Vol. 3

POLY(3-HYDROXYALKANOATES)

525

POLY(3-HYDROXYALKANOATES)

Introduction

Poly(3-hydroxyalkanoates) (PHAs) are a family of polyesters accumulated as in-
clusion bodies (Fig. 1) by a wide variety of bacteria. They are water-insoluble,
relatively resistant to aqueous hydrolysis but are readily biodegraded in any nat-
ural environment where microbial diversity exists. They can be produced from re-
newable resources and waste materials and, although temperature-sensitive, are
potentially recyclable. Many PHAs have mechanical properties similar to those of
common synthetic plastics.

Most PHAs contain exclusively 3-hydroxycarboxylate units (Fig. 2a)

but many microorganisms can also accumulate a polymer containing 4-
hydroxycarboxylate monomers [eg 4-hydroxybutyrate (4HB), where p

= 2 and

R

= H] when fed the appropriate substrate. The most common PHA is poly(3-

hydroxybutyrate) homopolymer (PHB, where R

= CH

3

). This was discovered and

studied by Lemoigne during the 1920s (1). It was first suspected that these gran-
ules were lipid bodies as they were stained with hydrophobic dyes. Lemoigne
proved that they were high molecular weight polyesters, years before the pioneer-
ing work of Carothers (2) on synthetic polyesters.

Although copolymers of 3-hydroxyalkanoates were identified as early as the

1960s (3), little importance was attributed to them. The term PHA only came into
use during the 1980s to refer to PHB and the other 3-hydroxyalkanoate polymers.
This definition had not been in use very long when copolymers containing 4-HB
(4) or 5-hydroxyvalerate (5-HV) (5) were discovered. Their production involved the
same enzymes as those for the synthesis of PHB but they contained units that had
the hydroxyl group on the fourth or fifth carbon rather than the third. For this
reason, the term PHA is now most commonly defined as polyhydroxyalkanoate.
This also includes polymers containing units with double bonds and other reactive
groups in their side chain. The term PHA is generally applied to the biologically
synthesized polymers.

Two major classes of PHAs are known to exist. PHAs whose polymeric

units are predominantly five carbons or less are classified as short-chain-length
PHAs (SCL PHAs) while all others are referred to as medium-chain-length PHAs
(MCL PHAs). The term MCL was coined because the number of carbons in the
monomers roughly corresponds to those of medium-chain-length carboxylic acids.
PHA nomenclature may still be in a state of flux as new structures continue to be
discovered.

Encyclopedia of Polymer Science and Technology. Copyright John Wiley & Sons, Inc. All rights reserved.

526

POLY(3-HYDROXYALKANOATES)

Vol. 3

Fig. 1.

PHB inclusion bodies in Alcaligenes latus (electron micrograph courtesy of Dr. G.

Braunegg, Technical University of Graz, Graz, Austria).

Fig. 2.

Typical PHA structures including (a) General structure of polyhydroxyalkanoate

subunits, P(HA), (b) P(3HB-co-4HB) copolymer, and (c) P(3HB-co-3HP) copolymer.

Properties

PHB exists in an amorphous state inside the bacterial cell. This seems to be
necessary for its biosynthesis and there are several theories explaining how this
state is maintained (6,7). Melting or solvent casting usually results in highly crys-
talline PHB because of its stereoregular nature. Crystallization may be avoided if
the granules are isolated from the bacterial cells by a gentle enzymatic method,

Vol. 3

POLY(3-HYDROXYALKANOATES)

527

which results in a stable latex suspension (6), but most PHB products are highly
crystalline in nature. The rate of crystallization after melt processing can be in-
creased by adding nucleating agents such as boron nitride, and by allowing the
product to stabilize at the optimal crystallization temperature (about 60

◦

C). The

main drawbacks against the use of PHB are its brittle nature and its tendency to
degrade near its melt temperature. Inclusion of a plasticizer (often a biodegradable
one, such as triacetin) and annealing after initial crystallization aid in increas-
ing flexibility. It has been demonstrated that impurities significantly increase the
degradation rate (8). Thus a short processing time in the melt and a highly pure
grade of polymer is recommended for PHB processing.

Effect of Side-Chain Length on PHA.

The use of copolymers is another

way to avoid degradation while increasing flexibility during melt processing of SCL
PHA. Increasing the HV content of poly(3-hydroxybutyrate-co-3-hydroxyvalerate)
[P(HB-co-HV)] from 0 to about 30% (expressed as mol% in this article) results in a
concommitant decrease in melting point and Young’s modulus (Table 1) but crys-
tallinity remains high as a result of a phenomenon known as cocrystallization or
isodimorphism (17). Above 30% HV the melting point begins to rise again. Increas-
ing 4HB or 3-hydroxypropionate (3HP) monomers also lowers the melting point
and gives a more flexible polymer. However, crystallinity decreases with increas-
ing 4HB and 3HP since isodimorphism does not occur in these copolymers (18,19).
Both 4HB and 3HP have much lower melting and glass-transition temperatures.

Conventional equipment used to process polyethylene can be used for pellets

of PHA plasticized with a nucleating agent to form blow-molded, injection-molded,
and film products. A key requirement is temperature control. It is recommended
to operate at the lowest possible temperature and to keep the residence time at a
minimum since the melting point of PHB is about 178

◦

C and thermal degradation

begins at about 180

◦

C (depending on the purity). P(HB-co-HV) with a high HV

content can have a melting point as low as 136

◦

C. For most applications, it is

suggested to process at a maximum temperature of 170

◦

C, a maximum residence

time of 3 min, and that molds and blow pin temperatures be kept at about 60

◦

C

(20). Mechanical properties deteriorate rapidly if the weight-average molecular
weight falls below 400,000 g/mol.

Homopolymers of MCL PHA have not been biosynthesized to date. Although

polymers designated as poly(3-hydroxyoctanoate) (PHO) have predominantly
3-hydroxyoctanoate repeating units (Fig. 2a where P

= 1 and R = C

5

H

11

),

they contain significant quantities (10% or greater) of other monomers, usually
3-hydroxydecanoate and 3-hydroxyhexanoate, depending on the bacterium and
the carbon source. Melting temperature, Young’s modulus, and tensile strength
decrease as the length of the MCL PHA side chain increases (Table 1). MCL PHA
can be toughened by annealing, which results in the formation of “crystalline
cross-links.”

Polymer Blends Incorporating PHA.

The mechanical properties,

morphology, biodegradability, and thermal and crystallization behavior of PHAs
melt-blended or solvent-cast with nonbiodegradable polymers [such as poly(vinyl
acetate)] and with biodegradable materials [such as wood cellulose fibers (21)
and starch] have been reviewed (22). PHB blends with poly(ethylene oxide),
poly(vinyl alcohol), poly (

L

-lactide), poly(

D

,

L

-lactide), poly(

ε-caprolactone), poly(3-

butyrolactone), P(HB-co-HV), and cellulose and starch derivatives have been

Table 1. Properties of Various PHAs

a

Compared to Polypropylene (PP)

b

P(3HB-3HV)

d

P(3HB-3HP)

f

P(3HB-4HB)

Physical properties

PP

c

P(3HB)

c

(0–25%)

P(3HO)

e

P(3HP)

f

(0–88%)

P(4HB)

g

(0–49%)

h

Crystalline melting point,

◦

C

171 to 186

171 to 182

179 to 137

61

77

177 to 61

53

177 to 150

Glass-transition temp.,

◦

C

−10

−5 to 5

i

10 to

−6

−35

−19

4 to

−15

−48

4 to 23

Crystallinity, %

65 to 70

60 to 80

80 to 30

30

37

60 to

>13

60 to 15

Molecular weight M

w

(10

5

)

2.2 to 7

1 to 8

3

5

2

4 to 15

5 to 8

Molecular weight

5 to 12

2.2 to 3

1.4

2.0 to 3.2

distribution

Density, g/cm

0.905 to 0.94

1.23 to 1.25

1.20

1.0

Tensile strength/MPa

j

38

40

40 to 30

6 to 10

104

43 to 10

Young’s modulus, GPa

k

1.7

3.5 to 4.0

3.5 to 0.7

149

Elongation to break, %

400

6 to 8

8 to 10

300 to 450

1000

5 to 511

a

For additional properties, see Ref. 9.

b

Structures are shown in Figures 2a–c.

c

Ref. 10.

d

Ref. 11.

e

Ref. 12.

f

Ref. 13.

g

Ref. 14.

h

Ref. 15.

i

Ref. 16.

j

To convert MPa to psi, multiply by 145.

k

To convert GPa to psi, multiply by 145,000.

528

Vol. 3

POLY(3-HYDROXYALKANOATES)

529

shown to be completely biodegradable (23). In a reactive blending process, the com-
ponents have been mixed with a peroxide. This decomposes to form free radicals
which cross-link the polymers (24).

Applications

The first commercial user of the P(HB-co-HV) resin was Wella AG, Darmstadt,
Germany. They packaged hair and skin care products in injection blow-molded
bottles. These products were introduced to the American and Japanese markets
in 1991 and 1992 respectively.

Since PHAs are isotactic, they can serve as a feedstock for enantiomeric com-

pounds, which can be used in the synthesis of chiral chemicals such as antibiotics,
vitamins, fragrances, and pheromones. Optically pure monomers may be obtained
from PHAs by chemical hydrolysis at 80–160

◦

C using a titanate catalyst (25) or by

enzymatic hydrolysis using extracellular bacterial depolymerases. It is possible to
synthesize the optically active monomers by using bacteria, which can make the
polymer but lack the 3-hydroxyalkanaote polymerase (Fig. 3a) or possess a high
activity of intracellular PHA depolymerase. Monomers such as 4HB have thera-
peutic applications as an intravenous anaesthetic, for the treatment of narcolepsy,
alcohol, heroin, and nicotine addiction (26).

PHB and P(3HB-co-4HB) have been evaluated for the controlled release of

drugs (26,27). The drug may be microencapsulated in PHB and injected subcuta-
neously, or may be pressed into a pill and administered orally. Degradation is slow
in the human body as it does not contain PHA depolymerases. Aqueous hydrolysis
is probably responsible for in vivo degradation. PHB tablets of 60 mg (6.0 mm in
diameter and 2.0 mm in thickness) degraded at a rate of 0.2 mg/week in the tis-
sue of mice. In its present form, it might be considered for long-term applications
in the body. PHB has been evaluated for use in surgical devices such as sutures,
prosthetics, pins, dressings, and a three-dimensional heart valve scaffold (28). The
monomers of PHB are natural metabolites found in the human body and are not
toxic to humans. PHB and P(3HB-co-4HB) have been shown to be compatible with
human tissue and mouse fibroblast cells. These materials elicit only a very mild
immunological response in mice or rats (27).

Films of PHB and P(HB-co-HV) have piezoelectric properties. A charge is

generated if a PHB crystal is deformed by hydrostatic force, thermal expansion,
or other forces. Some synthetic piezoelectric polymers such as poly(vinylidene
fluoride) are known to stimulate bone growth. It has been suggested that films of
PHB and P(HB-co-HV) wrapped around bone fractures may accelerate the healing
process (27).

PHAs are moisture-resistant and have gas barrier properties similar to the

best coated films. They have potential use as plastic moisture barriers. Possible
markets include mulch films, food and agricultural packaging, drink cartons, fem-
inine hygiene products, disposable diapers, and disposable kitchen items, such as
plastic films which are difficult to separate for recycling. Coatings may be ap-
plied with PHA as an amorphous latex which can be either reconstituted from the
crystalline form of the polymer or obtained directly from a bacterial suspension
by enzymatically dissolving the other cellular components. In the latter case, it

530

POLY(3-HYDROXYALKANOATES)

Vol. 3

Fig. 3.

(a) Biosynthesis of SCL PHA; 1, 3-ketothiolase; 2, NADPH-dependent acetoacetyl-

CoA reductase; 3, SCL PHA polymerase; 4, SCL PHA depolymerase; 5,

D

(

−)-3-

hydroxybutyrate-dimer hydrolase. (b) Biosynthesis of P(HB-co-HV), 1, 3-ketothiolase; 2a,
NADPH-dependent acetoacetyl-CoA reductase; 2b, NADH-dependent acetoacetyl-CoA re-
ductase; 3, SCL PHA polymerase; 4, fatty acyl-CoA dehydrogenase; 5, enoyl-CoA hydratase.

Vol. 3

POLY(3-HYDROXYALKANOATES)

531

is important that the granules are not dried prior to use as they would become
crystalline. The latex may be applied as a spray to a material such as paper, dried
and cross-linked using heat or uv light. The MCL PHAs are promising as coatings
(29) since they are less crystalline than the SCL PHAs, and the microorganisms
which synthesize them can introduce different functional groups (30).

Degradation of PHA

SCL PHA has been shown to degrade as rapidly as cellophane, much more rapidly
than poly-

L

-lactic acid and aliphatic polyesters such as poly(butylene succinate)

and poly(butylene succinate-co-ethylene succinate) under anaerobic conditions
(31), faster than polycaprolactone under aerobic conditions, and faster than paper
during composting (32). Although PHAs degrade rapidly in most nonsterile natu-
ral environments, they are much less susceptible to abiotic hydrolysis than most
other “environmentally degradable” polymers. For example, although P(HB-co-
HV) is less readily hydrolyzed under aseptic conditions (33), it generally degrades
much more rapidly than polycaprolactone in septic environments.

PHA Biodegradability.

One of the attractive features of PHAs products

is that they can be biodegraded in a variety of environments including compost,
soil, and sea water, under aerobic or anaerobic conditions. The rate of degradation
in nature depends on environmental factors such as pH, temperature, moisture
level, quantity of microorganisms capable of degrading PHAs, nutrients other
than a carbon source, and on characteristics of the PHA material such as its
thickness, surface area, molecular weight, crystallinity, and the presence of other
components such as a filler. Standard methods for the determination of biodegrad-
ability have been developed for international standards organizations. For exam-
ple, ISO/FDIS14855 is the compostability testing method. It is identical to ASTM
D5338 and to the DIN (German) Standard.

Enzymatic Degradation of PHAs.

The extracellular and intracellular

depolymerases for the degradation of PHAs function differently. In both cases, the
substrate is PHA but there is a difference in crystallinity. PHA granules inside the
cell are amorphous (the native form). However in extracellular degradation, PHA
has been removed from the cell and has been “processed,” becoming a more crys-
talline material with both amorphous and crystalline regions. Intracellular de-
polymerases demonstrate low activity with semicrystalline PHA and extracellular
depolymerases poorly attack amorphous PHA. Extracellular degradation is impor-
tant in an environmental context as it would be the primary mechanism for the
disappearance of a PHA product during composting or wastewater treatment.

Intracellular Degradation.

Native PHB granules isolated from Rhodospril-

lium rubrum (34), Azotobacter beijerinckii (35), and Zooglea ramigera (36) were
shown to be self-hydrolyzing while those of Bacillus megaterium (34) were not and
those of Alcaligenes eutrophus (now called Ralstonia eutropha) did so at a very
low rate (37). Depolymerization in B. megaterium requires a heat-labile factor
associated with the granule and three soluble components: a heat-stable protein
activator, PHB depolymerase, and a hydrolase. The concerted action of the first
three components resulted in monomers and 15–20% dimers with the hydrolase
converting the dimers to monomers.

532

POLY(3-HYDROXYALKANOATES)

Vol. 3

Active intracellular depolymerases could lead to “turn over” of accumulated

PHAs in the presence of excess carbon and/or PHA catabolism under carbon lim-
itation. When R. eutropha containing PHB was placed in a nitrogen-free medium
with valeric acid as the sole carbon source, many HB units were replaced by HV
units or “turned over” (38). Similarly when cells containing P(HB-co-HV) were
placed in the same medium with butyric acid as the sole carbon source, the HV
content of the cells decreased.

The production of PHA depolymerase is repressed in most PHA-degrading

bacteria when an easily degradable carbon source is present. When the carbon
substrate is exhausted, the depolymerase synthesis is derepressed. Once the PHA
monomers are released, they can be repolymerized, serve as a source of carbon
and energy for growth or be excreted. If the microorganism is lacking the hydrox-
yalkanoic acid dehydrogenase or this enzyme has a low activity, the monomer
will accumulate as hydroxycarboxylic acids. It has been demonstrated that this
approach can be used to produce a high yield of pure enantiomers (39).

Extracellular Degradation.

Although aqueous and enzymatic hydrolyses

contribute to degradation in the environment, aqueous hydrolysis is very slow
compared to enzymatic hydrolysis. It took about 16 weeks to produce a 30% weight
loss of a P(HB-co-9%4HB) film (12 mm in diameter and 70

µm in thickness) in

0.01 M phosphate buffer at 70

◦

C and pH 7.4 (16). Because of random hydrolytic

chain scission, the number-average molecular weight (M

n

) decreased from 226,000

to 13,000 in 55 days, with little change in actual weight (8%). At day 110, the extent
of weight loss had increased to 30%. Degradation (usually monitored as weight
loss) is typically much more rapid in the presence of active microorganisms. For
example, a 75-

µm-thick film completely degraded in anaerobic sewage in 1 week,

in a well-watered soil at 25

◦

C in 12 weeks, and in an estuarine sediment at 20

◦

C

in 8 weeks (40).

An initial sequence of events for a PHB film buried in a well-watered soil

has been described (40). The first step is a wetting-out process, in which aqueous
hydrolysis is probably responsible for the increase in the number of hydroxyl and
carboxylic acid groups at the surface. After 1 week, the plastic surface becomes
more hydrophilic. Bacterial and fungal colonization then occurs. Microbial attack
(primarily due to extracellular enzymes) takes place at the surface (41,42) which
becomes more pitted, increasing the surface area for degradation. Surface area is
further increased by blending with a more rapidly biodegradable polymer such as
starch. In a P(HB-co-19%HV)–starch composite, the degradation rate increased
with the starch content (43). Starch was lost more rapidly than the P(HB-co-
19%HV) starting at both surfaces, then toward the center. As the starch was
removed, a larger internal surface area was available for PHA degradation. At
least 31vol% starch granules is required to have continuity of all internal surfaces
in a polyethylene–starch composite (44).

The mechanism of extracellular degradation of PHA is not yet well under-

stood. A wide variety of microorganisms (fungi, aerobic and anaerobic bacteria)
have been implicated but only Pseudomonas lemoignei, Alcaligenes faecalis, and
Comamonas sp. have been studied in detail. Production of extracellular PHA de-
polymerases is induced by the presence of PHA as the sole carbon source. Gen-
erally, a PHB depolymerase acts from the hydroxyl terminus to form dimers (or
trimers, depending on the depolymerase) and a trace amount of monomer. An

Vol. 3

POLY(3-HYDROXYALKANOATES)

533

oligomer hydrolase further hydrolyzes the dimers (or trimers but not the poly-
meric material) to monomers which can then be used as a source of carbon and
energy. Since the depolymerase acts at the hydroxyl end (sequentially removing
monomers, dimers, or trimers), there should be a very slow change in the M

n

. The

rate of weight loss should not depend on the initial molecular weight and it has
been shown that high and low molecular weight PHBs were indeed degraded at
the same rate (41). There is no significant change in the molecular weight of the
bulk of a film or blow-molded product as degradation takes place at the surface.

Although extracellular depolymerases act on semicrystalline PHA, the

degradation rate has been shown to decrease with increasing crystallinity of PHB
films (45). For a blow-molded, P(HB-co-16%HV) product (a 20-

µm cross section

was exposed to partially purified PHB hydrolase from Penicillium funiculosum),
the degradation rate is faster at the less crystalline surface than at the more crys-
talline core (42). Furthermore, it has been shown that the amorphous regions on
the surface of PHB films are degraded preferentially whereas microbial coloniza-
tion takes place at spherical holes which tend to be at the crystal centers and at
the boundary between spherulites (46).

The enzyme specificity depends to some extent on the composition of the

PHA which induced the degradation. In general, enzymes induced by SCL PHAs
will not act on MCL PHAs and those induced by MCL PHAs will not act on SCL
PHAs. The specificity based on the composition of the PHA inducer is not as clear
within each group of depolymerases. P(HB-co-HV) was degraded more slowly than
PHB, and P(4HB) was degraded more quickly than PHB (41) by PHB-induced
depolymerases. However, PHB, PHV, and P(HB-co-HV) were degraded at com-
parable rates by purified PHV depolymerase (PHV as sole carbon source) from
P. lemoignei (47). It has been suggested that P(4HB) degraded faster because
4HB monomers were more sterically accessible to the enzyme than were 3HB
monomers (41).

PHA Synthesis

Synthetically, PHB is produced from racemic 3-butyrolactone and P(HB-co-HV)
is made from a mixture of 3-butyrolactone and 3-valerolactone, with triethyla-
luminum and water as catalyst (48). Although the stereospecificity of the re-
sulting polymer can be controlled to a certain extent by adjusting the ratio of
triethylaluminum and water, synthetic PHAs are usually optically inactive and
hence only partially biodegradable.

SCL PHA Biosynthesis.

During PHB synthesis, 3-ketothiolase cat-

alyzes the condensation of two acetyl-CoAs to produce acetoacetyl-CoA which
is stereospecifically reduced by an independent reductase to produce

D

(

−)-3-

hydroxybutyryl-CoA (35,49) (Fig. 3). First elucidated in A. beijerinckii, the pres-
ence of this synthetic route was confirmed in R. eutropha. These four-carbon
monomers are then added to the growing PHB chain by a polymerase. If pro-
pionic acid is supplied during the fermentation, some of the acid is converted to
proprionyl-CoA. A portion of this proprionyl-CoA is condensed with acetyl-CoA via
3-ketothiolase to produce five-carbon monomers (Fig. 3b). This is the most com-
mon pathway of P(HB-co-HV) production. Most PHB-accumulating bacteria will

534

POLY(3-HYDROXYALKANOATES)

Vol. 3

produce P(HB-co-HV). Since acetate is needed for this synthetic route and some
propionic acid is decarboxylated before monomer synthesis, usually less than 50%
of the monomers will be HV even if a feed of pure propionic acid is added. This yield
can be increased in R. eutropha by decreasing the dissolved oxygen concentration
in the bioreactor during the accumulation phase (50).

High yields of HV are obtained when pentanoic acid is used because acetyl-

CoA is no longer required since pentanoic acid is metabolized directly to

L

-(3)-

hydroxyvaleryl-CoA which is subsequently racemized (at least in R. eutropha)
to

D

-(3)-hydroxyvaleryl-CoA (51) (Fig. 3b). Although some acetate is produced,

leading to HB monomer formation, PHAs with greater than 90% HV content
can be obtained. SCL PHA producers have not been reported to accumulate
monomers with side chains longer than C

2

or C

3

but can produce P(3HB-co-

4HB) when fed 4-hydroxybutyric acid or 1,4-butanediol. R. eutrophus accumu-
lates terpolyesters of 3HB, 3HV, and 5HV monomers when fed a mixture of
pentanoic and 5-chloropentanoic acids. Even among SCL-accumulating microor-
ganisms, there may be differences in the PHA synthetic pathway. For example,
in R. rubrum,

L

-(

+)-3-hydroxybutyryl-CoA is racemized to

D

-(

−)-3-hydroxybutyrl-

CoA using enoyl-CoA hydrases.

In most SCL PHA-accumulating bacteria, high intracellular levels of NADH

and acetyl-CoA greatly stimulate the rate of PHA synthesis. Thus, if the growth
rate is limited by a nutrient other than the source of carbon and energy, PHA
is accumulated in the cell. By definition, the PHB content of the cells is stable
under balanced growth conditions, while carbon and energy limitations increase
depolymerase activity, leading to a decrease in cellular PHA content.

MCL PHA Biosynthesis.

MCL PHA was first identified in mixed cultures

in activated sludge (52) and marine sediments (53). The unusual manner in which
these materials crystallized led to some debate as to whether they were copolymers
or mixtures of homopolymers (54). The breakthrough in MCL PHA technology
came when it was discovered that Pseudomonas oleovorans would accumulate
MCL PHA when fed with the appropriate alkane, alkanoate, or alkanol (30,55,56).
Shortly thereafter, several other ribosomal RNA homology group I pseudomonads
were found that could also accumulate MCL PHA from the above substrates. Many
pseudomonads can produce MCL PHA from simple carbohydrates such as glucose
(57,58) although the yield is usually much lower. This led to the belief that MCL
PHA could be derived from fatty acid synthesis as well as fatty acid degradation
(59).

In fatty acid synthesis or degradation (Fig. 4), once the substrate is in the

form of alkanoyl-CoA, a two-step process involving a 3-ketoacyl-CoA reductase
and a PHA synthase occurs. The reduction step requires NADPH. A suitable sup-
ply of intracellular substrate appears to be the key factor controlling the rate of
biosynthesis. If P. oleovorans is fed octanoate, it accumulates much more PHA
than if hexanoate or dodecanoate are added. Nevertheless, the synthase is rela-
tively nonspecific. P. oelovorans can accumulate polymers in which the side chain
of the monomers contain double bonds, branches, cyclic structures, or even halogen
atoms if the appropriate substrate is supplied.

Much less polymer is produced from carbohydrates (ie fatty acid synthesis)

than from alkanes or related substrates (ie fatty acid degradation). This is pos-
sibly due to a shortage of acceptable substrate for the synthase. Pseudomonads

Vol. 3

POLY(3-HYDROXYALKANOATES)

535

Fig. 4.

Biosynthesis of MCL PHA. Substrate may result from fatty acid degradation, or

from fatty acid synthesis. The substrate must be transferred from the acyl carrier protein
(ACL) to Coenzyme A (CoA) if derived from fatty acid synthesis.

that produce PHA from glucose, accumulate a polymer where 3-hydroxydecanoate
units dominate. They also frequently use 3-hydroxydecanoate as an intermediate
metabolite for the synthesis of other materials such as biosurfactants (60). Block-
ing these pathways could result in increased polymer production.

The intracellular PHA depolymerase system can act very efficiently in MCL

PHA producers. Carbon limitation often leads to extremely rapid depolymeriza-
tion. The diversity of metabolic machinery among PHA-accumulating microor-
ganisms indicates that unusual and potentially valuable PHA copolymers await
discovery.

Synthesis in Transgenic Plants.

PHB can be synthesized in plants in

which the bacterial reductase and polymerase genes have been introduced. Al-
though virtually all organisms possess the third gene in the SCL PHA pathway,
coding for 3-ketothiolase, its activity and specificity vary, and it may not be present
in all eucaryote organelles. Production is usually increased by cloning the 3-
ketothiolase gene with the other two. Although transgenic plants are easily cre-
ated, productivity remains problematic. The SCL PHA pathway was first cloned
into Arabidopsis thaliana, the Escherichia coli of plant biotechnology (61). The
yield was low, 0.1% dry weight, and the plants were stunted. However, granules
similar to those found in bacteria were produced. The initial cloning strategy tar-
geted the cytoplasm and granules were also found in the cell nucleus. Subsequent
work has targeted the plastid which is also the site of starch and oil synthesis
in plants. This has resulted in increased yields of up to 14% dry weight in the
leaves of A. thaliana. These plants grew well but were chlorotic probably because
of large amounts of PHB in the chloroplasts. Present research centers on oil and
starch producing plants since these are already adapted to have large amounts of

536

POLY(3-HYDROXYALKANOATES)

Vol. 3

storage material in the plastids (62), but other strategies are also being assessed.
Although MCL PHA has been produced in transgenic plants (63), yield is low, par-
tially because of a lack of understanding of the link between fatty acid synthesis
and MCL PHA synthesis.

Separation of PHA from Biomass

All separation processes to date have been developed for microbial rather than
plant biomass but many of these approaches could also be applied to recovery
of PHA from plants. As a first step in the recovery from microbial biomass, the
cells are separated from the fermentation broth usually by centrifugation or floc-
culation. In the latter case, the pH of fermentation broth is adjusted to 9 and
the broth is heated; the pH is readjusted to 5 (64), and the biomass is recovered
by centrifugation or filtration. The wet biomass may be freeze- or spray-dried to
remove water. Processes to separate the intracellular PHA from the rest of the
microbial biomass may be divided into two categories: (1) selective solubilization
of PHA in an organic solvent, leaving the majority of the other biomass compo-
nents in suspension and (2) destruction (eg enzymatic solubilization) of cellular
components other than PHA, leaving the PHA granules in suspension. Physical
treatments such as drying, heating, freezing and thawing, repeated centrifuga-
tion, and exposure to acids and organic solvents may result in crystallization of
native PHA.

Solvent Extraction of PHAs.

The majority of the patented separation

processes describe the extraction of PHB from microbial biomass using organic
solvents such as chlorinated hydrocarbons (eg chloroform or 1,2-dichloroethane),
azeotropic mixtures [eg 1,1,2-trichloroethane with water (65)]; chloroform with
methanol, ethanol, acetone, or hexane (66), and cyclic carbonates [eg hot (120–
150

◦

C) ethylene carbonate or 1,2-propylene carbonate (67)] in which the polymer

is soluble. With the wider variety of PHAs which can be produced today, the choice
of solvents should be carefully considered. In general, solvents which are suitable
for PHB should be equally good for any SCL and MCL PHA. The reverse may
not be true. For example, while semicrystalline PHB is insoluble in acetone, MCL
PHA will dissolve in it.

Cells may be pretreated to make the polymer more accessible to the ex-

traction solvent. This may involve washing in a hydrophilic solvent in which the
polymer is insoluble (such as methanol or acetone for SCL PHA and methanol for
MCL PHA), or physical disruption such as grinding, wet-milling, French press, or
freezing and thawing cycles. Pretreatment with acetone has been shown to have
the added advantage of removing certain polar lipid impurities and increasing the
amount of PHB recovered (68).

After solvent extraction, the polymer solution is separated from the cel-

lular residue by filtration or centrifugation. PHB is then precipitated from the
chlorinated hydrocarbons or azeotropic mixtures with a cold nonsolvent, such as
methanol. In the cases of ethylene carbonate and 1,2-propylene carbonate, the
majority of the PHB precipitates on cooling (67).

PHA solutions become too highly viscous to process easily at concentrations

of greater than 5 wt%. As a result, large amounts of solvent are required to obtain
sufficiently dilute solutions to enable easy separation. A significant portion of

Vol. 3

POLY(3-HYDROXYALKANOATES)

537

the PHB may remain unextracted in the biomass (68). Solvent extraction is not
considered economical unless cheap solvent and energy for solvent recovery are
available. However, solvent extraction is useful when the PHA content of the
biomass is low or when high purity is required.

Although thermal degradation of PHB in air or a nitrogen atmosphere has

been well studied (69), there are few examples which show that PHB degrades
in organic solvents. The molecular weight is known to decrease in propylene car-
bonate and ethylene carbonate at 110–140

◦

C (67), in chloroform–methanol mix-

tures at room temperature (70), and in methylene chloride, chloroform, and 1,2-
dichloroethane at their boiling points (68).

Solvent extraction is a commonly used laboratory technique for PHA re-

covery, especially for molecular weight determination. Usually, methanol- (or
ethanol-) pretreated, lyophilized biomass is heated in chloroform under reflux
conditions for about 1 h. The polymer solution is recovered by filtration or centrifu-
gation, and the polymer is precipitated in cold ethanol or methanol. The polymer
may be further purified by redissolving in chloroform and reprecipitating in cold
ethanol. It should be noted that about 25% of the “original” molecular weight (de-
termined by chloroform extraction under ambient conditions) is lost when refluxed
in chloroform for 1 h (68).

Removal of Cellular Components Other Than PHA.

Enzymatic Methods.

Classical enzymatic treatment (involving lysozyme,

proteinases, DNAses, etc) with or without a surfactant step can be used to remove
the non-PHA biomass (71). Biologists have used these methods to prepare native
PHB granules to study intracellular enzymatic degradation. On a commercial
level, a process in which the fermentation broth may be first heated to 80

◦

C to

denature DNA and some proteins has been described (72). A series of enzymes
[such as lysozyme, phospholipase, lecithinase, or alcalase (proteinase)] are then
used to obtain a product which is 90–95% pure with 6–7% proteins and some
peptidoglycan as major impurities. There may be a final wash with hydrogen
peroxide to whiten the product. This was the method used for the production of
P(HB-co-HV) (73), which was sold under the trade name Biopol by Zeneca, and
later by Monsanto. When PHA granules recovered by enzymatic treatment are
freeze-dried, they have a hard, crystalline shell and an amorphous core (6).

Chemical Methods.

Strong oxidizing agents such as sodium hypochlorite

at pH 10 have been used to dissolve the cellular biomass under conditions which
result in significant loss of PHA molecular weight. However, by optimizing the
operating conditions, PHB with a weight-average molecular weight of 600,000
g/mol and 95% purity has been obtained from R. eutropha biomass containing
50% PHB with an initial molecular weight of 1,200,000 g/mol. When the biomass
is pretreated with a surfactant, less contact time with hypochlorite is needed to
obtain PHB with the same purity. This allows an even higher molecular weight
(800,000 g/mol) to be obtained (74). Never-dried granules obtained by surfactant-
hypochlorite treatment also have an external crystalline shell and an amorphous
core (75). Similar PHB recovery, molecular weights, and purity may be obtained
when sodium hypochlorite is omitted and only an NaOH solution at pH 13 is used.

Physical Methods.

Although there are many methods (eg reciprocating

pump, ball mill) to physically disrupt microbial cells, there have been few
processes which are based on these principles proposed for the large-scale
recovery of PHAs. A process in which a biomass suspension was heated to 220

◦

C

538

POLY(3-HYDROXYALKANOATES)

Vol. 3

under nitrogen in an autoclave and then forced through a fire-jet into a tank of
cold water has been patented (76). Cell breakage occurred, releasing PHB gran-
ules which were suspended in 50% aqueous acetone to remove lipids and to cause
flocculation.

Commercial Production

PHB remained an academic curiosity until the fifties and sixties when W.R. Grace
and Co. in the United States produced small quantities for commercial evaluation.
They obtained patents for production processes and manufactured articles such
as sutures and prosthetic devices (77) but eventually abandoned the project be-
cause of low PHB yields and difficulties in the separation process. About 10 years
later, Imperial Chemical Industries (ICI) began a program in PHA production and
product development. They patented (78) the first commercial process to produce
SCL PHA copolymer (P(HB-co-HV)).

Choice of Production Strain and Process.

The most important eco-

nomic considerations are substrate costs, reactor productivity, and separations
costs. SCL PHA can be produced from hydrolyzed starch (glucose), lactose (recom-
binant E. coli on cheese whey), cane or beet molasses (Azotobacter vinelandii or
Alcaligenes latus on sucrose), waste wood (fermented hydrolyzed hemicellulose),
methanol, and even mixtures of hydrogen and carbon dioxide (R. eutropha). Addi-
tion of an appropriate cosubstrate such as propionic acid results in P(HB-co-HV)
synthesis in any of these processes although the percentage of HV can vary con-
siderably depending on the bacterial strain. High productivity has been reported
for several of these substrates. Using R. eutropha growing on glucose, 164 g/L of
biomass containing 76% PHB has been produced in a fed-batch process in only 50 h
(79). Densities well in excess of 100 g/L have been reported for fed-batch processes
using E. coli strains with multicopy plasmids containing the key SCL PHA syn-
thetic genes (80) as well as for methylotrophs growing on methanol (81). In re-
combinant E. coli strains, PHB content is frequently reported as being in excess
of 90% of the biomass dry weight, making the separation process much simpler.

MCL PHA.

The highest productivity results from using octanoic or

nonanoic acids, but these are toxic at concentrations of several grams per liter
and their low solubility may limit the growth rate because of poor mass transfer.
Sodium salts of these acids are much more soluble but the final product concen-
tration in fed-batch culture is limited by the accumulation of toxic concentrations
of sodium ions in the medium. In any event, the present cost of both the salts and
the acids prohibit their use in commercial production. Most microorganisms which
make MCL PHA produce lipases and grow well on vegetable oils but the lipase con-
centration may limit the production rate. A Pseudomonas putida strain has been
grown in fed-bath culture from oleic acid (a fatty acid commonly found in plant
oils) to a biomass concentration of 92 g/L containing 45% PHA in 26 h (82). Because
of their low density, these PHAs float to the surface after breakage of the bacterial
cell and can thus be easily and cheaply separated from the rest of biomass.

Commercial Production Methods.

Tredegar Industries has used re-

combinant E. coli to produce over 100 g/L of biomass containing at least 80%
PHB by dry weight at the pilot-scale level. Mild detergent treatment and washing

Vol. 3

POLY(3-HYDROXYALKANOATES)

539

steps followed by spray drying result in a white product which is 98% pure (83).

Fermentation and separation processes were developed for the pilot-scale

production of PHB using an A. latus strain as well as its further processing into
consumer products. Up to 1 ton/week of PHB was produced in Austria in a 15,000-L
bioreactor (84). This technology is currently owned by the German company
Biomer. The Biomer fermentation process is relatively simple with multistaging
from the petri dish to a shake flask to a small fermentor which is then used to
inoculate the production reactor (Fig. 5a). If sufficient carbon source (molasses)
is added during the fed-batch process, up to 60 g/L of PHB may be obtained. The

Steam

Strain

Shaker flask

Fermentor I

Fermentor II

Seperation

unit

Seperation

unit

Concentrated

cell suspension

Washing unit

Water

Medium

(a)

Solvent recovery

Extraction vessel

Cell separation

Drying centrifuge

Dryer

PHB

Extracted
cells

Precipitation
vessel

Concentrated

cell suspension

(b)

Fig. 5.

(a) Biomer PHB fermentation process. (b) Biomer PHB separation process.

540

POLY(3-HYDROXYALKANOATES)

Vol. 3

resulting biomass is washed in water. The concentrated cell suspension is then
subjected to methylene chloride extraction (Fig. 5b). The resulting slurry is sepa-
rated in a decanter. The solvent phase is then injected into hot water in a precip-
itation vessel where methylene chloride evaporates and PHB precipitates. After
water removal (drying), PHB can then be processed into a marketable product.

Biopol [P(HB-co-HV)] has been produced from glucose and propionic acid in

35,000-L airlift reactors and in up to 200,000-L stirred tank reactors (85). Phos-
phate limitation (after about 60 h of growth) is used to trigger the accumulation
phase in a strain of R. eutropha. Phosphate limitation was found to give better
results than nitrogen limitation (86). Although solvents (chloroform or methylene
chloride extraction after a hot methanol wash) may be used when high purity is
required, the P(HB-co-HV) is usually obtained by treatment of the biomass with
a series of surfactants, enzymes, and possibly oxidizing agents to dissolve the
non-PHA biomass. The product is essentially a latex that may then be washed,
spray-dried, compounded with nucleating agents and plasticizers, and extruded in
the form of pellets. Until recently, several grades of Biopol were available at a cost
of about 17.60/kg. Monsanto reported that Biopol production costs were 25–30%
greater than those of polyolefins and has discontinued production.

Copersucar (Sao Paulo, Brazil) currently manufactures up to 5 tons of PHB

per month in a fed-batch process using a 20,000-L fermentor equipped with a
Frings aeration system. A Burkholderia sp. or R. eutropha mutant lacking PHA
depolymerase is grown on inverted sugar and mineral salts. The PHB accumula-
tion phase is triggered by phosphate limitation. The PHB is extracted in a mixture
of boiling alcohols (a waste from ethanol distillation) under reflux. The cellular
debris is removed and the polymer precipitates on cooling. Solvent is recovered by
injecting the solvent/PHB paste on boiling water under vaccum. Copersucar may
be able to sell PHB much more cheaply than its competitors as it has access to
inexpensive carbon (sucrose) and energy (bagasse) sources.

Future Prospects

Cost of production is not a major consideration if PHA is to be used exclusively
for specialty applications. However, if PHA is to be economically competitive with
synthetic plastics, it must be produced on a far larger scale than any other aer-
obically produced microbial product. An additional problem arises from the fact
that PHAs are intracellular products. Thus the amount that can be produced per
cell has physical limitations.

In a typical PHA fermentation process, the cost is divided fairly evenly be-

tween substrate cost, capital, and operating expenditures in the fermentation
process and the cost of separating the PHA from the microbial biomass. Typically
the yield of dry biomass from a hexose sugar such as glucose is about 42% for R.
eutropha. If the final biomass contains 75 wt% by weight of recoverable PHA, then
the yield of PHA from the total amount of substrate supplied is about 32%. Thus, if
carbon substrate was available at $500/t, the cost of the carbon source alone would
be $1560/t, making the product twice as expensive as bulk synthetic plastics.

New methods such as the incorporation of temperature-sensitive plasmids

for the production of autolytic enzymes should reduce separation costs. The direct

Vol. 3

POLY(3-HYDROXYALKANOATES)

541

use of PHA-containing bacteria in melt processing (ie no separation process apart
from water removal) has also been proposed. However, even assuming a negligible
cost for the carbon source by using a waste material such as cheese whey perme-
ate, and neglecting the separation costs, studies have shown that PHA produced in
bioreactors could not be economically competetive with synthetic bulk plastics in
the foreseeable future. The volumetric production rate cannot compete with chem-
ical processes and this rate is limited by the oxygen transfer rate of present day
bioreactors. New technologies must be developed for commercial PHA production.

In vitro synthesis of PHA granules from activated substrates may eventually

become commercially feasible. In R. eutropha, the maximum rate of PHA synthe-
sis occurs early in accumulation phase and deteriorates slowly thereafter. One
would therefore expect that the maximum amount of PHA accumulation would
depend on enzymatic activity or on NADPH or substrate supply. However the most
detailed published study on this process points to a physical limitation (86) where
polymer synthesis slows to a virtual stop simply because there is no more space
available. If this is indeed the case, then the way around such a limitation is to
produce the PHA granules outside the cells. This has been achieved (87,88). Com-
mercial exploitation of this process requires a source of the biosynthetic enzymes
(easily achievable through molecular biological techniques), substrates, NADP,
and a method for NADP reduction already used in commercial steroid transfor-
mation processes. This would eliminate or greatly reduce separation costs. The
present economic and technical barrier to this process is the lack of an inexpen-
sive method to link coenzyme A to the appropriate carboxylic acid.

It is likely that in the near future, starch, oilseed, or leguminous plants

will be used to produce PHAs. Other plant materials such as cellulose and starch
are already used to make plastics or plastic-like materials. Hemicelluloses have
been used directly and/or converted to organic acids for PHA production at a
laboratory scale. Thus any waste from the separation of PHA from plants could
be hydrolyzed, fermented to lactic, acetic, and propionic acids, and then fed into
bioreactors for the production of specialty PHAs. Polylactides and other polymers
of organic acids could be produced in the same plant as could polysaccharides.
Plant materials may become a major source of plastic materials in the next
century with PHAs leading the way.

BIBLIOGRAPHY

1. L. Lemoigne, Bull. Soc. Chim. Biol. (Paris) 8, 770–782 (1926).
2. W. H. Carothers, in H. Mark and G. S. Whitby, eds., Collected Papers of Wallace H.

Carothers, Interscience Publishers, New York, 1946.

3. J. Davis, Appl. Microbiol. 4, 301–304 (1964).
4. Y. Doi and co-workers, Macromolecules 21, 2722–2727 (1988).
5. Y. Doi and co-workers, Makromol. Chem., Rapid Commun. 8, 631–635 (1987).
6. C. Lauzier, J.-F. Revol, and R. H. Marchessault, FEMS Microbiol. Rev. 103, 299–310

(1992).

7. D. M. Horowitz and co-workers, Nature 363, 23 (1993).
8. P. Kopinke, M. Remmler, and K. Mackenzie, Polym. Degrad. Stab. 52, 25–38 (1996).
9. I. Noda, R. H. Marchessault, and M. Terada, in J. Mark, ed., Polymer Data Handbook,

Oxford University Press, New York, 1999, pp. 586–599.

542

POLY(3-HYDROXYALKANOATES)

Vol. 3

10. H. Brandl and co-workers, Adv. Biochem. Eng. Biotechnol. 41, 77–93 (1990).
11. H. Mitomo, P. Barham, and A. Keller, Polym. J. 19, 1241–1253 (1987).
12. K. D. Gagnon and co-workers, Macromolecules 25, 3723–3728 (1992).
13. Y. Doi, H. Abe, E. Shimamura, M. Hiramitsu, and S. Nakamura, in J. C. Salamone,

ed., Polymeric Materials Encyclopedia, CRC Press, Inc., Boca Raton, Fla., 1996, pp.
4304–4309.

14. S. Y. Lee, Biotechnol. Bioeng. 49, 1–14 (1996).
15. Y. Doi, Microbial Polyesters, VCH Publishers, Inc., New York, 1990.
16. P. Barham and co-workers, J. Mater. Sci. 19, 2781–2794 (1984).
17. T. Bluhm and co-workers, Macromolecules 19, 871–2876 (1986).
18. M. Kunioka, A. Tamaki, and Y. Doi, Macromolecules 22, 694–697 (1989).
19. A. Cao and co-workers, Macromol. Chem. Phys. 198, 3539–3557 (1997).
20. Biopol Nature’s Plastic: Properties and Processing, Monsanto, St. Louis, Mo., 1996, pp.

13–15.

21. H. Brandl and co-workers, Can. J Microbiol. 41(1), 143–153 (1995).
22. H. Verhoogt, B. Ramsay, and B. Favis, Polymer 35, 5155–5169 (1994).
23. L. Zhang and co-workers, Polymer 38, 6001–6007 (1997).
24. M. Grimaldi and co-workers, J. Mat. Sci. 31, 6155–6162 (1996).
25. D. Seebach and M. Zueger, Hel. Chim. Acta 65, 495 (1982).
26. K. Sudesh, H. Abe, and Y. Doi, Prog. Polym. Sci. 25, 1503–1555 (2000).
27. R. M. Lafferty, B. Korsatko, and W. Korsatko, in H.-J. Rehm and G. Reed, eds., Biotech-

nology, Vol. 6, VCH Publishers, Inc., Weinheim, 1988, pp. 135–176.

28. R. Sodain and co-workers, Tissue Eng. 6, 183–188 (2000).
29. U.S. Pat. 5451456 (Sept. 19, 1995), R. H. Marchessault, P. F. Lepoutre, and P. E. Wrist

(to Pulp and Paper Research Institute of Canada).

30. R. G. Lageveen and co-workers, Appl. Environ. Microbiol. 54, 2924–2932 (1988).
31. P. Shin, M. Kim, and J. Kim, J. Environ. Polym. Degrad. 5, 33–39 (1997).
32. U. Pagga, D. Beimborn, and M. Yamamoto, J. Environ. Polym. Degrad. 4, 173–178

(1996).

33. M. Yasin and B. J. Tighe, Biomaterials 13, 9–16 (1992)
34. J. M. Merrick and M. Doudoroff, Nature (London) 189, 890–892 (1961).
35. A. J. Anderson and E. A. Dawes, Microbiol. Rev. 54, 450–472 (1990).
36. T. Saito and co-workers, FEMS Microbiol. Rev. 103, 333–338 (1992).
37. H. Hippe and H. G. Schlegel, Arch. Mikrobiol. 56, 278–299 (1967).
38. Y. Doi and co-workers, FEMS Microbiol. Lett. 65, 165–170 (1990).
39. S. Y. Lee, Y. Lee, and F. Wang, Biotechnol. Bioeng. 65, 363–368 (1999).
40. P. A. Holmes, in D. C. Basset, ed., Developments in Crystalline Polymers, Elsevier, New

York, 1988, pp. 1–65.

41. Y. Doi and co-workers, Macromolecules 23, 26–31 (1990).
42. M. Parikh, R. A. Gross, and S. P. McCarthy, Polym. Mat. Sci. Eng. 66, 408–410 (1992).
43. B. A. Ramsay and co-workers, Appl. Environ. Microbiol. 59, 1242–1246 (1993).
44. R. P. Wool, J. S. Peanasky, J. M. Long, and S. M. Goheen, in S. A. Barenberg, J. L.

Brash, R. Narayan, and A. E. Redpath, eds., Degradable Materials, Proceedings of the
First International Scientific Concensus Workshop, CRC Press, Inc., Boca Raton, Fla.,
1990.

45. Y. Kumagai, Y. Kanesawa, and Y. Doi, Makromol. Chem. 193, 53–57 (1992).
46. H. Nishida and Y. Tokiwa, J. Environ. Polym. Degrad. 1, 65–80 (1993).
47. B. M ¨

uller and D. Jendrossek, Appl. Microbiol. Biotechnol. 38, 487–492 (1993).

48. S. Bloembergen, D. A. Holden, and R. H. Marchessault, Polym. Prepr. 29, 594–595

(1988).

49. A. Steinb ¨

uchel, in D. Byrom, ed., Biomaterials, Macmillan, New York, 1991, pp. 124–

213.

Vol. 3

POLY(3-HYDROXYALKANOATES)

543

50. G. Lefebvre, M. Rocher, and G. Braunegg, Appl. Environ. Microbiol. 63, 827–833,

(1997).

51. Y. Doi and co-workers, J. Chem. Soc., Chem. Commun. 1635–1636 (1987).
52. L. L. Wallen and W. K. Rohwedder, Environ. Sci. Technol. 8, 576–579 (1974).
53. R. H. Findlay and D. C. White, Appl. Environ. Microbiol. 45, 71–78 (1983).
54. R. H. Marchessault, H. Morikawa, and J.-F. Revol, Macromolecules 17, 1882–1884

(1984).

55. M. J. de Smet and co-workers, J. Bacteriol. 154, 870–878 (1983).
56. H. Brandl and co-workers, Appl. Environ. Microbiol. 54, 1977–1982 (1988).
57. A. Timm and A. Steinb ¨

uchel, Appl. Environ. Microbiol. 56, 3360–3367 (1990).

58. G. W. Haywood and co-workers, Appl. Environ. Microbiol. 56, 3354–3359 (1990).
59. G. N. M. Huijberts and co-workers, J. Bacteriol. 176, 1661–1666 (1994).
60. F. G. Jarvis and M. J. Johnson, J. Am. Chem. Soc. 71, 4124–4127 (1949).
61. Y. Poirier and co-workers, Science 256, 520–523 (1993).
62. C. Nawrath, Y. Poirier, and C. Somerville, Mol. Breed. 1, 105–122 (1995).
63. V. Mittendorf and co-workers, Proc. Natl. Acad. Sci. U.S.A. 95, 13397–13402 (1998).
64. U.S. Pat. 4358583 (Sept. 11, 1982), J. Walker and J. R. Whitton (to Imperial Chemical

Industries, Plc).

65. U.S. Pat. 4705604 (Nov. 10, 1987), N. Vanlautem and J. Gilain (to Solvay and Co.).
66. U.S. Pat. 4562245 (Dec. 31, 1985), R. F. Stageman (to Imperial Chemical Industries,

Plc).

67. U.S. Pat. 4101533 (July 18, 1978), R. M. Lafferty and E. Heinzle (to Agroferm AG).
68. J. A. Ramsay and co-workers, Biotechnol. Tech. 8, 589–594 (1994).
69. N. Grassie, E. J. Murray, and P. A. Holmes, Polym. Degrad. Stab. 6, 95–103, (1984).
70. A. H. de Mola, M. Marx-Figini, and R. V. Figini, Makromol. Chem. 176, 2655–2667

(1975).

71. R. Greibel, Z. Smith, and J. M. Merrick, Biochemistry 7, 3767–3681 (1968).
72. U.S. Pat. 4910145 (Mar. 20, 1990), P. A. Holmes and G. B. Lim (to Imperial Chemical

Industries, Plc).

73. D. Byrom, in E. A. Dawes, ed., Novel Biodegradable Microbial Polymers, Kluwer Aca-

demic Publishers, Dordrecht, the Netherlands, 1990, pp. 113–117.

74. J. A Ramsay and co-workers, Biotechnol. Tech. 4, 221–226 (1990).
75. C. A. Lauzier and co-workers, Tappi J. 76, 71–77 (1993).
76. Eur. Par. Appl. 46335 (Feb. 24, 1980), P. A. Holmes and E. Jones (to Imperial Chemical

Industries, Plc).

77. U.S. Pat. 3225766 (Dec. 28, 1965), J. N. Baptist and J. B. Ziegler (to W.R. Grace and

Co.).

78. Eur. Pat. 69497 (Apr. 15, 1987), P. A. Holmes, L. F. Wright, and S. H. Collins (to Imperial

Chemical Industries, Plc).

79. B. S. Kim and co-workers, Biotechnol. Bioeng. 43, 892–898 (1994).
80. B. S. Kim, S. Y. Lee, and H. N. Chang, Biotechnol. Lett. 14, 811–816 (1992).
81. D. Bourque, Y. Pomerleau, and D. Groleau, J. Microbiol. Biotechnol. 44, 367–376 (1993).
82. G. N. M. Huijberts, H. van der Wal, R. A. Weusthuis, and G. Eggink, in Abstract 5th In-

ternational Symposium on Bacterial Polyhydroxyalkanoates, Davos, Switzerland, Aug.
1996. (Abstract).

83. F. Merchant, D. E. Dennis, R. Merchant, and C. F. Batrell, in 4th International Sympo-

sium on Bacterial Polyhydroxyalkanoates, Montreal, Canada, Aug. 1994. (Abstract).

84. U. J. H ¨anggi, in Ref. 72, pp. 65–70.
85. D. Byrom, Int. Biodeter. Biodegrad. 31, 199–203 (1993).
86. D. G. H. Ballard, P. A. Holmes, and P. J. Senior, Recent Advances in Mechanistic and

Synthetic Aspects of Polymerization 215, 293–314 (1987).

87. R. J. Griebel and J. M. Merrick. J. Bacteriol. 108, 782–789 (1971).

544

POLY(3-HYDROXYALKANOATES)

Vol. 3

88. T. U. Grengross and D. P. Martin, Proc. Natl. Acad. Sci. U.S.A. 92, 6279–6283 (1995).

B

RUCE

A. R

AMSAY

Polyferm Canada,
J

ULIANA

A. R

AMSAY

Queen’s University

POLY(ACRYLIC ACID).

See A

CRYLIC

(

AND

M

ETHACRYLIC

) A

CID

P

OLYME

.

POLY(CYCLOHEXYLENE TEREPHTHALATE).

See C

YCLOHEXANEDIMETHANOL

P

OLYESTERS

.

POLY(ETHYLENE-NORBORNENE).

See E

THYLENE

-N

ORBORNENE

C

OPOLYMERS

.

POLY(METHACRYLIC ACID).

See A

CRYLIC

(

AND

M

ETHACRYLIC

)

A

CID

P

OLYME

.