth optimization of S. elongatus PCC 7942 has been limited but still provided useful data on the effects of growth by varying different conditions. For instance, Kajiwara performed batch cultivations of this strain in an external loop airlift reactor (Kajiwara et al., 1997). It was found that varying the gas flow rate from 0.2 L/min to 1 L/min did not affect growth when the inlet CO2 concentration, light intensity and temperature were fixed at 5%, 73 ?E/m2/s and 30?C, respectively. This indicated that mass transfer of dissolved CO2 was not a limiting factor of cell growth at CO2 concentrations of 5% or higher. In addition, for cultivations with the inlet CO2 concentration ranging from 5% to 30%, the highest ?max (0.044 h-1) and Xmax (0.88 g/L) was observed at 5%. This observation was similar to that of Chlorella vulgaris (Sasi, 2009). In the experiments where light intensity was varied between 41 ?E/m2/s and 135 ?E/m2/s, ?max and Xmax increased as the light intensity increased. However, the responses decreased at light intensities above 108 ?E/m2/s, which implied the presence of photoinhibition. Most importantly, cultivations grown in CO2-enriched air yielded greater specific growth rates and final cell mass concentrations, showing promise for applications centred on CO2 fixation with algae. 1.8    Modelling microalgae growth 1.8.1    Cell growth models In general, batch cultures of microorganisms experience 5 growth phases: lag phase, exponential phase, declining phase, stationary phase, and death phase (Figure 9). The duration of each phase 27  depends on the initial cell concentration, concentrations of the nutrients, pH, temperature, and other factors. During the lag phase, cells acclimatize to the medium. Cells synthesize new enzymes required for cell multiplication and there may be a small increase in cell mass but that is usually apparent near the end of the phase. Following the lag phase is the exponential growth phase. Cells adapt and multiply rapidly and a ?balanced growth? is achieved: all components of the cell increase at the same rate. The growth rate is independent of concentrations of medium components as they are in excess; cell numbers increase exponentially with time. Past the end of this phase, the growth rate begins to decline but the biomass concentration continues to increase. At the stationary phase, medium nutrients are exhausted and the cells stop multiplying but existing cells continue to grow in size. Formation of inhibitory products like organic acids may also occur. Cells are still metabolically active, and can produce secondary metabolites. Eventually, cells enter the final phase, the death phase, where cells lyse and the number of living cells decreases.  Figure 9: Typical growth curve for batch cultivation Mathematical models have been developed, based on time dependent changes of the specific growth rate ?, for prediction of cell growth in batch cultures (Lin et al., 2000). The most commonly used expression is the classic Monod equation (Monod, 1949): SKSS +=max??  11 Number of cellsTimeExponential phase Death phase Declining growth Lag phase Stationary phase 28  where ?max is the maximum specific growth rate, S is the substrate concentration, and Ks is the saturation constant of the rate-limiting substrate when the specific rate of growth is half of the maximum. The specific growth rate is defined as the rate of increase in cell mass per unit time as measured in batch culture. Traditional growth models for batch and fed-batch cultures mostly involved unstructured kinetic models and variations of the Monod equation (Goudar et al., 2005). Unstructured kinetic models are impractical because they require a large number of parameters from a system of differential equations. For example, in the model introduced by Dhir specifically for mammalian cell growth in a fed-batch culture (Dhir et al., 2000), the rate of change of live cell concentration X was XdtdX ?= 12 ? was a function of the concentration of four limiting substrates G,Q, L and A. ),,,( ALQGf=?  13 Equations 14-17 represented the rate of change of each substrate, with each equation involving several parameters. ),( XGfdtdG= 14 ),( XQfdtdQ= 15 ),,( XQLfdtdL= 16 ),,( XQAfdtdA= 17  Although Dhir?s model adequately described experimental data, like other unstructured kinetic models, it was computationally not practical for modelling growth because it involved the non-linear estimation of a large number of parameters. Analytical solutions of these equations are not applicable because they assume that specific growth rates are constant. This is not the case in batch mode. For heterotrophic microorganisms, the growth rate decreases over time because the concentration of the limiting nutrient decreases during growth (Ernst, 2004). In bacteria like E. 29  coli, the limiting substrate is typically glucose under aerobic conditions. Thus one can simply plot the Monod growth model with glucose concentration as the substrate. However it is difficult to model S. elongatus growth with Monod because, as autotrophs, it produces its own intracellular substrate from extracellular components and this is difficult to quantify. Another approach is polynomial fitting, which is useful over the whole duration of growth and the specific growth rates are simple to compute. However, cellular and product concentrations show exponential behaviour and are difficult to describe with polynomials. Polynomials are also especially sensitive to outliers, yielding unrealistic trends. 1.8.2    Logistic growth model Logistic equations have adequately described bacterial growth curves. A logistic growth model, as introduced by Yang (Yang et al., 2011), was chosen to model microalgae growth: XXXdtdX )11(maxmax +?= ?  18 Where ?max is the maximum specific growth rate of the microalgae, X is the biomass concentration, and Xmax is the maximum biomass concentration. The disadvantage of the model is that it does not apply for the declining phase; the model assumes that X approaches Xmax as t increases. But since experiments were not typically run long enough to observe a declining phase, the model adequately described the first four growth phases which were the primary focus of this research.    30  2    Research Objectives The objective of this project was to maximize the specific growth rate and biomass concentration of S. elongatus PCC 7942 cells in a 2D airlift photobioreactor and demonstrate the potential for production of high-value products using microalgae. A set of experiments was performed to complete this study. To evaluate whether microalgae is an ideal vehicle for production of value-added products, the following objectives were performed. First, the effect of laboratory-scale shake flask conditions on the growth of S. elongatus was studied. The independent variables that were included in the study were medium composition, light intensity, and temperature. To gain an understanding of how the three variables affected the specific growth rate and biomass concentration, a logistic growth model was used. This was followed by the determination of optimized parameters which resulted in the highest specific growth rate and biomass concentration for the ranges investigated. Next, to analyze the effects of CO2 concentration and mixing on cell growth, studies were conducted in a 2D airlift photobioreactor, operating in a batch configuration and in axenic conditions. Similar to the methodology from the shake flask experiments, a logistic growth model was implemented with experimental data to investigate the effects of different variables on the specific growth rate and biomass concentration. Optimized conditions from the shake flask scale were applied to the airlift reactor studies. The independent variables of interest were inlet CO2 concentration, aeration rate, and light intensity. Light/dark cycles were also investigated to see whether provision of regulated illumination intervals improved cell growth.31  3    Materials and Methods 3.1    Synechococcus elongatus PCC 7942 Wild type Synechococcus elongatus PCC 7942 was selected as the representative microalgal strain for experimentation. It was isolated and provided by Dr. Francis Nano from the University of Victoria (UVic). 3.1.1    Subculturing of S. elongatus in BG-11 liquid medium The method of conserving microalgal cultures was by continued maintenance under controlled conditions, which involved routine liquid media and agar media preparation using aseptic techniques. The purpose of using liquid media was to prepare inoculum for experiments. The purpose of using agar media was to preserve monoculture for future use, and these are typically kept for two months before they are replaced. Liquid media involved serial subculturing, in which inoculum from a late log phase culture was transferred to fresh, sterilized BG-11 medium. This organism was subcultured at 30?C under a continuous light intensity of 80 ?E/m2/s using Philips 6500k 60W compact fluorescent lamps. All experiments were conducted with the seed culture, taken during the exponential growth phase (approximately 96 hours from the inoculation time). To prepare a 100 mL flask of subculture, 10 mL of 96 hour-old culture is transferred to 90 mL of BG-11. This was cultured in 250 mL Erlenmeyer flasks and agitated at 120 rpm in a New Brunswick Innova? 42 shaking incubator (Figure 10).  Figure 10: Incubation of subcultures 32  3.1.2    Preparation of BG-11 liquid medium To prepare BG-11 medium, stocks of each component were made first according to the recipe from the University of Texas (UTEX). Components in BG-11 medium are listed in Table 7 below (UTEX, 2009). Following this, NaNO3, MgSO4?7H2O, CaCl2?2H2O, C6H8O7?H2O (citric acid), C6H11FeNO7 (ammonium ferric citrate), and Na2EDTA?2H2O were mixed in deionized water and then autoclaved in a Midmark? M11 autoclave. The other components were not added into the solution until after the autoclave process to prevent precipitation of components. HEPES pH buffer was not used in the experiments to simplify the factorial design and to reduce the number of potential carbon sources. Table 7: BG-11 medium stock solutions (UTEX, 2009) Component Quantity per litre of media Final Concentration NaNO3 10 mL 17.6 mM K2HPO4 10 mL 0.22 mM MgSO4?7H2O 10 mL 0.03 mM CaCl2?2H2O 10 mL 0.2 mM C6H8O7?H2O 10 mL 0.03 mM C6H11FeNO7 10 mL 0.02 mM Na2EDTA?2H2O 10 mL 0.002 mM Na2CO3 10 mL 0.18 mM BG-11 Trace Metals Solution 1 mL (see recipe below) Trace metals solution H3BO3 - 46 ?M MnCl2?4H2O - 9 ?M ZnSO4?7H2O - 0.77 ?M Na2MoO4?2H2O - 1.6 ?M CuSO4?5H2O - 0.3 ?M Co(NO3)2?6H2O - 0.17 ?M  3.1.3    Preparation of BG-11 agar plates To make BG-11 agar plates, agar medium was first made. The concentrations from Table 7 were added to distilled water for a total of 500 mL. In a separate container, 15g of agar were added to 500 of distilled water. After autoclaving both solutions, they were allowed to cool to 45-50?C. 3g Na2S2O3?5H2O, which served as a source of sulphur, was added to the agar solution. Next, both agar and liquid solutions were mixed together and quickly poured into 30 Petri plates, which  were left to cool. To transfer microalgal colonies from agar, a wire loop was sterilized over a Bunsen burner flame and allowed a colony of cells from one Petri plate to another and subsequently streaked across the plate into four quadrants, as shown in Figure about 2 weeks. To ensure that the subcultures used as inoculum contained only monoculture, 10 mL of BG-11 was inoculated with a colony was then allowed to mature into a culture growing in the exponential phase (96 hours after inoculation), and then used as seed culture for 100 mL medium.Figure 3.1.4    Sterilization To ensure that all working materials were not contaminated with foreign microorganisms, aseptic conditions had to be established. Sterilization was important in maintaining isolated strains of elongatus and keeping experiments free of potential unwantedtwo methods of sterilization were used: filter sterilization and autoclave sterilization. Autoclave sterilization was used to sterilize BGequipment that would contact with the medium. autoclave at a temperature of 121?C and 28 psia for 30 minutes. Filter sterilization was used to sterilize several BG-11 components that may precipitate under autoclave conditions. To filter sterilize, the mixture was passed through a 0.45 micron Millexphotobioreactor, the chamber was soaked in temperature, followed by thorough rinsing in sterile deto cool before use. Once cooled, the loop was used to transfer11. The time it took for cells to mature into colonies was S. elongatussingle colony from the Petri plate. The   11: Petri dish of S. elongatus  variables. In the shake flask scale, -11 medium, Erlenmeyer flasks and foam plugs, and all The medium and equipment were placed in the-GP Filter. To sterilize the 0.6% sodium hypochlorite for 60 minutesionised water in a sterile environment.33   S.   at room  34  The sodium hypochlorite solution was made by diluting 15 mL household bleach (containing 5% hypochlorite) in 1.25 L water. 3.2    Measurement of biomass concentration A reliable way of tracking cell growth is by measuring biomass concentration using a spectrophotometer. To determine the cell concentration, the optical density (OD) was measured at 600 nm using a Shimazu spectrophotometer. Before any estimation of biomass concentrations can be made, a calibration curve between the OD and dry cell weight was constructed (Figure 12). To determine the biomass concentration for a specific OD, a 2 mL sample was taken from a growth flask of 100 mL volume during a late exponential phase. Five 1:2 serial dilutions were made from this sample and the ODs were recorded. The volume of the remaining culture (98 mL) was recorded, then placed in an oven and dried overnight at 90?C (Yun et al., 1997). Following overnight drying, the sample was moved to a desiccator where it could cool without absorbing moisture. After 1 hour of cooling, a dry weight measurement was recorded. Dividing this by the recorded volume gave the biomass concentration of the original sample. The biomass concentrations of other dilutions were calculated. Sampling was performed in duplicate.  Figure 12: Determined relationship between dry cell concentration and optical density The biomass concentration was calculated using the following equation:  004.03542.0 ?= AX  19 y = 0.3543x - 0.004R? = 0.999400.050.10.150.20.250.30.350.40.450 0.5 1 1.5Biomass concentration (g/L)Optical density (600 nm)35  where X is biomass concentration (g/L) and A is optical density. The biomass concentration of all subsequent samples was measured by the optical density based on this relationship. 3.3    Logistic growth model and carbon uptake rate After integration of Eq. 18, the experimental data were fitted by non-linear least squares using multiple sets of data to the following model equation:  tootoeXXXeXXXmaxmaxmaxmax??+?=  20 where Xo is the initial biomass concentration. In Figure 13, the continuous line represented a logistic fit of growth from experimental data. ?max was determined by the solver function in Microsoft Excel. All other parameters can be directly retrieved from experimental data.   Figure 13: Example of a logistic growth curve Carbon uptake rates (CUR) are calculated indirectly from the biomass concentration, assuming that CO2 is converted into organic carbon via photosynthesis. Since S. elongatus PCC 7942 metabolized only CO2 as a carbon source, the amount of CO2 consumed was estimated directly from cell mass. Results from an elemental analysis of S. elongatus done by Kajiwara showed that 00.050.10.150.20.250.30.350.40.450.50 50 100 150 200Biomass concentration (g/L)Time (h)expt datalogistic modelExperimental dataFitted logistic model36  the cells comprised of 46.5% carbon. This led to an equation for calculating the CUR (Kajiwara et al., 1997). 1244465.0)//( ??=dtdXhLgCUR 21  where dX/dt is the biomass growth rate, 44 is the molecular weight of CO2, and 12 is the molecular weight of carbon. Elemental analysis showed that the 0.465 g of carbon per gram of dry cell mass (Kajiwara et al., 1997). It was observed to be constant among all growth phases. 3.3    Optimization of medium composition Growth media are an important determinant of cell growth and composition because they provide the nutrients for metabolism and growth. Some of the medium recipes like BG-11 are formulated for many freshwater algae species; others are derived from detailed studies of a particular organism. For the purpose of optimizing growth for a chosen species, it is necessary to establish specific nutrient requirements. 3.3.1    Defining conditions for screening medium composition A screening design was used to find components of BG-11 medium that had significant effects on ?max and Xmax. A full factorial experimental design was carried out for 5 components: ammonium ferrous citrate, NaNO3, CaCl2, K2HPO4, and Na2CO3. The concentrations used in the experimental design are outlined in Table 8. The other components that made up BG-11 were also present in the medium, and they had the same concentrations as listed in Table 7. Each run was carried out in 250 mL shake flasks, incubated at 30?C and at 120 rpm in a shaker. Four fluorescent lamps were mounted at the top of the shakers to provide an average incident light intensity of 80 ?E/m2/s. Table 8: Experimental design for screening of medium components Parameter Concentrations (mM) C6H11FeNO7 0, 0.02 NaNO3 0, 17.6 CaCl2 0, 0.2 K2HPO4 0, 0.22 Na2CO3 0, 0.18 37  3.3.2    Defining conditions for optimization of medium composition Following the screening experiments, a central composite experimental design was used to study the effects of the concentration of each BG-11 component on ?max and Xmax. Table 9 outlines the concentrations of each component used in the design. Analysis of growth in flasks was conducted in a randomized order. Temperature, light intensity, and agitation rate were identical to those used in the screening experiments. Table 9: Central composite design for optimization of BG-11 medium components Component Concentration (mM) NaNO3 4.4 8.8 17.6 35.2 70.4 K2HPO4 0.055 0.11 0.22 0.44 0.88 CaCl2 0.05 0.1 0.2 0.4 0.8 C6H11FeNO7 0.005 0.01 0.02 0.04 0.08 Na2CO3 0.045 0.09 0.18 0.36 0.72 3.4    Experimental setup of photobioreactor Large scale work was carried out in a 2D photobioreactor. The photobioreactor setup is illustrated in Figure 14. The photobioreactor was connected to a CO2 cylinder (supplied by Praxair) and a building-compressed air nozzle. Gas from the CO2 cylinder was regulated by two flowmeters; one to handle flow rates less than 150 mL/min and the other for higher flow rates. A manual switch valve allowed for gas to be directed to either flowmeter. Between the flowmeter setup and the reactor was a pressure relief valve set at 5 psig to prevent pressure build-up. A 0.45 ?m filter was also placed on the gas line to remove foreign microorganisms and therefore minimize the risk of contamination. An outlet gas port at the top flange of the reactor allowed for exhaust gases to be directed back to the fume hood. The photobioreactor was equipped with a dissolved CO2 electrode (VL-PC), an oxygen electrode (VWR), and one sampling port. Another port was also built in to the top flange if needed.  Figure The dissolved CO2 electrode was connected to calibration curve is shown in Figure measuring the electrode responses of serial dilutions of a 1000 ppm dissolved oxygen were measured using a VWR Symphony the reactor was not possible due to space constraints, so pH was measured except pH were recorded in-situ every 10 minutes by photobioreactor was illuminated continuously at the front surface by 4 CFL bulbs (Philips 60W, 6500k color temperature). Figure 15051015202530354045Potential difference (mV)Flowmeter manifold 14: Photobioreactor setup a Thermo Orion A710+ pH meter. 15. Calibration of the electrode was performed CO2 standard. meter. Installation of a pH probe in offlinethe Labtech Notebook software : Dissolved CO2 electrode calibration y = 8.2082ln(x) - 1.2R? = 0.95751 10 100CO2 concentration (mg/L)Light panel PhotobioreactorSampling port  and electrodes Filter 38  The electrode by Both pH and . All parameters . The  39  3.4.1    Calculation of the concentrations for each carbon species The CO2 electrode was an ion-selective type, and it had a membrane that allowed the passage of only dissolved CO2. When dissolved CO2 passed through the membrane, it entered the probe in the form of H2CO3. However, the probe does not have the ability to measure the concentrations of other inorganic carbon species in the solution. An understanding of the dissociation reactions of CO2 is required to determine such concentrations and therefore calculate the concentration of total dissolved inorganic carbon c, where c can be defined as: c = [H2CO3] + [HCO3-] + [CO32-] 22 Carbonic acid dissociates in two steps (Haynes et al., 2012): H2CO3 + H2O  H3O+ + HCO3-      pKa1  (33 ?C) = 6.316 23  HCO3- + H2O  H3O+ + CO32-        pKa2 (33 ?C) = 10.27   24 Where the pKa values are derived from a water-CO2 system. Figure 16 shows the concentrations of each CO2 species at pH between 0 and 14 for c=1.  Figure 16: CO2 species and pH (Haynes et al., 2012) By definition: x = 10-pH 25  0.00.10.20.30.40.50.60.70.80.91.00 1 2 3 4 5 6 7 8 9 10 11 12 13 14pHc (mol/l) [HCO3-][CO3 2-][H2CO3]=[CO2]l[HCO3-] [CO32-] [H2CO3] 40  To calculate c, we can derive the following equation from the equilibrium constants: 2211232 )](COH[xKKxKxc aaa++=  26 pH values and [H2CO3] were known from experimental data. As pKa values have not been determined for CO2-BG11 medium system, the pKa values for the water-CO2 system were used for calculations instead.  3.4.2    Specifications of photobioreactor A diagram of the airlift photobioreactor is shown in Figure 17. A vertical flat-plate airlift photobioreactor was selected as a suitable configuration because it provided a short light path length, high surface area to volume ratio for minimal dark zones, and a simple design (Q. Hu et al, 1998). The reactor had an internal depth of 2.54 cm, width of 16 cm, and an internal height of 28 cm. One of the rationales for choosing a ?two-dimensional? design was that a single vertical tubular photobioreactor bubble-column or airlift reactor cannot exceed about 0.2 m in diameter (Chisti & Moo-Young, 1993) or light availability would be hindered (Mir?n et al., 2000). Secondly, a thin and flat design allowed for simple measurement of light intensity that would otherwise not be possible with a curved reactor surface. Two-dimensional systems have made it easier to attain quantitative flow information such as bubble size, liquid superficial velocity, and gas hold-up. An internal loop circulation was induced by a sparger and a pair of 6? baffles, forming riser and downcomer regions of equal volumes. 41   Figure 17: Reactor dimensions An acrylic block was installed to provide a rounded bottom, with a partition in the centre for insertion of a glass sparger, pictured in Figure 18. This block removed blind zones and minimized the settling of cells. The sparger was supplied by Cansci Glass Ltd and used for all photobioreactor experiments in this study. The sparger had an average pore size of 4.0-5.5 microns. At a working volume of 1 L, there was a headspace of 5 cm and gas-liquid separator height of 2.5 cm. The temperature was regulated by a 2 cm water jacket surrounding the sides and back of the reactor, with circulating warm water from a water bath. The system was operated at atmospheric pressure for all experiments. 42   Figure 18: Front view of photobioreactor 3.5    Hydrodynamic parameters of the photobioreactor Macroscopic flow structures of gas-liquid fluidization systems were studied through flow visualization using the 2D airlift photobioreactor under various operating conditions. The gas flow rates in the reactor were adjusted by valves and measured by flowmeters. The effects of gas flow rate on superficial liquid velocity, bubble size, gas hold-up, and mixing times were examined. Eight gas flow rates were used: 0.25, 0.5, 0.75, 1, 1.25, 1.5, 1.75, 2 L/min. For non-Newtonian fluids such as biological suspensions, fluid behaviour is assumed to be similar to that of water due to the relatively low mass fraction of the solid phase (algae).  3.5.1    Procedure for determining mixing time A tracer experiment was designed for determining the mixing time of the reactor at various gas flow rates, and Bodenstein number. The acid tracer experiment was based on a procedure devised by Miron (2004). The method measured the mixing time, defined as the time to reach a 5% deviation of the steady state concentration. Before each experiment, the system was purged with N2 gas to strip out dissolved CO2. The pH of the medium was lowered to 2 with HCl and bubbled with air for 20 minutes to remove carbonates. Then it was raised to 4.5 by adding 12M sodium hydroxide. The acid tracer (0.66 mL of 35% HCl) was then added instantaneously into Dissolved oxygen probe Dissolved CO2 electrode Sparger Sampling port Gas outlet port 43  the 1L BG-11 medium at the centre of the surface of the sparger. Air was used as the sparging gas. The pH detection point was 2.5 cm below the gas-liquid interface, located above one of the downcomers, and the tracer injection point was 1.2 cm above the detection point. Changes in pH were detected with a Cole Parmer pH electrode. Readings were automatically recorded every 0.8 seconds, which was the minimum time permitted by the pH meter. The theoretical mixing time tm for 95% homogeneity was calculated using the following equation (Petrovi? et al., 1995):  26.05.019.012.031.05.53 ?? ??????= SDRGm VVVDHUt  27 where UG is the superficial gas velocity, H is the height of riser, D is the characteristic diameter of the whole reactor, VR is the riser volume, VD is the downcomer volume, and VS is the separator volume. The superficial liquid velocity was calculated by dividing the distance of the circulation path 2H by the peak interval time in pH readings. The pH readings were converted to hydrogen ion concentrations, and the dimensionless concentration CT was calculated using the following equation:   initialfinalinitialinstT HHHHC ][][][][++++??=  28         The experimental mixing time was found by determining the time taken for CT to reach 0.95 in the readings starting from the time of tracer addition. 3.5.2    Procedure for determining the Bodenstein number The Bodenstein number characterized axial mixing in the PBR, and this is described in the following equation (Blenke, 1979): ? ?+=?=????????????????==2222/14)(exp4)(UUUUUxx UUUUxrr BoxBocc?? ??pi? 29 Where cr is the predicted dimensionless concentration, xU is the dimensionless distance, and ?U is the dimensionless time. Experimental cr was calculated using the formula: ???=cctccr)( 30        44  Where c(t) is the concentration at time t and c? denotes the steady state concentration. xU is the relative flow path, which is defined by: UU LLx ?  31 which is the ratio of the distance travelled L and the characteristic length LU=2H of one circulation. In this case, the characteristic length was 12 inches (twice the length of each baffle). ?U is the ratio of time t to the mean circulation time tU. The mean circulation time was found by measuring the time interval between peaks of the tracer profile. Once xU and ?U were computed, the experimental cr was fitted to Eq. 29 with Bo as the fitting parameter. This was achieved using the solver function in Excel. Once Bo is determined in Eq. 29, Dax can be subsequently calculated in Eq. 2 (Onken & Tr?ger, 1990). This information can be used to compare mixing of the internal loop airlift reactor at air flow rates between 0.25 L/min and 1 L/min. This range for the flow rates was chosen because these have been commonly used in airlift photobioreactors for microalgae cultivation (Kajiwara et al., 1997). 3.5.3    Procedure for calculating gas hold-up Gas hold-up ?R was calculated by recording the change in dispersed liquid riser volume VDR relative to the initial liquid riser volume VR. Since the cross-sectional area of the riser was the same along the height of the riser h, the gas hold-up was calculated simply by measuring the change in height of the gas-liquid interface. This is expressed as: DRRDRDRRDRR hhhVVV ?=?=?  32 3.5.4    Procedure for determining volumetric mass transfer coefficient The medium used for mass transfer experiments was BG-11. Effects of cells were not analyzed to simplify the experiments. The pH value of the fresh medium was not adjusted and was initially at about 8.4. The volumetric mass transfer coefficient of oxygen (kLa) was determined by the dynamic gassing-out method. First, the medium is scrubbed free of oxygen and CO2/carbonates by nitrogen gas (supplied by Praxair). Aeration was then initiated at a constant air flow rate and the increase in dissolved O2 (DO2) and CO2 (DCO2) was logged by computer. To analyze the effects of temperature on kLa, a low temperature of 23?C and a high temperature of 33?C were observed. CO2-enriched air was introduced at CO2 concentrations of 5% and 10%, at 45  a constant total gas flow rate of 0.5 L/min. To analyze the effects of gas flow rate and CO2 concentrations on kLa, experiments were conducted at total gas flow rates of 0.5, 1, 1.5, and 2 L/min. CO2 concentrations were adjusted to 5% and 10%. The temperature was maintained at 33?C.  The following equation represents the carbon dioxide transfer rate CTR into the culture medium: )( CCakdtdCCTR SL ?==  33 Where C [mol/m3] is the dissolved CO2 concentration in the bulk liquid at any time t [s], CS is the saturation concentration of CO2 in the liquid, and kLa is the volumetric mass transfer coefficient. Integration of Eq. 33 yields an equation where kLa can easily be calculated: ????????=SSoL CCCCtak ln1 34 where Co is the initial dissolved CO2 concentration at time t=0. Calculation of kLa using Eq. 34 is justified when the following conditions are met: the liquid phase mixing time, the response time of the CO2/O2 electrodes, and the gas phase residence time must be lower than the time constant for CO2/O2 transfer [1/kLa] (Boogerd & Kuenen, 1990). To observe physical changes of the gas-liquid system, bubble size was measured by image processing software ImageJ, as shown in Figure 19. A total of 100 measurements were taken for each gas flow rate (Appendix B).   Figure 19: Measurement of bubble diameter with ImageJ 46  3.6    Procedure for optimization of ?max and Xmax in shake flasks A central composite design was used to investigate the effects of light intensity and temperature on ?max and Xmax. The experiment consisted of 8 different conditions and 5 center points for a total of 13 runs. Three replicate experiments were carried out for reproducibility. The reaction conditions used in the experimental design are outlined in Table 10. Light intensity was varied by changing the height of the light panel, which was suspended over the flasks. Incident light intensities (ILI) were measured using an Apogee MQ-200 quantum sensor. Temperature was controlled by the incubator and the agitation rate was set at 120 rpm. Air was used as the source of CO2. Gas exchange between the shake flask and the external environment occurred at the mouth of the flask, through foam plugs. The foam plugs maintained axenic conditions inside the flasks. In each experiment, 10 mL inoculum was added to 90 mL of BG-11 medium and grown in 250 mL flask. The flasks were kept in two Innova incubators (Models 4230 and 42). Initial biomass concentrations were 0.03-0.05 g/L. One mL samples were taken daily from each flask for OD measurements.  Table 10: Central composite design for optimization of ?max and Xmax using two factors Light intensity (?E/m2/s) Temperature (?C) 150 30 79 30 200 35 150 30 150 23 100 25 150 30 221 30 200 25 150 37 100 35 150 30 150 30 3.7    Procedure for optimization of ?max and Xmax in photobioreactor A full factorial experimental design was used to study the effects of light intensity, CO2 concentration, and gas flow rate on ?max and Xmax. The reactor operating conditions used in the experimental design are outlined in Table 11. Temperature was kept constant at 33?C by a water 47  bath. Light intensity was varied by changing the distance between the light panel and the reactor, as shown in Figure 20.   Figure 20: Photobioreactor and light panel setup The laboratory in which this reactor was located was not exposed to external sources of light; the reactor was subjected to light only from the light panel. Analysis of each run was conducted in a randomized order. ILI was measured using an Apogee MQ-200 quantum sensor. At the beginning and at the end of each experiment, ILI was recorded to verify whether the ILI was constant during cultivation time. The ILI was the average of measurements at twelve different locations of the illuminated surface. Exiting light intensities were recorded at twelve different locations of the back surface of the reactor, and then averaged to give the mean exiting light intensity. ILI were within 5% of the mean, showing little variance over the surfaces of the reactor. Table 11: Experimental design for optimization of ?max and Xmax using two factors in photobioreactor CO2 concentration (%) Inlet flow rate (L/min) Light intensity (?E/m2/s) Air 1 60 Air 1 120 5 1 60 5 1 120 10 1 60 10 1 120 Photobioreactor Light panel 48  4    Results for Shake Flask Scale Experiments 4.1.1    Screening of BG-11 medium components A set of experiments designed to test the effects of a broad range of BG-11 component concentrations was conducted. The effects of NaNO3, K2HPO4, CaCl2, C6H11FeNO7, and Na2CO3 on response variables ?max and Xmax were analyzed. Optimal BG-11 concentrations that maximized ?max and Xmax were then determined. The experiment consisted of two stages: a screening stage and a central composite design. The purpose of the screening experiment was to determine the range of concentrations that had significant impacts on the response variables. Using these concentrations, the next step was to integrate them into the optimization experiments and then determine the concentrations that gave the highest ?max and Xmax.  A full factorial design was developed for the screening study, as shown in Table 12, to assess the impact of each medium component on ?max and Xmax. Each independent variable was investigated at a high level and a low level which represented two different nutrient concentrations. The same initial concentrations for MgSO4, CaCl2, C6H8O7, Na2EDTA and trace metals from Table 6 of Materials and Methods were used in this design. The t test and regression analysis were done to determine whether the inclusion of a given component had a significant effect on ?max and Xmax. Table 12: Full factorial design of BG-11 components Independent variable Symbol Level -1 1 NaNO3 (mM) X1 0 17.6 K2HPO4 (mM) X2 0 0.22 CaCl2 (mM) X3 0 0.2 C6H11FeNO7 (mM) X4 0 0.02 Na2CO3 (mM) X5 0 0.18  For each shake flask, the algae were highly dispersed in the culture medium, and cells did not adhere to the surface throughout the cultivations. Thus, optical density measurement in each sample was a good representation of the biomass concentration.   49  As an example, Figure 21 shows the growth pattern obtained in a shake flask at high levels of each medium component listed in Table 12.  Figure 21: Growth curve and fitted logistic model for estimating ?max A lag phase was not apparent in the experimental data points in Figure 21. This absence may be explained by the lag phase being shorter in duration than the sampling rate (once per day), and thus was not detected. In most of the growth runs conducted, a lag phase was apparent in the experimental data points, and this meant that cells needed to acclimatize to new conditions. For the particular set of conditions in Figure 21, the exponential growth phase spanned approximately144 hours, and this was followed by a stationary phase. Most growth curves had exponential growth phases that ranged between 120 to 288 hours depending on the conditions. The biomass concentration was calculated from optical density measurements using Eq. 19. Next, ?max was calculated using Eq. 20 using non-linear least squares for model fitting. Xmax was taken from the highest optical density in each run. The results for each set of conditions are presented in Table 13.   00.10.20.30.40.50.60.70 50 100 150 200 250Biomass concentration (g/L)Time (h)Experimental dataFitted model50  Table 13: Results from the BG-11 medium factorial design Component concentrations Response Initial pH NaNO3 (mM) K2HPO4 (mM) CaCl2 (mM) C6H11FeNO7 (mM) Na2CO3 (mM) ?max (1/h) Xmax (g/L) 17.6 0.22 0.2 0.02 0.18 0.031 0.57 8.507 17.6 0 0.2 0.02 0.18 0.021 0.82 8.662 0 0 0.2 0 0 0.013 0.80 8.722 0 0.22 0 0.02 0.18 0.022 0.80 8.933 17.6 0.22 0.2 0.02 0 0.027 0.57 8.451 17.6 0 0.2 0 0 0.019 0.52 8.559 0 0 0 0 0 0 0.032 9.057 0 0.22 0 0 0 0.021 0.43 8.719 0 0.22 0.2 0 0 0.020 0.70 8.374 17.6 0 0 0 0.18 0 0.023 8.691 0 0 0 0.02 0 0.017 0.70 8.654 0 0.22 0.2 0 0.18 0.022 0.68 8.572 17.6 0 0 0 0 0 0.027 8.794 17.6 0 0.2 0.02 0 0.025 0.59 8.172 17.6 0 0 0.02 0 0.023 0.65 8.477 17.6 0.22 0 0.02 0.18 0.025 0.56 8.510 0 0 0.2 0.02 0.18 0.021 0.82 8.570 0 0 0.2 0 0.18 0.010 0.60 8.791 17.6 0.22 0.2 0 0 0.025 0.84 8.399 17.6 0 0.2 0 0.18 0.020 0.77 8.650 0 0.22 0.2 0.02 0 0.035 0.74 8.189 17.6 0.22 0 0 0.18 0.029 0.38 8.611 0 0.22 0.2 0.02 0.18 0.032 0.60 8.596 17.6 0.22 0.2 0 0.18 0.031 0.42 8.516 0 0 0 0 0.18 0.021 0.70 8.979 0 0.22 0 0 0.18 0.026 0.73 8.547 17.6 0 0 0.02 0.18 0.020 0.88 8.662 17.6 0.22 0 0 0 0.023 0.43 8.209 0 0 0.2 0.02 0 0.023 0.71 8.111 0 0.22 0 0.02 0 0.020 0.52 8.156 0 0 0 0.02 0.18 0.027 0.96 8.812 17.6 0.22 0 0.02 0 0.027 1.12 7.951    51  Table 14 lists the statistical values of each variable for the response to ?max. The relative magnitude of each effect coefficient indicates its impact on the response and the sign of the value indicates whether it has a positive or negative impact on the response. Three components had a significant positive effect on ?max: C6H11FeNO7, CaCl2, and K2HPO4. The same result was obtained by Deshmukh for optimization of BG-11 components using Microcystis (Deshmukh et al, 2013). Interactive effects between variables were not significant.  Table 14: Effects and coefficients of variables estimated using a screening design for response ?max Variable Effect Standard error t-value p-value Confidence level (%) NaNO3 0.000 0.000 0.55 0.588 41.2 K2HPO4 0.044 0.009 4.67 0.000 100.0* CaCl2 0.024 0.010 2.27 0.032 96.8* C6H11FeNO7 0.365 0.104 3.52 0.002 99.8* Na2CO3 0.014 0.012 1.24 0.228 77.2 *significant at 95% level (p<0.05) Table 15 lists the statistical values of each variable for the response to Xmax. C6H11FeNO7 had a significant positive effect on Xmax. Interactive effects were not significant. Based on the results of the screening design, the independent significances of factors C6H11FeNO7, K2HPO4, and CaCl2 on ?max and Xmax warranted further investigation. Although NaNO3 and Na2CO3 were insignificant, including them in the subsequent optimization experiments can provide useful information on its effects on cell growth as nitrogen and carbon are the major elements that make up the cell.  Table 15: Effects and coefficients of variables estimated using a screening design for response Xmax Variable Effect Standard error t-value p-value Confidence level (%) NaNO3 -0.005 0.005 -1.06 0.298 70.2 K2HPO4 0.133 0.369 0.36 0.721 27.9 CaCl2 0.566 0.406 1.39 0.176 82.4 C6H11FeNO7 11.005 4.063 2.71 0.012 98.8* Na2CO3 0.325 0.451 0.72 0.478 52.2 *significant at 95% level (p<0.05) 52  An investigation of BG-11 medium components provided additional insight into the effects of pH on growth of S. elongatus. Initial pH was important in the analyses of Xmax and ?max because it influenced the initial growth rate and the duration of the lag phase, as these are tabulated in Table 13. Different pH values were observed because the component concentrations varied for each run. Initial pH readings ranged between 8 and 9 due to the alkalinity of the medium. During exponential growth phase, S. elongatus increased the alkalinity of the medium (pH 10-11), and this was likely caused by the removal of dissolved CO2 (Summerfield & Sherman, 2008). The declining phase was associated with a pH decrease to 8-9. This was caused by cell death, during which photosynthetic activity ceased and, as a result, alkalinity levels dropped.  4.1.2    Optimization of BG-11 component concentrations To determine the optimal component concentrations that maximize ?max and Xmax for the tested domain, a central composite experimental design was conducted. These results helped to explain the effects of component concentrations on ?max and Xmax. The response can be predicted for a particular set of conditions using the following equation: ? ??<+++=ki jijiijii xxxy ????0  35 Where y is the predicted response, ?i is the coefficient of the equation, ?0 is the intercept of the plane, xi and xj are coded levels of variables, and ? is the error term. Table 16 shows the levels of the experimental variables used. Table 16: Values of independent variables in different levels of the optimization design Independent variable Symbol Level -2 -1 0 1 2 NaNO3 (mM) X1 4.4 8.8 17.6 35.2 70.4 K2HPO4 (mM) X2 0.055 0.11 0.22 0.44 0.88 CaCl2 (mM) X3 0.05 0.1 0.2 0.4 0.8 C6H11FeNO7 (mM) X4 0.005 0.01 0.02 0.04 0.08 Na2CO3 (mM) X5 0.045 0.09 0.18 0.36 0.72  Table 17 tabulates the responses for all 32 runs, including 10 axial points and 6 centre points. Table 18 and Table 19 summarize the significance of each medium component on Xmax and ?max, respectively. The effects of each variable were determined by JMP statistical software (V 7, SAS 53  Institute Inc., Cary, NC, U.S.A.). Coefficients with p < 0.05 identified the significant factors (a 95% confidence level).  Table 17: Experimental results from optimization of BG-11 component concentrations Component concentrations Response NaNO3 (mM) K2HPO4 (mM) CaCl2 (mM) C6H11FeNO7 (mM) Na2CO3 (mM) ?max (1/h) Xmax (g/L) 8.8 0.11 0.1 0.01 0.09 0.0284 0.366 8.8 0.11 0.1 0.04 0.36 0.0262 0.459 8.8 0.11 0.4 0.01 0.36 0.0270 0.305 8.8 0.11 0.4 0.04 0.09 0.0227 0.334 8.8 0.44 0.1 0.01 0.36 0.0384 0.416 8.8 0.44 0.1 0.04 0.09 0.0385 0.349 8.8 0.44 0.4 0.01 0.09 0.0376 0.408 8.8 0.44 0.4 0.04 0.36 0.0280 0.525 35.2 0.11 0.1 0.01 0.36 0.0392 0.219 35.2 0.11 0.1 0.04 0.09 0.0253 0.473 35.2 0.11 0.4 0.01 0.09 0.0256 0.327 35.2 0.11 0.4 0.04 0.36 0.0226 0.249 35.2 0.44 0.1 0.01 0.09 0.0377 0.362 35.2 0.44 0.1 0.04 0.36 0.0465 0.334 35.2 0.44 0.4 0.01 0.36 0.0423 0.367 35.2 0.44 0.4 0.04 0.09 0.0330 0.445 4.4 0.22 0.2 0.02 0.18 0.0315 0.436 70.4 0.22 0.2 0.02 0.18 0.0268 0.249 17.6 0.055 0.2 0.02 0.18 0.0293 0.399 17.6 0.88 0.2 0.02 0.18 0.0488 0.375 17.6 0.22 0.05 0.02 0.18 0.0421 0.318 17.6 0.22 0.8 0.02 0.18 0.0226 0.349 17.6 0.22 0.2 0.005 0.18 0.0410 0.400 17.6 0.22 0.2 0.08 0.18 0.0322 0.461 17.6 0.22 0.2 0.02 0.045 0.0348 0.381 17.6 0.22 0.2 0.02 0.72 0.0401 0.306 17.6 0.22 0.2 0.02 0.18 0.0360 0.382 17.6 0.22 0.2 0.02 0.18 0.0345 0.382 17.6 0.22 0.2 0.02 0.18 0.0283 0.381 17.6 0.22 0.2 0.02 0.18 0.0287 0.403 17.6 0.22 0.2 0.02 0.18 0.0326 0.402 17.6 0.22 0.2 0.02 0.18 0.0319 0.374 54  Table 18: Optimization of BG-11 component concentrations for response Xmax (R2=0.91) Variable Effect Standard error t-value p-value Confidence level (%) Base 0.386 0.0130 29.56 <0.0001 >99.99* NaNO3 -0.032 0.0067 -4.75 0.0006 99.94* C6H11FeNO7 0.022 0.0067 3.24 0.0079 99.21* K2HPO4 0.018 0.0067 2.67 0.0217 97.83* Na2CO3 0.022 0.0082 2.65 0.0228 97.72* CaCl2 0.0018 0.0067 0.26 0.7973 20.27 *significant at 95% level (p<0.05) From the analysis of the response for Xmax, NaNO3 had a significant negative effect. This may have been evidence of excess NaNO3 concentration in BG-11. The significant positive effect of C6H11FeNO7 highlighted the importance of iron uptake in photosynthesis. K2HPO4 had a significant positive effect on Xmax, and this showed that its role in the nitrate uptake process was important for the growth of S. elongatus. As expected, Na2CO3 was important in providing inorganic carbon and stabilizing pH to promote cell growth. However, CaCl2 had an insignificant effect on Xmax. These results indicated that the four components, NaNO3, C6H11FeNO7, K2HPO4, and Na2CO3 may act as a major or limiting factor in culture medium, and slight changes in their concentrations can affect Xmax. The predicted response of Xmax can be described by Eq. 36, using the effects coding from Table 16 (ie. -2, -1, 0, 1, 2.). 54321max 022.0022.00018.0018.0032.0386.0 XXXXXX ++++?=  36 From the analysis of the response for ?max in Table 19, K2HPO4 had a significant positive effect. This may indicate that the K2HPO4 concentration in BG-11 was a limiting factor for the growth rate. Literature has shown that nitrogen to phosphate molar ratios (N:P) higher than 25 gave lower biomass growth among cyanobacteria (Bulgakov & Levich, 1999). In BG-11, the N:P ratio was 80, which explained why K2HPO4 was a limiting factor. Furthermore, it was believed that nitrate uptake was influenced by the availability of the phosphate ion, which is important for cellular metabolism and influences enzyme activity and the nitrate uptake process (Q. Hu, Westerhoff, & Vermaas, 2000). Hu has shown that in a phosphate-deprived culture, nitrate uptake did not occur in S. elongatus. The significant positive effect of K2HPO4 implied that phosphates were very important towards the nitrogen uptake process. This was consistent with 55  the results from the analysis of the response of Xmax by NaNO3.  Conversely, CaCl2 and C6H11FeNO7 had significant negative effects on ?max. Because it was previously shown that C6H11FeNO7 had a positive effect on Xmax, it demonstrated that iron had a more significant role in maintaining cellular reproduction, but it did not enhance the growth rate. Table 19: Optimization of BG-11 component concentrations for response ?max (R2=0.89)          *significant at 95% level (p<0.05) Na2CO3 and NaNO3 had insignificant effects on ?max. Na2CO3 had a more significant role in providing inorganic carbon, which was the major building block for cell reproduction, but had no correlation with the cellular growth rate. The predicted response of ?max can be described in Eq. 37. 54321max 0013.00021.00033.00052.000067.00326.0 XXXXX +??++=?  37 The significant factors of both Xmax and ?max had small p-values, and these were associated with large t-values because they imply that the coefficient is greater than its standard error. This confirmed that the fitted surface had maximum points. Using JMP statistical analysis, the concentrations for optimum growth were determined in Table 20. To confirm that the findings were indeed the concentrations that gave the highest Xmax and ?max, an experiment was conducted to compare cultivation of S. elongatus in unmodified BG-11 medium and optimized media. Experiments were performed in triplicates.   Variable Effect Standard error t-value p-value Confidence level (%) Base 0.0326 0.00153 21.35 <0.0001 >99.99* K2HPO4 0.0052 0.00078 6.62 <0.0001 >99.99* CaCl2 -0.0033 0.00078 -4.28 0.0013 99.87* C6H11FeNO7 -0.0021 0.00078 -2.73 0.0197 98.03* Na2CO3 0.0013 0.00071 1.7 0.1165 88.35 NaNO3 0.00067 0.00078 0.86 0.4095 59.05 56  Table 20: Components of the original and modified media Component BG-11 concentrations Concentrations for optimal Xmax (mM) Concentrations for optimal ?max (mM) NaNO3 17.6 8.8 17.6 K2HPO4 0.22 0.44 0.44 CaCl2 0.2 0.2 0.14 C6H11FeNO7 0.02 0.04 0.01 Na2CO3 0.18 0.36 0.18  Figure 22 illustrates the cell concentrations and their fitted logistic curves in both optimized media and in the original BG-11 medium with respect to time. S. elongatus reached the maximum after 210 hours of cultivation. The growth was stable during the exponential growth phase and the culture pH ranged between 8.5 and 10.5.  Figure 22: Growth in optimized media. Lines added to visualize data. It was deduced that there was no significant difference in Xmax and ?max between the three runs as the standard deviations were relatively large.  00.10.20.30.40.50.60.70.80 50 100 150 200 250 300Cell concentration (mg/mL)Time (h)originalbiomassgrowthFitted model for original mediumSeries5Series6Original mediumModified medium for optimal XmaxModified medium for optimal ?maxFitted model for ori i l iFitted model for optimal XmaxFitted model for optimal ?max57  Table 21: Summary of Xmax and ?max Medium Xmax (g/L) ?max (h-1) Unmodified BG-11 medium 0.629?0.020 0.0318?0.0009 Adjusted medium for optimal Xmax 0.650?0.047 0.0314?0.0005 Adjusted medium for optimal ?max 0.643?0.001 0.0294?0.0001 It was determined that CO2 limitation in the medium was a likely cause of this effect, which was confirmed in subsequent carbon balance experiments (Figure 37). Under air, dissolved CO2 is so severely limited that any changes in the nutrient composition had no effect on growth. If more CO2 was available, it is likely that changes in nutrient composition would have a more profound effect on cell growth. Furthermore, Xmax was larger for both adjusted media compared to the unmodified medium but the difference was not significantly great enough to draw this conclusion. Unexpectedly, the adjusted medium for optimal ?max gave a slightly lower ?max compared to the other two media. This was likely due to the logistic model fitting of the data, in which the fitting underestimated the experimental values. For these reasons, the concentrations of the unmodified BG-11 medium were used for optimization experiments in shake flasks and in the photobioreactor. 4.2    Effects of light intensity and temperature on growth of S. elongatus in shake flasks For successful growth of microalgae in culture, the environment must be conditioned to meet the intrinsic requirements of S. elongatus for optimal growth. Following the initial results, which indicated that there were insignificant differences in Xmax and ?max for the original and optimized media, the original BG-11 component concentrations were used for subsequent experiments. A central composite design was developed for the optimization of light intensity and temperature on Xmax and ?max. This allowed for a reduced number of runs, while still producing statistically reliable results. Table 22 summarizes the optimization results obtained. The biomass concentration was calculated using the biomass calibration curve. The growth rate was calculated by fitting the biomass concentrations to the logistic growth model. Similar to the growth curves from the medium optimization experiment, each curve in these experiments revealed a short lag phase (? 24 h), followed by an exponential phase (duration between 96 to 144 h) and the beginning of the stationary phase. 58  Table 22: Results from the optimization of light intensity and temperature Temperature (?C) Light intensity (?E/m2/s) ?max (h-1 ) Xmax (g/L) X6 X7 30 150 0.0510 0.485 30 79.3 0.0533 0.444 35 200 0.0495 0.450 30 150 0.0492 0.485 23 150 0.035 0.335 25 100 0.0347 0.381 30 150 0.0493 0.485 30 220.7 0.0438 0.430 25 200 0.0337 0.358 37 150 0.0397 0.356 35 100 0.0499 0.505 30 150 0.0516 0.497 30 150 0.0507 0.497  Table 23 shows the significance of each factor on ?max. Only temperature had a significant positive effect on ?max. The 2nd order temperature effect showed that concavity was present on the response surface, i.e. there was an optimum temperature at 33?C. Interestingly, light intensity did not have a significant effect on ?max. This may indicate that light is not a limiting factor in conditions between 79.3 and 220.7 ?E/m2/s in the shake flasks. No significant interactive effects between temperature and light were present in the model. Table 23: Optimization of light intensity and temperature for response ?max Variable Effect Standard error t-value p-value Confidence level (%)  0.0504 0.00171 29.39 <0.0001 >99.99* Temperature*temperature -0.00669 0.00145 -4.60 0.0025 99.75* Temperature 0.00463 0.00135 3.42 0.0112 98.88* Light intensity -0.00185 0.00135 -1.37 0.2137 78.63 Light intensity*light intensity -0.00117 0.00145 -0.81 0.4454 55.46 Light intensity*temperature 0.000150 0.00192 0.08 0.9396 6.04 *significant at 95% level (p<0.05) Figure 23 displays the response surface for ?max. Within the tested range of conditions, ?max were achieved at 33?C for the tested range of light intensities. Below 30?C, the cells experienced a 59  dramatic drop in ?max. As previously mentioned, light intensity had no significant effect on ?max. A slight decrease in ?max was observed as the light intensity increased, which may indicate the presence of photoinhibition in the cells; however, this was not statistically significant. The highest ?max in this response curve was 0.0519 h-1, and this was attained at a temperature of 33?C and light intensity of 120 ?E/m2s. The model is represented by the following equation, where X6 is the effects coding for temperature (ie. -1.414, -1, 0, 1, 1.414): 266max 00669.000463.00504.0 XX ?+=?  38   Figure 23: Response curve for ?max as a function of temperature and light intensity (R2=0.83) Table 24 shows the significance of each factor on Xmax. Similar to the response curve for ?max, Xmax depends on the temperature. It also displays the same trend of concavity as observed in the response for ?max.   60  Table 24: Optimization of light intensity and temperature for response Xmax Variable Effect Standard error t-value p-value Confidence level (%)  0.490 0.0142 34.59 <0.0001 >99.99* Temperature*temperature -0.0641 0.0120 -5.34 0.0011 99.89* Temperature 0.0307 0.0112 2.74 0.0288 97.12* Light intensity*light intensity -0.0184 0.0120 -1.53 0.1702 82.98 Light intensity -0.0122 0.0112 -1.09 0.3110 68.90 Light intensity*temperature -0.008 0.0158 -0.51 0.6289 37.11 *significant at 95% level (p<0.05) The model is represented by the following equation: 266max 0641.00307.0490.0 XXX ?+=  39  Figure 24 is the response surface model of Xmax. Photoinhibition was observed to take effect at light intensities greater than 150 ?E/m2/s. The highest Xmax was estimated to be 0.496 g/L, at a temperature of 33?C and a light intensity of 120 ?E/m2/s. Similar to the ?max response curve, the response from temperature showed concavity. Xmax decreased drastically at temperatures below 30?C. A similar trend was observed by Kajiwara et al. (1997). Although optimization of temperature was not conducted, Kajiwara et al. used light intensities between 40 ?E/m2/s and 135 ?E/m2s and a mixture of 5% CO2 and air to cultivate S. elongatus. They found that 108 ?E/m2/s at 30?C gave the highest ?max and Xmax, 0.048 h-1 and 1.1 g/L, respectively. This apparent similarity in the optimal light intensities from two independent experiments hinted that responses of ?max and Xmax from changes in light intensity were not significantly affected by other factors. 61   Figure 24: Response curve for Xmax as a function of temperature and light intensity (R2=0.85) From these results, the highest ?max and Xmax predicted in the shake flask experiments were 0.0519 h-1 and 0.496 g/L, respectively. Both values were attained at a light intensity of 120 ?E/m2/s and a temperature of 33?C. These optimized parameters were subsequently used for the optimization experiments in the flat-plate photobioreactor.   62  5    Mass Transfer Studies in 2D Airlift Photobioreactor 5.1    Determination of Bodenstein number and gas hold-up in photobioreactor To determine if axial liquid mixing was sufficient for flow rates between 0.25 L/min and 2 L/min in the flat-plate photobioreactor, Bo of the overall photobioreactor was calculated using Equations 29-31. Table 25 lists Bo and the mixing time for the given gas flow rates.  Table 25: Bo and tm Gas flow rate (L/min) Bo tm (s) 0.25 48.7 4.0 0.5 37.1 3.2 0.75 17.8 2.8 1 11.9 2.6 1.25 11.5 2.4 1.5 11.5 2.3 1.75 11.3 2.2 2 11.3 2.1  As the gas flow rate increased from 0.25 L/min to 1 L/min, Bo decreased significantly. Beyond 1 L/min, Bo appeared to stabilize just above 11. This indicated that no further significant improvement in axial liquid mixing was achieved at gas flow rates above 1 L/min. It also was deduced that significant axial liquid mixing was present at flow rates above 0.5 L/min, but the photobioreactor did not have characteristics of a perfectly mixed reactor. Instead, the photobioreactor had Bo values typical of internal airlift reactors in the range between 10 and 60 (Gavrilescu & Tudose, 1999). The mixing time tm necessary to achieve 95% homogeneity in the reactor was then calculated from Bo using equation 27. tm was observed to decrease from 4 s to 2.1 s as gas flow rate increased from 0.25 L/min to 2 L/min. tm was also observed to reach an asymptotic value of approximately 2, indicating that gas flow rates beyond 2 L/min would have insignificant improvement liquid in mixing. Results from the determination of Bo and tm will provide insight into the volumetric mass transfer characteristics of dissolved CO2 in the photobioreactor, and will determine if the tested operating parameters are sufficient for cell growth. 63  Measurement of the gas holdup was also essential for understanding its effects on mixing and mass transfer. The gas holdup was measured only at the riser section and the gas entrainment space above it. Figure 25 showed that the gas holdup increased linearly with the air flow rate. The total gas holdup reached 6.5% at 2 L/min, which was the maximum gas flow rate investigated. The linear relationship suggested that a greater gas holdup can be attained at higher air flow rates, and this could result in greater liquid circulation and greater gas-liquid interfacial area available for mass transfer of CO2 and O2 (Chisti & Moo-Young, 1988). However, experiments at higher flow rates could not be conducted due to operational limitations of the experimental setup.  Figure 25: Relationship between gas holdup and gas flow rate 5.2    Aeration by CO2-enriched air for mass transfer characterization in 2D photobioreactor To optimize the aeration conditions for microalgal biomass production in an airlift photobioreactor, the effect of aeration rate on CO2 mass transfer was investigated under given conditions. To understand whether cell growth was CO2-limited under the experimental conditions, it was crucial to first analyze the mass transfer properties of the photobioreactor. The relationship between the volumetric mass transfer coefficient kLa and aeration rate was studied and parameters for enhanced kLa of CO2 were determined. Air enriched with 5% or 10% (vvm) CO2 was supplied for the experiments at rates of 0.5, 1, 1.5 and 2 L/min. y = 0.0289x + 0.0011R? = 0.97130.00%1.00%2.00%3.00%4.00%5.00%6.00%7.00%0 0.5 1 1.5 2 2.5Gas holdupAir flow rate (L/min)64  The following equation represents the CO2 transfer rate into the culture medium: )( CCakdtdCSL ?=  40 where C (mol/m3) is the dissolved CO2 concentration (DCO2) in the bulk liquid at any time t (s), and CS is the saturation concentration of DCO2. CS was obtained by determining the maximum concentration recorded in the experiments. In this experiment, the dynamic method is used to measure kLa. Integration of equation 1 yields an equation where kLa can easily be calculated: ????????=ssoL CCCCtak ln1 41 where Co is the initial DCO2 concentration at time t=0. The same equations can be used to calculate kLa values for dissolved oxygen (DO2). Measurements of DO2 were important in this study because they were required for calculations of kLa for DCO2 using the Rand and Marshall equations (Eq. 3). As an example, Figure 23 shows the concentrations of DCO2 and DO2 at 33?C, constant gas flow rate of 0.5 L/min, and at 5% CO2. These profiles were important for determining the saturation concentrations and volumetric mass transfer coefficients of DCO2 and DO2 for the specified conditions. The data was then regressed to fit the model described in Eq. 41 to yield kLa and CS.   65   Figure 26: DCO2 and DO2 at 33?C, constant gas flow rate of 0.5 L/min, and at 5% CO2 Figure 27, which shows kLa values for DCO2 at two temperatures and two inlet CO2 concentrations, indicated that kLa values increased as the temperature increased from 23?C to 33?C. kLa values for DCO2 with air were not available as CO2 concentrations were too low to be detected by the CO2 electrode. A rise in temperature led to a decrease in surface tension of the liquid, generating smaller bubbles and higher kLa values (Bird, 2002). kLa values for DO2 showed no significant change as inlet CO2 concentration increased.   Figure 27: CO2 kLa at 23?C and 33?C for a constant gas flow rate of 0.5 L/min   00.511.522.533.540510152025300 500 1000 1500 2000D O2(mg/L)D CO2(mg/L)Time (s)dCO2DODCO2DODCO2DO20.00000.00020.00040.00060.00080.00100.00120.00140.00160.00180.00200 10 20 30 40k Lafor DCO2in BG-11 medium (1/s)Temperature (?C)5% CO210% CO25% CO210% CO266  Figure 28 plots kLa values at various gas flow rates and CO2 concentrations at a constant temperature of 33?C. As gas flow rate increased, kLa values rose. This was expected due to the additional mixing power the gas provided to the medium. The kLa values for 5% CO2 were lower than the values at 10% CO2 by approximately 10-24%. The dissociation of carbonic acid decreased the surface tension of the water, reducing the average bubble diameter and increasing total surface area (Beltran et al, 1998; Dos Santos & Levin, 2010). In other words, increasing the inlet CO2 % increased kLa.   Figure 28: CO2 kLa at various gas flow rates and a constant temperature of 33?C For the same conditions, kLa values of DO2 in Figure 29 increased as inlet gas flow rate increased. These observations indicate that higher kLa values were possible at higher gas flow rates. The same relationship between the gas flow rate and kLa was also observed by Kazim (2011). However, the effect of shear stress on cells resulting from these higher gas flow rates was never previously investigated for this reactor configuration. 0.00000.00050.00100.00150.00200.00250.00300.00350.00400.00450.00500 0.5 1 1.5 2 2.5k Lafor DCO2in BG-11 medium (1/s)Gas flow rate (L/min)5% CO210% CO25% CO210% CO267   Figure 29: Oxygen kLa at various gas flow rates and a constant temperature of 33?C Figure 29 showed that as the gas flow rate increased, kLa for oxygen also increased. There was a significant difference in kLa between 10% CO2 and the other runs at flow rates below 1.5 L/min, but then kLa converged to similar values at 1.5 and 2 L/min. This was due to the higher gas flow rates that promoted greater liquid mixing by gas bubbles in the medium. In Table 26, saturation concentrations for both DCO2 dropped as the temperature increased, which complied with literature results. This can be explained by the changes in water crystallinity. Elevated temperatures increase the kinetic energy of water molecules, disturbing its crystalline structure and allowing gas molecules to escape, thereby lowering gas solubility (Bird, 2002). At both temperatures of 23?C and 33?C, CS values at 10% inlet CO2 were almost double that of 5% inlet CO2. Table 26: Saturation concentration for CO2 at 23?C and 33?C and a constant gas flow rate of 0.5 L/min Temperature CO2 concentration  5% CO2 10% CO2 23?C 79.9 mg/L 151.9 mg/L 33?C 62.0 mg/L 116.3 mg/L  In Figure 30, CS values for DCO2 at 5% and 10% CO2 did not change significantly at the tested range of gas flow rates. Similarly, CS values for DO2 were relatively constant. 0.00000.00500.01000.01500.02000.02500 0.5 1 1.5 2 2.5k Lafor DO2in BG-11 medium (1/s)Gas flow rate (L/min)air5% CO210% CO2air5% CO210% CO268   Figure 30: Saturation concentration of CO2 at various gas flow rates and a constant temperature of 33?C As shown in Figure 31, when varying the inlet CO2 concentrations and gas flow rates, DO2 remained relatively constant. Calculations of superficial gas velocity showed that the reactor was in bubbly flow regime at the tested gas flow rates.    Figure 31: Saturation concentration of O2 at various gas flow rates and a constant temperature of 33?C Like the dependence of kLa on the gas flow rate, the gas hold up was found to be directly proportional to the gas flow rate at the tested parameters, as shown in Figure 32.  0.020.040.060.080.0100.0120.0140.00 0.5 1 1.5 2 2.5C sfor DCO2in BG-11 medium (mg/L)Gas flow rate (L/min)5% CO210% CO25% CO210% CO20.000.501.001.502.002.503.003.504.004.500 0.5 1 1.5 2 2.5C sfor DO2in BG-11 medium (mg/L)Gas flow rate (L/min)air5% CO210% CO2air5% CO210% CO269   Figure 32: Superficial gas velocities and gas hold-ups at various flow rates A summary of kLa values for CO2 is tabulated in Table 27. Errors were obtained by calculating the standard deviation of the experimental data. Experiments were performed in duplicates. Table 27: Comparison of kLa values at different inlet gas flow rates and CO2 %. T=33?C Flow rate (L/min) kLa values for O2-BG11 medium system (h-1)  0.04% CO2 5% CO2 10% CO2 0.5 L/min 0.0088 ? 0.0004 0.0084 ? 0.0007 0.0027 ? 0.0003 1 L/min 0.0103 ? 0.0003 0.0111 ? 0.0000 0.0062 ? 0.0002 1.5 L/min 0.0147 ? 0.0004 0.0146 ? 0.0003 0.0161 ? 0.0007 2 L/min 0.0186 ? 0.0006 0.0192 ? 0.0006 0.0216 ? 0.0014  kLa values for CO2-BG11 medium system (h-1)  0.04% CO2 5% CO2 10% CO2 0.5 L/min - 0.0015 ? 0.0001 0.0017 ? 0.0001 1 L/min - 0.0019 ? 0.0003 0.0021 ? 0.0001 1.5 L/min - 0.0029 ? 0.0006 0.0036 ? 0.0008 2 L/min - 0.0036 ? 0.0001 0.0041 ? 0.0002  A rise in temperature resulted in increases of CO2 and O2 kLa in the BG-11 medium and drops in the maximum dissolved CO2 and O2 concentrations. An increase in gas flow rate also led to an increase in both kLa values due to the higher rate of bubbles and agitation supplied to the medium. Increasing CO2 concentration led to an increase in kLa value, possibly because of the 0.00%0.50%1.00%1.50%2.00%2.50%3.00%3.50%00.0020.0040.0060.0080.010.0120.0140.0160.0180.020 0.5 1 1.5 2 2.5Gas holdup %Superficial gas velocity (m/s)Gas flow rate (L/min)superficial gas velocitygas holdup70  formation of carbonic acid and the associated reduction in the gas-water surface tension. Experimental results suggest that operation of the airlift photobioreactor at 33?C, 5% CO2, and high gas flow rates will yield high volumetric mass transfer coefficients in the BG-11 system. However, light intensity and temperature optimization experiments conducted earlier in shake flasks suggested that studies at light intensities lower than 120 ?E/m2/s should be carried out to determine its effects on cell growth. Although increasing the gas flow has been shown to improve kLa values, providing high flow rates is energy intensive. Given the Bo values in Table 25, selecting a flow rate of 1 L/min provides the necessary CO2 transfer while sustaining good axial liquid mixing. This is equivalent to 1 vvm (gas volume flow per unit of liquid volume per minute), which is commonly used in airlift photobioreactors (Hong et al, 2013). Keeping the flow rate and temperature constant at 1 L/min and 33?C will allow for a reduced number of runs and greater focus on the effects of light intensity and inlet CO2 concentration on the growth of S. elongatus.  71  6    Optimization of Xmax and ?max in photobioreactor Following the initial results, which indicated that a temperature of 33?C yielded the highest Xmax and ?max and a gas flow rate of 1 L/min provided sufficient CO2 mass transfer, an experiment designed and conducted to test a range of growth conditions. The effect that light intensity and CO2 concentration had on Xmax and ?max in the photobioreactor was analyzed. Optimal reactor conditions that maximized total biomass yield were then determined. A factorial experimental design was developed for this study. The range of each variable was chosen based on results from the previous light and temperature experiments and the mass transfer experiment. Light intensities greater than 120 ?E/m2/s induced photoinhibition. Gas flow rates of at least 1 L/min were shown to provide good CO2 mass transfer. Conditions that maximize Xmax and ?max were therefore sought. Table 28 outlines the reactor conditions studied. The gas flow rate was kept constant at 1 L/min and the temperature was also maintained at 33?C. The results from this experiment on growth in the photobioreactor are presented in Table 29. Growth curves are plotted in Appendix C. Table 28: Factorial design of reactor conditions Independent variable Symbol Level -1 0 1 Light intensity (?E/m2/s) X7 60 n/a 120 CO2 % X8 0.04 5 10 The light intensity had a relatively insignificant effect on both ?max and Xmax in comparison to the inlet CO2 concentration. As the CO2 concentration increased from 0.04% (air) to 10%, ?max decreased by 62%, which was caused by the increasing acidity of the medium. However, the highest Xmax was observed at 5% CO2 cultivation. Further increasing the CO2 concentration above 5% appeared to inhibit cell growth because the cells experienced a lower pH (6.42-6.65) in the lag phase caused by formation of carbonic acid. That was because S.elongatus PCC 7942, an alkaliphilic cyanobacterium, exhibits the optimum growth at pH values between 7.0 and 9.0 (Billini, 2008). 72  Table 29: Results of full factorial design in airlift photobioreactor Light intensity (?E/m2/s) CO2 % Xmax (g/L) ?max (1/h) X7 X8   60 0.04 0.316 0.0504 120 0.04 0.405 0.0373 60 5 1.066 0.0255 120 5 1.006 0.0234 60 10 0.765 0.0189 120 10 0.879 0.0194  Table 30 shows the significance of each factor on ?max. Only CO2 % had a significant effect on ?max. The negative coefficient indicates that ?max decreases as CO2 % increases. Table 30: Optimization of light intensity and CO2 % for response ?max in photobioreactor (R2=0.85) Variable Effect Standard error t-value p-value Confidence level (%) Base 0.02915 0.00257 11.37 0.0015 99.85* CO2 % -0.0124 0.00314 -3.93 0.0293 97.07* Light intensity (?E/m2/s) -0.00245 0.00257 -0.96 0.410 59.00 *significant at 95% level (p<0.05) The statistical model for ?max was represented by the following equation in terms of CO2 %, where X8 is the effects coding for CO2 % (ie. -1, 0, 1): 8max 0124.002915.0 X?=?  42  73   Figure 33: Response curve for ?max as a function of light intensity and CO2 % (R2=0.85) Figure 33 displays the response curve for ?max. The light intensity had no significant effect on the response. On the other hand, increasing the CO2 concentration to 5% and 10% reduced ?max because excess carbonic acid in the medium dropped the pH below 7 for the first 60-70 hours, as shown in Figure 34. S. elongatus has been known to grow best at pH values between 7 and 9, and pH values below this range caused deterioration of cells and formation of toxic by-product (Billini, 2008). The low initial pH values at CO2-enriched conditions explained why low ?max values were observed. Secondly, any abrupt changes in the conditions that the cells were exposed to as they were transferred from inoculum conditions to shake flask and PBR conditions slowed growth. The more extreme the change, the longer it took for the cells to adapt to the new environment. 74   Figure 34: Culture pH at various conditions in the photobioreactor at 33?C and 1 L/min gas flow rate The same data were also used to develop another statistical model for Xmax. Table 31 tabulates the parameter estimates along with the corresponding confidence levels. Table 31: Optimization of light intensity and CO2 % for response Xmax in photobioreactor (R2=0.44) Variable Effect Standard error t-value p-value Confidence level (%)  0.7395 0.123 6.01 0.0092 99.1* CO2 % 0.2308 0.151 1.53 0.223 77.7 Light intensity (?E/m2/s) 0.0238 0.123 0.19 0.859 14.1  None of the response variables were significant at 95% confidence. Both factors had no significant effects on Xmax, and the relationship between the factors and Xmax was not linear. As a result, it was not possible to develop a statistical model for a first order response surface.  Figure 35 plots the data of Xmax for CO2 % of 0.04, 5, and 10% and light intensities of 60 and 120 ?E/m2/s. 0123456789100 50 100 150 200 250 300 350 400pHTime (h)33 deg C, 60 umol/m2/s, air @ 1 L/min33 deg C, 60 umol/m2/s, 5% @ 1 L/min33 deg C, 60 umol/m2/s, 10% @ 1 L/min33 deg C, 120 umol/m2/s, air @ 1 L/min33 deg C, 120 umol/m2/s, 5% @ 1 L/min33 deg C, 120 umol/m2/s, 10% @ 1 L/min60 ?E/m2/s, air60 ?E/m2/s, 5% CO260 ?E/m2/s, 10% CO2120 ?E/m2/s, air120 ?E/m2/s, 5% CO2120 ?E/m /s, 10% CO275   Figure 35: Xmax for 0.04, 5 and 10% CO2, and 60 and 120 ?E/m2/s light intensity Figure 36 plots the data of ?max for the same CO2 % and light intensities. The fitted model provided a good representation of ?max when inlet CO2 % varied.  Figure 36: Specific growth rate ?max for 0.04, 5 and 10% CO2, and 60 and 120 ?E/m2/s light intensity 6.1 Effects of light intensity on growth The first observation made was that light intensity had a small effect on Xmax and ?max. An increase from 60 to 120 ?E/m2/s resulted in an increase of Xmax of 28% in air, a drop of 5.5% in 0.0000.2000.4000.6000.8001.0001.2000 2 4 6 8 10 12X max(mg/mL)CO2%60 ?E/m2/s120 ?E/m2/s60 ?E/m2/s120 ?E/m2/s0.00000.01000.02000.03000.04000.05000.06000 2 4 6 8 10 12? max(h-1)CO2%60 ?E/m2/s120 ?E/m2/sPredicted umax for 60 and 120 uE/m2/s60 ?E/m2/s120 ?E/m2/sPredicted ?max for 60 and 120 ?E/ 2/s76  5% CO2, and an increase of 14.8% in 10% CO2. This result confirmed the findings of the light intensity experiments in shake flasks, in which the light-limiting regime was observed at light intensities below 120 ?E/m2/s. The same increase in light intensity decreased ?max by 25.9% in air, 8.3% in 5% CO2, and increased ?max by 2.4% in 10% CO2. ?max seemed to drop as CO2 % decreased. Furthermore, it appeared that under 120 ?E/m2/s, cell growth was severely limited if dissolved CO2 was at a limiting concentration of 0.24 ppm CO2. No research has been conducted on the effects of CO2 % on ?max under various light intensities for S. elongatus PCC 7942 so a comparison with published results could not be made. To ensure reproducibility of results, a duplicate run for 5% CO2 at 60 ?E/m2/s was conducted. As shown in Table 32, the percent change in the measured ?max was 5%. The replicate results generally agreed with the original results. Table 32: Reproducibility of growth rate data from the photobioreactor Conditions Run Xmax (g/L) ?max (h-1) 33?C, 5% CO2, 1 L/min, 60 ?E/m2/s 1 1.066 0.0255 Replicate 0.978 0.0239   6.2    Effects of inlet CO2 concentration on growth  A change in CO2 concentration at the investigated parameters was shown to have a greater effect on Xmax and ?max when comparing with a change in light intensity. A higher CO2 concentration resulted in a lower initial pH in the medium. This caused cells to grow at a slower rate. As cells adjusted to the more acidic environment, the growth rate began to accelerate by the 3rd day. It was at this point where growth curves diverged from each other. The highest Xmax observed was 1.066 g/L at conditions of 120 ?E/m2/s and 5% CO2. The lowest Xmax observed was 0.316 g/L at 60 ?E/m2/s and in air. A high ?max was observed when S. elongatus was cultivated in air. The relatively lower ?max was likely caused by the formation of carbonic acid, which lowered the pH, and this was only present in the growth runs with CO2-enriched air. To investigate which conditions made CO2 concentration a limiting factor on cell growth, the logistic equation (Eq. 20), CO2 mass balance equation (Eq. 21), and Rand and Marshall equation 77  (Eq. 3) were solved and graphed using Polymath 6.0. Appendix D details the defining equations, parameters, and initial values for each experiment. Figure 37 plots the total carbon content in medium and the total carbon content in biomass at 1 L/min air flow rate, 33?C and 60 ?E/m2/s. All of the CO2 absorbed into the medium was taken up by the algae cells, indicating that the growth was limited by the low availability of CO2. The total mass of carbon takes into account the CO2 absorbed from gas bubbles, CO2 originally saturated in the medium at the beginning of the experiment, sodium carbonate in the medium, and the carbon from biomass. A closer inspection showed that the total carbon content in the medium slightly exceeded the total amount of carbon in the biomass, indicating that an extremely small amount of dissolved CO2 in the medium was present throughout the cultivation.  Figure 37: Carbon balance from S. elongatus growth in air at 1 L/min, 33?C, 60 ?E/m2/s A similar trend was observed for 5% CO2 in Figure 38. There was a greater difference in mass of carbon between the two curves because of the higher dissolved CO2 concentration in the medium. This reinforced the hypothesis that CO2 % was a limiting factor of growth for cultivations with air. 00.020.040.060.080.10.120.140.160 50 100 150 200Total carbon content (g/L)Time (h)In mediumIn biomassIn medium+biomassIn bioma s78   Figure 38: Carbon balance from S. elongatus growth in 5% CO2 at 1 L/min, 33?C, 60 ?E/m2/s By comparing the carbon content of 5% in Figure 38 and 10% CO2 in Figure 39, a few observations were made. Increasing the inlet CO2 concentration from 5% to 10% increased the amount of dissolved CO2 but lowered Xmax. As concluded from observations made in Figure 33, dissolved CO2 was available in excess but the higher acidity limited cell growth. The effect of excess carbonic acid was more pronounced beyond 5% CO2, leading to the hypothesis that an optimal CO2 % must exist somewhere close to 5%. At this optimal concentration, the highest Xmax may be achieved, assuming all other parameters are held constant.  00.10.20.30.40.50.60 100 200 300 400Total carbon content (g/L)Time (h)In mediumIn biomassIn medium+biomassIn biomass79   Figure 39: Carbon balance from S. elongatus growth in 10% CO2 at 1 L/min, 33?C, 60 ?E/m2/s A review of literature has shown similar trends based on research using the same freshwater strain and Chlorella vulgaris. Kajiwara (1997) showed that 5% CO2 yielded the highest Xmax (0.98 g/L) amongst growth runs of 0.04%, 5%, 10%, and 15% CO2. However, because Kajiwara utilized HEPES buffer in the BG-11 medium, ?max was about 0.044 h-1, which was 73% higher than ?max obtained in this experiment. This can be attributed to the alkalinic pH and shorter lag phase in Kajiwara?s research. In Sasi?s research with Chlorella vulgaris, the highest Xmax was also recorded at 5% CO2 (Sasi, 2009). Any further increase in CO2 % led to a drop in biomass yield. This was evidenced that inhibition of cell growth was a common response of microalgae at high dissolved CO2 concentrations and it was important to determine the optimal inlet CO2 % for economical cultivations. However, growth of S.elongatus at other CO2 concentrations (between 0.04 and 10%) and other gas flow rates (between 1 and 2 L/min) merits further investigation as the concentration of dissolved CO2 can be adjusted by changing both parameters. This may lead to improved Xmax and ?max values. 6.3    Effect of 12:12 h light/dark cycle on S. elongatus growth To investigate whether the introduction of light/dark cycles can potentially improve cell growth of S. elongatus compared to conditions in continuous lighting, an experiment was implemented for the purpose of culturing cells under 12:12 h light/dark cycle. The experimental conditions 00.050.10.150.20.250.30.350.40.450 100 200 300 400 500Total carbon content (g/L)Time (h)In mediumIn biomassIn medium+biomassIn biomass80  were chosen based on the conditions that yielded the best biomass growth from the previous photobioreactor runs; ie. 33?C, 60 ?E/m2/s, 5% CO2 @ 1 L/min. Figure 40 showed that Xmax for growth under continuous lighting was 64.6% higher than Xmax for growth under the light/dark cycle. Similarly, ?max for growth under continuous lighting was 27.0% higher than ?max for growth under the light/dark cycle.  Figure 40: Comparison of growth in continuous light and 12:12 light/dark cycle at 33?C, 60 ?E/m2/s, 5% @1 L/min Table 33 lists Xmax and ?max values for each condition. It appeared that the light/dark cycle had a significant negative effect on cell growth using the response variables as a basis for comparison. Table 33: Results for growth under continuous light and 12:12 light/dark cycle Light intensity (?E/m2/s) CO2 % Xmax (g/L) ?max (h-1) X7 X8 60  - Continuous 5 1.066 0.0255 60  - 12:12 light/dark 5 0.648 0.0201  Nonetheless, it was also important to analyze the energy consumption in cultivating the cells under the two difference lighting conditions to see whether energy savings can be incurred without compromising cell growth. Normalizing input power per gram biomass gives: 00.20.40.60.811.20 50 100 150 200 250 300 350Biomass concentration (g/L)Time (h)33 deg C, 60 umol/m2/s, 5% @ 1 L/min33 deg C, 60 umol/m2/s, 5% @ 1 L/min, Light dark cycleContinuous lightingLight/dark cycle81  Table 34: Energy input required for growth at continuous light and 12:12 light/dark cycle Light intensity (?E/m2/s) Input power/g biomass (W/g) 60  - Continuous 30.0 60  - 12:12 light/dark 24.7 It appeared that growth under light/dark cycle was 17.7% more energy efficient in producing biomass. To put this in perspective, using a 12:12 light/dark cycle saved 17.7% in energy usage but resulted in a drop of 39.2% for Xmax and 21.3% for ?max. This probably would not be an economical trade-off. For the purpose of cultivating S. elongatus to achieve a high biomass yield under economical conditions, continuous light would be recommended.   82  7    Conclusion A response surface method was adequate for determining the effects of each BG-11 medium component in order to maximize growth of S. elongatus PCC 7942. At the investigated range of component concentrations, results from an optimization experiment to determine the specific concentrations of these components generated the following response surface model for Xmax (g/L) in terms of effects coding Xi:         5421max 022.0022.0018.0032.0386.0 XXXXX +++?=  36 as well as for ?max:          432max 0021.00033.00052.0326.0 XXX ???=? 37 Where: X1 = NaNO3 X2 = K2HPO4 X3 = CaCl2 X4 = C6H11FeNO7 X5 = Na2CO3 Comparison of cell growth in unmodified BG-11 medium and optimized media showed that there was no significant increase in Xmax and ?max, indicating that the unmodified medium was suitable enough for high algal productivity. Studies conducted in a shake flask scale demonstrated that an optimal temperature of 33?C and an optimal light intensity of 120 ?E/m2/s gave the highest Xmax, 0.496 g/L, and the highest ?max, 0.0519 h-1, for the tested range of growth parameters. Below 30?C, the cells experienced a remarkable drop in both Xmax and ?max. Light intensity had little effect on both responses. Cells experienced a slight decrease in both responses as the light intensity increased, which may possibly indicate the presence of photoinhibition at light intensities greater than 150 ?E/m2/s. Equations  represents the statistical model for Xmax and ?max, respectively. 266max 0641.00307.0490.0 XXX ?+=        38 266max 00669.000463.00504.0 XX ?+=? 39 83  Where X6 is the effects coding of temperature. These optimized parameters were subsequently used for the optimization experiments in the flat-plate photobioreactor. To identify the best operating range of inlet gas flow rates in the photobioreactor, it was important to analyze changes in axial mixing in terms of the Bodenstein number Bo. As the gas flow rate increased from 0.25 L/min to 1 L/min, Bo dropped from 48.7 to 11.9. Beyond 1 L/min, Bo was observed to stabilize at approximately 11, meaning that any further increase in gas flow rate had no significant improvement in axial mixing. To determine the photobioreactor conditions in which cell growth was CO2-limited, experiments which studied the effects of inlet CO2 %, gas flow rate, and temperature on the volumetric mass transfer coefficient kLa were conducted. In general, increases of each factor enhanced kLa values. kLa values increased 10-24% when the inlet CO2 % doubled from 5% to 10%. Ramping up the gas flow rate improved gas-liquid mixing and consequently kLa values. A rise in temperature from 23?C to 33?C resulted in 19-31% increase in kLa due to the decrease in surface tension of the gas-liquid interface, creating smaller bubbles. Additionally, increasing the inlet CO2 concentration from 5% to 10% doubled the saturation concentration of dissolved CO2. Similar changes in kLa values of dissolved O2 were also observed. These results led to the decision to cultivate algae at the optimal temperature of 33?C and gas flow rate of 1 L/min.  Results from the algae cultivation experiments in the photobioreactor presented several significant discoveries. As the CO2 concentration increased from 0.04% (air) to 10%, ?max dropped by 62%, and this was caused by the increasing acidity of the medium which hindered cell growth. However, the highest Xmax was observed at 5% CO2 for each investigated light intensity. In comparison, Xmax at 0.04% and 10% CO2 were noticeably less. Cell growth using air was CO2-limited, which was confirmed by performing a mass balance on carbon; cultivations were not CO2-limited at higher CO2 %. Formation of excess carbonic acid at 10% CO2 lowered pH and hindered reproduction. A change in light intensity did not significantly affect the responses. To attain the highest biomass concentration at a relatively fast rate, it is recommended to operate the photobioreactor at 33?C, 120 ?E/m2/s, and 1 L/min at 5% CO2. Including a 12 hour dark phase to the photobioreactor did not improve Xmax or ?max. However, when normalizing the input power per gram of biomass produced, growth under light/dark cycle 84  was 17.7% more energy efficient in producing biomass when compared to growth in continuous illumination. Yet this resulted in a drop of 39.2% for Xmax and 21.3% for ?max, which was not desirable from a productivity standpoint. Ultimately, this work has demonstrated that there is a future for commercial production of high-value microalgal products. By utilizing enhanced photobioreactor design and proper selection of operating parameters, the growth rate of S. elongatus can be significantly enhanced. Additional research is still required to realize the full potential of microalgae for production of recombinant product in a commercial scale. Nevertheless, continued studies on the effects of various parameters on S. elongatus growth will lead to more breakthroughs that can eventually establish economical production of microalgae-derived high-value products. Ecological engineering such as this serves as a prime example of the current paradigm shift towards the ?green? movement and places algae at the forefront of clean biotechnology.  85  8    Future Work The use of flashing lights to increase the frequency of light and dark cycles in photosynthesis and possibly biomass productivity can be investigated by using LEDs. The flashing light effect has the potential to increase growth rate by better matching the time scales of photosynthetic reactions or save energy by minimizing wasted irradiation (Lunka, 2013). Several areas have been identified which have the potential to improve the cultivation economics. Additional research on cultivation under sunlight or LEDs needs to be explored. Although limited by weather and geographic location, availability of sunlight is an attractive alternative to using CFL bulbs for cost savings. LEDs are also known to save in energy costs as they require relatively low power. Another approach is to use the microalgae to fixate CO2 from pre-treated flue gases. Power-plant flue gas can serve as a source of CO2; however the costs in removing toxic SOx and NOx gases harmful to microalgae are a subject of investigation. To assess the viability of using microalgae as a vehicle for production of value-added compounds, a trial run with a recombinant strain that produces an enzyme with commercial potential should be conducted. Examples of test enzymes may include cellulases and research enzymes with high profit margins. Finally, the possibility of using Aspen Plus to conduct the economic optimization of the photobioreactor conditions should be explored. Aspen Plus can be used to generate mass, energy, and utility balances over a range of cultivation conditions. In terms of scale-up production, design of a photobioreactor is crucial because a high cultivation volume is desired without compromising biomass productivity. These data can then be exported to Microsoft Excel where an economic analysis can be conducted. Economic optimization could be performed within Aspen Plus utilizing the optimization tool, and as a result a complete set of optimum process conditions could be determined. Improved optimization of reactor conditions would reduce cultivation costs.   86  Bibliography  Alabi, A., Tampier, M., & Bibeau, E. (2009). Microalgae technologies and processes for Biofuels/Bioenergy production in British Columbia. British Columbia: Seed Science Ltd.  Andersen, R. A. (2005). Algal culturing techniques Academic Press.  Beltran, F., Fernandez, L., Alvarez, P., & Rodriguez, E. (1998). Comparison of ozonation kinetic data from film and Danckwerts theories. Ozone: Science & Engineering, 20(5), 403-420.  Ben-Amotz, A. (2008). Bio-fuel and CO2 capture by algae. Seambiotic.  Benemann, J. R. (1997). CO2 mitigation with microalgae systems. Energy Conversion and Management, 38, Supplement(0), S475-S479.  Benemann, J. R., & Oswald, W. J. (1996). Systems and economic analysis of microalgae ponds for conversion of CO2 to biomass. Pittsburgh, PA: Pittsburgh Energy Technology Center.  Billini, M., Stamatakis, K., & Sophianopoulou, V. (2008). Two members of a network of putative Na+/H+ antiporters are involved in salt and pH tolerance of the freshwater cyanobacterium Synechococcus elongatus. Journal of Bacteriology, 190(19), 6318-6329. Bird, R. B., Stewart, W. E., & Lightfoot, E. N. (2007). Transport phenomena. Wiley.  Blenke, H. (1979). Loop reactors. In T. K. Ghose, N. Blakebrough & A. Fiechter (Eds.), (pp. 121-214). Springer Berlin Heidelberg.  Boogerd, F. C., & Kuenen, J. G. (1990). Oxygen and Carbon Dioxide Mass Transfer and the Aerobic, Autotrophic Cultivation of Moderate and Extreme Thermophiles: A Case Study Related to the Microbial Desulfurization of Coal. Biotechnology and Bioengineering. 35, 1111-1119. Borowitzka, M. (1995). Microalgae as sources of pharmaceuticals and other biologically active compounds. Springer Netherlands.  87  Bulgakov, G., & Levich, P. (1999). The nitrogen : Phosphorus ratio as a factor regulating phytoplankton community structure : Nutrient ratios. Stuttgart, ALLEMAGNE: Schweizerbart.  Carlsson, A. S., van Beilen, J., M?ller, R., & Clayton, D. (2007). Micro- and macro-algae: Utility for industrial applications. UK: University of York.  Chisti, Y., & Moo-Young, M. (1988). Gas holdup behaviour in fermentation broths and other non-Newtonian fluids in pneumatically agitated reactors. The Chemical Engineering Journal, 39(3), B31-B36.  Chisti, Y., & Moo-Young, M. (1993). Improve the performance of airlift reactors. Chemical Engineering Progress, , 38-45.  Dauvill?e, D., & Delhaye, S. (2010). Engineering the chloroplast targeted malarial vaccine antigens in chlamydomonas starch granules. PLoS ONE, 5(12)  De Philippis, R., Colica, G., & Micheletti, E. (2011). Exopolysaccharide-producing cyanobacteria in heavy metal removal from water: Molecular basis and practical applicability of the biosorption process. Applied Microbiology and Biotechnology, 92(4), 697-708.  Deshmukh, D., Mane, A., Vhare, P., & Harke, S. (2013). Optimization of media component affecting phycocyanin production from microcystsis sp. Isolated from salim ali lake, aurangabad. [null] Journal of Algal Biomass Utilization, 4(2), 42-46.  Dhir, S., Morrow, K. J., Rhinehart, R. R., & Wiesner, T. (2000). Dynamic optimization of hybridoma growth in a fed-batch bioreactor. Biotechnology and Bioengineering, 67(2), 197-205.  Doran, P. M. (2012). Bioprocess engineering principles Elsevier Science.  Dos Santos, A. P., & Levin, Y. (2010). Surface tensions and surface potentials of acid solutions. The Journal of Chemical Physics, 133, 154107.  88  Doucha, J., Straka, F., & L?vansk?, K. (2005). Utilization of flue gas for cultivation of microalgae chlorella sp.) in an outdoor open thin-layer photobioreactor. Journal of Applied Phycology, 17(5), 403-412.  Dunwell, J. M., & Ford, C. S. (2005). Technologies for biological containment of GM and non-GM crops . UK: University of Reading.  Environment Canada. (2012a). Canada's emission trends. Canada: Government of Canada.  Environment Canada. (2012b). Reducing Greenhouse Gases. Canada: Government of Canada. Environment Canada. (2012c). National inventory report 1990-2010. Canada: Government of Canada.  Enzing, C., & Nooijen, A. (2012). Algae and genetic modification. Netherlands: COGEM.  Espinosa, J., Forchhammer, K., & Contreras, A. (2007). Role of the synechococcus PCC 7942 nitrogen regulator protein PipX in NtcA-controlled processes. Microbiology, 153(3), 711-718.  Flaherty, J. M. (2009). Canada's economic action plan. Canada: Department of Finance.  Gavrilescu, M., & Tudose, R. Z. (1999). Modelling mixing parameters in concentric-tube airlift bioreactors. Bioprocess Engineering, 20(6), 491-497.  Geider, R. J., & Osborne, B. A. (1992). Algal photosynthesis Chapman and Hall.  Goudar, C. T., Joeris, K., Konstantinov, K. B., & Piret, J. M. (2005). Logistic equations effectively model mammalian cell batch and fed-batch kinetics by logically constraining the fit. Biotechnology Progress, 21(4), 1109-1118.  Guerin, M., Huntley, M. E., & Olaizola, M. (2003). Haematococcus astaxanthin: Applications for human health and nutrition. Trends in Biotechnology, 21(5), 210-216.  89  Haynes, W. M., Lide, D. R., & Bruno, T. J. (2012). CRC handbook of chemistry and physics CRC Press.  Higuchi, D. (2012). U.S. EPA recognizes kuehnle AgroSystems for innovative algae biofuel work. Retrieved Mar 18, 2013, from http://yosemite.epa.gov/opa/admpress.nsf/0/AD6501A1BEE2D38485257A5A007785BA  Hodge, N. (2009). Carbon intensity of algae biofuels. Retrieved 03/04, 2013, from http://www.greenchipstocks.com/articles/carbon-intensity-of-algae-biofuels/462  Hong, D., et al. (2013). Combined effects of nitrate concentration and illumination conditions on the growth of microalga Haematococcus pluvialis. Journal of Biology, 34(4), 493-499. Hu, Q., Westerhoff, P., & Vermaas, W. (2000). Removal of nitrate from groundwater by cyanobacteria: Quantitative assessment of factors influencing nitrate uptake.  Hu, Q., Kurano, N., et al. (1998). Ultrahigh-cell-density culture of a marine green alga chlorococcum littorale in a flat-plate photobioreactor Springer Berlin / Heidelberg.  Hu, Q., Guterman, H., & Richmond, A. (1996). A flat inclined modular photobioreactor for outdoor mass cultivation of photoautotrophs. Biotechnology and Bioengineering, 51(1), 51-60.  Jacob-Lopes, E., Scoparo, C. H. G., Lacerda, Lucy Mara Cacia Ferreira, & Franco, T. T. (2009). Effect of light cycles (night/day) on CO2 fixation and biomass production by microalgae in photobioreactors. Chemical Engineering and Processing: Process Intensification, 48(1), 306-310.  Jain, A., Verma, D., & Bagchi, D. (2006). Catalytic and regulatory properties of sulphur metabolizing enzymes in cyanobacterium synechococcus elongatus PCC 7942. Indian Journal of Experimental Biology, 44(9), 767-772.  90  Kajiwara, S., Yamada, H., Ohkuni, N., & Ohtaguchi, K. (1997). Design of the bioreactor for carbon dioxide fixation by synechococcus PCC 7942. Energy Conversion and Management, 38, Supplement(0), S529-S532.  Kilonzo, P. M., & Margaritis, A. (2004). The effects of non-newtonian fermentation broth viscosity and small bubble segregation on oxygen mass transfer in gas-lift bioreactors: A critical review. Biochemical Engineering Journal, 17(1), 27-40.  Kim, N., Suh, I. S., et al. (2002). Simple monodimensional model for linear growth rate of photosynthetic microorganisms in flat-plate photobioreactors  Kim, J., & Lee, C. (2005). Systemic optimization of microalgae for bioactive compound production The Korean Society for Biotechnology and Bioengineering.  Kobayashi, K., & Fujita, K. (1997). Tube diameter on tubular photobioreactor for microalgal culture and its biomass productivity. Journal of Chemical Engineering of Japan, 30(2), 339-341.  Konopka, A., & Brock, T. D. (1978). Effect of temperature on blue-green algae (cyanobacteria) in lake mendota. Applied and Environmental Microbiology, 36(4), 572-576.  Kr?ger, G. H. J., & Eloff, J. N. (1978). The effect of temperature on specific growth rate and activation energy of microcystis and synechococcus isolates relevant to the onset of natural blooms. Journal of the Limnological Society of Southern Africa, 4(1), 9-20.  Krupa, Z., ?quist, G., & Gustafsson, P. (1991). Photoinhibition of photosynthesis and growth responses at different light levels in psbA gene mutants of the cyanobacterium synechococcus. Physiologia Plantarum, 82(1), 1-8.  Kusnadi, A.,R., Nikolov, Z.,L., & Howard, J.,A. (1997). Production of recombinant proteins in transgenic plants: Practical considerations. Biotechnology and Bioengineering, 56(5), 473-484.  Lai, C., Dietrich, D., & Bowman, M. (2005). Global warming and the mining of oceanic methane hydrate. Topics in Catalysis, 32(3-4), 95-99.  91  Lee, C. (1999). Calculation of light penetration depth in photobioreactors. Biotechnology and Bioprocess Engineering, 4(1), 78-81.  Lee, J. N., Lee, J. S., Shin, C. S., Park, S. C., & Kim, S. W. (2000). Methods to enhance tolerances of chlorella KR-1 to toxic compounds in flue gas. Applied Biochemistry and Biotechnology, 84-86, 329-342.  Lee, J., & Lee, J. (2003). Review of advances in biological CO2 mitigation technology. Biotechnology and Bioprocess Engineering, 8(6), 354-359.  Lin, J., Lee, S., Lee, H., & Koo, Y. (2000). Modeling of typical microbial cell growth in batch culture. Biotechnology and Bioprocess Engineering, 5(5), 382-385.  L?nneborg, A., Kalla, S. R., Samuelsson, G., Gustafsson, P., & ?quist, G. (1988). Light-regulated expression of the psbA transcript in the cyanobacterium anacystis nidulans. FEBS Letters, 240(1?2), 110-114.  Loo, C. L., & Sudesh, K. (2007). Polyhydroxyalkanoates: Bio-based microbial plastics and their properties. Malaysian Polymer Journal, 2(2), 31-57.  Lundquist, T., Woertz, I., Quinn, N., & Benemann, J. R. (2010). A realistic technology and engineering assessment of algae biofuel production. Energy Biosciences Institute, , 1.  Lunka, A., & Bayless, D. (2013). Effects of flashing light-emitting diodes on algal biomass productivity. Journal of Applied Phycology, 1-7. Mayer, D. (2008). Simple photosynthesis overview. Mazel, D., Houmard, J., & Marsac, N. (1988). A multigene family in calothrix sp. PCC 7601 encodes phycocyanin, the major component of the cyanobacterial light-harvesting antenna. Molecular and General Genetics MGG, 211(2), 296-304.  Merchuk, J. C., & Gluz, M. (2002). Bioreactors, air-lift reactors. Encyclopedia of Bioprocess Technology,  92  Mir?n, A. S., Camacho, F. G., G?mez, A. C., Grima, E. M., & Chisti, Y. (2000). Bubble-column and airlift photobioreactors for algal culture. AICHE Journal, 46(9), 1872-1887.  Mir?n, A. S., G?mez, A. C., Camacho, F. G., Grima, E. M., & Chisti, Y. (1999). Comparative evaluation of compact photobioreactors for large-scale monoculture of microalgae. In R. Osinga, J. Tramper, J.G.Burgess and R.H.Wijffels (Ed.), Progress in industrial microbiology (pp. 249-270) Elsevier.  Molina , E., Belarbi, E., Aci?n Fern?ndez, F. G., Robles Medina, A., & Chisti, Y. (2003). Recovery of microalgal biomass and metabolites: Process options and economics. Biotechnology Advances, 20(7?8), 491-515.  Monod, J. (1949). The growth of bacterial cultures. Annual Review of Microbiology, 3(1), 371-394.  National Research Council. (2012). Sustainable development of algal biofuels in the united states The National Academies Press.  Oilgae. (2009). Oilgae report academic edition. India:  Onken, U., & Tr?ger, M. (1990). A simplified model of hydrodynamics in airlift loop reactors. Bioprocess Engineering, 5(3), 129-139.  Petrovi?, D. L., Posarac, D., & Dudukovi?, A. (1995). Prediction of mixing time in airlift reactors. Chemical Engineering Communications, 133(1), 1-9.  Phillips, J. N. Jr., & Myers, J. (1954). Growth rate of chlorella in flashing light. Plant Physiology, 29(2), 152-161.  Pond Biofuels. (2011). Algae biofuels. Retrieved Dec 3, 2012, from http://www.pondbiofuels.com/Technology/AlgaeBiofuels/AlgaeBiofuels.html  Pope, D. (1975). Effects of light intensity, oxygen concentration, and carbon dioxide concentration on photosynthesis in algae. Microbial Ecology, 2(1), 1-16.  93  PowerplantCCS. (2010). Algae CO2 Capture costs and cost components. Retrieved January 29, 2013, from http://www.powerplantccs.com/ccs/cap/fut/alg/algae_co2_capture_costs.html  Prezelin, B. B. (1981). Light reactions in photosynthesis. Can.Bull.Fish.Aquat.Sci, 210, 1-43.  Pulz, O., & Gross, W. (2004). Valuable products from biotechnology of microalgae. Applied Microbiology and Biotechnology, (65), Jan. 7, 2013-635-648.  Rasala, B. A., Muto, M., Lee, P. A., Jager, M., Cardoso, R. M., Behnke, C. A., et al. (2010). Production of therapeutic proteins in algae, analysis of expression of seven human proteins in the chloroplast of chlamydomonas reinhardtii. Plant Biotechnology Journal, 8(6), 719-733.  Sasi, D. (2009). Biokinetic behaviour of chlorella vulgaris in a continuously stirred bioreactor and a circulating loop photobioreactor. (M.Sc., University of Saskatchewan).  Schillberg, S., Fischer, R., & Emans, N. (2003). Molecular farming of recombinant antibodies in plants. Cellular and Molecular Life Sciences CMLS, 60(3), 433-445.  Schwarz, D., Nodop, A., H?ge, J., Purf?rst, S., Forchhammer, K., Michel, K., et al. (2011). Metabolic and transcriptomic phenotyping of inorganic carbon acclimation in the cyanobacterium synechococcus elongatus PCC 7942. Plant Physiology, 155(4), 1640-1655.  Seambiotic Ltd. (2010). Algae pilot plant. Ashkelon, Israel: Seambiotic Ltd.  Shcolnick, S., Summerfield, T. C., Reytman, L., Sherman, L. A., & Keren, N. (August 2009). The mechanism of iron homeostasis in the unicellular cyanobacterium synechocystis sp. PCC 6803 and its relationship to oxidative stress. Plant Physiology, 150(4), 2045-2056.  Shelp, B. J., & Canvin, D. T. (1980). Photorespiration and oxygen inhibition of photosynthesis in chlorella pyrenoidosa. Plant Physiology, 65(5), pp. 780-784.  Siegesmund, S. (2011). Stone in architecture: Properties, durability Springer.  94  Sierra, E., Aci?n, F. G., Fern?ndez, J. M., Garc?a, J. L., Gonz?lez, C., & Molina, E. (2008). Characterization of a flat plate photobioreactor for the production of microalgae. Chemical Engineering Journal, 138(1?3), 136-147.  Specht, E., Miyake-Stoner, S., et al. (2010). Micro-algae come of age as a platform for recombinant protein production Springer Netherlands.  Srivastava, A. K., Rai, A. N., & Neilan, B. A. (2013). Stress biology of cyanobacteria: Molecular mechanisms to cellular responses Taylor & Francis Group.  Suh, I. S., & Lee, S. B. (2001). Cultivation of a cyanobacterium in an internally radiating air-lift photobioreactor. Journal of Applied Phycology, 13(4), 381-388.  Suh, I. S., Park, C. B., Han, J., & Lee, S. B. (1998). Cultivation of cyanobacterium in various types of photobioreactors for biological CO2 fixation. Studies in surface science and catalysis (pp. 471-474) Elsevier.  Summerfield, T. C., & Sherman, L. A. (2008). Global transcriptional response of the alkali-tolerant cyanobacterium synechocystis sp. strain PCC 6803 to a pH 10 environment. Appl Environ Microbiol., 17(74), 5276-5284.  Suzuki, L., & Johnson, C. H. (2001). Algae know the time of day: Circadian and photoperiodic programs. Journal of Phycology, 37(6), 933-942.  Todar, K. (2012). Nutrition and growth of bacteria. Retrieved Feb 25, 2013, from http://textbookofbacteriology.net/nutgro.html  UTEX. (2009). BG-11 medium. Retrieved 07/08, 2010, from http://www.sbs.utexas.edu/utex/mediaDetail.aspx?mediaID=26  Wageningen UR. (2011). Research on microalgae within wageningen UR. Retrieved Feb 27, 2012, from http://www.algae.wur.nl/UK/factsonalgae/growing_algae/temperature/  95  Walker, T., Purton, S., Becker, D., & Collet, C. (2005). Microalgae as bioreactors. Plant Cell Reports, 24(11), 629-641.  Watkins, K. (2002). Fighting the clock. Retrieved January 28, 2013, from http://www.pubs.acs.org/cen/coverstory/8004/8004pharmaceuticals.html  Wyszecki, G., Stiles, W. (1982). Color Science: Concept and Methods, Quantitative Data and Formul?. Wiley. Yamada, H., Ohkuni, N., Kajiwara, S., & Ohtaguchi, K. (1997). CO2-removal characteristics of anacystis nidulans R2 in airlift bioreactors. Energy, 22(2?3), 349-352.  Yang, J., Rasa, E., Tantayotai, P., Scow, K. M., Yuan, H., & Hristova, K. R. (2011). Mathematical model of chlorella minutissima UTEX2341 growth and lipid production under photoheterotrophic fermentation conditions. Bioresource Technology, 102(3), 3077-3082.  Yun, Y., Lee, S. B., Park, J.M., Lee, C., & Yang, J. (1997). Carbon Dioxide Fixation by Algal Cultivation Using Wastewater Nutrients. J. Chem. Tech. Biotechnol, 69, 451-455. Zhang, K., Kurano, N., & Miyachi, S. (2002). Optimized aeration by carbon dioxide gas for microalgal production and mass transfer characterization in a vertical flat-plate photobioreactor. Bioprocess Biosyst Eng, 25(2), 97-101.     96  Appendices Appendix A: Sample calculations 1) Calculation for biomass concentration X from optical density reading X (g/L) = 0.3543*OD600 ? 0.004 For OD=1, X=0.3543*1-0.004 X=0.3503 g/L 2) CO2 electrode calibration Dissolved CO2 concentration (g/L) = exp[(mV+1.2)/8.2082] For mV=30, Dissolved CO2 conc. = exp[(30+1.2)/8.2082]             = 44.7 mg/L 3) Calculation of theoretical mixing time (95% homogeneity) 26.05.019.012.031.05.53 ?? ??????= SDRGm VVVDHUt Calculate characteristic diameter D: Assuming D is equal to the diameter of a circle with an area equivalent to that of the cross-sectional area of the reactor. WLD ?=?????? 22pi piWLD ?= 2 For a photobioreactor of 2.54 cm width and 16.61 cm: pi61.1654.21002 ?=D D = 0.0731 m  For UG = 0.25 cm/s, H = 0.17145 m, D = 0.0731 m  tm = 4 s 26.05.019.012.00.31- 00016.0000359.0000346.00731.017145.00.198753.5 ????????????=mt97  4) Calculation of Bodenstein number ? ?+=?=????????????????==2222/14)(exp4)(UUUUUxx UUUUxrr BoxBocc?? ??pi? An initial guess of 10 was first applied to Bo. For ?U of 4,  ( ) 997.010444exp44106222/1=????????????????= ?==UUxxUrxcpi The residual sum of squares taken from the theoretical cr and experimental cr was minimized. Using the solver function in Excel, Bo was determined to be 11.5. cr was calculated to be 0.999.   98  Appendix B: Bubble size in photobioreactor (33?C)    00.511.522.530 0.5 1 1.5 2 2.5Bubble diameter (mm)Gas flow rate (L/min)99  Appendix C: S. elongatus growth at various conditions in 2D photobioreactor   00.20.40.60.811.20 50 100 150 200 250 300 350 400Biomass concentration (g/L)Time (h)33 deg C, 60 umol/m2/s, air @ 1 L/min33 deg C, 60 umol/m2/s, 5% @ 1 L/min33 deg C, 60 umol/m2/s, 10% @ 1 L/min33 deg C, 120 umol/m2/s, air @ 1 L/min33 deg C, 120 umol/m2/s, 5% @ 1 L/min33 deg C, 120 umol/m2/s, 10% @ 1 L/min33?C, 60 ?E/m2/s, air @ 1 L/min33?C, 60 ?E/ 2/s, 5% @ 1 L/min3?C, 60 ?E 2/s, 10% @ 1 L/min3?C, 120 ?E/ 2/s, air @ 1 L/min3?C, 120 ?E/ 2/s, 5% @ 1 L/min3?C, 120 ?E/ 2/s, 10% @ 1 L/min100  Appendix D: Polymath program for calculating carbon content # air @ 1 L/min d(C)/d(t) = kLa * (Cs - C) - r # carbon dioxide mass balance in medium (g/L/h) X = Xo * Xm * exp(u * t) / (Xm - Xo + Xo * exp(u * t)) # logistic growth model (g/L) Xo = 0.0421617 # initial biomass concentration (g/L) C(0) = 0.000654 # initial dissolved CO2 concentration in medium (g/L) u = 0.05038517 # maximum specific growth rate (1/h) kLa = 45.50 # volumetric mass transfer coefficient  (1/h) Xm = 0.3160356 # maximum biomass concentration (g/L) t(0) = 0 # initial time (h) t(f) = 150 # final time (h) Cs = 0.000474 # estimated saturation concentration of CO2 in medium (g/L) r = u * (1 - X / (1 + Xm)) * X * 0.465 * 44 / 12 # carbon uptake rate (g/L/h) Ca = C * 12 / 44 + X * 0.465 + 0.00216 # Total carbon (from CO2 and Na2CO3) in medium (g/L) Cx = X * 0.465 # Total carbon content in biomass (g/L)  # 5% @ 1 L/min d(C)/d(t) = kLa * (Cs - C) - r X = Xo * Xm * exp(u * t) / (Xm - Xo + Xo * exp(u * t)) Xo = 0.049602 C(0) = 0.0367 u = 0.025514 kLa = 5.76 Xm = 1.065734 t(0) = 0 t(f) = 300 Cs = 0.0367 r = u * (1 - X / (1 + Xm)) * X * 0.465 * 44 / 12 Ca = C * 12 / 44 + X * 0.465 + 0.00216 Cx = X * 0.465  # 10% @ 1 L/min d(C)/d(t) = kLa * (Cs - C) - r X = Xo * Xm * exp(u * t) / (Xm - Xo + Xo * exp(u * t)) Xo = 0.053854 C(0) = 0.067 u = 0.018948 kLa = 5.37 Xm = 0.793632 t(0) = 0 t(f) = 300 Cs = 0.067 r = u * (1 - X / (1 + Xm)) * X * 0.465 * 44 / 12 Ca = C * 12 / 44 + X * 0.465 + 0.00216 Cx = X * 0.465 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            data-media="{[{embed.selectedMedia}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.24.1-0074247/manifest

Comment

Related Items