Biocatalytic Carbon Capture and Conversion

Home , Carbonic anhydrase, Formate dehydrogenase

BIOCATALYTIC CARBON CAPTURE AND CONVERSION

A DISSERTATION

SUBMITTED TO THE FACULTY OF THE GRADUATE SCHOOL

OF THE UNIVERSITY OF MINNNESOTA

SHUNXIANG XIA

IN PARTIAL FULFILLMENT OF THE REQUIREMENTS

FOR THE DEGREE OF

DOCTOR OF PHILOSOPHY

PING WANG, ADVISOR

MAY 2018

Acknowledgements

First and fore most I want to thank my advisor, Dr. Ping Wang in the Department of

Bioproducts and Biosystems Engineering at the University of Minnesota. It has been an honor to be his Ph.D. student. I appreciate all his contributions of time, ideas, and patience to make my Ph.D. experience productive and stimulating.

I am grateful to all my Dissertation Committee: Prof. Bo Hu, Prof. Kechun Zhang and

Prof. William Tai Yin Tze. Each of the members has provided me extensive personal and professional guidance and taught me a great deal about both scientific research and life in general.

The lab group has been a source of friendships as well as good advice and collaboration. I especially thank to my group members: Dr. Xueyan Zhao, Dr. Bo Tang,

Dr. Guoying Dai, Mr. Ben Frigo-Vaz and Mr. Zhen Qiao. I would like to acknowledge honorary Prof. Jungbae Kim group in Korean. We worked together on carbon capture and release experiments, and I very much appreciated his enthusiasm, intensity, willingness to share his knowledge.

Lastly, I would like to thank my family for all their love and encouragement. For my parents who raised me concerning science and supported me in all my pursuits. And most of all for my loving, supportive, and patient wife Fang Xu whose dedicated support during the final stages of this Ph.D. is so appreciated. Thank you.

Abstract

As the most significant anthropogenic greenhouse gas, carbon dioxide emitted as an industrial pollution waste can be probably most effectively handled by the Carbon

Capture and Storage (CCS) strategy. The energy efficiency of currently available CCS technologies has been, however, quite deterring for large scale practice of CCS.

Searching energy efficient CCS strategies via biological means, this research examines factors limiting reaction equilibrium conversion and reaction kinetics for biocatalytic carbon conversion reactions. It is assumed that, biocatalytic carbon conversion can be realized at energy cost close to theoretical efficiency, thanks to the unique energy transfer features of biological reactions; however, equilibrium thermodynamics, as well as kinetics, have to be improved to realize intensified reactions to match the scale and rate of industrial carbon emission.

Among the numerous CO2-conversion and CO2-elution reaction pathways found in biology, the isocitrate dehydrogenase (ICDH) reaction in reductive tricarboxylic acid

(RTCA) cycle was taken as our study model system due to its reversibility and again high reaction intensity. The overall objective of this research is to establish an ICDH based carbon capture and release strategy with high energy efficiency and reaction intensity, improved equilibrium conversion yield and enhanced catalyst stability.

For reaction equilibrium, the pH sensitivity of ICDH was first examined by studying the effect of system pH on reaction equilibrium both theoretically and experimentally.

As the theoretical prediction was relying on the Debye–Hückel theory, experimentally

determined reaction equilibrium was obtained by finding the equilibrium point with zero net reaction rate. The results showed that ICDH reaction exhibited a high pH sensitivity and pH was proven as an effective factor to manipulation the reaction direction. With the pH changed between 5 and 9, the reaction equilibrium constant can be shifted by a factor of ~500 fold.

To enhance reaction kinetics, screening ideal enzyme and optimizing reaction condition were conducted. ICDH from Chlorobium limicola revealed the highest specific activity: 15.3 U/mg at pH 6 for carboxylation and 90.2 U/mg at pH 9 for decarboxylation. The optimum temperature and ionic strength for ICDH was 45oC and

200 mM respectively. With higher reaction rate, the gas phase carbon dioxide instead of bicarbonate was considered as the primary substrate for carboxylation.

For carbon capture capacity, additional thermodynamic driving force can be created via designing of cascade reaction scheme. In this study, aconitase was introduced and isocitrate was further converted to citrate, lowing the overall Gibbs free energy of reaction. The cascade reactions had the potential to capture 0.48 mol of carbon for each mole of the substrate, approaching the intensities realized with chemical absorbents such as MEA.

Enzyme immobilization was conducted to improve system stability. During the process,

20 ~ 30 mg enzyme can be absorbed on 1 gram of mesoporous silicon foam (MSF) within 5 mins with the specific activity of decarboxylation in the order of 6 U per gram

MSF. After immobilization, enzyme stability against harsh environmental factors such as high temperature, extreme pH and high shear stress was improved significantly. iii

With economic analysis, the ICDH strategy can bring a 66% improvement in energy efficiency. The cost and stability of cofactor (NADPH) were the most significant barriers to the large-scale application of ICDH, mainly due to the poor stability of cofactor under the acid condition.

This work demonstrates the feasibility of a novel carbon capture strategy applying reversible enzyme reactions. It reveals that reaction equilibrium can be effectively by manipulating pH and reaction pathways. The reaction intensity and stability could also be improved through reaction condition optimization and enzyme immobilization.

With the discovery of new source enzymes and molecular modification of existing enzymes, the intensity and stable of biocatalytic CCS are expected to be proficient in comparison to traditional chemical carbon capture strategies, benefiting from the overall energy efficiency of biological reactions.

Table of Contents

Acknowledgements ...... i

Abstract ...... ii

Table of Contents ...... v

List of Tables ...... x

List of Figures ...... xii

Chapter 1. General Introduction ...... 1

1.1 Carbon dioxide emission and environment concerns ...... 1 1.2 Carbon capture and storage ...... 4 1.3 Biological CCS ...... 9 1.4 Biochemical CCS ...... 10 1.5 Objectives ...... 12

Chapter 2. Literature Review ...... 14

2.1 Biocatalysts based carbon capture...... 14 2.2 pH sensitivity of biocatalytic carbon capture ...... 20 2.3 Reaction intensity of biocatalytic CCS ...... 21 2.4 Reaction stability of biocatalytic CCS ...... 26 2.5 Economy feasibility of CCS ...... 32

Chapter 3. pH Sensitivity of Biocatalytic Carboxylation Reaction ...... 37

3.1 Introduction ...... 37 3.2 Materials and methods ...... 38 3.2.1 Materials ...... 38 3.2.2 Experimentally determination of Gibbs free energy of reaction ...... 38 3.2.3 Theoretical calculation of Gibbs free energy of reaction ...... 39 3.3 Results and discussion ...... 39 3.3.1 Experimental determination of equilibrium constant ...... 39 v

3.3.2 The effect of pH on equilibrium constant ...... 42 3.3.3 Debye–Hückel model of ICDH reaction ...... 43 3.3.4 Prediction of carbon dioxide fixation with ICDH ...... 47 3.3.5 pH sensitivity of other biochemical reactions for carbon fixing ...... 48 3.4 Conclusions ...... 50

Chapter 4. Characteristics of ICDH for Carbon Capture ...... 51

4.1 Introduction ...... 51 4.2 Material and methods ...... 52 4.2.1 Matieral ...... 52 4.2.2 Assay of total protein concentration ...... 52 4.2.3 Assay activity of ICDH ...... 53 4.2.4 Cell culture ...... 54 4.2.5 ICDH purification ...... 54 4.2.6 Determination of Michaelis constants of ICDH reaction ...... 55 4.2.7 Assay activity of CA ...... 56 4.2.8 Determination of selectivity of carbon substrate ...... 57 4.2.9 Carbon dioxide capture with bubbling gas mixture ...... 58 4.2.10 Carbon dioxide capture with surface covered with mixture gas ...... 59 4.3 Results and discussion ...... 60 4.3.1 ICDH expression and purification ...... 60 4.3.2 pH profile of different ICDHs ...... 65 4.3.3 Parameters effect the ICDH acitivties ...... 69 4.3.4 Optimum substrate for carboxylation reaction ...... 79 4.3.5 Gas phase carbon dioxide capture ...... 87 4.4 Conclusions ...... 90

Chapter 5. Carbon Dioxide Conversion with 2,3-dihydroxybenoic Acid Decarboxylase ...... 91

5.1 Introduction ...... 91

5.2 Material and methods ...... 92 5.2.1 Materials ...... 92 5.2.2 Assay of total protein concentration ...... 92 5.2.3 Expression and purification of the enzyme ...... 92 5.2.4 Enzyme activity assay ...... 93 5.2.5 Carbon capture experiment ...... 94 5.3 Results and discussion ...... 94 5.3.1 Purification of 2,3-DHBA ...... 94 5.3.2 The effect of pH on enzyme activity ...... 95 5.3.3 The effect of temperature on enzyme carboxylation activity ...... 97 5.3.4 Carbon capture with 2,3-DHBA ...... 98 5.4 Conclusions ...... 101

Chapter 6. Cascade Enzymatic Reaction System to Improve Carbon Capture Capacity ...... 103

6.1 Introduction ...... 103 6.2 Material and methods ...... 104 6.2.1 Matieral ...... 104 6.2.2 Assay activity of aconitase ...... 104 6.2.3 Determination of Michaelis constants of ICDH reaction ...... 105 6.2.4 Carbon capture capacity at different pH ...... 105 6.3 Results and discussion ...... 105 6.3.1 Thermodynamic prediction of Gibbs free energy of reaction ...... 105 6.3.2 Carbon capture with multiple enzyme system ...... 107 6.3.3 The effect of pH on aconitase activity ...... 109 6.3.4 Michaelis constant of aconitase ...... 110 6.3.5 Parameter optimization for aconitase ...... 111 6.3.6 Carbon captures and release cycle with cascade enzymes ...... 118 6.4 Conclusions ...... 119

vii

Chapter 7. Immobilization to improve the stability of ICDH ...... 121

7.1 Introduction ...... 121 7.2 Material and methods ...... 123 7.2.1 Material ...... 123 7.2.2 Enzyme immobilization with MSF adsorption ...... 124 7.2.3 Enzyme immobilization with chitosan entrapment ...... 124 7.2.4 Glutaraldehyde treatment of immobilized enzyme ...... 125 7.2.5 Activity assay of immobilized enzyme ...... 125 7.2.6 Detection of leakage behavior ...... 126 7.2.7 Enzyme stability tests ...... 126 7.2.8 Determination of michaelis constants and turnover number ...... 126 7.3 Results and discussion ...... 127 7.3.1 Optimization of immobilization conditions for MSF adsorption ...... 127 7.3.2 Optimization of immobilization condition for chitosan entrapment ...... 130 7.3.3 Immobilized enzyme characterization ...... 133 7.3.4 Leakage of adsorbed enzyme ...... 134 7.3.5 Stability of immobilized enzyme ...... 137 7.3.6 The effect of CA on activity of immobilized ICDH ...... 143 7.4 Conclusions ...... 145

Chapter 8. Economic analysis of ICDH based CCS ...... 146

8.1 Introduction ...... 146 8.2 Material and methods ...... 147 8.2.1 Materials ...... 148 8.2.2 Cofactor stability test ...... 148 8.2.3 Activity assay of formate dehydrogenase ...... 148 8.2.4 Cofactor regeneration ...... 149 8.2.5 Economy analysis ...... 149 8.3 Results and discussion ...... 151

viii

8.3.1 Cofactor stability ...... 151 8.3.2 Cofactor regeneration ...... 155 8.3.3 Economy analysis ...... 158

Chapter 9. Conclusions and Future work ...... 171

9.1 Conclusions ...... 171 9.2 Path ahead ...... 173

References ...... 175

List of Tables

Table 1- 1. Summary of different CCS technologies ...... 6 Table 2- 1. Summary of specific activity of carboxylation enzymes ...... 23 Table 2- 2. Summary of enzyme denature mechanism in CCS ...... 27 Table 2- 3. Summary with different enzyme immobilization methods ...... 27 Table 2- 4. Levelized avoided costs of electricity for different resource in expected in 2022...... 35 Table 3- 1. Equilibrium concentrations and constants determined experimentally for different pH...... 42

Table 3- 2. Gibbs free energy of formation of chemical species (ΔGf) at different pH calculated via Debye–Hückel theory...... 44 Table 3- 3. The dependence of Gibbs free energy on pH ...... 46 Table 3- 4. Carbon capture capacity of the ICDH reaction under different pressure. . 48 Table 3- 5. pH sensitivity of biochemical reactions involved in carbon fixation ...... 49 Table 4- 1. Michaelis constants of substrate for the ICDHs...... 68 Table 4- 2. The effect of cation on ICDH activity ...... 75 Table 4- 3. Comparison of flue gas quality of various fuels ...... 77 Table 4- 4. The effect of cation on ICDH activity ...... 78 Table 6- 1. The effect of cation on enzyme activity ...... 117 Table 6- 2. The effect of sulfate and nitrate on ICDH activity ...... 118 Table 7- 1. The effect of enzyme concentration on protein yield and activity yield. 128 Table 7- 2. Free and immobilized enzyme characterization...... 134 Table 8- 1. Degradation reaction constant and halftime of NADPH under various conditions ...... 154

Table 8- 2. Equilibrium constant (Km) of reaction catalyzed by formate dehydrogenase...... 156 Table 8- 3. Capital expense for both MEA and ICDH based CCS plant...... 159 Table 8- 4. Operation expense for both MEA and ICDH based CCS plant...... 162

Table 8- 5. The total annual cost of CCS for both MEA and ICDH case ...... 166

List of Figures

Figure 1- 1. Global carbon dioxide emission from 1990-2014...... 2 Figure 1- 2. Global carbon dioxide emission by economic sector...... 3 Figure 1- 3. Global electricity generation by different resources from 2012-2040...... 4 Figure 1- 4. Illusion of carbon capture and storage...... 5 Figure 2- 1. The scheme of RTCA cycle...... 16 Figure 3- 1. Determination of equilibrium point for reaction catalyzed by ICDH at pH 5...... 41 Figure 3- 2. Comparison between experimental determination and theoretical prediction for Gibbs free energy change for carbon conversion reaction...... 45 Figure 4- 1. Lineweaver-Burk plot ...... 56 Figure 4- 2. Scheme of carbon dioxide capture with bubbling gas mixture...... 59 Figure 4- 3. Scheme of carbon dioxide capture with surface covered with mixture gas...... 60 Figure 4- 4. Growth curve of recombined E. coli under 37 °C and 28 °C...... 61 Figure 4- 5. Cells dry weight (left) and total enzyme activity (right) with various induce temperature...... 62 Figure 4- 6. The pH profiles of ICDHs ...... 66

2+ 2+ Figure 4- 7. Km value of Mg (A) and Mn (B) determined with Lineweaver–Burk plot...... 71 Figure 4- 8. The effect of temperature on enzyme specific activity...... 72 Figure 4- 9. The effect of ionic strength on enzyme specific activity...... 74 Figure 4- 10. The progress of carboxylation reaction catalyzed by ICDH with different carbon resource ...... 84 Figure 4- 11. The progress of carboxylation reaction catalyzed by ICDH under pH 6.5...... 85 Figure 4- 12. Carbon dioxide capture rate with different concentration of carbon dioxide...... 87

xii

Figure 4- 13. Carbon dioxide capture rate with different surface area covered with carbon dioxide ...... 89 Figure 5- 1. The effect of pH on the specific activity of 2,3-DHBA ...... 96 Figure 5- 2. the effect of temperature on carboxylation activity of 2,3-DHBA...... 98 Figure 5- 3. The effect of pH on the conversion rate of 2,3-DHBA reaction...... 99 Figure 5- 4. The effect of temperature on the conversion rate of 2,3-DHBA reaction...... 100 Figure 5- 5. The conversion rate of phenol with different substrate concentration. .. 101 Figure 6- 1. The gap of Gibbs free energy of formation between substrate and product for reaction catalyzed by ICDH and aconitase...... 107 Figure 6- 2. The improvement of carbon captures capacity of ICDH system with aconitase under pH 6-9...... 108 Figure 6- 3. The scheme of conversion between citric and isocitric catalyzed by aconitase...... 109 Figure 6- 4. The specific activity of aconitase under different pH...... 110 Figure 6- 5. The Lineweaver–Burk plot of conversion reaction catalyzed by aconitase...... 111 Figure 6- 6. The improvement of specific activity of aconitase with different cofactors ...... 112 Figure 6- 7. The effect of temperature on aconitase activity ...... 114 Figure 6- 8. The effect of ionic strength on aconitase activity ...... 115 Figure 6- 9. Carbon capture and regeneration cycle with (circle) and without (square) aconitase, with pH switched between 7 and 8 for carboxylation and decarboxylation reactions, respectively...... 119 Figure 7- 1. Molecule structure of chitin and chitosan ...... 122 Figure 7- 2. Scheme of ICDH immobilization with chitosan ...... 124 Figure 7- 3. The amount of bound enzyme with MSF with incubation time...... 127 Figure 7- 4. Enzyme loading and specific activity for immobilization at different pH ...... 130

xiii

Figure 7- 5. The enzyme activity yield with different chitosan concentration ...... 131 Figure 7- 6. Molecule structure of (a) chitosan and tripolyphosphate and (b) chitosan-TPP crosslinking...... 133 Figure 7- 7. Enzyme activity and retain concentration with multiple reaction cycles...... 135 Figure 7- 8. a) Free enzyme adsorbed on MSF, b) CLEAs adsorbed on the MSF .... 136 Figure 7- 9. The retain concentration of enzyme with multiple reaction cycles with cross-linking treatment...... 137 Figure 7- 10. Effect of immobilization on enzyme stability against high temperature...... 139 Figure 7- 11. Effect of immobilization on enzyme stability against pH...... 141 Figure 7- 12. Effect of immobilization on enzyme stability against shear force...... 143 Figure 7- 13. The reaction rate of carboxylation with different CA concentration ... 144 Figure 8- 1. Model of NADH denature under acidic condition ...... 152 Figure 8- 2. NADPH stability at different pH and buffer category ...... 153 Figure 8- 3. Cofactor regeneration rate at different pH catalyzed by formate dehydrogenase...... 156 Figure 8- 4. The cost of CO2 avoided for MEA and ICDH cases ...... 166 Figure 8- 5. Electricity price from different resources ...... 168

xiv

Chapter 1. General Introduction

1.1 Carbon dioxide emission and environment concerns

Before the industrial revolution, carbon dioxide emitted from catabolism of living things was fixed in a balance with the photosynthesis of plants. From primitive hydroxypropionate/4-hydroxybutyrate cycle [1] to advanced Calvin cycle [2], carbon dioxide was utilized as a substrate to synthesize hydrocarbon compound in chloroplasts during the growing of plants. However, the transition of going from hand production methods to machines dramatically increased the consumption level of fossil fuels. Since the beginning of the Industrial Revolution, the annual emission of carbon dioxide has increased more than 10 times (Figure 1-1), corresponding a 40 percent increase of atmospheric CO2 levels [3], which was from about 280 parts per million (ppm) in the 1900s to 400 ppm today [4]. By effecting the exchange of incoming and outgoing radiation, the increased level of carbon dioxide causes the greenhouse phenomena, resulting in an elevation of the average surface temperature on the earth. As the consumption of fossil fuels over the past few decades has been kept at unprecedented high levels, there is an urgent need for reducing total CO2 emission into the atmosphere to embrace environmental and climate impacts of carbon emissions.

12000

10000

8000

6000

4000

2000

Carbon emisson ton/ year) M ( emisson Carbon 1900 1920 1940 1960 1980 2000 2020 Year

Figure 1- 1. Global carbon dioxide emission from 1990-2014. (Data from United States Environmental Protection Agency)

At the current stage, over 80 percent of energy demand for the world is supplied by fossil fuels, which also accounts more than 40 percent of CO2 emissions [5]. The carbon dioxide emissions by economic sectors in 2014 was summarized in Figure 1-2.

According to the date of the figure, electricity, heat production and industry sectors together contribute the half of total carbon emission. Different from the small and sporadic flows of emissions occurs in the transportation sector, the emissions from heat production and industry sections such as coal fire plant and ironworks possess much high flow rates. For example, a 400 MWe coal fire plant typically emits 375 tonne carbon dioxide in the form of flue gas per hour with the concentration around

30 percent. Regarding the high flow rate and concentration, the flue gas of coal fire plant and ironworks can be the primary resources for most carbon capture processes.

6 10 Electricity and heat 25 Agriculture and land use

14 Industry

Transpotation

Other energy 24 21 Building

Figure 1- 2. Global carbon dioxide emission by economic sector. (Date from United States Environmental Protection Agency).

Many efforts have been made to reduce the already high carbon dioxide concentration level. Since one-third of total carbon emission happens during electricity generation, shifting the electricity resource from fossil fuel to sources such as nuclear and renewable energy (solar, wind, hydro and biomass) can significantly reduce the carbon emission (Figure 1-3). According to the data of international energy information administration, the total electricity generation from nuclear and renewables has increased to 9.24 trillion kWh at 2017 from 7.07 trillion kWh at 2012, and will further increase to 15.13 trillion kWh by the end of 2040. However, the requirement of fossil fuel for electricity is not expected to decrease correspondingly.

By the end to 2040, coal and natural gas will still be the primary source of power generation (~ 60%). And again, the development of renewable energy is heavily limited by the local geographical environment and potential safety. Since the utilizing of fossil flue with a high level is unavoidable in foreseeable future, even with the

repaid development of renewable technologies, the environmental problems related to surplus emitted carbon dioxide will be more severe in the next decades.

40.00

30.00

20.00

10.00 ( trillion kwh) trillion (

Electricity generation Electricity 0.00 2012 2020 2025 2030 2035 2040 Year Petroleum Nuclear Natural gas Coal Renewables

Figure 1- 3. Global electricity generation by different resources from 2012-2040 (Data from International energy outlook 2017).

1.2 Carbon capture and storage

Carbon capture and storage (CCS) is proposed as a promising approach to reducing carbon dioxide emissions and both natural and artificial process of CCS are shown in

Figure 1-4. Plants are the main participators in the natural CCS process: through photosynthesis, carbon dioxide in the atmosphere is utilized as the substrate to synthesize hydrocarbons with carbon fixation [6]. Before the Industry Revolution, the emission rate of carbon dioxide from dissimilation process was in an equilibrium state with the carbon fixation rate of plants. The consumption of fossil fuel in recent decades dramatically increase the carbon emission rate, breaking the carbon cycle

equilibrium and bringing the needs for artificial CCS process. In artifical process,

CO2 from its gaseous mixtures like flue gas of power plants is separated, so that it can be subsequently released as a pure gas, liquefied and injected into underground storage sites or converted to valuable products [7].

Figure 1- 4. Illusion of carbon capture and storage.

Two common routes are proposed for carbon capture and sequestration, namely physicochemical routes and biological routes [8]. Physicochemical CCS is realized through carbon capture and release strategy and is more popular due to its high reaction extent and intensity, while biological routes can be conducted in mild condition and possess higher energy efficiency (Table 1-1).

Table 1- 1. Summary of different CCS technologies

Advantage Disadvantage

Biological routes Feed algae and Low capital expense; Low conversion and

enzyme Simple operation; capture rate;

conversion Environment friendly; Hard to control

High energy

efficiency

Physicochemical routes Membrane Low cost Poor stability;

separation Not suitable for low

concentration

Cryogenic High product purity Energy intensive;

distillation High cost

Adsorption High capture rate Low selectivity

Absorption High capture rate; Low energy efficient;

Compared low cost Mass transfer limited

The physicochemical routes involve the application of membrane separation, cryogenic distillation, adsorption, and absorption. During membrane separation process, carbon dioxide is separated from other components of flue gas (mainly nitrogen and oxygen). Permeability and selectivity are the two critical parameters to evaluate the performance of the membrane. While permeability determines the flux of

gas can pass through the membrane, selectivity is defined as the ratio of permeability between target component and other components. The ideal membranes selected for carbon dioxide separation normally possess both high permeability and selectivity.

According to the references, the permeability and selectivity of advanced polyvinyl amine membrane for CCS propose can be as high as 1000 GPU (gas permeability unit) and 200 respectively [9]. Nowadays, polymeric membranes are preferred in the market for their reliability and also low cost, while membranes are also fabricated with other materials such as carbon, alumina, silica and zeolite for special requirements of stability and permeability. Under high carbon dioxide concentration scenario in mild condition, membrane separation realizes the lowest cost, compared with other methods [10]. However, the high temperature of flue gas will destabilize most membranes and the permeability of the membrane will decrease sharply with the exhausting of carbon dioxide in flue gas during the fixation process.

For cryogenic distillation, carbon dioxide will be separated from the gas mixture by cooling it until it liquefies. During the process, energy is also wasted to cool down other components of flue gas. Although this energy-intensive technology is widely used for preparing pure air component in a lab, the high energy cost impedes its further application in the larger scale of carbon dioxide separation.

The large-scale CCS processes have been realized using chemical absorbents to capture flue gas carbon dioxide. Chemical absorption is the process that the carbon dioxide is selectively absorbed into the liquid phase and in practice, flue gases react

with the solvent absorption column at 50°C, and nearly 90 vol % of CO2 is absorbed.

The CO2-rich stream is fed to a regenerator at high temperature (100–120°C). In the regenerator, the solvent can be regenerated and the CO2 is released as pure form.

Different absorbents including Amines such as monoethanolamine (MEA), diethanolamine (DEA) and methyldiethanolamine (MDEA) are the most common agents for industrial processes [11]. In this process, carbon dioxide capture and release cycle is obtained with the absorbents regeneration by temperature shifting [12], resulting large portions of the energy are wasted during the heating–cooling cycles.

Adsorption is another proposed approach of physicochemical router, in which the molecules of carbon dioxide are adsorbed on the surface of the solid adsorbent with either physical forces or chemical reactions. Like the absorption process, adsorption and desorption are realized with by temperature shifting (60-200°C). Carbonaceous materials, mesoporous zeolites and modified polymeric materials are all reported as adsorbents with a ~10 mmol/g carbon loading level [13].

Nowadays, matured industrial CCS processes are purely based on physicochemical methods, especially through absorption. The CCS technology called PostCap has been developed by Abu Dhabi Future Energy Company recently with the selective absorption of CO2 from flue gas using amino acid salt solvent and subsequent desorption to gain high purity CO2. The project was successfully validated with more than 9000 operational hours [14]. The most obvious disadvantage of chemical CCS is its poor energy efficiency. As the regeneration of absorbent results in a huge

temperature shifting (~150 °C), most energy actually was wasted in heating the water portion of absorbent (~ 70 w/w %). The mainly waste of energy can be avoided in the biochemical based CCS realized with pH shifting. [15].

1.3 Biological CCS

Different from chemical processes, biological approaches can be operated at ambient conditions with high reaction specificity and unparalleled high energy efficiency.

Feeding energy crop with carbon dioxide can be the fundamental way of biological

CCS. The energy crops include willow, elephant grass, switchgrass, which are mostly lignocellulosic in nature. For example, timothy grass contains 43.4 wt % carbon, 6.1 wt % hydrogen, 1.3 wt % nitrogen, 0.1 wt % sulfur, and 45.4 wt % oxygen [16]. After pyrolysis of the biomass of energy crops, a part of carbon is fixed as a form of bio-oil and biochar which contains 49.2 wt % and 63.7 wt % carbon respectively. Bio-oil could be further refined as fuel and biochar could be mixed with soil as a long-term carbon sink. Growing energy crop, the simplest method, however, has some shortages: compared with food crops, the energy crops are not economically attractive in most areas of the world; and again, gathering the biomass of energy crops is energy- consuming as the yield per unit area is low. Considering the shortages, researchers tend to focus on the more efficiency biological route – culturing microalgae with carbon dioxide. The carbon fixation rate with algae can up to be 15 g carbon /m2 per day with flow rate of 0.3 L/min of 4% CO2 in mucilage culture, and yield of biomass is 20- 30 times higher than traditional energy corps [17]. After harvesting, the

triacylglycerides are released from the biomass and then be transesterified to produce biodiesel and the green waste residue can also be fermented into bioethanol or biobutanol. Although the process has been proven successful in the lab, the cost of recovering the lipid from algae is still too high for most cases. While biological approach proves that CCS can be realized in a mild condition, its slow capture rate and again limited room of manipulation make this approach improper to be applied in industrial large-scale CCS practices.

1.4 Biochemical CCS

Compared with in vivo, temperature, pressure, and pH are more controllable in vitro condition. Thus, researchers switch their focuses into converting carbon dioxide to other organic or inorganic products in vitro with enzymes. Small molecule compounds such as methanol [18], lactic acid, and pyruvate [19] have been de novo synthesized with enzymes involved in native carboxylation cycles. As far as direct conversion of carbon dioxide is concerned, biological processes, however, generally afford incompetent reaction intensity. Since Baskata found the pH controlled the reaction extension of carbon dioxide to methanol, more efforts have been done to improve the reaction intensity with parameters controlling according to Debye–

Hückel theory, a thermodynamic model of bioreactions [20]. Similar to formate dehydrogenase reaction, cofactor is involved in ICDH reaction, indicating the pH sensitivity of reaction. However, for ICDH reaction, whether enough driving force for

carbon capture can be created through manipulating system pH is still unknown and need to be investigated.

The design of biocatalytic approach begins with screening ideal carboxylation enzymes. Form primitive Acyl-CoA, Dicarboxylate-hydroxybutyrate and

3-Hydroxypropionate pathways to advanced Calvin cycle, biology has evolved several major reaction pathways for conversion of carbon dioxide. In particular, tricarboxylic acid (TCA) cycle was first found in aerobic organisms mostly for oxidation of carbohydrates into carbon dioxide [21]. For an extended time, people believed this cycle was irreversible, yet the discovery of the reductive tricarboxylic acid (RTCA) cycle in anaerobic bacteria Chlorobium and Desulfobacterium revealed the feasibility of switching the direction of this metabolic pathway [22]. Within the cycle, the key step is the reaction catalyzed by the enzyme isocitrate dehydrogenase

(ICDH), which uses one molecule of NADPH and 2-oxoglutarate to form isocitrate by using one molecule CO2.

One concern about biocatalytic carbon dioxide capture is the utilization efficiency of carbon species. The Km values of carbon dioxide for enzyme reaction is around 100

µM to 1 mM, in accord with the low solubility of carbon dioxide (350 µM with 1 v/v % carbon dioxide in atmosphere pressure). According to Henry’s law, the solubility of carbon dioxide is limited by the partial pressure of carbon dioxide. Moreover, the carbon species exhibits self-evolution in aqueous solution and four forms of carbon

- 2- dioxide (CO2, H2CO3, HCO3 and CO3 ) exist in equilibrium which is strongly

affected by the system pH. As the conversion rates between the carbon species are much faster than the enzyme reaction rates, the substrate selectivity of carbon species is rarely mentioned. Instead, CO2 (tot) is used to represent the overall characteristics of carbon species. The selectivity of carbon dioxide, however, may affect the carbon capture rate significantly in CCS process due to its high reaction intensity. The structure of active site of enzymes may bring some clue about its selectivity.

- According to the enzyme structure, HCO3 was proven to the primary substrate for propiony1-CoA carboxylase, while dissolved gas phase carbon dioxide was believed as the dominated substrate for most carboxylases originated from the yeast.

This work aims optimization and manipulation of biocatalytic reactions of ICDH via both theoretical analysis and experimental studies, therefore revealing the feasibility of controlling the direction of biological carboxylation reactions, and further demonstrating application potentials of such biological reactions toward more efficient and intensive carbon fixation.

1.5 Objectives

The overall goal of this research is to establish an ICDH-based carbon capture and release strategy. pH instead of temperature is the factor to manipulate the reversibility of the reaction, realizing high energy efficiency. The specific objectives are:

1. Understand factors affecting the equilibrium of the ICDH reactions and find the limited parameter of the conversion efficiency of ICDH reaction. Toward that, both theoretical calculation and experimental studies will be conducted. 12

2. Establish reaction conditions that allow procedures for carbon capture and release cycle. Firstly, ICDHs from different resource will be screened to find the best for the cycle; then this enzyme will be characterized for both kinetic and equilibrium parameters; finally, the condition of capture and release cycle will be optimized.

3. Improve the carbon capture capacity of ICDH reaction by extended cascade enzyme chain reactions. The carbon capture reactions with or without aconitase will be conducted to compare the amount of carbon dioxide captured, with the result explained by the thermodynamic model.

4. Improve the enzyme stability via enzyme immobilization. In order to improve the enzyme stability, several immobilized methods are proposed to improve the enzyme stability against pH shifting. The immobilized enzyme will be characterized and compared with native enzymes for its stability, specific activity and so on.

5. Evaluate the efficiency of the proposed biocatalytic CCS using ICDH in terms of cofactor regeneration and carbon capture efficiency.

Chapter 2. Literature Review

2.1 Biocatalysts based carbon capture

Recent Concerns have been raised about seeking alternative carbon capture approach.

Compared with chemical absorbent reactions, biological absorption can be proceeding at ambient conditions with high reaction specificity and unparalleled high energy efficiency [23]. Among these carbon fixations, biology has evolved several major reaction pathways for conversion of carbon dioxide, including the Calvin cycle

[24], 3-hydroxypropionate/4-hydroxybutyrate cycle [1], Reductive tricarboxylic acid

(RTCA) cycle, etc. Screening and separating enzymes from these cycles is the first step to evaluate their application in CCS.

Calvin cycle

Calvin cycle is the most well-known C3 carbon fixation pathway exists in high-level photosynthetic organisms. The reducing equivalents of carbon dioxide in this cycle come from the splitting water and photons of sunlight. In this cycle, the carbon dioxide is fixed with ribulose-1, 5-bisphosphate to synthesize 3-phosphoglycerate, catalyzed by ribulose-1, 5-bisphosphate carboxylase oxygenase (RuBisCO). Compare with current photovoltaics performance, RuBisCO is inefficient of carbon conversion due to the competitive inhibition of oxygen on its active site [25]. Recent research is focusing on improving the carboxylation activity of RuBisCO in the photosynthetic hosts, aiming to increase the yield of hosts [26]. For example, based on the theoretical

calculation, genetically inserting RuBisCO from red algae into C3 plants could increase 25% of the yield of plants [27].

Reductive tricarboxylic acid (RTCA) cycle

Calvin cycle widely exists in advanced plants, while reductive tricarboxylic acid

(RTCA) cycle only exists in unique lower plants. The reductive TCA cycle appears to operate in phylogenetically diverse autotrophic bacteria and archaea, including genera of anoxic phototrophic bacteria (Chlorobium), sulfate-reducing bacteria

(Desulfobacter), microaerophilic, hyperthermophilic hydrogen-oxidizing bacteria

(Aquifex and Hydrogenobacter), and sulfur-reducing Crenarchaeota

(Thermoproteus and Pyrobaculum) [28-33]. The most remarkable feature of RTCA cycle is its reversibility. Figure 2-1 shows the scheme of RTCA cycle in nature: the cycle begins with Acetyl-CoA as initial substrate and ends with oxaloacetate as the final product, with fixing of two molecules of carbon dioxide. The identical intermediates involved in TCA and RTCA cycle reveal their reversibility. In TCA cycle, the cell takes and oxidizes complex hydrocarbonates to CO2 and water, while

CO2 is the substrate of carbon compounds synthesis in the reverse cycle.

Figure 2- 1. The scheme of RTCA cycle.

In RTCA cycle, several enzymes catalyze the biosynthesis of hydrocarbon from carbon dioxide, and one of the key steps is catalyzed by isocitrate dehydrogenase

(ICDH). The enzyme uses one molecule CO2 and 2-oxoglutarate to form isocitrate, a reversible reaction in TCA cycle. Huge differences in enzyme structure, kinetics, stability and activity are referred with enzymes come from different specifies, although they are all called “ICDH”. Among these differences, their cofactor dependence, NAD- or NADP-, is most remarkable [34]. Both carboxylation and decarboxylation activities are detected in NADP-dependent ICDH, while the decarboxylation activity is preferred for a wide pH range [35] [36]. Interestingly, some NADP-depended monomeric ICDHs exist in very species like Azotobacter

Vinelandii, Vibrio Parahaemolyticus , Rhodomicrobium Vannielii , Desulfobacter

Vibrioformis, and Corynebacterium Glutamicum with a molecular weight of 80 kDa.

As involved in RTCA cycle, considerably higher carboxylation activity is expected for monomeric ICDHs. Kanao summarized the differences between the monomeric

ICDH and normal dimeric ICDH from S. cerevisiae (SC-ICDH), based on his study on characterizing the ICDH from green sulfur bacterium Chlorobium limicola

(CL-ICDH). At neutral condition, CL-ICDH and SC-ICDH revealed apparent carboxylation and decarboxylation activity respectively. Also, CL-ICDH had a lower

Km value for carbon dioxide than that of SC-ICDH (1.3 mM vs 8.2 mM). All of these differences were in accord with their different functions in metabolism: SC-ICDH was a decarboxylation enzyme while CL-ICDH involved in carboxylation [33].

Acyl-CoA pathways

Acyl-CoA pathways exist in archaea such as thermoacidophilic archaebacteria of the genera Sulfolobus [37] and the routine is encountered in two separate pathways, namely 3-hydroxypropionate/4-hydroxybutyrate cycle and 3-hydroxypropionate bicycle. In 3-hydroxypropionate bicycle, two steps with carboxylation are catalyzed by the same enzyme (acetyl-CoA/propionyl-CoA carboxylase). In the first

– carboxylation step, acetyl-CoA is carboxylated to form malonyl-CoA with HCO3 as substrate, along with the hydrolysis of ATP; in the second step, methylmalonyl-CoA

– is synthesized with propionyl-CoA and HCO3 [38]. In

3-hydroxypropionate/4-hydroxybutyrate cycle, acetyl-CoA is first carboxylated to form pyruvate, then the pyruvate is phosphorylated with ATP and form

phosphoenolpyruvate, and later is the substrate to form oxaloacetate catalyzed by phosphoenolpyruvate carboxylase (PEPC) [39].

Most carboxylation reactions involved in Acyl-CoA pathways were not suitable for large-scale carbon conversion due to the requirement of unique cofactors. Liu reported the application of purified PEPC for carbon capture in vitro, but no carbon conversion efficiency and reaction intensity were mentioned in his report [40].

Other reversible pathways

Through alternation of the reaction direction by reaction parameters manipulation, native decarboxylation enzyme may also be applicable in carbon capture. For example, native decarboxylase, formate dehydrogenase, can reversely convert carbon dioxide directly to a formate. With another two reversible enzymes, namely, formaldehyde dehydrogenase and alcohol dehydrogenase, the formate can be further converted to menthol at mid condition. Thanks to the protein engineering and immobilization technology, the activity and stability of FDH have be improved thousand times. And again, with cofactor regeneration, the cost of the reaction is also expected to decrease dramatically [41-43].

Decarboxylases are another enzyme group which are applied in carbon capture with their reversible reaction direction. Tong has reported the enzymatic synthesis of L-lactic acid from carbon dioxide and ethanol and the carboxylation step is catalyzed by pyruvate decarboxylation under alkali condition. The β-carboxylation of para-hydroxystyrene derivatives with excellent regio- and (E/Z)-stereoselectivity 18

catalyzed by phenolic acid decarboxylases was also reported [44]. For most cases, however, the carboxylation activity of decarboxylases was hundred times less than their native decarboxylation activity.

At present moment, the carbonic anhydrases (CA) based biocatalytic carbon capture is most close to practice [45-48]. In nature, CA catalyzes the rapid interconversion of bicarbonate to carbon dioxide and water, a reversible reaction that occurs most in the pulmonary endothelium for the elimination of CO2 in the lungs [49]. The native function of CA makes it an excellent catalysis for carbon capture and release strategy.

With water as substrate, carbon dioxide is fixed to form bicarbonate under high pressure and released again as pure form through lowering the pressure. The cost of chemicals in CA strategy is low due to no requirement of expensive cofactor and substrate. Moreover, the CA shows high specific activity: as the most power enzyme, the turnover rate of CA is about 106 reactions per second [50]. Now, the concerns are focusing on improving the CA stability against high temperature, toxic gas, and low pH through various methods including molecule evolution, enzyme immobilization, etc. [41-43]. Besides applying of CA solely, the co-utilization of CA and chemical absorbent, such as MDEA, to improve carbon capture rate was also reported [51].

Partnering with the high reaction, the unfavorable equilibrium of CA strategy hurts its economic feasibility. Since CA does not improve the equilibrium carbon capture capacity of water [52], a large amount of water is required to conduct the CCS process, resulting in the increasing cost of equipment.

2.2 pH sensitivity of biocatalytic carbon capture

To surpass the limitation of bio-based carbon capture efficiency, extensive studies on the factors of determining the extent of enzymatic carboxylation reaction were conducted. Different from chemical reactions, there is very limited room for enzyme reaction to manipulate the reaction kinetic and equilibrium through temperature and pressure changing. However, other factors such as pH and ionic strength still prove enough room for manipulation. For an extended period, the effects of pH and ionic strength on enzyme reaction limited on reaction rate, and their effects on reaction equilibrium are still rarely referred. In later research, Debye–Hückel theory provides a new way to analyze the effect of pH and ionic strength on the reaction equilibrium and Alberty further investigate common biochemical reactions’ equilibrium constant under different pH and ionic strength [53]. With this theory, pH and ionic strength could shift the reaction equilibrium by affecting the thermodynamic property of substrates and products. Baskata applied this theory to the enzymatic reduction of carbon dioxide to methanol catalyzed by formate dehydrogenase, acetaldehyde dehydrogenase and alcohol dehydrogenase and pointed out that pH was the critical parameter to determine the conversion rate. According to the theoretical calculation, the Gibbs free energy of total reaction was shifted from 9.66 kcal at pH 8 to -28.39 kcal at pH 2 [20]. The shifting of Gibbs free energy revealed that the carboxylation became more favorable at a acid condition. The pH sensitivity of other carboxylation enzymes was also observed during the carbon capture process. For example, isocitrate dehydrogenase showed apparent carboxylation activity at pH 6, while the direction of 20

reaction was decarboxylation at pH 9 [33]. On the other hand, the carboxylation reaction catalyzed by pyruvate decarboxylase can only be conducted at a pH higher than 8 [19]. Other carboxylation enzymes such as phenolic acid decarboxylases and phosphoenolpyruvate carboxylase also showed special requirement of pH to execute carboxylation [44]. Interestingly, biocatalytic carboxylation reactions show different pH preference, some prefer acid condition, while others can only take place in alkali condition. This contradiction is mainly due to the different thermodynamic characteristics of chemical species on both sides of the equation.

The sensitivity of biocatalytic carboxylation provides a novel approach to realizing the carbon capture and release strategy. pH, instead of temperature or pressure can be the driving force to leverage the reaction equilibrium. Instead of heating or pressuring the entire system, the shifting of carbon capture and release can be realized by just changing system pH with adding acid and alkali, resulting in massive improvement of energy efficiency.

2.3 Reaction intensity of biocatalytic CCS

To realize quick and low-cost carbon capture and release, high reaction intensity is required for enzyme-based CCS. Capture rate and capture capacity are two key parameters that determine the reaction intensity. While the capture rate is related to reaction kinetics, capture capacity is determined by the thermodynamic characteristics of chemical species in the reaction.

Carbon capture rate 21

The reaction rate is one of the most important parameters to evaluate the performance of CCS. As traditional chemical based CCS capture, there are two fundamental mechanisms for the reaction of amines (MEA) with carbon dioxide [54]:

+ - 2R-NH2 + CO2 ↔ R-NH3 + RNH-COO (2-1)

+ - R-NH2 + CO2 + H2O ↔ R-NH3 + HCO3 (2-2)

Assuming the concentration of MEA and dissolved carbon dioxide are 7 M and 0.035

M respectively, the carbon capture rate can be as high as 0.012 mole/L per second.

Different from chemical reactions, increasing substrate concentration does not necessarily result in higher reaction rate for enzyme reactions. Most enzyme reactions already achieve the maximum rate with comparable low carbon dioxide concentration, with the value of Michaelis–Menten constant (Km) of CO2 between10 μM to 1 mM

[55-57]. The further increase of carbon dioxide concentration cannot accelerate the carbon capture rate. Instead, increasing enzyme loading is a routine way to enhance the carbon capture rate. Table 2-1 summarized the specific activity of native enzymes involved in carbon capture, and in practice.

Table 2- 1. Summary of specific activity of carboxylation enzymes

Enzyme Specific activity (Ua/g)

2,6-dihydroxybenzoic acid decarboxylase 231 [58]

Isocitrate dehydrogenase 19500 [33]

Formate dehydrogenase 1300

Pyruvate decarboxylase 24000 [19]

phosphoenolpyruvate carboxylase 24800 [59]

Carbon anhydrase 1.2 X109 [60] a One unit of activity was defined as 1 μmol of carbon dioxide fixed per min.

According to the data of Table 2-1, most enzymes show inferior specific activity, and the native decarboxylase such as formate dehydrogenase and 2, 6-dihydroxybenzoic acid decarboxylase are the worst. For isocitrate dehydrogenase and formate dehydrogenase, the enzyme loading to catch up the capture rate of MEA are 37 g/L and 554 g/L respectively. In practice, the reactions with such high enzyme level are questionable due to mass transfer and stability issues.

Enzyme screening is an ordinary method to pick up the enzymes with high specific activities. Regarding their different source, the “same name” enzymes may be varied in their consequences of amino acid, molecule weight, optimum pH and more important, specific activity. Kanao reported the specific activity of ICDH form the

green sulfur bacterium Chlorobium limicola was three times higher than that from

S. cerevisiae. And the specific carboxylation activity of formate dehydrogenase from seudomonas oxalaticus was also significantly higher than that from a yeast source

[33]. In this case, enzymes with higher carboxylation activity are expected in the future as vast numbers of species are still undiscovered in the nature.

Besides discovering new enzymes, modification of known enzymes through molecule biology technologies also provides an approach to increasing the specific activities.

Enzymes execute their function by binding the substrate at the active site, which requires an unique 3D conformation. Modification of one or multiple amino acids consequence of enzyme may affect the interaction between enzyme and substrates, resulting in better substrate selectivity, cofactor dependence and specific activity

[60-63]. Modern enzyme database [64], technologies of DNA denovo synthesis [65] and gene expression [66] make the enzyme modification more efficient and economical.

Carbon capture capacity

Carbon capture capacity is another factor affects the feasibility of CCS process. More carbon can be fixed with the unit concentration of the substrate with increasing carbon capture capacity. Different from carbon absorption rate, which can be manipulated with controlling enzyme loading, the carbon capture capacity is solely determined by the reaction equilibrium. Generally speaking, biocatalytic reactions afford lower carbon capture capacity than chemical reaction routes. An almost complete

conversion would be achieved by chemical carbon capture while most biocatalytic reactions reach a much lower degree of conversion limited by reaction equilibrium.

The carbon capture capacity of the traditional chemical absorbent is around 2.1 mole

/L with 7 M substrate. Such high substrate concentration, however, may cause the salting out of enzyme, and therefore reduce the reaction rate for most enzyme reactions. The low carbon capture capacity becomes the most apparent disadvantage in CA-based strategy: the carbon capture capacity of water is only 35 mM with one atmosphere pressure of carbon dioxide. For flue gas, the capacity is even lower as only (15-30%) CO2 is in the flue gas. In practice, pressurizing the flue gas could somewhat increase the carbon capture capacity. However, the process itself increases cost.

To achieve high conversion rate in biocatalytic reactions, cascade enzyme reactions are designed for carbon capture in a homogeneous solution, which can further convert the substrate, thus improving the overall carbon capture capacity. In practice, cascade production of methanol from CO2 with formate dehydrogenase (FDH), formaldehyde dehydrogenase (FaldDH) and alcohol dehydrogenase (ADH) was examined by Luo.

In the first step, FDH catalyzed the conversion of carbon dioxide to formate. The conversion rate was meager due to both the low solubility of carbon dioxide in water and the unfavorable reaction equilibrium. By further converting formate to formaldehyde with FaldDH and then to methanol with ADH in the next two steps, the overall carbon conversion rate increased thousand times [67]. The steps of RTCA

cycle were also reviewed by Shi [68]: CO2 is first converted into isocitrate with a +8 kJ mol−1 of ΔG°, and the single ICDH reaction is thermodynamically unfavorable. As the step following ICDH could significantly decrease the total ΔG° to 0.39 kJ mol−1, the feasibility of RTCA cycle is realized with cascade enzyme reactions.

2.4 Reaction stability of biocatalytic CCS

Enzyme stability with immobilization

The large-scale of biocatalytic CCS outperforming traditional chemical processes is realized with the durability of enzymes. Different form chemical catalysts, enzymes require specific conformation to complete their activity with a comparatively mild condition. Various parameters including temperature, pH, oxidizing agents and mechanical forces may cause the unfolding of the subtle tertiary structure of the enzyme, resulting in the loss of enzyme activity (Table 2-2).

Table 2- 2. Summary of enzyme denature mechanism in CCS

Factors Mechanism Most susceptible

amino acid

Temperature [69][70] Maillard reaction Lys

Isomerization Asn Asp and Pro

Racemization All amino acid

Intermolecular interaction Cys

Amino acid decomposition Pro and Trp

Oxidation [71] Chemical alternation Cys, Met, His, Try. His

Mechanical Force [72-74] Pressure gradient, All

Dissociation of submits,

Destabilization of

non-Covalent interactions

pH [75] Electrostatic repulsion Charged groups

Metal Chelators [76] Stripping off the metal ions Metalloproteins

To increase the durability of enzymes, immobilized enzymes are used instead of free enzymes in the industry for large-scale projects. The immobilized enzyme is an enzyme that attached on an inert, insoluble material which can provide increased

resistance to changes in conditions such as pH or temperature [77]. Several immobilization methods are reported widely in biomedical, environmental and food industry: 1. Entrapment of enzyme in materials; 2. Cross-linking of enzyme aggregates. 3. Enzyme binding support material with hydrophobic, ionic or covalent force. During entrapment process, the enzyme is entrapped in matrices such as sol-gel matrices formed with the hydrolytic polymerization of metal alkoxides. In this way, the enzyme stability is improved by limiting the change of enzyme conformation during the reaction. The immobilization of Lipases [78], catalase and horseradish peroxidase [79] have been realized with this method. Immobilization with cross-linking is obtained through cross-linking of enzymes via their surface NH2 groups with bifunctional chemical cross-linker like glutaraldehyde. Enzyme molecules are accumulated to form significant size after the cross-linking. The successful application of cross-linking with Penicillin G dramatically reduces the production cost of penicillin in the pharmacy industry [80]. Binding the enzymes with support materials is the most extensive immobilization method in industry due to its easy preparation process and high efficiency. Compared with the entrapment and crosslinking methods, the binding method results in less limitation of mass transfer between enzyme and substrate and therefore retain high activity [81]. The binding types between enzyme and support material include hydrophobic, ionic and covalent.

The strongest binding, covalent binding, forms between the side chain amino acids like arginine, aspartic acid, histidine of the enzyme and the functional groups like imidazole, indolyl, phenolic hydroxyl of the material. The hydrophobic and ionic are 28

weaker, which may cause enzyme leaking during the later application. Except for hydrophobic, ionic and covalent binding, the interaction between molecules (Van der

Waals' force) also participates in the immobilization and single immobilization method may involve more than one types of force. Besides increasing enzyme stability, another significant advantage of immobilized enzyme is its capacity of recycling. Immobilization allows enzymes to be held in place throughout the reaction, and they are easy to be separated from the products and reused. A summary of enzyme immobilization methods is shown in Table 2-3.

Table 2- 3. Summary with different enzyme immobilization methods

Method Mechanism Characteristics

Entrapment Physical entrapped in insoluble Improve thermos stability but

matrix high mass transfer limitation

Crosslinking Covalently bonded to each other Easy prepaid, low cost, but not

suitable for all enzymes

Covalent Bond covalently to an insoluble Strong bonding, but may lose

binding support enzyme activity

Adsorption With hydrophobic, ionic, affinity Easy process, wide

and van der waals' force applicability and less effect on

enzyme activity, but high

leaking rate

Screening the best immobilization method of the target enzymes could be a time-consuming task. In practice, three key parameters are related to the performance of immobilization: enzyme loading level, specific activity, and leakage rate. Enzyme loading level represents the amount of enzyme can be bonded with solid support material [82-84]. General, Adsorption provides the higher loading level than covalent bonding. Specific activity is defined as the activity of enzyme per unit mass of protein.

Firstly, during the process of immobilization, the fluctuations of pH, temperature, and other factors may reduce the specific activity of enzymes. And again, more mass 30

transfer limitation between the enzyme and substrate is expected after immobilization can also cause the decline of specific activity [85] [86]. Leakage happens as the immobilized enzyme detaches from the support material, which are expected more in physical adsorption and entrapment methods with a weak bonding. Leakage will become severer with the degradation of the immobilization materials as the pH, temperature and ionic strength of the system changes.

Cofactor stability during reaction

Different form chemical reactions, most enzyme reactions utilize cofactors as an electron acceptor instead of using electrons directly. NADP/ NADPH, NAD/NADH are two most common cofactor systems widely exist in oxidoreductases and hydrolases. The cost of cofactors is much higher than the enzyme and other substrate.

The price of per mole of NAD+ and NADP+ would be $3162 and $13700 respectively, and the reductive form of them are even higher as $10641 and $210415 respectively

[87].

Extensive studies concerning the stability of cofactor have been done, as maintaining the activity of cofactor during the process is critical for the economic feasibility of enzymatic CCS. Similar to the enzyme, the activity of cofactor is sensitive to the environmental factors of the process. Wu has reported the degradation rates of

NADPH and NADH on a broad range of pH and temperature. Based on the data, the concentration of proton dramatically affected the stability of cofactor: the halftime of

NADH was only 4 mins at pH 4, and can be improved to 28800 mins with a shift

of pH to 8 [88]. Also, the degradation rate constant tripled for each 10 oC up, indicating the negative effect of temperature on the stability of cofactor. Besides temperature and pH, the ionic strength and buffer catalog were also reported to affect the stability of cofactors [89] [90].

2.5 Economy feasibility of CCS

Capital cost of CCS

The capital cost of CCS is the sum of the direct costs of equipment and star up chemicals and various indirect costs including project contingency, interesting during construction, royalty fees, etc. In traditional MEA based CCS, the flow rate and carbon dioxide concentration of flue gas are the two primary factors influence the capital cost through deciding the size of absorption and regeneration columns [91].

Among the capital expense, more than 30% is related to product dealing such as drying and pressurizing carbon dioxide to high pressure.

For MEA CCS, the startup chemicals include CaCO3 for SO2 absorption and MEA for carbon capture and release. The cost of startup chemicals is usually neglected due to its small portion of the capital cost. However, for potential enzyme based CCS, the cost of startup chemicals is expected to significantly increase due to the high price of substrate and cofactor.

Operation cost of CCS

The operation cost of CCS process can be separated into two main categories: the cost of chemicals and the cost of energy. The cost of chemicals occurs for the makeup of absorbent due to self-degradation and supplying chemicals for SO2 treating [92]. And the energy cost is related to both the heat duty for absorbent regeneration and the shaft duty for pumping and turbomachines. Regarding the different mechanism of CCS, the proportions of shaft and heat energy cost are various. For MEA case, heat duty takes up 75-80 percent of total energy cost [93], while For CA case, most energy is used for transporting and pressurizing the gas.

Cofactor cost reducing

As electrons carrier, organic cofactors frequently occur in enzyme reactions. The common cofactors include nicotinamide adenine dinucleotide (phosphate) (NAD(P)+),

Flavin adenine dinucleotide (FAD) and coenzyme A (CoA)[94]. There are increasing concerns about the regeneration and recycling of NAD(P)H since its close relationship with carboxylation (decarboxylation) reactions. For potential enzymatic

CCS, the reduced form of cofactor is the startup chemicals consumed in carbon fixation. Generally, the reduced form (NAD(P)H) is 3-10 times more expensive than the oxidized form (NAD(P)+). Therefore, cofactor regeneration can reduce the capital cost with replacing the reduced form of cofactor with the oxidized form. The major regeneration methods available now include chemical, enzymatic and electrochemical methods [95-100].

In practice, formate dehydrogenase based NADH/ NADPH regeneration method is promising in biocatalytic CCS for its compatibility and high reaction extensity. Firstly, the total turnover number (TTN) of formate dehydrogenase can achieve as high as

600,000 [101], providing high reaction rate. Also, as a weak acid, most enzymes can tolerate the formic acid and the product (CO2) hardly has any effects on enzymes and can be easily removed from the reaction system. The cofactor regeneration of formate dehydrogenase is widely used in the producing of L-Tert-Leucine [102] and

S-1-Phenyl-2-Propanol [103] in large-scale. Regarding the source of formic acid, a method of electrochemical synthesis formic acid from carbon dioxide was reported, and the concentration of formate can reach as high as ~10 mM with 80 % efficiency

[104]. By combining of formate synthesis and cofactor regeneration strategy, carbon dioxide can be utilized as substrate to achieve high efficient cofactor regeneration for carbon capture. Besides formate dehydrogenase, other enzymes include glucose dehydrogenase, alcohol dehydrogenase, and amino acid dehydrogenase [105-107] are also reported for using in cofactor regeneration.

Comparison of CCS with other renewable technologies

To predict the development of methods aimed at reducing carbon emission, several aspects including the resource potential, technical potential, and economic potential need be considered [108]. While renewable energy mainly aims at reducing the carbon emissions during electricity generation, CCS is expected to be applied not only in electricity generation, but also in industry activities including fertilizer, blast

furnace steel, and cement production [109] [110]. As a result, the wider application of carbon dioxide of CCS than renewable energy technologies can result in a quicker development. On the other hand, the development of renewable energy is also limited by the local environment such as solar intensive and wind power intensity.

The economic potential of carbon reducing technologies can be evaluated according to their value of levelized avoided costs of electricity (ACE) [111]. ACE is defined as the additional cost per MWh capacity for carbon emissions avoided compared with coal fire plant and the ACEs with different technologies expected in 2022 are summarized in Table 2-4.

Table 2- 4. Levelized avoided costs of electricity for different resource in expected in 2022

Levelized avoided costs (2016 $/MWh) a

Plant Type Minimum Non-weight average Maximum

Coal with CCS 47.4 58.7 80.2

Nuclear 47.6 57.3 64.5

Geothermal 51.5 65.3 79.8

Biomass 47.5 58.3 80.3

Wind-onshore 44.1 53.2 76.3

Wind-offshore 47.1 57.8 79.0

PV 42.5 64.7 82.9

Hydroelectric 46.1 57.4 79.5 a Source: U.S. Energy Information Administration, Annual Energy Outlook 2017, January 2017, DOE/EIA-0383(2017).

Based on data of Table 2-4, wind-onshore shows the lowest ACE as 53.2 $/MWh, and the ACE form CCS (chemical based) is close to that of nuclear, biomass and 35

hydroelectric, all are around 57-58 $/MWh. In the next five years, the economic advantage of CCS seems not obvious compared with conventional renewable energy technologies, but the novel combination of CCS and EOR [112] may dramatically reduce the ACE of CCS.

Chapter 3. pH Sensitivity of Biocatalytic Carboxylation Reaction

3.1 Introduction

The factors of determining the extent of carboxylation reaction catalyzed by enzymes have attracted wide attention [113]. Different from chemical reactions, there is very limited room to manipulate the reaction kinetic and equilibrium through temperature and pressure changing for enzyme reactions, however, other factors such as pH still prove the possibility of manipulation [20]. The recent study reveals that pH of the solution does not only affect the enzyme activity [114], but also the reaction equilibrium. With Debye–Hückel theory, a novel way to evaluate the effect of pH and ionic strength on the reaction equilibrium, Alberty calculated the equilibrium constant of common biochemical reactions under various pH and ionic strength [53]. The manipulation of pH can be conducted with adding acid or alkali, while the ionic strength of system is much harder to control. Aspiring to carbon capture, Baskata applied this theory on the analysis of the pH effect on the enzymatic reduction of carbon dioxide to methanol. The results showed that the pH can shift the reaction equilibrium by affecting the thermodynamic property of substrates and products, therefore affected the carbon dioxide conversion rate. The Gibbs free energy of reaction was shifted from 9.66 kcal at pH 8 to -28.39 kcal at pH 2[20], indicating the reaction catalyzed by formate dehydrogenase, acetaldehyde dehydrogenase, and alcohol dehydrogenase was more favorable at acid condition.

Aiming to reveal the pH sensitivity of the ICDH reaction, the influence of pH on

Gibbs free energy of ICDH reaction will be evaluated via both theoretical calculation and experimental determination.

3.2 Materials and methods

3.2.1 Materials

α-Ketoglutarate, NADPH, NADP+, D-Isocitrate, sodium bicarbonate, 2-(N morpholino) ethanesulfonic acid (Mes), N,N-Bis(2-hydroxyethyl)glycine(Bicine),

2-(Cyclohexylamino)ethanesulfonic acid (Ches) and, isocitrate dehydrogenase (ICDH) from porcine heart (EC 1.1.1.42) were purchased form Sigma Chemical Co. (St. Louis,

USA). Unless specially mentioned, all other reagents and solvents used in the experiments were of the highest grade commercially available

3.2.2 Experimentally determination of Gibbs free energy of reaction

The pH sensitivity of the above reaction was first examined by experimental meanusrements of the reaction equilibrium constant, which can be expressed:

[Cit][NADP] Ke  [Keto][NADPH ][CO2(tot)] (3-1)

' rG  RT ln(Ke) (3-2)

Enzymatic reactions were conducted with the concentrations of ketoglutarate, isocitrate, and sodium bicarbonate fixed at 8.0, 1.0 and 3.5 mM respectively, in a solution of a total volume of 1 ml. Reaction equilibrium was achieved by controlling

the ratio of NADPH and NADP+ (thereby controlling the direction of the reaction) added to the reaction solution, with the assumption that an equilibrium was achieved at the point when the net reaction rate approached zero. Carboxylation or decarboxylation reaction rate could be detected by monitoring changes in the concentration of NADPH at 340 nm. ICDH was diluted and used directly without any further purification, with concentration of 80 µg-protein/ml in the final reaction solution (determined with Bradford protein assay). Mg2+ at a concentration of 4 mM was also added to retain enzyme activity.

3.2.3 Theoretical calculation of Gibbs free energy of reaction

0 The theoretical prediction of Gibbs free energy of the reaction (ΔrG ) was calculated with applying the Debye–Hückel theory, which expresses the pH- and ionic strength

I-dependency of Gibbs free energy of formation (ΔGi,f) of a species as:

2 1/ 2 O   pH (zi  N H (i))I  Gi, f ( pH, I)  G i, f (I  0)  RT N H (i) ln10   1 BI 1/ 2   (3-3)

0 0 0  G ( pH, I)  G f , products ( pH, I)  G f ,substrates ( pH, I) r   (3-4)

Where zi is the electronic charge of species i, NH(i) the number of hydrogen atoms, R the gas constant, B an empirical term and is assumed to be 1.6 L/mol1/2 for most common biochemical temperatures [115], and α value is 1.17582 kg1/2mol-1/2.

3.3 Results and discussion

3.3.1 Experimental determination of equilibrium constant

Equilibrium constant (Ke) can be determined as the reaction reaches equilibrium in either direction [116], however, this conventional method would elicit large errors due to the slow reaction rate, degradation of reactants, and the enzyme activity loss. To improve the accuracy, Ke was determined by finding a zero-net reaction rate from varying compositions of the reactants and products, following a method suggested previously for a different reaction system [117]. In this study, the concentrations of chemical species on both sides were fixed except those of NADPH and NADP+. By controlling the ratio of NADPH and NADP+, the rate of carboxylation or decarboxylation reaction was detected by monitoring the change of concentration of

NADPH at 340 nm. High reaction rate was observed when the reaction was far away from the equilibrium, and the equilibrium point was detected as the rate approached zero.

The process of determining the equilibrium point at pH 5 was shown in Figure 3-1.

The reaction rate was -0.00524 mM/Min with 160 mM NADPH and 0.1 mM NADP+, indicating the carboxylation reaction was ongoing. As the concentration of NADP+ was fixed, the reaction rate of carboxylation was decreased with the reducing of

NADPH and finally closed to zero with 70 mM NADPH. The approach was also conducted from the decarboxylation side with a lower concentration of NADPH.

Similar to the first approach, a very low reaction rate was detected when the concentration of NADPH was increased to 100 mM. In this study, the equilibrium point was obtained when the NAPDH concentration is 85 mM, which is the average of 70 mM (reading form carboxylation) and 100 mM (reading for decarboxylation). 40

The difference between the results from carboxylation and decarboxylation approach was related to the minimum resolution of equipment.

Figure 3- 1. Determination of equilibrium point for reaction at pH 5. (Concentration of NADP+ was kept at 0.1 mM while initial reaction rate was measured for different concentration levels of NADPH. Positive reaction rate refers to consumption of NADPH, carboxylation reaction; negative reaction rate indicates generation of NADPH, decarboxylation reaction)

During the routine process of determining the equilibrium point, only the chemical species on one side are added to conduct the reaction. The equilibrium point is obtained by continuously monitoring the concentration of substrates for the whole reaction stage and the reaction is assumed to be completed with no changing of the concentrations of substrates. Due to equilibrium limitation, a low reaction rate is expected in the final stage of determination when the reaction state is close to the equilibrium. As a result, the determination of equilibrium point is a time-consuming process and inaccurate under harsh conditions for most enzyme reactions. Compared

with the conventional method, this new approach begins with adding the chemical species on both side and the process is completed more quickly as only the initial reaction rates are monitored. In this study, whether the reaction equilibrium was achieved can be determined in 15 seconds. The advantage of the new approach is more noticeable as the degradation of cofactor and enzyme is more severe at extreme pH. The minimum resolution of this method can be improved with increasing enzyme loading level to get better discrimination of initial reaction rates between each data points.

3.3.2 The effect of pH on equilibrium constant

The experimental determined Ke of ICDH reaction under different pH was executed with the novel method talked above, and the results are shown in Table 3-1. Based on the data, the Ke of ICDH reaction showed a strong dependence on the pH. The highest value of Ke was 42.02 under pH 5 and decreased to 0.089 when the pH was shifted to 9. Four orders of magnitude of equilibrium constant can be changed with manipulation of system pH. Table 3- 1. Equilibrium concentrations and constants determined experimentally for different pH

pH 5 pH 6 pH 7 pH 8 pH 9

[NADP+] (µM) 100 10 10 1 1

[NADPH] (µM) 85±15 60±20 260±20 100±40 400±100

Ke 42.02±6.30 5.95±1.49 1.37± 0.01 0.36±0.10 0.089±0.018

The manipulation of pH is more economic and feasible in this study, compared with the temperature or pressure. The denature of enzyme with extreme temperature and the high energy cost to shift the temperature of dilute solution hinder the manipulation of temperature for biochemical reactions. The aqueous condition also provides little room for controlling the system pressure. In contrast, shifting the pH of enzyme reaction can be easily conducted by adding acid or alkali. In practice, the dissolved carbon dioxide could naturally adjust the pH to acid, while alkali condition can be achieved by adding hydroxide.

3.3.3 Debye–Hückel model of ICDH reaction

According to the Debye–Hückel, the theoretical values of Gibbs free energy of the

0 reaction (ΔrG ) was predicted and listed in Table 3-2, along with the experimentally

’ determined Gibbs free energy for the reaction (ΔrG ) under the same reaction conditions. Based on the data, the ΔGf of all chemical species possessed more positive values at the higher pH, while the sensitivity of pH on the thermodynamic property of species varied with their different electronic charge and hydrogen atoms. The Large molecules such as NADPH experienced dramatic changing of ΔGf, between pH 5 and

9 (59.05 kJ/mol at pH 5 and 534.59 kJ/mol at pH 9), and the ΔGf, of small molecules such as water revealed less sensitive to the pH shifting. Finally, the effect of pH on

0 0 ΔrG was calculated: a negative ΔrG was revealed under acid condition, which

0 achieved the minimum value of -3.63 kJ/mol at pH 5; the value of ΔrG became

0 positive with the elevating of system pH, and the maximum value of ΔrG (16.71

0 kJ/mol) was obtained at pH 9. The manipulation room of ΔrG was 20.34 kJ/mol at the pH range of this study.

Table 3- 2. Gibbs free energy of formation of chemical species (ΔGf) at different pH calculated via Debye–Hückel theory 0 (ΔrG is the theoretical Gibbs free energy change for the reaction; ΔrG’ is Gibbs free energy change for the reaction calculated using experimental data of equilibrium constant Ke). Room temperature and I= 250mM

ΔGf (kJ/mol)

pH 5 pH 6 pH 7 pH 8 pH 9

Ketoglutarate -679.25 -656.42 -633.58 -610.75 -587.92

Isocitrate -1019.33 -989.91 -960.5 -931.08 -901.67

NADPH -59.05 89.36 237.77 386.18 534.59

NADP+ -108.72 33.98 176.68 319.38 462.08

CO2(tot) -564.61 -554.49 -547.1 -541.18 -535.8

H2O -178.49 -167.07 -155.66 -144.24 -132.83

0 ΔrG (kJ/mol) -3.63 -1.45 3.43 9.81 16.71

’ ΔrG (kJ/mol) -9.26±0.40 -4.42±0.71 -0.78±0.18 2.55±0.83 5.98±0.56

In our calculation, CO2 (tot) is assumed as the substrate of the enzyme. The thermodynamic properties of CO2 (tot) reflects those of the four carbonaceous species

- 2- (H2CO3, HCO3 , CO3 and CO2) at equilibrium in a solution. Although it has been speculated that dissolved CO2 is the favored substrate of the enzyme for reaction

[118], the carbonaceous species can realize equilibrium with a much faster reaction

- rate (rate for HCO3 degrade to CO2 is 6 mM/min). Accordingly, CO2(tot) should be a

better index of the entire system when an aqueous solution of carbon dioxide is concerned, as has been generally suggested for most similar biological reactions [53].

Both experimental measurement and theoretical analysis indicated the pH sensitivity

0 of carbon dioxide conversion catalyzed by ICDH. The values of ΔrG and ΔrG’were plotted in Figure 3-2. Under neutral and alkali conditions, the reaction affords positive

Gibbs free energy of reaction, and carboxylation is limited by the thermodynamic

0 favorability. However, both ΔrG and ΔrG’decrease sharply when pH decreases, and carboxylation reaction becomes favorable. The dependence of Gibbs free energy on

 G 0 / pH the pH was shown in Table 3-3. The values of r was between 3.44 and

4.84 kJ/mol and the values of rG'/ pH was between 4.54 and 6.04 kJ/mol. Both derivations were positive and close to each other in this pH range.

Figure 3- 2. Comparison between experimental determination and theoretical prediction for Gibbs free energy change for carbon conversion reaction.

0 (Filled circles are ΔrG and open circle are ΔrG’) room temperature and I= 250mM 45

Table 3- 3. The dependence of Gibbs free energy on pH

kJ/mol pH 5 pH 6 pH 7 pH 8 pH9

0  r G / pH 4.84 4.24 3.49 3.39 3.44

 r G'/ pH 4.54 4.64 5.31 5.73 6.04

In our comparison, there was a gap between the theoretical prediction and experimental observation. The main reason for the gap between theoretical value and experimental value, we assume, was the effect of activity coefficient (Γ)

  0 [Cit][NADPox ] Cit NADPOX rG  RT ln(Ke * )  RT ln( * ) [Keto][NADPred ][CO2(tot)]  keto NADP  CO Re d 2 (tot) (3.5)

[Cit][NADPox ] rG' RT ln(Ke)  RT ln( ) [Keto][NADP ][CO2(tot)] red (3.6)

0 rG  rG'RT ln()   , where   Cit NADPOX (3.7)    keto NADP Re d CO2 (tot)

Following such relationships, the value of Γ can be determined to be 0.10 at pH 5 and

0.013 for pH 9 for the current reaction system, according to equation (3.7).

Nevertheless, both the experimental data and theoretical predictions pointed out the same tendency of reaction equilibrium shifting as a function of pH. The pH dependency of the reaction Gibbs free energy agreed well between prediction and experimental determination. Overall, we believe this model predicts well the role of pH in regulating the directions of such reactions.

3.3.4 Prediction of carbon dioxide fixation with ICDH

Noticing the pH sensitivity of the reaction, we may conceive a carbon capture strategy by conducting carbon capture from gaseous mixtures at low pH and releasing pure

CO2 for storage or utilization at high pH by reversing the reactions. In practice, the conversion rate will be limited by the low solubility of carbon dioxide in water, which is governed by Henry’s law:

P = KHC （3.8）

Where P: solute vapor pressure; KH: Henry’s law constant which is 27.03 bar/M [119] for CO2 at room temperature; C: the solute concentration.

Table 3-4 tabulates the results of carbon capture capacity of the reaction at different pH values using thermodynamic analysis and carbon dioxide solubility value calculated from Henry’s law. With 1M of α-Ketoglutarate and NADPH and under 1 bar pressure, one capture and release cycle between pH 5 and pH 9 may have the potential to capture 1.25 M CO2 , over 30 times higher than that realized by physical adsorption of carbon dioxide in water. That prediction with the same concentration of substrate under 5 and 10 bars pressure was also shown in the table. Under high pressure, the solubility of carbon dioxide catches up the concentration of other substrates (1 M), and therefore shows less limitation of carbon conversion.

Table 3- 4. Carbon capture capacity of the ICDH reaction under different pressure

P (CO2,g) [ CO2(tot)] (bar) (M) pH 5 pH 6 pH 7 pH 8 pH 9

1 0.037 1.25 0.47 0.23 0.12 0.06

5 0.19 2.82 1.06 0.51 0.26 0.13

10 0.37 3.94 1.48 0.71 0.36 0.18

3.3.5 pH sensitivity of other biochemical reactions for carbon fixing

Similar to ICDH reaction, many decarboxylation reactions catalyzed by enzymes were proven to be reversible, and their carboxylation activities were repeated in vitro

[120-123]. To verify their potential application in carbon capture and release, theoretical prediction of their pH sensitivity with Debye–Hückel was conducted in this study and shown in Table 3-5.

Table 3- 5. pH sensitivity of biochemical reactions involved in carbon fixation

ΔrG (kj) ΔrG (kj) Enzyme Reaction at pH 5 at pH 9

Pyruvate carboxylase Pyruvate + CO2 (tot)+ ATP= oxaloacetate+ ADP+ Pi 0.75 -8.81

CO2 tot + NADred+ acetylcoA=

Pyruvate dehydrogenase pyruvate+CoA+NADOX+H2O 0.08 0.82

Ketoglurarate SuccinylcoA + CO2 tot + NADred= ketoglurarate+ dehydrogenase NADOX+ COA+ H2O 0.04 0.92

Phosphoenolpyruvate Phosphoenolphosphate+ ADP+ CO2 tot = carboxykinase oxaloacetate+ATP+H20 -1.15 -1.62

Formate dehydrogenases CO2 tot + NADred = formate + NADOX 192.58 174.97

Pyruvate decarboxylase Acetaldehyde + CO2 tot = pyruvate 201.18 172.26

Thermodynamic feasibility is the first parameter to predict their performance of CCS.

Based on the data, the reactions catalyzed by pyruvate decarboxylase and formate

0 dehydrogenase are thermodynamically unfavorable with the ΔrG around 200 kJ/mol

at the full pH range. The result indicates their slim application to carbon capture. The

phosphoenolpyruvate carboxykinase reaction, on the other hand, shows a negative

0 ΔrG , indicating the potential high conversion rate of carboxylation. No preference

between the carboxylation and decarboxylation are predicted in other reactions due to

0 their slight positive ΔrG .

Besides the thermodynamic feasibility, the pH based carbon fixation and release cycle

0 0 also requires the elevation of ΔrG with pH shifting. The slight elevation of ΔrG

between the pH 5 and 9 (0.76 kJ/mole) of phosphoenolpyruvate carboxykinase

reaction shows poor capacity of CCS. Although significant portion of carbon can be

fixed initially, most of it is hard to release with simple pH shifting. Pyruvate 49

0 carboxylase reaction, however, indicates the high potential with a 9.5 kJ/mole ΔrG

0 shifting. Interestingly, both negative and positive dependence of ΔrG on pH exist in

Table 3-5. While pyruvate carboxylase prefers alkali condition for carboxylation, the carboxylation of pyruvate dehydrogenase requires an acid condition. As most reactions involved cofactor, their pH sensitivity comes from the unique proton requirement of biochemical reactions

3.4 Conclusions

In conclusion, this work demonstrates that the reduction reaction of carbon dioxide catalyzed by ICDH has strong pH sensitivity, and can be reversibly manipulated by adjusting the pH of the reaction solution. Allowing non-ideality of the chemical species involved, the experimental data agreed well with theoretical thermodynamic predictions. Although manipulating pH in vivo is probably not an operational option, biochemical reactions using isolated enzyme may offer such convenience and promise processes to capitalize on the potentials of the reactions. Considering our growing capability of developing robust enzymes for extreme conditions, one can expect efficient carbon capture and conversion can eventually be realized by manipulating biological reaction pathways in vitro.

Chapter 4. Characteristics of ICDH for Carbon Capture

4.1 Introduction

As an enzyme participates in both TCA cycle and RTCA cycle, ICDH catalyzes the reversible conversion of ketoglutarate and isocitrate. The reversibility of ICDH makes it a sound candidate to realize the carbon capture and release strategy. Following the thermodynamic model, enzyme characterization and reaction optimization are conducted to improve the system efficiency. According to the cofactor dependence,

NADPH-dependent and NADH-dependent ICDHs exist in various microbiological species [124]. Besides the cofactor dependence, ICDHs from different resource also show vast varieties in their kinetics, structure, and stability [125-127]. For example,

ICDHs from the porcine heart, bovine heart, and yeast are multi-sub-unit enzymes composed of polypeptide chains with average molecular weights around 40,000 and have function in TCA cycles [128-130], while ICDH isolated from Chlorobium limicola is a monomer with apparent carboxylation activity [33].Generally, the ICDHs involved in carbon fixation recycle possess high carboxylation activity, while the ones in TCA cycles prefer a decarboxylation function.

Considering the high selectivity of enzyme reaction on substrate, the carbon substrate preference for the ICDH is also critical for carbon capture. There is an equilibrium

- 2- between the carbon species (CO2 dissolved, HCO3 and CO3 ) in water [131] and the preference of carbon species, that would affect not only the reaction rate but also the reaction equilibrium. Many references reported the combined utilization of carbonic

anhydrase (CA) and other carboxylation enzymes for carbon capture with the acceleration of the carbon dioxide equilibration[132][133]. As a powerful biocatalyst accelerating the transformation of carbon dioxide to bicarbonate ion, the first dissolve rate coefficient of carbon dioxide into the water can be improved from 5 × 10−2 to 1.6

× 106/s with the addition of CA, benefiting the mass transfer between gas phase carbon dioxide and reaction solution.

4.2 Material and methods

4.2.1 Matieral

Kanamycin B sulfate salt, isopropyl thio-b-D-galactoside (IPTG), Imidazole, p-nitrophenyl acetate And Bradford Reagent were purchased from purchased form

Sigma Chemical Co. (St. Louis, USA). The plasmid pET28A and Escherichia coli strain BL-21(DE3) were purchased from Novagen (Bilerical, MA, USA).

Others see the section 3.2.1

4.2.2 Assay of total protein concentration

The protein concentration in the aqueous solution was measured by Bradford protein assay. The mechanism was based on a color shift of the dye Coomassie blue reagent when bound to the protein being assayed. In a typical measurement, 0.5 mL of reagent was put into 1.4 mL of cuvette followed by 0.5 mL of enzyme solution. The solution was incubated at room temperature for 5 min and then the absorbance at 595 nm was

measured on UV–Vis spectrometer (50 Bio, Varian). BSA was used as the standards for calibration.

4.2.3 Assay activity of ICDH

In the decarboxylation reaction, the assay mixture contained 10 mM isocitrate, 2 mM

+ NADP , 40 mM MgCl2 and free enzyme in 1 mL of buffer. With prepaid time, the mixture was centrifuged quickly and the increase of NADPH in the supernatant was detected by absorbance at 340 nm, and one unit of decarboxylation activity was defined as 1 μmol of NADPH formed per min. The carboxylation activity was detected in the same way except the mixture was composed of 20 mM sodium ketoglutarate, 3 mM

NADPH, 40 mM MgCl2, 35mM NaHCO3 and one unit of carboxylation activity was defined as 1 μmol of NADPH oxidized per min. Mes buffers with pH values from 5.0 to

7.0; Bicine buffer with pH 8.0 and Ches buffer with pH 9.0 were used in this work.

Reaction rate at different temperature was conducted to evaluate the effect of temperature on enzyme activity. The assay mixture of decarboxylation without enzyme was kept in an electric-heated thermostatic water bath to get the targeted temperature.

Then enzyme solution was added to assay mixture to conduct the reaction and the initial reaction rate in the first 30 second was recorded, avoiding the effect of temperature on enzyme stability.

Similar with manipulation of pH, decarboxylation reaction was conducted under condition modified with various ionic strength and concentration of anions or cations to investigate their effect on enzyme activity.

4.2.4 Cell culture

The recombinant cells were cultured in Luria–Bertani medium at 37 °C, 250 rpm and

50 ug/ml kanamycin was added to prevent contamination. When the value of

OD@600 was reached 0.8, 0.1 mM isopropyl thio-b-D-galactoside (IPTG) was added to induce the expression of ICDH, and the temperature was kept at 28 °C. After 12 h, cells were harvested from a total 500 ml culture with centrifugation and washed twice with 0.1 M potassium phosphate buffer (pH 7.2), then resuspended in the same buffer.

The suspended cells were disrupted by sonication on ice and centrifuged for 15 min at

15 000 Xg to remove cell debris. The total protein concentration of cell extract was detected by Bradford method.

In order to optimize the expression process, only one parameter ( induce temperture, time of induce and IPTG concentration) was varied from the above standard culture condtion with 50 mL medium. The total cell weight was dried and weighted with scale and total enzyme acitivty (decarboxylation) in cell extract was detected at pH 7 with method described in 4.2.3.

4.2.5 ICDH purification

The expressed ICDH was purified through the Histidine Tag Column (His GraviTrap, bed size 7 mm × 25 mm, bed volume 1ml). 10 ml of binding buffer pH 7.4 contains 54

50 mM sodium phosphate, 300 mM NaCl was added to the column, followed by 10 ml of supernatant of the cell extract. Then 10 ml of wash buffer pH 7.4 contains 50 mM sodium phosphate, 300 mM NaCl and 20 mM imidazole was added to remove the untargeted protein. In the end, 5 ml of elution buffer contains 50 mM sodium phosphate, 300 mM NaCl, 100 mM imidazole was used to elute the ICDH from the column. The active fractions were examined for apparent homogeneity by

SDS/PAGE.

4.2.6 Determination of Michaelis constants of ICDH reaction

Determination of the Michaelis constants (Km) was conducted through

Lineweaver-Burk double reciprocal plots. For each experiment, substrates and enzyme except the target substrate were added at the concentration described in the standard assay of ICDH. By varying the concentration of target substrate, different

+ reaction rates were recorded. The Km value of isocitrate and NADP were obtained from decarboxylation reaction and the Km value of ketoglutarate and NADPH were obtained from carboxylation reaction through plotting the 1/ V vs 1/ [S] (Figure 4-1):

Vmax was determined by the point where the line crosses the 1/ V = 0 and

Km equaled Vmax timed the slope of line.

Figure 4- 1. Lineweaver-Burk plot. (Km: Michaelis constants; V: reaction velocity; Vmax: maximum reaction velocity; [S]: concentration of substrate)

4.2.7 Assay activity of CA

The CA catalyzed the hydrolysis reaction of p-nitrophenyl acetate to form p-nitrophenol and acetate, and the activity of CA was detected by monitoring the formation rate of p-nitrophenol. One unit of activity was defined as 1 μmol of p-nitrophenyl acetate hydrolyzed per min. Before the experiment, p-nitrophenyl acetate was dissolved in acetonitrile to get final concentration 60 mM. In a typical measurement, 10 μl of 60 mM p-nitrophenyl acetate solution was added into 990 μl of

50 mM pH 7.6 phosphate buffer, then 10 μg of CA was added into the system to initiate the reaction. The solution was mixed and incubated at room temperature in three mins and the slope of increase of absorbance at 348 nm was measured on UV–

Vis spectrometer (50 Bio, Varian). The assay mixture without adding enzyme was used as the standard for calibration.

4.2.8 Determination of selectivity of carbon substrate

The selectivity of carbon substrate was revealed by comparison of the initial carboxylation reaction rate of ICDH with pre-equilibrated and un-equilibrated

- HCO3 -CO2 solutions. The reactions were conducted with the standard concentration of substrate except NADPH (200μM) for carboxylation with both pH 6.5 and pH 7.5

MES buffer which was pretreated with pure nitrogen to remove the potential

- dissolved carbon dioxide. For pre-equilibrated HCO3 -CO2 group, before adding an enzyme, the solution was bubbled with gas phase carbon dioxide through the plastic tube settled in the bottom of the cuvette about 1-2 seconds and waited for the equilibrium of carbon species in 5 mins. After adding the enzyme, the decline curve of concentration of NADPH was recorded by spectrophotometer. For un-equilibrated

- HCO3 -CO2 group, the same process was conducted except that the enzyme was added immediately after bubbling carbon dioxide.

To further analyze the enzyme reaction rate with different carbon species, carbon anhydrase was introduced in the un-equilibrated group to accelerate the conversion of

- dissolved gas phase carbon dioxide to HCO3 . 0.1 mg of CA was added and mixed with the solution quickly with a pipette 15 s later than the initiation of the reaction.

The decline curve of the concentration of NADPH was recorded and compared with that of the previous group without CA to reveal the difference.

4.2.9 Carbon dioxide capture with bubbling gas mixture

The gas capture experiment was conducted by continuously bubbling mixing gas into absorption solvent. Carbon dioxide and nitrogen released from high-pressure cylinders and the total flow rate was fixed at 30 ml/min by a gas gauge. The gas was mixed and bubbled through a plastic tube at bottom of a 20-ml gas bottle filled with adsorption solvent. The concentrations of ketoglutarate, NADPH, and KHCO3 in adsorption solution were fixed at 5, 5 and 25 mM in 100 mM pH 7 MES buffer and the solvent was bubble nitrogen to remove the dissolved carbon dioxide before added enzyme. 1 U of

ICDH was added into the absorption solvent to initiate the reaction, and for a prepaid time, the concentration of NADPH was monitored at A340 with the spectrophotometer.

The concentration difference of NAPDH between experiment and control samples (not add enzyme) were compared to calculate the amount of carbon captured.

Figure 4- 2. Scheme of carbon dioxide capture with bubbling gas mixture.

4.2.10 Carbon dioxide capture with surface covered with mixture gas

The carbon dioxide capture experiment was also conducted with the following design

(Figure 4-3). The concentration of absorption solvent was the same as bubbling gas test but the carbon dioxide was supplied through interfacial mass transfer. The volume of solvent was 20 mL in a glass bottle, and the interfacial area was controlled with plastic foam. A 100-ml beaker covered the glass bottle and its chamber filled with gas mixture. After adding enzyme into the absorption solvent, the glass bottle was quickly moved to the chamber of the beaker. Gas mixture flow rate was kept at 30 ml/min to maintain the partial gas pressure of carbon dioxide, and the solvent was stirred at 200 rpm. For a prepaid time, 200 μl of the solvent was removed with pipette quickly, and the concentration of NADPH was monitored at A340 with the spectrophotometer. 59

Figure 4- 3. Scheme of carbon dioxide capture with surface covered with mixture gas.

4.3 Results and discussion

4.3.1 ICDH expression and purification

In this study, ICDH from Chlorobium limicola was expressed with E. coli BL21 system. The recombined E. coli was cultured at 37 °C and 28 °C. The cell suspension was monitored at OD at 600 nm for every two hours to determine the different phases of growth (Figure 4-4). As the value of OD closed to 0.6, E. coli entered the steady-state growth after 4 and 5 hours culture. After that, the OD value was increased until 3.2. With the high growth rate of E. coli at 37 °C, only 10 hours were needed to get the maximum OD value, while 14 hours were required at 28 °C.

3.5

2.5

1.5 OD @ 600 nm 600 @ OD 1

0.5

0 0 2 4 6 8 10 12 14 16 18 Time (hour) Figure 4- 4. Growth curve of recombined E. coli under 37 °C (Black square) and 28 °C (White square).

Lactose operon was designed for the protein expression of recombined E. coli. IPTG instead of lactose was the inducer to trigger transcription of the lac operon and to induce the ICDH expression. The effect of concentration of IPTG, temperature, and the time of inducement on the cell dry weight, and the total enzyme activity were detected and shown in Figure 4-5.

Figure 4- 5. Cells dry weight (left) and total enzyme activity (right) with various; time of inducement (a), induce temperature (b) and IPTG concentration (c).

Time of inducement is ctirical for the performance of expression when IPTG is utitlized as inducer . IPTG is widely used to trigger transcription of the lac operon for protein expression [134-136]. The concentration of IPTG during the culture can be kept at a constant level as it cannot be metabolized. In addition to the advantages, many researchers have reported the negative effects of IPTG on the cell metabolism[137][138], especially when the concentration is high. The potential causes include the overburden of the protein expression and also the quick accumulation of expressed proteins, especially when the protein is toxic for the cell.

In order to alleviate the negative effects, the inducement was conducted with OD values between 0.3 and 1, which was considered as a signal for the exponential phase of cell growth according to the growth curve. During this stage, most cells were energetic and resistible to stress pressure [139]. After optimization, the biomass and total activity achieved the highest values when the inducement happened at an early exponential stage with the OD value of 0.6. The delay of the inducement caused a slight decrease of both biomass yield and enzyme activities, while advance inducement was worse: only two-thirds of enzyme activity was obtained when the cell was induced with 0.4 OD.

According to the Figure 4-5(b), the inducing temperature also affected both biomass and total enzyme activities. Since cell growth and protein accumulation preferred high temperature[140], the highest cell weight (41.5 mg for 50 ml medium) was obtained at 36 °C, and lower inducing temperature resulted in a decline of cell weight. The enzyme activity, on the other hand, required a bit lower temperature to achieve the 63

highest value. The contradiction of optimum temperature between the yield of biomass and enzyme activity revealed the different synthesis pathway between the biomass and the target enzyme (ICDH). As an enzyme, the ICDH needed a special folding process after the linear sequence of its amino acid structural units was formed.

When the rate of forming linear sequence was beyond the maximum rate of enzyme folding, inclusion bodies were formed, instead of activity enzyme. These inclusion bodies were insoluble and lacked the right conformation to reveal enzyme activities.

Since the primary purpose of cell culture was to obtain enzyme activity instead of the protein itself, 28 °C was chosen as the optimum inducing temperature.

Finally, the IPTG concentration was optimized for enzyme expression. In order to induce the expression, the IPTG need to enter the cell by diffusion and bind with the targeted consequence of the plasmid with affinity. A suitable concentration of IPTG is critical for the successful inducement: a low concentration of IPTG cannot cover all the targeted plasmid, while high level reveals overcurrent of the cell metabolism.

[141]. Typically, 0.1-1 mM concentration of IPTG was considered as prevail level and in our study, and 0.3 mM IPTG displayed the best effects on both yield of biomass and enzyme activities. As shown in Figure 4-5 (c), 0.1 mM level of IPTG could support the inducement and only slight improvement of activity was achieved with 0.3 mM IPTG. On the other hand, a too high concentration of IPTG, like 3 mM, caused apparent inhibition effect on enzyme activities. Considering the high cost of

IPTG, 0.1mM instead of 0.3 mM IPTG was considered as optimum concentration.

In conclusion, the total activities of the enzyme were not linearly dependent on the total biomass produced. Both the cell growth and enzyme folding were critical to achieve the maximum enzyme activity. Based on the data, the inducement in the experiments was obtained when the OD 600 reached 0.6 value at temperature 28 °C and 0.1 mM IPTG.

4.3.2 pH profile of different ICDHs

To screen the ideal ICDHs for the carbon capture and release, ICDH from three different species, namely porcine heart (ICDH-ph), Chlorobium limicola (ICDH-cl) and bacillus subtilis (ICDH-bs) were chosen in this project. ICDH-ph and ICDH-bs were commercial forms of the enzyme and ICDH-cl was expressed in the lab. The pH profiles of these enzymes were shown in Figure 4-6.

2.5 2 1.5 b 1

0.5 Specific activity Specific activity (U/mg) 0 5 6 7 8 9 pH

Figure 4- 6. The pH profiles of ICDHs (a) ICDH-bc, (b) ICDH-ph, (c) ICDH-cl. Solid line: decarboxylation. Dash line: carboxylation.

Both carboxylation and decarboxylation activities were detected for ICDH-cl and

ICDH-ph, revealing the bi-function of these enzymes, different from the mono decarboxylation activity of ICDH-cl. Besides the reversibility, these three enzymes also revealed different optimum pHs for functions. For ICDH-cl and ICDH-ph, the optimum pH for decarboxylation was 8, while ICDH-bc preferred pH 7 for decarboxylation. For carboxylation, the optimum pH was 6 for both ICDH-cl and

ICDH-ph. Since the direction of the reaction was determined by the net reaction rate for both carboxylation and decarboxylation, ICDH-ph was more suitable for carbon release. For the full range of pH, the decarboxylation activity of ICDH-ph was much higher than the carboxylation activity, even at the optimum pH for carboxylation. On the other hand, for ICDH-cl the carboxylation activity exceeded the decarboxylation when the pH was lower than 6, enabling ICDH-cl as carboxylation enzyme.

The difference of pH profile of enzymes can be contributed to the variety of their structure and function in nature. ICDH-bs is a homodimeric enzyme that catalyzes the oxidative decarboxylation of isocitrate, to yield ketoglutarate and CO2 with concomitant reduction of NADP+ to NADPH. It lies at a critical juncture between the

Krebs cycle and the glyoxylate bypass, a pathway required for growth on substrates such as acetate, ethanol, or fatty acids [142]. In contrast, the ICDH-cl is a carboxylation enzyme in the RTCA cycle for a synthesis of hydrocarbons from carbon dioxide and water. It is believed that ICDH in ancient cells had both carboxylation and decarboxylation activities. During the biological evolution, the ICDH was modified at the molecule level to fit different growth condition. Some ICDHs focused 67

more on decarboxylation and lost the carboxylation activity, while some kept both activities.

After the pH profile, the values of Michaelis constants (Km) were also compared and shown in Table 4-1. Table 4- 1. Michaelis constants of substrate for the ICDHs

Reaction Properties ICDH-bc ICDH-ph ICDH-cl

Decarboxylation Km (μM)

isocitrate 11.2±4.5 75.8±4.3 45.4±6.8

NADP+ 33.4±6.5 17.4±2.3 28.9±7.4

Carboxylation ketoglutarate 73.2±11.7 24.5±8.6

NADPH 23.4±5.4 13.6±7.1

According to the Michaelis–Menten kinetics model

푣max [푆] 푣0 = (4-1) 퐾푚+[푆]

Km is defined as the concentration of substrate required when the reaction reaches half of Vmax and lower Km indicates high affinity of the substrate for the enzyme. In Table

4-1, the Km values of ketoglutarate and NADP for ICDH-cl were 24.5 uM and 13.6 uM respectively, much lower than them of ICDH-ph, which were 73.2 uM and 23.4 uM. These results explained the better performance of carboxylation with ICDH-cl.

For decarboxylation, ICDH-bc had the smallest value of Km for isocitrate and

ICDH-ph possessed that for NADP+, which revealed a preference for decarboxylation of these two species.

The ratio kcat/Km often referred to as the specificity constant, an index to compare the relative rates of enzyme acting on substrates. With the highest specific activity and again lower Km value, ICDH-cl was the best candidate for carbon capture and release strategy. Compared with the commercial forms of ICDH, the special activity of the expressed enzyme was 50 times higher for both carboxylation and decarboxylation.

ICDH-cl achieved the maximum carboxylation at pH 7, while ICDH-ph exhibited the maximum carboxylation activity in more acidic conditions.

According to the pH profile of ICDH, the carboxylation step of carbon capture and release was the time-limiting step, since the maximum carboxylation activity was four times less than that of decarboxylation. Hence with the same amount of enzyme, the time required for carbon capture will be much longer than that of carbon release.

4.3.3 Parameters effect the ICDH acitivties

Besides system pH, the performance of enzyme is also affected by various environmental factors including temperature, ionic strength, inhibitor concentration etc. To realize efficient carbon capture and release cycle, the effects of these factors on enzyme kinetics were studied in this chapter.

Metal cofactor concentration

In our study, the expressed ICDH-cl requires Mg2+ or Mn2+ as metal cofactor to maintain the enzyme activities, and similar results have been reported for ICDH from

other species [28-30]. The function of these divalent cations is to keep the stable structure of the active site of the enzyme and deliver the electron during the catalytic reactions. Without an additional supply of Mg2+ or Mn2+ during the reaction, the enzyme only retained 20 percent of the activity. The remaining activity can be entirely removed with the addition of EDTA, indicating the dependence of this residual activity on the magnesium or manganese coming from the culture. Based on the data

2+ of modified Lineweaver–Burk plot depicted in Figure 4-7, the Km values for Mg and

Mn2+ were detected as 213 μM and 667 μM respectively. The availability and lower

2+ Km value led to the selection of Mg in our work.

2+ 2+ Figure 4- 7. Km value of Mg (a) and Mn (b) determined with Lineweaver–Burk plot.

Temperature

The manipulation of the reaction temperature is a simple and effective way to improve the performance of enzyme. Generally, higher temperature translates to higher activity. According to molecular dynamics, a higher temperature increases the kinetic energy of molecules and also increases the probability of achieving the

activation energy of reaction when molecules collide with each other [143]. The increase in temperature also enhances the number of collisions of enzyme and substrate per unit time, resulting in the rise of the reaction rate. A 50-100% increase was observed in the enzyme activity with a 10 degree C increase in temperature[144].

However, as fragile protein, every enzyme has an optimum temperature. When the temperature is higher than the optimum value, the three-dimensional shape of enzyme becomes thermodynamically unstable, resulting in loss of activity [145][82]. The effect of temperature on enzyme activity was represented in Figure 4-8. The figure showed that ICDH-cl had an optimum temperature around 45 °C. At this temperature, the specific activity was 135 U, exhibiting a three times increase from the room temperature . A decline of specific activity was observed for the temperatures beyond

45 °C and a cooling treatment may be necessary for the performance of the enzyme, as the temperature of flue gas could be as high as 200 °C.

160

140 /mg)

120 U

（ 100

20 Specific activity activity Specific 0 20 25 30 35 40 45 50 55 Temperature ( °C )

Figure 4- 8. The effect of temperature on enzyme specific activity. 72

Ionic strength

The ionic strength is another important parameter to control the enzyme activity. As the ionic composition of the medium could influence both the binding of charged substrates to enzymes and the movement of charged groups to the active site of the enzyme, the apparent activity of the enzyme may vary with various ionic strengths[146]. In order to evaluate the effect of ionic strength on the enzyme decarboxylation activity, different concentrations of sodium chloride were added to

+ the standard working condition (1 mM isocitrate, 0.2 mM NADP , 4 mM MgCl2） and the new ionic strength was calculated with the following equation

1 퐼 = ∑푛 푐 푧2 (4.2) 2 푖=0 푖 푖

Where ci is the molar concentration of ion i, zi is the charge number of that ion.

The apparent enzyme activity under various ionic strength was depicted in Figure 4-9.

In this study, ICDH-cl obtained the maximum activity when the ionic strength was between 200 mM and 400 mM, exhibiting lower activity for both lower or higher ionic strengths. Before reaching the optimum ionic strength, increasing ionic concentration could assist the charged substrate moving into the active site of the enzyme. However, further increase the iconic strength beyond the optimum may result in the denaturing of enzyme and therefore loss of activity.

50 /mg)

U 40 （

10 Specific activity activity Specific

0 0 200 400 600 800 1000 1200 Ionic strength (mM)

Figure 4- 9. The effect of ionic strength on enzyme specific activity.

Concentration of various cations

During the carbon capture and release process, various cations may be introduced into the system through multiple pathways including pH adjustment, pipeline corrosion, water dilution, etc. The sensitivity of enzymes on these cations varies from species to species. To avoid the potential enzyme activity loss related with cations, remaining activity of enzyme with various cations was studied and shown in Table 4-2.

Table 4- 2. The effect of cation on ICDH activity

Cation Concentration (M)a Remaining activities (%)

None 20

Mg2+ 0.05 100

0.2 101

Mn2+ 0.05 102

0.2 100

Ca2+ 0.05 7

0.2 0

Ni2+ 0.05 0

0.2 0

Cu2+ 0.05 25

0.2 4

Na+ 0.05 95

0.2 55

K+ 0.05 90

0.2 50

Li+ 0.05 82

0.2 35 a Additional cation concentration beyond 40 mM Mg2+

For table 4-2, Mg2+ was considered as a necessary cation and 40 mM of Mg2+ was maintained in all the groups. No further activation effect was observed with an additional supply of Mg2+ and Mn2+, but the introduction of other cations may result in a decrease in enzyme activity. Ca2+ and Ni2+ possessed the highest inhibition effect on enzyme activity as only 50 mM of these cations could completely inhibit the enzyme activity. Compared with divalent cations, monovalent cations demonstrated less impact on enzyme activity. About 10% loss of enzyme activity was observed with

50 mM of Na+ and the inhibition became severe with higher Na+ concentration. To check the reversibility of denature, Ca2+ or Ni2+ was separated from the deactivated enzyme, and the enzyme was reactivated with Mg2+, restoring 50-60% of the activity.

The results suggested that the severe decline in enzyme activity with divalent cations was due to their substitution of the Mg2+ bound to the enzyme [147][148].

The negative effect of cation on enzyme activity brings concerns of cation control in the carbon capture and release cycle. During this cycle, adjustment of pH is realized by adding alkali, inevitably introducing cations. Alkalis with a monovalent cation such as sodium hydroxide are preferred due to the lower inhibition effect. However, cations such as Ca2+ and Ni2+ widely exist in enzyme purification and immobilization process, and the concentration of those needs to be carefully controlled.

Harmful gas

In addition to temperature and pH, the various anions and cations come from the flue gas would also affect the enzyme stability. The Table 4-3 lists the traditional components of the flue gas from the different sources [149-152]. 76

Table 4- 3. Comparison of flue gas quality of various fuels

Chemical species Natural gas Fuel oil Coal

Nitrogen (N2) 78-80% 78-80% 78-80%

Carbon dioxide (CO2) 10-12% 12-14% 10.6%

Oxygen (O2) 2-3% 2-6% 7%

Carbon monoxide (CO) 70-110 ppm 2000-5000 ppm

Nitrogen dioxide (NO2) ~1%

Nitric oxide (NO) ~1%

Sulphur dioxide (SO2) 2000-4000 ppm

Based on data of Table 4-3, the levels of conventional components (N2, CO2 and O2) are almost constant among the different sources of flue gas, but the portions of NOx and SOx are significantly higher in the flue gas from coal. These impurities are formed during the combustion of coal and their concentration highly depends on the combustion process and the native composition of coal. Before the flue gas reaches the enzyme, the following reaction could occur when the NOx and SOx contact the water of the solvent.

2SO2 + 2H2O + O2 = 2H2SO4 (4.3)

And 77

2NO2 + H2O + O2 = HNO2 + HNO3 (4.4)

To investigate the enzyme tolerance against these impurities, the enzyme activity was detected and listed in Table 4-4 with various concentrations of magnesium nitrate and magnesium sulfate in the system.

Table 4- 4. The effect of cation on ICDH activity

Anion Concentration (M) Remaining activities (%)

None 100

− NO3 0.01 100

0.05 100

0.1 91

0.2 88

2− SO4 0.01 100

0.05 90

0.1 84

0.2 69

- 2- In this study, magnesium was chosen as associated cation with NO3 and SO4 ，due to

- its nontoxicity to enzyme activity. The effect of low concentration of NO3 on enzyme activity was not obvious and the enzyme lost 12% of activity when the concentration

2- was 100 mM. SO4 , on the other hand, affected the enzyme activity more severely:

2- Only 50 mM of SO4 resulted in 10 percentage loss of enzyme activity and a loss of

2- 2- - 31% with 200 mM of SO4 . Since the more obvious influence of SO4 than NO3 on enzyme activity was observed in this study, the anion may affect the enzyme activity by changing the ionic strength. In practice, the influence of sulfate and nitrate would become more severe when combined with the effect of cations.

Besides the ionic strength, the existence of sulfate and nitrate could also cause other problems. Primarily, the impurity would dissolve in water and continuously lower the system pH. Although the CCS process may be conducted in a buffer system, the additional impurity would decrease the buffer capacity and increase the cost.

Moreover, as strong reducing agents, carbon monoxide and nitric oxide can directly denature the enzyme if lack of pretreatment.

In summary, the native ICDH-cl showed a poor resistant to both cations and anions originated from saline water or the impurity of flue gas. The installation of proper monitoring and controlling process can alleviate the issue. The recent development of molecular biology may also benefit the enzyme stability against the species ions by modifying the enzyme amino acid consequence.

4.3.4 Optimum substrate for carboxylation reaction

For carboxylation reactions, enzymes may prefer different carbon species as the

- primary substrate. While HCO3 was proven to the primary substrate for propiony1-CoA carboxylase [153], dissolved gas phase carbon dioxide was believed as the dominated substrate for most carboxylases originated from the yeast [154-156].

Since the carbon species exhibits self-evolution in the liquid solution with a higher rate than most carboxylation reactions, the enzyme preference on the carbon species is hard to detect and therefore neglected in most experiments. For example, as discussed in the previous chapter regarding the reaction thermodynamics, CO2 (tot) was introduced as the substrate. However, the neglection may cause high uncertainty during large-scale carbon capture, when the carboxylation reaction is conducted with high intensity. The selectivity of carbon species would significantly affect the reaction rate since carbon equilibrium may not be completed before the contact of enzyme and carbon species. In this section, the selectivity of ICDH on carbon species was studied for efficient carbon capture.

When the pH of the solution is between 5 to 7, three forms of carbon dioxide (CO2,

- H2CO3 and HCO3 ) exist in equilibrium according to the following reaction:

+ - CO2 + H2O ↔ H2CO3 ↔ H + HCO3 (4.5)

While the second ionization reaction is almost instantaneous, the rate of first hydration reaction is comparatively low. The ionization constant could be expressed as following:

[HCO3 -][H ] K1  (4.6) [H 2CO3 ]

-4 At room temperature, the value of K1 is about 3.0 X10 M, so the concentration of

H2CO3 can be neglected at our pH range and the equation (4.5) was simplified as

- CO2 ↔ HCO3 (4.7)

And the concentration of dissolved carbon dioxide was described as

푑[퐶푂 ] 2 = 푘[HC푂−] − 푘′[퐶푂 ] (4.8) 푑푡 3 2

Where 푘 = 0.015 sec-1 and 푘′ = 0.015 sec-1 at pH 6.5 and 푘 = 0.002 sec-1 and 푘′ = 0.015 sec-1 at pH 7.5.

- The value of HCO3 /CO2 ratio is calculated to be about 1 at pH 6.5 and 8 at pH 7.5.

Considering the reaction (4.7) is a first order reaction, the integrated rate equation for the approach to equilibrium from one side is

퐶푒 (푘 + 푘′ )푡 = 2.3log [( ] (4.9) 퐶푒−퐶푡

Where Ce is the concentration of product under equilibrium and Ct is the concentration of product at time t.

Based on the equation 4.9, the times required to reach 50% of equilibrium are 23 and

40 seconds under pH 6.5 and 7.5 respectively and times to reach 90% of equilibrium are 77 and 138 seconds.

The selectivity of carboxylation reaction with carbon species was revealed through the comparison of carboxylation reaction rate profiles with pre-equilibrated and

- un-equilibrated HCO3 -CO2 solutions (Figure 4-10). Although the total concentration of carbon species in these two kinds of solutions was the same, the ratio of CO2 and

- HCO3 decreased during the experiment in un-equilibrated solution but appeared constant in pre-equilibrated due to the carbon self-evolution. To get unequivocal kinetics, the reaction condition need fulfill three criterions：i) the window of observation is only several mins after the initiation of the reaction as the carbon evolution is not completed; ii) the concentration of carbon species must be the limited substrate to show the sensitive of reaction on the concentration of gas phase of CO2; 81

iii) the rate of carbon species consumption must all be small to neglect the change of total carbon concentration.

Carboxylation reaction progress was compared in Figure 4-6, taking the concentration of NADPH as an indicator. At both pH 6.5 and 7.5, the pre-equilibrated solution provided a constant reaction rate for complete two minutes of the reaction time, 35

μM and 64 μM of carbon was fixed respectively. In contrast to pre-equilibrated solution, a decline of the carboxylation rate was observed for un-equilibrated groups at two pHs. At pH 7.5, the reaction rate in the first min was 32 μM/min, much higher than that of pre-equilibrated (17.5 μM/min). In the second min, the reaction rate decreased to 16 uM/min, close to the equilibrium group. The difference of kinetics between pre-equilibrated and non-equilibrated was more obvious at pH 6.5: the reaction rate was 60 μM/min for the non-equilibrated solution in the first 30 seconds, almost twice of the rate for equilibrium group. However, the reaction rate decreased quickly and almost equalled the equilibrium group’s rates after one minute.

Figure 4- 10. The progress of carboxylation reaction catalyzed by ICDH with different carbon resource under pH 7.5(a) and pH 6.5(b). - (Dash line: pre-equilibrated HCO3 -CO2 solution; Solid line: bubbling the CO2 into buffer solution without equilibrium) These results indicated that dissolved carbon dioxide was the primary substrate for the carboxylation reaction catalyzed by ICDH. A higher reaction rate was obtained at pH

6.5 than that at pH 7.5, as more carbon exist in the form of dissolved gas phase to conduct the carboxylation reactions at pH 6.5. Moreover, for the un-equilibrium group, the reaction rate decrease was in accord with the conversion of dissolved gas carbon

- dioxide to HCO3 . At pH 6.5, the highest reaction rate was detected in the beginning when most of conversion did not happen. With time, gaseous CO2 began to evolve to

- HCO3 , lowering the concentration of gaseous CO2 and resulting in the decrease of the reaction rate. After two mins, the carbon species equilibrated and the reaction rate became constant and equaled that of pre-equilibrium group. At pH 7.5, the time required for carbon species to achieve equilibrium was longer than that at pH 6.5, representing a longer gap of reaction rate between the two groups.

In order to further confirm the assumption, carbon anhydrase (CA) was introduced to change the evolution profile of carbon species in solution. Acting as a catalyst, CA could make the reaction (4.7) reach the equilibrium almost instantaneously [157]. The effects of CA on the carboxylation reaction were described in Figure 4-11.

220

200

180

M) μ

160

[NADPH] ( [NADPH] 140

120

100 0 0.5 1 1.5 2 Time (min)

Figure 4- 11. The progress of carboxylation reaction catalyzed by ICDH under pH 6.5 (Solid line: no CA; Dash line: adding CA at 15 second).

In Figure 4-7, bubbled carbon dioxide was supplied as substrate and CA was added into the system at 15 second after the reaction start. After adding CA, the reaction rate

- decreased instantaneously with the conversion of gas phase carbon dioxide to HCO3 and the reaction rate became constant due to the equilibration of carbon species after another 30 seconds. Without CA, the equilibrium was achieved at 1.5 mins, much later than the CA group.

The preferred utilization of gas phase carbon dioxide was originated with the mechanism of the enzyme reaction. By labeling the single tritium atom in 2-oxo[3-3H] glutarate, Dalziel found that during the first step of carboxylation, enzyme bound and enolated the 2-oxoglutarate to form carbanion intermediate. In this step, Mg2+ and

NADPH were necessary to maintain the structure of intermediate. In the following

step, the carbanion intermediate was further converted to isocitrate by the selectively electrophilic attack of the gas phase CO2 instead of other carbon species [158].

This preference could provide guidance for improving the carbon fixation. First of all, locating the enzyme just close to the interface between flue gas and reaction solution can be helpful since more gas phase approaches the enzyme before its evolution.

Many membranes and foam bioreactors providing high mass transfer of carbon dioxide between gas and solution phase are suitable for this propose [159-161].

However, the high mass flux of carbon dioxide may bring severe shear stress [162], which negative effects on the enzyme stability. Also, this quick contact of the enzyme with flue gas requires stricter pretreatment of the toxic compounds of flue gas since there is less solution barrier between the gas and enzyme. Secondly, the carbon capture would benefit from addition of CA to accelerate the reversible conversion of

- dissolved CO2 and HCO3 . When the carboxylation reaction rate is much higher than

- the rate of conversion HCO3 to CO2, the local dissolved CO2 will deplete quickly, and therefore the conversion rate of carbon species limits the apparent carboxylation reactions. With CA, the local dissolved carbon dioxide can be replenished from

- HCO3 immediately to supply the carboxylation reactions. In the end, since the selectivity of carbon species for the enzyme is related to the enzyme structure, modification of amino acid of the enzyme through genetic method may change this selectivity of carbon species to improve overall carboxylation performance.

4.3.5 Gas phase carbon dioxide capture

When gas phase carbon dioxide is utilized as the substrate, the gas phase carbon dioxide needs to be transferred from gas phase to liquid phase forming dissolved carbon dioxide, then the dissolved CO2 and its hydration product can diffuse to the location of enzyme molecules. The carbon dioxide captures with different carbon dioxide concentrations were conducted and the rates were depicted in Figure 4-12.

4.8

4.6

4.4

[NADPH] mM) ( [NADPH] 4.2

4 0 10 20 30 40 50 Time (min)

Figure 4- 12. Carbon dioxide capture rate with different concentration of carbon dioxide ( ：15%; : 30%; : 60% ; : 100%).

After 40 mins, the reaction of all the samples reached equilibrium, representing a flat kinetic curve. The solution absorbed different amount of carbon dioxide with various carbon dioxide concentration. For pure carbon dioxide, 0.93 mM carbon was captured with 5 mM cofactor supplied, while the amount of carbon captured decreased sharply with lower carbon dioxide concentration. With 15% carbon dioxide, only 0.47 mM

carbon was captured. According to Henry’s law, low concentration of carbon dioxide brought low solubility of carbon dioxide, resulted in an unfavourability for the carbon capture equilibrium.

In the figure, a linear and sharp decline curve was observed at the beginning of the reaction with 30% - 100% carbon dioxide. In this period, the mass transfer of carbon dioxide from gas to liquid did not limit the overall carbon capture rate. With 15% carbon dioxide, however, the mass transfer limitation of carbon dioxide became obvious, as a smaller slope of the decline curve was observed in the initial 10 mins.

Theoretically, the mass transfer of carbon dioxide is related to transfer area (Ad), equilibrium point concentration (Ce), and mass transfer constant (KL) (equation 4.10).

As the carbon dioxide concentration affected the Ce, a lower mass transfer corresponded with lower carbon dioxide concentration.

푣푑 (푑퐶/푑푡) = 푘퐿퐴푑(퐶푒 − 퐶) (4.10)

To further reveal the effect of carbon dioxide transfer issue, the carbon capture curve with carbon dioxide covered the surface of solvent was shown in Figure 4-13.

4.9

4.8

4.7 [NADPH] mM) ( [NADPH] 4.6

4.5 0 10 20 30 40 50 Time (min)

Figure 4- 13. Carbon dioxide capture rate with different surface area covered with carbon dioxide ( : 2 cm2; : 1.5 cm2; : 1 cm2 ; : 0.5 cm2).

As the final equilibrium point of carbon capture was reached with 15% carbon dioxide, 0.47 mM carbon was captured for all the samples, in accord with the results of bubbling gas method. The mass transfer rates of carbon dioxide, controlled by the interfacial area, determined the initial rates of carbon capture. Based on the results, samples that had the interfacial area between 1 cm2 and 2 cm2 possessed a similar carbon capture rate, but a lower initial rate was observed for the sample with 0.5 cm2 interfacial area, revealing a mass transfer limited phenomenon. Only 0.7 mM carbon dioxide was theoretically transferred with 0.5 cm2 interfacial area in 5 mins, as the value of KL and Ce 0.0252 cm/s and 5.25 mM respectively. Regarding the Km value of ICDH (~1 mM), the concentration of carbon dioxide is insufficient to realize the

maximum reaction rate. With more interfacial area, more dissolved carbon dioxide was accumulated in the solvent and accelerated the reaction rate.

4.4 Conclusions

In this section, ICDHs from different species were screened with their pH profile and huge differences between their kinetics were observed. With better specific activity and Km values, ICDH from Chlorobium limicola is the comprising candidate for CCS.

Since the kinetics difference comes from their varied molecular structures, modification of the enzyme structure may be an approach to enhancing the specific activity.

Gas phase carbon dioxide is believed to be the primary substrate for ICDH reaction because it can provide higher reaction rate then bicarbonate. The preference of carbon species can bring clue about the enzyme reaction mechanism.

In gas capture experiment, both partial pressure of carbon dioxide and interfacial area affected the overall carbon capture rate and external mass transfer can be the limiting factor to the high carbon capture rate.

Chapter 5. Carbon Dioxide Conversion with 2,3-dihydroxybenoic

Acid Decarboxylase

5.1 Introduction

The 2,3-dihydroxybenoic acid decarboxylase (2,3-DHBA) (EC 4.1.1.46, DHBD) was discovered by its decarboxylation activity with aromatic nucleuses such as

2,3-dihydroxybenzoic acid and salicylic acid [163-165]. After that, multiple functions were reported for the enzymes in various species: for example, 2,3-DHBA functioned as a catalysis of conversion of indole to catechol in Aspergillus [166]; the enzyme was also found to participate in the metabolism of anthranilic acid in Trichosporon [167].

Recent research concluded that the decarboxylation reaction can be reversed, in the detoxification process of living cells. Although this enzyme was well known for its decarboxylation activities, limited data was available for the reversible reactions.

Wuensch reported the application of this enzyme in the regioselectivity carboxylation of phenols and hydroxystyrene derivatives [168]. After screening different substrate such as phenol, m-aminophenol, catechol, resorcinol, and 1-2-dihydroxybenzene,

Wuensch concluded that the carboxylation rate was determined by the chemical structure of the substrates.

Since both decarboxylation and carboxylation exist in 2, 3-DHBA, this enzyme indicates its potential application in carbon capture and conversion. Considering the atom economy, phenol could be the best substrate for carbon dioxide fixation and 91

conversion. With 2,3-DHBA, the salicylic acid can be synthesized from phenol and carbon dioxide in mild condition, replacing the existing complicated and costly process which requires two separate processes of acid-base catalysis [169][170]. In this chapter, both carboxylation and decarboxylation activity of 2,3-DHBA was studied to evaluate its potential application of carbon capture and conversion.

5.2 Material and methods

5.2.1 Materials

Phenol and salicylic acid were purchased from Sigma Chemical Co. (St. Louis, USA).

Others see the section 3.2.1

5.2.2 Assay of total protein concentration

See the section 4.2.2.

5.2.3 Expression and purification of the enzyme

Sililar to ICDH, recombinant cells were cultured in Luria–Bertani medium at 37 °C,

250 rpm and 50 ug/ml kanamycin was added to prevent contamination. After another

12 hours, 1 mM isopropyl thio-b-D-galactoside (IPTG) was added to induce the expression of ICDH, and the temperature stayed at 28 °C. After 12 hours, cells were harvested from a total 500 ml culture with centrifugation and washed twice with 0.1

M potassium phosphate buffer (pH 7.2), then resuspended in the same buffer. The suspended cells were disrupted by sonication on ice and centrifuged for 15 mins at 15

000 Xg to remove cell debris. The total protein concentration of cell extract was detected by Bradford method.

The expressed ICDH was purified through the Histidine Tag Column (His GraviTrap, bed size 7 mm × 25 mm, bed volume 1ml). 10 ml of binding buffer pH 7.4 contained

50 mM sodium phosphate, 300 mM NaCl was added to the column, followed by 10 ml of supernatant of the cell extract. Then 10 ml of wash buffer pH 7.4 contained 50 mM sodium phosphate, 300 mM NaCl and 20 mM imidazole was added to remove the untargeted protein. In the end, 5 ml of elution buffer contained 50 mM sodium phosphate, 300 mM NaCl, 200 mM imidazole was used to elute the 2,3-DHBA from the column. The enzyme was desalted and kept in pH 7.0 phoshpate buffer at

4 °C.

5.2.4 Enzyme activity assay

The activity of enzyme was detected by monitoring the changing of the concentration of salicylic acid for the prepaid time. In the decarboxylation reaction, the assay mixture contained 30 mM salicylic acid and free enzyme in 1 mL of buffer. With prepaid time, the mixture was centrifuged quickly and the decrease of salicylic acid in the supernatant was detected by absorbance at 310 nm with the spectrophotometer and one unit of decarboxylation activity was defined as 1 μmol of salicylic acid conversed per min. The carboxylation activity was detected in the same way except the mixture was composed of 20 mM phenol, 1 M KHCO3 and one unit of carboxylation activity was defined as 1 μmol of salicylic acid formed per min.

5.2.5 Carbon capture experiment

The amount of carbon dioxide captured was determined by detecting the increase of concentration of salicylic acid. Especially, the concentration of phenol and KHCO3 were fixed at 2.5 and 20 mM respectively and the enzyme solution was added to make the final concentration of protein as 1.5 mg/ml with a total volume of 10 ml in phosphate buffer. The reaction was monitored by retrieving 100 μl of the reaction solution for analysis of salicylic acid concentration (A310) after centrifugation.

5.3 Results and discussion

5.3.1 Purification of 2,3-DHBA

For 50 ml of culture medium, about 18 mg total crude protein with 0.93 U decarboxylation activity remained in the supernatant of cell extract. The target protein then was purified by Histidine Tag Column. During the purification, the target enzyme was initially attracted to the column with the affinity binding and then eluted from the column with 200 mM imidazole solution. After elution, 8.51 mg protein with

0.39 U decarboxylation activity was recovered. The recovery rate of enzyme activity was about 40%. The enzyme denaturation during the purification process [171] and failure to attach to the Histidine Tag Column may have contributed to the loss of enzyme activity.

In the study, the elution of 2,3-DHBA required 200 mM imidazole solution, much higher than ICDH ( 80 mM). Generally speaking, a stronger binding between the

histidine tag and the nickel of the column requires higher imidazole concentration

[172]. Although the sequence with same number of histidine was inserted into the origiral sequences of ICDH and 2,3-DHBA, the 3D structure of 2,3-DHBA may bring more interaction surface between the binding site( histidine) of 2,3-DHBA and the nickel of column [173].

5.3.2 The effect of pH on enzyme activity

To investigate the effect of pH on the activity of 2,3-DHBA, both carboxylation and decarboxylation experiments were conducted with phosphate buffer pH from 4 to 9 and the results were indicated in Figure 5-1.

a 2000

1500

1000

500 Specific activity activity U/g) ( Specific 0 3 4 5 6 7 8 9 10 pH

Figure 5- 1. The effect of pH on the specific activity of 2,3-DHBA (a. decarboxylation; b. carboxylation). According to the above data, 2,3-DHBA preferred the decarboxylation activity for the full pH range. The maximum decarboxylation activity was 1733 U/g at pH 6, a thousand times more than the maximum carboxylation activity, which was only 1.21

U/g at pH 7. Different from ICDH, the optimum pHs of carboxylation and decarboxylation were close to each other and enzyme activity decreased sharply when the pH was adjusted far from the optimum value.

Considering the pH profile, the native 2,3-DHBA has several disadvantages for application in carbon capture and release cycle. Primarily, the specific carboxylation activity is too low to afford high intensive carbon capture. The specific activity of

2,3-DHBA is thousand times less than that of ICDH, resulting in much higher enzyme loading level and longer reaction time to achieve the same carboxylation rate. Also, the 2,3-DHBA requires neutral pH to maximize the carboxylation rate, but the dissolved carbon species would lower the pH and therefore reduce the reaction rate.

Finally, the close optimum pHs of carboxylation and decarboxylation makes controlling the reaction direction by shifting pH, difficult. Instead of carbon capture and release, pure carbon dioxide conversion may be more suitable for 2,3-DHBA, as the carboxylation product (salicylic acid) is expensive than the substrate (phenol).

5.3.3 The effect of temperature on enzyme carboxylation activity

To evaluate the effect of temperature on enzyme activity, the carboxylation was conducted with various temperatures (Figure 5-2). According to the figure, the optimum temperature for carboxylation activity was 30 °C, much lower than the typical value (~ 50-60 °C) for most enzyme reactions, indicating the poor thermo-resistance of this enzyme. As a toxic substrate, phenol can denature the protein by interrupting its structure, and this interruption becomes severe with high temperature. In this study, the enzyme lost half of its activity at 40 °C and entirely denatured at the temperature higher than 50 °C. To maintain high carboxylation activity, the flue gas needs to be pretreated for cooling.

1.5

1.2

0.9

0.6

0.3 Specific activity activity U/g) ( Specific 0 0 10 20 30 40 50 60 Temperature (°C)

Figure 5- 2. The effect of temperature on carboxylation activity of 2,3-DHBA.

5.3.4 Carbon capture with 2,3-DHBA

With the enzyme kinetics data, carbon capture experiment was conducted at various pHs and temperatures to evaluate the feasibility of 2,3-DHBA enzyme in carbon capture and release strategy. During the CCS cycle, fluctuation of the system pH was expected during the carbon dioxide dissolving, so the effect of pH on the performance of carboxylation was first studied and the results were depicted in Figure 5-3. After 12 hours inculcation, about 10~12% of phenol (20 mM) was converted to salicylic acid with carbon dioxide fixation at neutral pH. Without strong sensitivity of pH, the same amount of carbon was captured with pH 6 to 8. At pH 9, the conversion rate decreased to 4.5% and the lowest conversion rate (1%) was observed at pH 4. The low conversion rate under extreme pH was thought to be related to enzyme activity loss

since the reaction solution became turbid after half hour of incubation.

Figure 5- 3. The effect of pH on the conversion rate of 2,3-DHBA reaction.

To further determine the influence of temperature on carbon capture, the reaction was conducted at a constant pH of 6 but at various temperatures (Figure 5-4). The conversion rate remained constant when the reaction temperature was lower than

40 °C, in accordance with the enzyme activity data. The decline of conversion rate occurred at 50 °C as the specific activity of the enzyme was strongly inhibited by the temperature.

） 12

% （ 9

3 Conversion rate rate Conversion

0 10 20 30 40 50 Temperature （°C）

Figure 5- 4. The effect of temperature on the conversion rate of 2,3-DHBA reaction.

According to previous data, the amount of carbon capture with fixed substrate concentration, was hardly improved with temperature or pH shifting. In other words, the carbon capture and release cycle cannot be realized with 2, 3-DHBA reaction through controlling the reaction temperature or pH. Manipulation of the partial pressure of carbon dioxide may be an alternative method for controlling CCS process, although the low carboxylation activity still impedes the large-scale application. The conversion rate doesn’t vary much with temperature and pH, however, this might benefit the pure carbon dioxide conversion, since there is no need for a strict control of reaction temperature and pH. In the end, the conversion rate with different phenol concentrations was studied and summarized in Figure 5-5. With 10 mM phenol and 1

M KHCO3 applied, 1.67 mM salicylic acid was synthesized after 12 hours inculcation.

A quick decline of conversion rate was observed with increased phenol concentration, 100

indicating a strong equilibrium limitation. At high substrate concentration, for example, 60 mM phenol, the conversion rate was only 4.18%.

Conversion rate (%)rate Conversion 4

0 0 10 20 30 40 50 60 70 [Phenol] (mM)

Figure 5- 5. The conversion rate of phenol with different substrate concentration.

5.4 Conclusions

As an enzyme has both carboxylation and decarboxylation activity, 2,3-DHBA can catalyze the reversible conversion between salicylic acid and phenol. About 40% of the enzyme activity was recovered with Histidine Tag purification.

The 2,3-DHBA preferred the decarboxylation activity at a full pH range, as the carboxylation activity was thousands of times less than the decarboxylation. Both decarboxylation and carboxylation activity obtained their maximum value at neutral pH and was inhibited strongly with extreme pH. A comparably low temperature

(30°C) was required for the 2,3-DHBA to maximize its activity, which may be related to the toxicity of substrate. 101

The insensitivity of conversion rate to temperature and pH made the enzyme unsuitable for carbon capture and release strategy. However, this enzyme provided an alternative synthesis pathway of salicylic acid with high energy efficiency. After 12 hours incubation, 10 % of 20 mM phenol was converted to salicylic acid in the mild conditions.

102

Chapter 6. Cascade Enzymatic Reaction System to Improve Carbon

Capture Capacity

6.1 Introduction

Carbon capture capacity, defined as the amount of carbon dioxide capture by a unit concentration of substrate, is one of these critical factors indicating the industrial feasibility of CCS. Generally, biocatalytic reactions offer lower carbon capture capacity than chemical reaction with a high concentration of substrate. The basic nature of the chemical absorbents allows chemical carbon capture reactions to be conducted almost entirely, while most biocatalytic carbon capture reactions reach a much lower degree of conversion as limited by reaction equilibrium. Considering the native carbon capture and conversion have been achieved through a cascade of reaction in biology, we assume here that cascade reactions conducted simultaneously in a homogeneous solution can further the conversion of substrate, improving the overall carbon capture capacity. Indeed, cascade reaction has been successfully demonstrated for conversion of CO2 to methanol, both enzymatically [20] [174] and chemically

[175].

In nature, isocitrate dehydrogenase (ICDH) and aconitase are the two cascade enzymes exist in reductive tricarboxylic acid (RTCA) for carbon fixation. Carbon dioxide is first fixed with ketoglutarate to form isocitrate, catalyzed by ICDH, and then further converted to citrate by aconitase. The previous carbon fixation experiment of the solo enzyme reveals a poor carbon capture capacity. As the aspect of reaction equilibrium,

103

the cooperation of ICDH and aconitase is thought to be critical for efficient carbon capture [176]. In this study, the improvement of carbon capture with cascade reactions will be evaluated with both thermodynamic model and direct experiment.

6.2 Material and methods

6.2.1 Matieral

Ferrous ammonium sulfate, L-cysteine Aconitase from the porcine heart (A5384) were purchased from Sigma Chemical Co. (St. Louis, MO, USA). See the section 3.2.1 for other materials.

6.2.2 Assay activity of aconitase

Aconitase was activated as a method of Sigma. Dissolve 15 mg of the enzyme in 2.0 ml of 0.1 M Tris buffer, pH 7.4, then add 50 µl of 1 mM ferrous ammonium sulfate (0.39 mg/ml in water) and 0.1 ml of 0.05 M L-cysteine (7.90 mg/ml in water, adjusted to pH

7.4 with NaOH). Incubate for one hour at 0 °C. The aconitase was now activated and ready for use.

For aconitase, one unit of activity is defined as 1 μmol of aconitate produced from isocitrate (forward) or citrate (reverse) per min. The activity assay reaction mixture contained 50 mM isocitrate or citrate and 10 μl enzyme solutions in 1 ml of buffer solution. The reaction solution was monitored continuously with a UV-Vis spectrometer (50 Bio, Varian) for changes in the concentration of aconitate via absorbance at 240 nm.

104

The effect of temperature, ionic strength, cofactor, different cations and impurity of flue gas on the enzyme activity was analyzed with reaction kinetics (see detail in

4.3.2.)

6.2.3 Determination of Michaelis constants of ICDH reaction

Determination of the Michaelis constants (Km) was conducted through

Lineweaver-Burk double reciprocal plots, similar to 4.2.6.

6.2.4 Carbon capture capacity at different pH

The concentration of carbon dioxide captured was determined by detecting the decrease of concentration of NADPH when the reaction approaches equilibrium at room temperature. Especially, the concentration of ketoglutarate, NADPH and

KHCO3 were fixed at 2.5, 2.5 and 25 mM, respectively. ICDH and aconitase solution was added to make the final concentration of protein as 20 μg/ml and 14 mg/ml with a total volume of 1 ml in phosphate buffer. The reaction was monitored by retrieving

100 μl of the reaction solution periodically for analysis of NADPH concentration

(A340) after centrifugation.

6.3 Results and discussion

6.3.1 Thermodynamic prediction of Gibbs free energy of reaction

Carbon capture capacity, related to reaction equilibrium, demonstrates how much carbon dioxide could be reacted and captured with a unit concentration of cofactor and ketoglutarate. In the aspect of thermodynamic, Gibbs energy of reaction (ΔGr),

105

representing the difference of the Gibbs free energy of formation (G0) between products and substrates, determines the carbon capture capacity of biochemical reaction. In most case, a high carbon capture capacity can be expected with a negative

ΔGr. For example, the carbon capture capacity of the MEA reaction for carbon capture is 0.3 M for each mole of substrate supply, with the ΔGr of a -8.8 kJ/mol, [177]. Many efforts have been [178] made to enhance the carbon capture capacity of biochemical reactions through lowering the value of ΔGr. With a shifting of pH from 8 to 6, the ΔGr of carbon dioxide to methanol can be decreased from 10.77 to -2.51 kJ/mol [20].

However, the single manipulation of pH or temperature cannot improve the capacity significantly. So, here is our new approach, designing ending product for enzyme systems to offer better thermodynamic driving forces for carboxylation.

The carbon capture capacity of the ICDH reaction is also impeded by the thermodynamic barrier, based on the study of Colman [179]. To achieve high carbon capture capacity, the overall ΔGr need be reduced. In this case, conducting reaction simultaneously with both ICDH and aconitase may benefit, by driving the carboxylation reaction of ICDH further through shifting the concentration balance, since it can readily convert isocitrate to citrate. To verify this, the gap of G0 between products and substrates for ICDH with aconitase was calculated and shown in Figure

6-1, according to Debye-Huckel theory [53].

106

Figure 6- 1. The gap of Gibbs free energy of formation between substrate and product for reaction catalyzed by ICDH and aconitase (data of Gibbs free energy of formation showing in the figure are calculated for substrate and product mixtures, according to the stoichiometry of related reaction).

With the introduction of aconitase, the carbon capture capacity will be enhanced significantly with a smaller value of ΔGr. For single ICDH, the ΔGr is only -0.34 kJ/mol with isocitrate as ending product. By involving aconitase, isocitrate is further converted to citrate, providing additional -6.64 kJ/mol (total -6.98 kJ/mol). With further calculation, the carbon capacity with ICDH can catch up that of prevailing chemical reaction (MEA) system. With 1 M substrate and 0.035 M dissolved carbon dioxide, cascade system can absorb 0.76 M and 1.92 M carbon dioxide at pH 6 and 5 respectively, compared with the 1.2 M carbon capture capacity of MEA.

6.3.2 Carbon capture with multiple enzyme system

In nature, carbon capture through RTCA cycle is achieved by cascade enzyme reaction [181]. To verify the theoretical calculation and also mimic the nature process,

107

carbon capture reaction with cascade enzymes system was conducted and the results were shown in Figure 6-2.

1.6

1.2 ] ] (mM)

+ 0.8 [NADP 0.4

0 6 7 8 9 pH

Figure 6- 2. The improvement of carbon captures capacity of ICDH system with aconitase under pH 6-9 (black: ICDH with aconitase; grey: ICDH only).

The improvement of carbon capture capacity with aconitase was evident for a full pH range. At pH 6, about 1.26 mM carbon was fixed for multiple enzymes system, 2.5 times more than single ICDH group (0.38 mM). The enhancements of capacity were also observed in other pHs, which revealed 1.7 and 1.8 times increase of carbon capture at pH 7 and 8 respectively.

Carbon capture capacity decreased sharply with the elevating of reaction pH in the cascade system, like single enzyme reaction. In practice, the difference of carbon capture capacity with pH shift is required for CCS. Assuming carbon dioxide is fixed at pH 6 and released at pH 9, ICDH with aconitase can treat 1.2 mM carbon dioxide 108

in one cycle, compared with 0.38 mM of the simple enzyme system. In conclusion, the experimental results were in accord with the theoretical prediction.

6.3.3 The effect of pH on aconitase activity

As aconitase significantly benefits the ICDH based carbon capture strategy, study on the kinetics characteristics of this enzyme is meaningful. Aconitase, another essential enzyme in RTCA cycle, catalyzes the conversion of isocitrate and citrate with aconitate as intermediate. Based on Meike’s research, the reaction with aconitate as product is the rate-limited step due to a comparable low catalytic efficiency (kcat/Km)

[182] (Figure 6-3).

Figure 6- 3. The scheme of conversion between citric and isocitric catalyzed by aconitase.

The specific activity of aconitase for both forward (isocitrate as substrate) and reversible (citrate as substrate) reaction was shown in Figure 6-4. The optimum pH of aconitase was 7 and the maximum specific activity was 11.3 U/g (forward) and 5.9

U/g (reversible) respectively. The activity of aconitase stayed stable at pH 6 to 8, while it decreased to 33% at pH 9. The direction of aconitase reaction did not reveal a strong sensitivity to pH, different from ICDH.

109

6 Specific activity ( U/ g) U/ ( activity Specific 3

0 4 5 6 7 8 9 10 pH

Figure 6- 4. The specific activity of aconitase under different pH (white square: isocitrate to citrate; black triangle: citrate to isocitrate).

6.3.4 Michaelis constant of aconitase

The Km values of aconitase reaction were obtained through Lineweaver–Burk plot and shown in Figure 6-5. Based on the plots, the Km values for isocitrate and citrate were

109 μM 487 μM respectively. Compared with ICDH, aconitase requires a higher concentration of isocitrate to maximize its activity. However, as an intermediate of cascade reaction, the too low concentration of isocitrate cannot afford of the maximum reaction rate. And again, due to the comparable low specific activity of aconitase, a high ratio of aconitase and ICDH is needed for efficient carbon capture.

110

Figure 6- 5. The Lineweaver–Burk plot of conversion reaction catalyzed by aconitase.

6.3.5 Parameter optimization for aconitase

Similar to ICDH, the effects of environmental factors such as cofactor, temperature, ionic strength, cations and gas impurity on the activity of aconitase were evaluated to prevent potential enzyme activity loss.

Cofactors concentration

As an enzyme, the aconitase requires specific cofactors to maintain its activity, and the importance of ferrous ammonium sulfate and cysteine on the activation of aconitase were reported widely in references. In this study, the improvement of 111

enzyme activity was also observed with the proper concentration of ferrous ammonium sulfate and cysteine. (Figure 6-6).

0 Control 1 μM Ferrous 100 μM L- 1 μM Ferrous 5 μM Ferrous Specific activity (U/g) activity Specific ammonium cysteine ammonium+ ammonium+ 0.5 100μM L- mM L-cycteine cycteine Cofactor concentration

Figure 6- 6. The improvement of specific activity of aconitase with different cofactors.

According to the Figure 6-6, the specific activity of aconitase was only 2.51 U/g at the control condition and, the addition of 1 μM ferrous ammonium and 100 μM

L-cysteine can increase the activity to 6.59 U/g and 3.55 U/g respectively. The highest specific activity was obtained as 6.59 U/g with 5 μM ferrous ammonium and 0.5 mM

L-cysteine, while no further improvement was observed with the cofactor concentration beyond that.

In the 1950s, Morrison found Fe2+ was an integral component of the aconitase to be active, since the purified no-iron aconitase showed no activity. As Fe3+ was useless for the enzyme, L-cysteine was also believed to have a function as reducing agent. Later

112

research proven that Fe2+ and L-cysteine both formed the [4F-4S] cluster in aconitase as the binding site to the substrate [183-185].

In practice, the requirement of cofactor will harm the feasibility of reaction, due to increase the cost and bring the stability issue. Firstly, maintaining the concentration of

L-cysteine ($ 10000 / ton) is costly; then, the oxides such as oxygen, sulfur dioxide in the system can cause the denature of enzyme by reacting with the reduced form of iron.

Temperature

Similar to ICDH, higher temperature translates to higher activity. As shown in Figure

6-7, the optimum temperature of the aconitase was around 35 °C to 40 °C, while enzyme activity decreased sharply beyond that temperature. As the optimum temperatures of aconitase and ICDH were close to each other, the manipulation of the temperature of cascade system should be accessible. By shifting the room temperature to 40 °C, the improvement of enzyme activity for ICDH and aconitase would be 70% and 40% respectively.

113

Specific activity U/g) ( Specific activity 2

0 15 25 35 45 55 65 Temperature (oC )

Figure 6- 7. The effect of temperature on aconitase activity.

Ionic strength

In this study, the ionic strength was adjusted between 200 -1000 mM with an addition of sodium chloride. Based on date of Figure 6-8, aconitase preferred 200 mM ionic strength to maximize its activity and a decline of activity was observed with higher ionic strength. Compared with ICDH, A more rigid control of the ionic strength was required for aconitase: with 1000 mM ionic strength, aconitase lost 48% of its activity, while the activity of ICDH only dropped 37%.

114

Specific activity (U/g) Specific activity 2

0 0 200 400 600 800 1000 1200 Ionic strength (mM)

Figure 6- 8. The effect of ionic strength on aconitase activity.

Cations concentration

Different kind of cations may also be introduced during the process and their effects on the aconitase activity were studied and summarized in Table 6-1. According to the table, Ca2+, Ni2+ and Cu2+ were strong inhibitors of the aconitase, as 50 mM of them can cause an 80% of enzyme activity loss. Compared with divalent, monovalent such as Na+ and K+ demonstrated less influence on the enzyme activity. The enzyme maintained its activity with 50 mM Na+ and K+, while lost 20% of its activity with

200 mM of them.

The cation control was more complicated in cascade system. While A proper level of

Mg2+ or Mn2+ is necessary to maintain the ICDH activity, the too high concentration of them will cause the denature of aconitase. For example, 200 mM Mn2+ will result a

50% activity loss of acontiae, based on the previous data. Luckily, aconitase can tolerate a low concentration, for example 50 mM, of Mn2+ and Mg2+ without activity

115

loss. In this study, the denature of aconitase caused by Mn2+ was reversible, different from the irreversible deactivation with Ca2+ or Ni2+. As a competitive inhibitor of aconitase, Mn2+can replace the Fe2+ in the enzyme activity center and this inhibitory effect was proven to be important to control the activity of mitochondrion in the brain cell [186][187].

116

Table 6- 1. The effect of cation on enzyme activity

Cation Concentration (M) Remaining activities (%)

None 100

Mg2+ 0.05 100

0.2 90

Mn2+ 0.05 78

0.2 54

Ca2+ 0.05 23

0.2 12

Ni2+ 0.05 26

0.2 14

Cu2+ 0.05 12

0.2 2

Na+ 0.05 100

0.2 80

K+ 0.05 97

0.2 84

Harmful gas

The anions come from the impurity of flue gas also was proven to affect the enzyme activity and their influence was studied and shown in Table 6-2. The obvious

117

inhibition only happened when the concentration of these anions was higher than 0.2

2− − M. The SO4 brought stronger inhibition than NO3 at the same concentration and

2− − that may because SO4 possesses more charge than NO3 . Aconitase showed better resistant to high concentration of sulfate and nitrate than ICDH in the study. Table 6- 2. The effect of sulfate and nitrate on ICDH activity

Anion Concentration (M) Remaining activities (%)

None 100

− NO3 0.01 100

0.05 100

0.1 93

0.2 91

2− SO4 0.01 100

0.05 93

0.1 89

0.2 73

6.3.6 Carbon captures and release cycle with cascade enzymes

In the end, the carbon capture and release strategy was checked by ICDH with/without aconitase. Figure 6-9 revealed the carbon capture amount with time in one cycle and carbon capture and regeneration cycle was conducted by switching the pH between 7 and 8 in an attempt to maintain the enzymes’ activities while demonstrating the cascade reaction cycle. For the data, carbon capture reaction could reach equilibrium within 5

118

min when only ICDH was applied; the reaction with both enzymes applied took a bit longer time to reach equilibrium (∼20 min), mostly due to the slower reaction associated with aconitase, but reached much higher degree of substrate conversion.

After adjusting the pH to 8, decarboxylation started immediately (as indicated by increase in [NADPH]), and eventually reached equilibrium with a similar final concentration of NADPH (∼2.4 mM) at the end of the 30-min reaction. The slight difference between the final concentrations of NADPH is most likely a result of the difference in reaction velocity, as the single step reaction of ICDH was faster. Based on the data shown in Figure 6-9, the cascade reaction could capture 0.62 mM carbon for each reaction cycle conducted (by calculating the difference in the equilibrium concentrations of the cofactor for carboxylation and decarboxylation reactions), that is about twice the amount observed for that one ICDH was applied alone.

Figure 6- 9. Carbon capture and regeneration cycle with (circle) and without (square) aconitase, with pH switched between 7 and 8 for carboxylation and decarboxylation reactions, respectively.

6.4 Conclusions

119

Cascade enzyme reaction with ICDH and aconitase revealed better carbon capture capacity than single enzyme system. Based on theoretical prediction, the ΔGr of reaction decreased to -6.98 kJ/mol from -0.34 kJ/mol by further converting isocitrate to citrate. The experiment result was in accord with prediction: For every mole of cofactor supply, 0.5 mM carbon can be fixed for cascade reactions, much higher than

0.15mM of single enzyme reaction.

Both enzymes possessed the similar optimum temperature and ionic strength making them preferable for cascade reaction. Aconitase showed tolerance for high concentrations of Na+, K+, and Mg2+, but totally inhibited with several divalent such as

Ca2+, Ni2+, Cu2+ and Mn2+. In order to achieve efficiency of cascade reaction, a high ratio of aconitase and ICDH was expected, due to the low specific activity of aconitase.

In practice, high carbon capture capacity can enhance the carbon capture intensity, reducing the reaction space. However, the additional requirements to control the cation and impurity in the system and to maintain the activity of aconitase will increase the cost. Hence, the further improvement of aconitase activity and stability would be critical to the successful application of cascade enzyme systems in large scale.

120

Chapter 7. Immobilization to improve the stability of ICDH

7.1 Introduction

The high activity, selectivity, and specificity of enzyme makes them powerful tool for biotransformation process in the industry. At a large-scale project, high stability and possible reusability are necessary for the economic feasibility of the process. Besides the application of protein engineering to improve the enzyme propensity by modifying the protein consequence, some apparently older techniques, such as immobilization are widely applied as a simple but powerful approach to increase the enzyme activities and reduction of inhibition in the system [188-190]. Considering the various native characterizations of enzyme and different immobilization proposes, hundreds of immobilization protocols were reported, while properly screening of the ideal immobilization material for the target enzyme is still an exciting goal in our project. Based on the enzyme characteristics, two immobilization methods, namely physical adsorption and entrapment were established to improve the ICDH stability.

Mesoporous materials, especially mesoporous silica foam (MSF), possess a highly ordered structure with nanometer pore size, therefore providing high surface area for enzyme adsorption [191][192]. As an immobilization material, mesoporous silicates exhibit high chemical, thermal resistance and inertia, and by controlling the amphiphilic block copolymer and surfactant self-assemblies, the pore size could be controlled between 2nm to 50 nm to meet different requirements [193]. Enzymes can directly bind the MSF through the hydrophilic attraction of the hydroxy group of the

121

surface of MSF, and for advanced application, various active groups modified on the surface of MSF can additionally facilitate enzyme loading with strong or specific binding [194].

Chitosan and chitin are natural polyaminosaccharides which are produced from shells of crustaceans, the exoskeletons of insects and the cell walls of fungi. Chitin is a long chain linear polymer which is composed of β(1→4) linked

2-acetamido-2-deoxy-β-D-glucose units [195]. The structure of chitosan is very similar to chitin, shown in Figure 7-1. As some d-glucosamine units are deacetylated, chitosan is soluble in acidic solutions.

Figure 7- 1. Molecule structure of chitin and chitosan.

As one kind of biopolymer, many unique features such as non-toxicity, good physiological inertness, and lower price make chitosan a suitable candidate for enzyme entrapment material [196-198]. The enzyme immobilization is conducted with two steps. First, dissolving the chitosan powder in dilute organic acid and mixing with

122

target protein; then gelation process would be triggered by adding gelation solvents such as tripolyphosphate (TPP) and the chitosan gels, that can be manufactured in the form of beads, membranes, coatings, capsules, fibers, hollow fibers and sponges.

Various enzymes are reported to be immobilized with this method, such as immobilization of glucose oxidase in sensor [199], amylase in the food industry [200] and carbonic anhydrase in carbon capture [201].

This investigation mainly aimed to obtain stable enzyme system for carbon dioxide capture. Regarding the enzyme stability and retaining activities, the immobilization process would be optimized to improve the performance of immobilized enzyme.

7.2 Material and methods

7.2.1 Material

Chitosan, acetic acid, sodium tripolyphosphate, glutaraldehyde solution and mesostructured silica foam (MSF, 560979) were purchased from Sigma Chemical Co.

(St. Louis, MO, USA) and the properties of MSF was shown as follow:

Form Powder

Cell window size ~ 15 nm

Unit cell size ~ 22 nm

Pore size 2.31 cm3/g pore volume

Surface area Spec. surface area 562 m2/g

See the section 3.2.1 for other materials.

123

7.2.2 Enzyme immobilization with MSF adsorption

The buffer solution was used to dilute the enzyme solution to the final protein concentration as 100 mg/ml, and then MSF (1 g) was added in 1ml solution. After mild shaking at 200 rpm for 5 min, the solid was recovered by centrifugation and washed with 5 ml buffer solution for three times. The immobilized enzyme was kept with MES buffer pH 7.0 at 4 °C.

7.2.3 Enzyme immobilization with chitosan entrapment

Chitosan powder was dissolved in acetic acid (1% v/v) solution and agitated for 1h and ICDH from the porcine heart was dissolved into pH 7 MES buffer to final protein concentration 10 mg/ml. Then ICDH solution was mixed thoroughly with chitosan in prepaid volume and the resulting solution was extruded through 1 ml pipet tip into a 3%

(w/v) sodium tripolyphosphate solution at a vertical distance about 20 cm (Figure 7-2).

After stirring for 10 min, the chitosan bead with immobilized enzyme was filtered from the solution and washed with Mes buffer for three times.

Figure 7- 2. Scheme of ICDH immobilization with chitosan. 124

7.2.4 Glutaraldehyde treatment of immobilized enzyme

In order to increase enzyme loading level and alleviate enzyme leaking, immobilized enzyme was treated with glutaraldehyde to form aggregations. After the initial five mins of adsorption of enzyme on the MSF, glutaraldehyde solution was added into the system with 0.1% w/w. For another five mins, the solid was recovered by centrifugation and washed with 5 ml buffer solution and Tris-Hcl buffer (100 mM Tris, pH 8.0) respectively.

7.2.5 Activity assay of immobilized enzyme

The decarboxylation assay mixture contained 10 mM isocitrate, 2 mM NADP+, 40 mM

MgCl2, 2 mg immobilized enzyme (protein weight) per 1 mL of buffer. The mixture was shaken at room temperature at 200 rpm. Samples were periodically taken and were centrifuged before NADPH concentration was detected by absorbance at 340 nm. One unit of special decarboxylation activity was defined as 1 μmol of NADPH produced per min. The carboxylation assay was conducted with same procedure with the same way except the mixture was composed of 20 mM sodium ketoglutarate, 3 mM NADPH, 40 mM MgCl2, and 35mM NaHCO3.

In order to study the effect of carbonic anhydrase on the carboxylation rate, CA was first dissolved in DI and the solution was added into reaction solution before adding

ICDH to conduct the carboxylation.

125

7.2.6 Detection of leakage behavior

To ascertain the non-leachable character of enzyme adsorbed on the MSF, the leakage investigation was performed in the following way during activity assays: Reaction solution with immobilized enzyme was shaken at room temperature at 200 rpm, and after 5 mins, the mixture was centrifuged and the concentration of protein in supernatant was detected with Bradford method. Then the immobilized enzyme was added into fresh reaction solution for another cycle.

7.2.7 Enzyme stability tests

Free or immobilized ICDH was kept in buffer solution for predetermined time, and the remaining activity of the enzyme for decarboxylation was detected with same methods as described for enzyme activity assay. Several factors such as pH, temperature, stirring rate were applied to the system to indicate the improvement of enzyme stability against these factors.

7.2.8 Determination of michaelis constants and turnover number

Determination of the Michaelis constants (Km) and maximum enzyme velocity (Vmax) was conducted through Lineweaver-Burk double reciprocal plots, and detail in section

4.2.6.

Turnover number (kcat) was calculated with the following equation

Vmax = [E] x kcat (7.1)

Where [E] was total enzyme mole concentration, and the MW of porcine ICDH was

58 kD. 126

7.3 Results and discussion

7.3.1 Optimization of immobilization conditions for MSF adsorption

As both the external and internal pore structure of MSF accessed high surface area for protein adsorption, the initial diffusion of enzyme molecules in bulk solution resulted in a quick adsorption on the external surface of MSF. The completion of adsorption on the internal surface required a longer time as slower internal diffusion. Many references have pointed out that performance of adsorption heavily depended on the pore size of the MSF [202][203]. In this study, the mean diameter of MSF cell window (15 nm) was a little bit larger than the diameter of enzyme molecules (10 nm), which can benefit the diffusion process and prevent too much leakage. The adsorption rate of ICDH on the MSF was plotted in Figure 7-3.

Figure 7- 3. The amount of bound enzyme with MSF with incubation time.

127

For the data, the adsorption process was rapid and was completed in 5 mins. In the first minute, 10.4 mg of enzyme was adsorbed, indicating a quick adsorption on the external surface of the material. Then the adsorption rate decreased to 4 mg/min. The amount of bound enzyme reached 23 mg/ml. During this period, most adsorption seemed to happen on the internal surface of MSF and the adsorption rate was limited by the diffusion of the enzyme into the internal room of MSF. After this stage, the amount of bound enzyme kept constant and the rate was smaller than 0.2 mg/ml/min.

This slight adsorption occurred in the place covered with a small cell window, where it was hard for the enzyme to enter.

According to references, pH value and bulk enzyme concentration were two critical parameters affecting the performance of physical adsorption. In this study, the influence of protein concertation was first studied and reported in Table 7-1.

Table 7- 1. The effect of enzyme concentration on protein yield and activity yield

Initial protein Bound protein Protein yield Expected activity Measured activity Activity yield

(mg /ml) (mg /ml) (%) (U/ ml) (U/ml) (%)

20 18.72 93.60 39.50 5.97 15.10

40 20.43 51.08 43.11 6.31 14.65

80 22.87 28.59. 48.26 6.43 13.32

100 23.51 23.51 49.61 6.54 13.18

200 25.80 12.90 54.44 6.94 12.75

Where expected activity was calculated by multiplying the amount of bond protein with specific activity of free enzyme Protein yield = bound protein/ initial protein Activity yield = measured activity / expected activity

128

As reported in the table, increasing the concentration of initial enzyme resulted in a more enzyme adsorption, but a lower protein yield. With 20 mg /ml enzyme, 18.72 mg of enzyme was bound on the MSF, representing a 93.60% of protein yield. While doubling the initial enzyme concentration can only get 2 mg more enzyme adsorbed, and the protein yield decreased to 51.08%. Increasing enzyme concentration beyond

20 mM showed limited improvement of adsorption, as most of the surface was covered with enzyme molecule. Besides adsorption amount, activity value of immobilization gave a better evaluation of its performance. The maximum measure activity was 6.94 U per gram of MSF with 200 mg/ml initial protein concentration and the value decreased to 5.97 with 20 mg/ml protein. The overall activity yield was around 15% in this study since a large part of the enzyme lost the access to the substrate due to conformation limitation originated from immobilization.

In addition to the initial concentration of enzyme, surface chemistry of a carrier also affects the performance of immobilization. The main binding force for physical adsorption, is the hydrophilic interaction between the amino group of enzyme molecules and the hydroxy group of MSF [204]. In this case, the pH of the solution can influence the surface charge of the enzyme and therefore the enzyme loading level. The effect of pH on enzyme loading level and activity were studied and shown in Figure 7-4. For the full range of pH (5 to 9), around 20 ~ 30 mg enzyme was absorbed on 1 g of MSF within 5 min. The lowest loading level happened at neutral pH and more enzyme was loaded with the material when the enzyme surface

129

possessed more charges under acidic or alkali conditions [205][206]. Although enzyme loading levels preferred extreme pH, the denature of enzyme became more severe during the immobilization, resulting in a lower effective activity. Based on the specific activity the optimum pH was 8 and 6.54 U specific activity was achieved with one gram of MSF.

8 35 7 30 6 5 25 4 3 20 2 15

1 Special ectivty U/g) ( Specialectivty 0 10 mg/g) ( Laoding Enzyme 5 6 7 8 9 pH Figure 7- 4. Enzyme loading and specific activity for immobilization at different pH (Column: specific activity; Dotted line: enzyme loading).

7.3.2 Optimization of immobilization condition for chitosan entrapment

After immobilization, the enzyme was entrapped in the matrix of polymer and the strength of the matrix was influenced by the concentration of chitosan. Enzyme solution with different concentrations of chitosan (w/v) were utilized to form the gel bead, and the enzyme activity yield was varying with the concentration and shown in

Figure 7-5.

130

2 Activity yield (%) yield Activity 1

0 3 3.5 4 4.5 5 5.5 Chitosan concentration (%)

Figure 7- 5. The enzyme activity yield with different chitosan concentration.

In this study, stable chitosan bead was failed to form when the concentration of chitosan was lower than 3% w/w, as chitosan-ICDH mixture got dispersed into the TPP solution or the beads were observed to be delicate and vulnerable to damage during the handling.

The highest activity yield (5.6%) was obtained with 3.5% chitosan and the gelation was completed in 30 s. The increasing of chitosan concentration resulted in a quicker gelation process, but later detection showed a decline in activity yield. With 5.5% of chitosan, only 2% of enzyme activity was retained after the immobilization. The decline might be related to the diffusional limitation of the substrate into the entrapped enzyme, which was more severe with a strong matrix. The overall retaining activity of

ICDH was only around 5% in this study, much lower than the value reported with references. For example, horseradish peroxidase can retain 80~90% activity with the

131

immobilization method [207]. The fragility of the enzyme with poor stability against pH, and again the chelation of TPP on the cofactor, may cause the low retain activity.

As a gelation agent, TPP caused the aggregation of chitosan with ionic attraction

(Figure 7-6). The high values of ionic strength were reported to cause the degradation of aggregation through interrupting the ionic interaction of TPP and amino group of chitosan. In this study, the gel bead swelled in the working condition after about one hour, and increasing the ionic strength could accelerate the progress. According to the reference, many factors such as the ratio of TPP and chitosan, cations category, and concentration all can affect the stability of the chitosan beads [208-210]. Since proper concentrations of magnesium and bicarbonate ions were necessary for the enzyme reactions, the negative effect of high concentration of ions on the bead stability was unavoidable. Recent research showed that the chemical crosslinking of the chitosan with chemical agents or coating the surface with silicon materials may alleviate the swelling issue, but their influence on enzyme activity and stability need to be investigated.

132

Figure 7- 6. Molecule structure of (a) chitosan and tripolyphosphate and (b) chitosan-TPP crosslinking.

7.3.3 Immobilized enzyme characterization

In most cases, the enzyme needs to be evaluated after immobilization due to the changing of kinetic behavior. After the process of immobilization, internal structural changes and restricted access to the active site would affect the kinetic constants Km,

Vmax of the enzyme. In order to evaluate the performance of immobilization, the kinetics constants of free ICDH from the porcine heart, entrapped ICDH with chitosan and adsorbed ICDH with MSF was detected and shown in Table 7-2.

133

Table 7- 2. Free and immobilized enzyme characterization

-1 -1 -1 Samples kcat [S ] Km [μM] kcat/ Km [mM S ]

Free ICDH 2.04 17.4 117.22

Adsorbed ICDH 0.298 70.4 4.24

Entrapped ICDH 0.116 156 0.74

Overall, both entrapment and adsorption methods achieved poor catalytic efficiency.

+ -1 -1 For adsorbed enzyme, the kcat/ Km value for NADP was 4.24 mM S , 28 times lower than that of free enzyme, but still 5.5 times higher than that of entrapped ICDH. The decline of catalytic efficiency was due to both kcat (14.6%) and increased value of Km

(4 times). The increased value indicated a stronger mass-transfer limitation between the substrate and immobilized enzymes, and enzyme denature during process attributed to the declining value of kcat. According to kinetic characteristics and also the stability, the MSF adsorbed method was preferred as the primary method for

ICDH immobilization.

7.3.4 Leakage of adsorbed enzyme

When applied physical adsorption for enzyme immobilization, the principal driving forces include hydrogen bonding, Van Der Waals, electrostatic forces and hydrophobic or hydrophobic interaction. Since these forces are affected strongly by the reaction conditions such as temperature/pH alterations, enzyme leakage is expected during the catalytic process [211].

134

120

100 ）

% 80 （

Relative value value Relative 20

0 0 2 4 6 8 10 12 Cycles Figure 7- 7. Enzyme activity and retain concentration with multiple reaction cycles (circle: relative enzyme activity; triangle: relative enzyme concentration).

To evaluate the leakage of immobilized enzyme, the relative enzyme activity and retain enzyme concentration were recorded and shown in Figure 7-7 at room temperature and pH 7. After ten cycles, about 55% of total enzyme activity was lost via both a gradual loss of binding protein (~31%) and enzyme inactivation (~24%).

Both binding enzyme concentration and activity declined quickly during the first three cycles, representing a fast leakage and denature of the enzyme binding on the outside surface of the MSF. Then, a mild decline occurred, indicating that the denature and leakage also happened on the enzyme adsorbed on the inside surface of the carrier.

Shear force was thought as the primary reason for the leakage and denature of immobilized enzyme, since high stirring rate resulted in a quicker decline of activity.

In practice, the leakage of ICDH may be more severe with additional existence of fluctuation of temperature and pH. Modification of the surface of the carrier to form

135

stronger covalent binding between enzyme and MSF can be helpful, but it may also cause the reducing of enzyme loading level.

Application of cross-linked enzyme aggregates (CLEAs) is another popular method of preventing enzyme leakage [212][213]. As the definition, cross-link agents such as glutaraldehyde are added into the system after the adsorption of enzyme on the surface of the carrier. The agents cause the crosslink of the mono enzyme molecule to form the CLEAs. With bigger aggregation size, it is harder for CLEAs to diffuse out of the MSF (Figure 7-8).

Figure 7- 8. a) Free enzyme adsorbed on MSF, b) CLEAs adsorbed on the MSF.

Similar to free enzyme, the leakage profile of CLEAs on MSF was studied and presented in Figure 7-9. Except for the initial 15% loss of binding enzyme during the first two cycles, no further leakage was observed in the following eight cycles, with the final enzyme retain concentration improved from 69% to 85%. However, the glutaraldehyde was proven to be toxic for ICDH. After adding 0.5% of glutaraldehyde,

136

only 0.5% of enzyme activity was retained after immobilization. Applying a lower concentration of glutaraldehyde could slight alleviate the inhibition, but also decreased the resistance to the leak of enzyme. As the crosslinking can happen at the active site of the enzyme, the necessary contact between enzyme and substrate was interrupted, with a decrease of activity.

120

） 100 %

（ 80

20 Relative valueRelative 0 0 2 4 6 8 10 12 Cycles

Figure 7- 9. The retain concentration of enzyme with multiple reaction cycles with cross-linking treatment.

7.3.5 Stability of immobilized enzyme

Improvement of enzyme stability was one of the primary proposes of enzyme immobilization. As the conformation of enzyme was limited by the scaffold, immobilized enzymes showed more resistance to harsh conditions such as extreme temperature, pH shift, shear force and a high level of chemical inhibitors [214-216].

137

Temperature

The improvement of thermo-stability was checked and presented in Figure 6-10. At room temperature, both free and immobilized enzymes were stable as more than 95% of the relative activity was maintained after 8 hours incubation. An obvious decline of relative activity was observed when the system temperature beyond 45 °C. Detailer, the relative activity was 93% after 2 hours and further decreased to 76% at 8 hours at

45 °C. Only a slight improvement was observed with immobilization at 45 °C. The benefit of immobilization became apparent at 55 °C: the half lifetime (t1/ 2) of immobilized ICDH was around 6 hours, higher than that of the free enzyme (4 hours).

70 Relative activity (%) activity Relative

50 0 2 4 6 8 Time (h)

138

Figure 7- 10. Effect of immobilization on enzyme stability against high temperature (Diamond: free enzyme; Triangle: Immobilized enzyme. (a) 20 °C; (b) 45 °C; (c) 55 °C).

Significant differences of the thermo-stability between ICDHs originated from different species were reported. For the data, ICDH from porcine heart showed a poor thermo-stability under high temperature, while some references pointed out ICDHs from other species such as thermus aquaticus can remain its activity at a much higher temperature. For example, Ramaley discovered that the ICDH form thermus qauaticus

139

can keep its activity at 60 °C, but the lack of carboxylation activity made this kind of

ICDH unsuitable for CCS [217]. The profile of enzyme stability with temperature was believed to be related to the enzyme’s native working condition. It is more likely that new ICDH with better thermo-stability can be screened from organisms lived in a hot environment such as hot springs [218].

Regarding the enzyme stability, although the enzyme accessed the highest reaction rate at 45°C, the quick decline of relative activity made this temperature unsuitable for long time process, even with immobilization. The ICDH required room temperature to maintain its activity.

System pH

Besides the thermo-stability, the stability of ICDH against pH was also critical for efficient carbon capture and release. In this design, the initial carbon capture was conducted under acidic conditions and then the pH of the system was adjusted to alkaline conditions to release carbon dioxide. Considering both kinetic and thermodynamic equilibrium data, pH 6 and 9 was chosen for carboxylation and decarboxylation separately, and the stability of free and immobilized enzyme at these two pHs was shown in Figure 7-11.

140

Figure 7- 11. Effect of immobilization on enzyme stability against pH (Diamond: free enzyme; Triangle: Immobilized enzyme. (a) pH 9; (b) pH 6). This enzyme revealed less stability against pH than stability in high temperature. A rapid decline of enzyme activity was detected when the free enzyme was kept in the pH 9 buffer and only 28% of activity was retained after 2 hours. During carboxylation step, ICDH was more stable at pH 6 and exhibited only 10% activity lost in the first 2 hours. The stability of the immobilized enzyme was apparently improved by the immobilization. At pH 6, the half lifetime (t1/ 2) of immobilized ICDH was around 18

141

hours, much higher than that of the free enzyme (8 hours). The stability of ICDH at pH 9 was worse than that at pH 6, yet immobilization improved its t1/ 2 from 1.5 to 3.5 hours (data not shown).

Shear force

When enzyme was exposed to the flow of reactor, the shear force is often regarded as a cause of protein denature. Many references have pointed out the severe enzyme activity loss with a high shear force [219-221]. Similar to thermo-denature, the loss of activity caused by shear stress is considered as irreversible [222]. In practice, immobilization can limit the conformation change of enzyme, therefore increases the enzyme stability against shear force. The improvement of immobilized enzyme against shear stress under 200 and 1000 rpm was shown in Figure 7-12.

a 100

Relative activity (%) activity Relative 20

0 0 2 4 6 8 Time (h)

142

Figure 7- 12. Effect of immobilization on enzyme stability against shear force (Diamond: free enzyme; Triangle: Immobilized enzyme. (a) 200 rpm; (b) 1000 rpm).

According to the figure, the activity of ICDH was stable at 200 rpm. After 8 hours incubation, the retaining activity was 85% and improved to 91% with immobilization.

However, a higher stirring rate can result in an obvious decline of enzyme activity.

For free enzyme, 15% enzyme activity was lost in the first 2 hours, and only 59% remained after the total 8-hour incubation at 1000 rpm. In this condition, immobilization can be helpful and only 12% activity was lost after 8 hours. Since the actual shear force during manipulation is less than 200 rpm case, shear stress will have less effect on activity.

7.3.6 The effect of CA on activity of immobilized ICDH

After ICDH is immobilized and concentrated on the surface of MSF, carboxylation is realized with the diffusion of carbon species to the active site of enzymes. Since the gas phase of carbon dioxide is the preferred substrate for carboxylation, the local conversion rate between bicarbonate and dissolved gas phase carbon dioxide may 143

limit the overall carboxylation reaction rate. Based on this assumption, carbonic anhydrase (CA) can accelerate the carbon species conversion and therefore enhance the reaction rate. The effect of different concentration of CA on the carboxylation activity was studied and exhibited in Figure 7-13.

100

90 M/min)

μ 80 （

60 Reaction rate rate Reaction

50 0 0.5 1 1.5 2 [CA] (mg/ml)

Figure 7- 13. The reaction rate of carboxylation with different CA concentration.

The reaction rate was about 65 μM per min for the control group and 0.5 mg /ml CA did not show noticeable improvement in the reaction rate. Since very limited adsorption of CA was detected, most CA was considered homogenously dispersed in the reaction solution and required higher concentration to show the benefits. When

CA concentration was 1 mg /ml, the carboxylation reaction rate increased to 75 μM per min, representing a 20% improvement. No further improvement of activity was observed beyond that CA concentration. The same experiment was conducted with free ICDH and CA, but no improvement of carbon capture was observed, indicating the heterogamy of the enzyme system was required for the benefit of CA.

144

7.4 Conclusions

In conclusion, ICDH immobilization was realized with both MSF adsorption and chitosan entrapment. For the fragile ICDH, adsorption method showed better performance than the entrapment method, which may due to less enzyme denature during the preparation process. The characteristics of enzyme can guide the screening of immobilization methods.

Being a simple immobilized method, the study resulted in leakage of enzyme after the immobilization. After 10 cycles, 30% of the binding enzyme was leaked from the carrier, combining with denaturing of the enzyme, caused 55% enzyme activity loss.

Although adding glutaraldehyde could alleviate the leakage of the enzyme, most enzyme lost its activity during the crosslink progress. In the end, additional CA in the adsorbed enzyme system can help to increase the carboxylation rate by 20% through enhancing the carbon species equilibrium near the enzyme active sites.

ICDH tended to lose a large portion of activity with temperature and pH fluctuation which commonly happened during carbon capture and release process. Although immobilization with physical adsorption, benefits enzyme stability somewhat, the stability of enzyme is still far away from the requirement for the industrial application.

Moreover, the immobilization itself also brings additional cost related to material and process and hurts the economic feasibility. In that case, the discovery of new enzymes through enzyme screening and also the modification of current enzymes with molecule methods may be helpful [223] [224].

145

Chapter 8. Economic analysis of ICDH based CCS

8.1 Introduction

Carbon capture and storage (CCS) has been pursued as the most promising strategy in recent years to mitigate carbon emissions from our still steadily-growing consumption of fossil fuels. Through a CCS process, CO2 is captured and separated from its gaseous mixtures (such as flue gas from coal-fired power plants) into a pure gas, which can be subsequently liquefied and injected into underground storage sites [225].

Alternatively, the purified CO2 can also be converted into value-added products, or applied to various industrial processes [226]. It has been reported that CCS processes

3 have reached a level of 2 billion ft CO2/yr in oil recovery industry by the end of 2010

[227]. Different approaches to CCS have been realized such as membrane separation, cryogenic distillation and physical adsorption on a carrier. Despite these prominent methods, the poor stability and energy efficiency hinder their broad application in industry. Most industrial CCS processes have been realized using chemical absorbents.

Among other options, amines such as monoethanolamine (MEA), diethanolamine

(DEA) and methyldiethanolamine (MDEA) were the most common agents for industrial processes [228]. Such weak basic chemicals generally react with CO2 at low temperatures (∼50 °C), realizing the separation of CO2 from its gaseous mixtures; the resulted products are required to be decomposed subsequently, thus regenerating the absorbents and releasing CO2 as a pure gas. However, chemical regeneration requires elevated temperatures, mostly above 120 °C. Since the process involves a large

146

amount of water (∼80% w/w in the solution), large portions of the energy are wasted during the heating-cooling cycles.

To seek energy-efficient CCS strategies, growing interests have been shown in directed to in vitro biocatalytic carbon capture, which can be operated in a way similar to chemical carbon sequestration but promises high energy efficiency as the regeneration and reuse of substrates can be realized at ambient conditions; the reactions can also be readily intensified through well-developed reaction engineering means. As an enzyme constituting the reductive tricarboxylic acid cycle (RTCA) for carbon fixation, ICDH catalyzes the conversion of CO2 and ketoglutarate to isocitrate under the acidic conditions and could also reverse the direction of reaction with pH shifting.

Regarding reaction mechanism, ICDH and MEA based CCS show the different requirement of energy and chemicals input. Generally speaking, enzyme-based CCS requires more input of chemical material but less energy, considering the complexity of reaction and high energy efficiency. To reveal the potential application of ICDH based CCS in large-scale, economic analysis can be helpful. By calculating the capital and operation cost with existing data and reasonable assumptions, the cost of avoiding unit of carbon dioxide through both MEA and ICDH based CCS could be calculated and compared to show their economic feasibility.

8.2 Material and methods

147

8.2.1 Materials

Sodium formate and Formate Dehydrogenase from Candida boidinii purchased from

Sigma Chemical Co. (St. Louis, MO, USA). See the section 3.2.1 for other materials.

8.2.2 Cofactor stability test

The pH of phosphate buffer, bicarbonate, and carbonate buffer were adjusted by varying the buffer species according to Henderson-Hasselbalch equation (7.1) [229] and the pH was carefully adjusted to the temperature specified. pH = pK + log [A]/[HA] (8.1)

Buffer concentration = [A]+[HA] (8.2)

NADPH was dissolved in the buffer at various pH to a final concentration of 0.1M and a constant temperature was applied with a water bath. For a prepaid time, the concentration of NADPH was detected at A 340 with a spectrophotometer. The degradation of NADPH was assumed as the first-order reaction, and the apparent first order constant (k) and half time (t 1/2) of NADPH could be calculated with following equations

푑[푁퐴퐷푃퐻] − = 푘[푁퐴퐷푃퐻] (8.3) 푑푡

푙푛2 And 푡 = (8.4) 1/2 푘

8.2.3 Activity assay of formate dehydrogenase

The assay mixture contained 50 mM sodium formate, 1mM NAD+ and 0.1 mg free

enzyme in 1 mL of 0.1M Tris buffer pH 7. At room temperature, the increase rate of 148

NADH in solution was detected by absorbance at 340 nm, and one unit of activity was

defined as 1 μmol of NADH formed per min by formate dehydrogenase.

8.2.4 Cofactor regeneration

In this study, NADH-depended formate dehydrogenase was used to realize the cofactor regeneration. The assay mixture contained 1 mM sodium formate, 1mM

NAD+ and 1 mg free enzyme in 1.5 mL of 100 mM bicarbonate or carbonate buffer with pH from 6 to 9 in a cuvette and the cuvette was capped to prevent the loss of releasing carbon dioxide. For every 10 mins, the concentration of NADH was recorded by monitoring absorbance at A340 nm by the spectrophotometer. The reaction was assumed to be completed when there is no changing of A340 between intervals, and the equilibrium constant (Km) was calculated with the following equation.

[퐶푂2(푡표푡)][푁퐴퐷퐻] 퐾푚 = − + (8.5) [HC푂2 ][푁퐴퐷 ]

8.2.5 Economy analysis

The purpose of economic analysis of carbon capture and sequestration is to better understand the technological options for carbon dioxide capture and their potential economic feasibility. The fire plant without carbon dioxide capture treatment was considered as the reference, and both chemical absorption (MEA) based and enzyme

(ICDH) based CCS strategy was analyzed with available data [230-234] and reasonable assumption with precious chapters.

149

Case study

400 MWe traditional coal fire plant and the main operation conditions as following:

 Coal burn rate: 221*103 kg/h

 The plant operates at 8000 h/ year.

 Carbon dioxide emission: 3’000’000 tons per year

 Flue gas temperature: 150 °C

 Carbon dioxide mole concentration of flue gas : 14.59%

Capital expense

The capital expense is related to property, plant, and equipment, initial chemicals and normally this cost will be charged to depreciation expense over the useful life of the asset. For MEA based carbon capture, the main capital expense includes absorption columns, heat exchangers, tanks and pumps, regeneration columns and piping. For

ICDH based carbon capture shares the same cost of plant and equipment, but brings more cost of chemicals. For both cases, the capital expense is amortized over 20 years with 7% interest rate and $0 salvage value.

Operating expense

Operation expense occurs during the carbon capture and release process and cost of chemicals and energy are two main components of operating expense.

Product value

In this study, the product is purified and compressed carbon dioxide and need to be transported to targeted locations. Regarding the geological feasibility, compressed

150

carbon dioxide can be sequestrated into deep sea or unmineable coal steam, or utilized for enhancing the recovery oil from mature oil fields. With solo sequestration, the value of product is negative since more cost occurs during transportation and sequestration, but a positive product value can be expected when the sequestration combines with enhance oil recovery strategy.

With the above data, annual cost of CO2 avoided and cost of electricity would be calculated with following equations:

Total annual cost = amortized capital cost + operation cost - annual avenue of product (8.6) Cost of CO2 avoided ($/ tonne) = Total annual cost / Total carbon dioxide sequestration （8.7） Cost of electricity ($ / kWh) = reference electricity price + total annual cost / annual electricity output （8.8）

8.3 Results and discussion

8.3.1 Cofactor stability

Due to the high cost, the stability of cofactor (NADPH and NADH) is critical for the overall economy of the system. References point out that the stability of cofactors can be influenced by many factors such as temperature, pH, ionic strength, and solution types. Since the 1930s, the effect of pH on cofactor stability has been widely studied

[235][236]. As NAPDH or NADH exist in acid solution, the ultraviolet absorption of the dehydronicotinamide ring at 340 nm decreases with the concomitant appearance of a new chromophore at 280 nm, indicating a continuous denature. The reaction can

151

represent with model compounds as Figure 8-1.

Figure 8- 1. Model of NADH denature under acidic condition.

According to the mechanism of denaturing, the concentration of proton should significantly affect the stability and cofactor. To verify it, the decline curve of cofactor under different pH with/without buffer condition was studied and showed in

Figure 8-2.

152

Figure 8- 2. NADPH stability at different pH and buffer category (a: carbonic buffer; b: phosphate buffer; : pH 9; : pH 8; : pH 7; : pH 6).

From the figure, NADPH was stable under the alkali conditions, but denatured quickly under neutral and acid conditions: only 8% of enzyme denatured under 10 hours cubulation at pH 8, but more than 53% of the cofactor lost its activity in the same conditions but at pH 6. In this study, phosphate buffer produced a negative effect on the stability at full pH range, in accord with Wu’s report [235].

In addition to the effect of pH and phosphate, ionic strength and temperature also affected the stability of the cofactor and their influences on the degradation reaction constant and halftime of NADPH were summarized in Table 8-1.

153

Table 8- 1. Degradation reaction constant and halftime of NADPH under various conditions

-1 pH T (°C) Concentration (M) K (min ) 푡0.5 (min) pH 6 20 0.1 0.00136 510

7 20 0.1 0.000428 1620

8 20 0.1 0.000192 3600

9 20 0.1 2.831x10-5 24480

T 6 30 0.1 0.00578 120

6 40 0.1 0.0198 35

Concentration 6 20 0.5 0.000924 750

6 20 1 0.000722 960

Phosphate 6 20 0.1 0.00495 140

6 20 1 0.00693 100

6 30 0.1 0.0139 50

According to date of above table, increasing reaction temperature resulted in a shorter halftime of the cofactor, as the values of t1/2 were 120 and 35 mins under 30 °C and

40 °C respectively with 0.1 M bicarbonate. Increasing bicarbonate concentration, on the other hand, decelerate the degradation of NADPH.

154

Compared with the enzyme, NADPH shows worse stability during carbon capture process which requires an acidic condition. Although the CCS cycle itself does not consume NADPH, the continuous degradation of cofactor during the cycle not only increase the cost but also complicate the process. By now, the approaches to increase

NADPH stability against pH were limited; however, many studies have reported synthetic nicotinamide cofactors employed with oxidoreductases in various enzymatic processes. The application of synthetic cofactor in CCS is also expected and these synthesized pyridine derivatives can be entirely redesigned to improve their stability under different conditions.

8.3.2 Cofactor regeneration

Formate dehydrogenase utilizes formate as the substrate to regenerate the cofactor, releasing carbon dioxide and the reaction equilibrium controls the regeneration rate

[237]. In CCS process, a high concentration of carbon species exists in the system, hindering the regeneration. In order to evaluate the performance of regeneration, the cofactor regeneration experiments were conducted at pH 6-9 and conversion rates were shown in Figure 8-3. The equilibrium constants (Km) were also calculated in

Table 8-2.

155

0.8

0.6

0.4

Conversion rate (unit) rate Conversion 0.2

0 6 7 8 9 pH Figure 8- 3. Cofactor regeneration rate at different pH catalyzed by formate dehydrogenase.

Table 8- 2. Equilibrium constant (Km) of reaction catalyzed by formate dehydrogenase pH 6 7 8 9

퐾푚 323 899 3809 10502

During the regeneration, carbon dioxide was the product and high concentration of it would lower the efficiency of cofactor regeneration and 100 mM carbon species was applied to mimic the carbon accumulation in this study. It was obvious that pH played an important role in regenerating the cofactor: at pH 6, about 580 uM of the cofactor

(58%) was regenerated and the regeneration rate can be improved to 90% when the pH was adjusted to pH 9. Considering the Km value, the reaction equilibrium constant was shifted by a factor of 32 when pH changed between 6 and 9. To achieve high

156

cofactor conversion, the regeneration reaction should be conducted under the alkali condition. Since the cofactor regeneration and carbon capture prefer different pH, the formate hydrogenase and isocitrate dehydrogenase should be immobilized on locations with different pH to prevent the interruption.

Besides the high reaction rate, another advantage of using formic acid to regenerate the NAPDH could be its easy synthesis method. Many references have pointed out the method of electrochemical reduction of carbon dioxide to formic acid [238-240]. The general reactions at cathode and anode were described as following:

+ - Cathode: 2CO2 + 4 H + 4 e → 2 HCOOH （8.9）

- + Anode: 2H2O → O2 + 4 e + 4 H （8.10）

With a low potential apply (~ 1-3V), the faradaic efficiency was about 90%. Different metals such as Cu, Ni, Pt, and Ag were mixed to make the electrode, to improve the reaction selectivity and efficiency. Interestingly, pH was reported to influence the production rate of formic acid as well. At alkali condition, both reaction selectivity and intensity decreased because most carbon species exist in the form of bicarbonate or carbonate, instead of the gas phase of carbon dioxide [241].

As the formic acid production rate, cofactor regeneration rate and also CCS process prefer the different values of pH, more efforts are required to find the optimum point to improve the overall efficiency and reduce the cost.

157

8.3.3 Economy analysis

Today, the coal fire plants with 400-500 MW capacity in the US provide almost half of all power generation and moreover, 80% of total carbon emissions from the electric utilities [242]. Although in laboratory scale, many technologies such as cryogenics, membranes separation are proven to successfully separation of carbon dioxide. The low concentration of carbon dioxide in the flue gas makes chemical based carbon absorption the only existed large-scale approach in the USA today. There is an increasing concern about realizing carbon capture and sequestration with a biocatalytic (enzyme) approach, but most of them are still under pilot scale. While biocatalytic approach brings high energy efficiency and greener process, the native reaction intensity and stability are too poor to scale up, and a lot of efforts are expected to be done to solve these critical problems.

Capital expense

The capital cost of the CCS is mainly related to equipment purchase and installment.

Consider the annual carbon dioxide emission and 90% carbon removal requirement, equipment size is optimized to complete the job. Table 8-3 summarized the capital cost of both MEA based and ICDH based case. While existing MEA plant provides detailed data for the MEA case, most data for ICDH data was obtained from correlation from MEA case.

158

Table 8- 3. Capital expense for both MEA and ICDH based CCS plant [243]

($) MEA based ICDH based

Absorption columns 29,536,400 29,536,400

Regeneration columns 10,960,000 10,960,000

Heat exchangers 4,416,100 0

Piping 6,830,130 6,830,130

SO2 scrubber 24,000,000 24,000,000

CO2 compressor 20,147,300 20,147,300

CO2 drying system 14,241,303 14,241,303

Solvent reclaimer 1,168,517 1,168,517

Auxiliary heat units 38,634,098 0

Startup and chemicals 8,544,782 108,144,782

Other cost 85,548,234 85,548,234

Total capital cost 244026864 337210764

For MEA case, the carbon capture and release cycles are conducted in the absorption and regeneration columns with auxiliary heat units to provide heat duty. The typical concentration of MEA in the solvent is about 30 w/w %, and for each mole of carbon dioxide, 4.54 mole of MEA is supplied for completing one CCS cycle. With 159

calculation, four separate sets of absorption and regeneration columns are installed, totally accounting about 15% of the total capital cost. For ICDH case, 1 mole of carbon need with 2 moles of substrate supply. It is reasonable that the same size of column sets matches the need for ICDH case. In the calculation, the cost of SO2 scrubber to deal with the SO2 in the flue gas, takes about 10% of the capital cost.

Regarding the sulfate content in the coal, the scrubber is changed in its size and can also be avoided with low sulfate coal or natural gas. CO2 compressor and CO2 drying system are used to treat with the purified carbon dioxide and compress it to 150 bars to pipeline transportation. The other cost in the project includes field indirect, home office engineering, contract fees and contingency.

Between MEA and ICDH case, the most significant difference of the capital cost would be the startup chemicals. A total 320 tone of MEA is filled with four sets of columns, which cost $0.4 M ($1250 per ton). For ICDH based case, the initial chemicals are NADP+ (with cofactor regeneration), ketoglutarate and magnesium chloride, which cost $5,400,000, $1000 and $110 per ton respectively. Based on calculation, the cost of startup chemicals for ICDH case is $ 10000 M at the current price level, almost 25000 times higher than MEA. Since the cofactor contribute the most to the chemicals cost, avoiding cofactor with high price is necessary for the economic feasibility of ICDH project. Many approaches to reduce the cost of cofactor have been reported by references, including changing the selectivity of the enzyme on cofactor. With genetic modification, ICDH may utilize NADH, instead of NAPDH, or

160

electron directly. To continue our analysis, the price of the cofactor is assumed to reduce to 1% of the current price with potential technology improvement and large-scale production effect, and the total cost of start-up chemicals for ICDH is

100.4 M.

For ICDH case, another extra cost occurs with using ICDH and aconitase. Regarding the enzyme kinetics, about 10 ton of ICDH and 100 ton of aconitase are required for treating the gas flow. With further enzyme screening and optimization, a thousand times of improvement of specific activity is expected here, reducing the need for the enzyme from 110 ton to 110 kg. Moreover, enzyme cost can be reduced significantly with molecular biology technology and modern fermentation process. As an industry enzyme, the price of cellulase can provide a good approximation to ICDH and aconitase, which is around $ 10000/ ton. As a result, the cost of the enzyme is only a small part of the capital expense.

As a result, the total capital cost of MEA and ICDH based case is $ 244 and $ 301 M respectively. The main increment for ICDH case is related to the high price of the cofactor.

Operation expense

For MEA case, the carbon capture and release cycle is realized by changing the reaction temperature, while shifting the reaction pH is the driving force for ICDH case.

Different mechanism results in unique characteristics of their operation cost and the

Table 8-4 summarized their annual operation cost. 161

Table 8- 4. Operation expense for both MEA and ICDH based CCS plant

($) MEA based ICDH based

Scrubber chemicals 7,000,000 7,000,000

Substrate makeup 4,656,720 0

Enzyme makeup 0 1,760,000

Shaft duty cost 9,600,000 9,600,000

Heat duty cost 28,800,000 0 pH shifting chemicals 0 1890

O & M cost 11,769,999 11,769,999

Total operation cost 35,888,719 26,308,825

As a big portion of operation cost, the energy consumption is expressed both shaft power and thermal energy. The first is mainly related to turbomachines, and the second is more about the input of energy into the re-boiler of the regeneration columns. The significant disadvantage of MEA based CCS is high energy input. For

400 MM coal fire plant, 40 MW of the shaft duty and 120 MW of the heat duty are required for MEA CCS. The annual cost of the shaft duty and heat duty is calculated as 38.4 M, regarding 0.03 kW/h cost of electricity today. For ICDH case, since no

162

energy was required to regenerate the absorbent, only shaft duty is accounted for energy cost.

Besides the energy cost, another significant portion of the operation cost is related to chemicals makeup. For MEA case, self-degradation of MEA necessitates continuous supply of MEA and the cost is $1.55/ton of carbon dioxide capture [234]. For ICDH case, the cost of chemicals includes makeup of the enzyme and cofactor and again, the alkali for adjusting the solution pH. With current data, the stability of cofactor during the process is the critical issue: although cofactor shows good stability under alkali condition (decarboxylation), the denaturing becomes server under acid condition

(carboxylation). Unluckily, there is no report about keeping the cofactor stable with a long-time operation. Designing artificial cofactor may be a solution for this issue, and the loss of stability of cofactor is neglected for this project. Finally, the enzyme activities loss is also considered and the annual cost of makeup enzyme is 1.76 M with 10% of enzyme activity loss for 5 hours (10 times improvement with current stability data).

For ICDH case, adjusting the pH of the system brings additional operating cost. While carbon dioxide can act as the acid to provide acidic condition, additional alkali is needed to elevate the pH for carbon release. Although calcium hydroxide is preferred in a large-scale project due to its low cost, both ICDH and aconitase show a poor stability with a high concentration of calcium. As a result, sodium hydroxide instead of calcium hydroxide would be the source for adjusting pH. The amount of alkali

163

required is mainly related to the substrate concentration. For 5 M substrate level, only

128 g of sodium hydroxide is needed for one cycle (release at pH 9), and the annual cost is only $1890. In practice, the buffer effect of carbon dioxide in the solution may result in a huge increase of the need of alkali. If 70 mM residual CO2 is retained after the absorption, the need of alkali for adjusting pH is 26.8 kg, represent an annual cost of $ 397,004. After adding the cost of O & M and cost of chemicals for sulfur dioxide, the operation cost of ICDH case is $ 26.3 M, $ 9M less than that of MEA case.

During the cycles, Na+ are accumulated in the solvent with the continuous supply of alkali. Without the buffer effect, only 10 µM sodium is introduced into the solvent during one cycle. This slow accumulation will bring limited effect on both reaction equilibrium and kinetics, which can be prevented by removing the sodium with membranes.

Product value

In our case, the product is compressed and dry carbon dioxide and as the last step of

CCS, the carbon dioxide need be transported within the pipeline and sequestrated into target locations and special geological feature such as trapping or adsorbing formation is needed to prevent the leakage of carbon dioxide. The CO2 transport cost is around

0.02/ tonne/ km and the disposal cost is 5/tonne. For 3,000,000 tons per year CO2，the annual revenue of the product is $ -21M. Interestingly, with the new technology from the petroleum industry, the enhanced oil recovery (EOR) with carbon dioxide injection would dramatically reduce the cost of CCS. During gas injection EOR,

164

carbon dioxide is continuously injected into the oil formation through an injection well, increasing the formation pressure. As a result, the initial oil in the formation is forced to move out from the formation through the production well. The performance of recovery is mainly depended on the oil price, reservoir size and characteristics.

According to reference, four barrels of oil can be recovered with one ton of carbon dioxide injected [244]. Considering the capital and operation cost of EOR project, the value of carbon dioxide per ton is between $ 124 per ton (new injection well) and

$ 280 per ton (after water flooding) with $ 80 / bbl oil (without royalty and tax). By now, the most critical factor which limits the gas injection EOR is the carbon dioxide feasibility: in the case there is no nearby fire plant close to the oil reservoir. If EOR is considered, the annual avenue of the product could increase from -21M to 840 M.

Cost of carbon dioxide avoided

The total annual cost of CCS with two cases was summarized in Table 8-5 and cost of

CO2 avoided with two cases was shown in Figure 7-4. For the data, the total annual cost of MEA case is about $ 107 M and results in a cost of $ 35.94 per tonne CO2 removed. ICDH case possesses a small value of annual cost, which is $ 79.5 M. For both cases, the operation cost weights the highest portion of total cost: 59% for MEA case and 38% for ICDH case. Although ICDH case shows a higher capital cost, the significant decrease of operation cost makes ICDH a better choice for CCS project with 20 years’ life.

165

Table 8- 5. The total annual cost of CCS for both MEA and ICDH case

MEA based ICDH based

Amortized capital cost 23034410 28372310

Operation cost 63776719 30131889

Sequestration cost 21000000 21000000

Total annual cost 107811129 79504200

$40.00

$7 $30.00

$20.00 avoided 2 2 $21.26 $10.04 $10.00

$ /Tonne CO /Tonne $ $7.68 $9.46 $0.00 MEA ICDH

Figure 8- 4. The cost of CO2 avoided for MEA and ICDH cases (Top: Sequestration cost; Middle: Operation cost; Bottom: Capital cost).

Many factors including the location of coal fire plant, quality of the fuel, and oil reservoir feasibility would affect the performance of CCS dramatically. CCS project can become economically feasible (make a profit) if the gas injection EOR is possible.

The value of recovered oil with purified carbon dioxide can totally make up the cost of capture and sequestration. With calculation, the project is at breakeven when the 166

price of oil is $20/bbl for a mature reservoir and $52/bbl for a new reservoir which is lack of infrastructure [245]. Also, among the cost of carbon dioxide sequestered, about $ 3.088 occurs in the treating sulfur content in the coal. Since a lower content of sulfate exists in natural gas, replacing coal with natural gas as fuel could results a

$ 2-3 decrease in the carbon avoided cost; using high sulfate coal, on the other hand, increase the cost.

In this case, ICDH case shows low carbon dioxide avoided cost due to higher energy efficiency and several optimizations are assumed to meet the target:

1. Cofactor price decreases to 1% of current level;

2. No cofactor activity loss during pH shifting;

3. 1000 times improvement of specific activity of ICDH and aconitate;

4. Cost of enzyme close to $ 10000 per ton.

All these assumptions focus on the reaction stability and intensity, which are comparatively poor in biocatalytic relations. Compared with the enzyme, cofactor brings more influence on the performance of ICDH based CCS, regarding its high price. Different from abundant references about the enzyme stability and activity improvement, the development of artificial cofactor is very limited and therefore brings more uncertainty about the application of ICDH based CCS in the future.

Cost of electricity

167

In the end, the generation cost of electricity based on the MEA and ICDH cased was calculated as $ 0.075 and 0.063 / kWh respectively (with low coal cost as base), and the comparison of electricity from different resources is shown in Figure 8-5.

Figure 8- 5. Electricity price from different resources.

Coal is the cheapest resource for electricity, and the cost of electricity from renewable resources such as wind and solar is about 0.12 / kWh, almost three times more than that from coal. To avoid carbon dioxide emission, CCS is applied in coal fired plant and increases the cost to 0.06~0.07 / kWh and its price is still lower than that from the renewable resources. Today, the tax credit is the primary approach to encouraging the society to utilize renewable energy source instead of fossil fuel [246-248]. However, the $ 0.01- 0.03 /kWh credit cannot entirely cover the difference of electricity price between coal and renewable resources and the development of wind and solar energy are heavily depended on the costume’s concerns about global warming. With the

168

current tax credit, the electricity cost from CCS based coal fire plant, however, can compete with traditional coal fire plant merely on economic aspect.

8.4 Conclusion

Formate dehydrogenase based cofactor regeneration was conducted to reduce the cost of the cofactor. The results showed that the regeneration preferred high pH: more than

90% of cofactor was regenerated at pH 9, but the regeneration rate decreased to 58% at pH 6. As the carboxylation and cofactor regeneration required different pH, it is required to separate the two processes to improve the overall performance,

Compared with MEA, ICDH case shows higher energy efficiency. Instead of regenerating chemical absorbent with heating, ICDH based CCS realizes the carbon capture and release cycle with shifting pH, thus saves the massive cost of heat. In ideal conditions, the cost of carbon dioxide avoided for ICDH case is $ 26.50 per ton, lower than that of MEA case ($ 35.94 per ton). By now, the stability and again high cost of cofactor are the biggest barriers impeding the application of ICDH.

With the CCS technology, the coal fire plant could generate electricity without carbon emission. The addition cost is can as low as $ 0.03-0.04 /kWh, bringing the total price to $ 0.07-0.08 /kWh. Although the development of new technologies dramatically reduces the cost of electricity from renewable resources such as wind and solar, the price of electricity from the renewable resource is still higher than that of coal with

CCS. As many factors including plant size, sequestration location, and sulfate content

169

affect the CCS performance significantly, the design of CCS project reply on the local situation.

170

Chapter 9. Conclusions and Future work

9.1 Conclusions

This research focused on the exploration of novel reversible biocatalytic reaction for carbon capture and release cycle with high energy efficiency. The major approach was manipulation of the direction of isocitrate dehydrogenase reaction by shifting the system pH. This study combined both theoretical model and experimental data to evaluate the potential feasibility of ICDH based CCS in large scale. To achieve this, carbon capture and release experiments were conducted and the results were verified with Debye–Hückel theory. To enhance the reaction intensity, ICDH from

Chlorobium limicola was screened for its highest specific activity and aconitase was introduced to increase the carbon capture capacity. ICDH was immobilized on the mesoporous silicon foam which improved its stability against temperature, pH, and shear force, resulting in better system stability and reduced reaction cost. In addition to that, the performance of cofactor regeneration catalyzed by formate dehydrogenase was also examined under carboxylation condition. The specific conclusions are:

1. This work demonstrated that the reduction reaction of carbon dioxide catalyzed by

ICDH had a strong pH sensitivity, and can be reversibly manipulated by adjusting the pH of the reaction solution. Allowing non-ideality of the chemical species involved, the experimental data agreed well with theoretical thermodynamic predictions.

Although manipulating pH in vivo was probably not an operational option, biochemical reactions using isolated enzyme may offer such convenience and promise

171

processes to capitalize on the potentials of the reactions. Considering our growing capability of developing robust enzymes for extreme conditions, one can expect efficient carbon capture and conversion can eventually be realized by manipulating biological reaction pathways in vitro.

2. In this study, ICDH from Chlorobium limicola was expressed with E. coli

BL21(DE3). The specific activity of carboxylation reached 15.23 U/mg purified enzyme, almost 40 times higher than the commercial form of ICDH from the porcine heart. Moreover, the optimum pH of carboxylation was 7 for the recombinant enzyme, higher than that of porcine heart (pH 6).

3. A novel carbon capture strategy was demonstrated through ICDH-catalyzed pH-sensitive reversible carboxylation reaction. Carboxylation reaction was preferred under acidic condition while decarboxylation reaction could be realized at basic condition, as indicated by both peak reaction rates and reaction equilibrium constant.

With kinetic analysis, gas phase carbon dioxide was the preferred substrate for the carboxylation and the partial pressure of carbon dioxide can determine both the reaction rate and reaction equilibrium with the fixed concentration of absorbent.

4. This work demonstrated a unique carbon capture strategy by integrating the reaction of ICDH and aconitase. Both theoretical prediction and experimental data showed a significant improvement of carbon capture capacity with such a cascade enzyme system. By manipulating the pH, the carbon capture capacity of biocatalytic carbon sequestration can approach that of chemical absorbents such as MEA,

172

although substantial efforts are needed to prepare inexpensive enzymes and cofactor, or their synthetic alternatives, to ultimately afford large-scale operations.

5. Silicon-based physical adsorption can modestly improve the stability against pH denaturation, with 130 % improvement in half lifetimes. As expected, the porous nature of the materials was efficient to achieve high enzyme loading (2.5 wt %), while maintain 13% of parent enzyme activity.

6. With several assumptions about improvements, ICDH based CCS can realize 66% improvement in energy efficiency, therefore bring the lower cost of carbon dioxide avoided than that of MEA based CCS. However, in present, the high cost of cofactor and the poor stability of cofactor under acid condition hindered the practical application of ICDH in CCS with large scale.

9.2 Path ahead

Biocatalytic reactions offer high reaction specificity and unparalleled high energy efficiency at ambient condition. The application of reversible enzyme reaction for carbon capture and release is attractive as less energy input for regenerating the absorbent. Taking ICDH reaction as an example, this research has set a solid basis toward technology down that direction. The future research regarding the application of enzyme reaction for carbon capture and conversion should be conducted.

The current barrier of the feasibility of enzyme reaction in large-scale CCS project would be the high material cost and low system stability. For most biochemical

173

reactions participating in native carboxylation cycle, the cofactor is needed for electron delivering, which possesses high price. Further Screening the reversible carboxylation reaction without cofactor may be an approach to reducing cost; otherwise, more effort need to be conducted to design artificial cofactors with low price and better stability. Regarding enzyme stability, the novel enzyme with high thermal or pH stability may be expected in the organisms living in extreme condition.

Also, the enzyme stability can be improved through more sophisticated immobilization and molecule modification methods.

174

References

[1] Berg, I. A., Kockelkorn, D., Buckel, W., & Fuchs, G. (2007). A 3-hydroxypropionate/4-hydroxybutyrate autotrophic carbon dioxide assimilation pathway in Archaea. Science, 318(5857), 1782-1786.

[2] Rao, I. M., & Terry, N. (1989). Leaf phosphate status, photosynthesis, and carbon partitioning in sugar beet I. Changes in growth, gas exchange, and Calvin Cycle enzymes. Plant Physiology, 90(3), 814-819.

[3] Schmalensee, R., Stoker, T. M., & Judson, R. A. (1998). World carbon dioxide emissions: 1950–2050. The Review of economics and statistics, 80(1), 15-27.

[4] Pawan, M. (2014). Impacts of Global Warming on Environment. International Research Journal of Environmental Sciences, 3(3), 72-78.

[5] Peters, G. P., Marland, G., Le Quéré, C., Boden, T., Canadell, J. G., & Raupach, M. R. (2012). Rapid growth in CO2 emissions after the 2008-2009 global financial crisis. Nature climate change, 2(1), 2.

[6] Haszeldine, R. S. (2009). Carbon capture and storage: how green can black be?. Science, 325(5948), 1647-1652.

[7] Gibbins, J., & Chalmers, H. (2008). Carbon capture and storage. Energy policy, 36(12), 4317-4322.

[8] Bowen, F. (2011). Carbon capture and storage as a corporate technology strategy challenge. Energy Policy, 39(5), 2256-2264.

[9] Kim, T. J., Vrålstad, H., Sandru, M., & Hägg, M. B. (2013). Separation performance of PVAm composite membrane for CO 2 capture at various pH levels. Journal of membrane science, 428, 218-224.

[10] Ho, M. T., Allinson, G. W., & Wiley, D. E. (2008). Reducing the cost of CO2 capture from flue gases using pressure swing adsorption. Industrial & Engineering Chemistry Research, 47(14), 4883-4890.

[11] Choi, S., Drese, J. H., & Jones, C. W. (2009). Adsorbent materials for carbon dioxide capture from large anthropogenic point sources. ChemSusChem, 2(9), 796-854.

[12] Rochelle, G. T. (2009). Amine scrubbing for CO2 capture. Science, 325(5948), 1652-1654.

[13] Yu, C. H., Huang, C. H., & Tan, C. S. (2012). A review of CO2 capture by absorption and adsorption. Aerosol Air Qual. Res, 12(5), 745-769.

175

[14] Pfaff, I., Oexmann, J., & Kather, A. (2010). Optimised integration of post-combustion CO2 capture process in greenfield power plants. Energy, 35(10), 4030-4041.

[15] Rao, A. B., & Rubin, E. S. (2002). A technical, economic, and environmental assessment of amine-based CO2 capture technology for power plant greenhouse gas control. Environmental science & technology, 36(20), 4467-4475.

[16] Nanda, S., Mohanty, P., Pant, K. K., Naik, S., Kozinski, J. A., & Dalai, A. K. (2013). Characterization of North American lignocellulosic biomass and biochars in terms of their candidacy for alternate renewable fuels. Bioenergy Research, 6(2), 663-677.

[17] Watanabe, Y., & Hall, D. O. (1996). Photosynthetic CO 2 conversion technologies using a photobioreactor incorporating microalgae-energy and material balances. Energy conversion and management, 37(6), 1321-1326.

[18] Obert, R., & Dave, B. C. (1999). Enzymatic conversion of carbon dioxide to methanol: Enhanced methanol production in silica sol− gel matrices. Journal of the American Chemical Society, 121(51), 12192-12193.

[19] Tong, X., El‐Zahab, B., Zhao, X., Liu, Y., & Wang, P. (2011). Enzymatic synthesis of L‐lactic acid from carbon dioxide and ethanol with an inherent cofactor regeneration cycle. Biotechnology and bioengineering, 108(2), 465-469.

[20] Baskaya, F. S., Zhao, X., Flickinger, M. C., & Wang, P. (2010). Thermodynamic feasibility of enzymatic reduction of carbon dioxide to methanol. Applied biochemistry and biotechnology, 162(2), 391-398.

[21] Barnett, J. A., & Kornberg, H. L. (1960). The utilization by yeasts of acids of the tricarboxylic acid cycle. Microbiology, 23(1), 65-82.

[22] Shiba, H., Kawasumi, T., Igarashi, Y., Kodama, T., & Minoda, Y. (1985). The

CO2 assimilation via the reductive tricarboxylic acid cycle in an obligately autotrophic, aerobic hydrogen-oxidizing bacterium, Hydrogenobacter thermophilus. Archives of Microbiology, 141(3), 198-203.

[23] Cheng, C. H., Du, T. B., Pi, H. C., Jang, S. M., Lin, Y. H., & Lee, H. T. (2011). Comparative study of lipid extraction from microalgae by organic solvent and supercritical CO 2. Bioresource Technology, 102(21), 10151-10153.

[24] Bassham, J. A., Benson, A. A., Kay, L. D., Harris, A. Z., Wilson, A. T., & Calvin, M. (1954). The path of carbon in photosynthesis. XXI. The cyclic regeneration of carbon dioxide acceptor1. Journal of the American Chemical Society, 76(7), 1760-1770.

176

[25] Schwender, J., Goffman, F., Ohlrogge, J. B., & Shachar-Hill, Y. (2004). Rubisco without the Calvin cycle improves the carbon efficiency of developing green seeds. Nature, 432(7018), 779.

[26] Parry, M. A., Andralojc, P. J., Scales, J. C., Salvucci, M. E., Carmo-Silva, A. E., Alonso, H., & Whitney, S. M. (2012). Rubisco activity and regulation as targets for crop improvement. Journal of Experimental Botany, 64(3), 717-730.

[27] Kromdijk, J., & Long, S. P. (2016, March). One crop breeding cycle from starvation? How engineering crop photosynthesis for rising CO2 and temperature could be one important route to alleviation. In Proc. R. Soc. B(Vol. 283, No. 1826, p. 20152578). The Royal Society.

[28] Chung, A. E., & Franzen, J. S. (1969). Oxidized triphosphopyridine nucleotide specific isocitrate dehydrogenase from Azotobacter vinelandii. Isolation and characterization. Biochemistry, 8(8), 3175-3184.

[29] Fukunaga, N., Imagawa, S., Sahara, T., Ishii, A., & Suzuki, M. (1992). Purification and characterization of monomeric isocitrate dehydrogenase with NADP+-specificity from Vibrio parahaemolyticus Y-4. The Journal of Biochemistry, 112(6), 849-855.

[30] LEYLAND, M. L., & KELLY, D. J. (1991). Purification and characterization of a monomeric isocitrate dehydrogenase with dual coenzyme specificity from the photosynthetic bacterium Rhodomicrobium vannielii. The FEBS Journal, 202(1), 85-93.

[31] Steen, I. H., Madsen, M. S., Birkeland, N. K., & Lien, T. (1998). Purification and characterization of a monomeric isocitrate dehydrogenase from the sulfate-reducing bacterium Desulfobacter vibrioformis and demonstration of the presence of a monomeric enzyme in other bacteria. FEMS microbiology letters, 160(1), 75-79.

[32] EiKmanns, B. J., Rittmann, D., & Sahm, H. (1995). Cloning, sequence analysis, expression, and inactivation of the Corynebacterium glutamicum icd gene encoding isocitrate dehydrogenase and biochemical characterization of the enzyme. Journal of bacteriology, 177(3), 774-782.

[33] Kanao, T., Kawamura, M., Fukui, T., Atomi, H., & Imanaka, T. (2002). Characterization of isocitrate dehydrogenase from the green sulfur bacterium Chlorobium limicola. The FEBS Journal, 269(7), 1926-1931.

[34] Chen, R. D., & Gadal, P. (1990). Structure, functions and regulation of NAD and NADP dependent isocitrate dehydrogenases in higher plants and in other organisms. Plant Physiology and Biochemistry (Paris), 28(3), 411-427.

177

[35] Kumar, S. M., Pampa, K. J., Manjula, M., Abdoh, M. M. M., Kunishima, N., & Lokanath, N. K. (2014). Crystal structure studies of NADP+ dependent isocitrate dehydrogenase from Thermus thermophilus exhibiting a novel terminal domain. Biochemical and biophysical research communications, 449(1), 107-113.

[36] Wang, P., Lv, C., & Zhu, G. (2015). Novel type II and monomeric NAD+ specific isocitrate dehydrogenases: phylogenetic affinity, enzymatic characterization, and evolutionary implication. Scientific reports, 5.

[37] Hügler, M., Huber, H., Stetter, K. O., & Fuchs, G. (2003). Autotrophic CO2 fixation pathways in archaea (Crenarchaeota). Archives of microbiology, 179(3), 160-173.

[38] Hawkins, A. B., Adams, M. W., & Kelly, R. M. (2014). Conversion of 4-hydroxybutyrate to acetyl coenzyme A and its anapleurosis in the Metallosphaera sedula 3-hydroxypropionate/4-hydroxybutyrate carbon fixation pathway. Applied and environmental microbiology, 80(8), 2536-2545.

[39] Andreo, C. S., Gonzalez, D. H., & Iglesias, A. A. (1987). Higher plant phosphoenolpyruvate carboxylase: structure and regulation. Febs Letters, 213(1), 1-8.

[40] Liu, R., Liang, L., Wu, M., Chen, K., Jiang, M., Ma, J., ... & Ouyang, P. (2013). CO 2 fixation for succinic acid production by engineered Escherichia coli co-expressing pyruvate carboxylase and nicotinic acid phosphoribosyltransferase. Biochemical engineering journal, 79, 77-83.

[41] Bolivar, J. M., Wilson, L., Ferrarotti, S. A., Fernandez-Lafuente, R., Guisan, J. M., & Mateo, C. (2006). Stabilization of a formate dehydrogenase by covalent immobilization on highly activated glyoxyl-agarose supports. Biomacromolecules, 7(3), 669-673.

[42] Bolivar, J. M., Wilson, L., Ferrarotti, S. A., Fernandez-Lafuente, R., Guisan, J. M., & Mateo, C. (2007). Evaluation of different immobilization strategies to prepare an industrial biocatalyst of formate dehydrogenase from Candida boidinii. Enzyme and microbial technology, 40(4), 540-546.

[43] Tishkov, V. I., & Popov, V. O. (2006). Protein engineering of formate dehydrogenase. Biomolecular engineering, 23(2), 89-110.

[44] Wuensch, C., Pavkov‐Keller, T., Steinkellner, G., Gross, J., Fuchs, M., Hromic, A. Faber, K. (2015). Regioselective Enzymatic β‐Carboxylation of para‐Hydroxy‐ styrene Derivatives Catalyzed by Phenolic Acid Decarboxylases. Advanced synthesis & catalysis, 357(8), 1909-1918.

[45] Supuran, C. T. (2013). Carbonic anhydrases: from biomedical applications of the inhibitors and activators to biotechnological use for CO2 capture. 178

[46] Bond, G. M., Stringer, J., Brandvold, D. K., Simsek, F. A., Medina, M. G., & Egeland, G. (2001). Development of integrated system for biomimetic CO2 sequestration using the enzyme carbonic anhydrase. Energy & Fuels, 15(2), 309-316.

[47] Mirjafari, P., Asghari, K., & Mahinpey, N. (2007). Investigating the application of enzyme carbonic anhydrase for CO2 sequestration purposes. Industrial & engineering chemistry research, 46(3), 921-926.

[48] Savile, C. K., & Lalonde, J. J. (2011). Biotechnology for the acceleration of carbon dioxide capture and sequestration. Current opinion in biotechnology, 22(6), 818-823.

[49] Zhu, X. L., & Sly, W. S. (1990). Carbonic anhydrase IV from human lung. Purification, characterization, and comparison with membrane carbonic anhydrase from human kidney. Journal of Biological Chemistry, 265(15), 8795-8801.

[50] Khalifah, R. G. (1971). The carbon dioxide hydration activity of carbonic anhydrase I. Stop-flow kinetic studies on the native human isoenzymes B and C. Journal of Biological Chemistry, 246(8), 2561-2573.

[51] Penders-van Elk, N. J., Derks, P. W., Fradette, S., & Versteeg, G. F. (2012). Kinetics of absorption of carbon dioxide in aqueous MDEA solutions with carbonic anhydrase at 298K. International Journal of Greenhouse Gas Control, 9, 385-392.

[52] ENICK, R. M., & KLARA, S. M. (1990). CO2 solubility in water and brine under reservoir conditions. Chemical Engineering Communications, 90(1), 23-33.

[53] Alberty, Robert A. Thermodynamics of biochemical reactions. John Wiley & Sons, 2005

[54] Hikita, H., Asai, S., Ishikawa, H., & Honda, M. (1977). The kinetics of reactions of carbon dioxide with monoethanolamine, diethanolamine and triethanolamine by a rapid mixing method. the chemical Engineering Journal, 13(1), 7-12.

[55] Axley, M. J., Böck, A., & Stadtman, T. C. (1991). Catalytic properties of an Escherichia coli formate dehydrogenase mutant in which sulfur replaces selenium. Proceedings of the National Academy of Sciences, 88(19), 8450-8454.

[56] Chen, R. F., & Plaut, G. W. E. (1963). Activation and inhibition of DPN-linked isocitrate dehydrogenase of heart by certain nucleotides. Biochemistry, 2(5), 1023-1032.

[57] Bringer-Meyer, S., Schimz, K. L., & Sahm, H. (1986). Pyruvate decarboxylase from Zymomonas mobilis. Isolation and partial characterization. Archives of microbiology, 146(2), 105-110.

179

[58] Ren, J., Yao, P., Yu, S., Dong, W., Chen, Q., Feng, J., Zhu, D. (2015). An Unprecedented Effective Enzymatic Carboxylation of Phenols. ACS Catalysis, 6(2), 564-567.

[59] Chang, K. S., Jeon, H., Gu, M. B., Pack, S. P., & Jin, E. (2013). Conversion of carbon dioxide to oxaloacetate using integrated carbonic anhydrase and phosphoenolpyruvate carboxylase. Bioprocess and biosystems engineering, 36(12), 1923-1928.

[60] Kumar, S., Tsai, C. J., & Nussinov, R. (2000). Factors enhancing protein thermostability. Protein engineering, 13(3), 179-191.

[61] Gavel, Y., & Heijne, G. V. (1990). Sequence differences between glycosylated and non-glycosylated Asn-X-Thr/Ser acceptor sites: implications for protein engineering. Protein Engineering, Design and Selection, 3(5), 433-442.

[62] Scrutton, N. S., Berry, A., & Perham, R. N. (1990). Redesign of the coenzyme specificity of a dehydrogenase by protein engineering. Nature, 343(6253), 38.

[63] Lehmann, M., Loch, C., Middendorf, A., Studer, D., Lassen, S. F., Pasamontes, L., ... & Wyss, M. (2002). The consensus concept for thermostability engineering of proteins: further proof of concept. Protein engineering, 15(5), 403-411.

[64] Buchholz, P. C., Vogel, C., Reusch, W., Pohl, M., Rother, D., Spieß, A. C., & Pleiss, J. (2016). BioCatNet: a database system for the integration of enzyme sequences and biocatalytic experiments. ChemBioChem, 17(21), 2093-2098.

[65] Hughes, R. A., & Ellington, A. D. (2017). Synthetic DNA Synthesis and Assembly: Putting the Synthetic in Synthetic Biology. Cold Spring Harbor perspectives in biology, 9(1), a023812.

[66] Roney, I. J., Rudner, A. D., Couture, J. F., & Kærn, M. (2016). Improvement of the reverse tetracycline transactivator by single amino acid substitutions that reduce leaky target gene expression to undetectable levels. Scientific reports, 6, 27697.

[67] Luo, J., Meyer, A. S., Mateiu, R. V., & Pinelo, M. (2015). Cascade catalysis in membranes with enzyme immobilization for multi-enzymatic conversion of CO 2 to methanol. New biotechnology, 32(3), 319-327.

[68] Shi, J., Jiang, Y., Jiang, Z., Wang, X., Wang, X., Zhang, S., ... & Yang, C. (2015). Enzymatic conversion of carbon dioxide. Chemical Society Reviews, 44(17), 5981-6000.

[69] Geiger, T., & Clarke, S. (1987). Deamidation, isomerization, and racemization at asparaginyl and aspartyl residues in peptides. Succinimide-linked reactions that contribute to protein degradation. Journal of Biological Chemistry, 262(2), 785-794.

180

[70] Gouda, M. D., Singh, S. A., Rao, A. A., Thakur, M. S., & Karanth, N. G. (2003). Thermal inactivation of glucose oxidase mechanism and stabilization using additives. Journal of Biological Chemistry, 278(27), 24324-24333.

[71] Davies, K. J., Lin, S., & Pacifici, R. E. (1987). Protein damage and degradation by oxygen radicals. IV. Degradation of denatured protein. Journal of Biological Chemistry, 262(20), 9914-9920.

[72] Weber, G., & Drickamer, H. G. (1983). The effect of high pressure upon proteins and other biomolecules. Quarterly reviews of biophysics, 16(1), 89-112.

[73] Perrett, S., & Zhou, J. M. (2002). Expanding the pressure technique: insights into protein folding from combined use of pressure and chemical denaturants. Biochimica et Biophysica Acta (BBA)-Protein Structure and Molecular Enzymology, 1595(1), 210-223.

[74] Royer, C. A. (2002). Revisiting volume changes in pressure-induced protein unfolding. Biochimica et Biophysica Acta (BBA)-Protein Structure and Molecular Enzymology, 1595(1), 201-209.

[75] Frankenberger, W. T., & Johanson, J. B. (1982). Effect of pH on enzyme stability in soils. Soil Biology and Biochemistry, 14(5), 433-437.

[76] Vallee, B. L., & Ulmer, D. D. (1972). Biochemical effects of mercury, cadmium, and lead. Annual review of biochemistry, 41(1), 91-128.

[77] Mateo, C., Palomo, J. M., Fernandez-Lorente, G., Guisan, J. M., & Fernandez-Lafuente, R. (2007). Improvement of enzyme activity, stability and selectivity via immobilization techniques. Enzyme and microbial technology, 40(6), 1451-1463.

[78] Sheldon, R. A. (2007). Enzyme immobilization: the quest for optimum performance. Advanced Synthesis & Catalysis, 349(8‐9), 1289-1307.

[79] Luo, X. L., Xu, J. J., Zhang, Q., Yang, G. J., & Chen, H. Y. (2005). Electrochemically deposited chitosan hydrogel for horseradish peroxidase immobilization through gold nanoparticles self-assembly. Biosensors and Bioelectronics, 21(1), 190-196.

[80] Cao, L., van Rantwijk, F., & Sheldon, R. A. (2000). Cross-linked enzyme aggregates: a simple and effective method for the immobilization of penicillin acylase. Organic Letters, 2(10), 1361-1364.

[81] Takahashi, H., Li, B., Sasaki, T., Miyazaki, C., Kajino, T., & Inagaki, S. (2000). Catalytic activity in organic solvents and stability of immobilized enzymes depend on 181

the pore size and surface characteristics of mesoporous silica. Chemistry of Materials, 12(11), 3301-3305.

[82] Liese, A., & Hilterhaus, L. (2013). Evaluation of immobilized enzymes for industrial applications. Chemical Society Reviews, 42(15), 6236-6249.

[83] Katchalski-Katzir, E., & Kraemer, D. M. (2000). Eupergit® C, a carrier for immobilization of enzymes of industrial potential. Journal of molecular catalysis B: enzymatic, 10(1), 157-176.

[84] Bolivar, J. M., Wilson, L., Ferrarotti, S. A., Fernandez-Lafuente, R., Guisan, J. M., & Mateo, C. (2007). Evaluation of different immobilization strategies to prepare an industrial biocatalyst of formate dehydrogenase from Candida boidinii. Enzyme and microbial technology, 40(4), 540-546.

[85] Guo, Z., Bai, S., & Sun, Y. (2003). Preparation and characterization of immobilized lipase on magnetic hydrophobic microspheres. Enzyme and Microbial Technology, 32(7), 776-782.

[86] D’Annibale, A., Stazi, S. R., Vinciguerra, V., Di Mattia, E., & Sermanni, G. G. (1999). Characterization of immobilized laccase from Lentinula edodes and its use in olive-mill wastewater treatment. Process Biochemistry, 34(6), 697-706.

[87] Chenault, H. K., Simon, E. S., & Whitesides, G. M. (1988). Cofactor regeneration for enzyme-catalysed synthesis. Biotechnology and Genetic Engineering Reviews, 6(1), 221-270.

[88] Wu, J. T., Wu, L. H., & Knight, J. A. (1986). Stability of NADPH: effect of various factors on the kinetics of degradation. Clinical chemistry, 32(2), 314-319.

[89] Pudney, C. R., Hay, S., Levy, C., Pang, J., Sutcliffe, M. J., Leys, D., & Scrutton, N. S. (2009). Evidence to support the hypothesis that promoting vibrations enhance the rate of an enzyme catalyzed H-tunneling reaction. Journal of the American Chemical Society, 131(47), 17072-17073.

[90] Markham, K. A., Sikorski, R. S., & Kohen, A. (2003). Purification, analysis, and preservation of reduced nicotinamide adenine dinucleotide 2′-phosphate. Analytical biochemistry, 322(1), 26-32.

[91] Karimi, M., Hillestad, M., & Svendsen, H. F. (2011). Capital costs and energy considerations of different alternative stripper configurations for post combustion CO 2 capture. Chemical engineering research and design, 89(8), 1229-1236.

[92] Shinoda, N., Tatani, A., Onizuka, M., Okino, S., & Shimizu, T. (1987). U.S. Patent No. 4,696,804. Washington, DC: U.S. Patent and Trademark Office.

182

[93] Singh, D., Croiset, E., Douglas, P. L., & Douglas, M. A. (2003). Techno-economic study of CO 2 capture from an existing coal-fired power plant: MEA scrubbing vs. O 2/CO 2 recycle combustion. Energy Conversion and Management, 44(19), 3073-3091.

[94] Wang, J., S. Zhu, and C. Xu, Biochemistry. 3rd ed. 2002, Beijing: China Higher Education Press.

[95] Wichmann, R., & Vasic-Racki, D. (2005). Cofactor regeneration at the lab scale. Technology transfer in biotechnology, 225-260.

[96] Wong, C. H., & Whitesides, G. M. (1981). Enzyme-catalyzed organic synthesis: NAD (P) H cofactor regeneration by using glucose-6-phosphate and the glucose-5-phosphate dehydrogenase from Leuconostoc mesenteroides. Journal of the American Chemical Society, 103(16), 4890-4899.

[97] Chenault, H. K., Simon, E. S., & Whitesides, G. M. (1988). Cofactor regeneration for enzyme-catalysed synthesis. Biotechnology and Genetic Engineering Reviews, 6(1), 221-270.

[98] Zhao, H., & Van Der Donk, W. A. (2003). Regeneration of cofactors for use in biocatalysis. Current opinion in biotechnology, 14(6), 583-589.

[99] Endo, T., & Koizumi, S. (2001). Microbial conversion with cofactor regeneration using genetically engineered bacteria. Advanced Synthesis & Catalysis, 343(6‐7), 521-526.

[100] Adlercreutz, P. (1996). Cofactor regeneration in biocatalysis in organic media. Biocatalysis and Biotransformation, 14(1), 1-30.

[101] Hummel, W., & KULA, M. R. (1989). Dehydrogenases for the synthesis of chiral compounds. The FEBS Journal, 184(1), 1-13.

[102] Bommarius, A. S., Schwarm, M., Stingl, K., Kottenhahn, M., Huthmacher, K., & Drauz, K. (1995). Synthesis and use of enantiomerically pure tert-leucine. Tetrahedron: Asymmetry, 6(12), 2851-2888.

[103] Hummel, W., Schütte, H., Schmidt, E., Wandrey, C., & Kula, M. R. (1987). Isolation of L-phenylalanine dehydrogenase from Rhodococcus sp. M4 and its application for the production of L-phenylalanine. Applied microbiology and biotechnology, 26(5), 409-416.

[104] Agarwal, A. S., Zhai, Y., Hill, D., & Sridhar, N. (2011). The electrochemical reduction of carbon dioxide to formate/formic acid: engineering and economic feasibility. ChemSusChem, 4(9), 1301-1310.

183

[105] Wong, C. H., & Whitesides, G. M. (1981). Enzyme-catalyzed organic synthesis: NAD (P) H cofactor regeneration by using glucose-6-phosphate and the glucose-5-phosphate dehydrogenase from Leuconostoc mesenteroides. Journal of the American Chemical Society, 103(16), 4890-4899.

[106] Lo, H. C., & Fish, R. H. (2002). Biomimetic NAD+ Models for Tandem Cofactor Regeneration, Horse Liver Alcohol Dehydrogenase Recognition of 1, 4‐ NADH Derivatives, and Chiral Synthesis. Angewandte Chemie International Edition, 41(3), 478-481.

[107] Kragl, U., Kruse, W., Hummel, W., & Wandrey, C. (1996). Enzyme engineering aspects of biocatalysis: cofactor regeneration as example. Biotechnology and bioengineering, 52(2), 309-319.

[108] Boyle, G. (1997). Renewable energy: power for a sustainable future (Vol. 2). OXFORD university press.

[109] Benson, S. M., Bennaceur, K., Cook, P., Davison, J., de Coninck, H., Farhat, K., ... & Wright, I. (2012). Carbon capture and storage. Global Energy Assessment-Toward a Sustainable Future, 993.

[110] Surampalli, R. Y., Zhang, T. C., Tyagi, R., Naidu, R., Gurjar, B., Ojha, C., ... & Kao, C. (2015). Carbon Capture and Storage.

[111] Rubin, E. S., Chen, C., & Rao, A. B. (2007). Cost and performance of fossil fuel power plants with CO 2 capture and storage. Energy policy, 35(9), 4444-4454.

[112] Marston, P. M., & Moore, P. A. (2008). From EOR to CCS: the evolving legal and regulatory framework for carbon capture and storage. Energy LJ, 29, 421.

[113] Farquhar, G. V., von Caemmerer, S. V., & Berry, J. A. (1980). A biochemical model of photosynthetic CO2 assimilation in leaves of C3 species. Planta, 149(1), 78-90.

[114] Bailey, E., Stirpe, F., & Taylor, C. B. (1968). Regulation of rat liver pyruvate kinase. The effect of preincubation, pH, copper ions, fructose 1, 6-diphosphate and dietary changes on enzyme activity. Biochemical Journal, 108(3), 427-436.

[115] Clarke, E. C. W., & Glew, D. N. (1980). Evaluation of Debye–Hückel limiting slopes for water between 0 and 150° C. Journal of the Chemical Society, Faraday Transactions 1: Physical Chemistry in Condensed Phases, 76, 1911-1916.

[116] Goldberg, R. N., Tewari, Y. B., & Bhat, T. N. (2007). Thermodynamics of enzyme-catalyzed reactions: Part 7—2007 update. Journal of Physical and Chemical Reference Data, 36(4), 1347-1397.

184

[117] Tufvesson, P., Jensen, J. S., Kroutil, W., & Woodley, J. M. (2012). Experimental determination of thermodynamic equilibrium in biocatalytic transamination. Biotechnology and bioengineering, 109(8), 2159-2162.

[118] Londesborough, J. C., and K. Dalziel. "The equilibrium constant of the isocitrate dehydrogenase reaction." Biochemical Journal 110.2 (1968): 217-222.

[119] Weiss, R. (1974). Carbon dioxide in water and seawater: the solubility of a non-ideal gas. Marine chemistry, 2(3), 203-215.

[120] Utter, M. F., & Keech, D. B. (1963). Pyruvate carboxylase I. Nature of the reaction. Journal of biological chemistry, 238(8), 2603-2608.

[121] Keech, D. B., & Utter, M. F. (1963). Pyruvate carboxylase II. Properties. Journal of Biological Chemistry, 238(8), 2609-2614.

[122] Linn, T. C., Pettit, F. H., Hucho, F., & Reed, L. J. (1969). α-Keto acid dehydrogenase complexes, XI. Comparative studies of regulatory properties of the pyruvate dehydrogenase complexes from kidney, heart, and liver mitochondria. Proceedings of the National Academy of Sciences, 64(1), 227-234.

[123] Ballard, F. T., & Hanson, R. W. (1967). Phosphoenolpyruvate carboxykinase and pyruvate carboxylase in developing rat liver. Biochemical Journal, 104(3), 866.

[124] Huh, T. L., Ryu, J. H., Huh, J. W., Sung, H. C., Oh, I. U., Song, B. J., & Veech, R. L. (1993). Cloning of a cDNA encoding bovine mitochondrial NADP+-specific isocitrate dehydrogenase and structural comparison with its isoenzymes from different species. Biochemical Journal, 292(3), 705-710.

[125] Nekrutenko, A., Hillis, D. M., Patton, J. C., Bradley, R. D., & Baker, R. J. (1998). Cytosolic isocitrate dehydrogenase in humans, mice, and voles and phylogenetic analysis of the enzyme family. Molecular biology and evolution, 15(12), 1674-1684.

[126] Gálvez, S., & Gadal, P. (1995). On the function of the NADP-dependent isocitrate dehydrogenase isoenzymes in living organisms. Plant Science, 105(1), 1-14.

[127] Sugden, P. H., & Newsholme, E. A. (1975). Activities of citrate synthase, NAD+-linked and NADP+-linked isocitrate dehydrogenases, glutamate dehydrogenase, aspartate aminotransferase and alanine aminotransferase in nervous tissues from vertebrates and invertebrates. Biochemical Journal, 150(1), 105-111.

[128] Ramaley, R. F., & Hudock, M. O. (1973). Purification and properties of isocitrate dehydrogenase (NADP) from Thermus aquaticus YT-1, Bacillus subtilis-168 and Chlamydomonas reinhardti-Y-2. Biochimica et Biophysica Acta (BBA)-Enzymology, 315(1), 22-36.

185

[129] Reeves, H. C., Daumy, G. O., Lin, C. C., & Houston, M. (1972). NADP+-specific isocitrate dehydrogenase of Escherichia coli: I. Purification and characterization. Biochimica et Biophysica Acta (BBA)-Enzymology, 258(1), 27-39.

[130] Steen, I. H., Lien, T., & Birkeland, N. K. (1997). Biochemical and phylogenetic characterization of isocitrate dehydrogenase from a hyperthermophilic archaeon, Archaeoglobus fulgidus. Archives of microbiology, 168(5), 412-420.

[131] Duan, Z., & Li, D. (2008). Coupled phase and aqueous species equilibrium of the H 2 O–CO 2–NaCl–CaCO 3 system from 0 to 250° C, 1 to 1000bar with NaCl concentrations up to saturation of halite. Geochimica et Cosmochimica Acta, 72(20), 5128-5145.

[132] Addo, P. K., Arechederra, R. L., Waheed, A., Shoemaker, J. D., Sly, W. S., & Minteer, S. D. (2011). Methanol production via bioelectrocatalytic reduction of carbon dioxide: role of carbonic anhydrase in improving electrode performance. Electrochemical and Solid-State Letters, 14(4), E9-E13.

[133] Moya, A., Tambutté, S., Bertucci, A., Tambutté, E., Lotto, S., Vullo, D., ... & Zoccola, D. (2008). Carbonic anhydrase in the scleractinian coral Stylophora pistillata characterization, localization, and role in biomineralization. Journal of Biological Chemistry, 283(37), 25475-25484.

[134] Baneyx, F. (1999). Recombinant protein expression in Escherichia coli. Current opinion in biotechnology, 10(5), 411-421.

[135] Donovan, R. S., Robinson, C. W., & Glick, B. R. (1996). Optimizing inducer and culture conditions for expression of foreign proteins under the control of thelac promoter. Journal of industrial microbiology & biotechnology, 16(3), 145-154.

[136] Grossman, T. H., Kawasaki, E. S., Punreddy, S. R., & Osburne, M. S. (1998). Spontaneous cAMP-dependent derepression of gene expression in stationary phase plays a role in recombinant expression instability. Gene, 209(1), 95-103.

[137] Jones, K. L., Kim, S. W., & Keasling, J. D. (2000). Low-copy plasmids can perform as well as or better than high-copy plasmids for metabolic engineering of bacteria. Metabolic engineering, 2(4), 328-338.

[138] Dvorak, P., Chrast, L., Nikel, P. I., Fedr, R., Soucek, K., Sedlackova, M., ... & Damborsky, J. (2015). Exacerbation of substrate toxicity by IPTG in Escherichia coli BL21 (DE3) carrying a synthetic metabolic pathway. Microbial cell factories, 14(1), 201.

[139] Lusk, J. E., Williams, R. J. P., & Kennedy, E. P. (1968). Magnesium and the growth of Escherichia coli. Journal of Biological Chemistry, 243(10), 2618-2624.

186

[140] Larentis, A. L., Nicolau, J. F. M. Q., dos Santos Esteves, G., Vareschini, D. T., de Almeida, F. V. R., dos Reis, M. G., ... & Medeiros, M. A. (2014). Evaluation of pre-induction temperature, cell growth at induction and IPTG concentration on the expression of a leptospiral protein in E. coli using shaking flasks and microbioreactor. BMC research notes, 7(1), 671.

[141] Lee, K. M., Rhee, C. H., Kang, C. K., & Kim, J. H. (2006). Sequential and simultaneous statistical optimization by dynamic design of experiment for peptide overexpression in recombinant Escherichia coli. Applied biochemistry and biotechnology, 135(1), 59-80.

[142] Singh, S. K., Matsuno, K., LaPorte, D. C., & Banaszak, L. J. (2001). Crystal Structure of Bacillus subtilis Isocitrate Dehydrogenase at 1.55 Å insights into the nature of substrate specificity exhibited by escherichia coli isocitrate dehydrogenase kinase/phosphatase. Journal of Biological Chemistry, 276(28), 26154-26163.

[143] Daniel, R. M., Peterson, M. E., Danson, M. J., Price, N. C., Kelly, S. M., Monk, C. R., ... & Lee, C. K. (2010). The molecular basis of the effect of temperature on enzyme activity. Biochemical journal, 425(2), 353-360.

[144] Behrisch, H. W., & Hochachka, P. W. (1969). Temperature and the regulation of enzyme activity in poikilotherms. Properties of rainbow-trout fructose diphosphatase. Biochemical Journal, 111(3), 287-295.

[145] Iyer, P. V., & Ananthanarayan, L. (2008). Enzyme stability and stabilization—aqueous and non-aqueous environment. Process Biochemistry, 43(10), 1019-1032.

[146] Nørby, J. G., & Esmann, M. (1997). The effect of ionic strength and specific anions on substrate binding and hydrolytic activities of Na, K-ATPase. The Journal of general physiology, 109(5), 555-570.

[147] Barratt, D. G., & Cook, R. A. (1978). Role of metal cofactors in enzyme regulation. Differences in the regulatory properties of the Neurospora crassa nicotinamide adenine dinucleotide specific isocitrate dehydrogenase depending on whether Mg2+ or Mn2+ serves as divalent cation. Biochemistry, 17(8), 1561-1566.

[148] Rutter, G. A., & Denton, R. M. (1989). Rapid purification of pig heart NAD+-isocitrate dehydrogenase. Studies on the regulation of activity by Ca2+, adenine nucleotides, Mg2+ and other metal ions. Biochemical Journal, 263(2), 445-452.

[149] KUCHEL, P. W., REYNOLDS, C. H., & DALZIEL, K. (1980). Determination of the Stability Constants of Mn2+ and Mg2+ Complexes of the Components of the

187

NADP‐Linked Isocitrate Dehydrogenase Reaction by Electron Spin Resonance. The FEBS Journal, 110(2), 465-473.

[150] Dinelli, G., Civitano, L., & Rea, M. (1990). Industrial experiments on pulse corona simultaneous removal of NO/sub x/and SO/sub 2/from flue gas. IEEE Transactions on Industry Applications, 26(3), 535-541.

[151] Aaron, D., & Tsouris, C. (2005). Separation of CO2 from flue gas: a review. Separation Science and Technology, 40(1-3), 321-348.

[152] Zevenhoven, R., & Kilpinen, P. (2001). Control of pollutants in flue gases and fuel gases (pp. 951-22). Espoo, Finland: Helsinki University of Technology.

[153] Prescott, D. J., & Rabinowitz, J. L. (1968). The Enzymatic Carboxylation of Propionyl Coenzyme A STUDY INVOLVING DEUTERATED AND TRITIATED SUBSTRATES. Journal of Biological Chemistry, 243(7), 1551-1557.

[154] Gameiro, P. A., Laviolette, L. A., Kelleher, J. K., Iliopoulos, O., & Stephanopoulos, G. (2013). Cofactor balance by nicotinamide nucleotide transhydrogenase (NNT) coordinates reductive carboxylation and glucose catabolism in the tricarboxylic acid (TCA) cycle. Journal of Biological Chemistry, 288(18), 12967-12977.

[155] Dalziel, K., & Londesborough, J. C. (1968). The mechanisms of reductive carboxylation reactions. Carbon dioxide or bicarbonate as substrate of nicotinamide-adenine dinucleotide phosphate-linked isocitrate dehydrogenase and ‘malic’enzyme. Biochemical Journal, 110(2), 223-230.

[156] Gameiro, P. A., Laviolette, L. A., Kelleher, J. K., Iliopoulos, O., & Stephanopoulos, G. (2013). Cofactor balance by nicotinamide nucleotide transhydrogenase (NNT) coordinates reductive carboxylation and glucose catabolism in the tricarboxylic acid (TCA) cycle. Journal of Biological Chemistry, 288(18), 12967-12977.

[157] Maren, T. H. (1967). Carbonic anhydrase: chemistry, physiology, and inhibition. Physiological Reviews, 47(4), 595-781.

[158] Dalziel, K. (1975). 1 Kinetics and Mechanism of Nicotinamide-Nucleotid-Linked Dehydrogenases. The enzymes, 11, 1-60.

[159] Lin, H., Zhang, M., Wang, F., Meng, F., Liao, B. Q., Hong, H., ... & Gao, W. (2014). A critical review of extracellular polymeric substances (EPSs) in membrane bioreactors: characteristics, roles in membrane fouling and control strategies. Journal of Membrane science, 460, 110-125.

188

[160] Yang, W., Cicek, N., & Ilg, J. (2006). State-of-the-art of membrane bioreactors: Worldwide research and commercial applications in North America. Journal of membrane Science, 270(1), 201-211.

[161] Kimura, K., Yamato, N., Yamamura, H., & Watanabe, Y. (2005). Membrane fouling in pilot-scale membrane bioreactors (MBRs) treating municipal wastewater. Environmental science & technology, 39(16), 6293-6299.

[162] Thomas, C. R., & Geer, D. (2011). Effects of shear on proteins in solution. Biotechnology letters, 33(3), 443-456.

[163] Wieser, Marco, et al. "Carbon dioxide fixation by reversible pyrrole‐2‐ carboxylate decarboxylase from Bacillus megaterium PYR2910." The FEBS Journal 257.2 (1998): 495-499.

[164] Wieser, Marco, Toyokazu Yoshida, and Toru Nagasawa. "Carbon dioxide fixation by reversible pyrrole-2-carboxylate decarboxylase and its application." Journal of Molecular Catalysis B: Enzymatic 11.4 (2001): 179-184.

[165] Glueck, Silvia M., et al. "Biocatalytic carboxylation." Chemical Society Reviews 39.1 (2010): 313-328.

[166] Rao, PV Subba, K. Moore, and G. H. N. Towers. "The conversion of tryptophan to 2, 3-dihydroxybenzoic acid and catechol by Aspergillus niger." Biochemical and biophysical research communications 28.6 (1967): 1008-1012.

[167] Anderson, Jeffrey John, and S. T. A. N. L. E. Y. Dagley. "Catabolism of aromatic acids in Trichosporon cutaneum." Journal of bacteriology 141.2 (1980): 534-543.

[168] Wuensch, Christiane, et al. "Regioselective enzymatic carboxylation of phenols and hydroxystyrene derivatives." Organic letters 14.8 (2012): 1974-1977.

[169] Métraux, Jean-Pierre. "Recent breakthroughs in the study of salicylic acid biosynthesis." Trends in plant science 7.8 (2002): 332-334.

[170] Deng, Youquan, et al. "Clean synthesis of adipic acid by direct oxidation of cyclohexene with H 2 O 2 over peroxytungstate–organic complex catalysts." Green Chemistry 1.6 (1999): 275-276.

[171] Kamath, Ajith V., Dipak Dasgupta, and C. S. Vaidyanathan. "Enzyme-catalysed non-oxidative decarboxylation of aromatic acids: I. Purification and spectroscopic properties of 2, 3 dihydroxybenzoic acid decarboxylase from Aspergillusniger." Biochemical and biophysical research communications 145.1 (1987): 586-595.

189

[172] Janknecht, Ralf, et al. "Rapid and efficient purification of native histidine-tagged protein expressed by recombinant vaccinia virus." Proceedings of the National Academy of Sciences 88.20 (1991): 8972-8976.

[173] Hoffmann, A., and R. G. Roeder. "Purification of his-tagged proteins in non-denaturing conditions suggests a convenient method for protein interaction studies." Nucleic Acids Research 19.22 (1991): 6337.

[174] El‐Zahab, B., Donnelly, D., & Wang, P. (2008). Particle‐tethered NADH for production of methanol from CO2 catalyzed by coimmobilized enzymes. Biotechnology and bioengineering, 99(3), 508-514.

[175] Huff, C. A., & Sanford, M. S. (2011). Cascade catalysis for the homogeneous hydrogenation of CO2 to methanol. Journal of the American Chemical Society, 133(45), 18122-18125.

[176] Beh, M., Strauss, G., Huber, R., Stetter, K. O., & Fuchs, G. (1993). Enzymes of the reductive citric acid cycle in the autotrophic eubacterium Aquifex pyrophilus and in the archaebacterium Thermoproteus neutrophilus. Archives of Microbiology, 160(4), 306-311.

[177] Xie, H. B., Zhou, Y., Zhang, Y., & Johnson, J. K. (2010). Reaction mechanism of monoethanolamine with CO2 in aqueous solution from molecular modeling. The Journal of Physical Chemistry A, 114(43), 11844-11852.

[178] Jana, A., Halder, S. K., Banerjee, A., Paul, T., Pati, B. R., Mondal, K. C., & Mohapatra, P. K. D. (2014). Biosynthesis, structural architecture and biotechnological potential of bacterial tannase: a molecular advancement. Bioresource technology, 157, 327-340.

[179] Colman, R. F. (1975). Mechanisms for the oxidative decarboxylation of isocitrate: implications for control. Advances in enzyme regulation, 13, 413-433.

[180] Costello, L. C., Franklin, R. B., Liu, Y., & Kennedy, M. C. (2000). Zinc causes a shift toward citrate at equilibrium of the m-aconitase reaction of prostate mitochondria. Journal of inorganic biochemistry, 78(2), 161-165.

[181] Shi, J., Wang, X., Jiang, Z., Liang, Y., Zhu, Y., & Zhang, C. (2012). Constructing spatially separated multienzyme system through bioadhesion-assisted bio-inspired mineralization for efficient carbon dioxide conversion. Bioresource technology, 118, 359-366.

[182] Kacser, H. (1952). The specificity and kinetics of aconitase. Biochimica et biophysica acta, 9, 406-415.

190

[183] Robbins, A. H., & Stout, C. D. (1989). The structure of aconitase. Proteins: Structure, Function, and Bioinformatics, 5(4), 289-312.

[184] Beinert, H. E. L. M. U. T. (1990). Recent developments in the field of iron-sulfur proteins. The FASEB Journal, 4(8), 2483-2491.

[185] Beinert, H., Kennedy, M. C., & Stout, C. D. (1996). Aconitase as iron− sulfur protein, enzyme, and iron-regulatory protein. Chemical Reviews, 96(7), 2335-2374.

[186] Bogusławski, W., Klimek, J., Tiałowska, B., & Żelewski, L. (1976). Inhibition by Mn2+ of citrate supported progesterone biosynthesis in mitochondrial fractions of human term placentae. Journal of steroid biochemistry, 7(1), 39-44.

[187] Rabe, T., Brandstetter, K., Kellermann, J., & Runnebaum, B. (1982). Partial characterization of placental 3ß-hydroxysteroid dehydrogenase (EC 1.1. 1.145), Δ4– 5isomerase (EC 5.3. 3.1) in human term placental mitochondria. Journal of steroid biochemistry, 17(4), 427-433.

[188] Garcia‐Galan, C., Berenguer‐Murcia, Á., Fernandez‐Lafuente, R., & Rodrigues, R. C. (2011). Potential of different enzyme immobilization strategies to improve enzyme performance. Advanced Synthesis & Catalysis, 353(16), 2885-2904.

[189] Sheldon, R. A., Schoevaart, R., & Van Langen, L. M. (2005). Cross-linked enzyme aggregates (CLEAs): A novel and versatile method for enzyme immobilization (a review). Biocatalysis and Biotransformation, 23(3-4), 141-147.

[190] Juang, R. S., Wu, F. C., & Tseng, R. L. (2002). Use of chemically modified chitosan beads for sorption and enzyme immobilization. Advances in Environmental Research, 6(2), 171-177.

[191] Ravikovitch, P. I., & Neimark, A. V. (2002). Density functional theory of adsorption in spherical cavities and pore size characterization of templated nanoporous silicas with cubic and three-dimensional hexagonal structures. Langmuir, 18(5), 1550-1560.

[192] Pandya, P. H., Jasra, R. V., Newalkar, B. L., & Bhatt, P. N. (2005). Studies on the activity and stability of immobilized α-amylase in ordered mesoporous silicas. Microporous and Mesoporous Materials, 77(1), 67-77.

[193] Suzuki, K., Ikari, K., & Imai, H. (2003). Synthesis of mesoporous silica foams with hierarchical trimodal pore structures. Journal of Materials Chemistry, 13(7), 1812-1816.

[194] Stein, A. (2003). Advances in microporous and mesoporous solids—highlights of recent progress. Advanced Materials, 15(10), 763-775.

191

[195] Kumar, M. N. R. (2000). A review of chitin and chitosan applications. Reactive and functional polymers, 46(1), 1-27.

[196] Krajewska, B. (2004). Application of chitin-and chitosan-based materials for enzyme immobilizations: a review. Enzyme and microbial technology, 35(2), 126-139.

[197] Guibal, E. (2005). Heterogeneous catalysis on chitosan-based materials: a review. Progress in Polymer Science, 30(1), 71-109.

[198] Peniche, C., Argüelles‐Monal, W., Peniche, H., & Acosta, N. (2003). Chitosan: an attractive biocompatible polymer for microencapsulation. Macromolecular Bioscience, 3(10), 511-520.

[199] Hui, Y., Nan, L., Jing‐Zhong, X., & Jun‐Jie, Z. (2005). A glucose biosensor based on immobilization of glucose oxidase in chitosan network matrix. Chinese Journal of Chemistry, 23(3), 275-279.

[200] Bayramoğlu, G., Yilmaz, M., & Arica, M. Y. (2004). Immobilization of a thermostable α-amylase onto reactive membranes: kinetics characterization and application to continuous starch hydrolysis. Food Chemistry, 84(4), 591-599.

[201] Simsek, F. A., Bond, G. M., & Stringer, J. (2001). Immobilization of carbonic anhydrase for biomimetic CO2 sequestration. World Resource Review, 13(1), 74.

[202] Hartmann, M., & Jung, D. (2010). Biocatalysis with enzymes immobilized on mesoporous hosts: the status quo and future trends. Journal of Materials Chemistry, 20(5), 844-857.

[203] Yiu, H. H., & Wright, P. A. (2005). Enzymes supported on ordered mesoporous solids: a special case of an inorganic–organic hybrid. Journal of Materials Chemistry, 15(35-36), 3690-3700.

[204] Cao, L., & Schmid, R. (2005). D. Carrier-Bound Immobilized Enzymes.

[205]Hanefeld, U., Gardossi, L., & Magner, E. (2009). Understanding enzyme immobilisation. Chemical Society Reviews, 38(2), 453-468.

[206] Hoffmann, F., Cornelius, M., Morell, J., & Fröba, M. (2006). Silica‐based mesoporous organic–inorganic hybrid materials. Angewandte Chemie International Edition, 45(20), 3216-3251.

[207] Monier, M., Ayad, D. M., Wei, Y., & Sarhan, A. A. (2010). Immobilization of horseradish peroxidase on modified chitosan beads. International Journal of Biological Macromolecules, 46(3), 324-330.

192

[208] López-León, T., Carvalho, E. L. S., Seijo, B., Ortega-Vinuesa, J. L., & Bastos-González, D. (2005). Physicochemical characterization of chitosan nanoparticles: electrokinetic and stability behavior. Journal of Colloid and Interface Science, 283(2), 344-351.

[209] Rampino, A., Borgogna, M., Blasi, P., Bellich, B., & Cesàro, A. (2013). Chitosan nanoparticles: preparation, size evolution and stability. International journal of pharmaceutics, 455(1), 219-228

[210] Rodrigues, S., da Costa, A. M. R., & Grenha, A. (2012). Chitosan/carrageenan nanoparticles: Effect of cross-linking with tripolyphosphate and charge ratios. Carbohydrate Polymers, 89(1), 282-289.

[211] Ulbrich, R., Golbik, R., & Schellenberger, A. (1991). Protein adsorption and leakage in carrier–enzyme systems. Biotechnology and bioengineering, 37(3), 280-287.

[212] Cao, L., van Rantwijk, F., & Sheldon, R. A. (2000). Cross-linked enzyme aggregates: a simple and effective method for the immobilization of penicillin acylase. Organic Letters, 2(10), 1361-1364.

[213] Sheldon, R. A., Schoevaart, R., & Van Langen, L. M. (2005). Cross-linked enzyme aggregates (CLEAs): A novel and versatile method for enzyme immobilization (a review). Biocatalysis and Biotransformation, 23(3-4), 141-147.

[214] Bornscheuer, U. T. (2003). Immobilizing enzymes: how to create more suitable biocatalysts. Angewandte Chemie International Edition, 42(29), 3336-3337.

[215] Rodrigues, R. C., Ortiz, C., Berenguer-Murcia, Á., Torres, R., & Fernández-Lafuente, R. (2013). Modifying enzyme activity and selectivity by immobilization. Chemical Society Reviews, 42(15), 6290-6307.

[216] Scouten, W. H., Luong, J. H., & Brown, R. S. (1995). Enzyme or protein immobilization techniques for applications in biosensor design. Trends in biotechnology, 13(5), 178-185.

[217] Ramaley, R. F., & Hudock, M. O. (1973). Purification and properties of isocitrate dehydrogenase (NADP) from Thermus aquaticus YT-1, Bacillus subtilis-168 and Chlamydomonas reinhardti-Y-2. Biochimica et Biophysica Acta (BBA)-Enzymology, 315(1), 22-36.

[218] Coolbear, T., Daniel, R. M., & Morgan, H. W. (1992). The enzymes from extreme thermophiles: bacterial sources, thermostabilities and industrial relevance. In Enzymes and Products from Bacteria Fungi and Plant Cells (pp. 57-98). Springer Berlin Heidelberg.

193

[219] Thomas, C. R., Nienow, A. W., & Dunnill, P. (1979). Action of shear on enzymes: studies with alcohol dehydrogenase. Biotechnology and bioengineering, 21(12), 2263-2278.

[220] Maa, Y. F., & Hsu, C. C. (1996). Effect of high shear on proteins. Biotechnology and bioengineering, 51(4), 458-465.

[221] Thomas, C. R., & Dunnill, P. (1979). Action of shear on enzymes: studies with catalase and urease. Biotechnology and bioengineering, 21(12), 2279-2302.

[222] Charm, S. E., & Wong, B. L. (1981). Shear effects on enzymes. Enzyme and Microbial Technology, 3(2), 111-118.

[223] Eijsink, V. G., Gåseidnes, S., Borchert, T. V., & van den Burg, B. (2005). Directed evolution of enzyme stability. Biomolecular engineering, 22(1), 21-30.

[224] Reetz, M. T. (2013). Biocatalysis in organic chemistry and biotechnology: past, present, and future. Journal of the American Chemical Society, 135(34), 12480-12496.

[225] Boot-Handford, M. E., Abanades, J. C., Anthony, E. J., Blunt, M. J., Brandani, S., Mac Dowell, N., ... & Haszeldine, R. S. (2014). Carbon capture and storage update. Energy & Environmental Science, 7(1), 130-189.

[226] Yu, C. H., Huang, C. H., & Tan, C. S. (2012). A review of CO2 capture by absorption and adsorption. Aerosol Air Qual. Res, 12(5), 745-769.

[227] Leach, A., Mason, C. F., & van‘t Veld, K. (2011). Co-optimization of enhanced oil recovery and carbon sequestration. Resource and Energy Economics, 33(4), 893-912.

[228] Freeman, S. A., Dugas, R., Van Wagener, D. H., Nguyen, T., & Rochelle, G. T. (2010). Carbon dioxide capture with concentrated, aqueous piperazine. International Journal of Greenhouse Gas Control, 4(2), 119-124.

[229] Po, H. N., & Senozan, N. M. (2001). The Henderson-Hasselbalch equation: its history and limitations. J. Chem. Educ, 78(11), 1499.

[230] Rao, A. B., & Rubin, E. S. (2002). A technical, economic, and environmental assessment of amine-based CO2 capture technology for power plant greenhouse gas control. Environmental science & technology, 36(20), 4467-4475.

[231] MacDowell, N., Florin, N., Buchard, A., Hallett, J., Galindo, A., Jackson, G., ... & Fennell, P. (2010). An overview of CO 2 capture technologies. Energy & Environmental Science, 3(11), 1645-1669.

194

[232] Rubin, E. S., Chen, C., & Rao, A. B. (2007). Cost and performance of fossil fuel power plants with CO 2 capture and storage. Energy policy, 35(9), 4444-4454.

[233] David, J., & Herzog, H. (2000, August). The cost of carbon capture. In fifth international conference on greenhouse gas control technologies, Cairns, Australia (pp. 13-16).

[234] Singh, D., Croiset, E., Douglas, P. L., & Douglas, M. A. (2003). Techno-economic study of CO2 capture from an existing coal-fired power plant: MEA scrubbing vs. O 2/CO 2 recycle combustion. Energy Conversion and Management, 44(19), 3073-3091.

[235] Wu, J. T., Wu, L. H., & Knight, J. A. (1986). Stability of NADPH: effect of various factors on the kinetics of degradation. Clinical chemistry, 32(2), 314-319.

[236] Markham, K. A., Sikorski, R. S., & Kohen, A. (2003). Purification, analysis, and preservation of reduced nicotinamide adenine dinucleotide 2′-phosphate. Analytical biochemistry, 322(1), 26-32.

[237] Kragl, U., Kruse, W., Hummel, W., & Wandrey, C. (1996). Enzyme engineering aspects of biocatalysis: cofactor regeneration as example. Biotechnology and bioengineering, 52(2), 309-319.

[238] Ahn, S. T., Abu-Baker, I., & Palmore, G. T. R. (2017). Electroreduction of CO2 on polycrystalline copper: Effect of temperature on product selectivity. Catalysis Today, 288, 24-29.

[239] Wang, Q., Dong, H., & Yu, H. (2014). Fabrication of a novel tin gas diffusion electrode for electrochemical reduction of carbon dioxide to formic acid. RSC Advances, 4(104), 59970-59976.

[240] Pletcher, D. (2015). The cathodic reduction of carbon dioxide—What can it realistically achieve? A mini review. Electrochemistry Communications, 61, 97-101.

[241] Lu, X., Leung, D. Y., Wang, H., Leung, M. K., & Xuan, J. (2014). Electrochemical reduction of carbon dioxide to formic acid. ChemElectroChem, 1(5), 836-849.

[242] Damen, K., van Troost, M., Faaij, A., & Turkenburg, W. (2006). A comparison of electricity and hydrogen production systems with CO 2 capture and storage. Part A: Review and selection of promising conversion and capture technologies. Progress in energy and combustion science, 32(2), 215-246.

[243] Icarus Process Evaluator. Icarus Process Evaluator User Manual. Cambridge, MA, US: Aspen Technology Limited; 2001.

195

[244] Kantzas, A., Chatzis, I., & Dullien, F. A. L. (1988, January). Enhanced oil recovery by inert gas injection. In SPE Enhanced Oil Recovery Symposium. Society of Petroleum Engineers.

[245] Rogers, J. D., & Grigg, R. B. (2001). A literature analysis of the WAG injectivity abnormalities in the CO2 process. SPE Reservoir Evaluation & Engineering, 4(05), 375-386.

[246] Barradale, M. J. (2010). Impact of public policy uncertainty on renewable energy investment: Wind power and the production tax credit. Energy Policy, 38(12), 7698-7709.

[247] Johnstone, N., Haščič, I., & Popp, D. (2010). Renewable energy policies and technological innovation: evidence based on patent counts. Environmental and resource economics, 45(1), 133-155.

[248] Beck, F., & Martinot, E. (2004). Renewable energy policies and barriers. Encyclopedia of energy, 5(7), 365-383.

196