PRODUCTION AND CHARACTERIZATION OF POLYHYDROXYALKANOATES (PHAS) FROM BURKHOLDERIA CEPACIA ATCC 17759 GROWN ON RENEWABLE FEEDSTOCKS

by Chengjun Zhu

A dissertation submitted in partial fulfillment of the requirements for the Doctor of Philosophy Degree State University of New York College of Environmental Science and Forestry Syracuse, New York August 2011

Approved: Department of Environmental and Forest Biology

James P. Nakas, Major Professor

Emanuel J. Carter, Jr., Chair Examining Committee

Donald J. Leopold, Department Chair

S. Scott Shannon, Dean The Graduate School

UMI Number: 3494520

All rights reserved INFORMATION TO ALL USERS The quality of this reproduction is dependent on the quality of the copy submitted. In the unlikely event that the author did not send a complete manuscript and there are missing pages, these will be noted. Also, if material had to be removed, a note will indicate the deletion.

UMI 3494520 Copyright 2012 by ProQuest LLC. All rights reserved. This edition of the work is protected against unauthorized copying under Title 17, United States Code.

ProQuest LLC. 789 East Eisenhower Parkway P.O. Box 1346 Ann Arbor, MI 48106 - 1346

 2011 Copyright C. J. Zhu All rights reserved

Acknowledgements
I would like to thank my major professor, Dr. James P. Nakas, for his advice, support and knowledge throughout my Ph.D study and with the manuscripts that we have submitted and will submit for publication. I also would like to thank Dr. Nomura for his guidance and discussion of our project and his kindness and great help for use and maintenance of his equipment. Likewise, I am very grateful to Dr. Arthur Stipanovic for his expert advice and assistance regarding physical-chemical property tests. I am also grateful to Dr. Patrick Mather, Ms. Erika D. Rodriguez and Mrs. Xinzhu Gu for their advice and support regarding mechanical property tests of PHAs. I wish to thank Mr. David Kiemle for assistance with NMR analysis, Mr. Daniel Nicholson and Dr. Kun Cheng for assistance with the physical characterization of PHAs. I want to thank Mr. David Sgroi, Ms. Laura Mateya, Ms. Giselle Kathryn Schlegel, Mr Matthew Michael Cleere and Mr. Sam Kogon for PHA isolation and purification from pilot-plant scale fermentations. I am grateful to Mr. Joseph K. Gredder, Ms. Anna Elyse Karczewski, etc. for assistance with the lab-scale research for PHA production. Lastly, I would like to thank my wife Qin for her support of my research and my son Felix who has brought great joy into my life. This research was supported by a grant from the New York State Energy Research and Development Authority (NYSERDA), the Welch Allyn Corp. (Skaneateles, NY), the Blue Highway LLC. (Syracuse, NY) and Tessy Plastics Corp. (Elbridge, NY).

iii

Table of Contents
List of Tables..viii List of Figures..ix Thesis Abstract..xii

1. Introduction..1 1.1 Microbial polyhydroxyalkanoates..1 1.2 Physicochemical and mechanical properties of PHAs.3 1.3 History and development of PHAs.8 1.4 Metabolic pathway for biosynthesis of PHAs from various carbon sources.13 1.5 Biodegradation and thermal degradation of PHAs.17 1.5.1 Biodegradation of PHAs.18 1.5.2 Thermal degradation of PHAs21 1.6 Considerations and rationale of renewable feedstocks and downstream processing 1.6.1 Renewable feedstocks for PHA production.23 1.6.2 Production, economics and renewability of glycerol and levulinic acid.26 1.6.3 Downstream processing for PHA isolation29

2. Materials and Methods..33 2.1 Microorganism and fermentation conditions.33 2.1.1 PHB production in shake flasks and fermentors.33 2.1.2 PHB-co-HV production in shake flasks and fermentors.34 2.1.3 P3HB -co-4HB production in shake flasks and fermentors..36 2.2 Concentration determination of glycerol, xylose, lactose and levulinic acid..37 2.3 Isolation and purification of PHAs from biomass.38 2.4 Sample preparation with nucleating agents40
iv

2.5 GC analysis for PHAs.41 2.6 Physicochemical property and mechanical property tests of PHAs42 2.6.1 Molecular mass determination42 2.6.2 Thermal analysis42 2.6.3 Tensile test.43 2.6.4 Nuclear magnetic resonance46

3. Results47 3.1 Renewable carbon sources for bacterial growth and PHA production by B. cepacia.47 3.2 Biodiesel-derived glycerol as a carbon source for PHA production.49 3.3 Production and characterization of PHB homopolymer using glycerol as a carbon source by B. cepacia.52 3.3.1 Effects of glycerol content on bacterial growth52 3.3.2 Effects of glycerol content on molecular mass of PHB.53 3.3.3 Characterization of PHB end-capped with glycerol by 1H NMR.55 3.3.4 Physical properties of PHB..56 3.3.5 Pilot scale (200-L) fermentation using biodiesel-glycerol..57 3.3.6 Injection-molding process for conversion of PHB to biodegradable eartips58 3.3.7 Thermal degradation of PHB..60 3.4 Production and characterization of PHB-co-HV copolymers using glycerol and levulinic acid as substrates in shake flasks by B cepacia..61 3.4.1 Bacterial growth and production of PHB-co-HV copolymers when co-feeding glycerol and levulinic acid in shake flasks.61 3.4.2
1

H and 13C-NMR analysis for the structure and composition of PHB-co-HV

copolymers62 3.4.3 Physical property test of PHB-co-HV copolymers..65 3.4.4 Mechanical property test of PHB-co-HV copolymers..67
v

3.5 Production and characterization of PHB-co-HV copolymers using glycerol and levulinic acid as substrates in fermentors by B cepacia72 3.5.1 Bacterial growth and PHB-co-HV copolymers when co-feeding glycerol and levulinic acid in fermentors by B. cepacia in fermentors..72 3.5.2 Crystallization temperatures of PHAs.74 3.5.3 Melting temperatures of PHAs75 3.5.4 Decomposition temperatures of PHAs..77 3.5.5 Decompositon temperature of PHAs with nucleating agents...78 3.5.6 Melting temperature and crystallization temperature of PHAs with ULTRATALC60979 3.6 Effects of different aging periods on melting temperatures of PHAs.82 3.7 Production of P3HB-co-4HB by B. cepacia using -butyrolactone/1,4-butanediol83 3.8 Confirmation of HV mol fraction in the PHB-co-HV copolymers by GC analysis and
1

H-NMR.85

3.9 Solvent extraction of PHAs..86 3.9.1 Determination of the volume of chloroform for maximum extraction efficiency 86 3.9.2 Determination of incubation temperature for maximum extraction efficiency.88 3.9.3 Determination of the incubation period for maximum extraction efficiency.89

Discussion.91

4.1 Renewable and inexpensive feedstocks for PHA production..91 4.2 Bacterial growth and Properties of PHB produced from glycerol99 4.2.1 Bacterial growth of B. cepacia using glycerol as a carbon source..99 4.2.2 Properties of PHB produced from glycerol as a carbon source.100

vi

4.3 Production and characterization of the PHB-co-HV copolymers produced from biodiesel-derived glycerol and levulinic acid103 4.3.1 Production of the PHB-co-HV copolymers using biodiesel-derived glycerol and levulinic acid103 4.3.2 Physical properties of the PHB-co-HV copolymers produced from biodieselderived glycerol and levulinic acid104 4.4 Mechanical properties of the PHB homopolymer and the PHB-co-HV copolymer107 4.5 Effects of nucleating agents on physical and mechanical properties of PHAs..111 4.6 Economic considerations of PHA production..112 4.6.1 Economic considerations of PHA production from renewable and inexpensive feedstocks.113 4.6.2 Economic considerations of PHA production by selective downstream isolation processes114 5 6 7 8 Conclusions118 References.123 Appendices134 Curriculum Vitae139

vii

List of Tables
Table 1. Comparison of physicochemical and mechanical properties of selected polyhydroxyalkanoate (PHA) polymers and petroleum-derived plastics4 Table 2. Physical and mechanical properties of PHB-co-HV copolymers.6 Table 3. The effect of concentrations and exposure times of glycerol on the numberaverage molecular weight (Mn) of PHB..54 Table 4. Physical-chemical properties of PHB produced from xylose and glycerol..56 Table 5. Comparison between the original PHB and the thermally degraded PHB..60 Table 6. Mechanical properties of PHB, PHB-co-HV copolymers and polypropylene..71 Table 7. Composition and molecular masses of PHB and copolymers of PHB-co-HV.73 Table 8. Comparison of melting temperatures (Tm) of PHAs detected at different aging times.82 Table 9. Comparison of HV mol% in the PHB-co-HV copolymers detected by GC and 1H NMR.85 Table 10. Renewable and inexpensive feedstocks used for PHA production.98 Table 11. Comparison of dry cell mass and PHA content from different bacterial strains grown on various carbon sources in shake flasks..100

viii

List of Figures
Figure 1. General structure of polyhydroxyalkanoates (a) and copolymers (b) poly-(3hydroxybutyrate-co-3-hydroxyvalerate) (abbr. PHB-co-HV) and poly-(3-hydroxybutyrateco-4-hydroxybutyrate) (abbr. P3HB-co-4HB)..1 Figure 2. Pathway of PHB biosynthesis.15

Figure 3. Proposed pathway for metabolism of various carbon sources and mediumchain-length PHA production.16 Figure 4. Figure 5. Figure 6. Thermal degradation of PHB.. 22 Production of biodiesel and glycerol. Catalysts include alkali and acids.27 Route of levulinic acid from lignocellulosic biomass28

Figure 7. Bacterial growth and PHA production using renewable carbon sources (tall oil fatty acids, biodiesel-derived glycerol and xylose) by B. cepacia in shake flask experiments48 Figure 8. Bacterial growth and PHB production from various sources of biodieselderived glycerol by B. cepacia in shake flask experiments50 Figure 9. Glycerol consumption during fermentation using pure glycerol, FutureFuel glycerol, Twin River glycerol and ESF glycerol by B. cepaica in shake flask experiments.51 Figure 10. Dry biomass of B. cepacia grown in shake flasks on different concentrations of pure glycerol...52 Figure 11. Number-average molecular weight (Mn) and weight-average molecular weight (Mw) of PHB produced by B.cepacia grown with different concentrations of glycerol..53
1 Figure 12. H-NMR of PHB produced by B .cepacia grown on 7% (v/v) glycerol as a carbon source. The expanded region indicates glycerol as the terminal end-group55

Figure 13. Changes in biomass, PHA% and glycerol concentration during a fed-batch fermentation, in a 400-L fermentor, with periodic additions () of biodiesel-glycerol57

ix

List of Figures (Continued)

Figure 14. (a) Purification and vacuum drying process of PHB, (b) Schematic flowsheet of injection-molding process, (c) Eartips made from polypropylene (black, left) and PHB (brown, right two). 59 Figure 15. Effects of levulinic acid concentration on bacterial growth and production of PHB-co-HV copolymers61 Figure 16. (a) Chemical shift expansions of 300 MHz 1H NMR spectra from PHB-co-HV copolymers, illustrating a progressive increase in the mol% of HV and quantified by the integrated areas of the HB doublet (methyl group of HB, 1.27 ppm) and the HV triplet (methyl group of HV, 0.90 ppm). (b) Chemical shift assignments of 300 MHz 13C NMR spectrum of PHB-co-35.8 mol% HV, which was produced by B. cepacia using 3% (v/v) glycerol and 0.9% (w/v) levulinic acid..63 Figure 17a. DSC curves displaying melting temperatures of PHB-co-HV copolymers66

Figure 17b. Melting temperatures and glass transition temperatures of PHB-co-HV copolymers with increasing HV mol%...............................................................................67 Figure 18a. Stress-strain curves of PHAs and polypropylene. Dot line: PHB (produced by B. cepacia using xylose as a carbon source); Dash dot line: PHB-co-17.6 mol% HV; Solid line: PHB-co-17.6 mol% HV; Dash line: polypropylene.69 Figure 18b. Real-time images of stretching for different types of PHAs in tensile testing. i represents PHB homopolymer, of which the dogbone was pulled for less than 0.7 mm to break; ii displays PHB-co-29.5 mol% HV, of which the dogbone was stretched for approximately 35 mm to break; iii represents for PHB-co-33 mol% HV, of which the dogbone was pulled for around 66 mm to break.70 Figure 19. Growth (dry cell mass, DCM, ), PHA yield () and HV content in the copolymer () produced by B. cepacia using glycerol and levulinic acid as carbon sources in a 7 L fermentor72 Figure 20. Crystallization temperatures (Tc) of PHB () and copolymers of PHB-co-HV () with increasing mol% HV..75 Figure 21. Melting temperatures (Tm) of PHB () and copolymers of PHB-co-HV () with increasing mol% HV76 Figure 22 Decomposition temperatures (Tdecomp, ) and melting temperatures (Tm, ) of PHB and copolymers of PHB-co-HV with increasing mol% HV. Bars indicate temperature differential (Tdecomp - Tm).77

List of Figures (Continued)

Figure 23. Effects of nucleating agents on decomposition temperatures of PHB (), PHB with 5% ULTRATALC609 (), and PHB with 1% HPN-68L ().79 Figure 24. Effects of 5% ULTRATALC609 on melting temperature (Tm) of PHAs containing increasing mol% HV (--PHB, --PHB-co-5.6 mol% HV, --PHB-co-11.4 mol% HV, --PHB-co-14.7 mol% HV, --PHB-co-17.9 mol% HV, --PHB-co-30.5 mol% HV, -PHB-co-32.6 mol% HV)...80 Figure 25. Crystallization temperatures of PHAs () and PHAs with 5% ULTRATALC609 (). Bars indicate temperature differentials.81 Figure 26. 1H NMR spectrum of P3HB-co-3 mol% 4HB produced from 1,4-butanediol. Peaks 2,3,4 are typical chemical shifts for P3HB and peaks 6,7,8 (highlighted by the triangles) represent protons of P4HB84 Figure 27. Relationship between PHA extraction efficiency and dosage of chloroform. Incubation of the mixture at 55 C overnight..87 Figure 28. Relationship between PHA extraction efficiency and incubation temperature. Incubation of the mixture at room temp. when stirring or 55 C while standstill.88 Figure 29 Relationship between PHA extraction efficiency and incubation period. Incubation of the mixture at room temperature when stirring.89

xi

Abstract
C. J. Zhu. Production and Characterization of Polyhydroxyalkanoates (PHAs) by Burkholderia cepacia ATCC 17759 Grown on Renewable Feedstocks. 141 Pages, 11 tables, 29 figures, 2011

This thesis describes the microbial production of polyhydroxyalkanoates (PHAs) from bench- and pilot-plant scale fermentations by B. cepacia, followed by physicalchemical and mechanical characterization of these polyesters. B. cepacia was evaluated to utilize several renewable and inexpensive feedstocks, e.g. wood hydrolysate, biodiesel-derived glycerol, cheese whey permeate and tall oil fatty acids from the paper pulping process. Glycerol was found to be the best carbon source to support bacterial growth and poly-3-hydroxybutyrate (PHB) production based on the highest dry cell mass (DCM, 5.8 g/L) and the highest PHB yield (82% of DCM). Increasing the glycerol concentration from 3% to 9% (v/v) resulted in a gradual reduction of biomass, PHB yield and molecular mass (Mn and Mw) of PHB. 1H-NMR revealed that molecular masses decreased due to the esterification of PHB with glycerol resulting in chain termination (end-capping). Supplementing levulinic acid, derived from lignocellulosic materials, with glycerol led to production of the poly-3-hydroxybutyrate-co-3-hydroxyvalerate (PHB-coHV) copolymer. Based on the concentration and timing of levulinic acid added in the medium, various mol fractions of HV were incorporated to form various PHB-co-HV copolymers. The copolymers exhibited a typical isodimorphic behavior (V-typed shape), where upon melting temperature decreased to a minimum point and then increased as mol% HV increased. Increasing mol% HV in the copolymer resulted in enhanced mechanical properties. Addition of heterologous nucleating agents improved industrial processability of PHAs by increasing crystallization temperature. However, HPN-68L was not used as a nucleating agent for the polyesters isolated in this study because it decreased the decomposition temperature of PHAs. Production of the PHB homopolymer and the PHB-co-HV copolymers were successfully scaled up for pilot-plant scale fermentations. Large quantities of PHAs were isolated, purified and used in the fabrication of eartips, which are biodegradable and environmentally friendly, through an injection-molding process. These renewable and inexpensive carbon sources and alternative downstream PHA isolation process may greatly reduce production cost of PHAs, by which the market price of PHAs becomes more competitive with that of conventional petroleum-derived plastics.
Key Words: renewable feedstocks, Burkholderia cepacia, biodiesel-derived glycerol, levulinic acid, polyhydroxyalkanoates, physical properties, mechanical properties, pilot-plant scale fermentation, downstream PHA recovery Author: Chengjun Zhu Candidate for the degree of Doctor of Philosophy, August 2011 Major professor: James P. Nakas, Ph.D. Department of Environmental and Forest Biology State University of New York College of Environmental Science and Forestry, Syracyse, New York James P. Nakas, Ph.D. xii

1. Introduction 1.1 Microbial polyhydroxyalkanoates

Polyhydroxyalkanoates (PHAs) are a class of polyesters which are accumulated as

microbial intracellular carbon and energy reserves. They exist as discrete inclusions in the cell cytoplasm, typically 0.2-0.5 m in diameter 1. A variety of PHAs is commonly classified into two major groups which are referred to as short-chain-length (scl, 3-5 carbons) and medium-chain-length (mcl, 6-14 carbons) 2. The scl-PHAs are semicrystalline thermoplastics, whereas mcl-PHAs are more elastomeric 3. A large number of microorganisms, encompassing Gram-positive 4, 5 and Gram-negative 2, 5-7 bacteria, have been reported to produce various types of PHAs.

b
O O

R O
m=1,

O CH2 C

CH

n
8 6 7

O
4

m
O

R=H R=CH3 R=CH2CH3 R=(CH2)xCH3, x=2,3,,11 R=H R=H

poly-3-hydroxypropionate poly-3-hydroxybutyrate poly-3-hydroxyvalerate mcl-poly-3-hydroxyalkanoates poly-4-hydroxybutyrate poly-5-hydroxyvalerate

m=2, m=3,

Figure 1. General structure of polyhydroxyalkanoates (a) and copolymers (b) poly-(3hydroxybutyrate-co-3-hydroxyvalerate) (abbr. PHB-co-HV) and poly-(3-hydroxybutyrateco-4-hydroxybutyrate) (abbr. P3HB-co-4HB)

Poly-3-hydroxyalkanoates are the most widely studied PHAs, which differ in the length of side chain (R group depicted in Figure 1a) and can also differ in composition (e.g. copolymers depicted in Figure 1b, terpolymers, etc.). However, poly-4hydroxyalkanoates
8, 9

and poly-5-hydroxyalkanoates

10

have also been observed in the

past several decades (Figure 1a). Over 150 different monomer units have been identified as constituents of these storage molecules
5, 11, 12

. The large variety of PHA

components results in an enormous diversity of material properties, which is beneficial for various potential applications. PHAs have recently received increased attention due to their physical and mechanical properties resembling those of petroleum-derived plastics. These characteristics make PHAs ideal substitutes for petroleum-based plastics, especially when global oil prices remain at relatively high levels. Also, PHAs are completely degraded in the environment to CO2 and H2O 13, 14, compared to recalcitrant, non-biodegradable conventional plastics, such as polypropylene, polyethylene and polystyrene. Since PHAs are generally biosynthesized using photosynthetically-based renewable carbon feedstocks and their end-use materials are biodegradable to carbon dioxide and water by microbial extracellular PHA depolymerases
15, 16

, production and

disposal of these PHA biopolymers constitute a sustainable, closed life cycle process with much less energy consumption and greenhouse gas emission 17-19. PHA polymers are synthesized as membrane-bound storage materials by a variety of microorganisms when exogenous carbon sources are provided in excess and their growth is impaired by the lack of at least one other nutrient, such as nitrogen, sulfate, phosphate, magnesium and oxygen
6, 20

. Therefore, microbial fermentation for PHA

2

production is regulated largely by the concentrations and contents of these nutrients in the medium. Optimal feeding strategies of these nutrients dramatically enhance microbial growth and PHA production yields. As reported by Steinbuchel et al. 21, PHAs are accumulated to levels as high as 90% (w/w) of the dry cell mass. Furthermore, as a PHA-producing microbe enters the starvation stage due to a carbon deficiency, PHA polymers will function as intracellular carbon and energy reserves and initiate the degradation process of PHAs to provide carbon and energy for microbial survival in a harsh environment
22

. Therefore, PHAs, under natural rules of

survival, have been designed and developed by a large number of microorganisms as a food and energy source during times of nutritional stress.

1.2 Physicochemical and mechanical properties of PHAs

PHAs are highly recommended as an ideal substitute for petroleum-derived

plastics in large part due to their similar material properties (Table 1), in terms of processability, strength and industrial fabrication into commodity plastic products 23, 24. Physicochemical and mechanical properties of PHA copolymers can be controlled and regulated by variation of the mole fractions of the monomeric constituents (Figure 1b), which could be achieved by careful selection and control of fermentation carbon sources.

Poly-3-hydroxybutyrate (PHB) is the most common type of PHA produced by bacteria. The PHB homopolymer is a highly crystalline, stiff, yet relatively brittle, material. Correspondingly, as shown in Table 1, PHB exhibits high tensile strength (43 MPa), but low elongation to break (5%) 25. The copolymer poly-3-hydroxybutyrate-co-20 mol%-3-hydroxyvalerate (PHB-co-20 mol% HV) exhibits lower crystallinity, less stiffness (20 MPa), but higher elasticity and flexibility (50% elongation to break)
25

compared to

the homopolymer PHB. Incorporation of different monomeric subunits, such as 4hydroxybutyrate (4HB)
25

3-hydroxyhexanoate

(HHx)

25

and

other

mcl-

hydroxyalkanoates26 (e.g. 3-hydroxyoctanoate [3HO], 3-hydroxydecanoate [3HD], 3hydroxydodecanoate [3HDD]), with 3HB will result in copolymers with varying material

Table 1 Comparison of physicochemical and mechanical properties of selected polyhydroxyalkanoate (PHA) polymers and petroleum-derived plastics 25 Tensile strength (MPa) 43 20 26 21 38 10 Elongation to break (%) 5 50 444 400 400 620

Polymers PHB PHB-co-20%HV PHB-co-16%4HB PHB-co-10%HHx Polypropylene Polyethylene (LDPE)d

Crystallinitya (%) 60 56 45 34 50-70 20-50

Tm b (C) 177 145 150 127 176 130

Tg c (C) 4 -1 -7 -1 -10 -36

c

Notes: a b degree of crystallinity melting temperature d LDPE: low density polyethylene 4

glass transition temperature

properties to address numerous applications in medical and engineering fields. Compared to commercially available polylactic acid (PLA) which is also renewable and biodegradable27, diverse combinations of PHA monomeric units are superior to PLA, which only consists of a single monomer-lactic acid. In addition to regulating material properties by controlling the composition of PHAs, it is possible to change material properties of PHAs with the same composition by incorporating various fractions of the co-monomer in the copolymers. This thesis mainly describes the production and characterization of poly-3-hydroxybutyrate-co-3hydroxyvalerate (PHB-co-HV) copolymers, biosynthesized from glycerol and levulinic acid as carbon sources. As the mol fraction of HV in the copolymer PHB-co-HV varies, the physical and mechanical properties of PHB-co-HV will correspondingly change (Table 2). When HV ratios of PHB-co-HV increased from zero (the homopolymer of PHB) to 25%, the melting temperatures (Tm) and glass transition temperatures (Tg) gradually decreased from 179 C to 137 C and 10 C to -6 C, respectively. As shown in Table 2, PHB-co-HV copolymers also became more flexible (as indicated by the decrease of Youngs modulus) and tougher (as demonstrated by the increase in impact strength) as the HV ratio increased. Besides PHB-co-HV, similar patterns were observed with copolymers PHB-coHHx28-30, P3HB-co-4HB20, PHB-co-HO31, PHB-co-HD31, when the ratios of these comonomers, such as HHx, 4HB, HO or HD, in their copolymers increased.

Table 2 Physical and mechanical properties of PHB-co-HV copolymers20. Mol fraction (mol%) 3HB 3HV 100 97 91 86 80 75 0 3 9 14 20 25 Youngs Modulus (GPa) 3.5 2.9 1.9 1.5 1.2 0.7 Tensile Strength (MPa) 40 38 37 35 32 30 Notched Izod Impact Strength (J/m) 50 60 95 120 200 400

Tm (C) 179 170 162 150 145 137

Tg (C) 10 8 6 4 -1 -6

PHB-co-HV copolymers with various HV mol fractions exhibit high degrees of crystallinity (between 55% and 70%) 32. PHB-co-HV is unique among the PHA family of copolymers in that the size and structure of HB and HV monomers are similar. Their similarity allows HB and HV to participate in a co-crystallization process, in which HV can be incorporated into the HB crystal lattice and vice versa. This phenomenon is termed isodimorphism32,
33

. As a result, the melting temperatures of PHB-co-HV gradually

decrease to a minimum point, then increase as the HV mol fraction increases. Therefore, isodimorphism and the transition between the HB crystal lattice and the HV crystal lattice typically demonstrate a V-shaped pattern (see melting temperatures of PHB-coHV in results and references
32, 33

). Incorporation of comonomers with PHB to form

6

copolymers can lower melting temperatures, by which PHA copolymers demonstrate an important advantage of industrial applications for melt processing at lower temperatures. However, the PHA family displays a rather slow crystallization process due to high purity and limited heterogeneous nuclei34. Basically, isolation of PHAs from cells using solvents excludes most of the impurities from cell components. During the meltquenching process, PHAs at high purity crystallize at a relatively low rate, which causes a longer manufacturing cycle and a less efficient industrial fabrication process for finished products. PHB and PHB-co-HV have been extensively studied for their nucleation behaviors34, 35. Pure PHB exhibited a very slow nucleation, though self seeding to some extent, when it was cooled from a melt34. The nucleation density of pure PHB is too low to initiate efficient crystallization. At the same time, limited nuclei formed limited quantities of spherulites so that the size of each PHB spherulite is relatively large, which makes PHB somewhat brittle and subject to cracking36, 37. PHB-co-HV also exhibited a slow crystallization behavior, which resulted in the films made from the copolymers with a higher HV ratio tacky and even sticky to themselves after cooling for an extended period of time35, 38. Impurities usually behave as external nuclei for PHAs. They can increase not only the rate of crystallization, but also the quantities of spherulites corresponding to the numbers of nuclei. The large quantities of nuclei make the size of spherulites relatively small, thus improving material mechanical properties36. Accounting for the slow crystallization process of PHB and PHB-co-HV, heterogeneous nucleating agents instead
7

of impurities, which are difficult for the quantification and qualification, need to be supplemented with these polymer melts to speed up the crystallization process by increasing the crystallization temperature and forming small spherulites. Various nucleating agents, such as orotic acid39, -cyclodextrin40, boron nitride, talc, terbium oxide, lanthanum oxide35, saccharin and phthalimide41 have been tested for enhancing the crystallization of PHAs. These nucleating agents will generally increase crystallization temperatures, accelerate crystallization rates and enhance PHA stability during heating. However, some nucleating agents have negative effects on PHAs by decreasing their decomposition temperatures (Tdecomp.). Hydroxyapatite, which is a naturally occurring mineral of calcium apatite, decreased the onset Tdecomp from 260 oC to 225 oC when the hydroxyapatite content in the hybrid PHB-hydroxyapatite composite increased from 0 to 10%42. These side effects of nucleating agents have not been described in many previous studies. Comprehensive physical property tests, including crystallization temperature, melting temperature, glass transition temperature and decomposition temperature, have been performed in this research (see results).

1.3 History and development of PHAs

was first identified by Lemoigne in 1926
43

Poly-3-hydroxybutyrate, as an intracellular storage material in Bacillus megaterium, . At that time, PHB was discovered

unexpectedly when Lemoigne attempted to determine the cause of acidification in an aqueous suspension of the bacterium Bacillus megaterium under oxygen-free conditions.
8

Although methods for PHB detection and quantification were limited, Lemoigne and coworkers published 27 papers from 1923 to 1951. Not until the late 1950s did microbial physiologists finally recognize the importance of PHB in the overall metabolism of bacterial cells44. The rediscovery of PHB occurred simultaneously and was reported independently in 1958 and 1959. Williamson et al. 45 at the University of Edinburgh in Scotland treated the cells of various Bacillus spp. with an alkaline solution of sodium hypochlorite, and found large amounts of PHB (89%) with the ether-soluble lipid (11%) in some Bacillus species, and elucidated the function of PHB in the cell. Doudoroff and Stanier 46 at the University of California at Berkeley discovered that PHB was the primary product of the oxidative and photosynthetic assimilation of organic compounds by Pseudomonas saccharophila and a phototropic bacterium, Rhodospirillum rubrum, respectively. When external carbon sources were removed from the medium, there occurred a fairly rapid intracellular breakdown of the storage PHB. Therefore, PHB was putatively thought to play a role analogous to that of starch and glycogen in the metabolism of other organisms. Doudoroff et al.46 attempted to clarify the biosynthesis and degradation mechanism of PHB in microbial cells. In the 1960s, these aforementioned authors
47, 48

isolated native PHB granules from a chemoheterotroph and phototroph, Bacillus megaterium and Rhodospirillum rubrum. The native PHB granules from R. rubrum retained the active PHB synthase and depolymerase that degrade PHB to the monomer, and the isolated granules from B. megaterium only exhibited the synthase associated with PHB. These authors also determined that PHB granules in bacteria actually serve as
9

an intracellular energy and carbon reserve and that PHB is produced in response to a nutrient limitation when at least one essential element in the environment is exhausted. In 1965, Marchessault and his co-workers49 at SUNY-ESF collected PHB samples from various microorganisms and employed X-ray diffraction to explore the crystal structure of PHB. All PHB samples precipitated from chloroform displayed an uniformity of crystal structure and molecular structure due to identical X-ray diffractograms and infrared spectra, respectively. The molecular mass of PHB isolated by alkaline hypochlorite was also found to be uniformly low compared to that of PHB prepared by direct solvent extraction. In 1974, Wallen and Rohwedder50 reported that the polyesters, isolated from activated sewage sludge, exhibited similar but not identical NMR and infrared spectra. Gas chromatographic analysis indicated a mixture of C4, C5, C6 and C7 components in these polyesters. The identification of a variety of PHAs in addition to PHB opened avenues of research for PHAs regarding their material properties. Since rising oil prices and an unstable oil supply became global issues and environmental concerns over limited fossil reserves drew more and more attention, the advantage of PHAs as biodegradable alternatives to petroleum-derived plastics was gradually recognized by researchers and much effort has been devoted to investigation of PHA properties. In addition, much time and effort has been dedicated to genetic engineering of microorganisms or plants for biosynthesis of PHAs from renewable feedstocks with improved material properties for different end-use applications.
10

Prompted by the oil crisis in the 1970s, the first industrial production of PHB was introduced in 1982 by Imperial Chemical Industries (ICI) in the UK, employing a twostage, fed-batch fermentation process and using a sugar-based carbon source for growth of Ralstonia eutropha. However, PHB was produced at a relatively high production cost due to expensive feedstocks and downstream processing for PHB isolation. Compared to conventional petroleum-derived plastics, higher cost has placed PHAs in a weaker economic position and mechanical properties, such as high crystallinity and brittleness, have resulted in a rather limited range of applications20. Based on these limitations of the homopolymer PHB, PHB-co-HV copolymer, which exhibits enhanced toughness and flexibility, attracted more attention and was first produced on a commercial scale in the late 1980s by ICI, which marketed their PHB-coHV products under the trade name of Biopol6, 51. Biopol (PHB-co-HV copolymer) was produced by R. eutropha, using glucose and propionate as carbon sources, at a reported production cost of US$ 16/kg, which was 18-fold higher than conventional polypropylene52. The prohibitively high price of PHAs and the relatively low prices of commodity plastics made the commercial-scale production of PHAs unrealistic. Eventually Biopol was acquired by Zeneca Ltd in 1990 until it was acquired in 1996 by Monsanto (St. Louis, MO), which utilized their considerable expertise in plants and initiated research to produce PHB and related copolymers photosynthetically in plants51. Subsequently, Monsantos right to Biopol was sold to Metabolix (Cambridge, MA) in 200153. Recombinant Escherichia coli strains, which exhibit broad nutritional diversity

11

and relatively fast growth, have been engineered to produce PHAs from inexpensive carbon sources. Mirel (trademark of Telles, an affiliated company of Metabolix and Archer Daniels Midland Company) bioplastics were announced to be produced using corn starch sugars, starting in 2009 in Clinton, IA (www.mirelplastics.com). Several other companies are also developing and commercializing various types of PHAs, including Procter & Gamble (Cincinnati, USA) which introduced novel PHA copolymers with HB as a monomeric unit and HHx (C6), HO (C8) or HD (C10) etc. as monomeric units, all of which are under the trademark of Nodax 54, 55. The projections for Nodax marketing consist of PHA production of 100-1000 metric ton in 2004 and the delivery plan of approximate $ 1/lb ($ 2.2/kg) in 2005/200656. Since the 1980s, BASF in Germany developed a pilotscale fermentation process for production of PHB and PHB-co-HV, which were supplemented and blended with its biodegradable polymer Ecoflex, which are petroleum-derived aliphatic-aromatic copolyesters57. Tianan Biologic in China started to produce PHB-co-HV in 2000 using glucose, fructose and organic acids as carbon sources (www.tianan-enmat.com), and have currently scaled up to commercial production of 2000 metric tons per year57. Most of the above-mentioned commercial PHAs have been produced by various microorganisms using pure carbon sources, such as glucose, sucrose, propionate and lauric acid, which are always more expensive than renewable feedstocks, including wood-based hydrolysate (xylose and levulinic acid)58, cheese whey permeate (including lactose)59, sugarcane molasses (sucrose, fructose, glucose and trace amount of maltose)60, 61, corn steep liquor (nitrogen sources)60, soybean oil62, and crude glycerol
12

from

biodiesel-producing

facilities63.

Regarding

environmental

concerns

of

commercializing PHAs, the production of PHAs from renewable and biobased feedstocks, e.g. wood hydrolysate and biodiesel-derived glycerol, need to be taken into account at a higher priority.

1.4 Metabolic pathway leading to the biosynthesis of PHAs from various carbon sources

Uptake of carbon sources by microorganisms supports their growth and other

basic vital functions. Carbon sources in the environment have to be translocated into the cytosol inside the cell, the location for nutrient metabolism. In this translocation process, transporters, which are transmembrane proteins, either actively transport the molecules of carbon sources by utilizing ATP (adenosine triphosphate) energy (e.g ATPbinding cassette transporters, ABC-transporters) or passively translocate these molecules without energy expenditure (e.g glycerol diffusion facilitator protein, GlpF). Therefore, whether a certain carbon source can be utilized for microbial growth is first determined by the transporters which can recognize and transport this molecule. Relevant to this research, xylose (a five carbon sugar) and/or glycerol were used as carbon sources by Burkholderia cepacia. D-xylose is transported by XylFGH transporter (possibly belonging to the ABC superfamily) into the cytosol and converted to D-xylulose by xylose isomerase (encoded by xylA gene), subsequently phosphorylated to Dxylulose-5-phosphate by xylulokinase (encode by xylB gene)64, 65 (Figure 3). Glycerol is

13

translocated by GlpF into the cell and phosphorylated by glycerol kinase (encoded by glpK), leading to glycerol-3-phosphate. Glycerol-3-phosphate dehydrogenase (encoded by glpD) oxidizes glycerol-3-phosphate to dihydroxyacetone phosphate (DHAP)66 (Figure 3). Further metabolism of these intermediates (e.g. D-xylulose-5-phoaphate and DHAP) derived from the carbon sources through either glycolysis (Embden-Meyerhof pathway) or the pentose phosphate pathway or Entner-Doudoroff pathway results in the end product pyruvate, which is converted through oxidative decarboxylation by pyruvate dehydrogenase to acetyl-CoA67. In this research, levulinic acid (4-ketovaleric acid) was employed as the cosubstrate to provide the precursor propionyl-CoA for biosynthesis of poly-3hydroxyvalerate. Metabolism of levulinic acid is not clearly elucidated, but tentatively considered to form one propionyl-CoA and one acetyl-CoA through -oxidation. The enzymes involved are still ill-defined21. Three key enzymes -ketothiolase, acetoacetyl-CoA reductase and PHA synthase, encoded by phaA, phaB and phaC, respectively, are involved in the last three steps of PHA biosynthesis. Two acetyl-CoA moieties are condensed to acetoacetyl-CoA by ketothiolase, and acetoacetyl-CoA is reduced to (R)-3-hydroxybutyryl-CoA by acetoacetyl-CoA reductase, followed by polymerizing the precursor 3-hydroxybutyrylCoA to the polymer poly-3-hydroxybutyrate68. Similarly for poly-3-hydroxyvalerate synthesis, one acetyl-CoA and one propionyl-CoA are condensed to 3-ketovaleryl-CoA,

14

with subsequent reduction to 3-hydroxyvaleryl-CoA, followed by polymerization for PHV synthesis (Figure 2).
O SCoA O O SCoA

2 acetyl-CoA

3-ketothiolase (PhaA)
CoASH

acetoacetyl-CoA
NADPH

acetoacetyl-CoA reductase (PhaB)

OH

O SCoA

NADP+

(R)-3-hydroxybutyryl-CoA PHA synthase (PhaC)

O O

CoASH

PHB
n

Figure 2 Pathway of PHB biosynthesis Biosynthesis of medium-chain-length PHAs, which are comprised of the constituents C6-C14 chains, recruits PHA-specific enzymes such as (R)-specific enoyl-CoA hydratase (PhaJ), putatively (R)-3-hydroxyacyl ACP thioesterase (PhaG) and acyl-CoA ligase (AlkK) (Personal communication with Qin Wang and Christopher Nomura at SUNYESF) to divert the intermediates (such as enoyl-CoA and (R)-3-hydroxyacyl acyl carrier protein (ACP), respectively) of fatty acid -oxidation pathway and fatty acid biosynthesis pathway for the formation of precursor (R)-3-hydroxyacyl-CoA (Figure 3), which leads to biosynthesis of medium-chain-length PHAs by PHA synthase (PhaC)69.
15

16

Figure 3 Proposed pathway for metabolism of various carbon sources and mediumchain-length PHA production69. GAP: glyceraldehyde-3-phospahte. DHAP: dihydroxyacetone phosphate. (1) or (2) represents one or more steps in glycolysis pathway or pentose phosphate pathway, respectively. : cell membrane

While the pathway for PHA biosynthesis has been gradually elucidated in the last several decades, construction of desired strains for PHA production seems to be a promising path and has greatly stimulated and enhanced PHA research. These engineered strains, such as E. coli70, Pseudomonas putida71 and Cupriavidus necator62 (formerly known as Ralstonia eutropha or Alcaligenes eutrophus) demonstrate the advantages of broad nutritional diversity (using various carbon sources, including inexpensive feedstocks), relatively rapid growth to high cell density, accumulation of high intracellular concentrations of PHAs and biosynthesis of novel PHAs which cannot be accomplished in native strains and which exhibit diverse material properties for various end-use applications.

1.5 Biodegradation and thermal degradation of PHAs

Although this thesis focuses on the production and characterization of PHAs using

inexpensive carbon sources, some consideration must be given to the biodegradability of PHAs, which is a distinct advantage of PHAs over conventional plastics. Also, since some applications of PHAs need to be conducted at high temperature, it is reasonable to examine the thermal degradation of PHAs.
17

1.5.1 Biodegradation of PHAs

Petroleum-derived plastics are xenobiotic compounds, which are recalcitrant to

degradation and take several decades, even over 100 years, to degrade in the natural environment72. One of the most important characteristics of PHAs as substitutes for conventional plastics is that PHAs are biodegradable. In nature, a vast consortium of microorganisms, by using intracellular or extracellular PHA depolymerases, will degrade PHAs, which either are stored inside of the cells or exist in the natural environment. It is noteworthy that intracellular PHA depolymerases cannot hydrolyze extracellular PHAs and extracellular PHA depolymerases are also unable to degrade intracellular PHA granules16. Apparently, these differences result from the biophysical structures of intracellular native PHA granules and extracellular denatured PHAs. The former are completely amorphous elastomers73, however, the latter are known for their high crystallinities2. Native PHA granules with a particular surface layer containing protein and phospholipids are denatured by losing this surface layer during the isolation processes of PHAs, leading to semicrystalline polymers that exhibit an ordered helical crystal structure, in which the remaining amorphous polymers are embedded13. In the remainder of this thesis, biodegradability of PHAs, if mentioned, will be concerned with the extracellular denatured PHAs using extracellular depolymerases in order to demonstrate the superiority of PHAs over petroleum-based plastics. Volova et al74 studied the biodegradability patterns of PHB and PHB-co-11 mol%HV in a tropical marine environment. After 160 days, the loss of mass of both PHA films,
18

submerged to a depth of 120 cm, approached 50% of the original dry weight of the PHA films. Due to a larger surface area, PHA films demonstrated a more rapid rate of biodegradation than compacted PHA pellets. The polydispersity increasing in all PHA samples suggested that the fragments of polymers with more diverse lengths were growing due to random scission of the hydrolyzed polymer chains. Under their study conditions, no significant differences were observed for the degradation rates between PHB and PHB-co-11 mol%HV, and the degree of crystallinity of both PHAs remained unchanged. Mergaert et al.75 tested microbial degradation of PHB and PHB-co-10 mol% HV in soils at 15 C, 28 C or 44 C for up to 200 days. These dog bone-shaped PHA samples were degraded at an erosion rate of 0.03% to 0.64% weight loss per day, depending on the polymer compositions, the types of soil and the incubation temperatures. In summary, the degradation was enhanced by incubation at higher temperatures, and in most cases the copolymer exhibited a higher erosion rate, based on weight loss, than the homopolymer. The degradation also resulted in the loss of mechanical properties (based on less elongation to break). In the studied soil (sandy soil, pH 6.5; clay soil, pH 7.1, loamy soil, pH 6.3; hardwood forest soil, pH 3.9; pinewood forest soil, pH 3.5), 295 dominant microbial strains capable of degrading PHB and PHB-co-HV in vitro were isolated and identified. This same research team76 investigated PHB, PHB-co-10 mol% HV and PHB-co-20 mol% HV in situ in natural waters. These polymers were degraded rather slowly (less than 7% weight loss after 6 months) in two freshwater ponds. However, after 358 days in a freshwater canal, 34% weight loss was recorded for PHB,
19

77% for PHB-co-10 mol% HV, and 100% for PHB-co-20 mol% HV. In seawater, within 270 days, PHB lost 31% of the initial weight and the copolymers lost between 49-52%. The degradation rate was observed to be more rapid during the summer due to higher temperature. Mergaert et al.77 also studied biodegradation of PHB, PHB-co-10 mol% HV and PHB-co-20 mol% HV in compost. PHB-co-20 mol% HV was degraded much faster (70% weight loss) than PHB (6% weight loss) and PHB-co-10 mol% HV (4% weight loss) within 150 days. From this biodegradation study, as well as two other composts, 109 microbial strains capable of degrading PHAs in vitro were isolated and identified. Although hundreds of microorganisms are able to secrete extracellular PHA depolymerases to degrade PHAs in the natural environment, the biodegradation rates of PHAs varied dramatically from several months up to several years as a function of PHA composition and shapes of the test samples, and environmental factors, such as temperature, types of soil and water, sunlight (UV radiation)78, pH16, etc.. In a word, the biodegradability of PHAs in the natural environment makes PHAtype thermoplastics ideal green substitutes for conventional plastics, which are resistant to degradation in the environment and therefore result in severe environmental pollution on a global scale.

20

1.5.2 Thermal degradation of PHAs

Considering the injection molding process for industrial fabrication of finished

products (packaging, eartips, thermometer covers, medical implants, etc.), thermal degradation of PHAs must be taken into account during the heating process due to their potential thermal instability79. PHB was extensively studied during melt processing and found to be rather unstable at temperatures above or even close to its melting temperature (177 C). Doi and his co-workers80 confirmed that PHB polyester suffered increasing reduction of molecular mass within 20 min at 175 C, even 2 C below its melting temperature. In addition, all copolyester samples (PHB-co-HV, HV=0-71 mol%; P3HB-co-4HB, 4HB=0-82 mol%) tested were thermally unstable at temperatures above 170 C, based on rapid loss of molecular masses. Below or equal to 160 C, all copolymer samples tested demonstrated thermal stability and their molecular masses decreased at a very slow rate within 20 min. Especially for the first 2 min of heating at 180 C , most of PHA polymers did not show significant decrease in molecular masses. Therefore, PHAs are suggested to be heated to melt and kept at that temperature for a short time period (2 min suggested in the study by Kunioka et al.80). It was reported that thermal degradation of PHB occurred by random chain scission, a widely accepted six-membered ring ester decomposition process, yielding crotonic acid (one product from PHB pyrolysis, Figure 4), 2-pentenoic acid (one product from PHV pyrolysis) and other related oligomers81-83.
21

CH3 R1

C H

H2 C

C O

CH3 CH
CH3

CH

C O

CH H CH

C O

R2

pyrolysis
CH3 R1

C H

H2 C

O C OH

CH3

CH3 O C OH

HC HC

CH HC C O R2

Crotonic acid
Figure 4 Thermal degradation of PHB

A major concern of PHB and other related PHAs for use as a thermoplastic is thermal sensitivity during melt processing82. A potential solution is grafting of certain compounds with PHAs. It has been reported that none of the conventional polyolefin stabilizers enhanced PHB stability84. However, radiation grafting of methyl methacrylate (MMA)85, 2-hydroxyethyl methacrylate (HEMA)85, acrylic acid (AAc)86 and styrene87, 88 onto PHB and its copolymers was found to enhance the thermal stability of PHAs. Also, grafting maleic anhydride onto PHB remarkably enhanced its thermal decomposition temperature79. In most cases, a low degree of grafting not only enhanced the thermal stability of PHAs, but also promoted the biodegradability of PHAs due to the wettability

22

between the polymer and the enzyme solutions. It was also noted that a high degree of grafting steeply decreased the biodegradability of PHAs. Therefore, it is necessary to strictly control the temperature to melt PHAs during heating. Because of the loss of molecular mass of PHAs, mechanical properties of PHAs definitely change after heating. In addition, a low degree of grafting helps stabilize PHAs when the processing temperature reaches the melting temperature.

1.6 Considerations and rationale of renewable feedstocks and downstream processing

Compared to the price of conventional petrochemical plastics (less than US $ 1/kg),

the price of PHAs was prohibitively high (ca. US $ 16/kg Biopol PHAs)89, which places PHAs at a distinct disadvantage in the market place. By examining the entire flowsheet of PHA production, the high price of PHAs is driven by high production costs of PHAs in two major areas. One is the cost of feedstocks, and secondly, the downstream recovery process of PHAs.

1.6.1 Renewable feedstocks for PHA production

A large number of microorganisms have been shown to produce various types of

PHAs. Although it is feasible to use pure carbon sources, such as glucose, galactose, xylose, fatty acids, etc., for PHA production in the laboratory, high production costs for carbon feedstocks hinder the scale-up of PHA production towards the commercial level.

23

Consequently, the cost of carbon sources accounts for as much as 50% of the overall production costs89, 90. In order to reduce the production costs of PHAs, renewable and inexpensive feedstocks have to be taken into consideration to substitute for pure and expensive carbon sources at the commercial level. As reviewed in 2009 by Chen57, the microbial production of PHAs at the industrial scale still uses some high purity sugars and fatty acids, such as glucose, sucrose, lauric acid, and propionate, during the fermentation process. Although the price of PHAs is expected to be approximately US $3-4/kg in the near future, it is obvious that PHAs are still more expensive than conventional plastics, especially when the price of oil is depressed. In order to further reduce the production costs of PHAs, some progress has been achieved in the last several years by using inexpensive agricultural or industrial wastes for PHA production. Khardenavis et al. 91 evaluated waste activated sludge generated from a combined dairy and food processing industry wastewater treatment plant for PHB production. Jowar or rice grain-based distillery spent wash was used as a carbon source for a PHB yield of 42.3% (w/w) and 40% (w/w), respectively. Addition of di-ammonium hydrogen phosphate further increased PHB production to 67% (w/w). Meanwhile, mixed culture production of PHAs from waste water was evaluated to be financially attractive in comparison to pure culture production of PHAs92.

24

Gouda et al.60 used sugarcane molasses and corn steep liquor, two of the most inexpensive substrates available in Egypt, as sole carbon and nitrogen sources for PHB production. The best growth of Bacillus megaterium was obtained with 3% molasses, while the maximum yield of PHB at 46.2% (w/w) occurred with 2% molasses. Corn steep liquor with a concentration equivalent to 0.05% NH4Cl was the best nitrogen source for PHB synthesis (32.7%, w/w). Yellore et al. 59 isolated a strain of Methylobacterium sp. ZP24 from a local pond, and tested this strain by using lactose from cheese whey, a byproduct of the dairy industry, for PHB production. Pure lactose at a concentration of 12 g/L resulted in biomass and PHB yield of 5.25 g/L and 59% (w/w) in 40 h, respectively. Cheese whey, used as a carbon source, led to 1.1 g/L PHB and further addition of ammonium sulphate increased PHB production from whey 2.5-fold. Nath et al.
93

employed a fed batch

process to attain a PHB yield of 4.58-fold using limiting dissolved oxygen in the fermentor with processed cheese whey supplemented with ammonium sulfate. Keenan et al.58 used aspen and maple wood hydrolysate, which contains xylose and glucose as major sugar components and low amounts of galactose, arabinose and mannose, supplemented with levulinic acid for PHB-co-HV production. These detoxified hydrolysates amended with 0.25%-0.5% levulinic acid were used as feedstocks, and resulted in PHB-co-HV yields, PHA% of dry cell mass and mol fraction of HV at 2.0 g/L, 40% (w/w), and 16 mol% - 52 mol%, respectively. From an economic standpoint, the substrate cost of hemicellulosic hydrolysate was reduced to US $ 0.34/kg PHB,

25

compared to US $ 0.58/kg PHB from hydrolyzed corn starch and US $ 1.30/kg PHB from pure glucose.

1.6.2 Production, economics and renewability of glycerol and levulinic acid

In this research, waste glycerol, a byproduct from the biodiesel-producing process

(Figure 5), was tested as a carbon source for PHA production. Biodiesel-derived glycerol has been produced in huge quantity as the production of biodiesel has significantly increased. Biodiesel production increased dramatically from 500,000 gallons in 1999 to 450 million gallons in 2007 (National Biodiesel Board, 2008). The major byproduct of the biodiesel industry is glycerol, which is product of approximately 10% of the final weight of biodiesel94. Consequently, in 2007, glycerol was produced in a quantity of 45 million gallons in the United States, and this crude glycerol is not suitable for use in the food, pharmaceutical, cosmetics and other industries due to low purity. It is expensive to refine crude glycerol to the purity needed for these applications95. Biodiesel production, as well as glycerol, are at an all-time high. Since the market is glutted, the price of glycerol has decreased and raw glycerol from biodiesel-producing companies is now at US $ 2.5 cents/lb96. Therefore, conversion of crude glycerol into higher-value products, such as PHAs, improves the economic viability of the biofuel industry by producing a value-added product as well as eliminating the cost of treatment for glycerol disposal.

26

O O O O O O R3 R1 O

R2

Alkali, Acids Methanol

OH

R1 O R2 O R3

OH

OH

Triacylglycerol

Glycerol

Biodiesel (Fatty acid methyl ester)

Figure 5 Production of biodiesel and glycerol. Catalysts include alkali and acids.

Besides glycerol, levulinic acid was employed as a cosubstrate for PHB-co-HV copolymer production by Burkholderia cepacia in this research. Levulinic acid (4ketovaleric acid, 4-oxopentanoic acid), an important biomass-derived feedstock, could be produced cost-effectively from a wide-array of renewable, hexose-containing materials, including forest-based lignocellulosic biomass 97, 98. Hexoses, such as glucose, fructose and mannose, from forest residues are converted under acid dehydration to 5hydroxymethylfurfural (HMF) and subsequently hydrated resulting in the final products of levulinic acid and formic acid (Figure 6).

27

Cellulose, Hemicellulose

OH HO HO O OH OH

H2O

O HO O

H2O

O OH O

OH

Glucose

5-hydroxymethylfurfural, HMF

levulinic acid

formic acid

Figure 6 Route of levulinic acid from lignocellulosic biomass

Levulinic acid, in addition to glycerol, is also among the top 12 value-added chemicals from biomass described by the US DOE99. It is especially attractive because a variety of lignocellulosic biomass, such as rice straw, wood, pulp slurry, corn starch, switch grass, sugarcane, etc., can be used for the direct production of this thermodynamically stable molecule100. The dehydration of C6 sugars by acids was reported to generate levulinic acid with a theoretical maximum yield of 64.5% (w/w) because of the concurrent production of equimolar amounts of formic acid101. Several technologies have been developed for the industrial-scale continuous production of levulinic acid, with yields reported between 20% and 48% (w/w)100. The most promising technology proposed for the large-scale production of levulinic acid was patented by the Biofine Corporation (South Glens Falls, NY)102, 103. This approach used a double-reactor system and minimized the formation of byproducts. The carbohydrate feedstocks and sulfuric acid catalyst (1-5 wt% of the feedstocks) were mixed at 210230 C for a short time of 13 to 25 s in the first reactor, in which C5 and C6 sugars are

28

produced. Meanwhile, dehydration of C6 sugars in the first reactor produces HMF. Subsequently, HMF is continuously removed from the first reactor and fed into the second reactor at 195-215 C for 15 to 30 min for a final conversion to levulinic acid at a yield of 50%-70% (w/w). As a result, levulinic acid can be produced at a low cost (US $ 0.04-0.1/kg)104, which depends on the scale of production and the economic climate of the cellulosic feedstocks. Besides levulinic aicd which can be obtained from lignocellulosic biomass, tall oil fatty acids, mainly consisting of C18 (i.e. 52% oleic acid and 45% linoleic acid58), are byproducts of the paper/Kraft pulping process, which is a technology for conversion of wood into almost pure cellulose fibers. These free fatty acids are isolated at a percentage of 30-40% from crude tall oil by distillation105, 106 and could be potential inexpensive precursors for forest-based PHA production107, 108.

1.6.3 Downstream processing for PHA isolation

In addition to the production costs contributed by feedstocks, another major cost

contributor is the recovery process of PHAs. PHAs are known to be stored inside cells and the native granules are surrounded by phospholipid membranes. However, only purified PHA polymers exhibit the desired physical and mechanical properties of thermoplastics. Therefore, removal of other cell components from PHAs presents a technical challenge and this isolation process serves to increase the production cost of PHAs.
29

Solvent extraction is the most common PHA recovery technique. Usually, halogenated solvents, such as chloroform and dichloromethane, display good solubility of PHAs and are widely used in the laboratory. In terms of high efficiency (around 90%) and high purity (over 99%), solvent extraction exhibits undoubted advantages over other recovery methods109. Meanwhile, this method can also remove bacterial endotoxins and causes negligible degradation of PHAs. Unfortunately, solvent extraction is not viewed as an environmentally benign method, and high costs of these solvents also hamper large-scale application in the industry. In addition to the solvents for dissolution of PHAs, anti-solvents need to be used for precipitation of PHAs from the solvents. Methanol, ethanol and heptane are usually employed to precipitate and purify PHAs in this extraction process, in which these solvents might contribute up to 50% of the entire production costs of PHAs110 and are not commercially economical if used at an industrial scale for PHA production. However, some non-chlorinated solvents could be acceptable alternatives in terms of environmental and economic concerns. Nonchlorinated solvents, including ethyl acetate, acetone (hot), cyclic carbonate, methyl tertiary butyl ether (MTBE), etc., can be used to partly reduce costs and water could also be used as an inexpensive anti-solvent to precipitate PHAs110, 111. Besides solvent extraction, another strategy for PHA isolation is to remove nonPHA cell mass (NPCM) from PHAs and keep PHAs in the solid state during the isolation process. In this regard, cell disruptions are necessary and can be performed in three ways: chemical digestion, enzyme disruption and mechanical disruption.

30

Chemical agents, such as NaOH, hypochlorite and surfactants, can destroy the cells and dissolve cell components into the supernatant, followed by centrifugation or filtration to collect PHA pellets112. However, this method is a non-selective process and polymer degradation occurs simultaneously. Compared to solvent extraction, the purity of PHAs isolated by chemical digestion is lower and varies from 68% to 98% based on the digestion conditions and the types of microorganisms109,
111

. A combination of

heating, H2SO4, NaOH and sodium hypochlorite at proper concentrations and duration could improve the quality of PHAs with purity higher than 97% and a recovery yield higher than 95%113. Enzymatic disruption, employing proteolytic enzymes, will break cells by hydrolysis of proteins through cleavage of peptide bonds. The proteases help to lyse the cells efficiently at 50 C and pH 9.0 and release PHAs from the cell at a purity of 88.8%. Supplemented with chloroform or sodium dodecyl sulphate-ethylenediaminetetraacetic acid (SDS-EDTA), higher purity was achieved with a recovery yield of 90%109,
111

Nevertheless, the high cost of enzymes and complexity of the recovery process outweigh advantages of this method. Mechanical disruption, using a bead mill, high-pressure homogenization (including microfluidization) or supercritical fluid (SCF), is favored mainly due to the elimination of harsh chemicals, which minimizes environmental effects. The recovery yield of this process can reach 89%111. Pretreatments using SDS or NaOH help to further purify PHAs.

31

The major drawbacks of this method are high capital investment cost and long processing time. Currently, a green, low-cost, highly efficient, and environmentally friendly PHA recovery process has not generally been accepted or implemented. Therefore, based on the characteristics and requirements for the end use of PHAs, a combination of several aforementioned recovery methods might be beneficial to reduce the production costs of PHAs.

32

2. Materials and Methods 2.1 Microorganism and fermentation conditions 2.1.1 PHB production in shake flasks and fermentors

Burkholderia cepacia ATCC 17759 was used in both shake flask as well as pilot

scale (200 L) fermentations. The nitrogen-limited mineral salts medium used in the fermentations was initially described by Bertrand et al.114. The recipe for this medium was slightly changed in our experiments as follows: 1.5 g of (NH4)2SO4, 3.6 g of Na2HPO4 , 1.5 g of KH2PO4, 75.5 mg of CaCl2, 60 mg of NH4-Fe(III) citrate, 200 mg of MgSO4 7H2O, and 1 ml of trace elements solution per liter. The trace elements solution contained 100 mg of ZnSO4 7H20, 30 mg of MnCl2 4H20, 300 mg of H3BO3, 200 mg of CoCl2 6H20, 20 mg of CuSO4 5H20, 20 mg of NiCl2 6H20, 30 mg of NaMoO4 2H20 per liter. For shake-flask experiments, all cultures were shaken at 30 C and 150 rpm. Glycerol (99.5%, EMD) and xylose (99% purity, Acros) were used to produce PHB for physical- chemical characterizations and were autoclaved separately in a solution of 50% (v/v) or 50% (w/v), respectively. All shake flask experiments were conducted in 500 mL baffled flasks containing 100 mL of medium with metal enclosures In experiments utilizing both xylose and glycerol as carbon sources, xylose (2.2%) was initially added into the medium. At 24 h or 48 h, 2% or 5% glycerol was added to the medium, and cultures were harvested at 72 h. Only 2.2% xylose was used as a carbon source for a control. When using the 400 L fermentor (Model No: IF 400, New Brunswick Scientific Co. Inc., NewBrunswick, NJ), a fed-batch method was used and glycerol (85%

33

purity) from a biodiesel-producing facility (Twin Rivers Technologies, Quincy, Mass.) was added as the primary carbon source for production of the homopolymer PHB. Tall oil fatty acids (MWV L-1A, MeadWestvaco Corporation, Richmond, VA) were autoclaved as received. Two milliliters of tall oil fatty acids were added as a carbon source into the 100 ml medium for bacterial growth and PHA production. Lactose (>99%, Sigma-Aldrich) was prepared as a stock solution at 20% (w/v) and sterilized separately. Cheese whey permeate (15%-20%, w/v) from Crowley Foods, a local subsidiary of HP Hood LLC in Arkport, NY, was adjusted to pH 7.0 and filtered by Whatman paper to remove the precipitate, followed by filter sterilization using Nalgene disposable filter units with 0.22 m PES membrane (ThermoFisher Scientific Inc.). The purities of other biodiesel-glycerol sources (from FutureFuel Corporation, Clayton, MO and SUNY-ESF biodiesel-producing facility, Syracuse, NY) were 93% and 40%, respectively. The ESF crude glycerol was adjusted to neutral pH before sterilization. All sources of biodiesel-derived glycerol were autoclaved at 121 C for 20 min before use in fermentation experiments.

2.1.2 PHB-co-HV production in shake flasks and fermentors

PHB-co-HV copolymers were produced by B. cepacia ATCC 17759 in shake flasks, a

7 L fermentor or a 400 L fermentor. The mineral salts-trace elements medium described previously was employed in these experiments, except for further addition of
34

ammonium sulfate to 10 g/L in the fermentor experiments. Reagent-grade glycerol (99.5%, EMD, Gibbstown, NJ) and levulinic acid (98%, Alfa Aesar, Ward Hill, MA) were used as carbon sources for bacterial growth and polymer production. Glycerol and levulinic acid were dissolved into ddH2O for a stock solution of 50% (v/v) and 50% (w/v), respectively. The stock solution of levulinic acid was adjusted to pH 7.0 before autoclaving. In shake flask experiments, Fernbach flasks (2800 ml in volume) were employed for copolymer production. Each Fernbach flask was prepared with 500 ml mineral salts medium and 3% (v/v) glycerol. Levulinic acid at a concentration of 0.1% (w/v) was added, except for the control group (only containing glycerol as a carbon source), when inoculated with a 5% (v/v) seed culture. At 24 h, the remaining levulinic acid at concentrations of 0.2%, 0.4%, 0.6% or 0.8% were added into the medium, to achieve final concentrations of levulinic acid in the medium of 0.3%, 0.5%, 0.7% and 0.9%, respectively, along with the control group without levulinic acid and the group with only 0.1% levulinic aicd. All Fernbach experiments were incubated at 30 C and shaken at 150 rpm for 72 h. Copolymers of PHB-co-HV, with a ratio of HV between 5% and 32.6%, were produced by B. cepacia in a 7-L fermentor (BIOFLO410, New Brunswick Scientific Co., Edison, NJ) with a working volume of 5 L. The concentration of glycerol was kept between 1% and 3% (v/v) during the fermentation and pure levulinic acid was continuously pumped into the fermentor at the rate of 0.5 g/Lh. Dissolved oxygen (DO) and pH were maintained at 40% and 7, respectively, by adjusting agitation/aeration and pumping 10 M sodium hydroxide. For this study, two stages of the fermentation were
35

employed: 1) growth of B. cepacia to as high an OD540 (optical density) as possible with sufficient glycerol; 2) initiating levulinic acid addition at the point of highest OD540 for HV production and providing additional glycerol for bacterial growth and PHB production. Scaling up to pilot-plant level, copolymers were produced in a 400 L fermentor (Model No: IF 400, New Brunswick Scientific Co.,Inc. NewBrunswick, NJ), using a fedbatch method and glycerol (85% purity), from a biodiesel-producing facility (Twin Rivers Technologies, Quincy, MA), was added as the primary carbon source and levulinic aicd ( 97% purity, Sigma-Aldrich) was added periodically at concentrations of approximately 1% (v/v) whenever the concentration of levulinic acisd was determined by HPLC to be less than 0.2%.

2.1.3 P3HB-co-4HB production in shake flasks and fermentors

P3HB-co-4HB copolymers were produced by B. cepacia ATCC 17759 in either shake

flasks or a 7-L fermentor. In shake flask experiments, 0.1%, 0.3%, 0.5%, 0.7% and 0.9% (v/v) of 1,4-butanediol (99% ReagentPlus, Sigma-Aldrich) or -butyrolactone ( 99% ReagentPlus, Sigma-Aldrich) was co-fed with 3% (w/v) xylose or 3% (v/v) glycerol as major carbon sources in mineral-salts medium, described previously in 2.1.1. After inoculation with a 5% (v/v) seed culture shaken for 48 h, all flasks were shaken at 30 C and 150 rpm for 72 h. In a 7 L fermentor, Luria-Broth (LB) medium, containing 10 g/L BactoTM tryptone (BD, Sparks, MD), 5 g/L yeast extract (Sensient Technologies
36

Corporation, Junean, WI) and 5 g/L sodium chloride (EM Science, Gibbstown, NJ), was used for bacterial growth, supplemented with 1,4-butanediol or -butyrolactone at a total concentration of 2% (v/v) for 72 h at 30 C with 30% dissolved oxygen. Cell harvest and polymer isolation followed the procedure described in section 2.3.

2.2 Concentration determination of glycerol, xylose, levulinic acid and lactose

Glycerol concentration was measured using the Free Glycerol Reagent kit (Catalog

No. F6428, Sigma, St,Louis, MO), basically containing glycerol kinase, glycerol phosphate oxidase, peroxidase, ATP, 4-aminoantipyrine (4-AAP) and sodium N-ethyl-N-(3sulfopropyl) m-anisidine (ESPA). The reactions were incubated for 5 min at 37 C and the absorbances were recorded spectrophotometrically at 540 nm (CARY 300

Spectrophotometer, Varian Inc. USA) as described in details per the instructions of the manufacturer (http://www.sigmaaldrich.com/etc/medialib/docs/Sigma/Bulletin/f6428 bul.Par.0001.File.tmp/f6428bul.pdf). Briefly, the Free Glycerol Reagent was pipetted at a volume of 0.8 ml into each cuvet. Water, Glycerol Standard (Catalog No. G7793, Sigma), and sample were added at a volume of 10 l to each cuvet labeled as blank, standard, and sample, respectively. The solution in each cuvet was mixed by gentle shake and all samples tested were incubated at 37 C for 5 minutes, and the absorbances at 540 nm were recorded of Blank, Standard, and Sample versus water as reference.

37

Note: the glycerol standard solution concentration is 0.26 mg glycerol/ml, with an equivalent triolein concentration of 2.5 mg/ml.

Sugars, including xylose and lactose, and organic acids, including levulinic acid, were analyzed by HPLC (Waters Corporation, Milford, MA) equipped with a Waters 717 plus autosampler, a 2996 photodiode array detector, a 2414 refractive index detector and a 1525 binary HPLC pump. For sugar analysis, a carbohydrate analysis column (WAT084038, 3.9 x 300 mm, Waters, Milford, MA) was employed with a mobile phase containing 20% Milli-Q water and 80% acetonitrile which had been filtered through a 0.22 m membrane before use. A Jordi Gel DVB organic acid column (Cat. 17001, 250 mm x 10 mm, The Nest Group, Inc., Southborough, MA) was used to determine the concentration of levulinic acid. The mobile phase (containing 0.02 M phosphoric acid/acetonitrile/methanol at a ratio of 90/5/5) was run at 1 ml/min and the spectrum was recorded at 214 nm. Empower Pro software (Waters, Milford, MA) was used to collect and analyze data.

2.3 Isolation and purification of PHAs from biomass

Cultures were harvested from shake flasks by centrifugation at 7000 x g for 10 min

(Sorvall model SS-3). The biomass was subsequently washed with distilled H2O, if
38

necessary, using additional methanol (10 ml CH3OH per 100 ml culture) for removal of residual long chain fatty acids, and centrifuged again to remove the supernatant and stored at -20 C until lyophilization. The biomass could be washed up to three times for further removal of residual glycerol, which could interfere with biomass drying after lyophilization. For cultures grown in the 400 L fermentor, the broth was centrifuged at 16,000 rpm using a continuous flow centrifuge (CEPA High Speed Centrifuge Z81 G, New Brunswick Scientific, USA). The harvested biomass was then frozen and lyophilized at -80 C and 200 millitorr for 24 h. The dry biomass was subsequently ground to a powder with a mortar and pestle, mixed with chloroform (10 ml chloroform / 1 g dry biomass), and either stirred at room temperature for 24 h for the biomass from the 400 L fermentor or placed in an incubator at 55-60 C overnight for the biomass from shake flask experiments and 7 L fermentor runs. For the purification of bench-scale quantities of PHAs, the general procedure of Keenan et al.115 was used for all biomass samples. In general, after PHA extraction, the mixture of biomass and chloroform was filtered through Whatman #1 filter paper seated in a porcelain Buchner funnel apparatus to remove cell debris. Subsequently, the PHA-containing filtrate solution was decanted into cold methanol or ethanol (at 4 C) by a ratio of 1: 10 (v/v) during stirring to precipitate PHAs. The powder-like PHAs were collected by filtration through a Pyrex glass Buchner funnel with a fritted disc (finegrade porosity with a maximum pore size of 4-5.5 m).

39

For the extraction and purification of polymers from cells produced at the pilotplant scale, lyophilized cell mass (1 kg) was stirred with 9 L of chloroform in a 10-L Pyrex glass container at room temperature for 24 h. This mixture was filtered as described previously for bench-scale PHA extraction. The polymer solution was concentrated by rotary evaporation using a Flash Evaporator (Buchler Instruments, Fort Lee, NJ) at 70 C to a highly viscous solution and a distillation system (Buchler Instruments, Fort Lee, NJ) was used to recover chloroform. The polymer was precipitated from this solution via slow addition to cold methanol (1 volume of PHA solution / 10 volumes of methanol). Precipitated polymer samples exhibited a fibrous, noodle-shaped morphology and, if necessary, were washed with additional methanol for further purification. After the large noodle-shaped PHA polymers were screened and removed directly by a wire sieve, the remaining floc of PHAs was filtered through a Pyrex glass Buchner funnel with a fritted disc (coarse-grade porosity with a pore size of 40-60 m). All purified PHA samples were air dried in glass petri dishes in a hood overnight, followed by vacuum drying in a desiccator until use.

2.4 Sample preparation with nucleating agents

The purified homopolymer PHB and various copolymers of PHB-co-HV (between

0.5 g and 1 g) were dissolved in chloroform, and mixed with either 1% (w/w) HyperformHPN-68L (Milliken Chemical, Spartanburg, SC) or 5% ULTRATALC609 (Barretts Minerals Inc., Dillon, MT). The mixtures were stirred for 2 h and poured into
40

Petri dishes and the solvent was allowed to evaporate overnight. The resulting films were dried in an incubator at 60 oC for at least 4 h prior to thermal analysis.

2.5 GC analysis for PHAs

Monomer composition and yield of PHAs were determined by gas chromatography

on a GC2010 equipped with AOC-20i autosampler (Shimadzu, Japan) as described previously by Nomura et al.116. Briefly, 15 mg of dry biomass were weighed, mixed with 2 ml of H2SO4/methanol (15:85) and 2 ml chloroform, and incubated at 100 C for 140 min to extract the polymer and subjected it to a methanolysis reaction to form the methyl esters. The mixture was then cooled to room temperature and 1 ml of distilled water was added and the solution was vortexed for 1 min. The mixture was allowed to separate into an aqueous and an organic phase and the organic layer (0.5 ml) was removed, filtered (PTFE membrane, 0.22 m), and mixed with 0.5 ml 0.1% caprylic acid in chloroform. This mixture was subjected to analysis by GC with a flame ionization detector (FID). Application of the sample was by split injection of 1 l onto a 30-m Rtx5 (5% diphenyl-95% dimethyl polysiloxane) column with a 0.25 mm ID (Restek, PA, USA ). The injection port was held at 280 C and the oven housing the column was held at 100 C for 3 min. The oven temperature was then raised 8 C min-1 to 280 C where the column was held for 2 min. The oven temperature was raised to 310 C at a rate of 20 C min-1 and held for 10 min to remove any residuals from the column before the next sample was applied. Products were detected by FID at a temperature of 310 C.
41

2.6 Physicochemical property and mechanical property tests of PHAs 2.6.1 Molecular mass determination

Number average molecular weight (Mn) and weight average molecular weight (Mw)

were determined by gel permeation chromatography (GPC) using an LC-20AD Liquid Chromatograph equipped with a SIL-20A auto-sampler and RID-10A refractive index detector (Shimadzu, Japan). PHB polymers were dissolved in chloroform to a final concentration of 0.7 mg/ml and filtered (PTFE membrane, 0.22 m) before analysis. A SDV 8 x 300 mm column with 5 m porosity was used as the stationary phase (Polymer Standards Service, USA) with an oven temperature of 40 C. The mobile phase was chloroform at a flow rate of 1 ml/min. Standard curves for molecular weight were created using polystyrene standards with a range from 682 to 1,670,000 Da and low polydispersities (Polystyrene High MW Standards Kit, Polymer Standards Service, Warwick, RI).

2.6.2 Thermal analysis

Thermogravimetric analysis (Hi-Res TGA 2950 or TGA QSeries-Q5000, TA

instruments, New Castle, DE, TA Instruments) was used to determine the decomposition temperature (Tdecomp.) of PHAs. Ten milligrams of PHA films were folded into platinum trays, which were pre-cleaned by burning on a gas burner followed by taring to calibrate before use. Samples were then subjected to a heating rate of 20 C /min from ambient
42

temperature to a final temperature of 500 C while purging with nitrogen. Tdecomp. was determined at the point where the polymer initiated weight loss Differential scanning calorimetry (DSC, model 2920 or QSeries-Q200, TA Instruments, New Castle, DE, TA Instruments) was used to characterize the melting temperature (Tm), glass transition temperature (Tg), and crystallization temperature (Tc) for all polymer samples. Aluminum (Tzero) pans were loaded with 10 mg of sample and sealed with Tzero lids (TA Instruments, New Castle, DE). The temperature was increased from room temperature to 200 oC at a heating rate of 10 oC/min, decreased at a cooling rate of 5.0 oC/min to -50 oC and then increased to 200 oC at a heating rate of 10 oC/min. Universal Analysis 2000 software (TA Instruments, New Castle, DE) was used for data analysis.

2.6.3 Tensile test

Polypropylene homopolymer pellets (Basell Pro-fax 6523 in natural color,

LyondellBasell Industries, Houston, TX) were generously provided by Tessy Plastics Corp. (Elbridge, NY). All pre-purified PHA samples were re-dissolved into chloroform and solvent-cast films were prepared in glass petri dishes, which were placed in a fume hood to allow for the evaporation of chloroform. All films were subsequently vacuum-dried in a desiccator overnight, followed by oven drying at 65 C overnight. A Carver Standard Press (Press No. 3851, Carver Inc, Wabash, IN) was used as a compression modeling to generate polymer sheets (40 mm x 30 mm x 0.4 mm, in length/width/thickness) from
43

molten films, which were heated to 360 F for 1.5 min in a Teflon spacer (0.4 mm in thickness) and subjected to an increasing pressure of 0.25 metric ton per 15 sec, and increased to 1 metric ton in order to remove the air bubbles trapped in the sheets. The polymer sheets were then cooled down by using tap water, if necessary, putting the whole polymer sheet sandwich in a refrigerator to make PHAs crystallize and solidify faster and left at room temperature overnight. All polymer sheets were pre-warmed at 200 F to slightly soften the materials and punched into a dogbone-type geometry using a custom dogbone die (Machined Tool Steel, TestResources Inc, Shakopee, MN), using a Carver press and minimum pressure (0.5 metric ton). The dimensions of the dogbone die were determined from a scaled down D638-03 type IV ASTM (American Society for Testing and Materials) standard, and had a width of 1.5 mm, gauge length of 6.25 mm and a total length of 28.75 mm. The thickness of each dogbone sample was measured by a PRO-MAX electronic digital caliper (Fred V. Fowler Co., Inc., Newton, MA) and averaged at three locations (both ends and the mid-point) The mechanical properties of PHA samples were evaluated by uniaxial tensile testing employing a Linkam TST 350 (20 N load cell at 0.01 N resolution, Linkam Scientific Instruments LTD, Tadworth, Surrey, UK)117. Samples were fixed by two clamping jaws and extensionally deformed until failure at a crosshead speed of 100 m/s (96%/min) at room temperature. SteREO Discovery V8 Stereo-microscope (Carl Zeiss Inc., Thornwood, NY) was employed for 3-D imaging. Tensile tests for each sample were conducted in triplicate. Youngs modulus was calculated by determining the initial slope of the stress versus strain curve (0<<10%) using linear regression. The tensile
44

strength was recorded as the stress at which the sample completely broke. The elongation to break is the strain on a sample when it breaks. Linksys 32 software (Linkam Scientific Instruments LTD, Tadworth, Surrey, UK) was used to communicate between the TST 350 system and the computer to collect and store data. All raw data were analyzed and converted by Linksys 32 to the compatible version of the SigmaPlot software (Systat Software Inc., San Jose, CA), which produced the stress-strain curve for a direct indication of the material mechanical properties. The engineering measures of stress and strain, denoted in this module as and respectively, are determined from the applied load or force (P) and the displacement () over the original cross sectional area (A) and the original gauge length (L), respectively (shown in Equation 1).

------ Equation 1

The cross sectional area A=width x thickness. In this research, the width is set to 1.5 mm and the thickness is equal to or approaches 0.4 mm based on the aforementioned measurements for dogbone thickness described in this test. As also mentioned previously, the original gauge length L is set to 6.25 mm. The calculated data were used to generate stress-strain curves in SigmaPlot 11.

45

2.6.4 Nuclear magnetic resonance

In additional to GC analysis for PHA composition, the monomeric composition of

13

PHAs was also determined by solution-state 1H and

C nuclear magnetic resonance

(NMR) spectroscopy using a Bruker BioSpin AVANCE 300 or 600 NMR spectrometer (Bruker BioSpin NMR Corporation, Billerica, MA). PHA samples were prepared for analysis by dissolving solvent-cast film segments in deuterated chloroform at a concentration of 1% (w/v) via mixing with mild heating. All samples were analyzed at 300 MHz, except for the analysis of the end-capped PHB, which was analyzed by 1H NMR spectroscopy using a Bruker BioSpin AVANCE 600 operating at 600 MHz. The analyses for glycerol, PHB and PHB-co-HV were accomplished using XWIN-NMR version 3.1 software (Bruker BioSpin NMR Corporation, Billerica, MA). The copolymer composition, expressed as the mol% 3HV in the PHB-co-HV copolymers, was determined according to the procedure described by Marchessault and his co-workers118, as a ratio of peak areas of the side-chain methyl resonance of HV and the sum of methyl resonances of HB and HV. The methyl resonances were determined by integration of the fully expanded and normalized spectra of these two methyl groups of HB and HV units.

46

3. Results 3.1 Renewable carbon sources for bacterial growth and PHA production by B. cepacia

Xylose (a five-carbon sugar present in wood hydrolysates), glycerol (a byproduct

from the biodiesel process), tall oil fatty acids (TOFAs) (byproducts in the paper pulping process, comprised of 52% oleic acid and 45% linoleic acid) and cheese whey permeate (containing 15-20 wt% lactose), were tested as potential carbon sources for bacterial growth and PHA production by B. cepacia. As shown in Figure 7, these three carbon sources (xylose, glycerol and TOFAs) can be metabolized by B. cepacia and used to support bacterial growth and PHA production, respectively. After 96 h, the amount of dry biomass from xylose(3%, w/v), glycerol (3%, v/v) and TOFA (2%, v/v) reached 4.92 g/L, 5.82 g/L and 3.78 g/L, respectively. When using xylose and TOFA as carbon sources, the highest PHA yield occurred at 72 h at 50.99% and 49.88% of dry biomass, respectively. When using glycerol as a carbon source, the maximum PHA yield at 81.88% of dry biomass occurred at 96 h. These data indicated that it is feasible to use these three renewable and inexpensive carbon sources for PHA production. Among these three feedstocks, glycerol is a much better carbon source than the other two, in terms of biomass and PHA yield. Pure lactose and cheese whey permeate did not support substantial growth of B. cepacia or PHA production (data not shown). It means that cheese whey permeate could not be used as an inexpensive carbon source by B. cepacia for PHA production. However, cheese whey permeate might be explored as an inexpensive nitrogen source for bacterial growth when supplementing other carbon sources due to the existing whey protein in the permeate.
47

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0 48 72 Time (h) 96

90 80 70 3HB yield (%) 60 50 40 30 20 10 0 Biomass(g/L) from tall oil Biomass(g/L) from glycerol Biomass(g/L) from xylose 3HB% from tall oil 3HB% from glycerol 3HB% from xylose

Figure 7 Bacterial growth and PHA production using renewable carbon sources (tall oil fatty acids, biodiesel-derived glycerol and xylose) by B. cepacia in shake flask experiments.

By GC analysis and 1H-NMR, the purified PHA polymers from these three carbon sources were comprised solely of 3HB as the repeating subunit (data not shown). However, some microorganisms, e.g. Pseudomonas putida, Pseudomonas oleovorans, Aeromonas hydrophila, can produce mcl-PHAs, consisting of monomeric units that include HHx (C6), HO (C8), HD (C10) and HDD (C12)107, 108, 119. B. cepacia will only produce the short chain length PHB, even when using long chain free fatty acids, e.g. oleic acid, as carbon sources.
48

Biomass (g/L)

3.2 Biodiesel-derived glycerol as a carbon source for PHA production

Four different purities of glycerol (pure glycerol, >99%; biodiesel-glycerol from

Future Fuel, 93%; biodiesel-glycerol from Twin River Technologies, 85%; biodieselglycerol from SUNY-ESF, 40%) were used as carbon sources for bacterial growth and PHB production, in order to test the ability of B. cepacia to utilize crude or highly pure glycerol. As shown in Figure 8, B. cepacia can utilize glycerol (30 g/L, v/v) as a carbon source from high to low purity. At 96 h, the highest biomass at 6.48 g/L was achieved from pure glycerol and the lowest biomass at 4.98 g/L from ESF glycerol. PHB was produced in high yield (over 80% of dry biomass) by B. cepacia from all four types of glycerol. For pure glycerol and FutureFuel glycerol, either PHB yield was close to 90% of dry biomass. For Twin River Technologies and ESF glycerol, both PHB yields approached 82% of dry biomass. Glycerol was consumed as a carbon source for bacterial growth and PHB production during the entire fermentation process (Figure 9A). After 120 h, glycerol content in the medium decreased by almost 20 g/L when using pure glycerol, FutureFuel glycerol and Twin River Technologies glycerol, respectively. However, ESF glycerol was consumed by only 10.7 g/L (Figure 9B), which consumption of glycerol is much lower than the other three more pure glycerol, probably due to the existence of free fatty acids in this ESF crude glycerol.

49

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0 48 72 Time (h) 96

100 90 80 70 PHB yield (%) 60 50 40 30 20 10 0 Biomass(g/L)-pure glycerol Biomass(g/L)-Future Fuel Biomass(g/L)-Twin River Biomass(g/L)-ESF PHB%-pure glycerol PHB%-Future Fuel PHB%-Twin-River PHB%-ESF

Figure 8 Bacterial growth and PHB production from various sources of biodiesel-derived glycerol by B. cepacia in shake flask experiments.

Biomass (g/L)

50

40 35 30 Glycerol content (g/L) 25 20 15 10 5 0 0

A
pure glycerol

ESF
B
glycerol (g/L) 30 20 10 0

esf futurefuel twinriver

Future Fuel
pure futurefuel twinriver esf

Twin River Pure glycerol

24

48 Time (h)

72

96

120

Figure 9 Glycerol consumption during fermentation using pure glycerol, FutureFuel glycerol, Twin River Technologies glycerol and ESF glycerol by B. cepacia in shake flask experiments. Graph A demonstrates the glycerol content at each time point after inoculation. Graph B indicates total glycerol consumption after 120-h fermentation.

51

3.3 Production and characterization of PHB homopolymer by B. cepacia using glycerol as a carbon source 3.3.1 Effects of glycerol content on bacterial growth
3.5 3 Dry cell mass (g/L) 2.5 2 1.5 1 0.5 0

3%

5%

7%

9%

Glycerol concentration

Figure 10 Dry biomass of B. cepacia grown in shake flasks on different concentrations of pure glycerol.

When different concentrations of pure glycerol were added to the medium (Figure 10), there occurred a gradual decrease in dry biomass with increasing glycerol concentration. For example, dry biomass reached 2.8 g/L at 3% (v/v) glycerol added in shake flask experiments, whereas when 9% (v/v) glycerol was added, the amount of dry biomass decreased to 1.3 g/L.

52

3.3.2 Effects of glycerol content on molecular mass of PHB

Number average molecular weights (Mn) and weight average molecular weights

(Mw) of thin PHB solvent-cast films were determined by GPC. When concentrations of glycerol were increased from 3% to 9%, both Mn and Mw decreased gradually from 173 kDa and 304 kDa to 87 kDa and 162 kDa, respectively (Figure 11).

350 300 Mn Mw

Molecular mass (kDa)

250 200 150 100 50 0 3% 5% 7% 9% Different concentrations of glycerol (v/v)

Figure 11 Number-average molecular weight (Mn) and weight-average molecular weight (Mw) of PHB produced by B.cepacia grown with different concentrations of glycerol.

Mixtures of xylose and glycerol were also examined as carbon sources. Initially, 2.2% (w/v) xylose was added to cultures at the time of inoculation and glycerol was added at time points of 24 h and 48 h at concentrations of either 2% (v/v) or 5% (v/v). Subsequently, the cultures were harvested at 72h. Mn of PHB from 5% glycerol was lower than that from 2% glycerol added to the medium at the same time point. Higher
53

glycerol concentrations resulted in the production of PHB with overall lower molecular mass (Table 1). Interestingly, when the cells were grown on the same concentration of glycerol (2% or 5%) with 2.2% xylose, PHB had lower Mn (133.6 or 114.0 kDa) as cells were exposed to glycerol for a longer time period (48 h) and higher Mn (174.6 or 138.4 kDa) when exposed to glycerol for a shorter time period (24h), respectively. The same trend was observed for Mw of PHB produced from 2% or 5% glycerol with 2.2% xylose (data not shown). When 2.2% xylose was provided as the sole carbon source, the Mn reached 468.3 kDa (Table 3). The xylose-based PHB polymer was found to be of greater molecular mass than PHA produced from or exposed to glycerol. High concentrations of glycerol increased the opportunity for incorporation of glycerol into the polymer. Also, exposing the cells to glycerol for a longer time periods increased the opportunity to terminate the elongation of PHB. Therefore, higher concentrations of glycerol or longer exposure to glycerol resulted in PHB with lower molecular masses. Table 3 The effect of concentrations and exposure times of glycerol on the number -average molecular weight (Mn) of PHB.

Exposure time in glycerol (h) b Glycerol concentration (v/v) a 0 24 48 c No glycerol (control) 468.3 / / 2% / 174.6 133.6 5% / 138.4 114.0 a. All treatments contained 2.2% (w/v) xylose as the primary carbon source. b. Glycerol additions were made at the time points of 24 and 48 h with all treatments harvested at 72 h. c. All data reported in kiloDaltons (kDa).
54

3.3.3 Characterization of PHB end-capped with glycerol by 1H NMR

1

H NMR shows the results from PHB synthesized in the presence of 7% glycerol

(Figure 12). The full spectrum showed the expected resonances for PHB as demonstrated by the methyl group at 1.25 ppm, the methylene group between 2.45 and 2.65 ppm and the methine group at 5.25 ppm. However, expansion of the spectral region between 3.0 and 4.5 ppm revealed the presence of additional resonances corresponding to terminal glycerol groups. The expanded region with three resonances

Figure 12 1H-NMR of PHB produced by B .cepacia grown on 7% (v/v) glycerol as a carbon source. The expanded region indicates glycerol as the terminal end-group. at 3.7 ppm, 3.95 ppm and 4.18 ppm showed the terminal esterification of glycerol to PHB through the primary hydroxyls (C1 or C3 positions of glycerol). Resonance at 3.86
55

ppm showed glycerol end-capping of PHB through the secondary hydroxyl group. Because glycerol is composed of 2 primary and 1 secondary hydroxyl groups, the possibility exists that glycerol termination of PHB polymers could also be the result of covalent bonding at the secondary hydroxyl group of glycerol. These results were identical to those reported by Ashby et al120.

3.3.4 Physical properties of PHB

Thermal characterization by DSC and TGA of PHB films are shown in Table 4.

Although the molecular mass of glycerol-based PHB was lower than that of xylose-based PHB, there was no significant change for both Tg and Tm; however, Tdecomp. increased from 268.6 C to 281.5 C. Higher decomposition temperature provides a broader separation between the required melting temperature for injection molding and thermal degradation of the polymers. Table 4. Physical-chemical properties of PHB produced from xylose and glycerol PHB Xylose-based PHB a Glycerol-based PHB b Tm (oC) 178.6 181.9 Tg (oC) -6.6 1.6 Tdecomp. (oC) 268.6 281.5

a. PHB produced from 2.2% (w/v) xylose as sole carbon source. b. PHB produced from 3% (v/v) glycerol as sole carbon source.

56

3.3.5 Pilot scale (200 L) fermentation using biodieselglycerol

The initial evaluation of physical properties of glycerol-based PHB demonstrated

the potential application of biodiesel-derived glycerol as a carbon source to support growth and PHB production by B.cepacia. Subsequently, biodiesel-glycerol was used as the carbon source for a 200 L pilot scale fermentation. Dissolved oxygen (DO) and pH were controlled at 35-40% and 7.0, respectively. As shown in figure 5, glycerol concentration

35 30 Biomass (g/L) and PHA% 25 20 15 10 5 0

Biomass (g/L) PHA% Glycerol content (g/L)

45 40 Glycerol content (g/L) 35 30 25 20 15 10 5

24

36

48

60

72 Time (h)

84

96

108

120

Figure 13 Changes in biomass, PHA% and glycerol concentration during a fed-batch fermentation, in a 400-L fermentor, with periodic additions () of biodiesel-glycerol.

57

was kept between 10 g/L and 40 g/L with periodic additions of glycerol. Biomass gradually increased and reached maximum density of 25.8 g/L at 108 h. PHA accumulated quickly from 0.7% at 36 h to 17.4% at 48 h, and PHA% fluctuated near 20% before increasing to the highest point of 31.4% at 120 h. In total, 26.5 L glycerol and 2.5 kg ammonium sulfate were added to the 400 L fermentor in a fed-batch mode and the cells were harvested at 120 h. The final yield of dry biomass and PHB were 23.6 g/L and 7.4 g/L, respectively.

3.3.6 Injection-molding process for conversion of PHB to biodegradable eartips

After cell harvest from the 200 L fermentation described previously in 3.3.5,

followed by isolation and purification of PHB homopolymer from freeze-dried biomass (see 2.3 in Materials and Methods), pure PHB (>99% purity, Figure 24a) was achieved through a second purification process at AMRI Syracuse Research Center (North Syracuse, NY) and then delivered to Tessy Plastics Corporation (Elbridge, NY), which finally converted PHB to medical eartips (Figure 24c) by an injection-molding process (Figure 24b).

58

Mold

Hopper for loading PHA granules

Nozzle for Injection of molten PHAs

Heater for PHA melting

Figure 14 (a) Purification and vacuum drying process of PHB, (b) Schematic flowsheet of injection-molding process (http://www.ppind.com/capabilities/injectionmolding.html), (c) Eartips made from polypropylene (black, left) and PHB (brown, right two).
59

3.3.7 Thermal degradation of PHB

The PHB polymer produced in the pilot plant fermentations was heated to

approximately 200 C for melting during the injection-molding process, and this polyester might thermally degrade at high temperature as indicated by weight loss. After eartips were made (Figure 14c), random pieces of eartips were dissolved into chloroform to determine their molecular masses. Table 5 Comparison between the original PHB and the thermally degraded PHB Sample PHB* PHB from eartip a# PHB from eartip b# PHB from eartip c# PHB from eartip d# Mn 137803 3373 44126 93683 51332 Mw 271807 32794 102485 210541 117542 Mz 456385 115311 207371 355760 214403 PDI (Mw/Mn) 2.0 9.7 2.3 2.2 2.3

Note: * represents the original PHB before heating; # represents PHB pieces from eartips after heating. Unit for molecular mass is Dalton.

As shown in Table 5, all PHB samples from eartips, lost weights by at least 32% of the original PHB mass for sample c, based on number average molecular weight (Mn). The worst thermal degradation occurred for sample a by almost 98% weight loss. All PHB-eartip samples exhibited higher polydispersity (PDI) than the original PHB,
60

accounting for more diverse polymer sizes resulting from polymer chain scission during heating.

3.4 Production and characterization of PHB-co-HV copolymers using glycerol and levulinic acid as substrates in shake flasks by B cepacia 3.4.1 Bacterial growth and production of PHB-co-HV copolymers when co-feeding glycerol and levulinic acid in shake flasks
90 80 PHA yield and HV fraction (%) 70 60 50 40 30 20 10 0 0 0.1 0.3 0.5 0.7 0.9 Levulinic aicd (%, w/v) 10 9 8 7 6 5 4 3 2 1 0 Dry cell mass (g/L)

HV% of PHA PHA% of DCM DCM (g/L)

Figure 15 Effects of levulinic acid concentration on bacterial growth and production of PHB-co-HV copolymers. DCM: dry cell mass.

61

In addition to glycerol (3%, v/v), levulinic acid was added to the medium for production of PHB-co-HV copolymers. By increasing the concentration of levulinic acid in the medium, the HV fraction in the copolymer increased correspondingly. As shown in Figure 15, when the concentration of levulinic acid gradually increased from 0 to 0.9% (w/v), the mol fraction of HV in the copolymer increased from 0 to 35.8%. However, levulinic acid at a concentration of greater than 0.1% (w/v) indeed displayed an inhibition effect on bacterial growth and PHA production. As a result, biomass and PHA yield decreased from 7.6 g/L to 5.5 g/L and from 77.8% to 51.7%, respectively. Therefore, the relationship between mol% HV in the copolymer and the concentration of levulinic acid in the medium should be carefully considered during the fermentation for production of PHB-co-HV copolymers.

3.4.2 1H and 13C-NMR analysis for the structure and composition of PHB-co-HV copolymers

Composition of the PHB-co-HV copolymers could be mediated by regulating the

ratio of levulinic acid to glycerol in the fermentation broth, adding co-substrate doses from 0.1% to 0.9% (w/v) (Figure 15). The structure and composition of these copolymers were determined by gas chromatography using known 3-hydroxyalkanoate standards. In addition, 1H and 13C-NMR were employed to further confirm the composition of PHB-coHV copolymers.

62

(a)
0.4 mol% HV

17.6 mol% HV

29.5 mol% HV

33.0 mol% HV

169.49 169.29 169.12

77.42 77.00 76.57 71.85 67.65 67.58

40.76 38.77 38.61

26.83

19.74

(b)
4
CH3 C H2 O

5
H2C CH

CH3 O

CH

C H2

HV3
m

HV2 HV4

HB

HV

HB3 HB1/HV1
169.5 169.0 ppm

HB2

HB4 HV5

220

200

180

160

140

120

100

80

60

40

20

9.34 9.31

ppm

28.62 7.72 10.96 1.10

11.80 7.49

6.67

1.51

6.10

63

11.55

6.49

Figure 16 (a) Chemical shift expansions of 300 MHz 1H NMR spectra from PHB-co-HV copolymers, illustrating a progressive increase in the mol% of HV and quantified by the integrated areas of the HB doublet (methyl group of HB, 1.27 ppm) and the HV triplet (methyl group of HV, 0.90 ppm). (b) Chemical shift assignments of 300 MHz
13

C NMR

spectrum of PHB-co-35.8 mol% HV, which was produced by B. cepacia using 3% (v/v) glycerol and 0.9% (w/v) levulinic aicd.

The progressive increase in mol fraction of HV is shown by the expanded and normalized 1H NMR spectra depicted in Figure 16a, which highlight the integrated peak area change of the methyl groups in HB and HV monomeric units, represented by a doublet peak (1.27 ppm) and a triplet peak (0.90 ppm), respectively. These 1H NMR chemical shifts of HB and HV, including all methyl, methylene (data not shown) and methane (data not shown) groups, match those reports published for bacterial and chemically synthesized PHB-co-HV copolymers33, 118, 121, 122. Although levulinic acid has shown to be a good co-substrate for the precursor of 3HV, 4HV (4-hydroxyvalerate) might be concurrently produced from levulinic acid with 3HV to form a terpolyester P(3HB-co-3HV-co-4HV) by R. eutropha
123, 124

. Additionally,

comparison between the 300 MHz 13C NMR spectrum in Figure 16b and the 125 MHz 13C NMR spectrum of P(3HB-co-3HV-co-4HV), which was characterized by Valentin et al.9, 125, demonstrated the lack of chemical shifts in the former spectrum at 31 ppm, 70 ppm and 172 ppm, indicating the methylene carbons of 4HV, the methine carbon of 4HV and the
64

carbonyl carbon of 4HV, respectively. This strong evidence supports the conclusion that polymers produced from glycerol and levulinic acid by B. cepacia do not contain the 4HV monomer and should be in a composition of P3HB-co-3HV, considered as a copolymer. The integrated peak areas determined for each carbon atom set (HB1/HV1 carbonyl diad, HB2/HV2, HB3/HV3 and HB4/HV4) in the fully relaxed
13

C NMR spectrum,

corresponding with the mol% of HB and HV, further confirmed that 3HV is indeed present and 4HV is absent in our polymers (Figure 16b). The three carbonyl resonance peaks occurred at 169.12 ppm, 169.29 ppm, 162.49 ppm, and are shown in the expanded inset of the 13C NMR spectrum depicted in Figure 16b, corresponding to HB-HB, HB-HV and HV-HV diad sequences in the sample PHB-co35.8 mol% HV. Based on the integrated areas of these three peaks, the relative intensities of the diad sequences were determined from this spectrum to be 40.4%, 36.6% and 23.0%, respectively (see the enlarged inset in Figure 16b), and are similar to other reports for random PHB-co-HV copolymers121, 122.

3.4.3 Physical properties of PHB-co-HV copolymers

PHB-co-HV copolymers with various mol fractions of HV were isolated, purified

and vacuum-dried for two weeks before DSC analysis. As the mol% of HV increased from 0 to 35.8 mol%, melting temperatures of the copolymers decreased from 179.19 C to a minimum point, 163.56 C and then increased to 168.32 C, which also demonstrated an isodimorphic behavior for PHB-co-HV copolymers (Figure 17a and 17b). Also, the glass
65

transition temperature decreased from 5.73 C to -6.90 C as a result of increasing HV mol% in the copolymers (Figure 17b).

33 mol% HV 35.8 mol% HV 29.5 mol% HV 17.6 mol% HV

0.4 mol% HV 0%

Figure 17a DSC curves displaying melting temperatures of PHB-co-HV copolymers

66

180 175 170 165 160 155 0 0.43 17.6

Tm

8 6 4 2 Tg (oC)

Tm (oC)

Tg

0 -2 -4 -6 -8 29.5 33 35.8

mol% HV in the copolymer

Figure 17b Melting temperatures and glass transition temperatures of PHB-co-HV copolymers with increasing HV mol%. Solid line with square and dot line with diamond represent Tm and Tg, respectively.

3.4.4 Mechanical properties of PHB-co-HV copolymers

The tensile strength test allows us to evaluate mechanical properties of PHAs and

elucidate more comprehensive material properties in addition to physical properties. The stress ()-strain () curve (Figure 18a) provide data on the mechanical properties of a material. This includes well known characteristics such as tensile strength, at which the sample is stretched to break, yield strength, at which a material begins to deform plastically and prior to which the material will deform elastically and will return to its original shape when the applied force is removed, Youngs Modulus, which displays

67

elasticity and stiffness of the material, elongation to break, exhibiting the strain on a sample when it breaks. The stress-strain curve begins as a linear relationship and the initial slope of the linear portion is defined as Youngs Modulus. If the initial slope, calculated from the curve by statistical regression (0< <10%), is larger, the material is more rigid and less elastic. As shown in Figure 18a, the curve for a PHB sample exhibited the sharpest rise at the beginning, corresponding to the highest Youngs Modulus, indicating the PHB homopolymer exhibits the greatest rigidity and the least elasticity among these four samples which included two PHB-co-HV copolymers and polypropylene. Meanwhile, the copolymer PHB-co-17.6 mol% HV exhibited similar elasticity to polypropylene due to these two initial curves being almost identical. The copolymer PHB-co-33 mol% HV exhibited the lowest Youngs Modulus indicating that it is the most elastic material among these four samples in Figure 18a under these test conditions. In addition, visual observation of the polymer samples during stretching at a speed of 100 m/s demonstrated the apparent difference between the PHB homopolymer and the PHB-co-HV copolymers with various mol% HV. Dogbone-typed PHB samples at a thickness of 0.4 mm were pulled to break within 10 seconds and exhibited rather limited displacement (low elasticity) (Figure 18b i). As the mol% HV in the copolymer increased, the copolymer dogbones demonstrated better strain without cleavage during stretching, and it usually took several minutes for the copolymers to break (Figure 18b ii and iii). Basically, the copolymer with higher HV mol% has better stretchability.

68

35 30 25

Polypropylene

Stress (MPa)

20 15 10 5 0 -5 -20 0 20 40

PHB-co-33 mol% HV PHB-co-17.6 mol% HV

PHB
60 80 100 120

Strain (%) Figure 18a Stress-strain curves of PHAs and polypropylene. Dotted line: PHB (produced
by B. cepacia using xylose as a carbon source); Dash-dotted line: PHB-co-17.6 mol% HV; Solid line: PHB-co-17.6 mol% HV; Dashed line: polypropylene. The material properties of each sample were determined by the stress-strain curve and are summarized in Table 6. It is noteworthy that the PHB homopolymer produced from xylose or glycerol as a carbon source exhibited different mechanical properties, possibly due to different molecular masses. Xylose-based PHB with a higher molecular mass exhibited higher tensile strength (26 MPa), slightly higher elasticity (Youngs Modulus, 304 MPa) and better elongation to break (12%) than glycerol-derived PHB with lower molecular mass (tensile strength, 17 MPa; Youngs Modulus, 277 MPa;

69

elongation to break, 8%). Furthermore, PHB-co-29.5 mol% HV exhibited tremendous improvement of elongation to break (28-fold higher than PHB-co-17.6 mol% HV). Therefore, the copolymer with mol% HV between 17.6 and 29.5 might be a good alternative for conventional petroleum-derived plastics.

PHB

iii

PHB-co-33 mol% HV

ii

PHB-co-29.5 mol% HV

Figure 18b Real-time images of stretching for different types of PHAs in tensile testing. i represents PHB homopolymer, of which the dogbone was pulled for less than 0.7 mm to break; ii displays PHB-co-29.5 mol% HV, of which the dogbone was stretched for approximately 35 mm to break; iii represents for PHB-co-33 mol% HV, of which the dogbone was pulled for approximately 66 mm to break.
70

Table 6 Mechanical properties of PHB, PHB-co-HV copolymers and polypropylene Mol% HV in PHB-co-HV copolymers 0 (Xylose-based PHB)a 0 (Glycerol-based PHB)b 17.6 29.5 33.0 35.8 Polypropylene Tensile Strength (MPa) 26 + 1 17 + 3 20 + 2 17 + 1 19 + 2 19 + 1 >32c1 Yield Strength (MPa) 25 + 1 16 + 3 16 + 1 13 + 2 14 + 2 10 + 0.4 30 + 2 Youngs Modulus (MPa) 304 + 32 277 + 20 228 + 2 61 + 1 68 + 10 45 + 14 189 + 16 Elongation to Break (%) 12 + 1 8+2 20 + 2 560 + 37 1046 + 4 1058 + 33 >1067c2

Mnd 468 122 244 284 205 280

PDI 2.0 2.4 2.2 2.3 2.6 2.1

Note: a and b represent PHB produced by B. cepacia using xylose or glycerol as a carbon source, respectively.
c1

and

c2

represent the maximum reading for tensile strength and

elongation to break, respectively, under the conditions set for this tensile test (see the tensile test part in Materials and Methods). weight in units of KiloDaltons. In summary, the PHB-co-HV copolymers demonstrated much better elasticity than the PHB homopolymer based on Youngs Modulus and elongation to break. For the copolymers, a higher mol fraction of HV resulted in greater elasticity and lower yield strength. In addition, all glycerol-derived PHB and PHB-co-HV exhibited similar tensile strength. Polypropylene, used as a reference, demonstrated higher tensile and yield strength than all PHA samples tested in this study; however, it was less elastic than PHBco-HV copolymers containing higher than 17.6 mol% of HV.
71
d

represents number average molecular

3.5 Production and characterization of PHB-co-HV copolymers using glycerol and levulinic acid as substrates in fermentors by B cepacia 3.5.1 Bacterial growth and production of PHB-co-HV copolymers when co-feeding glycerol and levulinic acid in fermentors by B. cepacia in fermentors
50 45 40 35 30 25 20 15 10 5 0 4 12 36 54 72 96 108 124 132 148 156 Time (h) 40 35 30 DCM (g/L) 25 20 15 10 5 0 PHA (% of DCM) and HV (% of PHA)

Figure 19 Growth (dry cell mass, DCM, ), PHA yield () and HV content in the copolymer () produced by B. cepacia using glycerol and levulinic acid as carbon sources in a 7 L fermentor. Previous work
63

has shown that B. cepacia utilizes glycerol as the sole carbon

source for bacterial growth and PHA production in the form of PHB. When provided with the co-substrate levulinic acid, B. cepacia will convert glycerol and levulinic acid to the precursors of 3-hydroxybutyrate and 3-hydroxyvalerate, respectively, and assemble these precursors to form the copolymer PHB-co-HV 115. The ratio of HV in the copolymer varies with the amount of levulinic acid provided to the organism. However, levulinic
72

acid at concentrations greater than 0.1% (v/v) will inhibit growth of B. cepacia and PHA production (data not shown). In this study, copolymers with varying ratios of HV were produced by a fed-batch process in a 7-L fermentor. This fermentation consisted of two stages: 1) A growth stage with glycerol as the sole source of carbon to obtain high cell density for PHB production over the first 120 h, and 2) copolymer (PHB-co-HV) biosynthesis initiated by supplementation of levulinic acid with the addition of glycerol after 120 h. Copolymers with a range of HV (0.97 mol% to 17.9 mol%) were produced as indicated in Figure 19. Table 7 Composition and molecular masses of PHB and copolymers of PHB-co-HV
Composition PHB PHB-co-5.6 mol% HV PHB-co-11.4 mol% HV PHB-co-14.7 mol% HV PHB-co-17.9 mol% HV PHB-co-20.0 mol% HV PHB-co-30.5 mol% HV PHB-co-32.6 mol% HV Mn (x10-3) 137.8 61.3 61.8 69.0 68.0 69.4 49.0 58.0 Mw (x10-3) 271.8 111.3 115.0 107.8 120.2 131.9 109.0 110.1 PDI 2.0 1.8 1.9 1.6 1.8 1.9 2.2 1.9

During the second phase of fermentation, high concentrations of levulinic acid not only inhibited growth of B. cepacia, but decreased PHA yield (Figure 19). Thus, a slow
73

feeding rate (approaching 0.5 g/Lh) for levulinic acid was found to be beneficial for bacterial growth and PHA production. All PHB and PHB-co-HV samples (Table 7) were collected from several independent fermentations described above. These

fermentations differed in the time at which the fermentation was supplemented with levulinic acid (72 h, 96 h and 108 h, instead of 120 h shown in Figure 19).

3.5.2 Crystallization temperatures of PHAs

Crystallization temperatures (Tc) measured by differential scanning calorimetry

indicated that the homopolymer PHB, without nucleating agent, exhibited the lowest Tc, 53.5 oC (Figure 20). However, crystallization temperatures for all copolymers were between 64.8 oC and 88.7 oC, at least 11 oC higher than PHB and the copolymers PHB-co30.5 mol% HV and PHB-co-32.6 mol% HV exhibit the highest Tc, over 80 oC (Figure 20). All copolymers crystallized at a higher temperature than the homopolymer PHB. Supercooling refers to the difference between the melting temperature and the temperature (Tco) at which crystallization of the polymer begins. Tcos of PHB and various copolymers of PHB-co-HV were always 17-20 oC higher than their corresponding peak crystallization temperatures (Tc) (data not shown). Copolymers also initiated crystallization at a higher temperature and exhibited a lower degree of supercooling than PHB, which allowed molten PHB-co-HV to crystallize within a shorter time frame than PHB.

74

100 90 80 70 Tc (C) 60 50 40 30 20 0 5.6% 11.4% 14.7% 17.9% 20.0% 30.5% 32.6% mol% HV in PHB-co-HV copolymers

Figure 20 Crystallization temperatures (Tc) of PHB () and copolymers of PHB-co-HV () with increasing mol% HV

3.5.3 Melting temperatures of PHAs

Melting temperature (Tm) is one of the most commonly reported properties of

PHAs. PHB exhibited the highest Tm (175.4 oC), compared to copolymers of PHB-co-HV (Figure 21). As shown in Figure 20, PHB exhibited the lowest Tc. Because PHB exhibited the greatest temperature differential between Tm and Tc, PHB will require the longest time to crystallize and solidify after melt-related processing. PHB-co-HV copolymers exhibited lower melting temperatures than the PHB homopolymer and when the ratio of HV in the copolymer increased from 0 to 33%, the Tm of copolymers gradually decreased from 175.4 oC to 168.5 oC, and subsequently increased to 171.7 oC. The lowest Tm, 168.5 oC, occurred for PHB-co-HV with 20% HV (Figure 21). The initial
75

decrease followed by an increase in the Tm, dependent on changes in the mol% of one of the monomer units, is a typical isodimorphic phenomenon for the copolymer PHB-coHV; however, the Tms did not change as much as other reports33, 121 which demonstrated a decrease in Tm from 179 oC for PHB to 84 oC for PHB-co-30%HV. Overall, the melting temperatures of copolymers in this study were lower than that of PHB.
176 174 172 Tm (C) 170 168 166 164 0 5.6% 11.4% 14.7% 17.9% 20.0% 30.5% 32.6% mol% HV in PHB-co-HV copolymers

Figure 21 Melting temperatures (Tm) of PHB () and copolymers of PHB-co-HV () with increasing mol% HV

76

3.5.4 Decomposition temperatures of PHAs

300 250 Temperature (C) 200 150 100 50 0 0 11.4% 17.9% 20.0% 30.5% mol% HV in PHB-co-HV copolymers 93.66 103.19 75.62 74.67

109.81

Representing the PHB homopolymer

Figure 22 Decomposition temperatures (Tdecomp, ) and melting temperatures (Tm, ) of PHB and copolymers of PHB-co-HV with increasing mol% HV. Bars indicate temperature differential (Tdecomp - Tm)

Decomposition temperature (Tdecomp.) was measured by thermogravimetric analysis and in this study was recorded as the point of initial weight loss. The bars shown in Figure 22 are the temperature differentials (Tdecomp-Tm) for various PHAs. These results show larger temperature differentials when the HV ratio was lower than 11.4% and higher than 20%. The Mn and Mw of PHB produced by B. cepacia using glycerol as a sole carbon source were lower than PHB produced from sugars as carbon sources, which is due to end capping of PHB with glycerol63, 120. The molecular masses of copolymers further
77

decreased by approximately 50% as compared to the PHB homopolymer (Table 7). When temperatures are raised above the Tm, polymers begin to melt with higher temperatures increasing the rate of polymer melting. For this reason, the practical processing temperature to melt polymers for injection molding is much higher than the theoretical Tm of polymers in order to reduce residence times during polymer processing. Great consideration was given to the Tdecomp because the loss of favorable mechanical properties of the polymers with low molecular masses will be more dramatic than polymers with high molecular masses. Therefore, higher Tdecomp of the polymers will be favorable because the larger the differential between Tdecomp and Tm offers more options for heating. In this study, copolymers with greater than 20% HV and the homopolymer PHB exhibited a higher differential (Tdecomp- Tm).

3.5.5 Decomposition temperature of PHAs with nucleating agents

In this study, HyperformHPN-68L, a heptane dicarboxylic acid derivative, and

ULTRATALC609 were used as nucleating agents to increase the crystallization temperature of PHAs. HPN-68L and talc have been known to elevate the onset temperature of crystallization, accelerate the crystallization rate, and function as preferred nucleating agents for polypropylene126, 127. However, the addition of 1% HPN68L to PHB dramatically decreased the Tdecomp. to 220.5 oC, compared to 269.1 oC for pure PHB and 273.2 oC for PHB with 5% ULTRATALC609 (Figure 23). The reduction in

78

the thermal window between Tdecomp. and Tm resulted in a narrower range for meltrelated processing. HPN-68L also showed a similar temperature reduction for PHB-co-HV (data not shown). Therefore, ULTRATALC609 was chosen as the sole nucleating agent for PHAs in this study to raise the crystallization temperature.

Figure 23 Effects of nucleating agents on decomposition temperatures of PHB (), PHB with 5% ULTRATALC609 (), and PHB with 1% HPN-68L ().

3.5.6 Melting temperature and crystallization temperature of PHAs with ULTRATALC609

The addition of ULTRATALC609 as a nucleating agent for PHAs had little effect on

the melting temperature of PHAs (Figure 24). The melting temperatures for the

79

homopolymer PHB and the copolymers of PHB-co-HV with 5% ULTRATALC609 were between 170 oC and 176 oC. However, the use of ULTRATALC609 did alter the crystallization temperature of PHAs. Bars in Figure 25 refer to the crystallization temperature increase caused by ULTRATALC609. For the homopolymer PHB, ULTRATALC609 increased the Tc from

Figure 24 Effects of 5% ULTRATALC609 on melting temperature (Tm) of PHAs containing increasing mol% HV (--PHB, --PHB-co-5.6 mol% HV, --PHB-co-11.4 mol% HV, --PHB-co-14.7 mol% HV, --PHB-co-17.9 mol% HV, --PHB-co-30.5 mol% HV, -PHB-co-32.6 mol% HV)

80

53.5 oC to 67.5 oC (Figure 25). The greatest increase for Tc occurred with PHB-co-20%HV which increased the Tc 30.2 oC. However, there was no significant change for Tc when the ratio of HV was greater than 30%.

120 100 Temperature (C) 80 60 40 20 0 0 5.6% 11.4% 14.7% 17.9% 20.0% mol% HV in the PHB-co-HV copolymer 14.04 24.48 29.67 17.29 10.95 30.17

1.44 30.5%

0.87 32.6%

Figure 25 Crystallization temperatures of PHAs () and PHAs with 5% ULTRATALC609 (). Bars indicate temperature differentials.

81

3.6 Effects of different aging periods on melting temperatures of PHAs

Due to the slow crystallization process of PHA-type polymers, it is possible that

different aging periods might change physical and mechanical properties of PHAs. In this research, PHA samples were isolated and purified from biomass, followed by vacuumdrying at room temperature for different periods, 1 day or 2 weeks. PHA samples were subject to be tested by DSC right after 1 day or 2 weeks. As shown in Table 8, each PHA sample exhibited similar melting temperatures and no significant change occurred whether the sample was aged for 1 day or 2 weeks.

Table 8 Comparison of melting temperatures (Tm) of PHAs detected at different aging times. Tm Tm

PHAs PHB PHB-co-0.43 mol% HV PHB-co-17.6 mol% HV PHB-co-29.5 mol% HV PHB-co-33 mol% HV PHB-co-35.8 mol% HV

1 d* 175.2 178.0 177.5 175.2 165.7 170.8

2 w# 179.3 179.2 176.2 172.8 163.6 168.3

PHAs PHB-co-11.4 mol% PHB-co-17.9 mol% HV PHB-co-20.0 mol% HV PHB-co-30.5 mol% HV PHB-co-32.6 mol% HV

1 d* 174.0 174.9 167.7 169.7 171.1

2 w# 174.2 171.7 168.5 169.9 171.7

Note: * and # represent the aging period for 1 day and 2 weeks, respectively.

82

3.7 Production of P3HB-co-4HB by B. cepacia using butyrolactone/1,4-butanediol.

Besides the P3HB-co-3HV copolymer, novel copolymers were explored to be

potentially produced by B. cepacia when proper cosubstrates were provided. Two cosubstrates, -butyrolactone (GBL) and 1,4-butanediol, were tested for the production of 4-hydroxybutyrate (4HB) monomer in the P3HB-co-4HB copolymer due to similar chemical structures between 4HB and GBL/1,4-butanediol. As shown in Figure 26, chemical shifts at 1.3 ppm, 2.5 ppm and 5.3 ppm indicated protons for the methyl group, methylene group and methane group of PHB, respectively. However, resonances at 1.9 ppm, 2.35 ppm (see the inset in Figure 26) and 4.1 ppm indicated that 4HB is present in the P3HB-co-4HB copolymer, and their corresponding multiplet, triplet and triplet strong supported the theoretical chemical structure of 4HB. In addition, these three peaks of 4HB were definitely not from the protons of residual 1,4-butanediol, which might exist in the purified polymer, because, in theory, the symmetrical structure of 1,4butanediol should only yield two peaks. When using -butyrolactone instead of 1,4butanediol as a cosubstrate, the purified polymer also exhibited a similar 1H NMR spectrum (data not shown). Furthermore, chemical shifts of P3HB-co-4HB matched those reports published by Dois group8, 20.

83

4 8 6 5 7

PHA-xylose+0.1% 1,4-butanediol

PHA-xylose+0.1% 1,4-butanediol

2.95

2.90

2.85

2.80

2.75

2.70

2.65

2.60

2.55

2.50

2.45

2.40

2.35

2.30

2.25

2.20

2.15

2.10

ppm

2 8

96.955

3.045

5.5

5.0

4.5

4.0

3.5

3.0

2.5
96.955 3.045

2.0

1.5

1.0

0.5

ppm

Figure 26 1H NMR spectrum of P3HB-co-3 mol% 4HB produced from 1,4-butanediol. Peaks 2,3,4 are typical chemical shifts for P3HB and peaks 6,7,8 (highlighted by the triangles) represent protons of P4HB.

However, the 4HB mol fraction in the P3HB-co-4HB copolymer did not increase while higher concentrations of cosubstrates were added to the medium. As a result, this copolymer produced by B. cepacia always contained less than 5 mol% 4HB, which may not be sufficient to effectively change material properties.

84

3.8 Confirmation of HV mol fraction in the PHB-co-HV copolymers by GC analysis and 1H-NMR.
NMR. HV mol% in PHB-co-HV copolymers detected by GC 0.43 5.6 11.4 14.7 17.6 17.9 detected by NMR 0.84 6.0 10.8 12.2 19.2 13.8 detected by GC 20 29.5 30.5 32.6 33.0 35.8

Table 9 Comparison of HV mol% in the PHB-co-HV copolymers detected by GC and 1H

detected by NMR 23.5 27.4 30.7 34.4 32.8 37.5

The composition and HV mol fraction of PHB-co-HV copolymers identified in this entire thesis were derived from gas chromatography (GC) analysis. Further confirmation of HV mol fraction in the copolymer was performed by 1H NMR, by the integration and normalization of methyl groups of HB and HV (for details see Figure 16a and section 2.6.4 in Materials and Methods). As shown in Table 9, some variation of the mol% HV for each copolymer sample detected by GC and 1H NMR indeed existed, differing from 0.2 mol% up to 4.1 mol%. The differential might result from accuracy and sensitivity of the equipment used, and
85

also from the integration of the peaks. These minor differences also demonstrated that it is feasible to use GC for compositional determination. Meanwhile, GC analysis is a time-saving process and exhibits greater convenience to analyze the known constituents of PHAs. Thus, the GC-derived HV mol fractions were used in this thesis for all PHB-coHV copolymers.

3.9 Solvent extraction of PHAs

Although the last but not least PHA extraction process by solvents seems a simple

and straightforward process, some details for the PHA isolation have to be explained here in terms of saving solvents, energy and time, especially for a scale-up pilot plant fermentation. In this whole session, chloroform was employed for PHA extraction, although dichloromethane also exhibited similar efficiency for PHA extraction.

3.9.1 Determination of the volume of chloroform for maximum extraction efficiency

In order to save solvents and reduce the environmental impact, different ratios of

chloroform used (ml) and the dry biomass (g) were tested to find an optimal point, at which the least volume of chloroform was used for maximum extraction efficiency. When using 5 ml or 2.5 ml chloroform to extract PHA from 1 g dry biomass at 55 C overnight, 6.9% or 12.3% PHA of the dry biomass still remained with biomass,

86

respectively (Figure 27). In other words, 6.9% or 12.3% PHA was lost after the extraction. As more chloroform was added for extraction from the ratio of 7.5 to 20, less PHAs, from 5.1% to 5.8%, were lost. However, increasing the dosage of chloroform from 7.5 to 20 did not significantly improve the extraction efficiency (on average of 5.4% PHA was still lost). Therefore, in order to save chloroform usage, it is reasonable to keep the dosage of chloroform between 7.5 to 10 ml/g DCM.

14 PHA% remaining in dry biomass 12 10 8 6 4 2 0 6.93 5.77 5.14 5.27 5.43 12.34

20

15

10

7.5

2.5

Ratio of chloroform (ml) / biomass (g)

Figure 27

Relationship between PHA extraction efficiency and dosage of chloroform.

Incubation of the mixture at 55 C overnight.

87

3.9.2 Determination of incubation temperature for maximum extraction efficiency

PHA extraction in the previous section occurred at 55 C when the mixture of

biomass and chloroform was incubated in an oven. This heating process could be completely avoided due to the high extraction efficiency at room temperature when stirring the mixture. As shown in Figure 28, when the mixture was stirred at room temperature overnight, 6.4% PHA still remained with biomass, which was comparable to 5.9% PHA loss when the stationary mixture was incubated at 55 C overnight. Therefore, in order to save energy, heating could be substituted by incubation at room temperature.

8 7 PHA% remaining in dry biomass 6 5 4 3 2 1 0 6.44% 5.85%

Stirring at room temp.

Still at 55 C

Figure 28 Relationship between PHA extraction efficiency and incubation temperature. Incubation of the mixture at room temperature when stirring or 55 C while standstill.
88

3.9.3 Determination of the incubation period for maximum extraction efficiency

6 PHA% remaining in biomass 5 4 3 2 1 0 5.23 4.54 5.3 4.67

6 Incubation time (h)

12

Figure 29 Relationship between PHA extraction efficiency and incubation period. Incubation of the mixture at room temperature when stirring.

Instead of heating, incubation of the mixture of chloroform and biomass occurred at room temperature for PHA extraction. The extraction efficiency was tested for incubation periods at 3 h, 6 h, 8 h and 12 h, respectively. 3-h incubation only resulted in 5.2% PHA loss, and longer incubation did not substantially increase the extraction efficiency (Figure 29).

89

Therefore, consideration of extraction efficiency and saving solvent, energy and time led to the conclusion that it is possible to extract PHA polymers at room temperature for 3 h incubation when using the least dosage of solvents, 7.5 to 10 ml chloroform/g DCM. In order to extract more PHA polymers from biomass, the second extraction was employed, by which most of PHAs was collected from biomass (over 99%, data not shown).

90

4. Discussion 4.1 Renewable and inexpensive feedstocks for PHA production

Environmental concerns about disposal of conventional petroleum-derived

commodity plasticss have resulted in increased attention to biodegradable plastics, e.g. PHAs, which utilize renewable sources of carbon and have a carbon neutral life cycle128. In terms of environmental sustainability and the focus on green technology, it is advantageous to produce these environmentally benign PHAs in large quantity to substitute for conventional plastics, which are recalcitrant for biodegradation in the natural environment. Two biological production systems, plants and microorganisms, have been investigated and employed for PHA production. The advantage of recombinant plants transformed with PHA synthetic genes is their ability to utilize photosynthetic pathways to directly convert carbon dioxide and water to sugars to support both growth and PHA production, without any capital expenditure for carbon sources. However, low PHA yield (4% PHB of plant fresh weight129, the highest level of PHA production in plants to date) in recombinant plants makes this system currently impractical in the near future to commercialize PHAs using plants as hosts19. In contrast to plants, microorganisms have been widely and almost exclusively used for PHA production using the laboratory- and industrial-scale fermentations; however, external carbon sources must be provided for basic microbial metabolism and PHA production.

91

However, prohibitively high production costs using microbial fermentation for PHA production greatly hinder commercial-level applications of PHAs, especially compared to the relatively low price of conventional commodity plastics. In fact, carbon sources may account for as much as 50% of the entire production cost of PHAs89, 90. Comprehensive consideration of environmental and economic concerns drives production of PHAs towards utilization of alternative feedstocks, which are both renewable and inexpensive. From the standpoint of reducing the environmental burden and reduction of production cost, the research described in this thesis utilized B. cepacia for bacterial growth and PHA production using renewable agricultural and forest residues as carbon sources. Wood hydrolysate from renewable lignocellulosic materials contains 5-carbon sugars (primarily xylose) and 6-carbon sugars (primarily glucose). These inexpensive pentose and hexose sugars were tested and determined to be acceptable carbon sources to support bacterial growth and PHA production by B. cepacia. Keenan130 reported that detoxified wood hydrolysate from aspen and maple could be used as a carbon source for PHA production. NREL CF (National Renewable Energy Laboratory Clean FractionationTM) hydrolysate from aspen supported bacterial growth for dry cell biomass (DCM) at 5.1 g/L and PHA yield at 40% (w/w) at 112 h, which were comparable to DCM at 5.3 g/L and PHA yield at 51% (w/w) at 75 h using pure xylose as a carbon source (similar to the results shown in Figure 7). However, detoxified ESF-derived maple hydrolysate exhibited some inhibition effect and led to lower DCM and PHA yield at 3.3

92

g/L and 18% (w/w) at 65 h, respectively. In this research, ESF-derived wood hydrolysate was examined for additional detoxification by overliming (adjusting pH to 10 using calcium hydroxide) and low-temperature sterilization in order to improve PHA production. However, no apparent differences in DCM and PHA yield were observed compared to Keenans results130. Dilution of this ESF hydrolysate at a ratio of at least 1:2 produced a less toxic hydrolysate, based on higher DCM, and may be a better solution than those above-mentioned detoxification processes (data not shown). During the preparation of wood hydrolysate, wood chips were cooked at high temperature and high pressure to hydrolyze polysaccharides and release sugars, as well as furan, furfural, furfuraldehydes, hydroxymethylfurfural, phenolics, etc., which are undesirable for microbial fermentation due to their toxicities, even at low concentrations131, 132. NREL CF hydrolysate seemed more fermentable by B. cepacia based on bacterial growth and PHA production, probably due to fewer toxic compounds in the hydrolysate produced by the NREL patented cooking technique. Careful control of cooking conditions of the hydrolysate is more important and a better strategy than detoxification processes, which can be complicated, economically challenging and timeconsuming. Tall oil fatty acids (TOFAs), comprised of 52% oleic acid (9 -C18) and 45% linoleic acid (9,12-C18), are produced as a byproduct from the paper pulping process, using renewable lignocellulosic material as the feedstocks. These long chain free fatty acids could be potential carbon sources for bacterial growth and PHA production because

93

various fatty acids, e.g. lauric acid, octanoic acid, undecenoic acid, oleic acid, palmitic acid, and estearic acid, have been shown to produce PHAs by several species of Pseudomonas133-135, only producing mcl-PHAs. B. cepacia was employed in this research project and tested for utilization of these TOFAs. The maximum amount of dry biomass (3.8 g/L) occurred at 96 h and the highest PHA yield of 49.9% at 72 h (Figure 7). By GC analysis and 1H NMR, the isolated polymer from B. cepacia only consisted of scl-PHB (C4) homopolymer (unpublished data), in contrast to mcl-PHAs (C6, C8, C10 and C12) by Pseudomonas68. These data provided useful information concerning the PHA synthase of B. cepacia, which only functionally incorporates C4 or C5 scl monomers into the polymer and may belong to Class I PHA synthases136, 137. In addition, larger molecules, such as 3phenylpropionic acid, 4-phenylbutyric acid, 5-phenylvaleric acid and 6-phenylhexanoic acid, seemed not to be incorporated into the copolymer by B. cepacia for synthesis of aromatic PHAs (based on 300 MHz 1H and
13

C NMR, data not shown). In contrast,

Pseudomonas synthesized poly-3-hydroxy-5-phenylvalerate, poly-3-hydroxy-6-phenylhexanoate, poly-3-hydroxy-7-phenylheptanoate, etc., when using corresponding substrates with similar chemical structures138-142. B. cepacia was also reported to directly use crude palm oil for PHB production68. It is conclusive that native B. cepacia can utilize those long chain fatty acids to support bacterial growth and PHA production, yet only scl-PHAs. Crude glycerol, a byproduct of the biodiesel fuel process, was produced in 2007 in the United States at a quantity of 45 million gallons, which was approximately 10% (v/v) of the total biodiesel output63. Such a large amount of glycerol needs an outlet to be
94

disposed of at low cost and, if possible, may contribute extra profits for revenue compensation of those biodiesel-producing companies if a value-added product from glycerol can be identified. B. cepacia is able to use various qualities of glycerol, from a high purity of 99% to a low purity of 40%, as a carbon source for bacterial growth, and exhibited high efficiency in the metabolism of glycerol for PHA production. Compared to the PHA yield at 40% and 49.9% when using wood hydrolysate and tall oil fatty acids as carbon sources, respectively, PHA yields were much higher and reached higher than 82% when using either pure or crude glycerol (Figure 8). Although ESF-derived biodiesel glycerol exhibited the lowest purity, 40% (v/v), it still supported bacterial growth well at 96 h for 4.98 g DCM/L and PHA production at 82% of DCM, which are comparable to data generated using high purity glycerol as a carbon source. Highly efficient metabolism of glycerol by B. cepacia and subsequent biosynthesis of PHA from those glycerol-based metabolites offers a means of utilizing crude glycerol as a renewable and inexpensive feedstock for production of value-added PHAs, and for disposal of low-value crude glycerol through an environmentally friendly fermentation process. Several microorganisms were reported to use glycerol as a carbon source for PHA production. Ashby and his coworkers143 employed Pseudomonas oleovorans NRRL B1468 to utilize crude glycerol for PHA production, and found that a maximum of 3.0 g/L and 38% occurred at 72 h for dry biomass and PHB yield, respectively. This same team

95

also reported that Pseudomonas corrugata could utilize glycerol for bacterial growth of 3.4 g DCM/L and mcl-PHA production of 19.7% at 72 h120. Recombinant E. coli transformed with PHA biosynthetic genes was used to produce PHAs from glycerol. Almeida et al
144

reported that E. coli with phaBAC from

Azotobacter sp. strain FA8 was grown on 3% glycerol and produced 5.6 g/L and 8.0 g/L dry biomass and 9.8% and 38.2% PHB in shake flasks and a bioreactor at 48 h, respectively. Nikel et al 145 demonstrated that a recombinant E. coli arcA mutant could produce 21.1 g/L dry biomass and 51% PHB at 60 h by a fed batch process under microaerobic conditions in a bioreactor. Mothes et al
146

used several different sources of biodiesel-derived glycerol as

carbon sources by Paracoccus denitrificans and Cupriavidus necator JMP 134, which exhibited an accumulation of PHB to 70% of dry biomass when using pure glycerol. However, crude glycerol including 5.5% NaCl led to a reduced PHB content of 48% at a high cell density of 50 g DCM/L. Compared to the aforementioned microorganisms utilizing glycerol for PHA production, B. cepacia displayed comparable growth (5-6.5 g DCM/L in flasks, 25.8 g/L in a 200 L fermentation in this thesis and 40 g/L under optimized conditions in a 7-L fermentor, unpublished data), and high PHA yield of higher than 82%, which is significantly higher than that described in the above-cited literature. Therefore, B. cepacia is a good choice to use various renewable and inexpensive feedstocks, such as tall oils (long chain fatty acids), plant oil, pentose (xylose) and
96

hexose (glucose, galactose) sugars from wood hydrolysate, and biodiesel-derived glycerol, for the production of scl-PHAs. Among these carbon sources, glycerol seems a better carbon source for PHA production in terms of the high yield of PHB. In addition, some common renewable and inexpensive feedstocks are summarized in Table 10. These potential feedstocks, including agricultural and forest residues, dairy food waste, animal fats, etc., could be used as carbon sources for PHA production on a commercial scale to greatly reduce production costs and make commercial PHAs competitive to conventional commodity plastics in the near future. The challenge to use these feedstocks is to find appropriate microorganisms with rapid growth and high yield of PHAs, and meanwhile, the impurities in the feedstocks (e.g salts in crude glycerol, furfural-related and aromatic compounds in wood hydrolysate) might have inhibitory effects on microbial metabolism, which should be taken into consideration carefully when choosing these renewable carbon sources for commercial scale production.

97

Table 10 Renewable and inexpensive feedstocks used for PHA production DCM (g/L) 4.4 91.3 2 14 10.7 3 3.3 11 21.2 5.2 PHA yield 55% 68% 34% 24% 57% 50% 62% 50% 76% 43%

Microorganisms Synechococcus sp. MA19 Ralstonia eutropha Azotobacter vinelandii UWD Azotobacter vinelandii UWD Ralstonia eutropha Bacillus megaterium Methylobacterium sp. ZP24 Unidentified strain Bacillus cereus Saccharophagus degradans ATCC 43961 Burkholderia cepacia

Feedstocks H2, CO2 H2, CO2 Wastewater Beet molasses Sugarcane molasses Date syrup Cheese whey Cheese whey Biodiesel glycerol Blue algae Tequila bagasse Wood hydrolysate, levulinic aicd Biodieselderived glycerol Glycerol, levulinic acid Tallow triacylglycerol Soybean oil Soybean oil

Type of PHAs PHB PHB PHB-co-7.9% HV PHB P3HB-co-3HV-co4HV PHB PHB PHB-co-8~10% HV unspecified PHB PHB-co-HV PHB PHB-co-HV mcl-PHAs PHB PHB-co-5% HHx

Ref.
147 148 149 150 123 151 93

152 153 154

5 5.8 ~6 0.9 ~120 ~130

40% 82% 50% 15% 72%

58

Burkholderia cepacia

This thesis

Pseudomonas resinovorans Ralstonia eutropha Recombinant Ralstonia eutropha with phaC from Aeromonas caviae

155

62

71%

Note: the major constituents available for microbial fermentation in molasses, wood hydrolysate and cheese whey are sucrose, xylose and lactose, respectively.

98

4.2 Bacterial growth and Properties of PHB produced from glycerol as a carbon source 4.2.1 Bacterial growth of B. cepacia using glycerol as a carbon source

When biodiesel-glycerol (from Twin River Technologies) was used as the sole

carbon source in shake flasks, B. cepacia produced 5.8 g/L dry biomass and contained up to 81.9% PHB of total dry biomass when using 3% (v/v) biodiesel-glycerol as a feedstock over 96 h of growth (Figure 8). B. cepacia showed higher dry cell weights and PHA production when grown on glycerol when compared to E. coli and various pseudomonads grown on glycerol (Table 11). However, increasing glycerol concentration from 3% to 9% (Figure 10) resulted in a decrease of microbial biomass of B. cepacia by almost 50%. In a recent study by Cavalheiro et al.156, similar effects of increasing concentrations of glycerol (on microbial growth) were observed. Specifically, at glycerol concentrations exceeding 30 g.L-1, the specific growth rate of Cupriavidus necator decreased when using either pure glycerol or biodiesel-glycerol as sole sources of carbon. In a recent publication by Cavalheiro et al.156, experiments were performed to establish the effects of gas composition and rate of addition to fed batch cultures of Cupriavidus necator. The authors demonstrated significant increases in dry cellular biomass (g L-1) by increasing the flow rate from 1.5 to 3.0 L of air min.-1. Additionally, exponential phase growth and final cell densities were achieved sooner by aeration with 2.0 L air min-1 supplemented with 1.0 L pure oxygen min-1. Preliminary experiments for this work established that initially 1 VVM (volume of air per volume of medium per
99

minute) was sufficient to maintain maximum growth. As the fermentation progressed, whenever the dissolved oxygen fell below 35%, the impeller speed and aeration (data not shown) were increased to maintain dissolved oxygen between 35-50%. The latter was found to be optimal for cell growth and PHA production by Burkholderia cepacia and this approach was used in all fermentor experiments. Table 11 Comparison of dry cell mass and PHA content from different bacterial strains grown on various carbon sources in shake flasks. Bacteria Substrates Polymer DCM (g/L) B. cepacia B. cepacia E. coli E. coli E. coli P. corrugata P.oleovorans Glycerol Xylose, Levulinic acid Glucose Glycerol Glycerol Glycerol (2%) Glycerol (1%) PHB PHB-co-PHV PHB PHB PHB PHB mcl-PHA 5.8 4.4-5.3 4.9 5.6 3.6 3.4 1.9 PHA content (%) 81.9 42-56 27.6 9.8 34 19.7 26.8 This study
115

Ref.

144 144 157 120 120

4.2.2 Properties of PHB produced from glycerol as a carbon source

The molecular mass of PHB varies from 50 to 3,000 kDa1, 2, depending on the

microorganism and growth conditions. Although the pathways of PHA synthesis using different carbon sources have been previously described1,
100
158

, the mechanism for

regulating molecular mass in PHA production is still unclear. The activity of PHA synthase (moles of substrate converted per unit time) may influence the molecular mass and the polydispersity of the polymer. It has been shown previously that PHA synthases with higher activity produced lower molecular weight polyesters159-161. On the other hand, enhancing the specific activity of PHA synthase in a phaC mutant increased not only PHA accumulation, but also weight-average molecular weight by 6% to 74%162. Thus, PHA synthase is not the only factor to control and regulate the molecular weight of PHAs, and other enzymes or compounds may also play an important role in chain termination of the polymer. In addition to providing a carbon source, glycerol has a unique function as a terminal end group for PHB synthesis. Both Ralstonia eutropha (formerly Alcaligenes eutrophus) and Pseudomonas oleovorans are known to produce short-chain-length PHAs; however, when using glycerol as the sole carbon source, the molecular mass of PHB was substantially lower than that produced from glucose and whey sugars (glucose and galactose from lactose hydrolysis)152,
163, 164

. The molecular weight gradually

decreased as glycerol concentration increased120. In this study, when glycerol content was increased from 3% (v/v) to 9% (v/v), both the molecular weight of PHB and dry cell weight decreased. One possible explanation for this phenomenon is that the organism was inhibited by osmotic stress as the glycerol content increased. The cells began to decrease their enzymatic efficiency and undergo chain termination earlier. PHB was capped by glycerol through covalent esterification in a chain termination position, resulting in lower molecular weights of the polymers produced. As the cells were in
101

contact with glycerol for longer time periods, there was a higher possibility for chain termination to form the lower molecular mass polymers. The osmotic stress may also have caused a slower growth rate and explains the decreasing cell dry weight. Interestingly, there has been no evidence to show that medium-chain-length PHA (produced from Pseudomonas corrugata) polymers can be esterified with glycerol to form an end-capped polymer120. Perhaps the PHA synthase of strains producing medium-chain-length PHAs (monomers greater than 6 carbons) cannot incorporate smaller molecules such as glycerol. However, the PHA synthase of B.cepacia may more readily incorporate glycerol as a terminal group since the enzyme may exhibit a specificity for shorter chain length substrates. The glycerol-based PHB exhibited similar Tg and Tm compared with PHB polymers produced in cells grown on xylose or compared with commercially available polypropylene115. Tdecomp., however, was 13 oC higher in glycerol-capped PHB compared to xylose-based PHB. This advantage could provide a broader range for temperature exposure in certain industrial applications. Although the thermal properties of glycerol end-capped PHB in this study were not markedly different from PHB without glycerol end-capping, the molecular weights of the polymers were significantly lower. The lower molecular weights of the end-capped polymers may affect the tensile strength of the polymers165. On the other hand, the diol of these end caps provides an interesting pair of functional groups (-OH) through which

102

these polymers have the potential to be further chemically modified to produce novel PHA derivatives with exceedingly high molecular masses .

4.3 Production and characterization of the PHB-co-HV copolymers produced from biodiesel-derived glycerol and levulinic acid 4.3.1 Production of PHB-co-HV copolymers using biodieselderived glycerol and levulinic acid

The homopolymer PHB and the copolymers of PHB-co-HV are the most widely

studied PHAs, especially for biosynthetic pathways, fermentation optimization, polymer properties and their biodegradation characteristics. It is well known that PHB exhibits highly crystalline and brittle characteristics and therefore has limited applications. However, the copolymer PHB-co-HV shows increased strength and better elongation to break as compared to PHB. Also, the mechanical properties of PHB-co-HV can be regulated by altering the ratio of HV in the copolymer115. The co-substrate, levulinic acid, was provided at different concentrations with glycerol as the primary carbon source, to synthesize PHB-co-HV copolymers with various mol% fractions of HV. However, levulinic acid is an inhibitor of bacterial growth when the concentration is increased beyond 0.1% (w/v) in the medium. Higher concentration of levulinic acid will suppress bacterial growth because the undissociated acid diffuses across the plasma membrane and lowers the intracellular pH in the cytosol132, 166, 167. However, optimization of nutrient concentration could lower the inhibitory effects of

103

levulinic acid. For the fermentation experiments in this study, levulinic acid was added at the optimal rate of 0.5 g/Lh (see Materials and Methods), which caused the least inhibition of bacterial growth and resulted in acceptable copolymer production (Figure 19). Further optimization of feeding strategy for glycerol (from 1% to 3% v/v) and nitrogen (between 0.14 g/Lh and 0.2 g/Lh, maximum 10 g/L) enhanced copolymer yield to 40.4% of the cell dry mass (data not shown).

4.3.2 Physical properties of the PHB-co-HV copolymers produced from biodiesel-derived glycerol and levulinic acid

Previous reports have shown that as the initial mol percentage of HV increases in

PHB-co-HV copolymers the melting temperature will gradually decrease to a minimum point (from 175 oC for PHB to 84 oC for PHB-co-30 mol% HV) and then increases as the mol percentage of HV increases further33, 121, typical of an isodimorphic pattern for a copolymer containing closely related monomeric structures. This is due to the statistically random copolymers where both monomers can crystallize and one type of monomer is incorporated into the crystal lattice of the other and vice versa. Doi et al.20,
32 o

observed this isodimorphic phenomenon for PHB-co-HV and the minimum value (75

C) occurred at approximately 40 mol% HV, where the crystal lattice transition from the

HB crystalline conformation to the HV crystalline conformation took place. PHB-co-HV produced by B. cepacia using xylose and levulinic acid as substrates was previously reported115 to exhibit a similar isodimorphic behavior; however, the minimum threshold

104

was close to 154 oC at 25 mol% HV115. Interestingly, when using ESF hemicellulosic hydrolysate instead of pure xylose for production of the PHB-co-HV copolymers by B. cepacia, the minimum Tm decreased to 80.8 C at 45 mol% HV130. In this study, PHB-co-HV produced by B. cepacia using glycerol and levulinic acid as substrates showed a minimum melting temperature at 168 oC for PHB-co-HV copolymer comprised of 20 mol% HV (Figure 21). PHB-co-HV produced by B. cepacia exhibited a higher minimum Tm with lower ratios of HV in this study. Although molecular size of PHA was thought to be a potential factor possibly affecting melting temperature of the PHBco-HV copolymers, there is no direct linear relationship between Tm and molecular mass, based on melting temperatures of the copolymers with high molecular mass (average Mv=687 kDa) produced from xylose and levulinic acid130 and with low molecular mass produced from glycerol and levulinic acid (average Mw=115 kDa, from Table 7). Further proof reported by Savenkova et al.168 demonstrated that the PHB-co-HV copolymers displayed a minimum Tm of 116 C at 20 mol% HV, yet average Mv of the copolymers reached a much higher level of 1426 kDa. And Ng et al.169 also reported that the copolymer with 22 mol% HV (Mn=240 kDa) exhibited a minimum Tm of 131 C. In summary, the copolymer with high molecular mass does not necessarily exhibit high melting temperature, and low molecular mass does not necessarily result in low melting temperature. Other potential factors for melting temperature are the aging process and thermal history of the PHAs. Due to the known slow crystallization rate of PHAs, it may take

105

several days up to several weeks for a PHA polymer to reach the equilibrium of crystallinity. Compared to the fresh PHA polymers which are isolated from biomass in a short time period, the aged PHA polymers (days or weeks or even several months) exhibited different material properties168. As shown in Table 8, the PHB homopolymer and the PHB-co-HV copolymers were aged for either 1 day or 2 weeks after PHA extraction at room temperature under vacuum-drying conditions. DSC tests for melting temperature were performed right after the specific aging period (1 day or 2 weeks). There were no substantial changes in melting temperatures (between 0.2 C and 4.1 C) for all 11 PHA samples and the lowest melting temperature for all these samples was still higher than 163 C. As reported by Keenan et al.115, when the PHB-co-HV copolymers were extensively aged for 60 days at room temperature, the lowest temperature was 154 C. The copolymers isolated from this study showed higher melting temperature than those in the above-mentioned reports. In addition, the thermal history of the polymers may also change thermal properties. The thermal history of the polymer refers to the process, during which the polymer is heated or cooled prior to physical property tests. All polymers shown in Table 8 were dissolved in chloroform and cast into films at room temperature, followed by vacuum-drying until use. As a result, these polymers should be in the same thermal condition. As mentioned above, the melting temperatures of the copolymers with the same thermal history did not change substantially. Therefore, it is still unclear which factors influence the dramatic change in Tm of the PHB-co-HV copolymers.

106

It is noteworthy that the PHB-co-HV copolymers exhibited slower crystallization rates than the PHB homopolymer and the copolymers with a high mol fraction of HV demonstrated much slower crystallization rates than the copolymers with a low mol fraction of HV
170

. Consequently, without the addition of heterogeneous nucleating

agents, it will take longer for the copolymers to crystallize and solidify after injectionmolding than the PHB homopolymer.

4.4 Mechanical properties of the PHB homopolymer and the PHB-co-HV copolymer

Mechanical properties of one material state this materials nature in terms of

toughness, ductility, etc., which determine whether this material is feasible to be used in a specific application field. The PHB homopolymer (derived from either xylose or glycerol as a carbon source) exhibited the highest Youngs Modulus among all polymers tested (Table 6), which means the least elasticity for PHB. The PHB-co-HV copolymers generally showed better elasticity than the PHB homopolymer, based on a lower value for Youngs Modulus. The elongation to break further demonstrated that the copolymers can generally withstand greater stretching than the homopolymer (Figure 18). The glycerol-derived PHB homopolymer and PHB-co-HV copolymers exhibited similar tensile strengths, meaning that they all had similar toughness, though the xylose-based PHB homopolymer was slightly tougher, based on a minimally higher tensile strength.
107

As shown in Table 3 and Figure 12, PHB produced from glycerol exhibited a lower molecular mass due to glycerol end-capping of PHB. Similar behaviors on reduction of molecular mass when using glycerol as a carbon source for PHB production were observed in several publications120,
152, 156, 163, 171

, which were discussed in detail in

section 4.2.2 of this thesis. The relatively low molecular mass of PHB produced from glycerol may result in less desirable mechanical properties, potentially leading to inferior quality of the finished products. This study demonstrated that glycerol-based PHB with low molecular mass (Mn=122 kDa) had a comparable Youngs modulus and elongation to break to those of xylose-based PHB with higher molecular weight (Mn=468 kDa), although glycerol-based PHB exhibited slightly lower tensile strength and yield strength than the xylose-based PHB (Table 6). Therefore, relatively low molecular mass might not adversely affect this glycerol-based PHB and could be used with the regular-sized PHB to substitute for conventional petroleum-derived plastics. Polypropylene, a typical example of petroleum-based plastics, was tested under similar conditions to PHA samples (Materials and Methods), and showed higher tensile strength and better elongation to break than PHA samples. However, as the mol fraction of HV increased to 17.6% of the copolymer, polypropylene and PHB-co-17.6 mol% HV exhibited similar elasticity, based on Youngs modulus, of 189 MPa and 228 MPa, respectively. The copolymers with higher HV fraction (>17.6 mol% HV) demonstrated much better elasticity (45-68 MPa of Youngs Modulus) than polypropylene (189 MPa).

108

Some thermal and mechanical techniques can be employed to improve the mechanical properties of PHAs. PHA films, which were drawn in silicone oil at 165 C by stretching and then annealed at 100 C for 2 h, exhibited higher ductility and flexibility based on higher tensile strength and elongation to break. Drawing and annealing resulted in PHA films with a highly ordered orientation of the crystal domain and generated a planar zigzag conformation consisting of extended molecular chains165, 172. The mechanical properties did not deteriorate for 60 days because the deterioration caused by secondary crystallization was avoided by a high degree of orientation165. Pure PHB is relatively brittle, partly due to formation of large spherulites of PHB crystals. Large spherulites tend to result in inter-spherulitic cracks. Addition of heterogeneous nucleating agents into PHB not only reduces the injection-molding processing cycle times, but also helps to increase the number of nuclei, leading to smaller average sizes of spherulites. Therefore, nucleating agents could improve the mechanical properties to some extent by increasing elongation to break by 5%173, 174. Meanwhile, plasticizers (acetyltriethylcitrate, acetyltributylcitrate, tributyrin, triacetin, glycerol, etc.) could enhance molecular motion, and thus increase elongation to break as well as impact strength36. As shown in Table 5, PHB homopolymer after being subjected to an injectionmolding process, was thermally degraded at high temperature (usually 20-30 C higher than the melting temperature). Weight loss of PHB was from 32% to 98% of the original PHB mass, which might cause additional deterioration in the quality of the finished

109

products made from the degraded PHB, correlating to less desirable mechanical properties of the thermally degraded PHB. By changing the conventional injectionmolding process to a reverse temperature injection-molding process, the plastic commodity products might exhibit better mechanical properties and qualities by much less weight loss of PHAs. Zhang et al.175 reported that the reverse temperature injectionmolding process largely reduced the thermal degradation of BiopolTM PHAs and therefore, the injection-molded PHAs showed better mechanical properties (higher tensile strength, lower Youngs modulus, better elongation to break and impact resistance) than PHAs processed through the conventional injection molding process. As a result, it is feasible to improve the quality of finished products by improving the engineering flowsheet for injection molding. It was noteworthy that the aging time for molded PHA samples would affect mechanical properties. Stiffness and tensile strength increased with storage time while elongation and impact strength decreased. The major aging process occurred in the first week after parts were molded, although mechanical properties slightly changed for up to 2 months175. Savenkova et al.
168

also found that the aging process of the PHB

homopolymer and the PHB-co-HV copolymers led to a less desirable elongation to break for 224 days. This unfavorable aging process, contributing to the embrittlement of PHB, might be ascribed to progressive crystallization that tightly constrains the amorphous phase of PHB between the crystals176, 177.

110

4.5 Effects of nucleating agents on physical and mechanical properties of PHAs

There are relatively few reports that demonstrate the effects of nucleating agents

on crystallization temperature, melting temperature and decomposition temperature of both PHB and PHB-co-HV. Nucleating agents which result in higher crystallization temperatures will benefit the overall injection molding process as it will require less time for molten PHAs to crystallize and solidify into finished products resulting in shorter cycling times for processing machinery. Usually, Tdecomp is approximately 75 oC to 100 oC higher than Tm of the copolymer. This greater temperature differential allows more flexibility to use higher temperatures during the polymer melt without thermal decomposition of the polymer. However, some nucleating agents decrease decomposition temperature of polymers. Hydroxyapatite, which is a naturally occurring mineral of calcium apatite, has almost no effect on Tm of PHB42, which is similar to our results (Figure 24). However, the onset Tdecomp decreased from 260 oC to 225 oC when the hydroxyapatite content increased from 0 to 10%42. HPN-68L also lowered the onset Tdecomp of both PHB (Figure 23) and PHB-co-HV by approximately 50 oC (data not shown). In this study, the use of ULTRATALC609 enhanced the thermal stability of PHB with a slightly higher Tdecomp (Figure 23). ULTRATALC609 exhibited negligible effects on Tm of both PHB and PHB-co-HV (Figure 21 and Figure 24) and the melting temperature of PHB was reported to be increased by less than 3 oC when -cyclodextrin (1% or 2%) or talc (2%) was used as a nucleating agent; however, crystallization temperatures were elevated by 18 oC, 27 oC
111

and 32 oC at 1% -cyclodextrin, 2% -cyclodextrin and 2% talc, respectively40. Also boron nitride (1%), talc (5%), terbium oxide (1%) and lanthanum oxide (1%) were tested as nucleating agents for PHB-co-6.6 mol% HV, which increased Tc by 25 oC, 21 oC, 11 oC and 10 oC, respectively 35.

4.6 Economic considerations of PHA production

High production costs of PHAs stem from two major sources, feedstock cost and

downstream processing. Concurrent cost reduction of feedstocks and extraction will lead to substantial reduction of the entire production costs of PHAs. Cost reduction of feedstocks could be achieved by using renewable and inexpensive feedstocks instead of pure and expensive carbon sources. Cost reduction of the PHA extraction process could be accomplished by using alternative PHA recovery methods, e.g microfluidization and sodium dedocyl sulfate (SDS). Once cost reduction from these two sources are implemented, commercial production of PHAs will be competitive to petroleum-derived plastics.

112

4.6.1 Economic considerations of PHA production from renewable and inexpensive feedstocks
The strategy for cost reduction must focus on using inexpensive and renewable feedstocks to replace expensive carbon sources.

Carbon sources may account for up to 50% of the entire production cost of PHAs.

Glycerol has been tested in this research to evaluate its use as a carbon source for bacterial growth and PHB production by B. cepacia. Instead of using pure and expensive glycerol as a carbon source, crude glycerol from the biodiesel-producing process has been produced in huge quantities every year in the United States, which results in the reduction in price of crude glycerol to as low as a few cents a pound. Crude glycerol (purity from 40% to 85%, v/v) has been shown to be a carbon source comparable to pure glycerol, based on dry biomass and PHA yield (Figure 8). Compared to xylose and tall oil fatty acids, glycerol as a carbon source supported higher biomass production (5-6 g DCM/L in flasks shown in Figure 8 and 40 g DCM/L in a 7-L fermentor, data not shown) and PHA yield (80%-90% in flasks and almost 50% in a 7-L fermentor, data not shown). Randall178 reported that B. cepacia produced 35.5 g DCM/L using xylose as a carbon source in a fermentor and a 52% PHA yield after optimization. In order to simplify the price estimation of PHAs, arbitrarily say no cost for the carbon source when using crude glycerol, though several cents per pound, the price of PHAs at $ 3-4/kg may decrease to approximately $1.5-2/kg, which is half of the current price, when assuming the cost of feedstocks accounts for 50% of the production cost.

113

Although this price is still 2 to 4 fold higher than the current price of polypropylene, additional cost reduction by cost-effective downstream processing may further reduce the price of PHAs closer to petroleum-derived plastics.

4.6.2 Economic considerations of PHA production by selective downstream isolation processes

In this thesis, PHA isolation was performed by solvent extraction. Although the

least volume of solvents and anti-solvents were used to isolate and purify PHAs, and most of the solvents were then recycled by a distillation system for the next extraction process, this process could still lose 15%-25% volume of the solvent during the entire extraction-distillation-purification process (data not shown). The solvent lost in such volume contributes to the prohibitively high production cost, even without consideration of other costs during the fermentation. Furthermore, these chlorinated solvents are not environmentally friendly chemicals. Release of these solvents can be potentially harmful to the environment and to human health, although PHAs isolated by these solvents is desirable due to their high purity. Alternative extraction processes for PHAs, which are low cost, highly efficient and environmentally benign, should be taken into account. Various techniques, such as chemical digestion and enzymatic and mechanical disruption, were tested for PHA isolation. Advantages and disadvantages of each method have already been discussed in the Introduction. The following will discuss the new isolation process for PHAs

114

developed in our laboratory using a microfluidizer for cell disruption and a surfactant (sodium dodecyl sulfate, SDS) to further lyse the cells and purify the PHA polymers. SDS will rupture cell membranes, but its efficiency is relatively low (30%-60%, based on the amount of protein released). When the slurry biomass concentration increased from 7 g DCM/L to 70 g DCM/L, protein release declined rapidly from 60% to 40%, even with increasing the SDS concentration from 0.5% to 10% (w/w)179. Because SDS did not release all available protein, further treatment became necessary. Usually sodium hypochlorite has been used to disrupt cells with SDS to a higher efficiency (higher than 80%). However, hypochlorite treatment required 24 h for cell disruption, compared to only 60 min for the SDS treatment179, and hypochlorite is known to degrade PHA polyesters, which is undesirable in terms of material properties of the intact PHA polymers. Microfluidization, a mechanical process instead of the chemical hypochlorite, was tested to lyse B. cepacia cells in our laboratory. Extremely high shear forces generated by this process will efficiently disrupt cells. Microfluidizer (M110L, Microfluidics International Corporation, Newton, MA) was used in our lab for cell rupture. The slurry biomass of B. cepacia at 35 g DCM/L was pumped through the chamber under a pressure of 3 psi three times, and most cells were lysed based on light microscopy. Furthermore, this is a time-saving process for cell disruption, considering that 200 L of fermentation broth generates approximately 27-L re-suspended cells takes around 30 minutes for complete cell lysis. Afterwards, SDS at a concentration of 5% (w/w) was

115

added to the broth and mixed with the lysed cells for 60 min. PHA pellets were collected by centrifugation and analyzed by GC to demonstrate their purity at higher than 90% (personal communication with Mr. Joseph A. Perrotta and Dr. James P. Nakas), and molecular masses of PHAs treated by both microfluidizer and SDS did not exhibit any apparent difference compared to those of PHAs through direct chloroform extraction. It is noteworthy that the microfluidization process was more efficient with high cell density broth and released over 90% of the available protein of the slurry biomass at 45 g DCM/L179. These characteristics make use of the microfluidizer a more efficient cell breakage process for PHA isolation from a high cell density fermentation than the bead mill process, which tends to break cells more efficiently at a relatively low cell density (17 g DCM/L)179. In addition to two publications from Metabolix Inc. (Cambridge, MA) describing that a microfluidizer was employed to break the cells for the broth by passage through this homogenizer180 or to form amorphous PHA-film polymer latex181 by a combination of mechanical and enzymatic treatments, few reports were found using a microfluidizer for downstream PHA isolation. It may be worthwhile to further consider research on mechanical cell disruption using a microfluidizer and, if necessary, combinations of this mechanical process with other suitable chemical or enzymatic processes for more efficient PHA isolation and purification. A rapid cell disruption process using microfluidization and efficient cell lysis by SDS may be low cost and highly efficient due to saving time and energy and eliminating

116

solvents, compared to the current lyophilization and solvent extraction process. There are many PHA recovery methods but no one procedure is a perfect solution with low cost, high efficiency and environmental compatibility. Therefore, one has to make compromises. Based on the end-use application requirements of PHAs (purity, recovery efficiency, molecular size, etc.), selection of proper PHA recovery methods or combinations of several methods for PHA isolation may be a reasonable solution for the PHA recovery conundrum.

117

5. Conclusions

In order to reduce the production cost of PHAs, renewable and inexpensive

feedstocks, such as wood hydrolysate, cheese whey permeate, tall oil fatty acids and biodiesel-derived glycerol, were evaluated as potential alternative carbon sources for bacterial growth and PHA production by B. cepacia. This study showed that xylose and tall oil fatty acids could support substantial growth of B. cepacia and PHA production, based on dry biomass of 4.9 g/L and 3.8 g/L at 96 h, and PHA yields of 51.0% and 49.9% of DCM at 72 h, respectively. When using glycerol as a carbon source, the highest biomass (5.8 g/L) and maximum PHA yield (81.9% of DCM) occurred at 96 h, which were significantly higher than those from xylose and tall oil fatty acids. Meanwhile, PHAs produced from these long chain tall oil fatty acids (C18) were identified as only the PHB homopolymer by GC and NMR. Therefore, PHA synthase of B. cepacia was deduced to only functionally synthesize scl-PHAs, although several species of Pseudomonas could utilize these long chain fatty acids for production of mcl-PHAs. These data indicated that it is feasible to use these three renewable and inexpensive feedstocks for PHA production. However, lactose and cheese whey permeate did not support substantial growth of B. cepacia or PHA production. Various sources (Twin River Technologies, Future Fuel and ESF) of biodiesel-derived glycerol from a high purity of 99% to a low purity of 40% were also tested for PHA production, and it was determined that all glycerol sources supported bacterial growth and PHA production, and the lowest biomass and PHA production from ESF biodiesel-derived glycerol (the lowest purity of

118

40%) still reached 5.0 g/L and 82%, respectively. Therefore, B. cepacia could efficiently utilize not only high purity glycerol, but also the crude glycerol at low purity for PHA production. When using glycerol as the carbon source, the concentration of glycerol needs to be strictly controlled. Increasing the glycerol concentration from 3% to 9% (v/v) resulted in a gradual reduction of biomass, PHB yield and molecular mass (Mn and Mw) of PHB.
1

H-NMR revealed that molecular masses decreased due to the esterification of PHB with

glycerol resulting in chain termination (end-capping). Therefore, by regulating the glycerol content, different lengths of PHB can be produced to meet the diverse criteria of various industrial and medical applications. Additionally, no significant change for material properties was observed for the glycerol-based PHB (low molecular mass) and the xylose-based PHB (high molecular mass), based on similar physical-chemical properties (Tm, Tg, Tdecomp) and mechanical properties (tensile strength, Youngs Modulus, elongation to break). This investigation here describes the regulation of molecular mass by end capping with glycerol by B.cepacia when using different concentrations of glycerol as a carbon source. Supplementation of glycerol with levulinic acid resulted in production of the PHBco-HV copolymers. Based on the concentration and timing of levulinic acid added to the medium, HV from 0.4 mol% to 35.8 mol% was incorporated to form various compositions of the PHB-co-HV copolymers. 1H and 13C NMR indicated that this polymer contained 3HB and 3HV monomeric units and 4HV was not present in this copolymer.

119

The copolymers exhibited a typical isodimorphic behavior (V-typed shape), by which melting temperature decreased from 175.4 C of PHB to a minimum point of 168.5 C of PHB-co-20 mol% HV and then increased as mol% HV increased further. Overall, the PHBco-HV copolymers exhibited more desirable melt and crystallization behavior since all copolymers had lower Tms and higher Tcs than the homopolymer PHB. The PHB-co-HV copolymers also exhibited better mechanical properties (higher elasticity) than the PHB homopolymer, and for the copolymer increasing mol% HV resulted in further improved elongation to break. However, levulinic acid at a concentration higher than 0.1% (w/v) in the medium resulted in inhibitory effects on bacterial growth of B. cepacia, which eventually increased production cost of the PHB-co-HV copolymers due to the relatively low biomass. Based on the consumption and conversion rate of levulinic acid into the HV monomer, the optimal feeding rate was set at 0.5 g/Lh, which exhibited the lowest degree of inhibition. By optimization of feeding strategy for glycerol (between 1% and 3%) and levulinic acid in a 7 L fermentor, dry biomass and PHA yield reached 40 g/L and 50% of DCM, respectively. Production of the PHB homopolymer and the PHB-co-HV copolymers was successfully scaled up from bench level to pilot-plant level fermentations (25.8 g DCM/L and 33% PHA of DCM). Large quantities of PHAs were then isolated, purified and subjected to injection molding for the fabrication of eartips, which are biodegradable

120

and environmentally compatible instead of using conventional petroleum-derived plastics. However, slow crystallization of PHAs impaired their commercial application due to longer injection molding processing. Addition of heterologous nucleating agents improved industrial processability of PHAs by increasing the crystallization temperature. However, HPN-68L was eliminated as a nucleating agent for the polyesters isolated in this study due to causing a decrease in decomposition temperature of PHAs, which is not desirable under high temperature heating during injection molding. Talc, a natural mineral, not only increased crystallization temperature, but also slightly enhanced decomposition temperature of PHAs. Downstream isolation processes for PHAs were performed using solvent extraction or newly developed mechanical and chemical treatments. These renewable and inexpensive carbon sources and alternative downstream PHA isolation processing may potentially reduce production cost of PHAs, by which the market price of PHAs might be competitive to that of conventional petroleum-derived plastics. PHB-co-20 mol% PHV was judged the best polymer tested in this study for injection molding purposes because it demonstrated high Tdecomp.and the lowest Tm. In addition, PHB-co-20 mol% HV was reported to exhibit greater toughness and better flexibility than PHB and PHB-co-10 mol% HV (20-fold and 5-fold higher for elongation at break, respectively) 182. Also, this study showed that the copolymer with 29.5 mol% HV exhibited dramatic improvement of mechanical properties as judged by Youngs
121

Modulus (3-fold decrease) and Elongation to break (28-fold increase), compared to the copolymer with 17.6 mol% HV. By examining the mechanical properties of various ratios of copolymers in the presence of a nucleating agent, ULTRATALC609, the copolymer PHB-co-HV consisting of approximately 20 mol% HV was determined to be more suitable for industrial processing.

122

6. References
1. Sudesh K, Abe H, Doi Y. Synthesis, structure and properties of polyhydroxyalkanoates: Biological polyesters. Prog Polym Sci. 2000; 25 (10): 1503-1555. 2. Madison LL, Huisman GW. Metabolic engineering of poly(3-hydroxyalkanoates): From DNA to plastic. Microbiol Mol Biol Rev. 1999; 63 (1): 21-53. 3. Suriyamongkol P, Weselake R, Narine S, Moloney M, Shah S. Biotechnological approaches for the production of polyhydroxyalkanoates in microorganisms and plants - a review. Biotechnol Adv. 2007; 25 (2): 148-175. 4. Valappil S, Boccaccini A, Bucke C, Roy I. Polyhydroxyalkanoates in gram-positive bacteria: Insights from the genera Bacillus and Streptomyces. Antonie van Leeuwenhoek. 2007; 91 (1): 117. 5. Lu J, Tappel RC, Nomura CT. Mini-review: Biosynthesis of poly(hydroxyalkanoates). Polym Rev. 2009; 49 (3): 226-248. 6. Anderson AJ, Dawes EA. Occurrence, metabolism, metabolic role, and industrial uses of bacterial polyhydroxyalkanoates. Microbiol Mol Biol Rev. 1990; 54 (4): 450-472. 7. Steinbuchel A, Fuchtenbusch B. Bacterial and other biological systems for polyester production. Trends Biotechnol. 1998; 16 (10): 419-427. 8. Doi Y, Kunioka M, Nakamura Y, Soga K. Nuclear magnetic resonance studies on unusual bacterial copolyesters of 3-hydroxybutyrate and 4-hydroxybutyrate. Macromolecules. 1988; 21 (9): 2722-2727. 9. Valentin HE, Schnebaum A, Steinbchel A. Identification of 4-hydroxyvaleric acid as a constituent of biosynthetic polyhydroxyalkanoic acids from bacteria. Appl Microbiol Biotechnol. 1992; 36 (4): 507-514. 10. Doi Y, Tamaki A, Kunioka M, Soga K. Biosynthesis of terpolyesters of 3-hydroxybutyrate, 3-hydroxyvalerate, and 5-hydroxyvalerate in Alcaligenes eutrophus from 5-chloropentanoic and pentanoic acids. Die Makromol Chem, Rapid Commun. 1987; 8 (12): 631-635. 11. Steinbchel A, Valentin HE. Diversity of bacterial polyhydroxyalkanoic acids. FEMS Microbiol Lett. 1995; 128 (3): 219-228. 12. Rehm BHA. Bacterial polymers: Biosynthesis, modifications and applications. Nat Rev Microbiol. 2010; 8 (8): 578-592. 13. Jendrossek D, Handrick R. Microbial degradation of polyhydroxyalkanoates. Annu Rev Microbiol. 2002; 56 (1): 403-432. 14. Abe H, Doi Y. Enzymatic and environmental degradation of racemic poly(3hydroxybutyric acid)s with different stereoregularities. Macromolecules. 1996; 29 (27): 86838688. 15. Elbanna K, Ltke-Eversloh T, Jendrossek D, Luftmann H, Steinbchel A. Studies on the biodegradability of polythioester copolymers and homopolymers by polyhydroxyalkanoate (PHA)-degrading bacteria and PHA depolymerases. Arch Microbiol. 2004; 182 (2): 212-225. 16. Jendrossek D, Schirmer A, Schlegel HG. Biodegradation of polyhydroxyalkanoic acids. Appl Microbiol Biotechnol. 1996; 46 (5): 451-463. 17. Akiyama M, Tsuge T, Doi Y. Environmental life cycle comparison of polyhydroxyalkanoates produced from renewable carbon resources by bacterial fermentation. Polym Degrad Stab. 2003; 80 (1): 183-194.

123

18. Kim S, Dale BE. Energy and greenhouse gas profiles of polyhydroxybutyrates derived from corn grain: A life cycle perspective. Environ Sci Technol. 2008; 42 (20): 7690-7695. 19. Snell KD, Peoples OP. PHA bioplastic: A value-added coproduct for biomass biorefineries. Biofuels, Bioprod Biorefin. 2009; 3 (4): 456-467. 20. Doi Y. Microbial polyesters. VCH Publishers, Inc.: New York City, 1990; p 118-120. 21. Steinbuchel A, Lutke-Eversloh T. Metabolic engineering and pathway construction for biotechnological production of relevant polyhydroxyalkanoates in microorganisms. Biochem Eng J. 2003; 16 (2): 81-96. 22. Ruiz JA, Lopez NI, Fernandez RO, Mendez BS. Polyhydroxyalkanoate degradation is associated with nucleotide accumulation and enhances stress resistance and survival of Pseudomonas oleovorans in natural water microcosms. Appl Environ Microbiol. 2001; 67 (1): 225-230. 23. Philip S, Keshavarz T, Roy I. Polyhydroxyalkanoates: Biodegradable polymers with a range of applications. J Chem Technol Biotechnol. 2007; 82 (3): 233-247. 24. Shen L, Worrell E, Patel M. Present and future development in plastics from biomass. Biofuels, Bioprod Biorefin. 4 (1): 25-40. 25. Tsuge T. Metabolic improvements and use of inexpensive carbon sources in microbial production of polyhydroxyalkanoates. J Biosci Bioeng. 2002; 94 (6): 579-584. 26. Nomura CT, Tanaka T, Eguen TE, Appah AS, Matsumoto K, Taguchi S, Ortiz CL, Doi Y. FabG mediates polyhydroxyalkanoate production from both related and nonrelated carbon sources in recombinant Escherichia coli LS5218. Biotechnol Prog. 2008; 24 (2): 342-351. 27. Sodergard A, Stolt M. Properties of lactic acid based polymers and their correlation with composition. Prog Polym Sci. 2002; 27 (6): 1123-1163. 28. Doi Y, Kitamura S, Abe H. Microbial synthesis and characterization of poly(3hydroxybutyrate-co-3-hydroxyhexanoate). Macromolecules. 1995; 28 (14): 4822-4828. 29. Shimamura E, Kasuya K, Kobayashi G, Shiotani T, Shima Y, Doi Y. Physical properties and biodegradability of microbial poly(3-hydroxybutyrate-co-3-hydroxyhexanoate). Macromolecules.1994; 27 (3): 878-880. 30. Asrar J, Valentin HE, Berger PA, Tran M, Padgette SR, Garbow JR. Biosynthesis and properties of poly(3-hydroxybutyrate-co-3-hydroxyhexanoate) polymers. Biomacromolecules. 2002; 3 (5): 1006-1012. 31. Noda I, Green PR, Satkowski MM, Schechtman LA. Preparation and properties of a novel class of polyhydroxyalkanoate copolymers. Biomacromolecules. 2005; 6 (2): 580-586. 32. Kunioka M, Tamaki A, Doi Y. Crystalline and thermal properties of bacterial copolyesters: Poly(3-hydroxybutyrate-co-3-hydroxyvalerate) and poly(3-hydroxybutyrate-co-4hydroxybutyrate). Macromolecules. 1989; 22 (2): 694-697. 33. Bluhm TL, Hamer GK, Marchessault RH, Fyfe CA, Veregin RP. Isodimorphism in bacterial poly(beta-hydroxybutyrate-co-beta-hydroxyvalerate). Macromolecules. 1986; 19 (11): 28712876. 34. Barham PJ. Nucleation behaviour of poly-3-hydroxy-butyrate. J Mater Sci. 1984; 19 (12): 3826-3834. 35. Liu WJ, Yang HL, Wang Z, Dong LS, Liu JJ. Effect of nucleating agents on the crystallization of poly(3-hydroxybutyrate-co-3-hydroxyvalerate). J Appl Polym Sci. 2002; 86 (9): 2145-2152. 36. El-Hadi A, Schnabel R, Straube E, Muller G, Henning S. Correlation between degree of crystallinity, morphology, glass temperature, mechanical properties and biodegradation of poly (3-hydroxyalkanoate) PHAs and their blends. Polym Test. 2002; 21 (6): 665-674. 124

37. Barham PJ, Keller A, Otun EL, Holmes PA. Crystallization and morphology of a bacterial thermoplastic: Poly-3-hydroxybutyrate. J Mater Sci. 1984; 19 (9): 2781-2794. 38. Kai W, He Y, Inoue Y. Fast crystallization of poly(3-hydroxybutyrate) and poly(3hydroxybutyrate-co-3-hydroxyvalerate) with talc and boron nitride as nucleating agents. Polym Int. 2005; 54 (5): 780-789. 39. Jacquel N, Tajima K, Nakamura N, Kawachi H, Pan PJ, Inoue Y. Nucleation mechanism of polyhydroxybutyrate and poly(hydroxybutyrate-co-hydroxyhexanoate) crystallized by orotic acid as a nucleating agent. J Appl Polym Sci. 2010; 115 (2): 709-715. 40. He Y, Inoue Y. Effect of alpha-cyclodextrin on the crystallization of poly(3hydroxybutyrate). J Polym Sci, Part B-Polym Phys. 2004; 42 (18): 3461-3469. 41. Withey RE, Hay JN. The effect of seeding on the crystallisation of poly(hydroxybutyrate), and co-poly(hydroxybutyrate-co-valerate). Polymer. 1999; 40 (18): 5147-5152. 42. Shishatskaya EI, Khlusov IA, Volova TG. A hybrid phb-hydroxyapatite composite for biomedical application: Production, in vitro and in vivo investigation. J Biomater Sci, Polym Ed. 2006; 17: 481-498. 43. Lemoigne M. Produits de deshydration et de polymerization de l' acide -oxybutyric. Bull Soc Chim Biol. 1926; 8: 770-782. 44. Lenz RW, Marchessault RH. Bacterial polyesters: Biosynthesis, biodegradable plastics and biotechnology. Biomacromolecules. 2005; 6 (1): 1-8. 45. Williamson DH, Wilkinson JF. The isolation and estimation of the poly-beta-hydroxybutyrate inclusions of Bacillus species. J Gen Microbiol. 1958; 19 (1): 198-209. 46. Doudoroff M, Stanier RY. Role of poly-beta-hydroxybutyric acid in the assimilation of organic carbon by bacteria. Nature. 1959; 183: 1440 - 1442. 47. Merrick JM, Doudoroff M. Enzymatic synthesis of poly-beta-hydroxybutyric acid in bacteria. Nature. 1961; 189: 890-892. 48. Lusty CJ, Doudoroff M. Poly-beta-hydroxybutyrate depolymerases of Pseudomonas lemoignei. Proc Natl Acad Sci USA. 1966; 56 (3): 960-965. 49. Lundgren DG, Alper R, Schnaitman C, Marchessault RH. Characterization of poly-{beta}hydroxybutyrate extracted from different bacteria. J Bacteriol. 1965; 89 (1): 245-251. 50. Wallen LL, Rohwedder WK. Poly-.Beta.-hydroxyalkanoate from activated sludge. Environ Sci Technol. 1974; 8 (6): 576-579. 51. Asrar J, Gruys KJ, Biodegradable polymer (biopol). In Polyesters iii: Applications and commercial products, Steinbchel, A., Ed. WILEY-VCH: 2004; Vol. 4, pp 53-84. 52. Reddy CSK, Ghai R, Rashmi, Kalia VC. Polyhydroxyalkanoates: An overview. Bioresour Technol. 2003; 87 (2): 137-146. 53. Bohlmann GM. Polyhydroxyalkanoate production in crops. In Feedstocks for the future, ACS Symposium series: 2006; Vol. 921, pp 253-270. 54. Noda I, Satkowski MM, Dowrey AE, Marcott C. Polymer alloys of Nodax copolymers and poly(lactic acid). Macromol Biosci. 2004; 4 (3): 269-275. 55. Noda I, Bond EB, Green PR, Melik DH, Narasimhan K, Schechtman LA, Satkowski MM. Preparation, properties, and utilization of biobased biodegradable Nodax copolymers. In Polymer biocatalysis and biomaterials, ACS Symposium series: 2005; Vol. 900, pp 280-291. 56. Narasimhan K, Green PR. In Nodax: Low cost delivery plans, The 227th ACS National Meeting, Anaheim, CA, 2004; Anaheim, CA, 2004. 57. Chen GQ. A microbial polyhydroxyalkanoates (PHA) based bio- and materials industry. Chem Soc Rev. 2009; 38 (8): 2434-2446. 58. Keenan TM, Nakas JP, Tanenbaum SW. Polyhydroxyalkanoate copolymers from forest biomass. J Ind Microbiol Biotechnol. 2006; 33 (7): 616-26. 125

59. Yellore V, Desai A. Production of poly-3-hydroxybutyrate from lactose and whey by Methylobacterium sp. Zp24. Lett Appl Microbiol. 1998; 26 (6): 391-4. 60. Gouda MK, Swellam AE, Omar SH. Production of PHB by a Bacillus megaterium strain using sugarcane molasses and corn steep liquor as sole carbon and nitrogen sources. Microbiol Res. 2001; 156 (3): 201-207. 61. Jones AM, Thomas KC, Ingledew WM. Ethanolic fermentation of blackstrap molasses and sugarcane juice using very high gravity technology. J Agri Food Chem. 1994; 42 (5): 12421246. 62. Kahar P, Tsuge T, Taguchi K, Doi Y. High yield production of polyhydroxyalkanoates from soybean oil by Ralstonia eutropha and its recombinant strain. Polym Degrad Stab. 2004; 83 (1): 79-86. 63. Zhu CJ, Nomura CT, Perrotta JA, Stipanovic AJ, Nakas JP. Production and characterization of poly-3-hydroxybutyrate from biodiesel-glycerol by Burkholderia cepacia ATCC 17759. Biotechnol Prog. 2010; 26 (2): 424-430. 64. Erlandson KA, Park JH, Wissam, El K, Kao HH, Basaran P, Brydges S, Batt CA. Dissolution of xylose metabolism in Lactococcus lactis. Appl Environ Microbiol. 2000; 66 (9): 3974-3980. 65. Zhang M, Eddy C, Deanda K, Finkelstein M, Picataggio S. Metabolic engineering of a pentose metabolism pathway in ethanologenic Zymomonas mobilis. Science. 1995; 267 (5195): 240-243. 66. Holmberg C, Beijer L, Rutberg B, Rutberg L. Glycerol catabolism in Bacillus subtilis: Nucleotide sequence of the genes encoding glycerol kinase (glpk) and glycerol-3-phosphate dehydrogenase (glpd). J Gen Microbiol. 1990; 136 (12): 2367-2375. 67. Voet D, Voet JG. Biochemistry. 3rd ed.; Wiley: 2004. 68. Verlinden RAJ, Hill DJ, Kenward MA, Williams CD, Radecka I. Bacterial synthesis of biodegradable polyhydroxyalkanoates. J Appl Microbiol. 2007; 102 (6): 1437-1449. 69. Wang Q, Nomura CT. Monitoring differences in gene expression levels and polyhydroxyalkanoate (PHA) production in Pseudomonas putida KT2440 grown on different carbon sources. J Biosci Bioeng. 110 (6): 653-659. 70. Fidler S, Dennis D. Polyhydroxyalkanoate production in recombinant Escherichia coli. FEMS Microbiol Lett. 1992; 103 (2-4): 231-235. 71. Li S, Dong C, Wang S, Ye H, Chen GQ. Microbial production of polyhydroxyalkanoate block copolymer by recombinant Pseudomonas putida. Appl Microbiol Biotechnol. 90 (2): 659669. 72. Derraik JGB. The pollution of the marine environment by plastic debris: A review. Mar Pollut Bull. 2002; 44 (9): 842-852. 73. Amor SR, Rayment T, Sanders JKM. Poly(hydroxybutyrate) in vivo: NMR and x-ray characterization of the elastomeric state. Macromolecules. 1991; 24 (16): 4583-4588. 74. Volova TG, Boyandin AN, Vasiliev AD, Karpov VA, Prudnikova SV, Mishukova OV, Boyarskikh UA, Filipenko ML, Rudnev VP, Xuan BB, Dung VV, Gitelson II. Biodegradation of polyhydroxyalkanoates (PHAs) in tropical coastal waters and identification of PHA-degrading bacteria. Polym Degrad Stab. 95 (12): 2350-2359. 75. Mergaert J, Webb A, Anderson C, Wouters A, Swings J. Microbial degradation of poly(3hydroxybutyrate) and poly(3-hydroxybutyrate-co-3-hydroxyvalerate) in soils. Appl Environ Microbiol. 1993; 59 (10): 3233-3238. 76. Mergaert J, Wouters A, Swings J, Anderson C. In situ biodegradation of poly(3hydroxybutyrate) and poly(3-hydroxybutyrate-co-3-hydroxyvalerate) in natural waters. Can J Microbiol. 1995; 41 (13): 154-159. 126

77. Mergaert J, Anderson C, Wouters A, Swings J. Microbial degradation of poly(3hydroxybutyrate) and poly(3-hydroxybutyrate-co-3-hydroxyvalerate) in compost. J Polym Environ. 1994; 2 (3): 177-183. 78. dos Santos Rosa D, Calil MR, das Graas Fassina Guedes C, Santos CEO. The effect of UVB irradiation on the biodegradability of poly-beta-hydroxybutyrate (PHB) and poly-caprolactone (PCL). J Polym Environ. 2001; 9 (3): 109-113. 79. Chen C, Fei B, Peng S, Zhuang Y, Dong L, Feng Z. The kinetics of the thermal decomposition of poly(3-hydroxybutyrate) and maleated poly(3-hydroxybutyrate). J Appl Polym Sci. 2002; 84 (9): 1789-1796. 80. Kunioka M, Doi Y. Thermal degradation of microbial copolyesters: Poly(3hydroxybutyrate-co-3-hydroxyvalerate) and poly(3-hydroxybutyrate-co-4-hydroxybutyrate). Macromolecules. 1990; 23 (7): 1933-1936. 81. Morikawa H, Marchessault RH. Pyrolysis of bacterial polyalkanoates. Can J Chem. 1981; 59 (15): 2306-2313. 82. Ariffin H, Nishida H, Shirai Y, Hassan MA. Determination of multiple thermal degradation mechanisms of poly(3-hydroxybutyrate). Polym Degrad Stab. 2008; 93 (8): 1433-1439. 83. Santos A, Polese L, Crespi M, Ribeiro C. Kinetic model of poly(3-hydroxybutyrate) thermal degradation from experimental non-isothermal data. J Therm Anal Calorim. 2009; 96 (1): 287-291. 84. Grassie N, Murray EJ, Holmes PA. The thermal degradation of poly(-(D)-[beta]hydroxybutyric acid): Part 3--the reaction mechanism. Polym Degrad Stab. 1984; 6 (3): 127-134. 85. Mitomo H, Watanabe Y, Yoshii F, Makuuchi K. Radiation effect on polyesters. Radiat Phys Chem. 1995; 46 (2): 233-238. 86. Mitomo H, Sasaoka T, Yoshii F, Makuuchi K, Saito T. Radiation-induced graft polymerization of acrylic acid onto poly(3-hydroxybutyrate) and its copolymer. Sen'i Gakkaishi. 1996; 52 (11): 623-626. 87. Bahari K, Mitomo H, Enjoji T, Yoshii F, Makuuchi K. Degradability of poly(3hydroxybutyrate) and its copolymer grafted with styrene by radiation. Polym Degrad Stab. 1998; 61 (2): 245-252. 88. Bahari K, Mitomo H, Enjoji T, Hasegawa S, Yoshii F, Makuuchi K. Radiation-induced graft polymerization of styrene onto poly(3-hydroxybutyrate) and its copolymer with 3hydroxyvalerate. Die Angew Makromol Chem. 1997; 250 (1): 31-44. 89. Choi Ji, Lee SY. Process analysis and economic evaluation for poly(3-hydroxybutyrate) production by fermentation. Bioprocess Biosyst Eng. 1997; 17 (6): 335-342. 90. Choi J, Lee S Y. Factors affecting the economics of polyhydroxyalkanoate production by bacterial fermentation. Appl Microbiol Biotechnol. 1999; 51 (1): 13-21. 91. Khardenavis A A, Suresh Kumar M, Mudliar SN, Chakrabarti T. Biotechnological conversion of agro-industrial wastewaters into biodegradable plastic, poly [beta]hydroxybutyrate. Bioresour Technol. 2007; 98 (18): 3579-3584. 92. Gurieff N, Lant P. Comparative life cycle assessment and financial analysis of mixed culture polyhydroxyalkanoate production. Bioresour Technol. 2007; 98 (17): 3393-3403. 93. Nath A, Dixit M, Bandiya A, Chavda S, Desai A J. Enhanced PHB production and scale up studies using cheese whey in fed batch culture of Methylobacterium sp. Zp24. Bioresour Technol. 2008; 99 (13): 5749-5755. 94. Pachauri N, He B. In Value-added utilization of crude glycerol from biodiesel production: A survey of current research activities, 2006 American Society of Agricultural and Biological Engineering Annual International Meeting, Portland, Oregon, 9 - 12 July, 2006; Portland, Oregon, 2006. 127

95. Yazdani SS, Mattam AJ, Gonzalez R. Fuel and chemical production from glycerol, a biodiesel waste product. In Biofuels from agricultural wastes and byproducts, Wiley-Blackwell: 2010; pp 97-116. 96. Yazdani SS, Gonzalez R. Anaerobic fermentation of glycerol: A path to economic viability for the biofuels industry. Curr Opin Biotechnol. 2007; 18 (3): 213-219. 97. Ragauskas AJ, Williams CK, Davison BH, Britovsek G, Cairney J, Eckert CA, Frederick WJ, Hallett JP, Leak DJ, Liotta CL, Mielenz JR, Murphy R, Templer R, Tschaplinski T. The path forward for biofuels and biomaterials. Science. 2006; 311 (5760): 484-489. 98. Huber GW, Iborra S, Corma A. Synthesis of transportation fuels from biomass: Chemistry, catalysts, and engineering. Chem Rev. 2006; 106 (9): 4044-4098. 99. Werpy T, Petersen G. Top value added chemicals from biomass. Vol. 1results of screening for potential candidates from candidates from sugars and synthesis gas. U.S. Dep. Energy, Off. Sci. Tech. Inf. http://www.nrel.gov/docs/fy04osti/35523.pdf 100. Serrano-Ruiz JC, West RM, Dumesic JA. Catalytic conversion of renewable biomass resources to fuels and chemicals. Annu Rev Chem Biomole Eng. 1 (1): 79-100. 101. Leonard R. Levulinic acid as a basic chemical raw material. Ind Eng Chem. 1956; 48 (8): 1330-1341. 102. Fitzpatrick SW. Lignocellulose degradation to furfural and levulinic acid. U.S. Patent 4897497. 1990. 103. Fitzpatrick SW. Production of levulinic acid from carbohydrate-containing materials. U.S. Patent 5608105. 1997. 104. Bozell JJ, Moens L, Elliott DC, Wang Y, Neuenscwander GG, Fitzpatrick SW, Bilski RJ, Jarnefeld JL. Production of levulinic acid and use as a platform chemical for derived products. Resour Conserv Recy. 2000; 28 (3-4): 227-239. 105. AltIparmak D, Keskin A, Koca A, Guru M. Alternative fuel properties of tall oil fatty acid methyl ester-diesel fuel blends. Bioresour Technol. 2007; 98 (2): 241-246. 106. Keskin A, Guru M, AltIparmak D. Biodiesel production from tall oil with synthesized Mn and Ni based additives: Effects of the additives on fuel consumption and emissions. Fuel. 2007; 86 (7-8): 1139-1143. 107. Tan IKP, Kumar KS, Theanmalar M, Gan SN, Gordon III, B. Saponified palm kernel oil and its major free fatty acids as carbon substrates for the production of polyhydroxyalkanoates in Pseudomonas putida PGA1. Appl Microbiol Biotechnol. 1997; 47 (3): 207-211. 108. Kellerhals MB, Kessler B, Witholt B, Tchouboukov A, Brandl H. Renewable long-chain fatty acids for production of biodegradable medium-chain-length polyhydroxyalkanoates (mclPHAs) at laboratory and pilot plant scales. Macromolecules. 2000; 33 (13): 4690-4698. 109. Jacquel N, Lo CW, Wei YH, Wu HS, Wang SS. Isolation and purification of bacterial poly(3-hydroxyalkanoates). Biochem Eng J. 2008; 39 (1): 15-27. 110. Chen GQ, Zhang G, Park SJ, Lee SY. Industrial scale production of poly(3hydroxybutyrate-co-3-hydroxyhexanoate). Appl Microbiol Biotechnol. 2001; 57 (1): 50-55. 111. Kunasundari B, Sudesh K. Isolation and recovery of microbial polyhydroxyalkanoates. eXPRESS Polym Lett. 2011; 5 (7): 620-634. 112. Dias JML, Lemos PC, Serafim LS, Oliveira C, Eiroa M, Albuquerque MGE, Ramos AM, Oliveira R, Reis MAM. Recent advances in polyhydroxyalkanoate production by mixed aerobic cultures: From the substrate to the final product. Macromol Biosci. 2006; 6 (11): 885-906. 113. Yu J, Chen LXL. Cost-effective recovery and purification of polyhydroxyalkanoates by selective dissolution of cell mass. Biotechnol Prog. 2006; 22 (2): 547-553.

128

114. Bertrand JL, Ramsay BA, Ramsay JA, Chavarie C. Biosynthesis of poly-betahydroxyalkanoates from pentoses by Pseudomonas pseudoflava. Appl Environ Microbiol. 1990; 56 (10): 3133-3138. 115. Keenan TM, Tanenbaum SW, Stipanovic AJ, Nakas JP. Production and characterization of poly-beta-hydroxyalkanoate copolymers from Burkholderia cepacia utilizing xylose and levulinic acid. Biotechnol Prog. 2004; 20 (6): 1697-704. 116. Nomura CT, Taguchi K, Taguchi S, Doi Y. Coexpression of genetically engineered 3ketoacyl-ACP synthase iii (fabH) and polyhydroxyalkanoate synthase (phaC) genes leads to shortchain-length-medium-chain-length polyhydroxyalkanoate copolymer production from glucose in Escherichia coli JM109. Appl Environ Microbiol. 2004; 70 (2): 999-1007. 117. Gu XZ, Wu J, Mather PT. Polyhedral oligomeric silsesquioxane (POSS) suppresses enzymatic degradation of PCL-based polyurethanes. Biomacromolecules. xxxx; xxx: 000-000 early view published online. 118. Bloembergen S, Holden DA, Hamer GK, Bluhm TL, Marchessault RH. Studies of composition and crystallinity of bacterial poly(-hydroxybutyrate-co--hydroxyvalerate). Macromolecules. 1986; 19 (11): 2865-2871. 119. Lee SH, Oh DH, Ahn WS, Lee Y, Choi Ji, Lee SY. Production of poly(3-hydroxybutyrate-co3-hydroxyhexanoate) by high-cell-density cultivation of Aeromonas hydrophila. Biotechnol Bioeng. 2000; 67 (2): 240-244. 120. Ashby RD, Solaiman DK, Foglia TA. Synthesis of short-/medium-chain-length poly(hydroxyalkanoate) blends by mixed culture fermentation of glycerol. Biomacromolecules. 2005; 6 (4): 2106-12. 121. Bloembergen S, Holden DA, Bluhm TL, Hamer GK, Marchessault RH. Isodimorphism in synthetic poly(beta-hydroxybutyrate-co-beta-hydroxyvalerate): Stereoregular copolyesters from racemic beta-lactones. Macromolecules. 1989; 22 (4): 1663-1669. 122. Shang L, Yim SC, Park HG, Chang HN. Sequential feeding of glucose and valerate in a fedbatch culture of Ralstonia eutropha for production of poly(hydroxybutyrate-co-hydroxyvalerate) with high 3-hydroxyvalerate fraction. Biotechnol Prog. 2004; 20 (1): 140-144. 123. Yu J, Stahl H. Microbial utilization and biopolyester synthesis of bagasse hydrolysates. Bioresour Technol. 2008; 99 (17): 8042-8048. 124. Yu J, Chen LXL, Sato S. Biopolyester synthesis and protein regulations in Ralstonia eutropha on levulinic acid and its derivatives from biomass refining. J Biobased Mater Bioenergy. 2009; 3: 113-122. 125. Valentin HE, Steinbchel A. Accumulation of poly(3-hydroxybutyric acid-co-3hydroxyvaleric acid-co-4-hydroxyvaleric acid) by mutants and recombinant strains of Alcaligenes eutrophus. J Polym Environ. 1995; 3 (3): 169-175. 126. Byelov D, Panine P, Remerie K, Biemond E, Alfonso GC, de Jeu WH. Crystallization under shear in isotactic polypropylene containing nucleators. Polymer. 2008; 49 (13-14): 3076-3083. 127. Ferrage E, Martin F, Boudet A, Petit S, Fourty G, Jouffret F, Micoud P, De Parseval P, Salvi S, Bourgerette C, Ferret J, Saint-Gerard Y, Buratto S, Fortune J. Talc as nucleating agent of polypropylene: Morphology induced by lamellar particles addition and interface mineral-matrix modelization. J Mater Sci. 2002; 37 (8): 1561-1573. 128. Sudesh K, Iwata T. Sustainability of biobased and biodegradable plastics. CLEAN Soil, Air, Water. 2008; 36 (5-6): 433-442. 129. Bohmert K, Balbo I, Kopka J, Mittendorf V, Nawrath C, Poirier Y, Tischendorf G, Trethewey RN, Willmitzer L. Transgenic Arabidopsis plants can accumulate polyhydroxybutyrate to up to 4% of their fresh weight. Planta. 2000; 211 (6): 841-845. 129

130. Keenan TM. Production and characterization of poly-beta-hydroxyalkanoate copolymers fron xylose and levulinic acid using Burkholderia cepacia. Ph.D. thesis. SUNY-College of Environmental Science and Forestry, Syracuse, NY, 2005. 131. Palmqvist E, Hahn-Hagerdal B. Fermentation of lignocellulosic hydrolysates. I: Inhibition and detoxification. Bioresour Technol. 2000; 74 (1): 17-24. 132. Palmqvist E, Hahn-Hagerdal B. Fermentation of lignocellulosic hydrolysates. Ii: Inhibitors and mechanisms of inhibition. Bioresour Technol. 2000; 74 (1): 25-33. 133. Hoffmann N, Rehm BHA. Regulation of polyhydroxyalkanoate biosynthesis in Pseudomonas putida and Pseudomonas aeruginosa. FEMS Microbiol Lett. 2004; 237 (1): 1-7. 134. Fernandez D, Rodriguez E, Bassas M, Vinas M, Solanas AM, Llorens J, Marques AM, Manresa A. Agro-industrial oily wastes as substrates for PHA production by the new strain Pseudomonas aeruginosa NCIB 40045: Effect of culture conditions. Biochem Eng J. 2005; 26 (2-3): 159-167. 135. Hartmann R, Hany R, Pletscher E, Ritter A, Witholt B, Zinn M. Tailor-made olefinic medium-chain-length poly[(R)-3-hydroxyalkanoates] by Pseudomonas putida GPO1: Batch versus chemostat production. Biotechnol Bioeng. 2006; 93 (4): 737-746. 136. Stubbe J, Tian J. Polyhydroxyalkanoate (PHA) homeostasis: The role of the PHA synthase. Nat Prod Rep. 2003; 20 (5): 445-457. 137. Solaiman DKY, Ashby RD. Rapid genetic characterization of poly(hydroxyalkanoate) synthase and its applications. Biomacromolecules. 2004; 6 (2): 532-537. 138. Tobin KM, O'Connor KE. Polyhydroxyalkanoate accumulating diversity of Pseudomonas species utilising aromatic hydrocarbons. FEMS Microbiol Lett. 2005; 253 (1): 111-118. 139. Abraham GA, Gallardo A, San Roman J, Olivera ER, Jodra R, Garcia B, Minambres B, Garcia JL, Luengo JM. Microbial synthesis of poly(beta-hydroxyalkanoates) bearing phenyl groups from Pseudomonas putida: Chemical structure and characterization. Biomacromolecules. 2001; 2 (2): 562-567. 140. de Eugenio LI, Garcia P, Luengo JM, Sanz JM, Roman JS, Garcia JL, Prieto MA. Biochemical evidence that phaz gene encodes a specific intracellular medium chain length polyhydroxyalkanoate depolymerase in Pseudomonas putida KT2442. J Biol Chem. 2007; 282 (7): 4951-4962. 141. Ward PG, O'Connor KE. Bacterial synthesis of polyhydroxyalkanoates containing aromatic and aliphatic monomers by Pseudomonas putida CA-3. Int J Biol Macromol. 2005; 35 (3-4): 127-133. 142. Choi MH, Xu J, Rho JK, Shim JH, Yoon SC. Shifting of the distribution of aromatic monomer-units in polyhydroxyalkanoic acid to longer units by salicylic acid in Pseudomonas fluorescens BM07 grown with mixtures of fructose and 11-phenoxyundecanoic acid. Biotechnol Bioeng. 2009; 102 (4): 1209-1221. 143. Ashby R, Solaiman D, Strahan G. Efficient utilization of crude glycerol as fermentation substrate in the synthesis of poly(3-hydroxybutyrate) biopolymers. J Am Oil Chem Soc. 2011; 88 (7): 949-959. 144. de Almeida A, Nikel PI, Giordano AM, Pettinari MJ. Effects of granule-associated protein PhaPon glycerol-dependent growth and polymer production in poly(3-hydroxybutyrate)producing Escherichia coli. Appl Environ Microbiol. 2007; 73 (24): 7912-7916. 145. Nikel P, Pettinari M, Galvagno M, Mndez B. Poly(3-hydroxybutyrate) synthesis from glycerol by a recombinant Escherichia coli arcA mutant in fed-batch microaerobic cultures. Appl Microbiol Biotechnol. 2008; 77 (6): 1337-1343. 146. Mothes G, Schnorpfeil C, Ackermann JU. Production of PHB from crude glycerol. Eng Life Sci. 2007; 7 (5): 475-479. 130

147. Nishioka M, Nakai K, Miyake M, Asada Y, Taya M. Production of poly--hydroxybutyrate by thermophilic cyanobacterium, Synechococcus sp. MA19, under phosphate-limited conditions. Biotechnol Lett. 2001; 23 (14): 1095-1099. 148. Ishizaki A, Tanaka K, Taga N. Microbial production of poly-D-3-hydroxybutyrate from co2. Appl Microbiol Biotechnol. 2001; 57 (1): 6-12. 149. Cho KS, Wook Ryu H, Park CH, Goodrich PR. Utilization of swine wastewater as a feedstock for the production of polyhydroxyalkanoates by Azotobacter vinelandii UWD. J Biosci Bioeng. 2001; 91 (2): 129-133. 150. Page WJ. Suitability of commercial beet molasses fractions as substrates for polyhydroxyalkanoate production by Azotobacter vinelandii UWD. Biotechnol Lett. 1992; 14 (5): 385-390. 151. Omar S, Rayes A, Eqaab A, Vo I, Steinbchel A. Optimization of cell growth and poly(3hydroxybutyrate) accumulation on date syrup by a Bacillus megaterium strain. Biotechnol Lett. 2001; 23 (14): 1119-1123. 152. Koller M, Bona R, Braunegg G, Hermann C, Horvat P, Kroutil M, Martinz J, Neto J, Pereira L, Varila P. Production of polyhydroxyalkanoates from agricultural waste and surplus materials. Biomacromolecules. 2005; 6 (2): 561-565. 153. Yan Q, Zhao M, Miao H, Ruan W, Song R. Coupling of the hydrogen and polyhydroxyalkanoates (PHA) production through anaerobic digestion from taihu blue algae. Bioresour Technol. 2010; 101 (12): 4508-4512. 154. Alva Munoz LE, Riley MR. Utilization of cellulosic waste from tequila bagasse and production of polyhydroxyalkanoate (PHA) bioplastics by Saccharophagus degradans. Biotechnol Bioeng. 2008; 100 (5): 882-888. 155. Cromwick AM, Foglia T, Lenz RW. The microbial production of poly(hydroxyalkanoates) from tallow. Appl Microbiol Biotechnol. 1996; 46 (5): 464-469. 156. Cavalheiro JMBT, de Almeida MCMD, Grandfils C, da Fonseca MMR. Poly(3hydroxybutyrate) production by Cupriavidus necator using waste glycerol. Process Biochem. 2009; 44 (5): 509-515. 157. Nikel P, Pettinari J, Mendez B. Poly(3-hydroxybutyrate) synthesis in microaerobic fedbatch cultures by a recombinant Escherichia coli arcA mutant using glycerol as a carbon source. J Biotechnol. 2007; 131 (2): S157-S158. 158. Yeh JI, Chinte U, Du S. Structure of glycerol-3-phosphate dehydrogenase, an essential monotopic membrane enzyme involved in respiration and metabolism. Proc Natl Acad Sci USA. 2008; 105 (9): 3280-5. 159. Gerngross TU, Martin DP. Enzyme-catalyzed synthesis of poly[(R)-(-)-3-hydroxybutyratel: Formation of macroscopic granules in vitro. Proc Natl Acad Sci USA. 1995; 92: 6279-6283. 160. Rehm BH, Steinbuchel A. Biochemical and genetic analysis of PHA synthases and other proteins required for PHA synthesis. Int J Biol Macromol. 1999; 25 (1-3): 3-19. 161. Sim SJ, Snell KD, Hogan SA, Stubbe J, Rha C, Sinskey AJ. PHA synthase activity controls the molecular weight and polydispersity of polyhydroxybutyrate in vivo. Nat Biotechnol. 1997; 15 (1): 63-7. 162. Amara AA, Steinbuchel A, Rehm BH. In vivo evolution of the Aeromonas punctata polyhydroxyalkanoate (PHA) synthase: Isolation and characterization of modified PHA synthases with enhanced activity. Appl Microbiol Biotechnol. 2002; 59 (4-5): 477-82. 163. Madden LA, Anderson AJ, Shah DT, Asrar J. Chain termination in polyhydroxyalkanoate synthesis: Involvement of exogenous hydroxy-compounds as chain transfer agents. Int J Biologl Macromol. 1999; 25 (1-3): 43-53. 131

164. Taidi B, Anderson AJ, Dawes EA, Byrom D. Effect of carbon source and concentration on the molecular-mass of poly(3-hydroxybutyrate) produced by Methylobacterium extorquens and Alcaligenes eutrophus. Appl Microbiol Biotechnol. 1994; 40 (6): 786-790. 165. Aoyagi Y, Doi Y, Iwata T. Mechanical properties and highly ordered structure of ultrahigh-molecular-weight poly[(R)-3-hydroxybutyrate] films: Effects of annealing and two-step drawing. Polym Degrad Stab. 2003; 79 (2): 209-216. 166. Almeida JRM, Modig T, Petersson A, Hhn-Hgerdal B, Lidn G, Gorwa-Grauslund MF. Increased tolerance and conversion of inhibitors in lignocellulosic hydrolysates by Saccharomyces cerevisiae. J Chem Technol Biotechnol. 2007; 82 (4): 340-349. 167. Zaldivar J, Ingram L O. Effect of organic acids on the growth and fermentation of ethanologenic Escherichia coli LY01. Biotechnol Bioeng. 1999; 66 (4): 203-210. 168. Savenkova L, Gercberga Z, Bibers I, Kalnin M. Effect of 3-hydroxyvalerate content on some physical and mechanical properties of polyhydroxyalkanoates produced by Azotobacter chroococcum. Process Biochem. 2000; 36 (5): 445-450. 169. Ng K-S, Wong Y-M, Tsuge T, Sudesh K. Biosynthesis and characterization of poly(3hydroxybutyrate-co-3-hydroxyvalerate) and poly(3-hydroxybutyrate-co-3-hydroxyhexanoate) copolymers using jatropha oil as the main carbon source. Process Biochem. 46 (8): 1572-1578. 170. Yoshie N, Saito M, Inoue Y. Structural transition of lamella crystals in a isomorphous copolymer, poly(3-hydroxybutyrate-co-3-hydroxyvalerate). Macromolecules. 2001; 34 (26): 8953-8960. 171. Taidi B, Anderson AJ, Dawes EA, Byrom D. Effect of carbon source and concentration on the molecular mass of poly(3-hydroxybutyrate) produced by Methylobacterium extorquens and Alcaligenes eutrophus. Appl Microbiol Biotechnol. 1994; 40 (6): 786-790. 172. Kusaka S, Iwata T, Doi Y. Properties and biodegradability of ultra-high-molecular-weight poly[(R)-3-hydroxybutyrate] produced by a recombinant Escherichia coli. Int J Biol Macromol. 1999; 25 (1-3): 87-94. 173. Koning Gd. Physical properties of bacterial poly((R)-3-hydroxyalkanoates). Can J Microbiol. 1995; 41 (13): 303-309. 174. Babel W, Steinbchel A, van der Walle G, de Koning G, Weusthuis R, Eggink G. Properties, modifications and applications of biopolyesters. In Biopolyesters, Steinbchel, Babel W & Steinbuchel A, Ed. Springer Berlin / Heidelberg: 2001; Vol. 71, pp 263-291. 175. Zhang J, McCarthy S, Whitehouse R. Reverse temperature injection molding of Biopol and effect on its properties. J Appl Polym Sci. 2004; 94 (2): 483-491. 176. de Koning GJM, Lemstra PJ. Crystallization phenomena in bacterial poly[(R)-3hydroxybutyrate]: 2. Embrittlement and rejuvenation. Polymer. 1993; 34 (19): 4089-4094. 177. de Koning GJM, Scheeren AHC, Lemstra PJ, Peeters M, Reynaers H. Crystallization phenomena in bacterial poly[(R)-3-hydroxybutyrate]: 3. Toughening via texture changes. Polymer. 1994; 35 (21): 4598-4605. 178. Randall RS. Strategies for the optimization of polyhydroxyalkanoate production utilizing xylose. M.S. thesis. SUNY-College of Environmental Science and Forestry, Syracuse, NY, 2008. 179. Tamer IM, Moo-Young M, Chisti Y. Disruption of Alcaligenes latus for recovery of poly(beta-hydroxybutyric acid): Comparison of high-pressure homogenization, bead milling, and chemically induced lysis. Ind Eng Chem Res. 1998; 37 (5): 1807-1814. 180. Boynton ZL, Koon JJ, Brennan EM, Clouart JD, Horowitz DM, Gerngross TU, Huisman GW. Reduction of cell lysate viscosity during processing of poly(3-hydroxyalkanoates) by chromosomal integration of the Staphylococcal nuclease gene in Pseudomonas putida. Appl Environ Microbiol. 1999; 65 (4): 1524-1529. 132

181. Horowitz DM, Brennan EM, Koon JJ, Gerngross TU. Novel thermal route to an amorphous, film-forming polymer latex. Macromolecules. 1999; 32 (10): 3347-3352. 182. Chen G-Q, Wu Q. The application of polyhydroxyalkanoates as tissue engineering materials. Biomaterials. 2005; 26 (33): 6565-6578.

133

Appendix I. Biotechnology Progress, 2010, 26(2): 424-430.

Production and Characterization of Poly-3-hydroxybutyrate From Biodiesel-Glycerol by Burkholderia cepacia ATCC 17759
Chengjun Zhu
Dept. of Environmental and Forest Biology, SUNY-College of Environmental Science and Forestry, Syracuse, NY 13210

Christopher T. Nomura
Dept. of Chemistry, SUNY-College of Environmental Science and Forestry, Syracuse, NY 13210

Joseph A. Perrotta
Dept. of Environmental and Forest Biology, SUNY-College of Environmental Science and Forestry, Syracuse, NY 13210

Arthur J. Stipanovic
Dept. of Chemistry, SUNY-College of Environmental Science and Forestry, Syracuse, NY 13210

James P. Nakas
Dept. of Environmental and Forest Biology, SUNY-College of Environmental Science and Forestry, Syracuse, NY 13210 DOI 10.1002/btpr.355 Published online December 1, 2009 in Wiley InterScience (www.interscience.wiley.com).

Glycerol, a byproduct of the biodiesel industry, can be used by bacteria as an inexpensive carbon source for the production of value-added biodegradable polyhydroxyalkanoates (PHAs). Burkholderia cepacia ATCC 17759 synthesized poly-3-hydroxybutyrate (PHB) from glycerol concentrations ranging from 3% to 9% (v/v). Increasing the glycerol concentration results in a gradual reduction of biomass, PHA yield, and molecular mass (Mn and Mw) of PHB. The molecular mass of PHB produced utilizing xylose as a carbon source is also decreased by the addition of glycerol as a secondary carbon source dependent on the time and concentration of the addition. 1H-NMR revealed that molecular masses decreased due to the esterication of glycerol with PHB resulting in chain termination (end-capping). However, melting temperature and glass transition temperature of the end-capped polymers showed no signicant difference when compared to the xylose-based PHB. The fermentation was successfully scaled up to 200 L for PHB production and the yield of dry biomass and C PHB were 23.6 g/L and 7.4 g/L, respectively. V 2009 American Institute of Chemical Engineers Biotechnol. Prog., 26: 424430, 2010 Keywords: biodiesel-glycerol, polyhydroxyalkanoates, Burkholderia cepacia, polyhydroxybutyrate, end-capped PHB

Introduction
Polyhydroxyalkanoates (PHAs) are accumulated as microbial intracellular carbon and energy reserves. These polymers represent a class of compounds with physical-chemical characteristics similar to petroleum-derived plastics such as polypropylene, polyethylene and polystyrene, but are environmentally compatible and totally biodegradable to carbon dioxide and water.13 A number of microorganisms, including Ralstonia eutropha, Alcaligenes latus, and several species of Pseudomonas,47 have been shown to produce various polyesters with different subunits.8 The homopolymer PHB and the copolymer poly-3-hydroxybutyrate-co-poly-3-hydroxyvalerate (PHB-co-PHV) are the most widely studied and have been produced commercially to manufacture some nished products, which are primarily used in medical applications such as tissue engineering.9 Burkholderia (formerly PseudoCorrespondence concerning this article to J. P. Nakas at jpnakas@esf.edu.
424

should

be

addressed

monas) cepacia has been shown to efciently synthesize short-chain-length (scl) PHAs, such as PHB, PHV, and PHBco-PHV. By incorporating PHV with PHB to form the copolymer, lower crystallinity and better elongation can be obtained, which have been shown to exhibit more desirable mechanical properties.10 Many carbon sources, including xylose, galactose, glucose, glycerol and levulinic acid, have been used to support growth and scl-PHA production by B. cepacia.10,11 Although it is feasible for these carbon sources to be used to produce PHAs in the laboratory, high production costs hamper largescale commercial production and the cost of fermentation feedstocks can account for up to 50% of the overall production cost.4,12 Several process stream feedstocks, such as cheese whey permeate,13 wood hydrolysate,14 sugarcane molasses, and corn steep liquor,15 have been used to produce PHAs in an attempt to reduce production costs. Glycerol (approximately 10% of the nal weight of biodiesel)16 is the major byproduct of the biodiesel industry. As

134

C V 2009 American Institute of Chemical Engineers

Appendix II. Manuscript will be submitted for publication in the journal Polymer Degradation and Stability (August, 2011).

The effect of nucleating agents on physical properties of poly-3-hydroxybutyrate (PHB) and poly-3-hydroxybutyrate-co-3-hydroxyvalerate (PHB-co-HV) produced by Burkholderia cepacia ATCC 17759
Chengjun Zhu1, Christopher T. Nomura2, Joseph A. Perrotta1, Arthur J. Stipanovic2 and James P. Nakas1* Department of Environmental and Forest Biology, 2 Department of Chemistry, State University of New York-College of Environmental Science and Forestry, Syracuse NY 13210 Abstract Polyhydroxyalkanoates (PHAs), in the absence of nucleating agents, generally exhibit slow crystallization rates which make them less favorable for injection molding purposes. The copolymer, PHB-co-HV (poly-3-hydroxybutyrate-co-3-hydroxyvalerate), exhibited increased Tc and Tco as compared to the homopolymer PHB (poly-3-hydroxybutyrate). Increasing the mol fraction of HV monomer in the PHB-co-HV initially led to a decrease in the Tm of the copolymer from 175.4 oC to a minimum of 168.5 oC, at 20 mol% of HV, and subsequently increased in PHBco-HV copolymers with higher fractions of HV, indicating a typical isodimorphic relationship. Two nucleating agents, heptane dicarboxylic derivative HPN-68L and ULTRATALC609, were tested to increase the Tc and reduce the time for crystallization necessary for injection molding processing. HPN-68L decreased the Tdecomp of the homopolymer and all copolymers by almost 50
o 1

C . However, the use of ULTRATALC609 as a nucleating agent slightly enhanced the Tdecomp and

had negligible effect on the Tms of all polymers. Also, PHB and PHB-co-HV with 5% (w/w) talc exhibited higher Tc and Tco than polymers without ULTRATALC609. A careful comparison of Tc, Tm and Tdecomp, for PHB-co-HV with 20 mol% of HV indicated this copolymer to be the best option for injection molding with both a high Tdecomp and more rapid crystallization. 135

Appendix III. Provisionary patent application filed with the US Patent and Trademark Office. November, 2010. Patent pending.

Methods for Producing Polyhydroxyalkanoates from Biodiesel-Glycerol

James P. Nakas Chengjun Zhu Joseph A. Perrotta Christopher T. Nomura State University of New York College of Environmental Science and Forestry Department of Environmental and Forest Biology 1 Forestry Drive Syracuse, NY, USA 13210

136

Appendix IV. Manuscript will be sent for publication in the journal Biomacromolecules (2011)

Glycerine and Levulinic Acid: Inexpensive Co-substrates for the Fermentative Synthesis of Short-Chain Poly(hydroxyalkanoate) Biopolymers
Richard D. Ashby1*, Daniel K. Y. Solaiman1, Gary D. Strahan1, Chengjun Zhu2, Ryan C. Tappel2, and Christopher T. Nomura2
1

Eastern Regional Research Center, Agricultural Research Service, U. S. Department of

Agriculture, 600 East Mermaid Lane, Wyndmoor, Pennsylvania 19038 USA

2

Department of Chemistry, State University of New York College of Environmental

Science and Forestry (SUNY-ESF), 1 Forestry Drive, Syracuse, New York 13210 USA

137

Appendix V. Manuscript will be sent for publication in the journal Applied and Environmental Microbiology (2011)

Identification and characterization of a biosynthetic pathway to produce medium-chain-length polyhydroxyalkanoates (MCL-PHAs) from unrelated carbon sources
Qin Wang , Ryan C. Tappel , Chengjun, Zhu , and Christopher T. Nomura * Department of Chemistry and Department of Biology , State University of New YorkCollege of Environmental Science and Forestry, Syracuse, NY, 13210
1 2 1 1 2 1

138

Curriculum Vitae

Chengjun Zhu
202 Illick Hall SUNY College of Environmental Science and Forestry 1 Forestry Drive Syracuse, NY 13210 Tel: 315-470-4843 Email: czhu05@syr.edu

EDUCATION: Ph.D, Microbiology

State University of New York--College of Environmental Science and Forestry, Syracuse, NY Completion date Aug. 2011. Thesis title: Fermentation and characterization of polyhydroxyalkanoates from renewable feedstocks by Burkholderia cepacia. 2007-present

Graduate Courses taken in Biogeochemistry

University of Oregon, Sept. 2006-Dec.2006

Graduate Studies in Environmental Biology

The Institute of Environmental Health, Nanjing University, China 2003-2006

B. S., Environmental Biology

Nanjing University, China 1999-2003

SKILLS:
Bacterial fermentations of biodegradable plastics (polyhydroxyalkanoates, PHAs) from lab scale (1 L and 7 L fermentors) to pilot scale (400 L fermentor), solvent extraction of PHAs, mechanical cell disruption, characterization of physical and mechanical properties for the polyester, biological production of the copolymers, optimization of bacterial growth. H & 13C NMR, DSC (Differential Scanning Calorimetry), TGA (Thermogravimetric Analysis), GPC (Gel Permeation Chromotography), GC, HPLC, microfluidizer, tensile testing, injection molding.
1

139

DNA extraction, PCR (Polymerase Chain Reaction), ligation of PCR products with vector, preparation of competent cells, transformation of the vector into competent cells. Biological treatment of petrochemical and pharmaceutical wastewater, toxicological evaluation, toxicity tests using mice

HONORS / AWARDS:

Eastman Chemical Company Graduate Student Award at SUNY-ESF, 2010 SUNY-ESF Student Travel Grant to the 2010 International Chemical Congress of Pacific Basin Societies (Pacifichem) in Honolulu, Hawaii Student Travel Award from American Society for Microbiology General Meeting in San Diego, CA 2010 SUNY-ESF Student Travel Grant to ASM General Meeting in San Diego, CA 2010 The Fourth Annual ESF International Student Award, SUNY-ESF 2009 The Excellent Graduate Fellowship of Nanjing University, China 2004. The Outstanding Paper in the 7th Annual Seminar of Jiangsu Association for Microbiology, China 2004.

PUBLICATIONS:

1.

2. 3. 4. 5.

Chengjun Zhu, Christopher T. Nomura, Arthur J. Stipanovic and James P. Nakas. Effects of nucleating agents on the mechanical properties of poly-3-hydroxybutyrate and poly-3hydroxybutyrate-co-poly-3-hydroxyvalerate. To be submitted to Polymer Degradation and Stability, August, 2011. Chengjun Zhu, Christopher T. Nomura, Joseph A. Perrotta, Arthur J. Stipanovic and James P. Nakas. Production and characterization of poly-3-hydroxybutyrate from biodiesel-glycerol by Burkholderia cepacia ATCC 17759. Biotechnology Progress, 2010,26 (2):424-430 Zhu Chengjun, Cheng Shupei. Effects of Added Factors on Fhhh Strain in Wastewater Degradation. Journal of Nanjing University (in Chinese),2004,40: 349-355 Zhu Chengjun, Cheng Shupei. PTA Wastewater Molecular Toxicity Detected with Gene Chip. Journal of Environment Science , 2006, 18 (3): 514-518 J. Yan, S. P. Cheng, X. X. Zhang, L. Shi and C. J. Zhu. Effects of Four Metals on the Degradation of Purified Terephthalic Acid Wastewater by Phanerochaete chrysosporium and Strain Fhhh. Bull. Environ. Contam. Toxicol. 2004, 72(2): 387-393

140

PATENTS:

1. James P. Nakas, Chengjun Zhu, Joseph A. Perrotta, Christopher T. Nomura. Methods for producing polyhydroxyalkanoates from biodiesel-glycerol. Provisional U.S. Patent filed November, 2010. 2. S. P. Cheng, X. X. Zhang, L. Shi, J. Yan, C. B. Hao, B. Wang, C. J. Zhu. NJU- Ebis3 Security Control Software of Effluent Wastewater V 1.0, Copyright Protection Center of China, Copyright N.O.: 2003SR4152 3. S. P. Cheng, X. X. Zhang, L. Shi, C. B. Hao, J. Yan, C. J. Zhu. NJU-Ebis4 Regulation Software of Wastewater Treatment Technics V 1.0, Copyright Protection Center of China, Copyright N.O.: 2003SR4153 4. S. P. Cheng, X. X. Zhang, S. L. Sun, G. Z. Wei, H. F. Yu, L. Zhang, C. J. Zhu. The Special Strain and Construction Method of Organnic Pharmacy Wastewater, State Intellectual Property Office of P. R. China, 2004

141