A NOVEL, ALGAL-BASED CHEMICAL ABSORPTION SYSTEM FOR POST-COMBUSTION CARBON DIOXIDE CAPTURE By Adam John Smerigan A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Biosystems Engineering - Master of Science 2021 ABSTRACT A NOVEL, ALGAL-BASED CHEMICAL ABSORPTION SYSTEM FOR POST-COMBUSTION CARBON DIOXIDE CAPTURE By Adam John Smerigan Post-combustion carbon dioxide capture using amine solutions is an integral technology for reducing carbon dioxide emissions from the energy sector. However, environmental impacts and economic costs are restricting the implementation of amine absorbents. This study investigated the development of a sustainable algal based chemical absorption process to capture post-combustion carbon dioxide efficiently. Microalgal biomass was hydrolyzed to amino acids under basic conditions at 134oC. The supernatant of the hydrolysate was purged with carbon dioxide following centrifugation, and then underwent a desorption process to regenerate a chemical absorption algae-based solvent. A mass balance of the process showed that 31% of the mass into the process was recovered as an algal amino acid product. Another 30% exited the process as wet potassium carbonate which could be recovered as potassium hydroxide. The algal amino acid absorbent product contained 0.592 mol amino acid/L composed primarily of alanine, glutamic acid, glycine, aspartic acid, leucine, lysine, proline, etc. A trickling filter absorption column was built to determine the absorption capacity of the algal amino acid solution. The algal absorbent (1.27 ± 0.061 mol CO2/mol amine) had a higher absorption capacity than a synthetic amino acid absorbent (0.747 ± 0.021mol CO2/mol amine) composed of glycine, alanine, proline, and lysine. Both solutions were regenerable showing no signs (p<0.05) of deterioration after multiple absorption and desorption cycles regarding the pH of the solution, absorption capacity, and ATR-FTIR spectra. Using algal biomass as sustainable source of amino acids is a viable alternative to synthetic amino acid absorbents to effectively capture carbon dioxide in flue gas. ACKNOWLEDGEMENTS I would like to thank all the individuals who helped me create this thesis. Your guidance and support are greatly appreciated. Thank you to my major advisor Dr. Wei Liao. Your guidance, compassion, and wisdom were invaluable. Your work ethic is inspiring, and I am glad I was able to share your enthusiasm for this project. Thank you to Dr. Yan (Susie) Liu for serving on my committee. I want to thank you for giving me my start on research in our lab group and sparking my interest for graduate school. Thank you to Dr. Milton Smith for serving on my committee. I appreciate the insight you gave into the chemistry and the time you devoted to help me succeed. Thank you to Po-Jen Hsiao for the NMR and ATR-FTIR spectra. Your assistance was greatly appreciated. Thank you to Dr. Anthony Schilmiller and the RTSF Mass Spectrometry Center for the LCMS data. The amino acid concentration data were invaluable for this study. I would also like to thank the entire research group. Without your support, friendship, and knowledge this would not have been possible. Thank you, Ashley Cutshaw, Annaliese Marks, Douglas Clements, Henry Frost, Blake Smerigan, Yurui Zheng, and Meicai Xu. Special thanks to Dr. Sibel Uludag-Demirer who went to great lengths to support me personally and academically. I greatly appreciate you and enjoyed the time we spent together. Finally, I would like to thank my family and Sophie Morin for their love and encouragement. You helped me move along and work through any issues I had along the way. iii TABLE OF CONTENTS LIST OF TABLES ...................................................................................................................................... v LIST OF FIGURES ................................................................................................................................... vi CHAPTER 1: Literature Review............................................................................................................... 1 INTRODUCTION .................................................................................................................................... 1 LITERATURE REVIEW ......................................................................................................................... 4 Post-combustion carbon dioxide capture ............................................................................................ 4 Chemical absorption ............................................................................................................................ 5 Alkanolamine solutions ....................................................................................................................... 6 Amino acid solutions ............................................................................................................................ 7 Microalgae as a source of amino acids ............................................................................................. 10 PILOT AND COMMERCIAL CHEMICAL ABSORPTION PROCESSES ......................................... 11 SUMMARY OF KNOWLEDGE GAPS ................................................................................................ 12 OBJECTIVE AND HYPOTHESIS ........................................................................................................ 12 CHAPTER 2: MATERIAL AND METHODS ....................................................................................... 14 CHEMICALS AND OTHER MATERIALS .......................................................................................... 14 NUCLEAR MAGNETIC RESONANCE (NMR) .................................................................................. 14 LIQUID CHROMATOGRAPH/MASS SPECTROMETER (LCMS) ................................................... 15 ABSORPTION ....................................................................................................................................... 17 DESORPTION ........................................................................................................................................ 21 MICROALGAL PROTEIN CONVERSION TO AMINO ACIDS AND PROCESSING..................... 23 CHAPTER 3: ALGAE BIOMASS CONVERSION AND PROCESSING.......................................... 27 CHAPTER 4: USING A SYNTHETIC AMINO ACID (GAPL) ABSORBENT FOR CO2 ABSORPTION FROM A SYNTHETIC FLUE GAS CONTAINING 10% CO2 ............................... 32 CHAPTER 5: ALGAL AMINO ACID SOLUTION ABSORPTION .................................................. 39 CHAPTER 6: CONCLUSIONS AND FUTURE WORK ..................................................................... 47 APPENDICES ........................................................................................................................................... 48 APPENDIX A: CHAPTER 3 SUPPLEMENTAL INFORMATION ..................................................... 49 APPENDIX B: CHAPTER 4 SUPPLEMENTAL INFORMATION ..................................................... 61 APPENDIX C: CHAPTER 5 SUPPLEMENTAL INFORMATION ..................................................... 71 APPENDIX D: R-MARKDOWN FILE AND STATISTICS ................................................................ 85 APPENDIX E: ALGAL BIOMASS INFORMATION ........................................................................ 111 APPENDIX F: DAIRY ONE FORAGE LABORATORY PROCEDURES........................................ 113 APPENDIX G: LAB IMAGES............................................................................................................. 120 REFERENCES ........................................................................................................................................ 124 iv LIST OF TABLES Table 1: MS/MS Parameters for Function 1 (0-4.5min) 16 Table 2: MS/MS Parameters for Function 2 (4.5-6.55 min) 16 Table 3: MS/MS Parameters for Function 3 (6.55-13 min) 17 Table 4: Amino Acid Concentrations under Various Parr Reactor Conditions 27 Table 5: Energy Consumption of Process Equipment 30 Table 6: Absorption Totals for Different G/L Flow Ratios 33 Table 7: Synthetic Absorbent CO2 Absorption and Desorption Capacities 34 Table 8: Synthetic Absorbent pH over Four Absorption Cycles 36 Table 9: ATR-FTIR Peak Identification Table 37 Table 10: Algal Amino Acid Absorbent CO2 Absorption and Desorption Capacities 40 Table 11: Literature Absorption Capacities for Amino Acid and MEA Absorbents 45 Table 12: Energy Consumption for Algal Absorption and Desorption 46 Table 13: Pairwise Comparisons of Mean Amino Acid Concentrations for Parr Reactor Conditions 49 Table 14: Pairwise Comparisons between Gas to Liquid Flow Ratios 61 Table 15: Synthetic Absorbent CO2 Absorption and Desorption Capacities 62 Table 16: Algal Amino Acid Absorbent CO2 Absorption and Desorption Capacities 71 Table 17: Algal Amino Acid Absorbent pH over Absorption Cycles 72 Table 18: Absorption Capacities for the Synthetic and Algal Amino Acid Absorbents 72 Table 19: C. sorokiniana Biomass Composition 111 Table 20: C. sorokiniana Fatty Acid Composition 112 v LIST OF FIGURES Figure 1: Carbon Capture Technologies Flow Diagram10 _____________________________________ 3 Figure 2: Technology tree for Carbon Dioxide Separation12 ___________________________________ 4 Figure 3: Zwitterion Reaction Mechanism ________________________________________________ 7 Figure 4: Base-catalyzed Reaction Mechanism _____________________________________________ 7 Figure 5: Generic Amino Acid Ionization in Acidic, Neutral, and Alkaline Conditions______________ 8 Figure 6: Products of the Reaction of Amino Acids with Carbon Dioxide ________________________ 9 Figure 7: Amino Acid Carbon Capture Reaction Scheme25 ___________________________________ 9 Figure 8: Absorption Experimental Setup ________________________________________________ 20 Figure 9: Desorption Experimental Setup ________________________________________________ 22 Figure 10: Microalgal Biomass Conversion and Processing Flow Diagram ______________________ 24 Figure 11: Amino Acid Concentrations after each Processing Step ____________________________ 29 Figure 12: Mass Balance of the Biomass Conversion Process ________________________________ 30 Figure 13: Potassium Mass Balance of the Biomass Conversion Process ________________________ 30 Figure 14: ATR-FTIR Spectra for the Algal Amino Acid Solution after a) Parr Reactor b) Centrifuge c) Acidification d) Desorption ___________________________________________________________ 31 Figure 15: Synthetic Amino Acid Absorbent Absorption Curve for a Gas to Liquid Flow Rate of 2.5 _ 32 Figure 16: Carbon Dioxide Absorbed for Varied Gas to Liquid Flow Ratios _____________________ 33 Figure 17: Synthetic Absorbent CO2 Absorption and Desorption over Four Cycles________________ 35 Figure 18: Synthetic Absorbent pH over Four Absorption and Desorption Cycles ________________ 36 Figure 19: Synthetic Amino Acid Absorbent ATR-FTIR Spectra a) Original, b) After Absorption, c) After Desorption ____________________________________________________________________ 38 Figure 20: Algal Amino Acid Absorbent Absorption Curves _________________________________ 39 Figure 21: Algal Amino Acid Absorbent Absorption and Desorption Capacities__________________ 41 Figure 22: Algal Amino Acid Absorbent pH over Absorption and Desorption Cycles _____________ 42 vi Figure 23: Algal Amino Acid Absorbent ATR-FTIR Spectra a) original b) after absorption c) after desorption _________________________________________________________________________ 43 Figure 24: Cyclic Absorption Capacities for Absorbents in the Literature _______________________ 46 Figure 25: Amino Acid Concentrations after Parr Reactor for each Experiment __________________ 49 Figure 26: Amino Acid Concentrations after Centrifuge for each Experiment ____________________ 50 Figure 27: Amino Acid Concentrations after Acidification for each Experiment __________________ 50 Figure 28: Amino Acid Concentrations after Desorption for each Experiment ___________________ 51 Figure 29: Pairwise Comparisons of Amino Acid Concentrations for the Algal Amino Acid Solution for each Processing Step _________________________________________________________________ 51 Figure 30: Algal Amino Acid Solution Percent Mass Liquid Yield after Centrifuge _______________ 52 Figure 31: Algal Amino Acid Solution Percent Mass Liquid Yield Pairwise Comparisons __________ 52 Figure 32: Algal Amino Acid Solution Liquid Mass after Acidification ________________________ 53 Figure 33: Parr Slurry 1 ATR-FTIR Spectrum ____________________________________________ 53 Figure 34: Parr Slurry 2 ATR-FTIR Spectrum ____________________________________________ 54 Figure 35: Parr Slurry 3 ATR-FTIR Spectrum ____________________________________________ 54 Figure 36: Parr Slurry 4 ATR-FTIR Spectrum ____________________________________________ 55 Figure 37: Parr Centrifugate 1 ATR-FTIR Spectrum _______________________________________ 55 Figure 38: Parr Centrifugate 2 ATR-FTIR Spectrum _______________________________________ 55 Figure 39: Parr Centrifugate 3 ATR-FTIR Spectrum _______________________________________ 56 Figure 40: Parr Centrifugate 4 ATR-FTIR Spectrum _______________________________________ 56 Figure 41: Acidified Centrifugate 1 ATR-FTIR Spectrum ___________________________________ 57 Figure 42: Acidified Centrifugate 3 ATR-FTIR Spectrum ___________________________________ 57 Figure 43: Acidified Centrifugate 4 ATR-FTIR Spectrum ___________________________________ 58 Figure 44: Algal Amino Acid Product 1 ATR-FTIR Spectrum________________________________ 58 Figure 45: Algal Amino Acid Product 2 ATR-FTIR Spectrum________________________________ 59 Figure 46: Algal Amino Acid Product 3 ATR-FTIR Spectrum________________________________ 59 Figure 47: Algal Amino Acid Product 4 ATR-FTIR Spectrum________________________________ 60 vii Figure 48: Mean comparisons for various Gas to Liquid Flow Ratios __________________________ 61 Figure 49: Synthetic Absorbent CO2 Absorption and Desorption over Four Cycles in mol CO2/L ____ 62 Figure 50: Pairwise Comparisons for Synthetic Amino Acid Absorbent Absorption Capacities for each Cycle _____________________________________________________________________________ 63 Figure 51: Pairwise Comparisons for Synthetic Amino Acid Absorbent Desorption Capacities for each Cycle _____________________________________________________________________________ 63 Figure 52: Pairwise Comparisons for Synthetic Amino Acid Absorbent Absorbed pH for each Cycle _ 64 Figure 53: Pairwise Comparisons for Synthetic Amino Acid Absorbent Desorbed pH for each Cycle _ 64 Figure 54: Synthetic Amino Acid Absorbent ATR-FTIR Spectrum a) prior to Absorption b) after Absorption 1 c) after Absorption 2 d) after Absorption 3 e) after Absorption 4 ___________________ 65 Figure 55: Synthetic Amino Acid Absorbent ATR-FTIR Spectrum a) after Desorption 1 b) after Desorption 2 c) after Absorption 3 d) after Desorption 4 _____________________________________ 67 Figure 56: Synthetic Amino Acid Absorbent NMR Spectrum a) prior to Absorption b) after Absorption 4 c) after Desorption 4 _________________________________________________________________ 69 Figure 57: Amino Acid Concentration within the Algal Amino Acid Absorbent after Each Absorption 71 Figure 58: Algal Amino Acid Absorbent CO2 Absorbed and Desorbed over 8 Cycles in mol/L ______ 72 Figure 59: Absorption Capacities for the Synthetic and Algal Amino Acid Absorbents in mol/L _____ 73 Figure 60: Algal Amino Acid Absorbent ATR-FTIR Spectrum prior to Absorption _______________ 73 Figure 61: Algal Amino Acid Absorbent ATR-FTIR Spectrum after Absorption 1 ________________ 74 Figure 62: Algal Amino Acid Absorbent ATR-FTIR Spectrum after Absorption 2 ________________ 74 Figure 63: Algal Amino Acid Absorbent ATR-FTIR Spectrum after Absorption 3 ________________ 75 Figure 64: Algal Amino Acid Absorbent ATR-FTIR Spectrum after Absorption 4 ________________ 75 Figure 65: Algal Amino Acid Absorbent ATR-FTIR Spectrum after Absorption 5 ________________ 76 Figure 66: Algal Amino Acid Absorbent ATR-FTIR Spectrum after Absorption 6 ________________ 76 Figure 67: Algal Amino Acid Absorbent ATR-FTIR Spectrum after Absorption 7 ________________ 77 Figure 68: Algal Amino Acid Absorbent ATR-FTIR Spectrum after Absorption 8 ________________ 77 Figure 69: Algal Amino Acid Absorbent ATR-FTIR Spectrum after Desorption 1 ________________ 78 Figure 70: Algal Amino Acid Absorbent ATR-FTIR Spectrum after Desorption 2 ________________ 78 viii Figure 71: Algal Amino Acid Absorbent ATR-FTIR Spectrum after Desorption 4 ________________ 79 Figure 72: Algal Amino Acid Absorbent ATR-FTIR Spectrum after Desorption 5 ________________ 79 Figure 73: Algal Amino Acid Absorbent ATR-FTIR Spectrum after Desorption 6 ________________ 80 Figure 74: Algal Amino Acid Absorbent ATR-FTIR Spectrum after Desorption 7 ________________ 80 Figure 75: Algal Amino Acid Absorbent ATR-FTIR Spectrum after Desorption 8 ________________ 80 Figure 76: Algal Amino Acid Absorbent ATR-FTIR Spectrum after Desorption 1 after Dilution _____ 81 Figure 77: Algal Amino Acid Absorbent ATR-FTIR Spectrum after Desorption 2 after Dilution _____ 81 Figure 78: Algal Amino Acid Absorbent ATR-FTIR Spectrum after Desorption 3 after Dilution _____ 82 Figure 79: Algal Amino Acid Absorbent ATR-FTIR Spectrum after Desorption 4 after Dilution _____ 82 Figure 80: Algal Amino Acid Absorbent ATR-FTIR Spectrum after Desorption 5 after Dilution _____ 83 Figure 81: Algal Amino Acid Absorbent ATR-FTIR Spectrum after Desorption 6 after Dilution _____ 83 Figure 82: Algal Amino Acid Absorbent ATR-FTIR Spectrum after Desorption 7 after Dilution _____ 84 Figure 83: Desorption Experimental Setup Photo a) not in use b) in use _______________________ 120 Figure 84: Absorption Experimental Setup Photo a) not in use b) gas flow control setup c) column in use d) setup in use _____________________________________________________________________ 121 Figure 85: Algal Amino Acid Absorbent Photo a) before absorption b) after absorption __________ 122 Figure 86: Precipitated Solids after Acidification Photos a) solids and liquids b) dried solids ______ 123 ix CHAPTER 1: Literature Review INTRODUCTION Global climate change is a major issue facing the next generation. The effects of global climate change will be expensive and compromise health and safety in many regions of the United States1. One of the main drivers to global climate change is the greenhouse effect from greenhouse gases emitted by human activities. Carbon dioxide originating from land use and industrial processes, as the primary greenhouse gas, accounts for 76% of all greenhouse gases emitted globally2 and 55% of the observed global warming3. The United States, one of the top three greenhouse gas emitters in the world (after China and European Union), produces 15% of the total global carbon dioxide emissions2. Within the United States, the electricity sector is the second largest contributor of greenhouse gas emissions after the transportation sector, at 25% of all emissions. These emissions are primarily in the form of carbon dioxide from non-renewable fuel sources. Electricity generation from fossil fuel sources accounts for 2,588 billion kWh in the United States compared to 1,529 billion kWh from nuclear and renewable sources4. Fossil fuels will continue to fill a large proportion of the energy needs in the United States. Though the emissions from this sector are trending slightly downward, down 11.8% from 1,819.95 million metric tons (MMmt) in 1990 to 1,606.02 MMmt of carbon dioxide equivalent5 in 2019, the electricity sector will continue to represent a significant portion of emissions and must be addressed to secure a sustainable climate future6. A variety of opportunities are available to reduce the emissions from the electricity sector. These include improved efficiency of existing power plants, further implementation of renewable energy, increased end-use energy efficiency, and carbon capture and sequestration6. Currently, renewables and nuclear energy account for 37.2% of electricity generation in the 1 United States and the remaining generation comes from fossil fuels7. Increasing the efficiency of equipment in the power plants and at end use sites can improve energy efficiency by 30% or more8. However, carbon capture technologies are still required to achieve zero emission from these power plants. Accordingly, numerous technologies have been researched and developed to capture and store carbon dioxide emitted from the power industry. There are three primary methods of carbon dioxide capture including oxy-combustion, pre-combustion, and post- combustion. Some fuels, such as coal and natural gas, are pretreated by gasification and the water gas shift reaction prior to combustion. Pre-combustion capture removes the carbon dioxide evolved from these reactions. Oxyfuel combustion uses a pure oxygen feed to the combustion process allowing for a relatively pure, 80-98% carbon dioxide, product stream that can be compressed and stored. Finally, post-combustion techniques capture carbon dioxide from the flue gas of existing power plants9. Figure 110 below shows the process flow for each of these treatments. Post-combustion carbon dioxide capture is the most easily implemented and is the focus of this study. Specifically, chemical absorption separations are investigated with observations of future research areas. 2 Figure 1: Carbon Capture Technologies Flow Diagram10 3 LITERATURE REVIEW Post-combustion carbon dioxide capture Post-combustion carbon capture technologies are of specific interest due to the ability to retrofit existing power plants with the technology. This will allow for very little disruption to the existing infrastructure and allow for continued use of fossil fuels for energy production. This is important because, economically, fossil fuels are still favored and will be for the foreseeable future11. The ability to be easily implemented is counteracted by the main drawback of post- combustion processes, the separation12. Carbon dioxide from the flue gas of power industry is more difficult to be separated than the other processes due to the low concentration of carbon dioxide in flue gas, 4-14% depending on fuel source9. There are numerous technologies attempting to address this challenge. Figure 212 below shows a tree of carbon dioxide separation technologies with the focus of this review circled in red. Figure 2: Technology tree for Carbon Dioxide Separation12 4 Chemical absorption Of these technologies, chemical absorption, using monoethanolamine (MEA), amino acid, and other amine based solutions, has been intensively studied, and is most likely to be the first implemented in the near future 12. Chemical absorption has a high absorption capacity, has regenerable absorbents, and is the most mature carbon dioxide separation technology9,13. Disadvantages of this technology include the environmental impact of absorbent degradation, a high heat requirement for regeneration of absorption absorbents, and variable efficiency of absorption at different carbon dioxide concentrations in the flue gas9,13. The most used absorbent today is (MEA) due to the low cost of the chemical and high absorption efficiency. However, MEA has disadvantages which include low carbon dioxide loading capacity, degradation of absorbent by sulfur dioxide and oxygen, high corrosivity, and high energy consumption12,14. Most research on the chemical absorption is being done on new types of absorbents that have a higher absorption capacity than MEA while also requiring less energy for absorbent regeneration12. In addition, solutions with less toxicity and fewer environmental impacts are highly sought after. These solutions include alkanolamine solutions (the class of compound that MEA belongs) and amino acid solutions. Mixtures of these chemicals with themselves and other compounds are also being tested to attempt to combine the strengths of each individual component of the mixture15,16. Life cycle assessments have been performed to understand the impacts of the chemical absorption process in the power industry. However, environmental impacts of chemical absorption are significantly varied between different fossil fuel power plants, such as coal and natural gas plants. Since burning coal is inherently less environmentally friendly, chemical absorption can have more benefits than with a less environmentally impactful fuel, like natural 5 gas. To fully understand the impacts of chemical absorption technology both types of plants must be considered. The introduction of chemical absorption to a coal power plant shows a reduction of about 50% in the global warming potential17. Other benefits include a reduction of 50% or greater in the impact categories of human toxicity potential, acidification potential, and marine aquatic ecotoxicity potential, since the absorbent captures many toxic compounds from the flue gas17. For a natural gas power plant, a greater reduction in global warming potential of 58-68% is observed18. However, almost all other impact categories (i.e., human toxicity potential, acidification potential, marine ecotoxicity potential, and terrestrial ecotoxicity potential) were increased due to the toxicity of the degraded MEA, chemical wastes from production, and the effects of waste disposal, among other issues18. If the energy requirement and absorbent degradation is reduced, many of these negative impacts attributed to the chemical absorption technology can be avoided18. Alkanolamine solutions Alkanolamines are alkanes that contain a hydroxyl group and an amine group at the end of their carbon chains. The amines of these compounds can come in several forms: primary, secondary, or tertiary. Monoethanolamine (MEA), diethanolamine (DEA), triethanolamine (TEA), methyl diethanolamine (MDEA), and adenosine monophosphate (AMP) are all commonly used alkanolamines for carbon dioxide capture19. There are three primary reaction mechanisms for absorption: the zwitterion mechanism, termolecular mechanism, and base- catalyzed mechanism. Most alkanolamines (primary, secondary, and sterically hindered) follow the zwitterion mechanism shown in Figure 3. 6 Figure 3: Zwitterion Reaction Mechanism Figure 4 below shows how tertiary amines follow the base-catalyzed hydration mechanism19. Figure 4: Base-catalyzed Reaction Mechanism Alkanolamines typically have issues with high vapor pressures and oxidative degradation20,21. The high vapor pressure indicates a high heat of desorption reducing the economic viability of the process. Further, the use of alkanolamines causes corrosion in equipment and creates higher capital expenses21. Additionally, the oxidative degradation of the compounds leads to environmental concerns by production of toxic compounds, such as formaldehyde and other heat soluble salts, and waste disposal after the absorbent experiences reductions in absorption capacity, enhanced corrosion, foaming, and other undesirable properties20,22. Amino acid solutions Amino acid salt solutions are a viable alternative to the alkanolamine solutions23,24 and have multiple advantages over their alkanolamine counterparts. They are resistant to oxidative degradation which allows for further reuse due to the ionic nature of the absorbent25. This, coupled with the fact that amino acid compounds are eco-friendly and found in nature, suggests the solution could be economically feasible and sustainable21. Amino acid solutions have also been shown to have similar absorption capacities as compared to alkanolamine solutions21,24 as 7 well as a higher surface tension21,23. Additionally, sterically hindered amine groups commonly found in amino acids require less heat for desorption26,27. This shows the ability of amino acid solutions for comparable carbon dioxide and acid gas reduction at a potentially lower environmental expense. Amino acids have an amine group and a carboxylic group at each end of their structure forming an amphoteric compound that has a charge change based on the pH of the solution, as shown in Figure 5. Figure 5: Generic Amino Acid Ionization in Acidic, Neutral, and Alkaline Conditions At high pH, amino acids act as a base which allows for the lone pairs on the nitrogen from the deprotonated amine group to attack carbon dioxide to form carbamate, as shown in Figure 6. This carbamate can then react with water to form bicarbonate and recover the deprotonated amine group required for further reaction with carbon dioxide. At low pH, when a deprotonated amine reacts with carbon dioxide and water, bicarbonate can be formed in addition to a zwitterion which is unable to react further with carbon dioxide. 8 Figure 6: Products of the Reaction of Amino Acids with Carbon Dioxide Therefore, pH is an important characteristic to observe when determining the performance of an amino acid solution. In addition, amino acids with a lower pKa have better kinetics and a large operational pH range26 due to this pH dependence. A condensed reaction scheme for amino acid absorption is shown below in Figure 725. Figure 7: Amino Acid Carbon Capture Reaction Scheme25 9 From Figure 7, when the solution with high pH is first introduced to carbon dioxide, the deprotonated amine group will react with carbon dioxide to produce carbamate. The carbamate exists in equilibrium with bicarbonate and carbonate with the equilibrium favoring bicarbonate. This carbamate then reacts with water to form bicarbonate and a regenerated amine group which can again react with carbon dioxide when in its deprotonated form. This reaction can continue as long as hydrolysis occurs, and the pH of the solution favors the deprotonated amine group25. Further absorption will continue to reduce the pH of the solution as bicarbonate and carbonate acidify the solution and equilibrium is reached with bicarbonate being the primary product. Microalgae as a source of amino acids One possible source of amino acids is microalgae. Microalgae contains large amounts of proteins28 that can be hydrolyzed into amino acids29 through thermal, chemical, and biological reactions. After the hydrolysis, the algae slurry containing a significant amount of free amino acids imitates an amino acid absorption absorbent and could potentially be used as an economically viable and sustainable alternative to replace synthetic amino acid solutions. Conveniently, microalgae achieve their maximum growth rate with higher than atmospheric carbon dioxide concentrations30 with the most biomass being produced at concentrations as high as 10%31. Introducing the flue gas from power plants to microalgae cultivation has been shown to increase the growth rate and therefore carbon removal of the microalgae32,33. Additionally, microalgae have the ability to remove acid gases from the flue gas32. The microalgae used for carbon dioxide capture from the flue gas could be harvested and hydrolyzed as mentioned above to produce an amino acid absorption absorbent. The absorbent would capture additional carbon dioxide from the flue gas which could then be fed back into the algae reactor to produce more microalgae and capture more carbon30. Adding carbon capture 10 absorbents back into the algal culture provides another carbon source in the form of bicarbonate which could increase the carbon fixation efficiency of the microalgae by up to eight times34. Using an algae based absorption solution can also avoid toxicity to the culture compared to other absorbents such as ammonia35. This approach could greatly increase the cost effectiveness of the post-combustion absorption process and reduce negative environmental externalities. The algae grown can also be used to create other value-added products (i.e., protein-rich animal feed and polymer precursor) to increase the economic viability of the process34. No research has yet been completed using an algal based amino acid solution for carbon dioxide absorption so important parameters such as absorption capacity and heat of regeneration for this solution need to be investigated. PILOT AND COMMERCIAL CHEMICAL ABSORPTION PROCESSES Several pilot and commercial plants have been developed by organizations including the University of Texas, CSIRO, University of Stuttgard (CASTOR), BASF, Hitachi, and DOW36. Post-combustion carbon dioxide capture processes would increase the cost of electricity by 80- 85%37,38 and incur an energy penalty of 35% or more but, with the introduction of new amine absorbents and process intensification, this energy penalty can be reduced to around 15%38. For plants, one major cost comes from absorbent regeneration which accounts for 50-80% of energy costs36. Several test plants have attempted to tackle this issue by selecting new absorbents with a low heat of absorption36,37,39,40. Absorbents with low heat of absorptions allow for reduced heating during absorbent regeneration and energy savings. Additionally, absorbents are selected for their absorption capacity, kinetics, and cyclic capacity41. Absorbents with higher capacities and kinetics require less solution and residence time in the column leading to smaller columns and reduced capital costs. Absorbents with high cyclic capacities will reduce costs related to 11 absorbent degradation including absorbent disposal, maintenance costs, and operational costs. Another route of increasing the economic viability of the plants include utilizing the captured carbon dioxide to create a value-added product through processes such as biogas upgrading42 or algae cultivation. SUMMARY OF KNOWLEDGE GAPS This review establishes that chemical absorption using amino acid absorbents is a viable process for post-combustion carbon dioxide capture. However, the lack of incentives for capturing carbon dioxide inhibits the widespread use of the process. Combined use of algae cultivation and biomass conversion to amino acid absorbents, among other products (i.e., polymer and biofuels), could provide an economical, sustainable, and environmentally friendly alternative to other types of post-combustion chemical absorbents. The complex composition of the algal based absorbent may even provide additional advantages due to synergistic chemistries between compounds43. Despite this, no research has been conducted on the use of algal based amino acid absorbents for post-combustion capture of carbon dioxide. There is limited literature on the optimal conditions for the conversion of algae biomass to amino acids. Additionally, there is no process for isolating these amino acids from the rest of the solution after conversion. The absorption capacity, absorption rate, and regenerability of an algal-based amino acid absorbent are also unknown. This information is of great importance to the design of equipment and determination of process costs. OBJECTIVE AND HYPOTHESIS The goal of this study is to develop an algal based chemical absorption process to capture post-combustion carbon dioxide efficiently and sustainably. The hypothesis is that algal-based amino acids should synergistically enhance absorption capacity and improve technical feasibility 12 of the amino acid absorbent for post-combustion carbon dioxide capture. Correspondingly, five objectives are investigated in this study: 1. Develop a process for converting algal biomass into a mixed amino acid salt absorbent 2. Create an absorption unit that can accommodate the algal amino acid absorbent for carbon dioxide absorption 3. Determine the absorption and cyclic capacity of the algal amino acid absorbent 4. Use analytical spectroscopy to observe changes in the composition of absorbents 5. Conduct a mass and energy balance of the process. 13 CHAPTER 2: MATERIAL AND METHODS CHEMICALS AND OTHER MATERIALS The amino acids Glycine, Alanine, Proline and Lysine HCl powder were purchased from the Bulk Supplements Company. Potassium Hydroxide (KOH) was purchased from Fisher Scientific with a purity of greater than or equal to 85%. The synthetic amino acid solution (GAPL Solution) was mixed with water to create concentrations of 0.25M of glycine, 0.25M alanine, 0.25M proline, and 0.15M lysine for a total of 0.9M amino acid. The additional water generated from amino acid hydrolysis was poured off prior to adding the potassium hydroxide. Some amino acids were lost in this process. Using the Liquid Chromatograph/Mass Spectrometer (LCMS) method described below, the concentrations of the amino acids were determined to be 0.232M glycine, 0.216M alanine, 0.213M proline, and 0.174M lysine for a total of 0.835M amino acid. Potassium hydroxide was added to a 1M concentration accounting for 85% purity. The solution had a starting pH around 13. Bone dry carbon dioxide and air were obtained in gas cylinders from Airgas. Polyethylene tubing and brass fittings were used to connect to the regulators on the compressed air cylinders. Algal amino acid processing samples were tested for potassium concentration using the Tetraphenylborate method (HACH, Method 8049, Loveland, CO). NUCLEAR MAGNETIC RESONANCE (NMR) The NMR spectra were collected either on a Bruker Avance III HD 500 MHz NMR spectrometer equipped with 5mm iProbe (X-nucleus optimized double resonance broad band probe) or on a Varian Inova 500 MHz NMR spectrometer equipped with 5mm PFG broad band switchable probe. Both spectrometers operate at a frequency of 125 MHz for 13C and 500 MHz for 1H. All the NMR experiments were run at 25 ºC. The NMR samples were prepared by 14 diluting 0.5 mL of the aqueous amino acid solutions after absorption or desorption with 0.1 mL of deuterium oxide (D2O) for the signal lock. 1,4-dioxane was added as chemical shift reference. Quantitative 13C NMR spectra were collected on either of the spectrometers (Bruker HDIII 500 MHz NMR and Varian Inova 500 MHz NMR). The samples ran with Bruker used the following parameters: pulse duration p1 = 5 µs for 45º pulse; number of scans, NS = 800-1600 (more number of scans were used for the three and four amino acid mixtures to get better signal- to-noise ratio); acquisition time, AQ = 1.10 s; and relaxation delay, D1 = 60 s (relaxation time of nuclei D1 ≥ 5T1, T1: the longest carbon nuclei relaxation time constant). The samples ran on the Varian were with the following parameters: pulse duration p1 = 5.25 µs for 45º pulse, number of scans, NS = 800-1600; acquisition time, AQ = 1.04 s; and relaxation delay, D1 = 60 s (relaxation time of nuclei D1 ≥ 5T1, T1: the longest carbon nuclei relaxation time constant). The 13C NMR spectra were phase corrected automatically, baseline corrected (Whittaker Smoother), and integrated using MestReNova software v.14.2.0. LIQUID CHROMATOGRAPH/MASS SPECTROMETER (LCMS) Samples were analyzed by LC/MS/MS using a Waters Xevo TQS Micro interfaced with a Waters Acquity H-class UPLC. 10 ul of sample was injected onto a Waters Acquity HSS-T3 column (2.1x100 mm; 1.7 um particle size). The 13-minute gradient for separation of amino acids was as follows: initial conditions were 100% mobile phase A (10 mM PFHA in water) and 0% mobile phase B (acetonitrile), hold for 1 min at 100% A, linear ramp to 65% B at 8 min, ramp to 90% B at 8.01 min, hold at 90%B until 9 min, return to initial condition of 100%A at 9.01 min, hold at 100% A until 13 min. The flow rate was 0.3 ml/min and the column temperature was 40oC. Compounds were ionized by electrospray ionization in positive ion mode with a capillary voltage of 1.0 kV. Source temperature was 150oC, desolvation temp was 350oC 15 and desolvation and cone gas flows were 800 L/hr and 40 L/hr respectively. MS/MS data were obtained using a multiple reaction monitoring method with parameters shown in Table 1, Table 2, and Table 3 below. 13C and 15N-labeled amino acid internal standards were from Sigma (767964-1EA). Data processing was done using the Targetlynx tool in Masslynx. Table 1: MS/MS Parameters for Function 1 (0-4.5min) Parent Daughter Dwell Time Cone Collision Amino Acid Ion Ion (s) Voltage Energy 76 30 0.1 17 8 Glycine 79 32 0.1 17 8 13C2,15N-Glycine 90.1 44 0.03 17 8 Alanine 94.1 47.1 0.03 17 8 13C3,15N-Alanine 106.1 60 0.03 19 10 Serine 110.1 63 0.03 19 10 13C3,15N-Serine 120.1 74 0.03 19 8 Threonine 122 76 0.03 18 15 Cysteine 125.1 78.1 0.03 19 8 13C4,15N-Threonine 126 79 0.03 18 15 13C3,15N-Cysteine 133.1 74 0.03 19 14 Asparagine 134.1 74 0.03 19 10 Aspartic acid 139.1 77 0.03 19 11 13C4,15N-Aspartic acid 147.1 84 0.03 16 14 Glutamine 148.1 84 0.03 19 14 Glutamic acid 154.1 89.1 0.03 17 14 13C5,15N-Glutamine Table 2: MS/MS Parameters for Function 2 (4.5-6.55 min) Parent Daughter Dwell Time Cone Collision Amino Acid Ion Ion (s) Voltage Energy 116 70 0.03 21 10 Proline 118.1 72 0.03 17 9 Valine 122.1 75.1 0.03 21 10 13C5,15N-Proline 124.1 77.1 0.03 17 9 13C5,15N-Valine 150.1 104 0.03 19 9 Methionine 156.1 109.1 0.03 19 9 13C5,15N-Methionine 182.1 136.1 0.03 20 12 Tyrosine 192.1 145.1 0.03 20 12 13C9,15N-Tyrosine 16 Table 3: MS/MS Parameters for Function 3 (6.55-13 min) Parent Daughter Dwell Time Cone Collision Amino Acid Ion Ion (s) Voltage Energy 132.1 86 0.03 19 9 Leucine 139.1 92 0.03 19 9 13C5,15N-Leucine 147.1 84 0.03 19 14 Lysine 155.1 90.1 0.03 19 14 13C6,15N2-Lysine 156.1 110 0.03 20 12 Histidine 165.1 118.1 0.03 20 12 13C6,15N3-Histidine 166.1 120 0.03 20 10 Phenylalanine 175.1 70 0.03 24 18 Arginine 176.1 129.1 0.03 20 10 13C9,15N-Phenylalanine 185.1 75 0.03 24 18 13C6,15N4-Arginine 205.1 146 0.03 19 14 Tryptophan 218.1 156 0.03 19 14 13C11,15N2-Tryptophan ABSORPTION Trickling Filter Absorption Column A trickling filter absorption column was made to address foaming issues when the gas contacted the algal amino acid solution. The column was made with 5.08cm diameter furniture grade clear PVC pipe. It was capped with a PVC plug connected by a PVC union. A primer and adhesive were used to bond the PVC pieces together. A drill press was used to bore holes for stainless steel fittings. Two male, 0.635cm NPT fittings were added to the top and bottom of the column for a total of four fittings. One female, 0.635cm NPT fitting was connected to a male fitting on the top and bottom of the column. Epoxy was used to ensure a seal between the fittings and the PVC. A Topfin small air stone was attached to the bottom stainless-steel fitting and was used to purge the gas into the reactor. This purge stone was covered by a PVC pipe cap to prevent direct liquid contact to avoid foaming. The reactor is about 45.72cm long and 5.08cm in diameter for a total volume of 926.7 cm3. P-series 16 Pall rings were added into the column for 17 increased surface area for the gas-liquid interface. The entire system including the column is shown in Figure 8 below. Experimental Setup The gas stream into the reactor was created by mixing the compressed high purity grade carbon dioxide and air (Airgas). First, the gas from each cylinder was fed to a rotameter (VWR, Radnore, PA), maximum flow rates of 0.5LPM and 2.5LPM for CO2 and air respectively, and then the outlets from each rotameter were connected to a tee connection and mixed to achieve 10% (v/v) carbon dioxide in the inlet stream. The gas then passed through the gas flow meter (OMEGA, FMA, LP1620A-V2, Digital, Stamford, CT). Before connecting the gas stream to the absorption column, the two rotameters were adjusted to achieve the correct gas flow rate and composition by reading the output from the gas flow meter and IR gas analyzer (Quantek, Model 908 IR Gas Analyzer, Grafton, MA). Synthetic amino acid absorbents and algal amino acid absorbents were run at gas flow rates of 0.6LPM and 1.0LPM, respectively. After connecting the gas stream, the outlet of the column was monitored until the concentration of carbon dioxide reached the desired level, 10% carbon dioxide (v/v), indicating that the absorbent is fully saturated. To setup the column, 300 mL of the absorbent was added to a beaker and the inlet to the metering pump (Iwaki America Inc., EWN-B16PCUR, Holliston, MA) was placed inside the beaker. The outlet of the pump was connected to one of the stainless-steel fitting at the top of the reactor. The other fitting connected the gas outlet to the IR gas analyzer. Once the pump was turned on, the liquid would fall through the column (200 mL/min), being dispersed by the pall rings, before reaching the bottom. At the bottom of the column, one of the two fittings was connected to the air purge stone using 3 inches of plastic tubing. The other fitting was the outlet 18 for the liquid. A shut-off valve (Grainger, 3ZLG9, Lake Forest, IL) was used to create a 100 mL hold-up in the bottom of the reactor to prevent bubbling out of the column. The liquid outlet returns the liquid back into the beaker for recirculation through the reactor. The pump was run until the carbon dioxide concentration increased to the original concentration (10% v/v CO2) before the introduction of the absorbent. Absorptions were run at 23oC and atmospheric pressure. At the end of the experiment, the pH of the absorbent was measured with a pH probe (Fisher Scientific, Accumet Basic A15, Waltham, MA), and a sample was taken for NMR and ATR- FTIR (Jasco, FT/IR-660 ATR PRO ONE, Oklahoma City, OK). For absorption cycles, the mass and pH of the solution was measured, and a sample was taken after each absorption and desorption. For the synthetic amino acid absorbent cycles, the experiment was terminated after the carbon dioxide concentration in the gas outlet stream increased to 2% to reduce experiment time. For the algal amino acid absorbent absorption cycles, the mass of solution lost after each cycle was remade using same pH, DI water and sampled to track the amino acid concentrations after dilution. The absorption system is shown below in Figure 8. 19 Figure 8: Absorption Experimental Setup Calculation During the absorption experiment, the outlet carbon dioxide concentration was measured to calculate the total carbon dioxide absorbed by the absorbent. Knowing the gas flow and inlet concentration, the total amount of carbon dioxide absorbed can be calculated by area of integration method using the CO2 concentration profiles with operation time. The first integral sums the total mass of carbon dioxide entering the column based on the gas flow, inlet concentration, and time. The second integral sums the total mass of carbon dioxide that leaves the column. This first integral is then subtracted by the second integral to determine the mass of carbon dioxide captured by the absorbent. This is represented as a single integral below. Q G t mCO2,abs = V ∙MW 0 (cin -cout )dt ---------------------------- Equation 1 L 20 Where mCO2,abs is the amount of carbon dioxide absorbed (mol/L), QG is the gas flow rate through the column (LPM), VL is the volume of absorbent used (L), MW is the molecular weight of carbon dioxide (g/mol), cin and cout are the concentrations of carbon dioxide going into and leaving the column respectively (g/L), and t is the total absorption time (min). Carbon dioxide concentration was converted from percent by volume to g/L by dividing the percentage by 100 and multiplying by 1.964 (the molecular weight of CO2 divided by the molar volume from the ideal gas law at standard temperature and pressure (STP). The mass of carbon dioxide absorbed was converted to mol/mol amine using a conversion factor of the number of amines per mol of amino acid and the amino acid concentrations in solution determined by LCMS. This calculation was performed using the trapezoidal rule in Excel (2019) and the trapz function in MATLAB (R2019b). RStudio (Version 1.3.1056) was used to run ANOVA using the lm function and pairwise comparisons were completed using the package emmeans and Tukey’s method. DESORPTION Experimental Setup For the synthetic amino acid absorbent, a 24/40, two-necked, 1L round bottom flask was used to hold 300mL of the absorbent for desorption. A 24/40, two-necked, 2L round bottom flask was used for the algal amino acid absorbent for increased head space to accommodate for the bubbling of the solution during desorption. In one neck of the flask, a Dimroth column was used to condense water vapor using cool tap water to prevent loss from the absorbent solution. The other neck of the flask was used as an inlet for an air sweep gas to force carbon dioxide out of the head space of the flask and maintain a constant gas flow through the system. The flow rate of this gas stream was adjusted to around 0.6LPM using a rotameter and recorded using a gas flow meter. This sweep gas allowed for more accurate measurement of carbon dioxide leaving the 21 system and prevented carbon dioxide accumulation in the headspace. From the top of the condenser, tubing was used to connect to the IR gas analyzer. The round bottom flask was placed on a heating mantle (Glas-Col, 0412, Terre Haute, IN). A stirring bar was placed in the flask and the heating mantle was placed on top of a stir plate which was set to low for homogenous boiling. The absorbent was heated until there was no significant change in the concentration of carbon dioxide in the outlet over time. The algal amino acid absorbent was heated using a ramped heating method by using a voltage controller (Glas-Col, PL-312 Minitrol, Terre Haute, IN) to reduce bubbling within the flask. The heat initially is set at medium (50%) heat for 30mins then is increased to 75% heat for 15min, 90% heat for 15min, and finally 100% heat until the endpoint is reached. The absorbent was allowed to cool before the pH was taken and a sample was gathered for NMR or FTIR. The desorption system is shown below in Figure 9. Figure 9: Desorption Experimental Setup 22 Calculation The amount of carbon dioxide released from the desorption was calculated using the carbon dioxide concentration in the outlet, the gas flow rate, and time. The integral shown below was used to calculate the total CO2 desorbed. Q G t mCO2,des = V ∙MW (c )dt ----------------------------- Equation 2 0 out L Where mCO2,des is the amount of carbon dioxide released from the absorbent (mol/L), QG is the gas flow rate through the column (LPM), VL is the volume of absorbent used (L), MW is the molecular weight of carbon dioxide (g/mol), cout is the concentrations of carbon dioxide leaving the column (g/L), and t is the total absorption time (min). This integral was evaluated using the trapezoidal rule in Excel (2019) and the trapz function in MATLAB (R2019b), as well. The mass of carbon dioxide desorbed was converted to mol/mol amine using a conversion factor of the number of amines per mol of amino acid and the amino acid concentrations in solution determined by LCMS. RStudio (Version 1.3.1056) was used to run ANOVA using the lm function and pairwise comparisons were completed using the package emmeans and Tukey’s method. MICROALGAL PROTEIN CONVERSION TO AMINO ACIDS AND PROCESSING Experimental Setup Algal biomass was converted to an algal amino acid product through four processing steps. To get 1L of product, four separate experiments were completed following the process explained in this section. The mass and pH of the solution was recorded after each step and a sample was taken for ATR-FTIR and LCMS analysis. Figure 10 below shows the processing steps for the algal biomass prior to the absorption experiments. 23 Figure 10: Microalgal Biomass Conversion and Processing Flow Diagram Algae biomass was collected from a recirculating algae culture44 and stored in the freezer. The characteristics of Chlorella sorokiniana are shown in Appendix E. Before use, about 250g of the frozen algal biomass thawed for a day. Two samples were taken for total solids analysis and were placed in an oven at 105oC for 12h. Then the mass of the biomass was recorded using a balance (OHAUS, Scout Pro, Parsippany, NJ) giving 0.21g dry algal biomass per gram of wet biomass. Using the Jones factor (6.25g Protein/g N)45 and the protein content of the algae46, the amount of potassium hydroxide required for the specified ratios are calculated as follows: 0.586g protein 1g N 1 mol N 0.00670mol N ∙ ∙ = g dry algal biomass 6.25g protein 14g g dry algal biomass 0.00670mol N 0.21g dry algal biomass 56.106g 0.0790g KOH ∙ ∙ = g dry algal biomass g wet algal biomass 1 mol KOH g wet algal biomass Based on the calculation above, 19.75g KOH per 250g algal biomass were added for a 1:1 molar ratio of protein to potassium hydroxide. For a 1:5 ratio, the mass of potassium hydroxide was multiplied by five to get 98.75g KOH per 250g algal biomass. The potassium hydroxide was added slowly and mixed until completely dissolved. The Parr reactor, consisting of the motor (Pacific Scientific, SR3642-4982-7-56BC-CU, Moline, IL), reaction vessel (Parr Instrument Company, MAWP 1900psi at 350oC, Moline, IL), and controller (Parr Instrument 24 Company, 4848, Moline, IL), was filled with the algal biomass and potassium hydroxide solution and set to the conditions specified in Table 5. After the reaction time was complete, the reactor was turned off and the solution was cooled for 24h. The reactor was then emptied, and a sample of the solution was taken for ATR-FTIR, LCMS, and potassium analysis. The rest of the Parr algae slurry was stored in the refrigerator at 4oC. RStudio (Version 1.3.1056) was used to run an ANOVA using the lm function and pairwise comparisons were completed using the package emmeans and Tukey’s method to test for the most ideal reactor conditions. Parr Slurries from runs G, H, I, and J were then put through several processing steps. First, the Parr algal slurry contains solids that must be removed. A centrifuge (Beckman Coulter, Allegra X-12R Centrifuge, Brea, CA) was used to separate these solids at 5oC and 10,000rpm for 10min. The liquid was collected from the tubes and the solids remaining were placed on a scale and their mass was recorded. Two samples were kept in the refrigerator at 4oC for ATR-FTIR, LCMS, and potassium analysis. The pH of each Parr centrifugate was recorded as well. The centrifugate was stored in the refrigerator between processing steps. The Parr centrifugate still has a high concentration of potassium hydroxide that must be removed to prevent solid precipitation during absorption and to recycle for conversion. An air purge stone was connected to rubber tubing and placed in a 2L beaker. Pure carbon dioxide was released using a rotameter and gas flow meter into the centrifugate from the pressurized gas cylinder. The carbon dioxide reacted with the free hydroxide in the solution to form potassium carbonate and bicarbonate acidifying the solution. A stir bar was placed in the centrifugate and the beaker was placed on a stir plate set on medium. The gas flow rate started at 2.5LPM before being gradually reduced over 45min to a minimum of 0.5LPM to alleviate excessive bubbling of the solution. The pH of the solution was measured in situ and the experiment was ended when 25 there was no significant pH change over time. The acidified centrifugate was then immediately poured into a beaker and allowed to sit for 24h. The liquid layer at the top of the beaker was poured off and sampled for ATR-FTIR, LCMS, and potassium analysis. The mass of the wet solid layer was measured on a scale and either stored in the refrigerator or placed in the oven at 105oC for 24h and measured for dry mass. The acidified centrifugate now has too low of a pH (typically around 8.5) to be effective in absorption and must be regenerated. This method is the same as the desorption process for algal amino acid solutions detailed above in the Desorption Experimental Setup Section. Once there is no significant change in carbon dioxide concentration over time, the solution was cooled. The pH was recorded, and a sample was taken for ATR-FTIR, LCMS, and potassium analysis. At this point, the four solutions after desorption were combined into a 2L beaker, mixed, the pH was measured, and a sample was taken for ATR-FTIR and LCMS. 26 CHAPTER 3: ALGAE BIOMASS CONVERSION AND PROCESSING Varying the conditions of protein to KOH ratio, reaction temperature, and reaction time, the most effective conditions were chosen to create the algal amino acid absorbent. The conditions with the highest amino acid concentration were considered the best conditions for this study. From Table 4 below, samples G, H, I, and J were used to create the absorbent and are included for more statistical power. The ratio of protein to KOH has the most significant effect on amino acid concentration (AA Conc.) when averaged across the other conditions with a mean of 68.7g/L. The 1:5 ratio of protein to KOH performed significantly better than the 1:1 ratio (p < 0.05). The levels of reaction temperature and reaction time tested in this study did not have a significant effect on the amino acid concentration when averaged across the other factors (p > 0.05). Therefore, the conditions of a 1:5 ratio, 3h reaction time, and 134oC temperature were selected for the Parr reactor to prepare algal based amino acid salt solution. Appendix A has a table including the significance of each pairwise comparison. Table 4: Amino Acid Concentrations under Various Parr Reactor Conditions Sample ID KOH mol Reaction Reaction AA Conc. Ratio Temp (oC) Time (h) (g/L) A 1:5 134 3 59.37 B 1:5 134 7 69.62 C 1:1 121 5 14.43 D 1:1 121 7 16.76 E 1:1 134 5 14.66 F 1:1 134 7 13.86 G 1:5 134 3 75.95 H 1:5 134 3 69.23 I 1:5 134 3 63.35 J 1:5 134 3 60.21 27 After reacting the algal biomass in the Parr reactor, excess solids and KOH were removed from the solution in the next three processing steps: centrifuge, acidification, and desorption. These were described previously in the Microalgal Protein Conversion to Amino Acids and Solution Processing Section in Chapter 2. Figure 11 below shows the concentration of individual amino acids after each processing step. There is no significant (p>0.05) difference between the mean amino acid concentrations of the four processing steps confirming that amino acids are not lost throughout the process. Some of the major amino acids in the biomass are alanine, glutamic acid, glycine, aspartic acid, leucine, lysine, and proline. Many of these amino acids are observed in similar quantities among microalgae species47–49. However, the quantity of arginine, threonine, and serine are relatively low compared to reported values, while alanine and glycine have higher values than some of those reported in the literature47–49. Glycine, alanine, proline, and lysine compose 45.8% of the amino acid concentration of the solution. Since this is such a large proportion, a synthetic solution of these four amino acids is studied in the next Chapter as a control absorption solution to the algal absorbent. 28 Figure 11: Amino Acid Concentrations after each Processing Step A mass balance was conducted to observe the flow of different materials throughout the process. This information could then be used to recycle compounds such as KOH as they leave the system. As can be observed from Figure 12, 31% of the mass that enters the system leaves as the algal amino acid absorbent. Another 30% leaves as wet solids after the acidification step. The solids from this are formed from the reaction between carbon dioxide and hydroxide to create carbonate in the solution. Due to the large amount of KOH added to the biomass, the carbonate is formed to an extent that exceeds the solubility within the solution and precipitates as a solid. Since the liquid was separated by simply pouring off the top layer, there is residual liquid within the solids (55% dry mass). This leaves 211.5g of dry solids, most of which (79.4%) are potassium carbonate, as shown by Figure 13, that can be recycled to KOH using a kilning process50. This would greatly reduce the costs of the biomass conversion process. The centrifuge 29 solids compose the final 34% of the mass leaving the process. The remainder of mass not accounted for is lost during the transfer between glassware. The energy consumed by each piece of equipment is shown in Table 5 and are overestimates calculated using the current and voltage listed on the devices. The conversion of algal biomass to amino acids in the Parr reactor accounts for most of the energy consumed during the process at 76% of the total energy consumed. Figure 12: Mass Balance of the Biomass Conversion Process Figure 13: Potassium Mass Balance of the Biomass Conversion Process Table 5: Energy Consumption of Process Equipment Energy Consumption (kWh/kg Solution) Parr Reactor 9.99 Centrifuge 1.19 Desorption 2.04 The ATR-FTIR spectra shown in Figure 14 below show the change in the algal amino acid solution after each process step. The biggest change occurs after the acidification step where 30 new peaks are recorded at around 1300 and 1350cm-1. These two peaks disappear after desorption showing that the absorbent is ready to be reused for absorption. The peaks at 1560 and 1630cm-1 are relatively unchanged throughout the process. The pH change likely accounts for the changes in rank order between the two peaks since the acidification step is around 3 pH units lower than the other steps. (a) (b) 100 100 90 90 1400 80 80 1400 70 %T 70 %T 1630 1630 60 60 1560 1560 50 50 40 40 2000 1900 1800 1700 1600 1500 1400 1300 1200 1100 1000 2000 1900 1800 1700 1600 1500 1400 1300 1200 1100 1000 Wavenumber (cm-1) Wavenumber (cm-1) (c) (d) 100 100 90 90 1350 80 80 1400 1300 70 %T 70 %T 60 60 1630 1560 50 1560 1400 50 1630 40 40 2000 1900 1800 1700 1600 1500 1400 1300 1200 1100 1000 2000 1900 1800 1700 1600 1500 1400 1300 1200 1100 1000 Wavenumber (cm-1) Wavenumber (cm-1) Figure 14: ATR-FTIR Spectra for the Algal Amino Acid Solution after a) Parr Reactor b) Centrifuge c) Acidification d) Desorption 31 CHAPTER 4: USING A SYNTHETIC AMINO ACID (GAPL) ABSORBENT FOR CO2 ABSORPTION FROM A SYNTHETIC FLUE GAS CONTAINING 10% CO2 The first objective of this research was to develop an absorption column that was able to utilize the algal amino acid absorbent. The trickling filter absorption column was first tested with a synthetic amino acid absorbent to ensure it was operating properly and to determine the effective gas to liquid (G/L) flow ratios. Several G/L flow ratios were tested to determine an effective range for carbon dioxide absorption experiments. Figure 15 shows the synthetic amino acid absorption curve for a G/L flow ratio of 0.5 LPM. All curves have a similar shape but differ in the length of the complete absorption region (CO2 at 0%) at different G/L flow ratios. 12 Carbon Dioxide Concentration (%) 10 8 6 4 2 0 0 50 100 150 200 250 Time (min) Figure 15: Synthetic Amino Acid Absorbent Absorption Curve for a Gas to Liquid Flow Rate of 2.5 As Table 6 and Figure 16 show below, a G/L flow ratio of 3 absorbs the most carbon dioxide and was used in the following synthetic amino acid absorption cycle experiments. The statistical analysis shows that the means of two G/L ratios, 0.5 and 4, were determined to be significantly (p<0.05) different from all other means. The means of G/L ratios 1.875, 2.5, and 3 32 are significantly (p<0.05) higher than the means of G/L ratios 0.5 and 4 but are not significantly (p>0.05) different from each other. The significance of each pairwise comparison can be found in Appendix B. Table 6: Absorption Totals for Different G/L Flow Ratios G/L Gas Flow Liquid Flow Amount Absorbed ratio (LPM) (LPM) (mol CO2/mol Amine) 0.5 0.1 0.2 0.223 ± 0.011a 1.875 0.375 0.2 0.662 ± 0.051c 2.5 0.5 0.2 0.687 ± 0.007c 3 0.6 0.2 0.746 ± 0.021c 4 0.8 0.2 0.535 ± 0.025b *Means sharing the same letter within the same column are not significantly different based on a Type I error rate of 5% 0.9 0.8 0.7 Total Amount Absorbed 0.6 0.5 (mol CO2/mol Amine) 0.4 0.3 0.2 0.1 0 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 G/L Flow Ratio Figure 16: Carbon Dioxide Absorbed for Varied Gas to Liquid Flow Ratios A baseline for the regenerability of the algal amino acid absorbent is necessary. Therefore, the synthetic amino acid absorbent was tested across 4 absorption and desorption 33 cycles. After the first cycle, the amount absorbed by the solution seemed to stabilize and there were no significant (p>0.05) differences between the means of the last three cycles. The significant difference between the amount of carbon dioxide absorbed in the first absorption and the following absorptions is likely due to excess hydroxide ions in the absorbent which are converted to carbonate after the first absorption. This carbonate is unable to be converted back into hydroxide during desorption as shown by the amount of carbon dioxide released in Table 7 and Figure 17 below. This results in an absorbent with a maximum absorption capacity that is directly related to the amount of carbon dioxide that can be released from the solution. Since there is no evidence of a significant (p>0.05) change in the amount desorbed across the 4 cycles, it was assumed that the absorbent was completely regenerable after the initial absorption. Table 7: Synthetic Absorbent CO2 Absorption and Desorption Capacities Cycle Number 1 2 3 4 Amount 0.466 ± 0.023a 0.348 ± 0.022b 0.338 ± 0.022b 0.354 ± 0.040ab Absorbed (mol/mol Amine) Amount 0.287 ± 0.0002a 0.296 ± 0.024a 0.301 ± 0.015a 0.310 ± 0.026a Desorbed (mol/mol Amine) *Means sharing the same letter within the same row are not significantly different based on a Type I error rate of 5% 34 Figure 17: Synthetic Absorbent CO2 Absorption and Desorption over Four Cycles The regenerability of the absorbent is further confirmed by the pH data, as shown in Table 8 and Figure 18. The capacity of an absorbent is directly correlated to its pH. At high pH, the amine groups in the absorbent are deprotonated and are free to continue to bond carbon dioxide. An increase in the pH of the absorbent indicates that the absorbent regained its ability to react with carbon dioxide. There is an initial unrecoverable drop in pH after the first absorption. This is likely due to excess hydroxyl groups in the absorbent bonding to carbon dioxide to form carbonate in the solution. These carbonate molecules are not removed during desorption because they are more stable than bicarbonate and carbamate and create a permanent decrease in the pH of the absorbent. 35 Table 8: Synthetic Absorbent pH over Four Absorption Cycles Initial Abs 1 Des 1 Abs 2 Des 2 Abs 3 Des 3 Abs 4 Des 4 pH 12.97a 9.82b 11.50d 9.73bc 11.36d 9.47c 11.45d 9.67bc 11.42d Standard 0.0566 0.0707 0.0283 0.0141 0.1626 0.1414 0.0919 0.0495 0.0636 Deviation *Means sharing the same letter within the same row are not significantly different based on a Type I error rate of 5% 14 13 12 pH 11 10 9 8 0 1 2 3 4 Cycle Number Absorption Desorption Initial Figure 18: Synthetic Absorbent pH over Four Absorption and Desorption Cycles ATR-FTIR was used to analyze the change in the absorbent after absorption and desorption. Figure 19 shows the composition of the original absorbent, absorbent after absorption, and absorbent after desorption. After absorption, two additional peaks are generated at around 1350cm-1 and 1300cm-1. After desorption, those peaks disappear, and the spectra very closely resembles the spectra of the original solution. The peaks at 1350cm-1 and 1300cm-1 correspond to the generation of bicarbonate51–53 and carbamate51,52 during absorption, respectively. Since these peaks disappear, it shows that the absorbent is regenerable and that the 36 absorbent regains the chemistry of the original solution. The peaks at around 1400cm-1 and 1563cm-1 are estimated to be carboxylates from carbonate since they do not disappear after desorption. The peaks at 1633cm-1 and 3000cm-1 are assigned to water51–53. Table 9 below shows the peak assignments Table 9: ATR-FTIR Peak Identification Table Species Frequency (cm-1) Type Literature Frequency (cm-1) Carbamate 1300 v N-COO- 128352, 132251 Bicarbonate 1350 vsy CO 136053 Carboxylate/Carbonate 1400 and 1560 vas and vs COO- 1595 and 140552 H2 O 1630 δd H-O-H 162553 37 (a) (b) 100 100 90 1350 90 1400 1400 1300 80 80 70 %T 70 %T 60 60 1630 1560 1560 50 1630 50 40 40 2000 1900 1800 1700 1600 1500 1400 1300 1200 1100 1000 2000 1900 1800 1700 1600 1500 1400 1300 1200 1100 1000 Wavenumber (cm-1) Wavenumber (cm-1) (c) 100 90 80 1400 70 %T 1560 60 1630 50 40 2000 1900 1800 1700 1600 1500 1400 1300 1200 1100 1000 Wavenumber (cm-1) Figure 19: Synthetic Amino Acid Absorbent ATR-FTIR Spectra a) Original, b) After Absorption, c) After Desorption 38 CHAPTER 5: ALGAL AMINO ACID SOLUTION ABSORPTION Another objective of the research was to determine the absorption and cyclic capacity of the algal amino acid absorbent. This information is important for determining the environmental and economic advantages of using algal amino acid absorbents for post combustion carbon dioxide capture. Figure 20 below shows the algal absorption curve for all 8 of the absorption cycles. The leftward trend of the plot is due to the dilution of the absorbent solution. Overall, the curve has the same shape across cycles and shows the repeatability of the experiment. 12 Carbon Dioxide Concentration (%) 10 8 6 4 2 0 0 50 100 150 200 250 300 Time (min) Abs 1 Abs 2 Abs 3 Abs 5 Abs 7 Abs 8 Figure 20: Algal Amino Acid Absorbent Absorption Curves The absorption capacity and regenerability of the algal amino acid absorbent was determined by completing 8 absorption and desorption cycles. There were atypical experimental conditions during cycles 4 and 6 and the data were removed for greater clarity of results. Table 10 and Figure 21 below show the amount of carbon dioxide absorbed and desorbed for the 39 cycles. Appendix C has more information on the amount of CO2 absorbed and desorbed for each cycle. Like the synthetic absorbent results (GAPL solution), there was a drop in absorption capacity of the solution after the first absorption. The absorption stabilizes after the first absorption and there does not appear to be a significant difference between the absorptions from cycles 2 through 8. This is likely due to excess hydroxyl groups in the solution that are used up during the first absorption. The following absorptions are representative of the regenerability of the absorbent. It appears that the absorbent is highly regenerable after the first absorption. The desorption values match the absorption values showing little accumulation of carbon dioxide in the absorbent. Also, there does not appear to be any downward trend of the absorption capacity. Table 10: Algal Amino Acid Absorbent CO2 Absorption and Desorption Capacities Cycle Number 1 2-8 Average CO2 Absorbed 2.021 1.27 ± 0.061 (mol/mol Amine) Average CO2 desorbed 0.9671 1.18 ± 0.093 (mol/mol Amine) *: 1 indicates that the standard deviation cannot be computed for cycle 1 due to no replicates 40 Figure 21: Algal Amino Acid Absorbent Absorption and Desorption Capacities The pH of the algal amino acid absorbent is quite steady with a change of about 0.4 pH units after 8 cycles. However, Figure 22 shows a slight downward trend due to the dilutions. Appendix C shows the pH values of the solution after each cycle. Interestingly, there is not a large drop in pH after the first absorption like was observed in the synthetic absorbent. This is likely due to the precipitation of the formed carbonate and bicarbonate in the algal absorbent. During the acidification step, the algal absorbent was oversaturated with carbonate causing it to precipitate from solution as a solid. The synthetic absorbent lacks this step and retains the carbonates in solution causing a reduction in pH not seen in the algal absorbent results. This theory is validated by the observation of solids at the bottom of the absorption beaker after the 41 first algal absorption. Therefore, acidification of the algal absorbent should be completed twice or for a longer period to ensure all free potassium hydroxide is reacted prior to absorption. 12 11.5 11 10.5 10 pH 9.5 9 8.5 8 7.5 7 0 1 2 3 4 5 6 7 8 Cycle Number Initial Absorption Desorption Figure 22: Algal Amino Acid Absorbent pH over Absorption and Desorption Cycles The ATR-FTIR spectra in Figure 23 below are very similar to the synthetic absorbent spectra in Chapter 4. They both share peaks at wavenumbers 1630cm-1 and 1560cm-1. Additionally, the same absorption and desorption trend is observed. After absorption, two additional peaks are observed at around 1350 and 1300cm-1 which correspond to bicarbonate and carbamate since they are generated during absorption and disappear after desorption. The peak at 1400cm-1 is assigned to the carboxylate in carbonate since it is present in both the desorbed and absorbed spectra51–53. Overall, the spectra show that the absorbent regains a similar composition to itself prior to absorption. This suggests that the absorbent is highly regenerable as seen in Figure 21 and Figure 22 above. 42 (a) (b) 100 100 90 90 80 80 70 %T 1350 70 %T 60 1400 1300 60 1630 1560 50 1400 50 1630 1560 40 40 2000 1900 1800 1700 1600 1500 1400 1300 1200 1100 1000 2000 1900 1800 1700 1600 1500 1400 1300 1200 1100 1000 Wavenumber (cm-1) Wavenumber (cm-1) (c) 100 90 80 70 %T 60 1630 1560 1400 50 40 2000 1900 1800 1700 1600 1500 1400 1300 1200 1100 1000 Wavenumber (cm-1) Figure 23: Algal Amino Acid Absorbent ATR-FTIR Spectra a) original b) after absorption c) after desorption The final objective of this research was to compare the absorption capacity of the algal amino acid absorbent to that of a synthetic amino acid (GAPL) absorbent. Figure 24 below contains the absorption capacities of the two absorbents in mol/mol amine. Appendix C includes a table of the exact values along with a plot in mol/L. The algal amino acid absorbent had a significantly higher absorption capacity (p < 0.05) compared to the synthetic absorbent (1.27 to 0.747 mol CO2/mol amine, respectively). This suggests that there are other factors increasing the absorption capacity. The algal absorbent has other components of the biomass remaining in the 43 solution such as carbohydrates54, polypeptides, and lipids55 that may have a synergistic effect with the amino acids when used for carbon dioxide absorption and desorption. This could be useful in the development of better absorbents for further reduced costs and environmental impact. Other studies have reported similar absorption capacities to those determined here. Figure 24 shows the average cyclic absorption capacity of some amino acid absorbents and MEA from the literature along with standard deviation error bars. Solutions of 1M glycine, alanine, and proline with KOH had cyclic absorption capacities of 0.465, 0.535, and 0.412 mol/mol amine, respectively27. Another study that used an organic base with 2.5M glycine and alanine determined cyclic absorption capacities of 0.519 and 0.518 mol/mol amine, respectively24. A 0.5M equimolar KOH and lysine solution captured 0.378 mol/mol amine3. Overall, the numbers reported in this study are similar to those found in related literature shown in Table 11. When looking to compare cyclic absorption capacity against the most common absorbent (MEA), the algal absorbent greatly outperforms by over 300%. MEA had a very comparable cyclic absorption capacity to the amino acid absorbents at around 0.385 mol/mol amine24,27,56. 44 Table 11: Literature Absorption Capacities for Amino Acid and MEA Absorbents Absorbent Absorbent Temperature CO2 Absorption Cyclic Reference Concentration (oC) Concentration Capacity Capacity (M) (kPa) (mol/mol (mol/mol amine) amine) 27 MEA + 1.0 40 15 0.736 0.483 KOH 27 Glycine + 1.0 40 15 0.738 0.465 KOH 27 Alanine + 1.0 40 15 0.670 0.535 KOH 27 Proline + 1.0 40 15 0.746 0.412 KOH 24 MEA + 2.5 40 10 0.529 0.303 KOH 24 Glycine + 2.5 40 10 0.519 0.338 MAPA 24 Alanine + 2.5 40 10 0.518 0.308 MAPA 25 MEA + 2.5 22 4.8 0.5 n/a KOH 25 Glycine + 2.5 22 4.8 0.49 n/a KOH 25 Alanine + 2.5 22 4.8 0.52 n/a KOH 21 Alanine + 1.5 40 9.6 0.7238 n/a KOH + Piperizine 57 Glycine + 1.0 20 5.6 0.689 n/a KOH 57 MEA + 2.5 40 41 0.635 n/a KOH 56 MEA + 1.0 40 9.5 0.593 0.368 KOH 45 1.4 1.2 Cyclic Absorption Capacity 1 0.8 0.6 (mol/mol Amine) 0.4 0.2 0 MEA (Literature) Synthetic AA (Literature) Synthetic AA (This Study) Algal AA (This Study) Figure 24: Cyclic Absorption Capacities for Absorbents in the Literature Also, an energy balance was completed for the algal absorption and desorption cycle. Table 12 below shows the consumption during each process. The desorption process consumes about 4.5 times more energy than the absorption process per cycle. The desorption energy consumption was entirely composed of the cost for the heating mantle. For absorption, the only energy consuming piece of equipment was the metering pump. Energy consumption was calculated using the listed values on the equipment and are overestimates of the actual value. Table 12: Energy Consumption for Algal Absorption and Desorption Energy Consumption Energy Consumption (kWh/cycle) (kWh/kg CO2) Absorption 0.184 30.6 Desorption 0.833 139 46 CHAPTER 6: CONCLUSIONS AND FUTURE WORK A novel process was used to convert microalgal biomass to an algal amino acid absorbent for post-combustion carbon dioxide capture using a trickling filter absorption column. The results show that the algal amino acid absorbent captures significantly more carbon dioxide than a synthetic amino acid absorbent (1.27 to 0.747 mol CO2/mol amine, respectively).). ATR-FTIR showed no significant change within the absorbents after multiple absorption cycles indicating high regenerability of the absorbents. A mass balance of the biomass conversion process shows that 168g of potassium carbonate can be recovered and recycled back into potassium hydroxide. This would greatly reduce the cost of the process and potentially provide a cheap alternative to synthetic amino acids allowing for increased implementation of amino acid absorbents in post- combustion carbon dioxide capture. Future work should focus on investigating the interaction between algal amino acids and other algal compounds in the algal based amino acid solution on CO2 absorption, and determining the economics of the process. A techno-economic analysis needs to be conducted on the process to determine the cost of producing the algal amino acid absorbent compared to synthetic absorbents. More work needs to be done to optimize the process and scale it up into a pilot scale operation as well. These steps would help determine the viability of mass implementation of algal amino acid absorbents for post-carbon dioxide capture and sequestration. 47 APPENDICES 48 APPENDIX A: CHAPTER 3 SUPPLEMENTAL INFORMATION Table 13: Pairwise Comparisons of Mean Amino Acid Concentrations for Parr Reactor Conditions Comparison Estimate P-value KOH:Protein 1:1 – 5:1 -55.0 0.0012 Temperature 121 - 134 1.33 0.8396 Reaction Time 3–5 -3.234 0.9361 3–7 -3.998 0.8347 5-7 -0.764 0.9918 Figure 25: Amino Acid Concentrations after Parr Reactor for each Experiment 49 Figure 26: Amino Acid Concentrations after Centrifuge for each Experiment Figure 27: Amino Acid Concentrations after Acidification for each Experiment 50 Figure 28: Amino Acid Concentrations after Desorption for each Experiment Figure 29: Pairwise Comparisons of Amino Acid Concentrations for the Algal Amino Acid Solution for each Processing Step 51 Figure 30: Algal Amino Acid Solution Percent Mass Liquid Yield after Centrifuge Figure 31: Algal Amino Acid Solution Percent Mass Liquid Yield Pairwise Comparisons 52 Figure 32: Algal Amino Acid Solution Liquid Mass after Acidification Figure 33: Parr Slurry 1 ATR-FTIR Spectrum 53 Figure 34: Parr Slurry 2 ATR-FTIR Spectrum Figure 35: Parr Slurry 3 ATR-FTIR Spectrum 54 Figure 36: Parr Slurry 4 ATR-FTIR Spectrum Figure 37: Parr Centrifugate 1 ATR-FTIR Spectrum Figure 38: Parr Centrifugate 2 ATR-FTIR Spectrum 55 Figure 39: Parr Centrifugate 3 ATR-FTIR Spectrum Figure 40: Parr Centrifugate 4 ATR-FTIR Spectrum 56 Figure 41: Acidified Centrifugate 1 ATR-FTIR Spectrum Figure 42: Acidified Centrifugate 3 ATR-FTIR Spectrum 57 Figure 43: Acidified Centrifugate 4 ATR-FTIR Spectrum Figure 44: Algal Amino Acid Product 1 ATR-FTIR Spectrum 58 Figure 45: Algal Amino Acid Product 2 ATR-FTIR Spectrum Figure 46: Algal Amino Acid Product 3 ATR-FTIR Spectrum 59 Figure 47: Algal Amino Acid Product 4 ATR-FTIR Spectrum 60 APPENDIX B: CHAPTER 4 SUPPLEMENTAL INFORMATION Table 14: Pairwise Comparisons between Gas to Liquid Flow Ratios Contrast Estimate p-value 0.5 - 1.875 -0.439 <0.01 0.5 - 2.5 -0.464 <0.01 0.5 - 3 -0.523 <0.01 0.5 - 4 -0.312 <0.01 1.875 – 2.5 -0.025 0.887 1.875 - 3 -0.084 0.131 1.875 - 4 0.127 0.029 2.5 - 3 -0.059 0.330 2.5 - 4 0.152 0.014 3-4 0.211 <0.01 Figure 48: Mean comparisons for various Gas to Liquid Flow Ratios 61 Table 15: Synthetic Absorbent CO2 Absorption and Desorption Capacities Cycle Number 1 2 3 4 Amount 0.470 ± 0.024a 0.352 ± 0.022b 0.341 ± 0.023b 0.357 ± 0.041ab Absorbed (mol/L) Amount 0.290 ± 0.0002a 0.298 ± 0.024a 0.303 ± 0.015a 0.313 ± 0.027a Desorbed (mol/L) Figure 49: Synthetic Absorbent CO2 Absorption and Desorption over Four Cycles in mol CO2/L 62 Figure 50: Pairwise Comparisons for Synthetic Amino Acid Absorbent Absorption Capacities for each Cycle Figure 51: Pairwise Comparisons for Synthetic Amino Acid Absorbent Desorption Capacities for each Cycle 63 Figure 52: Pairwise Comparisons for Synthetic Amino Acid Absorbent Absorbed pH for each Cycle Figure 53: Pairwise Comparisons for Synthetic Amino Acid Absorbent Desorbed pH for each Cycle 64 (a) (b) (c) Figure 54: Synthetic Amino Acid Absorbent ATR-FTIR Spectrum a) prior to Absorption b) after Absorption 1 c) after Absorption 2 d) after Absorption 3 e) after Absorption 4 65 Figure 54 (cont’d) (d) (e) 66 (a) (b) (c) Figure 55: Synthetic Amino Acid Absorbent ATR-FTIR Spectrum a) after Desorption 1 b) after Desorption 2 c) after Absorption 3 d) after Desorption 4 67 Figure 55 (cont’d) (d) 68 (a) (b) Figure 56: Synthetic Amino Acid Absorbent NMR Spectrum a) prior to Absorption b) after Absorption 4 c) after Desorption 4 69 Figure 56 (cont’d) (c) 70 APPENDIX C: CHAPTER 5 SUPPLEMENTAL INFORMATION Figure 57: Amino Acid Concentration within the Algal Amino Acid Absorbent after Each Absorption Table 16: Algal Amino Acid Absorbent CO2 Absorption and Desorption Capacities Cycle Number 1 2 3 5 7 8 Amount Absorbed 1.30 0.889 0.900 0.774 0.764 0.800 (mol/L) Amount Desorbed 0.615 0.697 0.739 0.728 0.728 .742 (mol/L) Amount Absorbed 2.02 1.34 1.32 1.28 1.21 1.20 (mol/mol Amine) Amount Desorbed 0.967 1.09 1.08 1.30 1.21 1.20 (mol/mol Amine) 71 Figure 58: Algal Amino Acid Absorbent CO2 Absorbed and Desorbed over 8 Cycles in mol/L Table 17: Algal Amino Acid Absorbent pH over Absorption Cycles Cycle Number 1 2 3 4 5 6 7 8 Absorbed 9.25 9.04 8.99 9.55 8.93 10.35 8.87 8.88 pH Desorbed 11.65 11.62 11.58 11.55 11.4 11.37 11.43 11.25 pH Table 18: Absorption Capacities for the Synthetic and Algal Amino Acid Absorbents Synthetic AA Absorbent Algal AA Absorbent Amount CO2 Absorbed 0.754 ± 0.021 0.825 ± 0.065 (mol/L) Amount CO2 Absorbed 0.747 ± 0.021 1.27 ± 0.061 (mol/mol Amine) 72 1 0.9 Carbon Dioxide Absorbed 0.8 0.7 0.6 0.5 (mol/L) 0.4 0.3 0.2 0.1 0 Synthetic Absorbent Algal Absorbent Figure 59: Absorption Capacities for the Synthetic and Algal Amino Acid Absorbents in mol/L Figure 60: Algal Amino Acid Absorbent ATR-FTIR Spectrum prior to Absorption 73 Figure 61: Algal Amino Acid Absorbent ATR-FTIR Spectrum after Absorption 1 Figure 62: Algal Amino Acid Absorbent ATR-FTIR Spectrum after Absorption 2 74 Figure 63: Algal Amino Acid Absorbent ATR-FTIR Spectrum after Absorption 3 Figure 64: Algal Amino Acid Absorbent ATR-FTIR Spectrum after Absorption 4 75 Figure 65: Algal Amino Acid Absorbent ATR-FTIR Spectrum after Absorption 5 Figure 66: Algal Amino Acid Absorbent ATR-FTIR Spectrum after Absorption 6 76 Figure 67: Algal Amino Acid Absorbent ATR-FTIR Spectrum after Absorption 7 Figure 68: Algal Amino Acid Absorbent ATR-FTIR Spectrum after Absorption 8 77 Figure 69: Algal Amino Acid Absorbent ATR-FTIR Spectrum after Desorption 1 Figure 70: Algal Amino Acid Absorbent ATR-FTIR Spectrum after Desorption 2 78 Figure 71: Algal Amino Acid Absorbent ATR-FTIR Spectrum after Desorption 4 Figure 72: Algal Amino Acid Absorbent ATR-FTIR Spectrum after Desorption 5 79 Figure 73: Algal Amino Acid Absorbent ATR-FTIR Spectrum after Desorption 6 Figure 74: Algal Amino Acid Absorbent ATR-FTIR Spectrum after Desorption 7 Figure 75: Algal Amino Acid Absorbent ATR-FTIR Spectrum after Desorption 8 80 Figure 76: Algal Amino Acid Absorbent ATR-FTIR Spectrum after Desorption 1 after Dilution Figure 77: Algal Amino Acid Absorbent ATR-FTIR Spectrum after Desorption 2 after Dilution 81 Figure 78: Algal Amino Acid Absorbent ATR-FTIR Spectrum after Desorption 3 after Dilution Figure 79: Algal Amino Acid Absorbent ATR-FTIR Spectrum after Desorption 4 after Dilution 82 Figure 80: Algal Amino Acid Absorbent ATR-FTIR Spectrum after Desorption 5 after Dilution Figure 81: Algal Amino Acid Absorbent ATR-FTIR Spectrum after Desorption 6 after Dilution 83 Figure 82: Algal Amino Acid Absorbent ATR-FTIR Spectrum after Desorption 7 after Dilution 84 APPENDIX D: R-MARKDOWN FILE AND STATISTICS Thesis Statistics Adam Smerigan 07/27/2021 Clear the Environment rm(list=ls()) Gather Packages and Data from Metadata Sheet # Load Libraries library(readxl) library(ggplot2) ## Warning: package 'ggplot2' was built under R version 4.0.4 ## Registered S3 methods overwritten by 'tibble': ## method from ## format.tbl pillar ## print.tbl pillar library(dplyr) ## ## Attaching package: 'dplyr' ## The following objects are masked from 'package:stats': ## ## filter, lag ## The following objects are masked from 'package:base': ## ## intersect, setdiff, setequal, union library(emmeans) # Import Data from Metadata Sheet g.l.ratio.data <- read_excel("Metadata_072621.xlsx", sheet = "g.l.rati o") GAPL.cycles.data <- read_excel("Metadata_072621.xlsx", sheet = "GAPL.c ycles") algae.cycles.data <- read_excel("Metadata_072621.xlsx", sheet = "algae .cycles") TS.data <- read_excel("Metadata_072621.xlsx", sheet = "TS") 85 centrifugate.data <- read_excel("Metadata_072621.xlsx", sheet = "Centr ifugate") acidification.data <- read_excel("Metadata_072621.xlsx", sheet = "Acid ification") desorption.data <- read_excel("Metadata_072621.xlsx", sheet = "Desorpt ion") LCMS.processing.data <- read_excel("Metadata_072621.xlsx", sheet = "LC MS.processing") LCMS.algae.cycles.data <- read_excel("Metadata_072621.xlsx", sheet = " LCMS.algae.cycles") Parr.Cond.data <- read_excel("Metadata_072621.xlsx", sheet = "Parr Con ditions") Microalgae Conversion and Processing Parr Reactor Conditions Analysis col <- c("protein.to.KOH","temperature","reaction.time") Parr.Cond.data[col] <- lapply(Parr.Cond.data[col], factor) summary(Parr.Cond.data) ## sample.ID protein.to.KOH temperature reaction.time AA. conc ## Length:10 1:4 121:2 3:5 Min. :13.86 ## Class :character 5:6 134:8 5:2 1st Qu .:15.19 ## Mode :character 7:3 Median :59.79 ## Mean :45.74 ## 3rd Qu .:67.76 ## Max. :75.95 parr.cond.lm <- lm(AA.conc ~ protein.to.KOH + temperature + reaction.t ime, Parr.Cond.data) joint_tests(parr.cond.lm) ## model term df1 df2 F.ratio p.value ## protein.to.KOH 1 5 44.106 0.0012 ## temperature 1 5 0.045 0.8396 ## reaction.time 2 5 0.178 0.8424 lsmeans.parr.cond.koh = emmeans(parr.cond.lm, "protein.to.KOH") pairs(lsmeans.parr.cond.koh) 86 ## contrast estimate SE df t.ratio p.value ## 1 - 5 -55 8.28 5 -6.641 0.0012 ## ## Results are averaged over the levels of: temperature, reaction.time multcomp::cld(lsmeans.parr.cond.koh, by = NULL, Letters = "abcdefg", a lpha = .05) # Tukey ## protein.to.KOH emmean SE df lower.CL upper.CL .group ## 1 13.7 4.01 5 3.41 24.0 a ## 5 68.7 5.69 5 54.06 83.3 b ## ## Results are averaged over the levels of: temperature, reaction.time ## Confidence level used: 0.95 ## significance level used: alpha = 0.05 lsmeans.parr.cond.temp = emmeans(parr.cond.lm, "temperature") pairs(lsmeans.parr.cond.temp) ## contrast estimate SE df t.ratio p.value ## 121 - 134 1.33 6.26 5 0.213 0.8396 ## ## Results are averaged over the levels of: protein.to.KOH, reaction.t ime multcomp::cld(lsmeans.parr.cond.temp, by = NULL, Letters = "abcdefg", alpha = .05) # Tukey ## temperature emmean SE df lower.CL upper.CL .group ## 134 40.5 2.67 5 33.7 47.4 a ## 121 41.9 5.17 5 28.6 55.2 a ## ## Results are averaged over the levels of: protein.to.KOH, reaction.t ime ## Confidence level used: 0.95 ## significance level used: alpha = 0.05 lsmeans.parr.cond.time = emmeans(parr.cond.lm, "reaction.time") pairs(lsmeans.parr.cond.time) ## contrast estimate SE df t.ratio p.value ## 3 - 5 -3.234 9.28 5 -0.348 0.9361 ## 3 - 7 -3.998 6.86 5 -0.583 0.8347 ## 5 - 7 -0.764 6.26 5 -0.122 0.9918 ## ## Results are averaged over the levels of: protein.to.KOH, temperatur e 87 ## P value adjustment: tukey method for comparing a family of 3 estima tes multcomp::cld(lsmeans.parr.cond.time, by = NULL, Letters = "abcdefg", alpha = .05) # Tukey ## reaction.time emmean SE df lower.CL upper.CL .group ## 3 38.8 5.00 5 26.0 51.6 a ## 5 42.0 6.06 5 26.5 57.6 a ## 7 42.8 4.14 5 32.2 53.4 a ## ## Results are averaged over the levels of: protein.to.KOH, temperatur e ## Confidence level used: 0.95 ## P value adjustment: tukey method for comparing a family of 3 estima tes ## significance level used: alpha = 0.05 par(mfrow=c(2,2)) plot(lsmeans.parr.cond.koh, comparisons=TRUE, xlab = "Mean AA Concentr ation (g/L)", ylab = "KOH:Protein Ratio") plot(lsmeans.parr.cond.temp, comparisons=TRUE, xlab = "Mean AA Concent ration (g/L)", ylab = "Temperature (Celsius)") plot(lsmeans.parr.cond.time, comparisons=TRUE, xlab = "Mean AA Concent ration (g/L)", ylab = "Reaction Time (Hours)") par(mfrow=c(1,1)) Plot of Amino Acid Concentration after Each Processing Step # Organize the Data for Plotting parr.slurry.averaged <- LCMS.processing.data %>% group_by(Name) %>% filter(Type == "Parr Slurry") %>% summarize(m.1 = mean(AA.Conc.M), sd.1 = sd(AA.Conc.M), .groups = 'dr op') centrifugate.averaged <- LCMS.processing.data %>% group_by(Name) %>% filter(Type == "Centrifugate") %>% summarize(m.2 = mean(AA.Conc.M), sd.2 = sd(AA.Conc.M), .groups = 'dr 88 op') acid.cent.averaged <- LCMS.processing.data %>% group_by(Name) %>% filter(Type == "Acidified Centrifugate") %>% summarize(m.3 = mean(AA.Conc.M), sd.3 = sd(AA.Conc.M), .groups = 'dr op') prod.averaged <- LCMS.processing.data %>% group_by(Name) %>% filter(Type == "AA Product") %>% summarize(m.4 = mean(AA.Conc.M), sd.4 = sd(AA.Conc.M), .groups = 'dr op') y.proc.1 <- c(parr.slurry.averaged$m.1, centrifugate.averaged$m.2, aci d.cent.averaged$m.3, prod.averaged$m.4) x.proc.1 <- c(rep(sort(unique(LCMS.processing.data$Name)), 4)) error.proc.1 <- c(parr.slurry.averaged$sd.1, centrifugate.averaged$sd. 2, acid.cent.averaged$sd.3, prod.averaged$sd.4) fill.proc.1 <- c(rep("Parr Slurry", 20), rep("Centrifugate", 20), rep( "Acidified Centrifugate", 20), rep("AA Product", 20)) proc.data.frame <- as.data.frame(y.proc.1) proc.plot.1 <- data.frame(x.proc.1, y.proc.1, error.proc.1, fill.proc. 1) # Plot the Data ggplot(proc.plot.1, aes(x= x.proc.1, y = y.proc.1, fill = fill.proc.1) ) + geom_bar(position = position_dodge(), stat = "identity") + geom_errorbar(aes(ymin=y.proc.1-error.proc.1, ymax=y.proc.1+error.pr oc.1), width=0.2, position=position_dodge(0.9)) + labs(x = "Amino Acid", y = "Concentration (M)") + theme_classic() + theme(axis.text.x = element_text(angle = 90), legend.position="botto m", text = element_text(size = 12)) + guides(fill=guide_legend(title="")) Analysis of Algal Amino Acid Solution Amino Acid Concentrations after each Processing Step # Organize Data for ANOVA lcms.processing.data <- LCMS.processing.data lcms.processing.data$Type <- factor(lcms.processing.data$Type) summary(lcms.processing.data) 89 ## Column1 Type Experiment Name ## Min. : 1 AA Product :80 Min. :1.000 Length:30 0 ## 1st Qu.: 4 Acidified Centrifugate:60 1st Qu.:1.000 Class :ch aracter ## Median : 8 Centrifugate :80 Median :3.000 Mode :ch aracter ## Mean : 8 Parr Slurry :80 Mean :2.533 ## 3rd Qu.:12 3rd Qu.:4.000 ## Max. :15 Max. :4.000 ## MW g/mol N Raw Dilution LCMS Dilution ## Min. : 75.07 Min. :1.00 Min. :5000 Min. :1 ## 1st Qu.:118.60 1st Qu.:1.00 1st Qu.:5000 1st Qu.:1 ## Median :132.60 Median :1.00 Median :5000 Median :1 ## Mean :136.90 Mean :1.40 Mean :5000 Mean :1 ## 3rd Qu.:150.70 3rd Qu.:1.25 3rd Qu.:5000 3rd Qu.:1 ## Max. :204.20 Max. :4.00 Max. :5000 Max. :1 ## Diluted Conc. (uM) Undiluted Conc. (uM) AA.Conc.gperL AA.Con c.M ## Min. : 0.0000 Min. : 0 Min. : 0.0000 Min. : 0.000000 ## 1st Qu.: 0.3267 1st Qu.: 1634 1st Qu.: 0.1855 1st Qu.: 0.001634 ## Median : 3.7105 Median : 18553 Median : 2.9021 Median : 0.018552 ## Mean : 5.7316 Mean : 28658 Mean : 3.5012 Mean : 0.028658 ## 3rd Qu.: 9.0510 3rd Qu.: 45255 3rd Qu.: 5.7940 3rd Qu.: 0.045255 ## Max. :25.9380 Max. :129690 Max. :15.4220 Max. : 0.129690 # Conduct ANOVA and Pairwise Comparisons algae.proc.lm <- lm(AA.Conc.M ~ Type, lcms.processing.data) joint_tests(algae.proc.lm) #Type 3 ANOVA Table ## model term df1 df2 F.ratio p.value ## Type 3 296 0.07 0.9758 lsmeans.algae.proc = emmeans(algae.proc.lm,"Type") summary(lsmeans.algae.proc) ## Type emmean SE df lower.CL upper.CL ## AA Product 0.0297 0.00353 296 0.0228 0.0367 ## Acidified Centrifugate 0.0291 0.00408 296 0.0211 0.0371 ## Centrifugate 0.0283 0.00353 296 0.0214 0.0353 ## Parr Slurry 0.0276 0.00353 296 0.0206 0.0345 90 ## ## Confidence level used: 0.95 plot(lsmeans.algae.proc, comparisons=TRUE, xlab = "Mean Amino Acid Con centration (M)", ylab = "Processing Step") pairs(lsmeans.algae.proc) ## contrast estimate SE df t.ratio p.value ## AA Product - Acidified Centrifugate 0.000613 0.0054 296 0.114 0.9995 ## AA Product - Centrifugate 0.001399 0.0050 296 0.280 0.9923 ## AA Product - Parr Slurry 0.002174 0.0050 296 0.435 0.9724 ## Acidified Centrifugate - Centrifugate 0.000786 0.0054 296 0.146 0.9989 ## Acidified Centrifugate - Parr Slurry 0.001561 0.0054 296 0.289 0.9916 ## Centrifugate - Parr Slurry 0.000775 0.0050 296 0.155 0.9987 ## ## P value adjustment: tukey method for comparing a family of 4 estima tes multcomp::cld(lsmeans.algae.proc, by = NULL, Letters = "abcdefg", alph a = .05) # Tukey ## Type emmean SE df lower.CL upper.CL .group ## Parr Slurry 0.0276 0.00353 296 0.0206 0.0345 a ## Centrifugate 0.0283 0.00353 296 0.0214 0.0353 a ## Acidified Centrifugate 0.0291 0.00408 296 0.0211 0.0371 a ## AA Product 0.0297 0.00353 296 0.0228 0.0367 a ## ## Confidence level used: 0.95 ## P value adjustment: tukey method for comparing a family of 4 estima tes ## significance level used: alpha = 0.05 Percent Mass Liquid Yield after Centrifuge Plot and Analysis # Organize data for plotting cent.exp.1 <- centrifugate.data %>% group_by(experiment.number) %>% 91 summarize(m = mean(perc.mass.liquid, na.rm = T), stdev = sd(perc.mas s.liquid, na.rm = T)) ## `summarise()` ungrouping output (override with `.groups` argument) cent.exp.1$experiment.number <- as.factor(cent.exp.1$experiment.number ) # Plot Percent Mass Liquid Yield after Centrifuge Data ggplot(cent.exp.1, aes(fill = experiment.number, x= experiment.number, y = m)) + geom_bar(position = position_dodge(), stat = "identity") + geom_errorbar(aes(ymin=m-stdev, ymax=m+stdev), width=0.2, position=p osition_dodge(0.9)) + labs(x = "Experiment Number", y = "Percent Mass Yield") + theme_classic() + theme(legend.position = "none") # ANOVA Analysis summary(aov(perc.mass.liquid~as.factor(experiment.number), data = cent rifugate.data)) ## Df Sum Sq Mean Sq F value Pr(>F) ## as.factor(experiment.number) 3 0.14588 0.04863 359.4 <2e-16 *** ## Residuals 92 0.01245 0.00014 ## --- ## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1 ## 16 observations deleted due to missingness cent.mass.sas <- lm(perc.mass.liquid~as.factor(experiment.number),data = centrifugate.data) summary(cent.mass.sas) ## ## Call: ## lm(formula = perc.mass.liquid ~ as.factor(experiment.number), ## data = centrifugate.data) ## ## Residuals: ## Min 1Q Median 3Q Max ## -0.042226 -0.007986 0.001870 0.009704 0.021917 ## ## Coefficients: ## Estimate Std. Error t value Pr(>|t|) ## (Intercept) 0.661556 0.002480 266.759 < 2e-16 *** 92 ## as.factor(experiment.number)2 -0.079112 0.003507 -22.557 < 2e-16 *** ## as.factor(experiment.number)3 0.030482 0.003433 8.878 5.23e-14 *** ## as.factor(experiment.number)4 -0.008276 0.003314 -2.497 0.0143 * ## --- ## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1 ## ## Residual standard error: 0.01163 on 92 degrees of freedom ## (16 observations deleted due to missingness) ## Multiple R-squared: 0.9214, Adjusted R-squared: 0.9188 ## F-statistic: 359.4 on 3 and 92 DF, p-value: < 2.2e-16 lsmeans_cent.mass = emmeans(cent.mass.sas,"experiment.number") summary(lsmeans_cent.mass) ## experiment.number emmean SE df lower.CL upper.CL ## 1 0.662 0.00248 92 0.657 0.666 ## 2 0.582 0.00248 92 0.578 0.587 ## 3 0.692 0.00237 92 0.687 0.697 ## 4 0.653 0.00220 92 0.649 0.658 ## ## Confidence level used: 0.95 plot(lsmeans_cent.mass, comparisons=TRUE) pairs(lsmeans_cent.mass) # Shows which comparison has the most signifi cant difference between means. All except 1-4 are very significant ## contrast estimate SE df t.ratio p.value ## 1 - 2 0.07911 0.00351 92 22.557 <.0001 ## 1 - 3 -0.03048 0.00343 92 -8.878 <.0001 ## 1 - 4 0.00828 0.00331 92 2.497 0.0670 ## 2 - 3 -0.10959 0.00343 92 -31.920 <.0001 ## 2 - 4 -0.07084 0.00331 92 -21.375 <.0001 ## 3 - 4 0.03876 0.00324 92 11.978 <.0001 ## ## P value adjustment: tukey method for comparing a family of 4 estima tes Algal Amino Acid Solution Liquid Mass after Acidification # Calucate mean and standard deviation for plotting mean(acidification.data$mass.liquid[c(1,4)]) 93 ## [1] 459.05 sd(acidification.data$mass.liquid[c(1,4)]) ## [1] 45.32554 # Plot the mean liquid mass after acidification with error bars ggplot(data = acidification.data, aes(x = "Acidification", y = mean(m ass.liquid[c(1,4)]))) + geom_bar(position = position_dodge(), stat = "identity") + geom_errorbar(aes(ymin=mean(mass.liquid[c(1,4)])-sd(mass.liquid[c(1, 4)]), ymax=mean(mass.liquid[c(1,4)])+sd(mass.liquid[c(1,4)])), width=0 .2, position=position_dodge(0.9)) + labs(x = "", y = "Liquid Mass After Acidification (g)") + theme_classic() theme(legend.position = "none") ## List of 1 ## $ legend.position: chr "none" ## - attr(*, "class")= chr [1:2] "theme" "gg" ## - attr(*, "complete")= logi FALSE ## - attr(*, "validate")= logi TRUE Plot of Amino Acid Concentrations (M) after each Processing Step for Each Experimental Replicate # Plot Amnio Acid Concentrations after Parr Reactor for each Experimen t ggplot(filter(select(LCMS.processing.data, Type, Experiment, Name, AA. Conc.M), Type == "Parr Slurry"), aes(fill = as.factor(Experiment), x = Name, y = AA.Conc.M)) + geom_bar(position = position_dodge(), stat = "identity") + theme_classic() + theme(axis.text.x = element_text(angle = 90), text = element_text(si ze = 12))+ labs(x = "Amino Acid", y = "Concentration (M)") + scale_fill_discrete(name = "Experiment Number") # Plot Amnio Acid Concentrations after Centrifuge for each Experiment ggplot(filter(select(LCMS.processing.data, Type, Experiment, Name, AA. Conc.M), Type == "Centrifugate"), aes(fill = as.factor(Experiment), x = Name, y = AA.Conc.M)) + geom_bar(position = position_dodge(), stat = "identity") + theme_classic() + theme(axis.text.x = element_text(angle = 90), text = element_text(si 94 ze = 12))+ labs(x = "Amino Acid", y = "Concentration (M)") + scale_fill_discrete(name = "Experiment Number") # Plot Amnio Acid Concentrations after Acidification for each Experime nt ggplot(filter(select(LCMS.processing.data, Type, Experiment, Name, AA. Conc.M), Type == "Acidified Centrifugate"), aes(fill = as.factor(Exper iment), x = Name, y = AA.Conc.M)) + geom_bar(position = position_dodge(), stat = "identity") + theme_classic() + theme(axis.text.x = element_text(angle = 90), text = element_text(si ze = 12))+ labs(x = "Amino Acid", y = "Concentration (M)") + scale_fill_discrete(name = "Experiment Number") # Plot Amnio Acid Concentrations after Desorption for each Experiment ggplot(filter(select(LCMS.processing.data, Type, Experiment, Name, AA. Conc.M), Type == "AA Product"), aes(fill = as.factor(Experiment), x = Name, y = AA.Conc.M)) + geom_bar(position = position_dodge(), stat = "identity") + theme_classic() + theme(axis.text.x = element_text(angle = 90), text = element_text(si ze = 12))+ labs(x = "Amino Acid", y = "Concentration (M)") + scale_fill_discrete(name = "Experiment Number") Synthetic Amino Acid Plots and Analysis Gas to Liquid Flow Ratio (G/L) Analysis # Convert G/L ratio as factor g.l.ratio.data$g.l.ratio <- as.factor(g.l.ratio.data$g.l.ratio) summary(g.l.ratio.data) ## row.number covariate.value g.l.ratio CO2.abs. initia l.ph. ## Min. : 1.00 Min. :1 0.5 :2 Min. :0.2176 Min. :12.63 ## 1st Qu.: 3.25 1st Qu.:2 1.875:2 1st Qu.:0.5316 1st Qu. :12.86 ## Median : 5.50 Median :3 2.5 :2 Median :0.6603 Median 95 :12.88 ## Mean : 5.50 Mean :3 3 :2 Mean :0.5762 Mean :12.88 ## 3rd Qu.: 7.75 3rd Qu.:4 4 :2 3rd Qu.:0.7032 3rd Qu. :12.93 ## Max. :10.00 Max. :5 Max. :0.7678 Max. :12.96 ## abs.ph. ## Min. :8.020 ## 1st Qu.:8.605 ## Median :8.670 ## Mean :8.695 ## 3rd Qu.:8.845 ## Max. :9.580 # Conduct ANOVA Analysis g.l.ratio.lm <- lm(CO2.abs./1.009356 ~ g.l.ratio, g.l.ratio.data) joint_tests(g.l.ratio.lm) #Type 3 ANOVA Table ## model term df1 df2 F.ratio p.value ## g.l.ratio 4 5 115.047 <.0001 # Look at Pairwise Comparisons lsmeans.g.l.ratio = emmeans(g.l.ratio.lm,"g.l.ratio") summary(lsmeans.g.l.ratio) ## g.l.ratio emmean SE df lower.CL upper.CL ## 0.5 0.223 0.0195 5 0.173 0.274 ## 1.875 0.662 0.0195 5 0.612 0.712 ## 2.5 0.687 0.0195 5 0.637 0.737 ## 3 0.746 0.0195 5 0.696 0.796 ## 4 0.535 0.0195 5 0.485 0.586 ## ## Confidence level used: 0.95 par(cex.lab=2, cex.axis=2) plot(lsmeans.g.l.ratio, comparisons=TRUE, xlab = "Mean Absorption (mol CO2/L)", ylab = "G/L Flow Ratio") pairs(lsmeans.g.l.ratio) ## contrast estimate SE df t.ratio p.value ## 0.5 - 1.875 -0.4389 0.0275 5 -15.939 0.0001 ## 0.5 - 2.5 -0.4635 0.0275 5 -16.835 0.0001 ## 0.5 - 3 -0.5225 0.0275 5 -18.975 <.0001 ## 0.5 - 4 -0.3120 0.0275 5 -11.330 0.0005 96 ## 1.875 - 2.5 -0.0247 0.0275 5 -0.897 0.8870 ## 1.875 - 3 -0.0836 0.0275 5 -3.036 0.1305 ## 1.875 - 4 0.1269 0.0275 5 4.608 0.0292 ## 2.5 - 3 -0.0589 0.0275 5 -2.140 0.3302 ## 2.5 - 4 0.1516 0.0275 5 5.505 0.0140 ## 3 - 4 0.2105 0.0275 5 7.645 0.0033 ## ## P value adjustment: tukey method for comparing a family of 5 estima tes multcomp::cld(lsmeans.g.l.ratio, by = NULL, Letters = "abcdefg", alpha = .05) # Tukey ## g.l.ratio emmean SE df lower.CL upper.CL .group ## 0.5 0.223 0.0195 5 0.173 0.274 a ## 4 0.535 0.0195 5 0.485 0.586 b ## 1.875 0.662 0.0195 5 0.612 0.712 c ## 2.5 0.687 0.0195 5 0.637 0.737 c ## 3 0.746 0.0195 5 0.696 0.796 c ## ## Confidence level used: 0.95 ## P value adjustment: tukey method for comparing a family of 5 estima tes ## significance level used: alpha = 0.05 Synthetic Amino Acid Absorption Cycle Plots GAPL.amine.conc <- 1.009356 # from LCMS data (averaging the results fo r each AA) GAPL.cycles.data$cycle.number <- factor(GAPL.cycles.data$cycle.number) # make cycle number a factor # Organize Synthetic Absorption Cycle data for mol/mol Amine plot GAPL.averaged <- GAPL.cycles.data %>% group_by(cycle.number) %>% summarize(m.abs = mean(CO2.abs/GAPL.amine.conc), m.des = mean(CO2.de s/GAPL.amine.conc), sd.abs = sd(CO2.abs/GAPL.amine.conc), sd.des = sd( CO2.des/GAPL.amine.conc), .groups = 'drop') x.GAPL <- GAPL.cycles.data$cycle.number y.GAPL <- c(GAPL.averaged$m.abs, GAPL.averaged$m.des) error.GAPL <- c(GAPL.averaged$sd.abs, GAPL.averaged$sd.des) fill.GAPL <- c(rep("Absorption", 4), rep("Desorption", 4)) GAPL.plot.1 <- data.frame(x.GAPL, y.GAPL, error.GAPL, fill.GAPL) # Plot Absorption Cycle Data for Synthetic AA Absorbent in mol/mol Ami 97 ne ggplot(GAPL.plot.1, aes(x= x.GAPL, y = y.GAPL, fill = fill.GAPL)) + geom_bar(position = position_dodge(), stat = "identity") + geom_errorbar(aes(ymin=y.GAPL-error.GAPL, ymax=y.GAPL+error.GAPL), w idth=0.2, position=position_dodge(0.9)) + labs(x = "Cycle Number") + theme_classic() + theme(legend.position="bottom", text = element_text(size = 16)) + guides(fill=guide_legend(title="")) + ylab(expression(atop("Carbon Dioxide Absorbed/", paste("Released (mo l/mol Amine)")))) # Organize Synthetic Absorption Cycle data for mol/L plot GAPL.averaged.2 <- GAPL.cycles.data %>% group_by(cycle.number) %>% summarize(m.abs = mean(CO2.abs), m.des = mean(CO2.des), sd.abs = sd( CO2.abs), sd.des = sd(CO2.des), .groups = 'drop') x.GAPL.2 <- GAPL.cycles.data$cycle.number y.GAPL.2 <- c(GAPL.averaged.2$m.abs, GAPL.averaged.2$m.des) error.GAPL.2 <- c(GAPL.averaged.2$sd.abs, GAPL.averaged.2$sd.des) fill.GAPL.2 <- c(rep("Absorption", 4), rep("Desorption", 4)) GAPL.plot.2 <- data.frame(x.GAPL.2, y.GAPL.2, error.GAPL.2, fill.GAPL. 2) # Plot Absorption Cycle Data for Synthetic AA Absorbent in mol/L ggplot(GAPL.plot.2, aes(x= x.GAPL.2, y = y.GAPL.2, fill = fill.GAPL.2) ) + geom_bar(position = position_dodge(), stat = "identity") + geom_errorbar(aes(ymin=y.GAPL.2-error.GAPL.2, ymax=y.GAPL.2+error.GA PL.2), width=0.2, position=position_dodge(0.9)) + labs(x = "Cycle Number") + theme_classic() + theme(legend.position="bottom", text = element_text(size = 16)) + guides(fill=guide_legend(title="")) + ylab(expression(atop("Carbon Dioxide Absorbed/", paste("Released (mo l/L)")))) Synthetic Amino Acid Absorbent Absorption Cycle Analysis # Convert cycle number to a factor GAPL.cycles.data$cycle.number <- as.factor(GAPL.cycles.data$cycle.numb 98 er) summary(GAPL.cycles.data) ## row.number cycle.number CO2.abs CO2.des pH .initial ## Min. :1.00 1:2 Min. :0.3252 Min. :0.2814 Min. :11.24 ## 1st Qu.:2.75 2:2 1st Qu.:0.3340 1st Qu.:0.2899 1st Qu.:11.45 ## Median :4.50 3:2 Median :0.3623 Median :0.2935 Medi an :11.49 ## Mean :4.50 4:2 Mean :0.3801 Mean :0.3011 Mean :11.82 ## 3rd Qu.:6.25 3rd Qu.:0.4030 3rd Qu.:0.3143 3rd Qu.:11.87 ## Max. :8.00 Max. :0.4867 Max. :0.3317 Max. :13.01 ## pH.abs pH.des ## Min. :9.370 Min. :11.24 ## 1st Qu.:9.615 1st Qu.:11.38 ## Median :9.710 Median :11.46 ## Mean :9.671 Mean :11.43 ## 3rd Qu.:9.748 3rd Qu.:11.49 ## Max. :9.870 Max. :11.52 # Conduct ANOVA and Pairwise Comparisons for Absorption Capacity acros s Cycles GAPL.cycles.lm <- lm(CO2.abs/GAPL.amine.conc ~ cycle.number, GAPL.cycl es.data) joint_tests(GAPL.cycles.lm) #Type 3 ANOVA Table ## model term df1 df2 F.ratio p.value ## cycle.number 3 4 8.99 0.0299 lsmeans.GAPL.cycles.abs = emmeans(GAPL.cycles.lm,"cycle.number") summary(lsmeans.GAPL.cycles.abs) ## cycle.number emmean SE df lower.CL upper.CL ## 1 0.466 0.0199 4 0.410 0.521 ## 2 0.348 0.0199 4 0.293 0.404 ## 3 0.338 0.0199 4 0.283 0.393 ## 4 0.354 0.0199 4 0.299 0.409 ## ## Confidence level used: 0.95 99 par(cex.lab=2, cex.axis=2) plot(lsmeans.GAPL.cycles.abs, comparisons=TRUE, xlab = "Mean Absorptio n Capacity (mol CO2/L)", ylab = "Cycle Number") pairs(lsmeans.GAPL.cycles.abs) ## contrast estimate SE df t.ratio p.value ## 1 - 2 0.11720 0.0282 4 4.159 0.0466 ## 1 - 3 0.12753 0.0282 4 4.526 0.0353 ## 1 - 4 0.11149 0.0282 4 3.957 0.0547 ## 2 - 3 0.01033 0.0282 4 0.367 0.9809 ## 2 - 4 -0.00571 0.0282 4 -0.203 0.9966 ## 3 - 4 -0.01604 0.0282 4 -0.569 0.9363 ## ## P value adjustment: tukey method for comparing a family of 4 estima tes multcomp::cld(lsmeans.GAPL.cycles.abs, by = NULL, Letters = "abcdefg", alpha = .05) # Tukey ## cycle.number emmean SE df lower.CL upper.CL .group ## 3 0.338 0.0199 4 0.283 0.393 a ## 2 0.348 0.0199 4 0.293 0.404 a ## 4 0.354 0.0199 4 0.299 0.409 ab ## 1 0.466 0.0199 4 0.410 0.521 b ## ## Confidence level used: 0.95 ## P value adjustment: tukey method for comparing a family of 4 estima tes ## significance level used: alpha = 0.05 # Conduct ANOVA and Pairwise Comparisons for Desorption Capacity acros s Cycles GAPL.cycles.des.lm <- lm(CO2.des/GAPL.amine.conc ~ cycle.number, GAPL. cycles.data) joint_tests(GAPL.cycles.des.lm) #Type 3 ANOVA Table ## model term df1 df2 F.ratio p.value ## cycle.number 3 4 0.494 0.7056 lsmeans.GAPL.cycles.des = emmeans(GAPL.cycles.des.lm,"cycle.number") summary(lsmeans.GAPL.cycles.des) ## cycle.number emmean SE df lower.CL upper.CL ## 1 0.287 0.0136 4 0.249 0.325 ## 2 0.296 0.0136 4 0.258 0.333 100 ## 3 0.301 0.0136 4 0.263 0.338 ## 4 0.310 0.0136 4 0.272 0.348 ## ## Confidence level used: 0.95 par(cex.lab=2, cex.axis=2) plot(lsmeans.GAPL.cycles.des, comparisons=TRUE, xlab = "Mean Desorptio n Capacity (mol CO2/L)", ylab = "Cycle Number") pairs(lsmeans.GAPL.cycles.des) ## contrast estimate SE df t.ratio p.value ## 1 - 2 -0.00837 0.0192 4 -0.436 0.9690 ## 1 - 3 -0.01343 0.0192 4 -0.700 0.8924 ## 1 - 4 -0.02281 0.0192 4 -1.188 0.6642 ## 2 - 3 -0.00507 0.0192 4 -0.264 0.9926 ## 2 - 4 -0.01444 0.0192 4 -0.752 0.8718 ## 3 - 4 -0.00937 0.0192 4 -0.488 0.9577 ## ## P value adjustment: tukey method for comparing a family of 4 estima tes multcomp::cld(lsmeans.GAPL.cycles.des, by = NULL, Letters = "abcdefg", alpha = .05) # Tukey ## cycle.number emmean SE df lower.CL upper.CL .group ## 1 0.287 0.0136 4 0.249 0.325 a ## 2 0.296 0.0136 4 0.258 0.333 a ## 3 0.301 0.0136 4 0.263 0.338 a ## 4 0.310 0.0136 4 0.272 0.348 a ## ## Confidence level used: 0.95 ## P value adjustment: tukey method for comparing a family of 4 estima tes ## significance level used: alpha = 0.05 # Conduct ANOVA and Pairwise Comparisons for Absorption pH across Cycl es GAPL.cycles.abs.ph.lm <- lm(pH.abs ~ cycle.number, GAPL.cycles.data) joint_tests(GAPL.cycles.abs.ph.lm) #Type 3 ANOVA Table ## model term df1 df2 F.ratio p.value ## cycle.number 3 4 6.377 0.0527 101 lsmeans.GAPL.cycles.abs.ph = emmeans(GAPL.cycles.abs.ph.lm,"cycle.numb er") summary(lsmeans.GAPL.cycles.abs.ph) ## cycle.number emmean SE df lower.CL upper.CL ## 1 9.82 0.0588 4 9.66 9.98 ## 2 9.73 0.0588 4 9.57 9.89 ## 3 9.47 0.0588 4 9.31 9.63 ## 4 9.66 0.0588 4 9.50 9.83 ## ## Confidence level used: 0.95 par(cex.lab=2, cex.axis=2) plot(lsmeans.GAPL.cycles.abs.ph, comparisons=TRUE, xlab = "Mean pH", y lab = "Cycle Number") pairs(lsmeans.GAPL.cycles.abs.ph) ## contrast estimate SE df t.ratio p.value ## 1 - 2 0.090 0.0831 4 1.082 0.7174 ## 1 - 3 0.350 0.0831 4 4.210 0.0449 ## 1 - 4 0.155 0.0831 4 1.864 0.3666 ## 2 - 3 0.260 0.0831 4 3.127 0.1109 ## 2 - 4 0.065 0.0831 4 0.782 0.8595 ## 3 - 4 -0.195 0.0831 4 -2.345 0.2306 ## ## P value adjustment: tukey method for comparing a family of 4 estima tes multcomp::cld(lsmeans.GAPL.cycles.abs.ph, by = NULL, Letters = "abcdef g", alpha = .05) # Tukey ## cycle.number emmean SE df lower.CL upper.CL .group ## 3 9.47 0.0588 4 9.31 9.63 a ## 4 9.66 0.0588 4 9.50 9.83 ab ## 2 9.73 0.0588 4 9.57 9.89 ab ## 1 9.82 0.0588 4 9.66 9.98 b ## ## Confidence level used: 0.95 ## P value adjustment: tukey method for comparing a family of 4 estima tes ## significance level used: alpha = 0.05 # Conduct ANOVA and Pairwise Comparisons for Desorption pH across Cycl es 102 GAPL.cycles.des.ph.lm <- lm(pH.des ~ cycle.number, GAPL.cycles.data) joint_tests(GAPL.cycles.des.ph.lm) #Type 3 ANOVA Table ## model term df1 df2 F.ratio p.value ## cycle.number 3 4 0.736 0.5829 lsmeans.GAPL.cycles.des.ph = emmeans(GAPL.cycles.des.ph.lm,"cycle.numb er") summary(lsmeans.GAPL.cycles.des.ph) ## cycle.number emmean SE df lower.CL upper.CL ## 1 11.5 0.0705 4 11.3 11.7 ## 2 11.4 0.0705 4 11.2 11.6 ## 3 11.4 0.0705 4 11.2 11.6 ## 4 11.4 0.0705 4 11.2 11.6 ## ## Confidence level used: 0.95 par(cex.lab=2, cex.axis=2) plot(lsmeans.GAPL.cycles.des.ph, comparisons=TRUE, xlab = "Mean pH", y lab = "Cycle Number") pairs(lsmeans.GAPL.cycles.des.ph) ## contrast estimate SE df t.ratio p.value ## 1 - 2 0.145 0.0997 4 1.455 0.5340 ## 1 - 3 0.055 0.0997 4 0.552 0.9413 ## 1 - 4 0.085 0.0997 4 0.853 0.8286 ## 2 - 3 -0.090 0.0997 4 -0.903 0.8054 ## 2 - 4 -0.060 0.0997 4 -0.602 0.9264 ## 3 - 4 0.030 0.0997 4 0.301 0.9892 ## ## P value adjustment: tukey method for comparing a family of 4 estima tes multcomp::cld(lsmeans.GAPL.cycles.des.ph, by = NULL, Letters = "abcdef g", alpha = .05) # Tukey ## cycle.number emmean SE df lower.CL upper.CL .group ## 2 11.4 0.0705 4 11.2 11.6 a ## 4 11.4 0.0705 4 11.2 11.6 a ## 3 11.4 0.0705 4 11.2 11.6 a ## 1 11.5 0.0705 4 11.3 11.7 a ## ## Confidence level used: 0.95 ## P value adjustment: tukey method for comparing a family of 4 estima 103 tes ## significance level used: alpha = 0.05 Algal Amino Acid Absorbent Plots and Analysis Algal Amino Acid Absorbent Amino Acid Concentration throughout Absorption Cycles # Organize Data for Plotting cycle.0.conc <- LCMS.algae.cycles.data %>% filter(Cycle == 0 & Type =="Original") %>% select(Name, AA.Conc.M) cycle.1.conc <- LCMS.algae.cycles.data %>% filter(Cycle == 1 & Type =="Absorption") %>% select(Name, AA.Conc.M) cycle.2.conc <- LCMS.algae.cycles.data %>% filter(Cycle == 2 & Type =="Absorption") %>% select(Name, AA.Conc.M) cycle.3.conc <- LCMS.algae.cycles.data %>% filter(Cycle == 3 & Type =="Absorption") %>% select(Name, AA.Conc.M) cycle.4.conc <- LCMS.algae.cycles.data %>% filter(Cycle == 4 & Type =="Absorption") %>% select(Name, AA.Conc.M) cycle.5.conc <- LCMS.algae.cycles.data %>% filter(Cycle == 5 & Type =="Absorption") %>% select(Name, AA.Conc.M) cycle.6.conc <- LCMS.algae.cycles.data %>% filter(Cycle == 6 & Type =="Absorption") %>% select(Name, AA.Conc.M) cycle.7.conc <- LCMS.algae.cycles.data %>% filter(Cycle == 7 & Type =="Absorption") %>% select(Name, AA.Conc.M) cycle.8.conc <- LCMS.algae.cycles.data %>% filter(Cycle == 8 & Type =="Absorption") %>% select(Name, AA.Conc.M) value.plot.2 <- c(cycle.0.conc$AA.Conc.M, cycle.1.conc$AA.Conc.M, cycl e.2.conc$AA.Conc.M, cycle.3.conc$AA.Conc.M, cycle.4.conc$AA.Conc.M, cy cle.5.conc$AA.Conc.M, cycle.6.conc$AA.Conc.M, cycle.7.conc$AA.Conc.M, 104 cycle.8.conc$AA.Conc.M) cycle.plot.2 <- c(rep("0", 20), rep("1", 20), rep("2", 20), rep("3", 2 0), rep("4", 20), rep("5", 20), rep("6", 20), rep("7", 20), rep("8", 2 0)) name.plot.2 <- LCMS.algae.cycles.data %>% filter(Cycle == 1 & Type =="Absorption") %>% select(Name) plot.2.data <- data.frame(cycle.plot.2, name.plot.2, value.plot.2) plot.2.data$Cycle <- as.factor(plot.2.data$cycle.plot.2) # Plot Algal Amino Acid Absorbent Amino Acid Concentrations throughout Absorption Cycles ggplot(plot.2.data, aes(x = Name, y = value.plot.2, fill = cycle.plot. 2)) + geom_bar(position = position_dodge(), stat = "identity") + theme_classic() + theme(axis.text.x = element_text(angle = 90), legend.position="botto m")+ labs(x = "Amino Acid", y = "Concentration (M)") + scale_fill_discrete(name = "Absorption Number") Algal Amino Acid Absorption Cycles Plots # Organize Data for Calculation algae.cycles.amine <- LCMS.algae.cycles.data %>% mutate(aa.conc.amine = AA.Conc.M*N) abs.aa.conc.amine.1 <- algae.cycles.amine %>% filter(Type == "Original" & Cycle == "0") abs.aa.conc.amine.2 <- algae.cycles.amine %>% filter(Type == "Desorption Diluted" & Cycle == "1") abs.aa.conc.amine.3 <- algae.cycles.amine %>% filter(Type == "Desorption Diluted" & Cycle == "2") abs.aa.conc.amine.4 <- algae.cycles.amine %>% filter(Type == "Desorption Diluted" & Cycle == "3") abs.aa.conc.amine.5 <- algae.cycles.amine %>% filter(Type == "Desorption Diluted" & Cycle == "4") abs.aa.conc.amine.6 <- algae.cycles.amine %>% filter(Type == "Desorption Diluted" & Cycle == "5") abs.aa.conc.amine.7 <- algae.cycles.amine %>% filter(Type == "Desorption Diluted" & Cycle == "6") abs.aa.conc.amine.8 <- algae.cycles.amine %>% filter(Type == "Desorption Diluted" & Cycle == "7") aa.conc.amine.1 <- algae.cycles.amine %>% 105 filter(Type == "Absorption" & Cycle == "1") aa.conc.amine.2 <- algae.cycles.amine %>% filter(Type == "Absorption" & Cycle == "2") aa.conc.amine.3 <- algae.cycles.amine %>% filter(Type == "Absorption" & Cycle == "3") aa.conc.amine.4 <- algae.cycles.amine %>% filter(Type == "Absorption" & Cycle == "4") aa.conc.amine.5 <- algae.cycles.amine %>% filter(Type == "Absorption" & Cycle == "5") aa.conc.amine.6 <- algae.cycles.amine %>% filter(Type == "Absorption" & Cycle == "6") aa.conc.amine.7 <- algae.cycles.amine %>% filter(Type == "Absorption" & Cycle == "7") aa.conc.amine.8 <- algae.cycles.amine %>% filter(Type == "Absorption" & Cycle == "8") # Calculate the Amino Acid Concentration after each Absorption/Desorpt ion amine.vec <- c(sum(abs.aa.conc.amine.1$aa.conc.amine), sum(abs.aa.conc .amine.2$aa.conc.amine), sum(abs.aa.conc.amine.3$aa.conc.amine), NA, s um(abs.aa.conc.amine.5$aa.conc.amine), NA, sum(abs.aa.conc.amine.7$aa. conc.amine), sum(abs.aa.conc.amine.8$aa.conc.amine)) des.amine.vec <- c(sum(aa.conc.amine.1$aa.conc.amine), sum(aa.conc.ami ne.2$aa.conc.amine), sum(aa.conc.amine.3$aa.conc.amine), NA, sum(aa.co nc.amine.5$aa.conc.amine), NA, sum(aa.conc.amine.7$aa.conc.amine), sum (aa.conc.amine.8$aa.conc.amine)) # Calculation to adjust for water dilution volume.adj.vec.abs <- c(0.3, 0.24, 0.223024421, NA, 0.199450877, NA, 0 .18205885 , 0.17740077) volume.adj.vec.des <- c(0.205, 0.229875, 0.213459057, NA, 0.193800216, NA, 0.180512888, 0.175345759) algae.cycles.data <- algae.cycles.data %>% mutate(CO2.abs.amine = CO2.abs/amine.vec/volume.abs*volume.adj.vec.a bs) %>% mutate(CO2.des.amine = CO2.des/des.amine.vec/volume.des*volume.adj.v ec.des) # Organize Data for Plotting cycle.number.plot.3 <- c(algae.cycles.data$cycle.number, algae.cycles. data$cycle.number) value.plot.3 <- c(algae.cycles.data$CO2.abs.amine, algae.cycles.data$C O2.des.amine) name.plot.3 <- c(rep("Absorption", 8), rep("Desorption", 8)) 106 plot.3.data <- data.frame(cycle.number.plot.3, name.plot.3, value.plot .3) plot.3.data$name.plot.3 <- as.factor(plot.3.data$name.plot.3) # Plot Algal Amino Acid Absorbent Absorption Cycle Data in mol/mol Ami ne ggplot(plot.3.data, aes(x = cycle.number.plot.3, y = value.plot.3, fil l = name.plot.3)) + geom_bar(position = position_dodge(), stat = "identity") + labs(x = "Cycle Number") + theme_classic() + theme(legend.position="bottom", text = element_text(size = 16)) + guides(fill=guide_legend(title="")) + scale_x_continuous(breaks = seq(1, 8, by = 1)) + ylab(expression(atop("Carbon Dioxide Absorbed/", paste("Released (mo l/mol Amine)")))) ## Warning: Removed 4 rows containing missing values (geom_bar). # Output Mean and Standard Deviation for mol/mol Amine Absorption and Desorption Results excluding Absorption 1 mean(algae.cycles.data$CO2.abs.amine[-1], na.rm = T) ## [1] 1.271044 sd(algae.cycles.data$CO2.abs.amine[-1], na.rm = T) ## [1] 0.06145801 mean(algae.cycles.data$CO2.des.amine[-1], na.rm = T) ## [1] 1.177415 sd(algae.cycles.data$CO2.des.amine[-1], na.rm = T) ## [1] 0.09356362 # Organize Data for Plotting cycle.number.plot.4 <- c(algae.cycles.data$cycle.number, algae.cycles. data$cycle.number) value.plot.4 <- c(algae.cycles.data$CO2.abs, algae.cycles.data$CO2.des ) name.plot.4 <- c(rep("Absorption", 8), rep("Desorption", 8)) plot.4.data <- data.frame(cycle.number.plot.4, name.plot.4, value.plot .4) plot.4.data$name.plot.4 <- as.factor(plot.4.data$name.plot.4) 107 # Plot Algal Amino Acid Absorbent Absorption Cycle Data in mol/L ggplot(plot.4.data, aes(x = cycle.number.plot.4, y = value.plot.4, fil l = name.plot.4)) + geom_bar(position = position_dodge(), stat = "identity") + labs(x = "Cycle Number") + theme_classic() + theme(legend.position="bottom", text = element_text(size = 16)) + guides(fill=guide_legend(title="")) + scale_x_continuous(breaks = seq(1, 8, by = 1)) + ylab(expression(atop("Carbon Dioxide Absorbed/", paste("Released (mo l/L)")))) ## Warning: Removed 4 rows containing missing values (geom_bar). # Output Mean and Standard Deviation for mol/L Absorption Results excl uding Absorption 1 mean(algae.cycles.data$CO2.abs[-1], na.rm = T) ## [1] 0.8252807 sd(algae.cycles.data$CO2.abs[-1], na.rm = T) ## [1] 0.06463184 Synthetic and Algal Amino Acid Absorbent Absorption Capacity Analysis # Data Organization GAPL.comparison.L <- c(0.391181771, 0.376117742) GAPL.comparison.Amine <- c(0.387555799, 0.372631403) algal.comparison.L <- plot.4.data$value.plot.4[c(2,3,5,7,8)] algal.comparison.Amine <- plot.3.data$value.plot.3[c(2,3,5,7,8)] comparison.L <- c(GAPL.comparison.L, algal.comparison.L) comparison.Amine <- c(GAPL.comparison.Amine, algal.comparison.Amine) comparisons.factor <- as.factor(c("synthetic", "synthetic", rep("algal ", 5))) comparison.L.data <- data.frame(comparisons.factor, comparison.L) comparison.Amine.data <- data.frame(comparisons.factor, comparison.Ami ne) # ANOVA and Pairwise Comparisons for Absorption Capacity in mol/L comparison.L.lm <- lm(comparison.L ~ comparisons.factor, comparison.L. data) joint_tests(comparison.L.lm) #Type 3 ANOVA Table ## model term df1 df2 F.ratio p.value ## comparisons.factor 1 5 82.813 0.0003 108 lsmeans.comparison.L = emmeans(comparison.L.lm,"comparisons.factor") summary(lsmeans.comparison.L) ## comparisons.factor emmean SE df lower.CL upper.CL ## algal 0.825 0.0259 5 0.759 0.892 ## synthetic 0.384 0.0410 5 0.278 0.489 ## ## Confidence level used: 0.95 plot(lsmeans.comparison.L, comparisons=TRUE) pairs(lsmeans.comparison.L) ## contrast estimate SE df t.ratio p.value ## algal - synthetic 0.442 0.0485 5 9.100 0.0003 multcomp::cld(lsmeans.comparison.L, by = NULL, Letters = "abcdefg", al pha = .05) # Tukey ## comparisons.factor emmean SE df lower.CL upper.CL .group ## synthetic 0.384 0.0410 5 0.278 0.489 a ## algal 0.825 0.0259 5 0.759 0.892 b ## ## Confidence level used: 0.95 ## significance level used: alpha = 0.05 # ANOVA and Pairwise Comparisons for Absorption Capacity in mol/mol Am ine comparison.Amine.lm <- lm(comparison.Amine ~ comparisons.factor, compa rison.Amine.data) joint_tests(comparison.Amine.lm) #Type 3 ANOVA Table ## model term df1 df2 F.ratio p.value ## comparisons.factor 1 5 372.539 <.0001 lsmeans.comparison.Amine = emmeans(comparison.Amine.lm,"comparisons.fa ctor") summary(lsmeans.comparison.Amine) ## comparisons.factor emmean SE df lower.CL upper.CL ## algal 1.27 0.0247 5 1.21 1.33 ## synthetic 0.38 0.0390 5 0.28 0.48 ## ## Confidence level used: 0.95 plot(lsmeans.comparison.Amine, comparisons=TRUE) 109 pairs(lsmeans.comparison.Amine) ## contrast estimate SE df t.ratio p.value ## algal - synthetic 0.891 0.0462 5 19.301 <.0001 multcomp::cld(lsmeans.comparison.Amine, by = NULL, Letters = "abcdefg" , alpha = .05) # Tukey ## comparisons.factor emmean SE df lower.CL upper.CL .group ## synthetic 0.38 0.0390 5 0.28 0.48 a ## algal 1.27 0.0247 5 1.21 1.33 b ## ## Confidence level used: 0.95 ## significance level used: alpha = 0.05 110 APPENDIX E: ALGAL BIOMASS INFORMATION Table 19: C. sorokiniana Biomass Composition Components As Fed Dry Matter Moisture (%) 6.9 Dry Matter (%) 93.1 Crude Protein (%) 54.9 59.0 Soluble Protein (% CP) 25 NDICP (%) 1.3 1.4 ADF (%) 5.6 6.0 Lignin (%) 3.8 4.1 Starch (%) 2.3 2.5 ESC “Simple Sugars” (%) 3.0 3.2 Total Fatty Acids (%) 8.27 8.89 RUFAL (%) 3.82 4.10 Ash (%) 8.91 9.57 Calcium (%) 0.52 0.56 Phosphorus (%) 1.70 1.83 Magnesium (%) 0.27 0.29 Potassium (%) 0.92 0.98 Sodium (%) 0.014 0.015 Iron (ppm) 12,500 13,400 Zinc (ppm) 266 285 Copper (ppm) 218 234 Manganese (ppm) 34 36 Molybdenum (ppm) <0.1 <0.1 Sulfur (%) 0.73 0.78 Chloride Ion (%) 0.06 0.06 DCAD (mEq/100g) -25 *: Results from DairyOne, Inc. 111 Table 20: C. sorokiniana Fatty Acid Composition Fatty Acid Percent of Total Fatty Percent of Dry Matter Acids C12:0 Lauric 0.05 0.01 C14:0 Myristic 0.37 0.03 C16:0 Palmitic 18.42 1.67 C16:1 Palmitoleic 1.10 0.10 C18:0 Stearic 1.06 0.09 C18:1 Oleic 11.15 0.98 C18:2 Linoleic 22.92 2.02 C18:3 Linolenic 12.57 1.10 C20:0 Arachidic 0.11 0.01 C20:1 Gadoleic 0.09 0.01 C20:5 Eicosapentaenoic 0.00 0.00 (EPA) C22:0 Behenic 0.09 0.01 C22:6 Docosahexanoic 0.00 0.00 (DHA) C24:0 Lignoceric 0.00 0.00 Other 32.06 2.85 Total Fatty Acids 100.00 8.89 Saturated 20.11 MUFA 12.34 PUFA 35.49 RUFAL 4.10 *: Results from DairyOne, Inc. 112 APPENDIX F: DAIRY ONE FORAGE LABORATORY PROCEDURES Ash, Total AOAC Method 942.05 – Ash of Animal Feed. Carbohydrates, Soluble Ethanol Soluble Carbohydrates (ESC) Hall, M.B., W.H. Hoover, J.P. Jennings and T.K. Miller Webster. 1999. A method for partitioning neutral detergent soluble carbohydrates. J. Sci. Food Agric. 79: p.2079-2086. Samples shaken for 4 hours at 180 epm with 80% ethanol to extract ethanol soluble carbohydrates comprised of simple sugars. ESC determined using a Thermo Scientific Genesys 10S Vis Spectrophotometer after a colorimetric phenol-sulfuric acid reaction. Water Soluble Carbohydrates (WSC) West Virginia University Procedure by W.H. Hoover and T.K. Miller Webster. Determination of Nonstructural Carbohydrates. Hall, M.B., W.H. Hoover, J.P. Jennings and T.K. Miller Webster. 1999. A method for partitioning neutral detergent soluble carbohydrates. J. Sci. Food Agric. 79: p.2081. Samples incubated with water in a 40ºC bath for 1 hour extracting water soluble carbohydrates comprised of simple sugars and fructan. WSC determined using a Thermo Scientific Genesys 10S Vis Spectrophotometer after acid hydrolysis with sulfuric acid and colorimetric reaction with potassium ferricyanide. Dry Matter (DM) Oven – 60ºC for 4 hours (forced air) Goering, H.K. and P.J. Van Soest. 1970. Forage Fiber Analyses (apparatus, reagents, procedures, and some applications). ARS/USDA Handbook No. 379, Superintendent of Documents, US Government Printing Office, Washington, D.C. 20402. P15. NFTA Method 2.2.1.1 – Partial Dry Matter using Forced-air Drying Ovens. Oven – 135ºC for 2 hours AOAC 930.15 – Loss on Drying (Moisture) for Feeds. Oven – 105ºC for 3 hours NFTA Method 2.2.2.5 – Dry Matter by Oven Drying for 3hr at 105C. Fat 113 Crude, Acid Hydrolysis AOAC 954.02 – Crude Fat in Pet Food. Crude, Ether Extraction AOAC 2003.05 – Crude Fat in Feeds, Cereal Grains, and Forages. Dairy One Forage Lab, Equi-Analytical, Zooquarius Analytical Procedures Page 3 of 10 Extraction by Soxtec HT6 System using anhydrous diethyl ether. Crude fat residue determined gravimetrically after drying. Foss North America, 8091 Wallace Road, Eden Praire, MN 55344. www.foss.us Crude, Roese-Gottlieb Method (Base Hydrolysis) AOAC 932.06 A (b) and 932.06 B – Fat in Dried Milk. Used for milk (liquid and powder), whey, and milk based byproducts. Fatty Acids Total Fatty Acids (TFA) Direct FAME synthesis O’Fallon, J.V., J.R. Busboom, M.L. Nelson and C.T. Gaskins. 2007. A direct method for fatty acid methyl ester synthesis: Application to wet meat tissues, oils, and feedstuffs. J. Anim. Sci. 85: p.1511-1521. Fatty acid methyl esters (FAME) determined directly from fresh tissue, oils, or feedstuffs, without the need for prior organic solvent extraction. FAME synthesis is conducted in the presence of up to 33% water. Wet tissues or other samples are permeabilized and hydrolyzed for 1.5 hr. at 55C in 1N KOH in MeOH containing C13:0 as an internal standard. The KOH is neutralized, and the FFA are methylated by H2SO4 catalysis for 1.5 hr. at 55C. Hexane is the added to the reaction tube, vortex-mixed and centrifuged. The hexane layer pipetted into gas chromatography (GC) vials and then analyzed using a Thermo Trace 1310 Gas Chromotograph fitted with a Supelco SP-2560, 100m x 0.25mm x 0.20um capillary column and a Flame Ionization Detector (FID). Thermo Fisher Scientific Inc., 81 Wyman Street, Waltham, MA 02454. www.thermoscientific.com Fiber Acid Detergent Fiber (ADF) ANKOM Technology Method 12 – Acid Detergent Fiber in Feeds – Filter Bag Technique (for A2000 and A2000l), 05/19/2017. 114 Solutions as in AOAC 973.18 – Fiber (Acid Detergent) and Lignin (H2SO4) in Animal Feed. Samples individually weighed at 0.5g into filter bags and digested for 75 minutes as a group of 24 in 2L of ADF solution in ANKOM A2000 Digestion Unit. Samples are rinsed three times with boiling water for 5 minutes in filter bags followed by a 3 minute acetone soak and drying at 105ºC for 2 hours. ANKOM Technology, 2052 O’Neil Road, Macedon, NY 14502. www.ankom.com Lignin ANKOM Technology Method 9 – Method for Determining Acid Detergent Lignin in the DaisyII Incubator – 01/24/2017. Solution as in AOAC 973.18 – Fiber (Acid Detergent) and Lignin (H2SO4) in Animal Feed. ADF performed as above and residue digested as a group of 24 in 72% w/w sulfuric acid for 3 hours in ANKOM DaisyII Incubator at ambient temperature. Minerals Ca, P, Mg, K, Na, Fe, Zn, Cu, Mn, Mo, Co, S, Al, B, Cr, Sr Samples digested using CEM Microwave Accelerated Reaction System (MARS6) with MarsXpress Temperature Control using 50ml calibrated Xpress Teflon PFA vessels with Kevlar/fiberglass insulating sleeves then analyzed by ICP using a Thermo iCAP 6300 Inductively Coupled Plasma Radial Spectrometer. Sample weights – 0.5g for forages, ingredients, byproducts (1.0g for Co or Cr); 0.5g for grain mixes; 0.2g for mineral mixes; Manure - 0.5g dried, ground or 2-10g wet sample. Samples first pre-digested at ambient temperature 10 minutes with 8ml nitric acid (HNO3) and 2ml hydrochloric acid (HCl) and then an additional 10 minutes with 1ml 30% hydrogen peroxide (H2O2). After pre-digestion complete, samples digested in two stages: Stage one - 10-minute ramp to 135ºC and held for 3 minutes at 1500W. Stage two - 12-minute ramp to 200ºC and held for 15 minutes at 1600W. Vessels brought to 50-ml volume, aliquot used for analysis. Method utilized based upon CEM Application Notes for Acid Digestion on the following matrices - Feed Grain, Alfalfa, Corn Flour, Milk Powder, Soybean Meal, Flour, Hair, Potato Chips, Wheat Crackers, Peanut Butter, Urine, Dog Feces, Wine. Water – 35ul concentrated nitric acid added to 14ml of water, mixed, then aspirated on ICP for analysis. Manure Reference: Wolf, Ann, M. Watson, and N. Wolf. 2003. Digestion and dissolution methods for P, K, Ca, Mg and trace elements. Recommended methods of manure analysis. ed J. Peters, pp30-39. University of Wisconsin Extension Publication. A3769 CEM, 3100 Smith Farm Road, Matthews, NC 28106. www.cem.com 115 Thermo Fisher Scientific Inc., 81 Wyman Street, Waltham, MA 02454. www.thermoscientific.com Chloride Ion (Cl-) Potentiometric Titration – 0.2-0.5g dried, ground sample or 1-5g wet sample extracted for 15 minutes in 50ml 0.1N HNO3, followed by potentiometric titration with AgNO3 (0.01N or 0.10N) using a Metrohm 905 Titrando Titration Unit equipped with an Ag-ring electrode controlled by Metrohm Tiamo software. For water samples, 25ml of 0.2N HNO3 added to 25ml of sample. Metrohm Application Bulletin No. 130 by Metrohm Ltd., C-H-9101 Herisau, Switzerland Metrohm USA, 6555 Pelican Creek Circle, Riverview FL, 33578. www.metrohmusa.com The method by Metrohm is similar to the concepts found in: Cantliffe, D.J., MacDonald, G.E. and Peck, N.H. 1970. The potentiometric determination of nitrate and chloride in plant tissue. New York’s Food and Life Sciences Bulletin. No.3, September 1970. Plant Sciences. Vegetable Crops Geneva. No. 1: 5-7. Selenium (Se) Subcontracted to Michigan State University Veterinary Diagnostic Laboratory 4125 Beaumont Road, Lansing MI 48910-8104 Wahlen R, EvansL, Turner J, Hearn R: The use of collision/reaction cell ICP-MS for the determination of elements in blood and serum samples. Spectroscopy 20 (12): 84-89, 20050.5g aliquots of dried, ground feed samples are digested overnight at 95°C in 5mL of nitric acid. The digested samples are diluted with water to 100x the initial feed mass. 200uL of each diluted digest is pipetted and diluted with a solution containing 0.5% EDTA and Triton X-100, 1% ammonium hydroxide, 2% propanol and 20ppb of scandium, rhodium, indium and bismuth as internal standards. An Agilent Inductively Coupled Plasma – Mass Spectrometer (ICP/MS)1 is used for the analysis. The ICP/MS is tuned to yield a minimum of 7500 cps sensitivity for 1ppb yttrium (mass 89), less than 1.0% oxide level as determined by the 156/140 mass ratio and less than 2.0% double charged ions as determined by the 70/140 mass ratio. Selenium concentration is calibrated using a 6-point linear curve of the analyte-internal standard response ratio. Standards were from Inorganic Ventures2. A NIST3 Typical Diet standard was used as a control 1 Agilent Technologies Inc, Santa Clara CA 95051 2 Inorganic Ventures, Christainsburg, VA 24073 3 National Institute of Standards and Technology, Gaithersburg MD 20899 Nitrates (%NO3 or ppm NO3-N) RQflex® Reflectometer Method 116 1g of dried, ground sample or 10g of wet sample is extracted in 50ml deionized water for 20 minutes by shaking at 280 oscillations/minute. Samples are filtered through Whatman 934-AH (1.5um) filter paper, then analyzed by RQflex® Reflectometer using Reflectoquant® Nitrate test strips. When the Nitrate test strip is immersed in the aqueous sample, a reducing agent reduces nitrate ions to nitrite ions. In the presence of an acidic buffer, the nitrite ions react with an aromatic amine to form a diazonium salt. The salt reacts with N-(1-naphthyl)-ethyelene-diamine to form a red-violet azo dye that is measured reflectometrically. Nitrate concentration is proportional to the color reaction. Each strip contains two reaction zones generating dual replicate analyses per sample. The RQflex® Reflectometer’s double optic system measures the analyte concentration based on the light reflected from the dual reaction zones. Barcode controlled software calculates the mean of those two measurements. EMD Chemicals Inc., One International Plaza, Suite 300, Philadelphia, PA, 19113. www.emdmillipore.com Protein Acid Detergent Insoluble Crude Protein (ADICP) ADF residue analyzed using a Leco TruMac N Macro Determinator to determine the protein fraction bound to the acid detergent fiber. Crude Protein (CP) and Total Nitrogen (N) Dry, 1mm ground samples analyzed by combustion using a CN628 Carbon/Nitrogen Determinator. Liquid samples analyzed using a TruMac N Macro Determinator. AOAC 990.03 – Protein (Crude) in Animal Feed AOAC 992.15 – Crude Protein in Meat and Meat Products including Pet Foods AOAC 992.23 – Crude Protein in Cereal Grain and Oilseeds Leco Application Note – “Nitrogen/Protein in Feeds, Grains, and Pet Food” Form 20X-821-485, 03/15 – Rev0. Leco Application Note – “Nitrogen in Soil and Plant Tissue” Form 203-821-443, 11/14 – Rev2. Manure –Watson, M., A. Wolf, and N. Wolf. 2003. Total nitrogen. Recommended methods of manure analysis. ed J. Peters, pp18, 23-24. University of Wisconsin Extension Publication. A3769. Leco Corporation, 300 Lakeview Avenue, St. Joseph, MI 49085. www.leco.com Degradable Protein (Rumen Degradable Protein - RDP) 117 Cornell Streptomyces griseus (SGP) enzymatic digestion. Enzyme concentration held constant. Residues containing undegradable protein analyzed using Leco TruMac N Macro Determinator. Concentrates incubated for 18 hrs. Cornell Nutrition Conference Proceedings, 1990. pp. 81-88. Forage samples incubated for 2 hrs. at higher SGP concentration. J. Dairy Sci. 1999. 82: 343- 354. Leco Application Note – “Nitrogen/Protein in Feeds, Grains, and Oil Seeds” Form No. 203-821- 392, 01/16 – Rev2 Dairy One Forage Lab, Equi-Analytical, Zooquarius Analytical Procedures Page 8 of 10 Neutral Detergent Insoluble Crude Protein (NDICP) aNDF performed without sodium sulfite then residue analyzed using a Leco TruMac N Macro Determinator to determine the protein fraction bound to the neutral detergent fiber. Soluble Protein (SP) Cornell Sodium Borate-Sodium Phosphate Buffer Procedure. Soy products incubated at 39°C. All other samples incubated at ambient temperature. Residue containing insoluble protein analyzed using Leco TruMac N Macro Determinator. Cornell Nutrition Conference Proceedings, 1990, pp. 85-86. Leco Application Note – “Nitrogen/Protein in Feeds, Grains, and Oil Seeds” Form No. 203-821- 392, 01/16 – Rev2 Starch, Total YSI 2950D-1 or 2700 SELECT Biochemistry Analyzers YSI Incorporated Life Sciences, 1725 Brannum Lane, Yellow Springs, Ohio 45387 Application Note Number 319. www.ysilifescience.com Dairy One Forage Lab, Equi-Analytical, Zooquarius Analytical Procedures Page 9 of 10 Samples are pre-extracted for sugar by incubation in 40ºC water bath and filtration on Whatman 41 filter paper. Residues are thermally solubilized using an autoclave, then incubated with glucoamylase enzyme to hydrolyze starch to produce dextrose (glucose). Prepared samples injected into sample chamber of YSI Analyzer where dextrose diffuses into a membrane containing glucose oxidase. The dextrose is immediately oxidized to hydrogen peroxide and D-glucono-4-lactone. The hydrogen peroxide is detected amperometrically at the platinum electrode surface. The current flow at the electrode is directly proportional to the hydrogen peroxide concentration, and hence to the dextrose concentration. Starch is determined by multiplying dextrose by 0.9. 118 Volatile Fatty Acids (VFA) and Lactic Acid Extraction – 50g samples blended at 20000 rpm for 2 min. in 750ml deionized water (Manure 50g and 450ml water), filtered through cheesecloth, then filtered through disposable syringe filter. Adapted from Personal Communication, L.E. Chase, Ph.D., Cornell University. Gas Chromatography – Acetic, Propionic, Butyric, Iso-butyric acids Aliquot of extract mixed 1:1 ratio with 0.06M oxalic acid containing 100ppm trimethylacetic acid (internal standard). Samples injected into a Perkin Elmer Clarus 680 Gas Chromatograph containing a Supelco packed column with the following specifications: 2m x 2mm Tightspec ID, 4% Carbowax 20M phase on 80/120 Carbopack B-DA. Procedure based upon: - “GC Separation of VFA C2-C5” Supelco GC Bulletin 749F, 1975. - “Analyzing Fatty Acids by Packed Column Gas Chromatography” Supelco GC Bulletin 856A, 1990. - “Volatile Fatty Acid SOP” W.H. Miner Institute, Chazy, NY. Sigma Aldrich (Supelco), 3050 Spruce Street, St. Louis, MO 63103. www.sigmaaldrich.com Perkin Elmer, 940 Winter Street, Waltham, MA 02451. www.perkinelmer.com Biochemistry Analyzer – Lactic acid Aliquot of extract analyzed for L-Lactate using YSI 2950D-1 or 2700 SELECT Biochemistry Analyzer equipped with an LLactate membrane. YSI User’s Manual, page 4-7. Samples injected into sample chamber of YSI Analyzer where L-Lactate diffuses into a membrane containing L-Lactate oxidase. The L-Lactate is immediately oxidized to hydrogen peroxide and pyruvate. The hydrogen peroxide is detected amperometrically at the platinum electrode surface. The current flow at the electrode is directly proportional to the hydrogen peroxide concentration, and hence to the L-Lactate concentration. Total lactic acid is determined by multiplying L-Lactate by 2.0. YSI Incorporated Life Sciences, 1725 Brannum Lane, Yellow Springs, Ohio 45387. www.ysilifescience.com 119 APPENDIX G: LAB IMAGES (a) (b) Figure 83: Desorption Experimental Setup Photo a) not in use b) in use 120 (a) (b) (c) (d) Figure 84: Absorption Experimental Setup Photo a) not in use b) gas flow control setup c) column in use d) setup in use 121 (a) (b) Figure 85: Algal Amino Acid Absorbent Photo a) before absorption b) after absorption 122 (a) (b) Figure 86: Precipitated Solids after Acidification Photos a) solids and liquids b) dried solids 123 REFERENCES 124 REFERENCES (1) Jackson, R. The Effects of Climate Change https://climate.nasa.gov/effects (accessed Oct 10, 2020). (2) US EPA, O. Global Greenhouse Gas Emissions Data https://www.epa.gov/ghgemissions/global-greenhouse-gas-emissions-data (accessed Oct 10, 2020). (3) Mazinani, S.; Ramazani, R.; Samsami, A.; Jahanmiri, A.; Van der Bruggen, B.; Darvishmanesh, S. Equilibrium Solubility, Density, Viscosity and Corrosion Rate of Carbon Dioxide in Potassium Lysinate Solution. Fluid Phase Equilibria 2015, 396, 28–34. https://doi.org/10.1016/j.fluid.2015.03.031. (4) Electricity in the U.S. - U.S. Energy Information Administration (EIA) https://www.eia.gov/energyexplained/electricity/electricity-in-the-us.php (accessed Nov 16, 2020). (5) Greenhouse Gas Inventory Data Explorer | US EPA https://cfpub.epa.gov/ghgdata/inventoryexplorer/#electricitygeneration/entiresector/allgas/catego ry/all (accessed Apr 19, 2021). (6) US EPA, O. Sources of Greenhouse Gas Emissions https://www.epa.gov/ghgemissions/sources-greenhouse-gas-emissions (accessed Oct 10, 2020). (7) Frequently Asked Questions (FAQs) - U.S. Energy Information Administration (EIA) https://www.eia.gov/tools/faqs/faq.php (accessed Oct 10, 2020). (8) Akorede, M. F.; Hizam, H.; Ab Kadir, M. Z. A.; Aris, I.; Buba, S. D. Mitigating the Anthropogenic Global Warming in the Electric Power Industry. Renewable and Sustainable Energy Reviews 2012, 16 (5), 2747–2761. https://doi.org/10.1016/j.rser.2012.02.037. (9) Leung, D. Y. C.; Caramanna, G.; Maroto-Valer, M. M. An Overview of Current Status of Carbon Dioxide Capture and Storage Technologies. Renewable and Sustainable Energy Reviews 2014, 39, 426–443. https://doi.org/10.1016/j.rser.2014.07.093. (10) Kanniche, M.; Gros-Bonnivard, R.; Jaud, P.; Valle-Marcos, J.; Amann, J.-M.; Bouallou, C. Pre-Combustion, Post-Combustion and Oxy-Combustion in Thermal Power Plant for CO2 Capture. Applied Thermal Engineering 2010, 30 (1), 53–62. https://doi.org/10.1016/j.applthermaleng.2009.05.005. 125 (11) Herzog, H. J. Scaling up Carbon Dioxide Capture and Storage: From Megatons to Gigatons. Energy Economics 2011, 33 (4), 597–604. https://doi.org/10.1016/j.eneco.2010.11.004. (12) Wang, M.; Lawal, A.; Stephenson, P.; Sidders, J.; Ramshaw, C. Post-Combustion CO2 Capture with Chemical Absorption: A State-of-the-Art Review. Chemical Engineering Research and Design 2011, 89 (9), 1609–1624. https://doi.org/10.1016/j.cherd.2010.11.005. (13) Bhown, A. S.; Freeman, B. C. Analysis and Status of Post-Combustion Carbon Dioxide Capture Technologies. Environ. Sci. Technol. 2011, 45 (20), 8624–8632. https://doi.org/10.1021/es104291d. (14) Resnik, K. P.; Yeh, J. T.; Pennline, H. W. Aqua Ammonia Process for Simultaneous Removal of CO2, SO2 and NOx. IJETM 2004, 4 (1/2), 89. https://doi.org/10.1504/IJETM.2004.004634. (15) Rinker, E. B.; Ashour, S. S.; Sandall, O. C. Absorption of Carbon Dioxide into Aqueous Blends of Diethanolamine and Methyldiethanolamine. Ind. Eng. Chem. Res. 2000, 39 (11), 4346–4356. https://doi.org/10.1021/ie990850r. (16) Nwaoha, C.; Saiwan, C.; Tontiwachwuthikul, P.; Supap, T.; Rongwong, W.; Idem, R.; AL-Marri, M. J.; Benamor, A. Carbon Dioxide (CO2) Capture: Absorption-Desorption Capabilities of 2-Amino-2-Methyl-1-Propanol (AMP), Piperazine (PZ) and Monoethanolamine (MEA) Tri-Solvent Blends. Journal of Natural Gas Science and Engineering 2016, 33, 742–750. https://doi.org/10.1016/j.jngse.2016.06.002. (17) Petrescu, L.; Bonalumi, D.; Valenti, G.; Cormos, A.-M.; Cormos, C.-C. Life Cycle Assessment for Supercritical Pulverized Coal Power Plants with Post-Combustion Carbon Capture and Storage. Journal of Cleaner Production 2017, 157, 10–21. https://doi.org/10.1016/j.jclepro.2017.03.225. (18) Singh, B.; Strømman, A. H.; Hertwich, E. Life Cycle Assessment of Natural Gas Combined Cycle Power Plant with Post-Combustion Carbon Capture, Transport and Storage. International Journal of Greenhouse Gas Control 2011, 5 (3), 457–466. https://doi.org/10.1016/j.ijggc.2010.03.006. (19) Vaidya, P. D.; Kenig, E. Y. CO2-Alkanolamine Reaction Kinetics: A Review of Recent Studies. Chemical Engineering & Technology 2007, 30 (11), 1467–1474. https://doi.org/10.1002/ceat.200700268. (20) Goff, G. S.; Rochelle, G. T. Monoethanolamine Degradation: O2 Mass Transfer Effects under CO2 Capture Conditions. Ind. Eng. Chem. Res. 2004, 43 (20), 6400–6408. https://doi.org/10.1021/ie0400245. 126 (21) Kang, D.; Park, S.; Jo, H.; Min, J.; Park, J. Solubility of CO2 in Amino-Acid-Based Solutions of (Potassium Sarcosinate), (Potassium Alaninate + Piperazine), and (Potassium Serinate + Piperazine). J. Chem. Eng. Data 2013, 58 (6), 1787–1791. https://doi.org/10.1021/je4001813. (22) Wang, T.; Hovland, J.; Jens, K. J. Amine Reclaiming Technologies in Post-Combustion Carbon Dioxide Capture. Journal of Environmental Sciences 2015, 27, 276–289. https://doi.org/10.1016/j.jes.2014.06.037. (23) Hamborg, E. S.; Niederer, J. P. M.; Versteeg, G. F. Dissociation Constants and Thermodynamic Properties of Amino Acids Used in CO2 Absorption from (293 to 353) K. J. Chem. Eng. Data 2007, 52 (6), 2491–2502. https://doi.org/10.1021/je700275v. (24) Aronu, U. E.; Svendsen, H. F.; Hoff, K. A. Investigation of Amine Amino Acid Salts for Carbon Dioxide Absorption. International Journal of Greenhouse Gas Control 2010, 4 (5), 771– 775. https://doi.org/10.1016/j.ijggc.2010.04.003. (25) Hook, R. J. An Investigation of Some Sterically Hindered Amines as Potential Carbon Dioxide Scrubbing Compounds. Ind. Eng. Chem. Res. 1997, 36 (5), 1779–1790. https://doi.org/10.1021/ie9605589. (26) Hu, G.; Smith, K. H.; Wu, Y.; Mumford, K. A.; Kentish, S. E.; Stevens, G. W. Carbon Dioxide Capture by Solvent Absorption Using Amino Acids: A Review. Chinese Journal of Chemical Engineering 2018, 26 (11), 2229–2237. https://doi.org/10.1016/j.cjche.2018.08.003. (27) Song, H.-J.; Park, S.; Kim, H.; Gaur, A.; Park, J.-W.; Lee, S.-J. Carbon Dioxide Absorption Characteristics of Aqueous Amino Acid Salt Solutions. International Journal of Greenhouse Gas Control 2012, 11, 64–72. https://doi.org/10.1016/j.ijggc.2012.07.019. (28) Hussain, J.; Wang, X.; Sousa, L.; Ali, R.; Rittmann, B. E.; Liao, W. Using Non-Metric Multi-Dimensional Scaling Analysis and Multi-Objective Optimization to Evaluate Green Algae for Production of Proteins, Carbohydrates, Lipids, and Simultaneously Fix Carbon Dioxide. Biomass and Bioenergy 2020, 141, 105711. https://doi.org/10.1016/j.biombioe.2020.105711. (29) Costanzo, W.; Jena, U.; Hilten, R.; Das, K. C.; Kastner, J. R. Low Temperature Hydrothermal Pretreatment of Algae to Reduce Nitrogen Heteroatoms and Generate Nutrient Recycle Streams. Algal Research 2015, 12, 377–387. https://doi.org/10.1016/j.algal.2015.09.019. (30) Thomas, D. M.; Mechery, J.; Paulose, S. V. Carbon Dioxide Capture Strategies from Flue Gas Using Microalgae: A Review. Environ Sci Pollut Res 2016, 23 (17), 16926–16940. https://doi.org/10.1007/s11356-016-7158-3. 127 (31) Ruangsomboon, S.; Prachom, N.; Sornchai, P. Enhanced Growth and Hydrocarbon Production of Botryococcus Braunii KMITL 2 by Optimum Carbon Dioxide Concentration and Concentration-Dependent Effects on Its Biochemical Composition and Biodiesel Properties. Bioresource Technology 2017, 244, 1358–1366. https://doi.org/10.1016/j.biortech.2017.06.042. (32) Yen, H.-W.; Ho, S.-H.; Chen, C.-Y.; Chang, J.-S. CO2, NOx and SOx Removal from Flue Gas via Microalgae Cultivation: A Critical Review. Biotechnology Journal 2015, 10 (6), 829–839. https://doi.org/10.1002/biot.201400707. (33) Lam, M. K.; Lee, K. T.; Mohamed, A. R. Current Status and Challenges on Microalgae- Based Carbon Capture. International Journal of Greenhouse Gas Control 2012, 10, 456–469. https://doi.org/10.1016/j.ijggc.2012.07.010. (34) Song, C.; Qiu, Y.; Xie, M.; Liu, J.; Liu, Q.; Li, S.; Sun, L.; Wang, K.; Kansha, Y. Novel Regeneration and Utilization Concept Using Rich Chemical Absorption Solvent As a Carbon Source for Microalgae Biomass Production. Ind. Eng. Chem. Res. 2019, 58 (27), 11720–11727. https://doi.org/10.1021/acs.iecr.9b02134. (35) Hsueh, H. T.; Chu, H.; Yu, S. T. A Batch Study on the Bio-Fixation of Carbon Dioxide in the Absorbed Solution from a Chemical Wet Scrubber by Hot Spring and Marine Algae. Chemosphere 2007, 66 (5), 878–886. https://doi.org/10.1016/j.chemosphere.2006.06.022. (36) Vega, F.; Baena-Moreno, F. M.; Gallego Fernández, L. M.; Portillo, E.; Navarrete, B.; Zhang, Z. Current Status of CO2 Chemical Absorption Research Applied to CCS: Towards Full Deployment at Industrial Scale. Applied Energy 2020, 260, 114313. https://doi.org/10.1016/j.apenergy.2019.114313. (37) Hitachi’s Carbon Dioxide Scrubbing Technology with New Absorbent for Coal Fired Power Plants. Energy Procedia 2011, 4, 245–252. https://doi.org/10.1016/j.egypro.2011.01.048. (38) Yeh, J. T.; Pennline, H. W.; Resnik, K. P. Study of CO2 Absorption and Desorption in a Packed Column. Energy Fuels 2001, 15 (2), 274–278. https://doi.org/10.1021/ef0002389. (39) Notz, R.; Asprion, N.; Clausen, I.; Hasse, H. Selection and Pilot Plant Tests of New Absorbents for Post-Combustion Carbon Dioxide Capture. Chemical Engineering Research and Design 2007, 85 (4), 510–515. https://doi.org/10.1205/cherd06085. (40) Final Report Summary - CASTOR (CO2, from Capture to Storage) | Report Summary | CASTOR | FP6 | CORDIS | European Commission https://cordis.europa.eu/project/id/502586/reporting (accessed Nov 2, 2020). 128 (41) Wang, M.; Joel, A. S.; Ramshaw, C.; Eimer, D.; Musa, N. M. Process Intensification for Post-Combustion CO2 Capture with Chemical Absorption: A Critical Review. Applied Energy 2015, 158, 275–291. https://doi.org/10.1016/j.apenergy.2015.08.083. (42) Baena-Moreno, F. M.; Rodríguez-Galán, M.; Vega, F.; Reina, T. R.; Vilches, L. F.; Navarrete, B. Synergizing Carbon Capture Storage and Utilization in a Biogas Upgrading Lab- Scale Plant Based on Calcium Chloride: Influence of Precipitation Parameters. Science of The Total Environment 2019, 670, 59–66. https://doi.org/10.1016/j.scitotenv.2019.03.204. (43) Aroonwilas, A.; Veawab, A. Characterization and Comparison of the CO2 Absorption Performance into Single and Blended Alkanolamines in a Packed Column. Ind. Eng. Chem. Res. 2004, 43 (9), 2228–2237. https://doi.org/10.1021/ie0306067. (44) Daiek, C.; Link to external site, this link will open in a new window. Effects of Water Recirculation on Pilot-Scale Microalgae Cultivation Using Flue Gas CO2. M.S., Michigan State University, United States -- Michigan, 2020. (45) JONES, D. B. Factors for Converting Percentages of Nitrogen in Foods and Feeds into Percentages of Protein. USDA Circ. 1931, 183, 1–21. (46) Cutshaw, A.; Daiek, C.; Zheng, Y.; Frost, H.; Marks, A.; Clements, D.; Uludag-Demirer, S.; Verhanovitz, N.; Pavlik, D.; Clary, W.; Liu, Y.; Liao, W. A Long-Term Pilot-Scale Algal Cultivation on Power Plant Flue Gas – Cultivation Stability and Biomass Accumulation. Algal Research 2020, 52, 102115. https://doi.org/10.1016/j.algal.2020.102115. (47) The Amino-Acid and Sugar Composition of 16 Species of Microalgae Used in Mariculture. Journal of Experimental Marine Biology and Ecology 1991, 145 (1), 79–99. https://doi.org/10.1016/0022-0981(91)90007-J. (48) Growth and Ω3 Fatty Acid and Amino Acid Composition of Microalgae under Different Temperature Regimes. Aquaculture 1989, 77 (4), 337–351. https://doi.org/10.1016/0044- 8486(89)90218-4. (49) Brown, M. R.; Jeffrey, S. W.; Volkman, J. K.; Dunstan, G. A. Nutritional Properties of Microalgae for Mariculture. Aquaculture 1997, 151 (1–4), 315–331. https://doi.org/10.1016/S0044-8486(96)01501-3. (50) Tran, H.; Vakkilainen, E. THE KRAFT CHEMCIAL RECOVERY PROCESS. 2016. (51) Jackson, P.; Robinson, K.; Puxty, G.; Attalla, M. In Situ Fourier Transform-Infrared (FT- IR) Analysis of Carbon Dioxide Absorption and Desorption in Amine Solutions. Energy Procedia 2009, 1 (1), 985–994. https://doi.org/10.1016/j.egypro.2009.01.131. 129 (52) Robinson, K.; McCluskey, A.; Attalla, M. I. An FTIR Spectroscopic Study on the Effect of Molecular Structural Variations on the CO2 Absorption Characteristics of Heterocyclic Amines. ChemPhysChem 2011, 12 (6), 1088–1099. https://doi.org/10.1002/cphc.201001056. (53) Diab, F.; Provost, E.; Laloué, N.; Alix, P.; Souchon, V.; Delpoux, O.; Fürst, W. Quantitative Analysis of the Liquid Phase by FT-IR Spectroscopy in the System CO2/Diethanolamine (DEA)/H2O. Fluid Phase Equilibria 2012, 325, 90–99. https://doi.org/10.1016/j.fluid.2012.04.016. (54) Qaroush, A. K.; Alshamaly, H. S.; Alazzeh, S. S.; Abeskhron, R. H.; Assaf, K. I.; Eftaiha, A. F. Inedible Saccharides: A Platform for CO2 Capturing. Chem Sci 2018, 9 (5), 1088–1100. https://doi.org/10.1039/c7sc04706a. (55) Seifried, B.; Temelli, F. Density of Marine Lipids in Equilibrium with Carbon Dioxide. The Journal of Supercritical Fluids 2009, 50 (2), 97–104. https://doi.org/10.1016/j.supflu.2009.05.011. (56) Aronu, U. E.; Hoff, K. A.; Svendsen, H. F. CO2 Capture Solvent Selection by Combined Absorption–Desorption Analysis. Chemical Engineering Research and Design 2011, 89 (8), 1197–1203. https://doi.org/10.1016/j.cherd.2011.01.007. (57) Portugal, A. F.; Sousa, J. M.; Magalhães, F. D.; Mendes, A. Solubility of Carbon Dioxide in Aqueous Solutions of Amino Acid Salts. Chemical Engineering Science 2009, 64 (9), 1993– 2002. https://doi.org/10.1016/j.ces.2009.01.036. 130