A Dissertation Entitled Microalgae Fractionation and Production Of

Home , Ricinoleic acid

A Dissertation

entitled

Microalgae Fractionation and Production of High Value Nylon Precursors

Godwin Ameh Abel

Submitted to the Graduate Faculty as partial fulfillment of the requirements for the

Doctor of Philosophy Degree in

Engineering

______Dr. Sasidhar Varanasi, Committee Chair

______Dr. Sridhar Viamajala, Committee Member

______Dr. Kana Yamamoto, Committee Member

______Dr. Maria Coleman, Committee Member

______Dr. Patricia Relue, Committee Member

______Dr. Amanda Bryant-Friedrich, Dean College of Graduate Studies

The University of Toledo

August 2017

This document is copyrighted material. Under copyright law, no parts of this document may be reproduced without the expressed permission of the author. An Abstract of

Microalgae Fractionation and Production of High Value Nylon Precursors

Godwin Ameh Abel

Submitted to the Graduate Faculty as partial fulfillment of the requirements for the Doctor of Philosophy Degree in Engineering

The University of Toledo

August 2017

Liquid fuels from microalgal biomass have become less attractive recently due to the fall in prices of petroleum and natural gas. However, interest in microalgae as a renewable feedstock for value-added bioproducts such as oleo-chemicals and sugar- derived platform molecules has been on the rise. This is due to the economic and environmental benefits associated with the processing of these higher value products from microalgae biomass. Microalgae biomass consists mainly of carbohydrate, protein and lipids. When efficiently fractionated, the lipid and carbohydrate portions could be used for synthesizing oleochemicals and sugar-based platform molecules, respectively.

The protein rich residue could serve as feed for animals.

This study is a part of the microalgae processing methodologies being explored in our group, with the purpose of developing an integrated approach for producing renewable high value chemicals from microalgae via environmentally sustainable pathways. For this purpose, a specific pathway for processing microalgae after it is harvested following cultivation is proposed. The proposed pathway entails subjecting harvested microalgae to enzymatic digestion to fractionate and affect depolymerization of

iii the individual fractions: Simple sugars resulting from the hydrolysis of the carbohydrate fraction can be converted to value-added products via fermentation using product-specific microorganisms. Similarly, oleic acid, the major component of the lipid fraction, can form the feed-stock for the production of value-added compounds such as industrial nylon precursors.

As conventional (chemical/physical) methods of isolating carbohydrates from microalgae are deemed unsustainable, in our first project we explored an enzymatic hydrolysis method for isolation of simple sugars directly from microalgae biomass.

Lipids are isolated from the enzymatically digested algal slurry through immiscible solvent extraction. Following phase separation, the aqueous phase retains the oligomeric/monomeric sugars from carbohydrates and the protein fragments, and serves as a tailor-made broth for converting the sugars to value-added platform molecules via fermentation methods. Alternately, the monomeric sugars can be selectively extracted from this medium and can separately be converted into specific chemicals/fuels through appropriate chemical conversion strategies. This new fractionation method lowers cost and energy required over the traditional approaches due to the mild processing conditions and allows for the isolation of the simple sugars in high yield.

Our second project provides the synthesis of precursors for nylon 12 and 13 from methyl oleate/microalgal lipids. These nylons, developed in the 1980s for use in the automotive industries, are high-strength polymers that find applications in many industrial sectors (medical, electronic and sport industries). The production of these nylons currently uses petroleum feedstock or exotic fatty acids and involves 4 – 6 steps.

We have developed a new simple 2-step procedure featuring use of cross metathesis

(CM) reaction as the key step. This approach starts with cross-metathesis (CM) of methyl oleate (methyl ester of oleic acid) with alkenyl nitriles (allyl cyanide or homoallyl cyanide), to form cyano esters. This step is followed by hydrogenation of the unsaturated alkene and nitrile groups of the cyano-esters to produce the desired amino esters (nylon precursors).

We have also developed a ring-closing metathesis approach as an alternative pathway to cross-metathesis for producing nylon 12 precursors. This alternative pathway had been extended for the synthesis of precursors for nylons 11, 12, and 13 from oleic acid, the most abundant natural fatty acid in microalgae. The alternative pathway is also simple, but 3-step procedure featuring use of ring-closing metathesis (RCM) reaction as the key step. The first step (amide formation) and third step of this approach

(hydrogenation) can be performed with no major issues. Therefore, only the key ring- closing metathesis step was optimized. This second strategy (ring closing metathesis approach) involves first converting the oleic acid to alkenyl amides. Cross-coupling of free amines is not possible because of catalyst poisoning of amines. The amides were then subjected to ring closing metathesis generating the unsaturated lactams, which are then hydrogenated to lactams. This approach avoids the use of high-pressure hydrogenation and produces fewer undesired by-products than CM approach.

This dissertation is dedicated to my beloved parents (Abel Ameh and Anthonia Ameh), as a tribute to their sacrifices, prayers, support, encouragement and unceasing love.

Acknowledgements

This dissertation could not have been possible without the grace and blessings of

God, acceptance and guidance of my advisors and committee members, assistance from lab mates and friends, and prayers and support from my parents, brothers and sisters.

To my advisors, Drs. Sasidhar Varanasi and Sridhar Viamajala. I would like to express my deepest gratitude for accepting me into their group and giving me the opportunity to perform research. I would like to thank them for their guidance, genuine care, patience, and providing the necessary atmosphere for research.

To my committee members, Drs. Kana Yamamoto, Maria Coleman and Patricia

Relue. Thank you so much for your guidance, understanding and willingness to participate in my defense committee. Your time is very much appreciated.

To all of my friends, lab mates and postdocs, Christopher Anukwu, Engr. Patrick

Nwokolo, Dr. Ajith Yapa, Yaser Shirazi, Xiaofei Zhao, Matin Hanifzadeh, Ravi Gogar,

Arsalan Sepehri, Alessandra Krusciel, Jayachandra Kopalli, Drs. Brahmaiah Pendyala,

Jehad Almaleety, and Pramod Poudel. Thank you so much for your friendship, support and guidance. It was my pleasure to work with you guys.

To my beloved parents (Abel Ameh and Anthonia Ameh), brothers and sisters, thank you so much. They were always supporting me with their prayers and best wishes.

Finally, I would like to thank God almighty for making everything beautiful in

His time. To Him be the glory. vi

Table of Contents

Abstract…………………………… ...... iii

Acknowledgements…………...... vi

Table of Contents………… ...... vii

List of Tables………...... x

List of Figures…………...... xi

List of Abbreviations……………...... xiii

1 Downstream Processing of Microalgae Biomass ...... 1

1.1 Background ...... 1

1.1.1 Microalgae ...... 1

1.1.2 Advantages of using microalgae for the production of bioproducts……………… ...... 2

1.1.3 Microalgae harvesting and biomass concentration………… ...... 2

1.1.4 Microalgae cell disruption ...... 3

1.2 Use of renewable feedstock in the production of bioproducts ...... 5

1.2.1 Nylons ...... 7

1.2.2 Olefin metathesis ...... 11

1.3 Dissertation overview ...... 12

References for chapter 1…………………...... 15

vii

2 Microalgae Fractionation and Recovery of Native Components through

Application of Low-Cost Enzymes…………...... 23

2.1 Introduction ...... 23

2.2 Materials and methods ...... 25

2.2.1 Algae biomass ...... 25

2.2.2 Enzymatic digestion ...... 25

2.2.3 Total carbohydrates ...... 26

2.2.4 Elemental (CHN) and total protein analysis ...... 27

2.2.5 Fatty acid methyl esters (FAMEs) ...... 27

2.2.5 Lipid extraction after enzymatic digestion ...... 28

2.2.7 Ash content ...... 28

2.2.8 Analytical methods ...... 28

2.3 Results and discussion ...... 30

2.3.1 Microalgae biomass composition ...... 30

2.3.2 Enzymatic digestion and sugar yield after enzymatic digestion ...... 31

2.3.3 Lipid extraction after enzymatic digestion ...... 32

2.3.4 Protein distribution after enzymatic digestion...... ………….……...33

2.3.5 Enzyme stability……………………………………………………34

2.3.6 Techno-economic analysis…………………………………….…...35

2.4 Conclusions…………………………………………………………………..38

References for chapter 2…………………...... 39

viii

3 Cross-Metathesis Approach to Produce Precursors of Nylon 12 and Nylon 13 from Microalgae……………………………………………………………...………….43

3.1 Introduction ...... 43

3.2 Materials and methods ...... 45

3.2.1 General ...... 45

3.2.2 Experimental procedures ...... 46

3.3 Results and discussion ...... 54

3.4 Conclusions ...... 67

References for chapter 3…………………...... 69

4 Toward Sustainable Synthesis of PA12 (Nylon 12) Precursor from Oleic Acid

Using Ring-Closing Metathesis ...... 73

4.1 Introduction ...... 73

4.2 Materials and methods ...... 75

4.2.1 General ...... 75

4.2.2 Experimental procedures ...... 76

4.3 Results and discussion ...... 78

4.4 Conclusions ...... 94

References for chapter 4…………………...... 96

5 Summary……...... 103

References…………………...... 106

A Supplemental information accompanying Chapter 3 ...... 126

List of Tables

1.1 Comparison of cell disruption methods in terms of key aspects ...... 4

2.1 SLA-04 biomass composition ...... 30

2.2 Enzymatic fractionation and dilute acid pretreatment methods experimental parameters ...... 37

2.3 Variable operation cost for processing one ton of microalgae biomass ...... 37

3.1 Cross-metathesis of methyl 9-decenoate with allyl cyanide ...... 58

3.2 Solvent screening on cross metathesis of 9-decenoate with allyl cyanide ...... 59

3.3 Cross-metathesis of methyl oleate with allyl cyanide...... 61

3.4 Cross-metathesis of 9-decenoate with homoallyl cyanide ...... 62

3.5 Cross-metathesis of methyl oleate with homoallyl cyanide ...... 63

3.6 Hydrogenation of methyl 11-cyanoundec-9-enaote and 12-cyanododec-9- enoate...... 65

4.1 Ring-closing metathesis of homoallyloleamide (2) – solvent comparison ...... 83

4.2 Ring-closing metathesis of homoallyloleamide (2) – temperature comparison ....86

4.3 Stability evaluation of the selected Ru catalysts ...... 90

4.4 Demonstration of recovery and reuse of C9 catalyst immobilized on silica gel ...94

List of Figures

1-1 Downstream processing of microalgae biomass ...... 5

1-2 Microalgae inputs and outputs ...... 7

1-3 Current steps in the production of nylon 12 precursor ...... 9

1-4 Current steps in the production of nylon 11 precursor ...... 10

1-5 Olefin metathesis reaction...... 11

1-6 Common olefin metathesis Ru-catalysts ...... 12

1-7 Chauvin olefin metathesis mechanism ...... 12

2-1 Enzymatic hydrolysis of 10% (w/v) microalgae biomass solution with a reaction time of 2 h for different enzyme cocktails and their combinations ...... 32

2-2 Protein distribution after enzymatic digestion using protease (325U/g-biomass) and glucoamylase (169U/g-biomass) ...... 34

2-3 Stability of enzyme combination, protease (325U/g-biomass) and glucoamylase

(169U/g-biomass) in a fed-batch catalysis mode of usage ...... 35

3-1 Cross-metathesis approach for the synthesis of nylon precursors ...... 45

3-2 (a) Previously reported cross metathesis approach. (b) Our approach to nylon precursors from methyl oleate ...... 55

3-3 Mass spectrum of the cross metathesis of methyl dec-9-enoate and allyl cyanide 56

3-4 Chromatograms of the cross metathesis of methyl dec-9-enoate and allyl cyanide without and with benzoquinone ...... 57 xi

3-5 (a) A crude algal lipid containing a mixture of FAMEs. (b) Cross-metathesis with acrylonitrile. (c) Cross-metathesis with allyl cyanide ...... 67

4-1 Commercially available metathesis catalysts screened in this study ...... 81

4-2 Ring-closing metathesis of homoallyloleamide (2) – catalyst comparison when the highest conversion is achieved...... 88

A-1 13C NMR spectrum for Methyl 12-aminododecanoate (5) ...... 126

A-2 1H NMR spectrum for Methyl 12-aminododecanoate (5) ...... 127

A-3 13C NMR spectrum for Methyl 13-aminotridecanoate (6) ...... 128

A-4 1H NMR spectrum for Methyl 13-aminotridecanoate (6) ...... 128

A-5 HRM spectrum for Methyl 13-aminotridecanoate (6) ...... 129

A-6 13C NMR spectrum for (E/Z) – 11-Cyano-9-dodecenoic acid methyl ester (9) ..130

A-7 1H NMR spectrum for (E/Z) – 11-Cyano-9-dodecenoic acid methyl ester (9) ....130

A-8 HRM spectrum for (E/Z) – 11-Cyano-9-dodecenoic acid methyl ester (9) ...... 131

A-9 13C NMR spectrum for (E/Z) – 12-Cyano-9-tridecenoic acid methyl ester (10) .132

A-10 1H NMR spectrum for (E/Z) – 12-Cyano-9-tridecenoic acid methyl ester (10) ..132

A-11 HRM spectrum for (E/Z) – 12-Cyano-9-tridecenoic acid methyl ester (10) ...... 133

A-12 GC chromatogram of FAMEs after reactive extraction of microalgae ...... 134

A-13 GC chromatogram of cross metathesis (algal lipids with acrylonitrile) ...... 134

A-14 GC chromatogram of cross metathesis (algal lipids with allyl cyanide) ...... 135

xii

List of Abbreviations

CM ...... Cross-metathesis

DCM ...... Dichloromethane DCW ...... Dry cell weight DSC ...... Differential Scanning Calorimeter

EtOAc ...... Ethyl acetate

FAME ...... Fatty Acid Methyl Esters FID ...... Flame Ionization Detector

GC ...... Gas Chromatography GHG ...... Greenhouse Gas

HPLC ...... High Performance Liquid Chromatography

NHC ...... N-Heterocyclic carbene

PA ...... Polyamide

RCM ...... Ring-closing Metathesis

SLA ...... Soap Lake Algae

TCD...... Thermaconductivity Detector TGA ...... Thermogravimetric Analysis TLC ...... Thin Layer Chromatography

xiii

Chapter 1

Downstream Processing of Microalgae Biomass

1.1 Background

Biomass is a term used to describe biological materials which can be used as a source of bioenergy/fuel. Lignocellulosic biomass (terrestrial plants) comprises of lignin, hemicellulose and cellulose, e.g. agricultural residues, forest residues, wood waste, animal waste, sewage sludge. Non-lignocellulosic biomass includes oilseeds and microalgae.

1.1.1 Microalgae

Microalgae are photosynthetic microorganisms and can be prokaryotic (e.g. cyanobacteria) or eukaryotic (e.g. chlorophyta and diatoms) [1, 2] and are present in diverse environments [3]. Microalgae can serve as feedstocks in the production of value- added products (such as fine chemicals, pharmaceutical, food, cosmetics) as well as bulk commodities (e.g. biofuels and oleochemicals) in a bio-refinery. In addition to photoautotrophic growth (where energy derived from light is converted to chemical energy through photosynthetic reactions), microalgae can also be grown heterotrophically

(where organic compounds can be used as a source of carbon and energy) or

1 mixotrophically (where both organic compounds and CO2 are utilized, with photosynthesis as the major source of energy) [4].

Microalgae has five growth phases: (1) lag or induction phase (zero growth rate);

(2) exponential growth phase, indicating maximum growth rate under the specific conditions; (3) linear growth phase (increasing growth rate); (4) stationary phase (zero growth rate); (5) decline or death phase (negative growth rate). The growth phase can be affected by so many factors such as quality and quantity of light, temperature, nutrients, pH, salinity, CO2, O2, pathogens (bacteria), as well as operational factors such as mixing, addition of bicarbonates etc.

1.1.2 Advantages of using microalgae for production of bioproducts

Advantages of using microalgae biomass as feedstock in the production of bioproducts have been reported in many previous studies [5-17]. Microalgae can be grown easily using waste water and nutrients, and on low-quality land [17]. While microalgae typically grow slowly in natural environments, their growth rate can be increased by supplementation of certain nutrients and aeration [18-20]. As such, the productivity of microalgae can be much higher than terrestrial plants/agricultural feedstocks (such as rapeseed, soybeans) in well-designed reactors [15]. In addition, microalgae can remove CO2 from flue gases mitigating greenhouse gas (GHG) emissions

[21]. They can be used in waste water treatment as they utilize the contaminants as nutrients for their growth [22].

1.1.3 Microalgae harvesting and biomass concentration

Recovery of microalgae biomass from the culture media is termed harvesting and accounts for about 20 to 30% of the total production cost [23]. This can be done using

2 physical, chemical or biological methods. Most commonly used techniques include centrifugation, filtration, sedimentation, ultrafiltration and/or flocculation. The appropriate technique(s) could be selected depending on the desired product quality or moisture level.

1.1.4 Microalgae cell disruption

Microalgae components are not recovered effectively by applying techniques conventionally designed for food crops such as soybeans because of their small size and thick cell walls. The complex nature of microalgae cell wall also has significant effect on the disruption efficiency [24, 25]. Relationship between cell wall compositions, disruption efficiency and energy consumption has not been extensively explored.

To maximize product recovery, an efficient cell disruption is important. A feasible energy-efficient cell disruption method would comprise less operating cost, high yield of the product, and high quality of the recovered products.

The techniques for microalgae cell disruption can be broadly classified as mechanical and non-mechanical [26]. Mechanical techniques include bead milling, high speed homogenization, high pressure homogenization, ultrasonication, microwave and pulsed electric field. Non-mechanical techniques include chemical and enzymatic methods. These methods are compared in Table 1.1 below.

Table 1.1. Comparison of cell disruption methods in terms of key aspects (adapted from

[26]).

Selective Optimum Disruption Energy Practical Mildness product DCW Repeatability method consumption scalability recovery concentration Bead milling Yes/no No Conc. High/medium Yes High High Yes/no No Diluted/conc. High/medium Yes High pressure homogenization High No No Diluted High/medium Yes High/medium speed homogenizer Ultrasound Yes/no No Diluted Medium/low Yes/no Medium Microwave Yes/no No Diluted High/medium Yes/no Medium Enzymatic Yes Yes Diluted Low Yes High lysis Chemical Yes/no Yes Diluted/conc. Medium/low Yes High treatment Pulsed Very High/medium Yes/no No Yes/no medium dil./diluted /low electric field

In this study, we explored an enzymatic fractionation method for isolation of simple sugars (cell disruption and carbohydrate hydrolysis) directly from microalgae biomass (Figure 1-1). Lipids could be isolated from the enzymatically digested algal slurry through immiscible solvent extraction. Following phase separation, the aqueous phase retains the oligomeric/monomeric sugars from carbohydrates and the protein fragments, and could serve as a tailor-made broth for converting the sugars to value- added platform molecules via fermentation methods. Alternately, the monomeric sugars can be selectively extracted from this medium and can separately be converted into specific chemicals/fuels through appropriate chemical conversion strategies. Lipids can be used as renewable feedstock in the production of fuels or higher value bioproducts such as nylons.

Figure 1-1. Downstream processing of microalgae biomass.

1.2 Use of renewable feedstock in the production of bioproducts

Synthesis of commodity chemicals from renewable sources is gaining a lot of interest for several reasons. Conventional oil production, as predicted by a recent theoretical model, is predicted to follow a bell shape curve, with peak production expected to occur between the years 2010 and 2025 [27]. On the other hand, the demand for oil is growing due to increase in population. To cater for the growing demand with depleting petroleum resources, there’s a need to replace or supplement energy and materials traditionally manufactured from petroleum feedstock with alternatives from renewable resources. Specifically, biomass is the only renewable carbon-based resource that can serve as a substitute for petroleum. However, out of the 170 billion tons of

5 biomass produced worldwide annually, only 6 billion ton (3.5%) were utilized in the form of timber, grain, and other foodstuffs [28]. Current global research is focused on the production of additional bio-based materials such as bio-polymers, bio-surfactants, and bio-lubricants increased over the last few decades. These bio-based materials have lower environmental impact when compared to petroleum based materials [29, 30].

Algae biomass has three major components: carbohydrates, lipids and proteins

(Figure 1-2). Lipids are especially useful for the production of bio-fuels, bio-surfactants, bio-polymer, and bio-lubricants. Polyesters [31, 32], polyamides [33], polyurethanes

[34], polyepoxides [35], and polycarbonates [32] are some examples of bioproducts reported from lipids [36]. Despite the progress in these bio-based technologies, chemical industries have not been able to replace existing technologies, due to the high cost associated with bio-based feedstock/processing. Therefore, innovative technologies that are economically feasible are still required. A significant portion of the work presented in this thesis is related to the synthesis of lipid-derived monomers for the production of polymers.

Figure 1-2. Microalgae inputs and outputs.

1.2.1 Nylons

Polymers are macromolecules that can be natural (proteins, polysaccharides, and gums), synthetic (thermoplastic, thermosetting or elastomers). Natural rubber was developed as the first major industrial polymer in the 19th century [37]. Of the synthetic polymers, nylons are one of most widely used and are polyamides with the general formula–[(CH2)n-CONH]− or –[(CH2)n-CONH-(CH2)mNHCO]−. They are named according to the number of carbon atoms separating the amide linkages. In particular, nylon 11 and 12 possess excellent chemical and mechanical properties and are widely used in automotive, medical, and electronic industries as well as for sport equipment, back panel for solar cells, and lenses for glasses. Nylon 13 is reported to have similar

7 properties to nylon 12. These nylons are produced from amino acids or their derivatives with a corresponding number (11, 12 or 13) of carbons in their backbone.

Currently, the precursor of nylon 12 is manufactured from petroleum-derived butadiene in a six-step process (Figure 1-3) [38]. However, there is an increasing interest in use of renewable sources for production of these amino acid precursors, because of concerns over environmental sustainability of petrochemical products [39]. As such, synthetic approaches that use natural fatty acids and esters from plant- or algae-derived feedstocks are especially attractive [40-42]. Among the natural fatty acids, oleic acid is the predominant component of lipids in a large variety of vegetable oils (e.g. canola) as well as oleaginous microalgae [43].

Figure 1-3. Current steps in the production of nylon 12 precursor.

The precursor of nylon 11 is manufactured from ricinoleic acid, a natural fatty acid available only in castor beans (Figure 1-4) [38]. In this method, the acid is first converted to its methyl ester and then subjected to pyrolysis to produce 10-undecyleate, which is then converted to the nylon 11 precursor in two additional steps. Several other multi-step (>4 steps) approaches to produce nylon 11 precursors from either ricinoleic acid or oleic acid derivatives are reported in patents [44-48] and journals [49-56].

Figure 1-4. Current steps in the production of nylon 11 precursor.

While there is one study on a concise three-step metathesis-based approach for producing nylon 11 from oleic acid [51], the method has previously not been applied for producing nylons 12 or 13, likely due to low selectivity (discussed in more detail in

Chapter 3). The other reported methods to produce nylons 12 and 13 from non- petrochemical sources require several steps and exotic fatty acids as starting materials.

For example, there are several reported nylon 12 precursor syntheses that use ricinoleic acid from castor beans as the feedstock and require 4–6 steps [49, 50, 54, 55, 57].

Synthesis of nylon 13 has not been fully studied except for one reported approach that involved several steps from either erucic or lesquerolic acid, another atypical acid available from rapeseed oil and mustard oil[58]. In this approach, esterification of erucic acid to methyl erucate is followed by oxidative ozonolysis to produce half ester of

10 brassylic acid. The brassylic acid was then successfully converted to amide ester, nitrile ester, and amino ester. Finally, hydrolysis of the amino ester produces 13- aminotridecanoic acid.

In this study, we have developed the synthesis of nylon 12 and 13 precursors from oleic acid, a major component of microalagal lipids via olefin metathesis as the key step.

Cross-metathesis and ring closing metathesis approaches were used.

1.2.2 Olefin metathesis

Olefin metathesis (Figure 1-5) is a chemical transformation that can redistribute the carbenes or carbynes of alkenes or alkynes by scission and regeneration of the double or triple carbon bonds [59].

Figure 1-5. Olefin metathesis reaction

First reported in the 1950s, the name “olefin metathesis” was coined by Calderon and the team who were working in The Goodyear Tire and Rubber Company in Akron,

Ohio [60]. Olefin metathesis reactions such as cross-metathesis (CM), ring-closing metathesis (RCM), ring-opening metathesis polymerization (ROMP), and acyclic diene metathesis (ADMET) are very essential synthetic tools employed today. A wide variety of organometallic catalysts have been developed, which can be use in olefin metathesis reactions. Alkylidene complexes of tungsten, molybdenum, and ruthenium (Figure 1-6) are prominent examples. Though well-defined ruthenium-based catalysts are the most commonly used olefin metathesis catalysts due to their higher substrate functional group tolerance, stability, and high reactivity [61]. Chauvin described the first viable

11 mechanism (Figure 1-7) for olefin metathesis. The reaction begins with the formation of a metal carbene, proceeds through a [2+2] cycloaddition forming the metallocyclobutane intermediate and a reverse [2+2] to give the product.

Figure 1-6. Common olefin metathesis Ru-catalysts.

Figure 1-7. Chauvin olefin metathesis mechanism.

1.3 Dissertation overview

The study is a part of the microalgae processing methodologies being explored in our group, with the purpose of developing an integrated approach for producing

12 renewable high value chemicals from microalgae via environmentally sustainable pathways. Our overall goal is the production of chemicals of industrial value from algae/algae constituents. For this purpose a strategy for processing microalgae after it is harvested from the cultivation media is studied. In our strategy, harvested microalgae is subjected to enzymatic digestion to deconstruct and fractionate microalgae cell constituents. Simple sugars resulting from the hydrolysis of the carbohydrate fraction can be converted to value-added products via fermentation approaches involving product- specific microorganisms. Also, lipids can form the feed-stock for the production of industrial nylon precursors. Two new strategies for the production of nylon 12 -13 precursors from oleic acid are described.

In Chapter 2, we describe an enzymatic hydrolysis method for recovery of monomeric sugars and lipids from microalgae biomass. This method uses low cost industrial enzymes and requires less energy in comparison to traditional approaches due to the mild processing conditions.

Chapter 3 describes the synthesis of precursors for nylon 12 and 13 from methyl oleate and mixed microalgae lipids. We have developed a new simple 2-step procedure featuring use of cross metathesis (CM) reaction as the key step. This approach starts with cross-metathesis (CM) of methyl oleate (methyl ester of oleic acid) with alkenyl nitriles

(allyl cyanide or homoallyl cyanide), an inexpensive, abundant and renewable material to form cyano esters. This step is followed by hydrogenation of the unsaturated alkene and nitrile groups of the cyano-esters to produce the desired amino esters (nylon precursors).

In Chapter 4, we have demonstrated a simple 3-step ring-closing metathesis approach for the synthesis of precursors for nylons 11, 12, and 13. In this strategy, oleic

13 acid was first converted to alkenyl amides (step 1) since cross-coupling of free amines is not possible because of catalyst poisoning of amines. The amides were then subjected to ring closing metathesis (step 2) to generate unsaturated lactams, which are then hydrogenated to lactams (step 3). This approach avoids the use of high-pressure hydrogenation and produces fewer undesired by-products. The first step (amide formation) and third step of this approach (hydrogenation) were carried out using standard chemistry protocols and the study focused on optimizing the key second step that involved ring-closing metathesis.

Chapter 5 contains the summary of our accomplishments and future work.

References for chapter 1

1. Li, Y., et al., Effects of nitrogen sources on cell growth and lipid accumulation of green alga Neochloris oleoabundans. Applied microbiology and biotechnology, 2008.

81(4): p. 629-636.

2. Li, Y., et al., Biofuels from microalgae. Biotechnology progress, 2008. 24(4): p.

815-820.

3. Richmond, A., Handbook of microalgal culture: biotechnology and applied phycology. 2008: John Wiley & Sons.

4. Chojnacka, K. and F.-J. Marquez-Rocha, Kinetic and stoichiometric relationships of the energy and carbon metabolism in the culture of microalgae. Biotechnology, 2004.

3(1): p. 21-34.

5. Jacob-Lopes, E., L.M.C.F. Lacerda, and T.T. Franco, Biomass production and carbon dioxide fixation by Aphanothece microscopica Nägeli in a bubble column photobioreactor. Biochemical engineering journal, 2008. 40(1): p. 27-34.

6. Murakami, M., et al., The biological CO 2 fixation using Chlorella sp. with high capability in fixing CO 2. Studies in Surface Science and Catalysis, 1998. 114: p. 315-

320.

7. Bilanovic, D., et al., Freshwater and marine microalgae sequestering of CO 2 at different C and N concentrations–response surface methodology analysis. Energy

Conversion and Management, 2009. 50(2): p. 262-267.

8. Jacob-Lopes, E., et al., Effect of light cycles (night/day) on CO 2 fixation and biomass production by microalgae in photobioreactors. chemical engineering and

Processing: Process Intensification, 2009. 48(1): p. 306-310.

9. Hodaifa, G., M.E. Martínez, and S. Sánchez, Use of industrial wastewater from olive-oil extraction for biomass production of Scenedesmus obliquus. Bioresource technology, 2008. 99(5): p. 1111-1117.

10. Tsukahara, K. and S. Sawayama, Liquid fuel production using microalgae. J Jpn

Petrol Inst, 2005. 48(5): p. 251-259.

11. Hu, Q., et al., Microalgal triacylglycerols as feedstocks for biofuel production: perspectives and advances. The plant journal, 2008. 54(4): p. 621-639.

12. Schenk, P.M., et al., Second generation biofuels: high-efficiency microalgae for biodiesel production. Bioenergy research, 2008. 1(1): p. 20-43.

13. Rosenberg, J.N., et al., A green light for engineered algae: redirecting metabolism to fuel a biotechnology revolution. Current opinion in Biotechnology, 2008. 19(5): p.

430-436.

14. Rodolfi, L., et al., Microalgae for oil: Strain selection, induction of lipid synthesis and outdoor mass cultivation in a low‐cost photobioreactor. Biotechnology and bioengineering, 2009. 102(1): p. 100-112.

15. Chisti, Y., Biodiesel from microalgae. Biotechnology advances, 2007. 25(3): p.

294-306.

16. Hossain, A.S., et al., Biodiesel fuel production from algae as renewable energy.

American journal of biochemistry and biotechnology, 2008. 4(3): p. 250-254.

17. Sheehan, J., et al., A look back at the US Department of Energy’s aquatic species program: biodiesel from algae. National Renewable Energy Laboratory, 1998. 328.

18. Pratoomyot, J., P. Srivilas, and T. Noiraksar, Fatty acids composition of 10 microalgal species. Songklanakarin J Sci Technol, 2005. 27(6): p. 1179-1187.

19. Aslan, S. and I.K. Kapdan, Batch kinetics of nitrogen and phosphorus removal from synthetic wastewater by algae. Ecological Engineering, 2006. 28(1): p. 64-70.

20. Renaud, S.M., L.-V. Thinh, and D.L. Parry, The gross chemical composition and fatty acid composition of 18 species of tropical Australian microalgae for possible use in mariculture. Aquaculture, 1999. 170(2): p. 147-159.

21. EU-Commission, Directive 2003/30/EC of the European Parliament and of the

Council of 8 May 2003 on the promotion of the use of biofuels or other renewable fuels for transport. Official Journal of the European Union, 2003. 5.

22. Wang, B., et al., CO2 bio-mitigation using microalgae. Applied microbiology and biotechnology, 2008. 79(5): p. 707-718.

23. Grima, E.M., et al., Recovery of microalgal biomass and metabolites: process options and economics. Biotechnology advances, 2003. 20(7): p. 491-515.

24. Scholz, M.J., et al., Ultrastructure and composition of the Nannochloropsis gaditana cell wall. Eukaryotic cell, 2014. 13(11): p. 1450-1464.

25. Gerken, H.G., B. Donohoe, and E.P. Knoshaug, Enzymatic cell wall degradation of Chlorellavulgaris and other microalgae for biofuels production. Planta, 2013. 237(1): p. 239-253.

26. Günerken, E., et al., Cell disruption for microalgae biorefineries. Biotechnology advances, 2015. 33(2): p. 243-260.

27. Mohr, S.H. and G.M. Evans, Peak oil: Testing Hubbert's curve via theoretical modeling, in Nat. Resour. Res. (N. Y., NY, U. S.). 2008, Springer. p. 1-11.

28. Warwel, S., et al., Polymers and surfactants on the basis of renewable resources.

Chemosphere, 2001. 43(1): p. 39-48.

29. Hermann, B.G., K. Blok, and M.K. Patel, Producing bio-based bulk chemicals using industrial biotechnology saves energy and combats climate change. Environ. Sci.

Technol., 2007. 41(22): p. 7915-7921.

30. Tokiwa, Y., et al., Biodegradability of plastics. Int. J. Mol. Sci., 2009. 10(9): p.

3722-3742.

31. Ebata, H., K. Toshima, and S. Matsumura, Lipase-catalyzed synthesis and curing of high-molecular-weight polyricinoleate. Macromol. Biosci., 2007. 7(6): p. 798-803.

32. More, A.S., et al., Aliphatic polycarbonates and poly(ester carbonate)s from fatty acid derived monomers. Polym. Chem., 2011. 2(12): p. 2796-2803.

33. Kreye, O., et al., Structurally Diverse Polyamides Obtained from Monomers

Derived via the Ugi Multicomponent Reaction. Chem. - Eur. J., 2012. 18(18): p. 5767-

5776, S5767/1-S5767/64.

34. More, A.S., et al., AB type polyaddition route to thermoplastic polyurethanes from fatty acid derivatives. Polym. Chem., 2012. 3(6): p. 1594-1605.

35. Stemmelen, M., et al., A fully biobased epoxy resin from vegetable oils: From the synthesis of the precursors by thiol-ene reaction to the study of the final material. J.

Polym. Sci., Part A: Polym. Chem., 2011. 49(11): p. 2434-2444.

36. Lebarbe, T., et al., Methyl 10-undecenoate as a raw material for the synthesis of renewable semi-crystalline polyesters and poly(ester-amide)s. Polym. Chem., 2012.

3(10): p. 2842-2851.

37. Scott, G., Polymers in modern life. In polymers and the Environment. 1999: The

Royal Society of Chemistry.

38. Chauvel, A. and G. Lefebvre, Petrochemical processes, ed. E. Technip. 1989.

39. Ragauskas, A.J., et al., The path forward for biofuels and biomaterials. Science

(Washington, DC, U. S.), 2006. 311(5760): p. 484-489.

40. Hannon, M., et al., Biofuels from algae: challenges and potential. Biofuels, 2010.

1(5): p. 763-784.

41. Murzin, D.Y., P. Maeki-Arvela, and D.A.G. Aranda, Processing microalgae: beyond lipids. Biofuels, 2014. 5(1): p. 29-32.

42. Mata, T.M., A.A. Martins, and N.S. Caetano, Microalgae for biodiesel production and other applications: A review. Renewable Sustainable Energy Rev., 2009. 14(1): p.

217-232.

43. Behr, A., A. Westfechtel, and J. Perez Gomes, Catalytic processes for the technical use of natural fats and oils. Chem. Eng. Technol., 2008. 31(5): p. 700-714.

44. Dubois, J.-L., Method for synthesizing 9-aminononanoic acid or the esters thereof from natural unsaturated fatty acids. 2010, Arkema France, Fr. p. 22pp.; Chemical

Indexing Equivalent to 152:168701 (FR).

45. Couturier, J.-L. and J.-L. Dubois, Process for the synthesis of C11 and C12 omega-aminoalkanoic acid esters comprising a nitrilation step. 2013, Arkema France, Fr.

. p. 21pp.; Chemical Indexing Equivalent to 158:332172 (FR).

46. Couturier, J.-L. and J.-L. Dubois, Preparation of α,ω-amino acids or esters using cross-metathesis. 2013, Arkema France, Fr. p. 20pp.; Chemical Indexing Equivalent to

158:301890 (WO).

47. Dubois, J.-L. and J.-L. Couturier, Process for synthesis of omega-unsaturated nitrile acid/ester in which consecutively two types of swing cross metathesis processes are alternated. 2013, Arkema France, Fr. p. 39pp.; Chemical Indexing Equivalent to

161:186797 (WO).

48. Couturier, J.-L. and J.-L. Dubois, Preparation of an ω-amino acid or ester from a monounsaturated fatty acid or ester. 2013, Arkema France, Fr. p. 24pp.; Chemical

Indexing Equivalent to 158:301891 (WO).

49. Malacea, R., et al., Renewable materials as precursors of linear nitrile-acid derivatives via cross-metathesis of fatty esters and acids with acrylonitrile and fumaronitrile. Green Chem., 2009. 11(2): p. 152-155.

50. Miao, X., et al., Polyamide precursors from renewable 10-undecenenitrile and methyl acrylate via olefin cross-metathesis. Green Chem., 2012. 14(8): p. 2179-2183.

51. Miao, X., et al., Ruthenium-alkylidene catalyzed cross-metathesis of fatty acid derivatives with acrylonitrile and methyl acrylate: a key step toward long-chain bifunctional and amino acid compounds. Green Chem., 2011. 13(10): p. 2911-2919.

52. Rybak, A. and M.A.R. Meier, Cross-metathesis of fatty acid derivatives with methyl acrylate: renewable raw materials for the chemical industry. Green Chem., 2007.

9(12): p. 1356-1361.

53. Rybak, A. and M.A.R. Meier, Cross-metathesis of oleyl alcohol with methyl acrylate: optimization of reaction conditions and comparison of their environmental impact. Green Chem., 2008. 10(10): p. 1099-1104.

54. Miao, X., et al., Tandem catalytic acrylonitrile cross-metathesis and hydrogenation of nitriles with ruthenium catalysts: direct access to linear α,ω-aminoesters from renewables. ChemSusChem, 2012. 5(8): p. France1410-1414, S1410/1-S1410/7.

55. Miao, X., et al., Ruthenium-Benzylidenes and Ruthenium-Indenylidenes as

Efficient Catalysts for the Hydrogenation of Aliphatic Nitriles into Primary Amines.

ChemCatChem, 2012. 4(12): p. 1911-1916, S1911/1-S1911/5.

56. Kajetanowicz, A., A. Sytniczuk, and K. Grela, Metathesis of renewable raw materials-influence of ligands in the indenylidene type catalysts on self-metathesis of methyl oleate and cross-metathesis of methyl oleate with (Z)-2-butene-1,4-diol diacetate.

Green Chem., 2014. 16(3): p. 1579-1585.

57. Ternel, J., et al., Rhodium-Catalyzed Tandem Isomerization/Hydroformylation of the Bio-Sourced 10-Undecenenitrile: Selective and Productive Catalysts for Production of Polyamide-12 Precursor. Adv. Synth. Catal., 2013. 355(16): p. 3191-3204.

58. Greene, J.L., Jr., R.E. Burks, Jr., and I.A. Wolff, 13-Aminotridecanoic acid from erucic acid. Ind. Eng. Chem., Prod. Res. Develop., 1969. 8(2): p. 171-6.

59. Astruc, D., The metathesis reactions: from a historical perspective to recent developments. New J. Chem., 2005. 29(1): p. 42-56.

60. Calderon, N., H.Y. Chen, and K.W. Scott, Olefin metathesis, a novel reaction for skeletal transformations of unsaturated hydrocarbons. Tetrahedron Lett., 1967(34): p.

3327-9.

61. Dias, E.L., S.T. Nguyen, and R.H. Grubbs, Well-Defined Ruthenium Olefin

Metathesis Catalysts: Mechanism and Activity. J. Am. Chem. Soc., 1997. 119(17): p.

3887-3897.

Chapter 2 Microalgae Fractionation and Recovery of Native Components through Application of Low-Cost Enzymes

2.1 Introduction

There is increasing global concern regarding climate change related to growing demand for fossil carbon sources. Biofuels and bioproducts derived from renewable biomass sources have the potential to sustainably displace fossil carbon use. Microalgae are promising biomass feedstocks due to their high growth rates [1]. Microalgae transform light to chemical energy (biomass) more efficiently (0.5–2.0% efficiency) than energy crops such as switch grass (0.2% efficiency) [2, 3]. Microalgae production also does not compete with food crops for land and freshwater[1, 4]. However, bio-based production using microalgae carbohydrates has not yet been commercialized, in part, due to challenges associated with downstream conversion methods [5-7].

Microalgae contain three primary biopolymers - carbohydrates, proteins and lipids. When separately recovered, the carbohydrates yield monomeric sugars for fermentation (e.g. to ethanol, butanol or succinic acid) or other chemical conversion processes (e.g. to hydroxymethyl furfural or levulinic acid). Lipids can be converted into fuels (e.g. biodiesel and hydrocarbons) or oleochemicals. Finally, protein can serve as animal feed or as a source of amino acids for the human health industry. To recover

23 microalgae components, however, cells must first be disrupted. The cell disruption step can be done mechanically through bead milling[8], ultrasonication [9, 10], high pressure homogenization [11, 12], or high speed homogenization [13]. However, mechanical processes are capital- and energy- intensive, and volumetric scale-up is likely to prove challenging [14]. Alternately, thermal treatments including microwave[15] or chemical/thermochemical methods using acid[16, 17] or alkali [18] reactions can be employed, but also suffer from high energy inputs and loss of product and product quality. For example, thermochemical methods degrade protein to forms that are not usable for animal/human consumption[14]. Carbohydrates also degrade to furans that are toxic to fermentative organisms. As a result of the drawbacks of other methods, enzymatic cell disruption has gained interest [19] due to low energy consumption and high product quality. Enzymatic methods for microalgae cell disruption use mild reaction conditions and produce low (or no) by-products due to enzyme specificity and high selectivity. In addition, the mild temperature and pH reaction conditions employed in enzymatic processes eliminates maintenance cost associated with corrosion of equipment occurring in thermochemical processes.

Microalgae cell wall degradation and release of carbohydrate in the form of monomeric sugars have been reported using different enzymes and their combinations with varying sugar yields [19-21]. In one study, hydrolysis of carbohydrate was performed by Rodrigues and Bon (2011) using a combination of cellulose, xylanase and amylase enzymes on Chlorella sp. A moderate yield of 67.5% and 53.2% were obtained following a pretreatment step that involved dehydration and grinding of the biomass. In another study, Choi et al., 2010 reported high sugar yield (94%), though, via a two-step

24 method. Liquefaction of biomass by α-amylase was followed by saccharification using glucoamylase. Lee et al. (2013) reported 80.9% and 71.3% sugar yield on residual biomass after lipid extraction using AMG 300L and Viscozyme L, respectively. These studies were on the recovery of carbohydrates. These studies, however, did not consider the fate of other microalgae components, lipids and protein.

Herein, we present our studies aimed at a whole cell microalgae fractionation.

Enzymatic digestion of microalgae biomass to release sugars is followed by lipid extraction, and leaving behind a protein rich residue. First, cell disruption and carbohydrate hydrolysis (enzymatic fractionation) is combined in a single step using combination of low-cost industrial enzymes (protease, α-amylase and glucoamylase). The three enzymes’ cocktails were evaluated to identify the enzyme that are sufficiently stable to promote the reaction with good activity. A potential concern of this method would be the cost of enzymatic cocktails. Hence, the ability to reuse the enzymes was also investigated.

2.2 Materials and Methods

2.2.1 Algal biomass

Chlorella sorokiniana str. SLA-04 obtained from Soap Lake in the State of

Washington (USA) was used in this study and cultivated under outdoor conditions [22].

2.2.2 Enzymatic digestion

The digestion experiments were performed at a solid loading of ~9% (w/w) with

10 mL working volume (1g biomass). The enzymatic digestions were performed in a buffer solution containing 50 mM citric acid adjusted to a pH of 4.5 using a 10 M NaOH solution. Protease, α-amylase and glucoamylase purchased from Sigma-Aldrich and their

25 combinations were used. The amount of each enzyme was fixed as follows: 650µL

(325U/g-biomass) protease; 65 mg (1950U/g-biomass) of α-amylase, and 65µL (169U/g- biomass) of glucoamylase based on results from our preliminary screening data [23]. One protease unit is the amount of enzyme which hydrolyzes 1 µmol of L-leucine-p- nitroanilide per minute; one α-amylase unit corresponds to the amount of enzyme which liberates 1 µmol maltose per minute at pH 6.0 and 25 ℃; and one glucoamylase unit is defined as the amount of enzyme which cleaves 1µmol of maltose per min at pH 4.3 and

25 ℃. Enzymatic digestions were performed in sealed 50 mL Erlenmeyer flasks agitated at 200 rpm in a shaking water bath (Model C76, New Brunswick Scientific, NJ, USA) for

2 h at 55 ℃. At the end of the reaction, the supernatant (liquid) was collected after centrifugation of the reaction mixture and analyzed by HPLC as described in Section

2.2.7.1. The post-digestion solids (residue) were washed with deionized water, freeze dried and the total carbohydrate was determined as described in section 2.2.3 to close mass balances.

2.2.3 Total carbohydrates

Total carbohydrate (both structural carbohydrate and storage starch) present in the microalgae biomass and post-digestion residue was estimated according to NREL

Laboratory Analytical procedure [24]. 25±2.5 mg (actual weight recorded) of freeze- dried microalgae biomass was added to a 15 mL autoclavable glass tube. 0.25 mL of 72%

H2SO4 solution was then added to the glass tube containing the biomass. The glass tube was capped and placed in a water bath for 60 min at 30 ºC with mixing at regular intervals (10 min) using a vortex. After 60 min, 7 mL of deionized water was added to the reaction mixture bringing the acid concentration down to 4% before capping tightly to

26 prevent evaporation and placing in an autoclave at 121 ºC. After 60 min of autoclaving, the glass tubes containing the reaction mixture were removed and cooled to room temperature. A 3 mL aliquot (for faster/easier neutralization) of the reaction mixture was then taken, and neutralized to pH 6-8 using CaCO3. The sample was then prepared for analysis by filtering through a 0.2 µm filter into a 1.5 mL dram vial.

2.2.4 Elemental (CHN) and total protein analysis

3-5 mg (actual weight recorded) of microalgae biomass was added to a tin boat and tightly sealed to prevent air from entering. A CHN analyzer (Flash 2000 series, CE

Elantech Inc.) equipped with an auto sampler and a thermal conductivity detector was then used to determine the N-content in the samples after combustion at 950 °C in presence of oxygen. The flow rate of the carrier gas (helium) was set at 140 mL∙min-1.

Acetanilide standards were used for calibration. Protein content was obtained by multiplying the nitrogen content value with a factor of 5.99.

2.2.5 Fatty acid methyl esters (FAMEs)

Lipid content was determined as fatty acid methyl esters (FAMEs) following an in situ transesterification protocol [22]. 3 to 5 mg of freeze dried biomass was added to a 1.5 mL dram vial. 0.5 mL of acidified methanol (5% H2SO4 v/v) and 0.5 mL of hexane were then added to vial containing the biomass. The vials were tightly capped to prevent evaporation and incubated for 60 min at 90 °C in a water bath. The samples mixed at regular intervals (10 min) with a vortex. After 60 min, the vials were removed and allowed to cool to room temperature. Thereafter, the top hexane phase was removed for analysis by gas chromatography (GC) as described in Section 2.2.7.2. Calibration curves were prepared using FAME standards purchased from Sigma-Aldrich.

2.2.6 Lipid extraction after enzymatic digestion

After enzymatic digestion, 1 mL of the digested slurry was withdrawn and used for lipid extraction experiment. 1.7 mL mixture of hexane and isopropanol (ratio 3:2) was added to the vial containing the digested slurry. The vial was tightly capped to avoid evaporation and extraction was performed at 90 °C for 60 minutes with mixing at regular intervals (10 minutes) using a vortex. After 60 minutes, the reaction was allowed to cool and 0.5 mL of the hexane layer containing lipids was withdrawn. The lipid content was then determined as FAMEs following an in situ transesterification as described in section

2.2.5.

2.2.7 Ash content

To an empty crucible with a known weight, was added ~ 300 mg of freeze-dried microalgae biomass and placed in a ramping 600ºC muffle furnace for 24 h. After 24 h, the sample was removed from the furnace and allowed to cool to room temperature. The weight of the ash content was then estimated by subtracting the weight of the empty crucible from the final weight (weight of crucible plus ash).

2.2.8 Analytical methods

Carbohydrate analysis

Total carbohydrate yield from acid hydrolysis and enzymatic digestion experiments, respectively, were analyzed on an Agilent 1100 HPLC (Agilent

Technologies Inc., Santa Clara, CA) equipped with two Shodex SP8010 column and a refractive index (RI) detector. The column and detector were maintained at 80 °C and 35

°C, respectively. Ultrapure water was used as the mobile phase at a flow rate of 0.6

28 mL·min-1. The run time was set at 60 min with a sample injection volume of 30 µL. The total carbohydrate content was compared with calibration curve (a calibration range of

0.25 g/L to 4 g/L) for glucose, xylose, galactose, arabinose and mannose standards purchased from Sigma-Aldrich.

FAME analysis

Lipids obtained as FAMEs were analyzed on a Shimadzu 2010 GC equipped with an Rtx Bio-diesel column (15 m × 0.32 mm ID × 0.1 µm, Restek Corp., Bellefonte, PA) and a flame ionization detector (FID). The carrier gas, nitrogen, was set to a constant flow rate of 50 cm·s-1. The column oven temperature was maintained at 60 °C for 1 min, raised to 370 °C at a ramp rate of 10 °C·min-1 and then held at 370 °C for 6 min upon completion of the analysis run. For the entire time of the analysis, the injector and FID temperatures were set to 370 °C.

2.3 Results and Discussion

2.3.1 Microalgae biomass composition

Table 2.1. SLA-04 biomass composition

Algae biomass SLA-04

Nitrogen (%) 3.7

Moisture (%) 8.2

Ash content (%) 9.4

Nucleic acids (%) ~5.0

FAMEs (%) 26.6

Total Carbohydrates (%) 19.1

Glucose (%) 14.3

Galactose (%) 4.7

Protein (%) 22.2

Total mass (%) 90.5

Microalgae biomass composition varies depending on growth and harvest parameters. Microalgae biomass harvested at the early growth stage are rich in proteins due to high nitrogen in the culture. Those harvested from mid-growth stage are rich in carbohydrate. While microalgae biomass harvested at the late growth phase are richer in lipids due to nitrogen depletion in the culture. Table 2.1 showed that the microalgae biomass used in this study is rich in lipids and have been harvested at the late growth stage. The microalgae biomass contains 26.6% lipids, ~ 19% of carbohydrate and protein

~ 22% protein.

2.3.2 Enzymatic digestion and sugar yield after enzymatic digestion

The effects of different individual enzyme cocktails on carbohydrate hydrolysis are presented in Figure 2-1. Glucoamylase achieved the highest carbohydrate hydrolysis releasing 76.3% of the total glucose present in the microalgae biomass, while the lower glucose release was obtained by α-amylase (30.0%) and protease performed the weakest

(6.5%). Glucoamylase, a starch-breaking enzyme, cleaves both α-1, 4 and α-1, 6 glucosidic bonds present in the polymer of glucose (starch), α-amylase has the ability to break α-1, 4 – glucosidic bond only and protease cleaves peptide bonds. The small activity shown by protease is likely a result of the presence of some α-amylase in the cocktail [23]. Also, the activities shown by glucoamylase and α-amylase could be attributed to presence of protease in their cocktails, the enzymes made their way into the cell through protein channels or osmotic shock due to the acidic environment.

Galactose a component of the structural carbohydrate was not seen in the product.

Galactose exist as a component of galactolipids making up the cell membrane in actively growing cells [25] or in glycoproteins (β (1-6) linkages) making up the cell wall [26].

Galactose was likely not seen due to the absence of enzymes that can cleave β (1-6) linkages in the cocktails used in our study. Galactose remained in the post-digestion solid and was confirmed after performing acid hydrolysis to close the mass balances.

100 90 80 70 60 50 40

Glucose yield Glucose (%) yield 30 20 10 0 Protease α-amylase GlucoamylaseGlucoamylase α-amylase, & protease glucoamylase and protease Enzymes

Figure 2-1. Enzymatic hydrolysis of 10% (w/v) microalgae biomass solution with a reaction time of 2 h for different enzyme cocktails and their combinations. The amount of each enzyme was fixed as follows: 650µL (325U/g-biomass) protease; 65 mg (1950U/g- biomass) of α-amylase, and 65µL (169U/g-biomass) of glucoamylase.

Although, significant amount of glucose was released using glucoamylase only, the combination of glucoamylase and protease provided higher yield (Figure 2-2). And a combination of all three enzymes (α-amylase, glucoamylase and protease) showed no further increase in the yield (Figure 2-2). Hence, the results showed that α-amylase was not required to obtain maximum glucose release under our conditions, and a combination of protease and glucoamylase only were used for the remainder of this study.

2.3.3 Lipid extraction after enzymatic digestion

The lipid content quantified as FAMEs in our microalgae biomass was 26.56 wt%

(Table 2.1). After enzymatic digestion, lipids were isolated from the digested slurry through immiscible solvent extraction (hexane/isopropanol). Hexane and isopropanol

32 ratio of 3 to 2 was used. Isopropanol would reduce the surface tension between water and hexane. Also, isopropanol having a partial solubility in hexane could help increase the transfer of lipids into the hexane phase. 19.4 wt% of lipids measured as fatty acid methyl ester were recovered. This result represents 72.96% recovery of the initial lipids present in the microalgae biomass.

2.3.4 Protein distribution after enzymatic digestion

Protein present in microalgae biomass could be used as a nitrogen source for fermentation of sugars (e.g. to succinic acid or ethanol) after cell disruption and carbohydrate hydrolysis. Protease used in breaking glycoprotein (cell disruption) could release amino acids into the media, and determining the quantity of solubilized protein from microalgae biomass would be necessary for usage in fermentation. Protein can serve as a nitrogen source for fermentative organisms such as yeast. Therefore, the protein fraction in our supernatant and digestate were measured and shown in Figure 4. The result showed that using the optimal enzyme loading (Figure 2-2, ~ 4.7% of the total proteins present in the microalgae biomass were released into the media, and using twice the enzyme loading only increased the amount of solubilized protein to 17.1% in the media (Figure 2-2).The low protein solubility in the media is due to the acidic condition used, which is nearer to protein isoelectric point, hence, the protein having a less net charge [23].

100 90 80 70 60 50 40 30

Protein percentage Protein 20 10 0 Digestate Solubilized Digestate Solubilized (solids) protein (solids) protein Single enzyme loading Double enzyme loading

Figure 2-2. Protein distribution after enzymatic digestion using protease (325U/g- biomass) and glucoamylase (169U/g-biomass).

2.3.5 Enzyme stability

We investigated the stability of the enzyme combination (protease and glucoamylase) in a “fed-batch reaction” mode of usage (Figure 2-3). The experiments were conducted with successive additions of the microalgae biomass (500 mg of biomass in 1 ml of the solution) into the reaction mixture after every 2 h (sequential feeds). The reactions were performed starting with 5% (w/v) biomass loading to enable at least four sequential feeds since the slurry can become viscous ( >20% (w/v) biomass loading) and stirring becomes an issue. So, the enzymes’ loadings were adjusted as follows: 325µL

(325U/g-biomass) protease and 33µL (169U/g-biomass) of glucoamylase. The experiments are independent with analysis of the reacting mixture performed only at the end of the sequence.

This fed-batch approach allows the evaluation of enzyme stability/activity over multiple feeds. The enzymes stability/activity estimates shown per each cycle are based on the % glucose yield of the combined amounts of microalgae biomass put into the

34 reaction medium till the completion of that cycle. In this sense, the percentages reported reflect the “average activity” displayed by the enzymes over those cycles. The results showed that the enzymes were active over multiple feeds under the reaction conditions. It retains enough activity to achieve above 90% yield of glucose even after four cycles of reuse (Figure 2-3).

100 90 80 70 60 50 40 30

20 Average glucose yield glucose (%) yield Average 10 0 1 2 3 4 Number of sequential feeds

Figure 2-3. Stability of enzyme combination, protease (325U/g-biomass) and glucoamylase (169U/g-biomass) in a fed-batch catalysis mode of usage.

2.3.6 Techno-economic analysis

The experimental parameters for enzymatic fractionation and dilute acid pretreatment methods [27] are shown in Table 2.2. Based on these parameters, the variable operating cost for the two methods were estimated (Table 2.3). The dilute acid pretreatment method was chosen for comparison because the mechanism of cell disruption is same as enzymatic fractionation. In dilute acid pretreatment, acid is used as catalyst. While in enzymatic fractionation, enzymes serve as catalyst. The process

35 assumption was consistent with that of dilute acid pretreatment method for a meaningful comparison; that is 20 wt% solid loading. The cost assumptions are: electricity at 10.1 cents/kWh, glucoamylase at $30/kg and protease at $3.5/kg. In comparison, the total variable operation cost for dilute acid pretreatment method was estimated to be

$71.51/ton-biomass, approximately 20% lower than our enzymatic fractionation cost without recycle of the enzymes. The cost of heating and enzyme cost are the drawbacks for the dilute acid pretreatment and enzymatic fractionation methods, respectively. Each accounting for roughly 84% of the total variable operation cost. Lower temperature gave a lower sugar yield in the acid pretreatment method, and higher temperature causes product degradation and would increase the energy cost. However, in our case, we demonstrated the possibility of recycling the enzymes (this may require additional cost for enzyme recovery). Therefore, considering using the enzymes for five cycles drops the total operation cost for enzymatic fractionation method to roughly one third ($28.78) of the initial cost ($87.54). This significant drop in cost and ability to recover biopolymers in their native form (recovery of high value products) are benefits of enzymatic fractionation. In addition to these, the high capital cost associated with acid pretreatment method due to the need for equipment (Incoloy) with corrosion resistance and strength at high temperature could make our enzymatic fractionation method superior and thus, preferable.

Table 2.2. Enzymatic fractionation and dilute acid pretreatment methods experimental parameters

Experimental parameters Enzymatic fractionation Dilute acid pretreatment

Temperature 55 ℃ 145 ℃

Time (min) 120 10

Sulfuric acid - 2% (w/w)

Ammonia - 0.7% (w/w)

Glucoamylase 0.17% (w/w) -

Protease 0.65% (w/w) -

Table 2.3. Variable operation cost for processing one ton of microalgae biomass

Operation parameters Cost in enzymatic Cost in dilute acid

fractionation pretreatment

Electricity $14.09 ($14.09)* $60

Sulfuric acid - $4.40

Ammonia - $7.11

Glucoamylase $50.7 ($10.14)* -

Protease $22.75 ($4.55)* -

Total cost $87.54 ($28.78)* $71.51

* assumes the enzymes can be reused five times.

2.4 Conclusions

In conclusion, we demonstrated an enzymatic hydrolysis method for isolation of simple sugars directly from microalgae biomass. The new method lowers cost and energy required over the traditional approaches due to the mild processing conditions and would allow for the isolation of the simple sugars in high yield. After the isolation/purification, they can be separately converted into specific chemicals/fuels through appropriate conversion methods. We also demonstrated the possibility of recycling/reusing the enzymes in our developed process for the enzymatic fractionation of microalgae biomass. The enzymes were stable for four cycles without loss in activity.

Finally, we examined the possibility of protein released into the media during enzymatic fractionation process, twice the protease loading is required for a significant release of the protein present in the microalgae biomass.

References for chapter 2

1. Wijffels, R.H. and M.J. Barbosa, An outlook on microalgal biofuels. Science,

2010. 329(5993): p. 796-799.

2. Ducat, D.C., J.C. Way, and P.A. Silver, Engineering cyanobacteria to generate high-value products. Trends in biotechnology, 2011. 29(2): p. 95-103.

3. Melis, A., Solar energy conversion efficiencies in photosynthesis: minimizing the chlorophyll antennae to maximize efficiency. Plant science, 2009. 177(4): p. 272-280.

4. Dismukes, G.C., et al., Aquatic phototrophs: efficient alternatives to land-based crops for biofuels. Current opinion in biotechnology, 2008. 19(3): p. 235-240.

5. Laurens, L., et al., Acid-catalyzed algal biomass pretreatment for integrated lipid and carbohydrate-based biofuels production. Green Chemistry, 2015. 17(2): p. 1145-

1158.

6. Jones, S., et al., Process design and economics for the conversion of algal biomass to hydrocarbons: whole algae hydrothermal liquefaction and upgrading. Pacific

Northwest National Laboratory, 2014.

7. Vadlamani, A., et al., Enzymatic digestion of microalgal biomass for lipid, sugar, and protein recovery. 2015, Google Patents.

8. Doucha, J. and K. Lívanský, Influence of processing parameters on disintegration of Chlorella cells in various types of homogenizers. Applied microbiology and biotechnology, 2008. 81(3): p. 431-440.

9. Halim, R., et al., Microalgal cell disruption for biofuel development. Applied

Energy, 2012. 91(1): p. 116-121.

10. Lee, A.K., D.M. Lewis, and P.J. Ashman, Disruption of microalgal cells for the extraction of lipids for biofuels: processes and specific energy requirements. Biomass and bioenergy, 2012. 46: p. 89-101.

11. Samarasinghe, N., et al., Algal cell rupture using high pressure homogenization as a prelude to oil extraction. Renewable Energy, 2012. 48: p. 300-308.

12. Samarasinghe, N.U. and S. Fernando. Effect of high pressure homogenization on aqueous phase solvent extraction of lipids from Nannochloris oculata. in 2011 Louisville,

Kentucky, August 7-10, 2011. 2011. American Society of Agricultural and Biological

Engineers.

13. Balasubramanian, R.K., T.T.Y. Doan, and J.P. Obbard, Factors affecting cellular lipid extraction from marine microalgae. Chemical Engineering Journal, 2013. 215: p.

929-936.

14. Günerken, E., et al., Cell disruption for microalgae biorefineries. Biotechnology advances, 2015. 33(2): p. 243-260.

15. Prabakaran, P. and A.D. Ravindran, A comparative study on effective cell disruption methods for lipid extraction from microalgae. Letters in applied microbiology,

2011. 53(2): p. 150-154.

16. Harun, R. and M.K. Danquah, Influence of acid pre-treatment on microalgal biomass for bioethanol production. Process Biochemistry, 2011. 46(1): p. 304-309.

17. Miranda, J., P.C. Passarinho, and L. Gouveia, Pre-treatment optimization of

Scenedesmus obliquus microalga for bioethanol production. Bioresource technology,

2012. 104: p. 342-348.

18. Harun, R., et al., Exploring alkaline pre-treatment of microalgal biomass for bioethanol production. Applied Energy, 2011. 88(10): p. 3464-3467.

19. Choi, S.P., M.T. Nguyen, and S.J. Sim, Enzymatic pretreatment of

Chlamydomonas reinhardtii biomass for ethanol production. Bioresource Technology,

2010. 101(14): p. 5330-5336.

20. Rodrigues, M.A. and E.P.d.S. Bon, Evaluation of Chlorella (Chlorophyta) as source of fermentable sugars via cell wall enzymatic hydrolysis. Enzyme research, 2011.

2011.

21. Lee, O.K., et al., Chemo-enzymatic saccharification and bioethanol fermentation of lipid-extracted residual biomass of the microalga, Dunaliella tertiolecta. Bioresource technology, 2013. 132: p. 197-201.

22. Vadlamani, A., et al., Cultivation of microalgae at extreme alkaline pH conditions–a novel approach for biofuel production. ACS Sustainable Chemistry &

Engineering, 2017.

23. Vadlamani, A.K., et al., Enzymatic Digestion of Microalgal Biomass for Lipid,

Sugar, and Protein Recovery. 2014, Google Patents.

24. Van Wychen, S. and L.M. Laurens, Determination of Total Carbohydrates in

Algal Biomass: Laboratory Analytical Procedure (LAP). 2013, National Renewable

Energy Laboratory (NREL), Golden, CO.

25. Siegenthaler, P.-A. and N. Murata, Lipids in photosynthesis: structure, function and genetics. Vol. 6. 2006: Springer Science & Business Media.

26. Noda, K., et al., A water-soluble antitumor glycoprotein from Chlorella vulgaris.

Planta medica, 1996. 62(05): p. 423-426.

27. Dong, T., et al., Combined algal processing: A novel integrated biorefinery process to produce algal biofuels and bioproducts. Algal Research, 2016. 19: p. 316-323.

Chapter 3

Cross-Metathesis Approach to Produce Precursors of Nylon 12 and Nylon 13 from Microalgae

A paper published in RSC Advances1

Godwin Abel, Kim Nguyen, Sridhar Viamajala, Sasidhar Varanasi,and Kana

Yamamoto

3.1 Introduction

Nylons are synthetic polyamides with extensive applications in modern society.

In particular, nylon 11 and 12 possess excellent chemical properties and are widely used

in automotive, medical, and electronic industries as well as for sport equipment, back

panel for solar cells, and lenses for glasses. Nylon 13 is reported to have similar

properties to nylon 12. These nylons are produced from amino acids or their derivatives

with a corresponding number (11, 12 or 13) of carbon-chain backbone [1].

______

2014Royal Society of Chemistry

Currently, the precursor of nylon 12 is manufactured from petroleum-derived butadiene in a six-step process [2]. However, there is an increasing interest in use of renewable sources for production of these amino acid precursors, because of concerns over environmental sustainability of petrochemical products [3]. As such, synthetic approaches that use natural fatty acids and esters from plant- or algae-derived feedstocks are especially attractive [4-6]. Among the natural fatty acids, oleic acid is the predominant component of lipids in a large variety of vegetable oils (e.g. soy or canola) as well as oleaginous microorganisms such as algae [7]. The precursor of nylon 11 is manufactured from ricinoleic acid, a natural fatty acid available only in castor beans.[2]

In this method, the acid is first converted to its methyl ester and then subjected to pyrolysis to produce 10-undecyleate, which is then converted to the nylon 11 precursor in two additional steps. Several other multi-step (>4 steps) approaches to produce nylon 11 precursors from either ricinoleic acid or oleic acid derivatives are reported in patents [8-

14] and journals [15-22], but have likely not been utilized for industrial process.

While there is one study on a concise three-step metathesis-based approach for producing nylon 11 from oleic acid [17], the method has previously not been applied for producing nylons 12 or 13, likely due to challenges associated with selectivities

(discussed in more detail in the Results section). The other reported methods to produce nylons 12 and 13 from non-petrochemical sources are lengthy, and also employ exotic fatty acids as starting materials. For example, there are several reported nylon 12 precursor syntheses that use ricinoleic acid from castor beans as the feedstock and require

4–6 steps [15, 16, 20, 21, 23]. Synthesis of nylon 13 has not been fully studied except for

44 one reported approach that involved several steps from either erucic or lesquerolic acid, another atypical acid available from rapeseed oil [24].

Herein, we report the successful application of cross metathesis to synthesize nylon 12 and 13 precursors from oleic acid. Our method provides a shorter and simpler route for production of both nylon 12 and nylon 13 from an abundant and inexpensive, and thus preferred starting material (Figure 3-1).

Figure 3-1.Cross-metathesis approach for the synthesis of nylon precursors

3.2 Materials and methods

3.2.1 General

1 13 H-NMR and C-NMR spectra were acquired in CDCl3, on either Varian Unity

Inova 600 (600 MHz) or Brucker Avance 600 (600 MHz) NMR spectrometers. Chemical shifts (δ) were reported as parts per million (ppm) with reference to tetramethylsilane

(TMS) or solvent unless otherwise stated. The coupling constants (J) are reported in Hz.

Mass spectra were obtained with Hewlett-Packard Esquire Ion Trap LC-MS

(electrospray). GC analyses were performed with HP 5890 series II equipped with FID and an auto-sampler HP controller 7672A and Biodiesel TG column (5%diphenyl, 95% dimethyl polysiloxane, 15m, 0.33mm ID and 0.10µm dF) or MXT biodiesel TG (Siltek – treated stainless steel). Samples were analyzed with the following method: 60°C to 370°C at 10°C/min; 6min hold time.

Most reagents were purchased from commercial suppliers and used without further purification. Thin layer chromatography (TLC) was carried out on glass backed silica plates, purchased from Sorbent Technology. The plates were visualized under UV

(254 nm) light, and by staining with either Potassium permanganate or Ninhydrin and gentle heating. Silica gel column chromatography was carried out using 20‒60 micron dry silica purchased from Sorbent Technology.

All the solvents used for CM were deoxygenated by passing through dry nitrogen gas for 20 min before being used.

3.2.2 Experimental procedures

Methyl 12-aminododecanoate (5)

12-Cyano-9-dodecenoic acid methyl ester (19.9 mg, 0.08916 mmol) was dissolved in 2 mL of chlorobenzene, and transferred into a 2-dram vial containing t-

BuOK (30 mol%, 2.9 mg, 0.02584 mmol). 2nd generation Grubbs catalyst (3 mol%, 2.2 mg, 0.002591 mmol) was dissolved in chlorobenzene (1mL) and transferred into the vial.

The vial was placed in a Parr reactor which was purged 3 times with hydrogen gas before

46 being pressurized at 20 bars and heated at 80 °C with stirring for 17hrs. Subsequently, the reaction mixture was purified by silica gel chromatography using a mixture of hexane/methanol/DCM (6:2:2) as eluent and concentrated under reduced pressure to give methyl 12-aminododecanoate (13.8 mg, 67.55% isolated yield). NMR data match with reported data.1

1 H-NMR (CDCl3, 600 MHz) δ/ppm: 3.65 (s, 3H, -OCH3), 2.67 (t, J=6.6Hz, 2H, -

CH2-NH2), 2.28 (t, J=7.5Hz, 2H, -CH2-CO-), 1.57 – 1.62 (m, 4H, -CH2-,), 1.41 – 1.43 (m,

2H, -NH2), 1.25 – 1.27 (m, 14H, -CH2-).

13 C-NMR (CDCl3, 600 MHz) δ/ppm: 174.5, 51.6, 42.3, 34.2, 33.8, 29.7, 29.6,

29.6, 29.5, 29.4, 29.3, 27.0, 25.1.

Methyl 13-aminotridecanoate (6)

13-Cyano-9-tridecenoic acid methyl ester (20.1 mg, 0.08476 mmol) was dissolved in 2 mL of chlorobenzene, and transferred into a 2-dram vial containing t-BuOK (30 mol%, 2.9 mg, 0.02584 mmol). 2nd generation Grubbs catalyst (2 mol%, 1.4 mg,

0.001649 mmol) was dissolved in chlorobenzene (1mL) and transferred into the vial. The vial was placed in a Parr reactor which was purged 3 times with hydrogen gas before being pressurized at 20 bars and heated at 80 °C with stirring for 19hrs. Subsequently, the reaction mixture was purified by silica gel chromatography using a mixture of hexane/methanol/DCM (6:2:2) as eluent and concentrated under reduced pressure to give methyl 13-aminotridecanoate (13.7 mg, 66.5% as an uncorrected isolated yield).

1 H-NMR (CDCl3, 600 MHz) δ/ppm: 3.66 (s, 3H, -OCH3), 2.68 (t, J=6.6Hz, 2H, -

CH2-NH2), 2.29 (t, J=7.5Hz, 2H, -CH2-CO-), 1.58 – 1.63 (m, 4H, -CH2-), 1.41 – 1.44 (m,

2H, -NH2), 1.25 – 1.28 (m, 16H, -CH2-).

13 C-NMR (CDCl3, 600 MHz) δ/ppm: 174.5, 51.6, 42.3, 34.3, 33.9, 29.8, 29.7,

29.6, 29.6, 29.5, 29.4, 29.3, 27.0, 25.1.

+ HRMS: C14H29NO2 [M+H] calc. 244.2271 found 244.2279

(E/Z)-11-Cyano-9-dodecenoic acid methyl ester (9)

a. Methyl oleate (0.538 mmol) was added into a dry three-necked round bottom flask. Chlorobenzene (5 mL) was then added into the flask with a syringe followed by allyl cyanide (0.220 mL, 2.735 mmol). 1,4-Benzoquinone (29.1 mg, 0.269 mmol, 50 mol%) was dissolved in chlorobenzene (2.5 mL) and transferred to the reaction mixture by a syringe. The flask containing the mixture was purged for 10 min with nitrogen, and then heated at 110 °C with stirring (400 rpm) for 20 min. The 2nd Generation Hoveyda-

Grubbs catalyst (2 mol%, 7.5 mg, 0.01198 mmol) was dissolved in chlorobenzene (5 mL) and transferred into a syringe. The solution containing the catalyst was transferred to the reaction mixture drop wise over a period of 3h while keeping the reaction temperature at

110 °C with stirring (400 rpm). After the completion of the addition, the reaction mixture was kept at this temperature for additional 1 h before being cooled to room temperature.

The crude product was purified by column chromatography using hexane/ethylacetate (9/1) as eluent and concentrated under reduced pressure to provide the desired compound (E/Z mixtures) as brown oil (66.6 mg, 55.47 % isolated yield).

b. Methyl 9-decenoate (0.4388 mmol) was added into a dry three-necked round bottom flask. Chlorobenzene (4 mL) was then added into the flask with a syringe followed by allyl cyanide (0.176 mL, 2.188 mmol). 1,4-Benzoquinone (23.2 mg, 0.2146 mmol, 50 mol%) was dissolved in chlorobenzene (2 mL) and transferred to the reaction mixture by a syringe. The flask containing the mixture was purged for 10 min with nitrogen, and then heated at 110 °C with stirring (400 rpm) for 20 min. The 2nd

Generation Hoveyda-Grubbs catalyst (2 mol%, 6.1 mg, 0.00974 mmol) was dissolved in chlorobenzene (4 mL) and transferred into a syringe. The solution containing the catalyst was transferred to the reaction mixture drop wise over a period of 3h while keeping the reaction temperature at 110 °C with stirring (400 rpm). After the completion of the addition, the reaction mixture was kept at this temperature for additional 1 h before being cooled to room temperature.

The crude product was purified by column chromatography using hexane/ethylacetate (8/2) as eluent and concentrated under reduced pressure to provide the desired compound (E/Z mixtures) as brown oil (35.4 mg, 36.16 % isolated yield).

(Note: because of the difficult removal of impurities from the reaction mixture, only a part of the fractions containing the product was isolated).

1 H-NMR (CDCl3, 600 MHz) δ/ppm: 5.65 – 5.83 (m, 1H, -CH=CH-), 5.30 – 5.40

(m, 1H, -CH=CH-), 3.66 (s, 3H, -OCH3), 3.04 – 3.08 (m, 2H, -CH2CN), 2.29 (t, J=7.5Hz,

2H, -CH2CO-), 2.02 – 2.06 (m, 2H, -CH2-), 1.58 – 1.63 (m, 2H, -CH2-), 1.25 – 1.40 (m,

8H, -CH2-).

13 C-NMR (CDCl3, 600 MHz) δ/ppm: 174.4, 174.4, 136.4, 136.2, 118.4, 118.0,

118.4, 117.2, 116.9, 51.6, 34.2, 34.2, 32.2, 29.2, 29.1, 29.0, 29.0, 29.0, 28.9, 27.4, 25.0,

25.0, 25.0, 20.6, 15.7.

(E/Z)-12-cyano-9-tridecenoic acid methyl ester (10)

a. Methyl oleate (0.543 mmol) was added into a dry three-necked round bottom flask. Chlorobenzene (5 mL) was then charged into the flask with a syringe followed by

4-pentenenitrile (0.130 mL, 1.346 mmol). 1,4-Benzoquinone (29.0 mg, 0.268 mmol, 50 mol%) was dissolved in chlorobenzene (2.5 mL) and transferred to the reaction mixture by a syringe. The flask containing the mixture was purged for 10 min with nitrogen, and then heated at 110 °C with stirring (400 rpm) for 20 min. The 2nd Generation Hoveyda-

110 °C. After the completion of the addition, the reaction was kept at this temperature for additional 1 h before being cooled to room temperature.

b. Methyl 9-decenoate (0.4383 mmol) was added into a dry three-necked round bottom flask. Chlorobenzene (4 mL) was then charged into the flask with a syringe followed by 4-pentenenitrile (0.064 mL, 0.6627 mmol). 1,4-Benzoquinone (23.2 mg,

0.2146 mmol, 50 mol%) was dissolved in chlorobenzene (2) and transferred to the

50 reaction mixture by a syringe. The flask containing the mixture was purged for 10 min with nitrogen, and then heated at 110 °C with stirring (400 rpm) for 20 min. The 2nd

Generation Hoveyda-Grubbs catalyst (2 mol%, 6.1 mg, 0.00974 mmol) was dissolved in chlorobenzene (4 mL) and transferred into a syringe. The solution containing the catalyst was transferred to the reaction mixture drop wise over a period of 3h while keeping the reaction temperature at 110 °C. After the completion of the addition, the reaction was kept at this temperature for additional 1 h before being cooled to room temperature.

1 H-NMR (CDCl3, 600 MHz) δ/ppm: 5.54 – 5.59 (m, 1H, -CH=CH-), 5.34 – 5.42

(m, 1H, -CH=CH-), 3.65 (s, 3H, -OCH3), 2.28 – 2.40 (m, 6H, -CH2-), 1.98 – 2.05 (m, 2H,

-CH2-), 1.58 – 1.64 (m, 2H, -CH2-), 1.26 – 1.37 (m, 8H, -CH2-).

13 C-NMR (CDCl3, 600 MHz) δ/ppm: 174.4, 174.4, 134.3, 133.7, 125.7, 125.1,

119.6, 117.2, 51.6, 34.2, 32.5, 29.5, 29.2, 29.2, 29.1, 29.0, 28.5, 27.4, 25.0, 23.4, 17.9,

17.7.

+ HRMS: C14H23NO2 [M+Na] calc. 260.1627 found 260.1623

Cross metathesis using algal lipid

Microalgae: Freeze dried cultures of a chlorella sp. Strain SLA04 were used in this study. The alga, originally isolated from Soap Lake, WA, was grown in a medium that contained the following: NaNO3 (2.94mM), KH2PO4 (1.43mM), Na2CO3 (2.35mM),

NH4Cl (0.93mM), MgSO4.7H2O (0.30mM), CaCl2.2H2O (0.17mM), NaCl (0.42mM), ferric ammonium citrate (10 mg/L) and 1Ml of trace metal solution which contains

H3BO3 (0.6g/L), MnCl2.4H2O (0.25g/L), ZnCl2 (0.02g/L), CuCl2.2H2O (0.015g/L),

Na2MoO4.2H2O (0.015g/L), CoCl2.6H2O (0.015g/L), NiCl2.6H2O (0.01g/L), V2O5

(0.002g/L), and KBr (0.01g/L). Cultures were grown phototropically in 3L cytostir reactors illuminated by fluorescent lamps with an incident light intensity of

300µmoles/m2. GC Analyses: GC conversion was calculated by integrating all the peaks in the chromatogram except the peaks that do not interfere the mass balance: by-products

((E/Z)-undec-2-enenitrile, dimer of acrylonitrile or allyl cyanide, 10-decene), unreactive substrates (C14:0, C15:0, C16:0 FAME), and impurities introduced from the starting material. This method is comparable to the method used to analyze the GC conversion of the reaction with pure oleic acid (1) used throughout this manuscript.

(1) Cross-metathesis with acrylonitrile:

Using pure oleic acid (1) (control experiment):

Methyl oleate (1) (0.1 mmol), acrylonitrile (35 μL, 0.534 mmol) and 1 mL of dry toluene were placed in a round bottom flask. A 2nd-Generation Grubbs-Hoveyda catalyst

(0.8 mg, 0.0013 mmol) was dissolved in dry toluene (1 mL) and transferred to a syringe.

Additional 1 mL of toluene was used to rinse the vial. The catalyst solution was transferred to the reaction mixture using a syringe pump, over a period of 1 h under nitrogen atmosphere with magnetic stirring (400 rpm) at 95 ̊C. At the end of the addition, the mixture was left to react for 1.5 h at 95 ̊C. The reaction mixture was analyzed by gas chromatography, indicating the conversion being >99 % by area. GC analysis showed the four major peaks, methyl 10-cyano-9-decenoate and 2-undecenitrile (both as mixtures of

E and Z isomers). The mixture was passed through a plug of silica gel (0.5cm in Pasteur pipette) with hexane/ethyl acetate (7/3) in order to remove the catalyst, concentrated, and

52 the residue was purified by silica gel chromatography, yielding (E)-10-cyano-9-decenoate

(4.69 mg, 0.022 mmol, 22%), (Z)-10- cyano-9-decenoate (12.3 mg, 0.059 mmol, 59%),

(E)-2-undecenitrile (4.76 mg, 0.029 mmol), and (Z)-2- undecenitrile (11.5 mg, 0.070 mmol) all as clear oil.

Using algal lipid:

Algal lipids, containing 55% methyl oleate (0.1119 mmol), obtained from microalgae (Chlorella species) by reactive extraction method,2 was added into a dry three-necked round bottom flask. Toluene (1.5 mL) was then added into the flask with a syringe followed by acrylonitrile (0.034 mL, 0.519 mmol). The flask containing the mixture was purged for 10 min with nitrogen, and then heated at 95 °C with stirring (400 rpm) for 20 min. The 2nd Generation Hoveyda-Grubbs catalyst (1 mol%, 0.8 mg, 0.0013 mmol) was dissolved in toluene (1 mL) and transferred into a syringe. The solution containing the catalyst was transferred to the reaction mixture drop wise over a period of

1h while keeping the reaction temperature at 95 °C with stirring (400 rpm). After the completion of the addition, the reaction mixture was kept at this temperature for additional 1h before being cooled to room temperature.

The reaction mixture was passed through a pipette of silica using hexane/ethyl acetate (7/3) as eluent, and the crude product was analyzed by GC.

(2) Cross-metathesis with allyl cyanide:

53 with a syringe followed by allyl cyanide (0.044 mL, 0.5470 mmol). 1,4-Benzoquinone

(6.3 mg, 0.0583 mmol, 50 mol%) was dissolved in chlorobenzene (0.5 mL) and transferred to the reaction mixture by a syringe. The flask containing the mixture was purged for 10 min with nitrogen, and then heated at 110 °C with stirring (400 rpm) for 20 min. The 2nd Generation Hoveyda-Grubbs catalyst (2 mol%, 1.4 mg, 0.0022 mmol) was dissolved in chlorobenzene (1 mL) and transferred into a syringe. The solution containing the catalyst was transferred to the reaction mixture drop wise over a period of 2h while keeping the reaction temperature at 110 °C with stirring (400 rpm). After the completion of the addition, the reaction mixture was kept at this temperature for additional 2h before being cooled to room temperature.

The reaction mixture was passed through a pipette filled with silica gel using hexane/ethyl acetate (7/3) as eluent, and the crude product was analyzed by GC.

3.3 Results and Discussion

The previously reported cross metathesis approach to produce the nylon 11 precursor (Figure 3-4, (a)) involved reacting methyl oleate (1) or ricinoleate (2) with acrylonitrile [15, 17, 20]. The resulting 10-cyano-9-decenoate (7) was subjected to high- pressure hydrogenation to remove all unsaturation and deliver the nylon 11 precursor (3)

(Figure 3-2, eq. 1). Nylon 12 precursor was also prepared in analogous fashion by cross metathesis with acrylonitrile and methyl 10-undecylenate (4) prepared from pyrolysis of

2 (Figure 2-1, eq. 2). In related approaches, methyl oleate (1) or ricinoleate (2) was first converted to their derivatives such as dimers or terminal alkenes before subjecting to metathesis – these routes showed improved throughput to the original approach, although lengthened the overall synthetic steps [15, 16, 20, 21].

Since our intention was to find a direct pathway toward nylon 12 and 13 precursors from oleic acid, we attempted cross metathesis of methyl oleate with alkenylcyanides with longer chain length (Figure 1, eq. 3). In the first step, we subjected methyl oleate to cross metathesis with allyl or homoallyl cyanides with the intent of generating the corresponding cyano ester intermediates 9 and 10. Subsequently, the double and triple bonds of these intermediates could be hydrogenated to deliver the nylon

12 and 13 precursors 5 and 6, respectively. Our approach, although related to the reported nylon 11 synthesis [17], behaved completely differently when their reaction conditions were used [25]. In this report, we disclose our studies to overcome the issue, as well as eventual success in delivering the desired nylon precursors.

(a) Cross me tathesis approach by Arkema 9 CO2Me acrylonitrile 10 9 11 9 7 CO Me CN NC CO2Me 2 (eq. 1) H2N 7 10 7 10 7 methyl 11-aminounedecanoate (3) X (nylon 11 precursor) X = H: methyl oleate (1) X = OH: methyl ricinolate (2)

acrylonitrile 11 9 9 CO2Me H N CO Me NC 2 9 2 11 CO2Me CN 7 7 (eq. 2) 10 12 10 10 7 8 methyl 12-aminododecanoate (5) methyl 10-undecylenate (4) (nylon 12 precursor) (b) THIS WORK 9 CN 9 11 9 CO2Me n 10 CO2Me H2N CO2Me 7 n (eq. 3) NC 7 n 7 10 10 n = 1: allyl cyanide n = 1: 9 n = 1: methyl 12-aminododecanoate (5) n = 2: homoallyl cyanide methyl oleate (1) n = 2: 10 n = 2: methyl 13-aminotridecanoate (6) (nylon 13 precursor)

Figure 3-2. (a) Previously reported cross metathesis approach. (b) Our approach to nylon precursors from methyl oleate.

We began our investigation of the cross metathesis step using allyl cyanide and methyl 9-decenoate (11), a model substrate for methyl oleate (Table 1). Under the reported reaction conditions (entries 1–3), while the reaction conversion is good (>92%) at the temperature >95 °C, only 25–30% (by GC area) of the desired product was seen.

The complex mixture of other products comprises unsaturated cyano esters with various

55 alkyl chain lengths, with molecular weight varying by 14 (Figure 3-3), which led us to speculate that they arose from olefin isomerization. Other major side-products of the reaction were series of dimers with structures yet to be determined.

Figure 3-3. Mass spectrum of the cross metathesis of methyl dec-9-enoate and allyl cyanide.

In order to suppress the undesired isomerization, the reaction was examined with several additives.[22, 26] We found that addition of 1,4-benzoquinone nearly completely suppressed the side reaction in our system (Table 3.1, entries 4). In order to fully suppress the isomerization, 50 mol% of the additive was required (entries 12 vs 14, Figure 3-4)

Figure 3-4. Chromatograms of cross metathesis reactions of methyl 9-decenoate and allyl cyanide without and with benzoquinone.

Subsequently, the other reaction parameters were screened to suppress dimerization and to further improve the reaction conversion. The reaction conversion was good at temperature >95 °C; however, the best product profile was observed at 110 °C

(entries 5 vs 7). The conversion was further improved by a continuous injection of the ruthenium catalyst over 1–2 h period (entries 5 vs 8). The reactant concentration ~0.033

M was found to provide the best selectivity for the desired reaction (entries 11, 13 and

14). The preferred molar ratio of allyl cyanide was found to be five equivalents (entries

2 vs 3, 10 vs 12). Use of cross metathesis for oleochemical production from fatty acids has been a subject of several reviews [19, 22, 27]. Consistent with other reports, Hoveyda-

Grubbs second generation catalyst showed better conversion than Grubbs catalyst (entries

5 and 9) for our system. At least 2 mol% of this catalyst was required to achieve full conversion under the reaction conditions shown (entries 4 vs 5).

Table 3.1. Cross-metathesis of Methyl 9-decenoate with allyl cyanide[a]

9 9 CO Me CO Me 2 + 2 7 CN NC 7 10 10 methyl 9-decenoate (11) allyl cyanide 9

Additive[h] Catalyst Temperature Time Conversion GC area% Entry (mol%) (mol%) (°C) (h) (%) 9 dimers 1 / 1 80 17 6.9 6.1 0

2[b] / 1 95 21 94.7 25.5 35.8

3 / 1 95 8 92.2* 28.2 21.7

4 BQ (10) 1 95 5 78.7* 52.0 7.8

5 BQ (10) 2 95 4 92.2 49.2 17.0

6 AA (10) 2 95 6 48.1 20.7 4.5

7 BQ (10) 2 110 6 84.4 56.4 9.0

8[c] BQ (10) 2 95 6 42.7 23.5 7.8

9[g] BQ (10) 2 95 6 49.1 27.8 3.3

10[b] BQ (10) 2 110 4 93.8 44.2 11.1

11[d] BQ (10) 2 95 6 64.6 39.2 11.0

12[e] BQ (10) 2 110 6 37.7 16.9 4.0

13[f] BQ (50) 2 110 4 84.7 42.2 3.0

14 BQ (50) 2 110 4 78.8 58.3 5.5

[a] Reaction conditions: To a flask containing methyl 9-decenoate (0.1 mmol), allyl cyanide (0.5 mmol), and toluene (2 mL) was added Hoveyda-Grubbs 2nd generation catalyst (1–2 mol%) in toluene (1 mL) dropwise over 1h. [b] 2 equiv. of allyl cyanide was used. [c] The catalyst added in one portion. [d] 0.5mL of solvent was used for catalyst delivery; no other solvent was used. [e] 10 equiv. of allyl cyanide was used. [f] 12 mL (11mL + 1mL) of solvent for the whole reaction was used instead of 3 mL (2 mL + 1 mL). [g] Grubbs 2nd generation catalyst was used. [h] BB: Benzoquinone; AA: Acetic acid. * Approximate values.

It is known that solvent selection significantly influences conversion and selectivity in metathesis reactions. In particular, halogenated solvents are considered to be superior

[28]. In our system, solvents with lower boiling point were not suitable (Table 3.2, entry

2) and fluorinated solvents increased the dimer formation (entries 3 and 5). However, use of chlorobenzene provided good reaction con version and better selectivity than other solvents (entries 4 and 6).

Table 3.2. Solvent screening on cross metathesis of 9-decenoate with allyl cyanide[a]

9 9 CO2Me CN 10 CO2Me 9 7 NC CO2Me allyl cyanide 7 10 + 10 7 9 methyl 9-decenoate (11) methyl oleate (1)

Conv. GC area% Entry Solvent Time (h) (%) 9 dimers

1 C6H5CH3 4 78.8 58.3 5.5

[b] 2 (CH2Cl)2 8 41.2 17.8 11.6

3 C6F5Cl 2 95.3 60.1 20.6

4 C6H5Cl 2 90.3 62.6 13.8

5 C6F5CF3 5 77.2 18.6 46.0

[c] [d] 6 C6H5Cl 3 93.5 71.2 (58) 4.1 [a] Reaction condition: To the flask containing methyl oleate (0.1 mmol), allyl cyanide (0.5 mmol), 1,4-benzoquinone (0.05 mmol), in chlorobenzene (2 mL) was added Hoveyda-Grubbs 2nd generation catalyst (2 mol%) in chlorobenzene (1 mL) was added drop wise at 110 °C. [b] Reaction temperature: 80 °C.[c]4.5 mol% of the catalyst was used. [d] Isolated yield.

Subsequently, these optimized cross metathesis conditions for methyl 9-decenoate

(11) were applied to methyl oleate (1) (Table 3.3). It was observed that dimerization was

59 suppressed with this substrate likely due to its lower reactivity. Use of 1 instead of 11 made the product analysis more complex, because of formation of non-volatile by- products, including methyl 9-decenoate (11).

We have examined several reaction parameters in the attempt to optimize the reaction. We first examined use of lower temperature (95 °C) with hope to lower the catalyst deactivation and increase the selectivity. However, the lower temperature reactions resulted in increased formation of 11 and dimers (entry 2). To suppress the catalyst decomposition, the catalyst addition time was further extended. This attempt was partially successful and extending the time to 3h improved the conversion as well as the selectivity for 9 (entries 3–4). Attempt to drive 11 to desired 9 with use of excess of cyanide led to low reaction conversion, possibly due to catalyst poisoning (entry 5).

Increasing the catalyst loading up to 4.5 mol% only provided similar reaction profiles

(entry 6–7). To date, we concluded that the current best reaction conditions for this conversion to be the one used in entry 3 or 4. This procedure reliably provides the desired

8 in ~55% yield while suppressing olefin isomerization and dimer formation (<10 %).

Table 3.3. Cross-metathesis of methyl oleate with allyl cyanide[a]

Time Time Conv. GC area% Entry [b] (h) (h) (%) 9 dimers 11

1 3 1 78.5 46.5 3.9 14.5

2[c] 3 1 70.2 34.4 4.0 14.9 3 4 2 85.3 49.0 5.6 9.3

4 4 3 87.0 55.3 (55)[h] 7.5 9.2 5[d] 4 2 38.5 16.0 0.0 9.8

6[e] 3 2 86.3 37.5 (56)[g] 7.1 7.0

7[f] 4 2 96.3 47.1 4.1 11.7

9 9 CO2Me CN 10 CO2Me 9 7 NC CO2Me allyl cyanide 7 10 + 10 7 9 methyl 9-decenoate (11) methyl oleate (1) [a] Reaction conditions: To the flask containing methyl oleate (0.1 mmol), allyl cyanide (0.5 mmol), 1,4-benzoquinone (0.05 mmol), in chlorobenzene (2 mL) was added Hoveyda-Grubbs 2nd generation catalyst (2 mol%) in chlorobenzene (1 mL) added dropwise at 110 °C. [b] Catalyst addition time. [c] Reaction temperature: 95 °C. [d] 10 equiv. of allyl cyanide was used. [e] 4.5 mol% of catalyst was used. [f] 3 mol% of catalyst was used. [g] GC yield (quantified). [h] Isolated yield.

Cross-coupling with homoallyl cyanide was also investigated using both 9- decenoic acid (Table 3.4) and methyl oleate (Table 3.5). Some reaction tuning was required in order to achieve a good conversion with this coupling partner. In particular, we found that the reaction conversion was sensitive to equivalency of homoallyl cyanide.

Under the optimal conditions used for allyl cyanide metathesis, the reaction conversion was as low as 4–5 % (Table 3.4, entries 5–6) or 23% if using methyl oleate (1) (Table

3.5, entry 6). We suspected that this behaviour may relate to the faster dimerization of

61 homoallyl cyanide. In any event, lowering the homoallyl cyanide equivalency to 1–2 equiv. resulted in good reaction conversion and selectivity with both substrates (Table

3.4, entry 2; table 3.5, entry 4). Thus, both methyl 9-decenoate (11) and methyl oleate (1) effectively underwent cross-metathesis reaction with non-conjugated nitriles. We were able to obtain the desired product at a comparable isolated yield of 43% from this reaction.

Table 3.4. Cross-metathesis of 9-decenoate with homoallyl cyanide 9 CO Me 9 2 NC CO Me 7 + CN 2 10 7 10 methyl 9-decenoate (11) homoallyl cyanide 10 Time cyanide Conv. GC area % Entry (h) (mmol) (%) 10 dimers 1 6 0.1 82.2 43.8 31.8 2 6 0.15 83.5 51.2 27.6 3 4 0.15 78.0 48.1 25.0 4 6 0.25 78.9 48.9 24.1 5 6 0.25 56.3 38.5 15.9 6 4 0.5 4.7 4.0 0.0 7 6 0.5 4.2 4.2 0.6 [a] Reaction condition: 0.1 mmol of methyl oleate, 50 mol% 1,4- Benzoquinone, 2 mol% of Hoveyda-Grubbs 2nd generation in chlorobenzene (1 mL) added dropwise into 1.5mL of chlorobenzene at 110 oC.

Table 3.5. Cross-metathesis of methyl oleate with homoallyl cyanide[a]

9 9 9 CO2Me 10 CN CO2Me CO2Me 7 7 homoallyl cyanide 10 + 10 7 CN 10 methyl 9-decenoate (11) methyl oleate (1) Time cyanide Conv. GC area % Entry (h) (mmol) (%) 10 Dimers 11

1 6 0.15 68.6 31.5 15.3 7.8 2 6 0.25 78.8 42.9 10.9 10.1 3 8 0.25 74.4 34.5 12.6 9.9 4 4 0.25 88.7 52.3 (42.9)[b] 8.7 16.2 5 6 0.3 77.6 38.0 7.5 16.8 6 6 0.5 23.3 0.4 6.6 11.7 [a] Reaction condition: 0.1 mmol of methyl oleate, 50 mol% 1,4- Benzoquinone, 2 mol% of Hoveyda-Grubbs 2nd generation in chlorobenzene (1 mL) added dropwise into 1.5mL of chlorobenzene at 110 oC. [b] Isolated yield.

In the second step, although hydrogenation of olefins is an established reaction that many catalysts are available, the reduction of nitriles to amines conventionally uses stoichiometric strong hydride reducing agents such as lithium aluminium hydride or borane, or hydrosilylation with Lewis acids such as titanium isopropoxide. Alternately, strong heterogeneous catalysts such as Raney nickel or cobalt can be used to catalyze addition of molecular hydrogen, but reaction conditions are harsh and only moderate selectivity is afforded. Studies on hydrogenation of nitriles with homogenous catalysts have been limited although catalysts based on rhodium, iridium, rhenium, and ruthenium have recently been investigated.[29] Most recent studies used ruthenium complexes, which by far gave the best selectivity under mild reaction conditions. Many studied catalytic systems use phosphine ligands as well as potassium tert-butoxide additive.

Typical reactions conditions are 80–140 ºC at H2 pressure of 14–75 bars. It was also shown that milder reaction conditions could be used when the phosphine ligands of the metal complex are replaced with carbene ligands. Finally, the metathesis catalyst has also been shown to facilitate the hydrogenation reaction, and that it is possible to use residual catalyst from metathesis for hydrogenation [20, 21].

The previous studies for hydrogenation of fatty acid derivatives indicated that either the Grubbs or Hoveyda-Grubbs second generation catalyst would be effective.[20,

21] However, only Grubbs second generation catalyst provided the desired product in our experiments. Using this catalyst, we found that toluene, benzene, or chlorobenzene solvents were all suitable and the reaction temperature of 80 ºC provided the best overall conversion. Thus these reaction conditions were adopted for the subsequent studies

(Table 3.6). The base additive is essential in this catalyst system, and 30 mol% of potassium tert-butoxide was sufficient to ensure the full reaction conversion (entries 1–2 vs 3–11). The catalyst loading of >2 mol% was found essential for good conversion

(entries 3–5; 8–10). The reaction mixture was kept for 17–44 h under hydrogen pressure in between 20–25 bar. These reaction conditions consistently provided >50% (by GC area) of the desired methyl 12-aminododecanoate (5) (entries 4, 6). The same reaction conditions, when used to hydrogenate C13 cyano ester (10), resulted in comparable yields

(entry 6,7 vs 9,10). These products were readily isolated and purified by column chromatography to yield 62% and 53% of the nylon 12 and nylon 13 precursors.

Table 3.6. Hydrogenation of methyl 11-cyanoundec-9-enoate and 12-cyanododec-9- enoate[a]

n 9 11 9 n 9 CO2Me H2N CO2Me CO Me NC + 2 7 n 7 NC 7 10 10 10 n = 1: 9 n = 1: methyl 12-aminododecanoate (5) n = 1: 12 n = 2: 10 n = 2: methyl 13-aminotridecanoate (6) n = 2: 13

Cat. t-BuOK Pressure Time GC area (%) Entry n (mol %) (mol %) (Bar) (h) 5 or 6 12 or 13 By products 1 1 1 15 20 20 1.4 71.6 15.4 2 1 3 15 25 20[b] 6.8 55.6 29.6 3 1 1 30 20 17 10.4 83.0 1.6 4 1 2 30 20 17 78.3 10.6 6.2 5 1 3 30 20 17 68.5 2.3 17.4 6 1 3 30 25 20[b] 56.8 4.1 27.0 7 1 3 30 24 44 68.6 (62)[c] 2.6 16.2

8 2 1 30 20 19 41.0 (53)[c] 30.3 16.9

9 2 2 30 20 19 58.2 2.9 30.0

10 2 3 30 20 19[d] 43.3 2.1 42.2 [a] Reaction condition: 20 mg of methyl 11-cyano-9-undecenoate, 30 mol% of t-BuOK, 3 mol% of Grubbs 2nd generation catalyst, 3mL of chlorobenzene at 80 °C, 30 bars during 20 h with stirring. [b] Isolated yield. [c] Extending the reaction period for additional 45 h under otherwise same conditions did not improve the product profile.

While the starting material, oleic acid, can be supplied economically from many renewable resources, our method would be particularly useful if a crude lipid derived from microalgae can be used [4-6]. Thus the cross-metathesis reaction was tested using crude algal lipid containing a mixture of fatty acid methyl esters (FAMEs) obtained from algal biomass by our recently developed “reactive-extraction” technology which enable direct isolation of algal FAMEs in a single step.[30] We are pleased to find that both

65 cross-metathesis with acrylonitrile and allyl cyanide proceeded smoothly consuming only unsaturated FAMEs and leaving the saturated FAMEs behind (Table 3.6 and Figure 3-5).

Even under unoptimized conditions, both reactions with acrylonitrile (entry 1) and allyl cyanide (entry 2) provided the desired cyano ester 7 or 9 in close to comparable conversion but lower selectivity to control reactions (acrylonitrile: entry 3; allyl cyanide

Table 3.3, entry 3). To the extent of our knowledge, this is the first example when cross- metathesis was demonstrated using crude algal lipid. It is also of note that use of crude algal lipid that is rich in C9-unsaturated FAMEs should further increase the throughput from this biomass.

Data File C:\HPCHEM\1\DATA\KIM\SIG11419.D Sample Name: KN-03-01

======Injection Date : 9/5/2013 11:23:11 AM Seq. Line : 1 Sample Name : KN-03-01 Location : Vial 1 Acq. Operator : KIM Inj : 1 Acq. Instrument : Instrument 1 Inj Volume : 1 µl Acq. Method : C:\HPCHEM\1\METHODS\FAME.M Last changed : 8/7/2013 6:18:31 PM by YAPA Analysis Method : C:\HPCHEM\1\METHODS\COOLING.M Last changed : 9/23/2014 6:02:54 PM by ABEL (modified after loading) 15" 1" cooling

FID1 A, (KIM\SIG11419.D) 14" counts CO2Me (a)$ 7

50000

methyl oleate (1) 40000

Data File C:\HPCHEM\1\DATA\KIM\SIG11420.D Sample Name: KN-03-04

===30=0=0=0 ======CO2Me Injection Date : 9/5/2013 5:51:41 PM Seq. Line : 1 Sample Name : KN-03-04 Location : Vial 1 7 Acq. Operator : KIM Inj : 1 Acq. Instrument : Instrument 1 Inj Volume : 1 µl Acq20.00M0 ethod : C:\HPCHEM\1\METHODS\FAME.M Last changed : 8/7/2013 6:18:31 PM by YAPA Analysis Method : C:\HPCHEM\1\METHODS\COOLING.M Last changed : 9/23/2014 6:00:51 PM by ABEL methyl plamitoleate (14) (modified after loading) coo10l0i0n0 g

FID1 A, (KIM\SIG11420.D) 0 2 4 6 8 10 12 14 16 18 min counts

======CO Me (b)$ 27500 15" 2 Area Percent Report ======(=Z==)='=7==="======7 25000 Sorted By : Signal Multiplier (Z)'1:6" 1.0000 Dil2u2t5i00on : 1.0000 Use Multiplier & Dilut(ioEn )F'a1ct6or"with (IEST)Ds'7" methyl plamitate (15) No 2p0e0a00ks found Data File C:\HPCHEM\1\1DA"TA\KIM\SIG11425.D Sample Name: KN-03-08 ======17500 ======*=*=*==E=n=d==o=f==R=e=p=o=r=t==*=*=*======Injection Date : 9/12/2013 4:20:32 PM Seq. Line : 2 Sample Name : KN-03-08 Location : Vial 2 15000 Acq. Operator : KIM Inj : 1 Acq. Instrument : Instrument 1 Inj Volume : 1 µl NC CO2Me Acq1.250M0ethod : C:\HPCHEM\1\METHODS\FAME.M Last changed : 8/7/2013 6:18:31 PM by YAPA 7 Analysis Method : C:\HPCHEM\1\METHODS\COOLING.M Las1t000c0hanged : 9/23/2014 6:05:57 PM by ABEL 7 (modified after loading) cooling 7500 FID1 A, (KIM\SIG11425.D) (c)$ counts 0 2 4 6 8 10 12 14 16 18 min CO2Me NC 7 ===2=75=0=0======Area Percent Report15" 9 ======PDF2 5Creat000 or - PDF4Free v3.0 http://www.pdf4free.com Instrument 1 9/23/2014 6:02:56 PM ABEL Page 1 of 1 Sorted By : Signal Mul2t25i0p0lier : 1.0000 Dilution : 1.0000 (Z)'9" 1" Use Multiplier & Dilution Factor with ISTDs CO2Me 20000 No peaks found 11" (E)'9" 7 ===1=75=0=0======11 *** End of Report ***

15000

12500 NC 10000 7

7500 pent-2-enenitrile (16)

0 2 4 6 8 10 12 14 16 18 min

======Area Percent Report ======PDF Creator - PDF4Free v3.0 http://www.pdf4free.com InstSrourmteendt B1y9/23/2014 6:0:0:57 PMSiAgBnEaLl Page 1 of 1 Figure 3-5 (a)Multi pAlie rcrude algal: 1 .lipid0000 containing a mixture of FAMEs. (b) Cross-metathesis Dilution : 1.0000 with acrylonitrile.Use Multipl i(c)er & DCrossilution Fac-tmetathesisor with ISTDs with allyl cyanide. No peaks found

======3.4 Conclusions *** End of Report ***

We have demonstrated synthesis of nylon 12 and nylon 13 precursors from methyl oleate via cross metathesis reactions with allyl and homoallyl cyanides,

PDF Creator - PDF4Free v3.0 http://www.pdf4free.com respectively.Instr um eOurnt 1 9/ 2work3/2014 6: 0represents5:59 PM ABEL the shortest syntheticP aroutege 1 of 1 to these precursors from the abundantly available oleic acid. The key feature of our approach includes use of 1,4-

67 benzoquinone as additive, which completely suppressed the undesired isomerization side reaction during the metathesis step. Our approach not only enables preparation of nylon

12 and 13 precursors directly from methyl oleate, but also offers new pathways to access precursors of other nylons of different chain-length, economically and sustainably.

We acknowledge Kim Nguyen for his contribution to the project.

References for chapter 3

1. Palmer, R.J., Polyamides, plastics. Kirk-Othmer Encycl. Chem. Technol. (5th

Ed.), 2006. 19: p. 772-797.

2. Chauvel, A. and G. Lefebvre, Petrochemical Processes. Vol. 2. 1989: Editions

Technip.

3. Ragauskas, A.J., et al., The path forward for biofuels and biomaterials. Science,

2006. 311(5760): p. 484-489.

4. Hannon, M., et al., Biofuels from algae: challenges and potential. Biofuels, 2010.

1(5): p. 763-784.

5. Murzin, D.Y., P. Mäki-Arvela, and D.A. Aranda, Processing microalgae: beyond lipids. Biofuels, 2014. 5(1): p. 29-32.

6. Mata, T.M., A.A. Martins, and N.S. Caetano, Microalgae for biodiesel production and other applications: A review. Renewable and Sustainable Energy Rev., 2010. 14(1): p. 217-232.

7. Behr, A., A. Westfechtel, and J. Pérez Gomes, Catalytic Processes for the

Technical Use of Natural Fats and Oils. Chem. Eng. Technol., 2008. 31(5): p. 700-714.

8. Dubois, J.L., Method for synthesising omega-amino-alkanoic acids or the esters thereof from natural fatty acids. 2013, US Patent Office.

9. Couturier, J.-L. and J.-L. Dubois, Process for the synthesis of C11 and C12 omega-aminoalkanoic acid esters comprising a nitrilation step. 2013, (Arkema France,

Fr.). Application: WO. p. 21pp ; Chemical Indexing Equivalent to 158:332172 (FR).

10. Couturier, J.-L. and J.-L. Dubois, Preparation of α,ω-amino acids or esters using cross-metathesis. 2013, (Arkema France, Fr.). Application: FR. p. 20pp ; Chemical

Indexing Equivalent to 158:301890 (WO).

11. Couturier, J.-L. and J.-L. Dubois, Preparation of an ω-amino acid or ester from a monounsaturated fatty acid or ester. 2013, (Arkema France, Fr.). Application: WO. p.

27pp ; Chemical Indexing Equivalent to 158:301893 (FR).

12. Couturier, J.-L., et al., Preparation of saturated amino acids or saturated amino esters via metathesis reaction. 2011, (Arkema France, Fr.; Universite de Rennes 1; Centre

National de la Recherche Scientifique). Application: WO. p. 23 pp.

13. Dubois, J.-L. and J.-L. Couturier, Process for synthesis of omega-unsaturated nitrile acid/ester in which consecutively two types of swing cross metathesis processes are alternated. 2013, (Arkema France, Fr.). Application: FR. p. 39pp.

14. Dubois, J.-L. and J.-L. Couturier, Process for the synthesis of C11 and C12 omega-aminoalkanoic acid esters comprising a nitrilation step. 2013, (Arkema France,

Fr.). Application: FR. p. 18pp ; Chemical Indexing Equivalent to 158:360706 (WO).

15. Malacea, R., et al., Renewable materials as precursors of linear nitrile-acid derivatives via cross-metathesis of fatty esters and acids with acrylonitrile and fumaronitrile. Green Chem., 2009. 11(2): p. 152-155.

16. Miao, X., et al., Polyamide precursors from renewable 10-undecenenitrile and methyl acrylate via olefin cross-metathesis. Green Chem., 2012. 14(8): p. 2179-2183.

17. Miao, X., et al., Ruthenium–alkylidene catalysed cross-metathesis of fatty acid derivatives with acrylonitrile and methyl acrylate: a key step toward long-chain bifunctional and amino acid compounds. Green Chem., 2011. 13(10): p. 2911-2919.

18. Rybak, A. and M.A.R. Meier, Cross-metathesis of fatty acid derivatives with methyl acrylate: renewable raw materials for the chemical industry. Green Chem., 2007.

9(12): p. 1356-1361.

19. Rybak, A. and M.A.R. Meier, Cross-metathesis of oleyl alcohol with methyl acrylate: optimization of reaction conditions and comparison of their environmental impact. Green Chem., 2008. 10(10): p. 1099-1104.

20. Miao, X., et al., Tandem Catalytic Acrylonitrile Cross-Metathesis and

Hydrogenation of Nitriles with Ruthenium Catalysts: Direct Access to Linear α,ω-

Aminoesters from Renewables. ChemSusChem, 2012. 5(8): p. 1410-1414.

21. Miao, X., et al., Ruthenium-Benzylidenes and Ruthenium-Indenylidenes as

Efficient Catalysts for the Hydrogenation of Aliphatic Nitriles into Primary Amines.

ChemCatChem, 2012. 4(12): p. 1911-1916.

22. Kajetanowicz, A., A. Sytniczuk, and K. Grela, Metathesis of renewable raw materials—influence of ligands in the indenylidene type catalysts on self-metathesis of methyl oleate and cross-metathesis of methyl oleate with (Z)-2-butene-1,4-diol diacetate.

Green Chem., 2014. 16(3): p. 1579-1585.

23. Ternel, J., et al., Rhodium-Catalyzed Tandem Isomerization/Hydroformylation of the Bio-Sourced 10-Undecenenitrile: Selective and Productive Catalysts for Production of Polyamide-12 Precursor. Adv. Synth. Catal., 2013. 355(16): p. 3191-3204.

24. Greene, J.L., Jr., R.E. Burks, Jr., and I.A. Wolff, 13-Aminotridecanoic acid from erucic acid. Ind. Eng. Chem., Prod. Res. Develop., 1969. 8(2): p. 171-6.

25. Gurusamy, R., C. Zhang, and A. Gaffney, Nylon polymer and process. 2013, US

Patent Office.

26. Hong, S.H., et al., Prevention of Undesirable Isomerization during Olefin

Metathesis. J. Am. Chem. Soc. , 2005. 127(49): p. 17160-17161.

27. Bruneau, C., et al., Cross-metathesis with acrylonitrile and applications to fatty acid derivatives. Eur. J. Lipid Sci. Technol., 2010. 112(1): p. 3-9.

28. Samojłowicz, C., et al., The Doping Effect of Fluorinated Aromatic Solvents on the Rate of Ruthenium-Catalysed Olefin Metathesis. Chem.-Eur. J., 2011. 17(46): p.

12981-12993.

29. Werkmeister, S., K. Junge, and M. Beller, Catalytic Hydrogenation of Carboxylic

Acid Esters, Amides, and Nitriles with Homogeneous Catalysts. Org. Proc. Res. Dev.,

2014. 8 (2): p. 289–302.

30. Viamajala, S., D. Nelson, and R. Sims, Transesterified lipid recovery and methods thereof. 2011, (Utah State University, USA). Application: WO.

Chapter 4 Toward Sustainable Synthesis of PA12 (Nylon 12) Precursor from Oleic Acid Using Ring-Closing Metathesis

A paper published in ACS Sustainable Chemistry & Engineering1

Godwin A. Abel, Sridhar Viamajala, Sasidhar Varanasi and Kana Yamamoto

4.1 Introduction

There has been increasing appreciation for the use of renewable bio-sourced materials to complement or replace petroleum-based specialty chemicals [1, 2]. In particular, olefin metathesis is regarded as an effective tool for refining bio-based lipids for this purpose [3, 4]. In this context, we have recently reported [5, 6] new approaches for synthesizing high-order polyamide (PA, nylon) precursors from oleic acid (1), an abundant feedstock available from various renewable sources. One of our approaches[5] involves three steps using ring-closing metathesis to provide macrolactams having carbon length ranging from 11–13, monomers of PA11–13 (Scheme 4.1). Although the synthesis is concise, the ring-closing metathesis (step 2) requires a halogenated solvent

(chlorobenzene) [7], high temperature (120 °C), and relatively high catalyst loading(2 mol%) [3].

Scheme 4.1. Previously demonstrated route from oleic acid (1) to PA precursors [5].

While chlorinated solvents (chlorobenzene, dichloromethane, 1,2-dichloroethane) and aromatic solvents (benzene, toluene, xylene) are traditionally employed in metathesis reactions [8, 9], a recent study showed that less harmful solvents could replace these traditional solvents in olefin metathesis [10]. Also, significant progress has been made in improving catalyst efficiency for olefin metathesis in the past decade [11]. Catalysts that possess a wide range of electronically and sterically different N-heterocyclic carbene

(NHC) ligands and/or alkylidene ligands are now commercially available, allowing laboratory-scale screening of industrially-relevant catalysts and metathesis reactions [12].

Studies on catalyst immobilization [13] to facilitate their recovery and reuse have also been reported [14-18].

In this article, we present our studies aimed at establishing improved reaction conditions for ring-closing metathesis step (step 2) in our synthesis (Scheme 4.1), from homoallyloleamide (2) to ene-lactam (3), the precursor of PA12 monomer. First, a series of commercially available ruthenium alkylidene catalysts (Figure 4-1) were evaluated to identify the catalysts that are sufficiently active to promote the reaction with good selectivity at lower temperatures. With low temperature conditions, formation of

74 undesirable side-products may be minimized, and the stability and life-time of the catalyst improved [10]. The use of the identified catalysts that enable use of low-boiling and green solvents [10], could favorably impact the process from economic as well as environmental viewpoints. Next, the stability of each of the screened-catalysts in these reaction media was evaluated to establish the potential for its recyclability. Finally, the best catalyst was immobilized on several types of silica gel support and its recovery and reuse were demonstrated up to several cycles.

4.2 Materials and methods

4.2.1 General

The reagents were purchased from Sigma-Aldrich, STREM chemicals, or Alfa

Aesar, and used without further purification. All the solvents used for ring-closing metathesis were purchased from Sigma-Aldrich and deoxygenated by bubbling dry nitrogen gas for 20 min before being used. Reactions were carried out under inert conditions (argon) in a fume hood.

Thin layer chromatography (TLC) was carried out on glass-backed silica plates purchased from Sorbent Technologies Inc., Norcross GA. The reaction products were identified by visualizing the plates under UV (254 nm) light, and also by staining with potassium permanganate followed by gentle heating. Silica gel column chromatography was carried out using 20–60 micron dry silica purchased from Sorbent Technologies Inc.

1 13 H- and C-NMR spectra were acquired in CDCl3, on Brucker Avance 600 (600 MHz)

NMR spectrometers. Chemical shifts (δ) were reported as parts per million (ppm) with reference to tetramethylsilane (TMS) or solvent unless otherwise stated. The coupling constants (J) are reported in Hz. Mass spectra were obtained with Hewlett-Packard

Esquire Ion Trap LC-MS (electrospray). GC analyses were performed with HP 5890 series II equipped with FID and an auto-sampler (HP controller 7672A) and Biodiesel TG column (5% diphenyl, 95% dimethyl polysiloxane, 15m, 0.33mm ID and 0.10 µm dF) or

MXT biodiesel TG (Siltek – treated stainless steel) column. GC analysis of samples was carried out using the following protocol: 60 – 370 °C at 10 °C/min and 6 min hold.

4.2.2 Experimental procedures

Representative reaction conditions for ring-closing metathesis of homoallyloleamide (2)

This work, which is an effort to optimize the reaction conditions for RCM approach for the synthesis of PA12 precursors from oleic acid, builds on our previous publication in which the details regarding the product identification and characterization were provided [5]. Hence, only a brief outline of the batch reaction conditions is given here. The procedure described represents the protocol for a specific catalyst (C9) in ethyl acetate. A similar procedure is implemented with all the catalysts screened and the solvents used in the study.

N-(But-3-en-1-yl)oleamide (44.4 mg, 0.1323 mmol) was dissolved in ethyl acetate (31 mL) and heated to 60 °C and was maintained at this temperature for 20 min.

0.8 mg (0.00097 mmol) of catalyst C9 (M74SIPr; see Fig 4-1) was dissolved in ethyl acetate (1 mL) and added to the reaction mixture. The solution was kept for 15 min at this temperature, before quenching the reaction with ethyl vinyl ether. After being cooled to room temperature, the reaction mixture was concentrated usign a rotary evaporator under reduced pressure. The crude residue was purified by column chromatography using first hexanes/ethyl acetate (7/3) and then acetone/hexanes (2/8) as eluents to provide the

76 desired ene-lactam as a white crystalline solid (19.6 mg, 75.9%). 1H NMR spectra matched those reported in the literature [5].

1 H NMR (600 MHz, CDCl3) δ 5.32–5.55 (m, 4H), 3.26–3.36 (m, 2H), 2.02–2.34 (m, 8H),

1.62–1.65 (m, 2H), 1.18–1.48 (m, 12H).

The reaction conversions and yields were reported in GC area%. They were calculated by integrating all the peaks in the GC chromatogram, excluding the known peaks that should not be taken into mass balance (such as 1-decene, 1-decene dimers).

There may be trace amounts of other oligomers that formed from isomerization and subsequent metathesis of 1-decene, which were not taken into account in the calculation.

However, we believe that our approach provides data that are adequate for this qualitative study.

Procedure for testing the stability of the homogeneous catalysts

For establishing the catalyst stability, we used the same catalyst repeatedly in a fed-batch process. Accordingly, after the reaction is carried out for 15 minutes following the above procedure, 1mL of the reaction mixture is withdrawn from the reactor to determine the conversion of the substrate and product yield, and a fresh batch of the substrate in 1 mL of reaction solvent is simultaneously loaded into the reactor for the next reaction cycle.

Catalyst Immobilization

Immobilization of metathesis catalysts on silica gel was carried out according to the following procedure from literature [14]. To a round-bottom flask containing toluene

(5 mL), silica gel (223 mg) and M74SiPr catalyst (9 mg) were added at room temperature and kept under N2 atmosphere for 4 h. The mixture was then vacuum-filtered and washed

77 several times with hexane and placed under reduced pressure overnight, prior to usage.

The immobilized-catalysts used in Table 4.4 were prepared with the following reagents:

M74SiPr (13.1 mg)/SBA-15 (422 mg); M74SiPr (12.7 mg)/MCM-41 (397 mg)).

Procedure for testing the stability of the immobilized catalyst

The dried catalyst-loaded silica gel (31.5 mg) was placed in a three-necked flask, then ethyl acetate (15 mL) was added and the mixture was heated to 60C. Subsequently,

N-(But-3-en-1-yl)oleamide (2, 11 mg) in ethyl acetate (1 mL) was charged to initiate the reaction and the mixture was maintained at 60C for 15 min. After 15 min, the reaction mixture was cooled to room temperature, and the clear liquid portion was removed with a syringe. The liquid was passed through a short pad of silica gel (with a mixture of ethyl acetate/hexanes and then with acetone/hexanes) to remove any leached catalyst that entrained with the liquid portion prior to GC analysis. The combined eluents from the silica pad were concentrated and re-dissolved in methanol for GC analysis. The subsequent experiments were performed by addition of a fresh solution of N-(But-3-en-1- yl)oleamide (2, 11 mg) in ethyl acetate (16 mL) to the flask containing the catalyst- loaded silica gel from the previous run, and repeating the above procedure. This approach allowed the use of the same immobilized catalyst for multiple runs for establishing the catalyst stability.

4.3 Result and discussion

Rationale behind catalyst library selection for screening

A series of commercially-available ruthenium catalysts were screened based on a semi-empirical rationale that is governed by (i) our own previous experience and (ii)

78 other studies in literature on the comparative evaluation of catalysts for metathesis reactions using model experiments [12, 19].

We have first excluded the catalysts without N-heterocyclic carbene (NHC) ligands because of their lower reactivity and stability [20]. In addition, we have previously observed that Hoveyda-Grubbs II catalyst (Fig.4-1, C1), but not Grubbs II catalyst, was active for the specific RCM reaction system being studied;[5], [21] therefore, the boomerang-type catalysts [22] as well as the catalyst with labile ligands were selected for this study (Figure 4-1).

In general, it has been recognized that catalysts with protective ligand

(boomerang) that is easier to release led to faster catalyst initiation, allowing the metathesis reaction to be conducted at lower temperature. However, fast catalyst initiation does not necessarily correlate to high activity/product yield due to the potential for decomposition of the active-form of the initiated catalyst [23-27]. In addition, many studies on tuning NHC ligands to improve catalyst performance have also been reported

[28], and catalysts with sterically hindered NHC ligands have shown better selectivity and stability for certain ring-closing metathesis reactions [12, 27, 29].

It is evident from the foregoing discussion that olefin metathesis is a rapidly advancing field of research wherein new types of catalysts are being continually developed, however there is not yet a standardized method for catalyst evaluation.

Moreover, the large number of variables to consider when developing new processes makes the prediction of appropriate choice of catalyst rather difficult. We, therefore, reasoned that when developing new metathesis methodologies, the evaluation of a judiciously chosen library of catalysts can be extremely valuable [10, 12, 19].

The following ruthenium complexes were ultimately chosen in this study (Figure

4-1): (a) Bench-mark catalyst: Hoveyda-Grubbs II (C1) [30], (b) “Fast initiating” or

“highly active”: Grela catalyst (C2) [25, 31, 32]; GreenCat catalyst (C4) [33][34];

Umicore M7 series [27, 35, 36]; M71SiPr (C6), M72SiPr (C7), M73SiPr (C8), M74SiPr

(C9); (C10) [26]; (C11) [37][38]; (c) Catalysts effective withunreactive or hindered olefins: Stewart-Grubbs catalyst (C3) [39, 40], (C5) [40]; (C10) [26], (d) Others: (C12)

[41].

Figure 4-1. Commercially available metathesis catalysts screened in this study. The major structural elements of these catalyst systems include a ruthenium cluster with an NHC (N-heterocyclic carbene) ligand and a protective ligand (as indicated in color with the bench-mark catalyst C1). Release of the protective ligand in the reaction medium generates the active catalyst - a process known as catalyst initiation. The NHC ligand contributes to the stability of the catalyst.

A detailed screening of this catalyst library was undertaken with the aim of answering the following questions:

(1) Whether the chlorinated solvents (chlorobenzene) could be replaced with more sustainable low-boiling solvents, such as ethyl acetate or hexane; (2) The lowest temperature that allows the reaction to go to completion while maintaining reactivity and selectivity; (3) The most stable catalysts, among the ones that provide high conversion and selectivity at low temperatures; (4) The possibility of immobilizing the catalyst while retaining its activity and stability.

Optimization of reaction conditions

Since one of our goals was to replace the chlorinated solvent (chlorobenzene) employed in the original procedure with greener solvents, we have evaluated two other alternative solvents [10] as reaction media during the catalyst evaluation study. While more specifics regarding detailed catalyst screening were presented later, Table 4.1 summarizes results pertinent to reaction-solvent choice.

Table 4.1. Ring-closing metathesis of homoallyloleamide (2) – solvent comparison.

Entry[a] Cat. Time Temp. Solvent Conv. 3 Dimers

(min) (C)[b] (%)[c] (%)[c] (%)[c]

1[d] C1 15 120 PhCl 96 70 11

2 C1 15 60 EtOAc 95.1 55.1 21.1

3[e] C9 15 60 EtOAc 93.2 71.5 (76)[f] 16.5

4 C9 15 60 PhCl 42.7 – –

5 C9 60 60 PhCl 88.9 70.2 16.3

6[e][h] C9 15 60 Hexanes 90.3 77.4 12.9

(56)[f][h]

[a] Reaction conditions: 1 mol% of the catalyst in a solvent (1 mL) was added into a solution of homoallyloleate (0.033 mmol) in EtOAc (15 mL) in one portion (2 mM final concentration). The reaction was kept at indicated temperatures before samples were taken for GC analyses. [b] Oil bath temperature. [c] GC area%. [d] Taken from ref.[5] [e] 4 mM concentration. [f] Isolated yield. [g] The sample solution was prepared by re- dissolving the concentrated reaction mixture in chloroform. [h] Isolated yield from the reaction run at 2 mM concentration.

First, low temperature reactions in ethyl acetate(60 °C) showed conversions similar to reactions performed in chlorobenzene at a much higher temperature (120 °C) with the same bench-mark catalyst (C1), although a slight increase of substrate dimerization was observed in ethyl acetate. (Table 4.1, entries 1–2) [42]. Indeed, when we compared the same two solvents with one other promising catalyst from our library

(i.e., C9) we found that the reaction proceeds four times faster in ethyl acetate under otherwise identical conditions, while providing higher conversions and yields with no relative increase in substrate dimerization (entries 3–5). We also investigated hexane as prospective reaction medium because, as discussed later, the non-polar nature of hexane also proves advantageous with regards to implementing the reaction with immobilized- versions of the catalysts. As can be noted from entries 3 and 6 in Table 4.1, the reaction kinetics and conversions are very similar in both ethyl acetate and hexanes. Interestingly, however, in hexanes the side-products (dimers) precipitated out, presumably as a mixture of isomers, allowing their isolation by filtration [43]. As seen in Table 4.1, the isolated and GC-based product yields agree very closely in case of ethyl acetate (entry 3)where as there is some discrepancy between these two yields in case of hexanes (entry 6).We suspect the formation of higher-order oligomers in hexane that precipitated out of the reaction mixture and were not accounted for in the GC area computations are the reason for the observed discrepancy. Nevertheless, this feature of oligomer precipitation in hexanes would make the product purification less cumbersome as the reaction medium has fewer soluble impurities. Overall, because of higher reaction rates, lower toxicity, and the price, both ethyl acetate and hexanes are attractive alternative solvents for our chemistry. Based on the ease of analysis, majority of the screening experiments reported below were performed using only ethyl acetate as the reaction solvent.

As already noted, reaction temperature has a profound effect on the catalyst degradation rates and lower reaction temperatures could significantly extend the catalyst life. There is evidence in literature that when olefin metathesis reaction is conducted at

80 C using immobilized Hoveyda-Grubbs II catalyst (C1), the catalyst could be recycled

84 successfully for multiple cycles [16]. Since our initial experiments with low-boiling solvents as reaction media have established the feasibility of conducting the reaction at temperatures lower than 80C, we wanted to verify how much further we could lower the reaction temperature without experiencing significant reduction in kinetics and yields.

Table 4.2 compares the performance of three different catalysts (C2, C4 and C6) at 60C and room temperature (22C).

Table 4.2. Ring-closing metathesis of homoallyloleamide (2) – temperature comparison.

Entry[a] Cat. Temp. Time Conv. 3 Dimers

(C)[b] (min) (%)[c] (%)[c] (%)[c]

1 C2 60 15 92.9 54.7 20.6

2 C2 22 300 93.1 33.8 53.9

3 C4 60 15 96.4 70.0 26.4

4 C4 22 300 95.8 33.7 61.3

5 C6 60 15 90.1 66.3 17.2

6 C6 22 300 93.7 58.9 31.7

[a] Reaction conditions: 1 mol% of the catalyst in EtOAc (1 mL) was added into a solution of homoallyloleate (0.033 mmol) in EtOAc (15 mL) in one portion (2 mM final concentration). The reaction was kept at indicated temperatures before samples were taken for GC analyses. [b] Oil bath temperature. [c] GC area%.

Two general trends could be seen in the data. First, at room temperature, the reaction kinetics became unacceptably slow. Second, lower temperatures lead to increased oligomer formation, the known trend explained by the entropy difference leading to the two products [44]. Thus, it appears that 60C forms the threshold temperature for conducting the oleamide metathesis reaction, and further significant 86 reduction of reaction temperatures below 60C may not be viable. At 60 C, extending the reaction time beyond 15 minutes was not beneficial, as there is no further improvement in the conversion and the yield of the desired product (not shown).

Overview of optimal performance results for the catalyst library screened

The results on the optimal performance (time that reached the highest conversion) of the individual members of the chosen catalyst library are compiled in Figure 4-2. The figure shows the selectivity towards desired ene-lactam (3) at near-complete conversion of the substrate for all the catalysts that displayed activity at 60C (when used fresh) in ethyl acetate. The benchmark catalyst C1 and catalyst C2 were the two least selective catalysts, while catalysts C3, C4, C5, and C9 showed high selectivity. It is worth noting that some of the catalysts required longer reaction times than others to achieve optimal performance. Also, unlike the rest of the catalysts in the figure, only ~50% conversion was achieved with catalyst C10, though it did display catalytic activity at the chosen reaction conditions. The results using catalyst C11 and C12 is not included in the figure, as the reaction does not occur under the said reaction conditions even after several hours.

Although several of the catalysts of the library showed promise when freshly used for producing ene-lactam from oleamide, for economic viability they should also retain their stability during reuse for multiple cycles of the reaction. Accordingly, we also studied the stability of all the catalysts in the library.

Figure 4-2. Ring-closing metathesis of homoallyloleamide (2)– catalyst comparison when the highest conversion is achieved. [a] Reaction conditions: See Table 2. [b] % conversion at 15 min with catalysts C5: 82.6%; C7: 51.4 %; C8: 91.6%; C9: 42.7%. [c] Reaction using C10 stalled after 15 min.

Assessment of catalyst stability

We investigated the stability of each of the catalysts of our library in a “fed-batch homogeneous catalysis” mode of usage (Table 4.3). The experiment was conducted with successive additions of the substrate into the reaction mixture after every 15 minutes.

This fed-batch approach allows the evaluation of catalyst stability over multiple runs. A

88 portion of the reaction mixture was sampled for analysis at the start of run 1 and also at the end of each consecutive cycle (Table 4.3) [45].

The catalyst stability estimates shown in Table 4 per each cycle are based on the

% conversion of the combined amounts of substrate put into the reaction medium until the completion of that cycle. In this sense, the percentages reported reflect the average activity displayed by the catalyst over those cycles. The effectiveness of a given catalyst is assessed based on its stability as well as its selectivity to the desired product. As our method of stability estimate of the catalyst is based on “percent of substrate converted” alone, this data has to be evaluated in combination with the selectivity data (Figure 4-2) for fresh catalysts, to establish the value of a given catalyst for this reaction.

Since catalysts C3, C4, C5, and C9 showed high selectivity (Figure 4-2), a comparison of the stability of these catalysts (Table 4.3) reveals that catalyst C9 is the most effective of all for oleamide metathesis reaction: it retains enough activity to achieve above 70% conversion of substrate even after five cycles of reuse with a selectivity exceeding 72%. C3 and C4 appear to be the next two promising candidates.

They display very similar stability and selectivity characteristics: both display high selectivity and reasonable stability. In addition, C3, C4 and C5 all have fast reaction kinetics. Thus, these catalysts could be viable candidates to consider for immobilization on supports for large-scale processes. Finally, catalysts C5, C7, and C8 also display good stability and selectivity, but would require longer reaction times (Figure 4-2) under the tested reaction conditions. Of all the catalysts evaluated, catalyst C10, with the lowest activity, was the least stable and deactivated after a single cycle (entry 10) and was deemed non-viable for our application.

Since catalyst C9 showed the best performance characteristics in ethyl acetate, we studied its stability in hexanes as well (entry 11). It is gratifying to observe that it displayed comparable stability in this solvent as in ethyl acetate (entries 9 vs 11).

Table 4.3. Stability evaluation of the selected Ru catalysts.

Cycle / Conv. (%) [b]

Entry[a] Catalyst 1 2 3 4 5

1 C1 95.1 82.6 58.4 41.1 30.3

2 C2 92.9 76.2 53.9 40.1 31.7

3 C3 92.4 89.7 76.5 62.8 50.0

4 C4 96.4 92.8 84.7 69.1 52.6

5 C5 82.3 87.1 82.5 84.1 63.7

6 C6 90.1 87.5 80.8 70.5 58.7

7 C7 51.4 78.2 73.9 65.0 55.8

8 C8 91.6 87.0 85.7 80.5 74.0

9 C9 93.2 88.6 84.3 80.8 72.1

10 C10 52.4 16.3 10.8 7.6 5.9

11[c][d] C9 95.5 88.9 82.1 70.6 61.5

[a] Homoallyloleamide (2, 0.033 mmol) in EtOAc (1 mL) was added every 15 min to a solution containing 1 mol% (initial conc.) of the catalyst in EtOAc (15 mL) at 60 C. A portion (1 mL) of the solution was taken for GC analyses at 15 min intervals prior to addition of the next portion (1 mL) of substrate solution. [b] GC area %. [c] Hexanes were used as solvent. [d] Because of the low solubility of oligomers in hexanes, precipitated oligomers were not taken in GC samples, which were prepared from the supernatant; therefore the actual % conversions could be higher than the observed % conversions.

Structure-activity relationship

Hoveyda-Grubbs II catalyst (C1) has been the most-widely used catalyst to-date for ring-closing metathesis (RCM) macrocyclizations [46]. It has been established that this mechanism leads to improved catalyst initiation and stability in both CM and RCM reactions [47]. More recently, Umicore M7 catalysts(in which the mesityl groups in the

NHC ligand of C1 are replaced by 2,6-diisopropylphenyl groups and/or H and electronegative substituents are attached to isopropoxystyrene ligand)(C6–C9) have been reported [12] to be more effective than C1for certain RCM macrocyclizations. It is expected that the higher steric hindrance of the NHC groups in M7 catalysts contribute to improved stability, while the higher electronegativity of the protective ligand contributes to faster catalyst initiation over C1. Our results showed that several structurally distinct catalysts (C3, C4, C5, and C9) out-performed the most popular Hoveyda-Grubbs II catalyst (C1) for oleamide to ene-lactam RCM transformation. Among them, C9 does belong to M7 class of catalysts, while C5 is structurally analogous to M7 class catalyst.

An important structural feature that distinguishes the two other viable catalysts C3 and

C4, which do not belong to M7 class, is the steric hindrance of its NHC ligand. Our observation that C3 is less stable than C4 (Table 4.3: entries 3 and 4) is consistent with this feature. Indeed, C3 has originally been developed for metathesis reactions involving hindered-olefins to provide easier access of the active site to the substrate [39, 40]. In our specific system, which involves non-hindered olefin (oleamide), the substrate is able to access the catalytic active site even with bulky NHC ligands, while at the same time such bulky NHC ligands impart better stability to the catalyst.

Overall, our results confirmed the difficulty of predicting catalyst performance a priori. Although the literature provides useful guidance for selecting subset of catalysts through qualitative reasoning, the ultimate catalyst selection requires experimental evaluation with respect to the specific substrate.

Immobilization of the Catalyst

With the best-performing catalyst identified, we have investigated the possibility of immobilizing catalyst C9 in porous supports and conducting the RCM reaction in heterogeneous catalytic mode amenable to continuous operation. Mesoporous silica molecular sieves such as MCM-41, SBA-15 and SBA-1 are well-characterized and highly stable supports ideally suited for catalyst immobilization. Their structure involves cage- like nano-cavities connected via nano-pores, and the relative size difference between the nono-cavities and the 3-D structure of the catalyst determines the number of catalyst molecules that could fit in a cavity. As dimerization of ruthenium clusters is known to lead to loss of catalyst activity, molecular sieves that could accommodate only a single catalyst molecule per cavity were expected to provide more stable catalysts; this was demonstrated for C1 catalyst immobilized in SBA-1 molecular sieve with a few specific

RCM and CM reactions [48]. Based on these reports, catalyst C9 was adsorbed on two different types of mesoporous silica gels SBA-15 and MCM-41 using previously established procedures [14, 48], as described in the Experimental section. Although the immobilized catalysts were as active as the homogenous catalysts, as the immobilization mechanism involved bonding of the catalyst to silica surface only through physical forces, the catalyst would desorb from silica in polar reaction media. Accordingly, we conducted the metathesis reaction in the non-polar hexanes, as catalyst C9 was effective

92 in both ethyl acetate and hexanes. The reactions were performed in a batch mode, as outlined in the Experimental section using C9 immobilized on SBA-15 as well as MCM-

41 (Table 4.4). We did not observe any drop in activity with MCM-41supported catalyst for three cycles, while there is some drop in activity during the third cycle with the SBA-

15 supported catalyst. This observation may be due to the average size of the nono-pores in MCM-41 (2.6 nm), which is known to be smaller than those in SBA-15 (6.2 nm).

Thus, it is likely that the MCM-41 does not permit more than one C9 catalyst molecule per cavity on the average, while with SBA-15 multiple molecules can occupy a single cavity leading to catalyst deactivation via dimerization [48].

An issue specific to hexanes as reaction medium, as already noted, is the precipitation of substrate oligomerization products from the reaction medium. While this phenomenon simplifies the product isolation, the precipitated products could be passed on to the next reaction cycle along with the catalyst particles. Use of other non-polar solvents in which oligomers stay soluble could alleviate or eliminate this issue. In addition, catalyst leaching and decomposition could also contribute to loss of activity.

On the other hand, issues with catalyst leaching from mesoporous silica by polar solvents can be circumvented by use of physical immobilization of catalyst, a more recently developed molecular sieves encapsulation method [49].

Thus, while we were able to establish the feasibility of catalyst immobilization at a proof-of-concept level, testing of other novel immobilization methods that may provide better stability to the catalyst [48-50] and would also allow the use of other polar and nonpolar solvents as reaction media [9] should be investigated for future development of nylon precursor synthesis using RCM.

Table 4.4. Demonstration of recovery and reuse of C9 catalyst immobilized on silica gel [a] Cycle / Conv. (%) [b]

Entry Support 1 2 3

1 SBA-15 98 96 56

2 MCM-41 >99 92 93

[a] Reaction conditions: Homoallyloleamide (2, 0.033 mmol) in Hexanes (16 mL) was added to a three-neck flask containing the heterogeneous (solid) catalyst and the reaction was run at 60 C. [b] GC area %

4.4 Conclusions

In conclusion, improved reaction conditions for our recently developed process for the production of PA12 precursor were established. Through the screening of commercially available metathesis catalysts under various reaction conditions, we identified a catalyst with high activity, selectivity and stability for our substrate. The screening information may be useful for optimizing other ring-closing metathesis reactions generating macrocycles. Our new process features (a) lower reaction temperature (60 °C vs 120 °C) and (b) use of environmentally benign solvent (ethyl acetate/hexanes vs chlorobenzene) at (c) relatively low catalyst loading (1 mol%).The identified catalyst was adsorbed on mesoporous silica gel to aid its recovery and reuse. Its recycling was demonstrated up to three cycles using hexanes without loss of catalytic activity. Although the reaction conditions established in this study require further refinement before the process can be scaled economically, we believe they represent an

94 important step forward toward environmentally sustainable production of polyamide precursors from oleic acid.

______

1 Adapted with permission from ACS Sustainable Chem. Eng. 2016, 4 (10), 5703-5710.

References for chapter 4

1. Yelchuri, V., B.N.P. Rachapudi, and S.L.K. Mallampalli, Synthesis of industrially important platform chemicals via olefin metathesis of palash fatty acid methyl esters. Eur.

J. Chem., 2014. 5(3): p. 532-535.

2. Murzin, D.Y., P. Mäki-Arvela, and D.A. Aranda, Processing microalgae: beyond lipids. Biofuels, 2014. 5(1): p. 29-32.

3. Chikkali, S. and S. Mecking, Refining of Plant Oils to Chemicals by Olefin

Metathesis. Angew. Chem. Int. Ed. , 2012. 51(24): p. 5802-5808.

4. Winkler, M. and M. Meier, Olefin cross-metathesis as a valuable tool for the preparation of renewable polyesters and polyamides from unsaturated fatty acid esters and carbamates. Green Chem., 2014. 16: p. 3335-3340.

5. Yapa Mudiyanselage, A., et al., Simple Ring-Closing Metathesis Approach for

Synthesis of PA11, 12, and 13 Precursors from Oleic Acid. ACS Sustainable Chem. Eng.,

2014. 2(12): p. 2831-2836.

6. Ameh Abel, G., et al., Cross-metathesis approach to produce precursors of nylon

12 and nylon 13 from microalgae. RSC Adv., 2014. 4(98): p. 55622-55628.

7. Chlorinated or aromatic solvents posed regulatory issues due to safety and environmental concerns.

8. Kotha, S. and M.K. Dipak, Strategies and tactics in olefin metathesis.

Tetrahedron, 2012. 68(2): p. 397-421. 96

9. Buchmeiser, M.R., Theory and Practice, in Olefin Metathesis, K. Grela, Editor.

2014, John Wiley & Sons. . p. 495-514.

10. Skowerski, K., et al., An attempt to provide an environmentally friendly solvent selection guide for olefin metathesis. Green Chem., 2014. 16(3): p. 1125-1130.

11. Naresh, P., et al., ChemInform Abstract: Recent Advances of Olefin Metathesis and It′s Applications in Organic Synthesis. Int. J. Pharm. Technol. , 2012. 43(26): p.

1748-1767.

12. Caijo, F., et al., Screening of a selection of commercially available homogeneous

Ru-catalysts in valuable olefin metathesis transformations. Cat. Sci. Technol. , 2013.

3(2): p. 429-435.

13. Buchmeiser, M. R., In Olefin Metathesis: Theory and Practice. Grela, K.; Editor,

John Wiley & Sons. : 2014; pp 495-514.

14. Van Berlo, B., et al., Silica Immobilized Second Generation Hoveyda-Grubbs: A

Convenient, Recyclable and Storageable Heterogeneous Solid Catalyst. Angew. Chem.

Int. Ed., 2008. 350(13): p. 1949-1953.

15. Monge-Marcet, A., et al., Catalytic applications of recyclable silica immobilized

NHC-ruthenium complexes. Tetrahedron, 2013. 69(1): p. 341-348.

16. Monge-Marcet, A., et al., Sol–gel immobilized Hoveyda–Grubbs complex through the NHC ligand: A recyclable metathesis catalyst. J. Mol. Catal. A: Chem.

, 2012. 357: p. 59-66.

17. Zhang, H., et al., Grubbs-type catalysts immobilized on SBA-15: A novel heterogeneous catalyst for olefin metathesis. J. Mol. Catal. A: Chem. , 2013. 372: p.

35-43.

18. Yao, Q., A Soluble Polymer‐Bound Ruthenium Carbene Complex: A Robust and Reusable Catalyst for Ring‐Closing Olefin Metathesis. Angew. Chem. Int. Ed.,

2000. 39(21): p. 3896-3898.

19. Bieniek, M., et al., In an attempt to provide a user's guide to the galaxy of benzylidene, alkoxybenzylidene, and indenylidene ruthenium olefin metathesis catalysts.

Chem.-Eur. J., 2008. 14(3): p. 806-818.

20. Nelson, D.J., et al., Key processes in ruthenium-catalysed olefin metathesis.

Chem. Commun. , 2014. 50: p. 10355-10375.

21. For a mechanistic study that showed a higher reactivity of the two catalysts:

Bates, J. M.; Lummiss, J. A. M.; Bailey, G. A.; Fogg, D. E., Operation of the Boomerang

Mechanism in Olefin Metathesis Reactions Promoted by the Second-Generation Hoveyda

Catalyst. ACS Catalysis 2014, 4 (7), 2387-2394.

22. Bates, J.M., et al., Operation of the Boomerang Mechanism in Olefin Metathesis

Reactions Promoted by the Second-Generation Hoveyda Catalyst. ACS Catalysis, 2014.

4(7): p. 2387-2394.

23. Bieniek, M., et al., Advanced Fine-Tuning of Grubbs/Hoveyda Olefin Metathesis

Catalysts: A Further Step toward an Optimum Balance between Antinomic Properties. J.

Am. Chem. Soc. , 2006. 128(42): p. 13652-13653.

24. Borré, E., et al., Recent advances for controlling the activity and the recoverability of Hoveyda-Grubbs type olefin metathesis catalysts. Chimica oggi, 2008.

26(5): p. 89-92.

25. Olszewski, T., et al., A New Tool in the Toolbox: Electron-Withdrawing Group

Activated Ruthenium Catalysts for Olefin Metathesis. Synlett, 2013. 24(08): p. 903-919.

26. Love, J.A., et al., A Practical and Highly Active Ruthenium‐Based Catalyst that

Effects the Cross Metathesis of Acrylonitrile. Angew. Chem. Int. Ed. , 2002. 41(21): p.

4035-4037.

27. Clavier, H., et al., Towards Long-Living Metathesis Catalysts by Tuning the N-

Heterocyclic Carbene (NHC) Ligand on Trifluoroacetamide-Activated Boomerang Ru

Complexes. Liesbig Chem Annalen, 2009. 2009(25): p. 4254-4265.

28. Lummiss, J.A.M., et al., The divergent effects of strong NHC donation in catalysis. Chem. Sci. , 2015. 6: p. 6739-6746.

29. Doppiu, A., et al., Synthesis Optimization and Catalytic Activity Screening of

Industrially Relevant Ruthenium-Based Metathesis Catalysts. Top. Catal. , 2014. 57(17-

20): p. 1351-1358.

30. Garber, S.B., et al., Efficient and Recyclable Monomeric and Dendritic Ru-Based

Metathesis Catalysts. J. Am. Chem. Soc. , 2000. 122(34): p. 8168-8179.

31. Grela, K., S. Harutyunyan, and A. Michrowska, A Highly Efficient Ruthenium

Catalyst for Metathesis Reactions. Angew. Chem. Imt. Ed. , 2002. 41(21): p. 4038-4040.

32. Michrowska, A., et al., Nitro-Substituted Hoveyda−Grubbs Ruthenium

Carbenes: Enhancement of Catalyst Activity through Electronic Activation. J. Am.

Chem. Soc. , 2004. 126(30): p. 9318-9325.

33. Skowerski, K., et al., Efficient, durable and reusable olefin metathesis catalysts with high affinity to silica gel. Tetrahedron, 2013. 69(35): p. 7408-7415.

34. Affinity to silica gel that aids recovery.

35. Mauduit, M., H. Clavier, and I. Laurent, Ruthenium-based catalytic complexes and the use of such complexes for olefin metathesis. 2013, (Umicore AG & CO KG):

France.

36. Rix, D., et al., Aminocarbonyl Group Containing Hoveyda−Grubbs-Type

Complexes: Synthesis and Activity in Olefin Metathesis Transformations. J. Org. Chem. ,

2008. 73(11): p. 4225-4228.

37. Keitz, B.K., et al., Improved Ruthenium Catalysts for Z-Selective Olefin

Metathesis. J. Am. Chem. Soc. , 2012. 134(1): p. 693-699.

38. Z-olefin selective.

39. Stewart, I.C., et al., Highly Efficient Ruthenium Catalysts for the Formation of

Tetrasubstituted Olefins via Ring-Closing Metathesis. Org. Lett. , 2007. 9(8): p. 1589-

1592.

40. Stewart, I.C., C.J. Douglas, and R.H. Grubbs, Increased Efficiency in Cross-

Metathesis Reactions of Sterically Hindered Olefins. Org. Lett. , 2008. 10(3): p. 441-444.

41. Ung, T., et al., Latent Ruthenium Olefin Metathesis Catalysts That Contain an N-

Heterocyclic Carbene Ligand. Organometallics, 2004. 23(23): p. 5399-5401.

42. We observed many side-products in chlorobenzene, which may be due to solvent impurities. Such a case in toluene has been reported: Nicola, T.; Brenner, M.; Donsbach,

K.; Kreye, P., Org. Proc. Res. Dev. 2005, 9 (4), 513-515.

43. The isolated oligomers appreared to be a mixture of cyclic dimers (head-to- head/head-to-tail, as well as E/Z isomers) from MS and NMR analyses.

100

44. A similar trend has been reported: Yamamoto, K.; Biswas, K.; Gaul, C.;

Danishefsky, S. J., Effects of temperature and concentration in some ring closing metathesis reactions. Tetrahedron Lett. 2003, 44 (16), 3297-3299.

45. In this study, the sampling loss of the catalyst is 6.25% per cycle, which accounts for ~23% loss at 5th cycle. However, we think the major culprit for % conversion decrease is catalyst degrading, based on the control experiments: (1) Comparable % conversion at 0.5 and 1 mol% catalyst loading under the otherwise identical reaction conditions; (2) No product inhibition was seen when isolated products (including oligomers and impurities) were spiked in.

46. van Lierop, B.J., J.A.M. Lummiss, and D.E. Fogg, Ring‐Closing Metathesis, in

Olefin Metathesis: Theory and Practice. 2014, John Wiley & Sons, Inc.: Hoboken,

NJ, USA. p. 85-152.

47. For a mechanistic study that showed a higher reactivity of the two catalysts:

Bates, J. M.; Lummiss, J. A. M.; Bailey, G. A.; Fogg, D. E., Operation of the Boomerang

Mechanism in Olefin Metathesis Reactions Promoted by the Second-Generation Hoveyda

Catalyst. ACS Catalysis 2014, 4 (7), 2387-2394.

48. Yang, H., et al., Hoveyda–Grubbs catalyst confined in the nanocages of SBA-1: enhanced recyclability for olefin metathesis. Chem. Commun. , 2010. 46(45): p. 8659-3.

49. Li, Q., T. Zhou, and H. Yang, Encapsulation of Hoveyda–Grubbs 2ndCatalyst within Yolk–Shell Structured Silica for Olefin Metathesis. ACS Catalysis, 2015. 5(4): p.

2225-2231.

101

50. Yang, H., et al., Encapsulation of an Olefin Metathesis Catalyst in the Nanocages of SBA-1: Facile Preparation, High Encapsulation Efficiency, and High Activity.

ChemCatChem, 2013. 5(8): p. 2278-2287.

102

Chapter 5

Summary

Three projects with the purpose of developing an integrated approach for producing renewable high value chemicals from microalgae via environmentally sustainable pathways were studied. Our overall goal is the production of chemicals of industrial value from algae/algae constituents. Herein, is the summary of the results of each chapter/project.

In the study described in chapter two, we investigated an enzymatic fractionation method for cell disruption/isolation of native components directly from microalgae biomass. The new method lowers cost and energy required over the traditional approaches due to the mild processing conditions and would allow for the isolation of the simple sugars in high yield. After the isolation/purification, they can be separately converted into specific chemicals/fuels through appropriate conversion methods.

In the study described in chapter three, a two-step synthesis for producing methyl

12-aminododecanoate and 13-aminotridecanoate, the precursors of nylon 12 and nylon

13, from methyl oleate is described．First, methyl 11-cyano-9-undecenoate or 12-cyano-

9-dodecenoate were prepared by cross metathesis of methyl oleate with either allyl cyanide or homoallyl cyanide, respectively. Subsequently, all the unsaturation of the two

103 intermediates was hydrogenated to deliver the final products. This method represents the first synthesis of nylon 12 and 13 precursors from methyl oleate, an ester of an abundant and renewable natural fatty acid present in vegetable oil or microalgae. It also represents the shortest synthesis of nylon precursors from fatty acids.

The efficient reaction conditions for ring-closing metathesis of homoallyloleamide (2) for the synthesis of PA12 (Nylon 12) precursor (4) are presented in chapter four. Systematic screening of a series of commercially available metathesis catalysts identified a catalyst that promotes the reaction at low temperature (60 °C), enabling use of low-boiling non-halogenated solvents, ethyl acetate and hexanes. The catalyst was adsorbed on mesoporous silica gel to aid its post-reaction recovery and reuse, which was demonstrated in hexanes solvent up to three cycles without significant loss of reaction conversion (> 90%). The identification of reaction-compatible green solvents as well as the demonstration of the potential for recycling of the supported catalyst are steps further toward establishing environmentally sustainable production of the polyamide precursors from oleic acid.

Future directions:

With regards to enzymatic fractionation of microalgae biomass to recover principal components, the challenge remain the high cost of enzymes. Enzyme immobilization could reduce the enzyme loadings and cost of downstream processing.

This would also eliminate the need for isolation/separation of the enzymes from the products.

Our cross –metathesis approach for the synthesis of nylon precursors involves use of high catalyst loading, halogenated solvent (not green, requiring expensive precautions

104 for storage and use), high temperature (contributes to catalyst deactivation/decomposition) and high pressure (not ideal at least from safety and economic perspective, requiring special equipment). The high pressure was due to incompatibility of amines to metathesis. These issues could be addressed with the design of a robust catalyst and development of an alternative approach that would require green solvents, lower temperature and atmospheric pressure. Ring closing metathesis approach addressed some of the issues.

The ring closing metathesis approach, a follow-up study, works best under high dilution. There is the need to improve the batch efficiency by increasing the batch concentration.

105

References