Progress in Polymer Science 33 (2008) 820–852

Contents lists available at ScienceDirect

Progress in Polymer Science

journa l homepage: www.e lsev ier .com/ locate /ppolysc i

Processing technologies for poly(lactic acid)

L.­T. Lim a,∗, R. Aurasb, M. Rubinob

a Department of Food Science, University of Guelph, Guelph, Ontario N1G 2W1, Canadab School of Packaging, Michigan State University, East Lansing, MI 48824­1223, USA

a r t i c l e i n f o

Article history:

Received 6 June 2007

Received in revised form 6 May 2008

Accepted 7 May 2008

Available online 19 June 2008

Keywords:

Polylactide

Poly(lactic acid)

PLA

Processing

Converting

Review

a b s t r a c t

Poly(lactic acid) (PLA) is an aliphatic polyester made up of lactic acid (2­hydroxy propionic

acid) building blocks. It is also a biodegradable and compostable thermoplastic derived from

renewable plant sources, such as starch and sugar. Historically, the uses of PLA have been

mainly limited to biomedical areas due to its bioabsorbable characteristics. Over the past

decade, the discovery of new polymerization routes which allow the economical produc­

tion of high molecular weight PLA, along with the elevated environmental awareness of the

general public, have resulted in an expanded use of PLA for consumer goods and packaging

applications. Because PLA is compostable and derived from renewable sources, it has been

considered as one of the solutions to alleviate solid waste disposal problems and to lessen

the dependence on petroleum­based plastics for packaging materials. Although PLA can

be processed on standard converting equipment with minimal modifications, its unique

material properties must be taken into consideration in order to optimize the conversion of

PLA to molded parts, films, foams, and fibers. In this article, structural, thermal, crystalliza­

tion, and rheological properties of PLA are reviewed in relation to its converting processes.

Specific process technologies discussed are extrusion, injection molding, injection stretch

blow molding, casting, blown film, thermoforming, foaming, blending, fiber spinning, and

compounding.

© 2008 Elsevier Ltd. All rights reserved.

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 821

2. Structural composition. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 821

3. Thermal properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 823

4. Crystallization behavior . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 824

5. Rheological properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 826

6. Thermal degradation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 827

7. Processing of PLA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 828

Abbreviations: BD, 1,4­butanedial; BDI, 1,4­butane diisocyanate; DSC, differential scanning calorimetry; BUR, blow­up­ratio; 1Hrel , endothermic

enthalpy relaxation; 1Hc , heat of crystallization; 1Hm , heat of fusion; HDPE, high density polyethylene; HIPS, high impact polystyrene; HMDI, hexamethy­

lene diisocyanate; ISBM, injection stretch blow molding; LDPE, low density polyethylene; MD, machine direction; MDO, machine direction orientation; MFI,

melt flow index; MMT, montmorillonite; Mn , number­average molecular weight; Mw , weight­average molecular weight; OPLA, oriented poly (lactic acid);

OPP, oriented polypropylene; OPS, oriented polystyrene; PEG, poly(ethylene glycol); PET, poly(ethylene terephthalate); PDI, polydispersity index; PDLLA,

poly(d,l­lactic acid); PHA, polyhydroxyalkanoate; PHO, poly(3­hydroxyloctanoate); PLA, poly(lactic acid); PLLA, poly(l­lactic acid); PP, polypropylene; PS,

polystyrene; PVT, pressure–volume–temperature; TD, transverse direction; TDO, transverse direction orientation; Tg , glass transition temperature; Tm ,

melting temperature; WAXS, wide angle X­ray scattering; WVTR, water vapor transmission rate; �0 , zero­shear viscosity.∗ Corresponding author. Tel.: +1 519 824 4120x56586; fax: +1 519 824 6631.

E­mail address: llim@uoguelph.ca (L.­T. Lim).

0079­6700/$ – see front matter © 2008 Elsevier Ltd. All rights reserved.

doi:10.1016/j.progpolymsci.2008.05.004

L.­T. Lim et al. / Progress in Polymer Science 33 (2008) 820–852 821

7.1. Drying and extrusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 828

7.2. Injection molding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 830

7.3. Stretch blow molding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 833

7.4. Cast film and sheet . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 835

7.5. Extrusion blown film . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 836

7.6. Thermoforming . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 837

7.7. Foaming . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 838

7.8. Fiber spinning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 840

7.9. Electrospinning of ultrafine fibers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 841

7.10. PLA blends with other polymers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 844

7.11. Compounding of PLA composites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 846

7.12. PLA nanocomposites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 847

8. Conclusion: prospects of PLA polymers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 849

Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 849

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 849

1. Introduction

Thermoplastic polymers exhibit many properties ideal

for use in packaging and other consumer products, such

as light weight, low process temperature (compared to

metal and glass), variable barrier properties to match end­

use applications, good printability, heat sealable, and ease

of conversion into different forms. Today, most plastics

are derived from non­renewable crude oil and natural

gas resources. While some plastics are being recycled and

reused, the majority are disposed in landfills due to end­use

contamination. In 2005, plastics were recovered at a rate

lower than 10% in the USA [1]. Over the past decade, there

has been a sustained research interest on compostable

polymers derived from renewable sources as one of the

solutions to alleviate solid waste disposal problems and to

lessen the dependence on petroleum­based plastics.

Poly(lactic acid) (PLA) is a compostable polymer derived

from renewable sources (mainly starch and sugar). Until

the last decade, the main uses of PLA have been limited to

medical applications such as implant devices, tissue scaf­

folds, and internal sutures, because of its high cost, low

availability and limited molecular weight. Recently, new

techniques which allow economical production of high

molecular weight PLA polymer have broadened its uses

[2]. Since PLA is compostable and derived from sustain­

able sources, it has been viewed as a promising material

to reduce the societal solid waste disposal problem [3,4].

Its low toxicity [5], along with its environmentally benign

characteristics, has made PLA an ideal material for food

packaging and for other consumer products [6].

PLA belongs to the family of aliphatic polyesters derived

from a­hydroxy acids. The building block of PLA, lactic

acid (2­hydroxy propionic acid), can exist in optically active

d­ or l­enantiomers. Depending on the proportion of the

enantiomers, PLA of variable material properties can be

derived. This allows the production of a wide spectrum of

PLA polymers to match performance requirements. PLA has

reasonably good optical, physical, mechanical, and barrier

properties compared to existing petroleum­based poly­

mers [7]. For instance, the permeability coefficients of CO2,

O2, N2, and H2O for PLA are lower than for polystyrene (PS),

but higher than poly(ethylene terephthalate) (PET) [8–10].

The barrier properties of PLA against organic permeants,

such as ethyl acetate and d­limonene, are comparable to

PET [11]. Mechanically, unoriented PLA is quite brittle, but

possesses good strength and stiffness. Oriented PLA pro­

vides better performance than oriented PS, but comparable

to PET [9]. Tensile and flexural moduli of PLA are higher than

high density polyethylene (HDPE), polypropylene (PP) and

PS, but the Izod impact strength and elongation at break val­

ues are smaller than those for these polymers [12]. Overall,

PLA possesses the required mechanical and barrier proper­

ties desirable for a number of applications to compete with

existing petroleum­based thermoplastics.

Today, the main conversion methods for PLA are based

on melt processing. This approach involves heating the

polymer above its melting point, shaping it to the desired

forms, and cooling to stabilize its dimensions. Thus, under­

standing of thermal, crystallization, and melt rheological

behaviors of the polymer is critical in order to optimize

the process and part quality. Some of the examples of

melt processed PLA are injection molded disposable cut­

lery, thermoformed containers and cups, injection stretch

blown bottles, extruded cast and oriented films, and melt­

spun fibers for nonwovens, textiles and carpets [6,13,14].

PLA also finds uses in other less conventional applications,

such as for the housing for laptop computers electronics

[14–17]. Recently, PLA has also been processed in conjunc­

tion with other filler materials to form composites which

possess various unique properties, including those based

on nanoclays [18–23], biofibers [16,24,25], glass fibers [26]

and cellulose [27,28]. The aim of this review is to discuss

the key process technologies for PLA and summarize the

properties of PLA related to the processing techniques used.

2. Structural composition

The basic building block of PLA, lactic acid, can be pro­

duced by carbohydrate fermentation or chemical synthesis.

Currently, the majority of lactic acid production is based

on the fermentation route. Various purification technolo­

gies for lactic acid and lactide can be found in a recent

review by Datta and Henry [2]. One of the main drivers

for the recent expanded use of PLA is attributable to the

economical production of high molecular weight PLA poly­

mers (greater than ∼100,000 Da). These polymers can be

produced using several techniques, including azeotropic

822 L.­T. Lim et al. / Progress in Polymer Science 33 (2008) 820–852

Fig. 1. Synthesis of PLA from l­ and d­lactic acids. Adapted from Auras et al. [3] by permission of Wiley–VCH Verlag GmbH & Co. KGaA.

dehydrative condensation, direct condensation polymer­

ization, and/or polymerization through lactide formation

(Fig. 1). By and large, commercially available high molecular

weight PLA resins are produced via the lactide ring­opening

polymerization route [3,4,29].

Commercial PLA are copolymers of poly(l­lactic acid)

(PLLA) and poly(d,l­lactic acid) (PDLLA), which are pro­

duced from l­lactides and d,l­lactides, respectively [3]. The

l­isomer constitutes the main fraction of PLA derived from

renewable sources since the majority of lactic acid from

biological sources exists in this form. Depending on the

composition of the optically active l­ and d,l­enantiomers,

PLA can crystallize in three forms (a, b and g). The a­

structure is more stable and has a melting temperature Tm

of 185 ◦C compared to the b­structure, with a Tm of 175 ◦C

[3]. The optical purity of PLA has many profound effects

on the structural, thermal, barrier and mechanical proper­

ties of the polymer [30–36]. PLA polymers with l­content

greater than ∼90% tend to be crystalline while those with

lower optical purity are amorphous. Moreover, Tm, glass

transition temperature Tg, and crystallinity decrease with

decreasing l­isomer content [30,34,37]. Tsuji et al. reported

that the optical impurity of PLLA films ranging from 0–50%

was insignificant in affecting the water vapor transmis­

sion rate (WVTR) of the polymer; nevertheless, the WVTR

values decreased with increasing film crystallinity in the

0–20% range [31]. Thus, judicious selection of appropriate

PLA resin grade is important to match the conversion pro­

cess conditions used. Usually, PLA articles which require

heat­resistant properties can be injection molded using

PLA resins of less than 1% d­isomer. Alternatively, nucle­

ating agents may be added to promote the development of

crystallinity under relatively short molding cycles. In con­

trast, PLA resins of higher d­isomer contents (4–8%) would

be more suitable for thermoformed, extruded, and blow

molded (e.g., injection molded preform for blow molding)

products, since they are more easily processed when the

crystallinity is low [38].

When exposed to elevated temperatures, PLA is known

to undergo thermal degradation, leading to the formation

of lactide monomers (Section 3). It has been suggested

that this property may be leveraged for the feedstock

recycling of PLA [39,40]. However, the propensity for the

lactide monomer to undergo racemization to form meso­

lactide can impact the optical purity and thus the material

properties of the resulting PLA polymer [39–43]. Recently,

Tsukegi et al. reported that at temperature less than

200 ◦C, conversion of PLLA into meso­lactide and oligomers

was minimal. However, above this temperature, the for­

mation of meso­lactide became quite significant (4.5 wt%

at 200 ◦C and 38.7 wt% at 300 ◦C for 120 min heating).

Oligomers were reported to form at temperatures higher

than 230 ◦C [39]. These authors also reported that the

oligomerization proceeded rapidly in the presence of MgO,

to reach an equilibrium between monomers and oligomers;

the l,l:meso:d,d lactide composition ratio converged to

1:1.22:0.99 (w/w/w) after 120 min heating at 300 ◦C [39].

Fan et al. reported that the racemization at 250–300 ◦C

L.­T. Lim et al. / Progress in Polymer Science 33 (2008) 820–852 823

Fig. 2. Comparison of glass transition and melting temperatures of PLA

with other thermoplastics.

can be controlled by adding calcium oxide to PLLA, which

reduces the pyrolysis temperature, and more importantly,

leads to predominant l,l­lactide formation [40].

3. Thermal properties

Similar to many thermoplastic polymers, semicrys­

talline PLA exhibits Tg and Tm. Above Tg (∼58 ◦C) PLA is

rubbery, while below Tg, it becomes a glass which is still

capable to creep until it is cooled to its b transition tem­

perature at approximately −45 ◦C, below which it behaves

as a brittle polymer [44]. Fig. 2 compares PLA’s Tg and Tm

values with other polymers. As shown, PLA has relatively

high Tg and low Tm as compared to other thermoplastics.

The Tg of PLA is dependent on both the molecular

weight and the optical purity of the polymer (Fig. 3). The

Tg increases with molecular weight to maximum values

at infinite molecular weight of 60.2, 56.4 and 54.6 ◦C for

PLA consisting of 100, 80, and 50% l­stereoisomer contents,

respectively. Furthermore, PLA with higher content of l­

lactide has higher Tg values than the same polymer with

the same amount of d­lactide [37]. Similar relationships

were reported by Tsuji and Ikada [34]. Table 1 shows the

Fig. 3. Glass transition temperatures for PLAs of different l­contents as a

function of molecular weight. Curves are created based on the original data

published by Dorgan et al. [37] by permission of The Society of Rheology.

Table 1

Primary transition temperatures of selected PLA copolymers

Copolymer ratio Glass transition

temperature (◦C)

Melting temperature (◦C)

100/0 (l/d,l)­PLA 63 178

95/5 (l/d,l)­PLA 59 164

90/10 (l/d,l)­PLA 56 150

85/15 (l/d,l)­PLA 56 140

80/20 (l/d,l)­PLA 56 125

Adapted from Bigg [33].

glass transition and melting temperatures of different PLA

polymers produced with different ratios of copolymer.

In general, the relationship between Tg and molecular

weight can be represented by the Flory–Fox equation:

Tg =T∞g − K

M̄n

(1)

where T∞g is the Tg at the infinite molecular weight, K is a

constant representing the excess free volume of the end

groups for polymer chains, and M̄n is the number aver­

age molecular weight. The values of T∞g and K are around

57–58 ◦C and (5.5–7.3)×104 as reported in the literature

for PLLA and PDLLA, respectively [45].

The glass transition behavior of PLA is also dependent

on the thermal history of the polymer. Quenching the poly­

mer from the melt at a high cooling rate (>500 ◦C/min, such

as during injection molding) will result in a highly amor­

phous polymer. PLA polymers with low crystallinity have a

tendency to undergo rapid aging in a matter of days under

ambient conditions [46,47]. The phenomenon is an impor­

tant contributor to the embrittlement of PLA. This topic will

be discussed in greater details in Section 7.2.

The Tm of PLA is also a function of its optical purity.

The maximum practical obtainable Tm for stereochemically

pure PLA (either l or d) is around 180 ◦C with an enthalpy

of 40–50 J/g. The presence of meso­lactide in the PLA struc­

ture can depress the Tm by as much as 50 ◦C, depending

on the amount of d­lactide incorporated to the polymer.

Fig. 4 shows the variation of the Tm as a function of % meso­

lactide introduced in the PLA based on data from Witzke

Fig. 4. Peak melting temperature of PLA as a function of % meso­lactide.

(©) Represents values reported by Witzke [48]; (d) represents values

reported by Hartmann [49]; solid line is calculated based on Eq. (2).

824 L.­T. Lim et al. / Progress in Polymer Science 33 (2008) 820–852

[48] and Hartmann [49]. The relationship of Tm and meso­

lactide content can be approximated reasonably well by the

following expression [48]:

Tm (◦C) ≈ 175 ◦C− 300 Wm (2)

where Wm is the fraction of meso­lactide below 0.18 level,

and 175 ◦C is the melting temperature of PLA made of 100%

l­lactide. Typical Tm values for PLA are in the range of

130–160 ◦C. The Tm depression effect of meso­lactide has

several important implications as it helps expand the pro­

cess windows, reduce thermal and hydrolytic degradation,

and decrease lactide formation.

Pyda et al. determined the heat capacity of PLA in solid

and liquid states ranging from 5 to 600 K [36]. The heat

capacity (Cp­liquid, J K−1 mol−1) can be represented in a sim­

ple form: Cp­liquid = 120.17 + 0.076T, where T is in Kelvin (K).

4. Crystallization behavior

The physical, mechanical and barrier properties of

PLA are dependent on the solid­state morphology and

its crystallinity. Accordingly, the crystallization behaviors

of PLA have been studied in detail by many researchers

[4,32,50–55]. PLA can be either amorphous or semicrys­

talline depending on its stereochemistry and thermal

history. The crystallinity of PLA is most commonly deter­

mined using the differential scanning calorimetry (DSC)

technique. By measuring the heat of fusion 1Hm and heat

of crystallization 1Hc, the crystallinity can be determined

based on the following equation:

crystallinity (%) =1Hm −1Hc

93.1× 100 (3)

where the constant 93.1 J/g is the 1Hm for 100% crystalline

PLLA or PDLA homopolymers.

On quenching the optically pure PLA polymer from the

melt phase (e.g., during injection molding process), the

resulting polymer will become quite amorphous. As shown

in Fig. 5, quenching the polymer from melt at a high cooling

rate resulted in an exothermic crystallization peak on the

DSC thermogram during the subsequent reheat, while slow

cooling produced a polymer with higher crystallinity with

much lower enthalpy of crystallization. The tendency for

PLA to crystallize upon reheat also depended on the heating

rate (Fig. 6), as well as the optical purity of the PLA polymer

(Fig. 7). As shown in Fig. 7, PLA polymers with greater than

∼8% d­isomer level remained amorphous even after 15 h of

isothermal treatment at 145 ◦C. In contrast, at 1.5%d­isomer

level, although the quenched sample (“Quenched PLA­l”)

has a minimal crystallinity, the isothermal treatment at

145 ◦C resulted in a large endothermic melting peak around

450 K (Fig. 7). In general, the crystallization half­time of

PLA increases about 40% for every 1% (w/w) meso­lactide

in the polymerization mixture, which is mainly driven by

the reduction of the melting point for the copolymer [56].

Nucleation parameters for PLLA crystallization under

isothermal and nonisothermal conditions were determined

by Kishore and Vasanthakumari using DSC and microscopy

[54]. They reported that the radius growth rate of the crys­

tals decreased as molecular weight increased, as observed

in many other polymers. The nucleation parameters are

Fig. 5. DSC thermograms of water quenched, air­annealed (cooled from

220 ◦C to ambient temperature in 5 min), and full­annealed (cooled from

220 ◦C to ambient temperature in 105 min) PLLA samples. DSC scans were

performed at a heating rate of 10 ◦C/min. Adapted from Sarasua et al. [32]

by permission of John Wiley & Sons, Inc.

Fig. 6. DSC scans for 1.5% d­lactide PLA samples cooled from the melt at

10 K/min and then reheated at different heating rates from 30 to 0.3 K/min.

Adapted from Pyda et al. [36] by permission of Elsevier B.V.

related in the following form [54,57]:

Kg =4b��eTm

1Hfk(4)

where Kg is the nucleation constant, b is the layer thickness

of the crystal, � is the lateral surface energy, �e is the

fold surface energy, 1Hf is the heat of fusion per unit

volume, and k is the Boltzmann constant. Table 2 shows

the nucleation parameters from isothermal and non­

Table 2

Nucleation parameters from isothermal and nonisothermal kinetic anal­

yses for PLLA

Parameter Isothermal Non­isothermal

Nucleation parameter, Kg (×105) 2.44 2.69

Lateral surface energy, � (×103 J/m2) 12.0 13.6

�×�e (×106 J2/m4) 753 830

Adapted from Kishore and Vasanthakumari [54].

L.­T. Lim et al. / Progress in Polymer Science 33 (2008) 820–852 825

Fig. 7. DSC scans at 20 K/min for PLA with 1.5% (PLA­L), 8.1% (PLA­M),

and 16.4% (PLA­H) d­isomers. All samples were cooled quickly from the

melt and isothermally crystallized at 145 ◦C for 15 h. The quenched PLA­L

sample was cooled similarly from the melt but did not undergo the 15 h

isothermal crystallization. Thermograms are recreated based on the data

originally published by Pyda et al. [36] by permission of Elsevier B.V.

isothermal kinetic analysis of PLLA. Solving Eq. (4) with

Tm = 480 K, 1Hf = 111.083×106 J/m3; b = 5.17×10−8 cm,

12.03×10−3 J/m2, and �e = 6.089×10−4 J/m2, Kg can be

determined. This value can be used to evaluate the transi­

tion between two types of crystallization behavior in PLA.

In the first type of crystallization, the nucleation rate is

low and axialite morphology in the films is prevalent. In

the second type, the nucleation rate is high, so multinu­

cleation occurs and spherulitic morphology in the films

is observed [57]. For PLLA, both crystallization processes

have been observed depending on the molecular weight

of the samples. The infinite dissolution temperature T0d

(determined by the extrapolation of dissolution tempera­

ture Td versus crystallization temperature Tc plots to the

intersection where Td = Tc) for PLLA in p­xylene solution

was determined by Kalb and Pennings to be 126.5 ◦C [58].

This temperature is relevant for fiber formation processes,

since fibers prepared from solution near this temperature

have ultra­high strength properties [58].

The formation of crystallinity may or may not be favor­

able depending on the end­use requirements of the PLA

articles. For instance, high crystallinity will not be opti­

mal for injection molded preforms which are intended

for further blow molding since rapid crystallization of the

polymer would hamper the stretching of the preform and

optical clarity of the resulting bottle. In contrast, increased

crystallinity will be desirable for injection molded articles

for which good thermal stability is important. Crystal­

lization of PLA articles can be initiated by annealing at

temperatures higher than Tg and below the melting point

to improve their thermal stability. For instance, Perego et

al. showed that crystallization of injection molded PLLA

parts by annealing at 105 ◦C for 90 min increased tensional

and flexural elasticity, Izod impact strength, and heat resis­

tance [59]. After annealing PLA copolymers, the presence

of two melting peaks in a DSC scan is quite common, as

previously observed by Yasuniwa et al. [60]. They reported

that the low temperature Tm peak height increased with

Fig. 8. Development of crystallinity in biaxially stretched PLA at 80 ◦C

using 100% s−1 strain rate. Data are adapted from Drumright et al. [38]

by permission of Wiley–VCH Verlag GmbH & Co. KGaA.

increasing heating rate, whereas the high temperature Tm

decreased. In contrast, increasing the cooling rate reduced

the low Tm peak, while the high Tm peak increased. The

double­melting peak behavior was explained based on

melt­recrystallization model, in which small and imper­

fect crystals changed successively into more stable crystals

through the melting and recrystallization [60].

Another strategy to increase the crystallinity of PLA is

by incorporating nucleating agent in the polymer during

extrusion. This lowers the surface free energy barrier for

nucleation and enables crystallization at higher tempera­

ture to take place upon cooling. Kolstad showed that talc

can be added to PLLA to effectively modify the crystalliza­

tion rate of the polymer [56]. With 6% talc added to PLLA,

the crystallization half­time of the polymer reduced from

3 min at 110 ◦C to approximately 25 s. At the same percent of

talc, for 3% mesolactide copolymerized with the l­lactide,

the half­time reduced from about 7 min to about 1 min [56].

Li and Huneault compared the crystallization kinetics of

talc and montmorillonite (MMT, Cloisite® Na+) for 4.5% d­

PLA. They reported that the lowest crystallization induction

period and maximum crystallization speeds were observed

around 100 ◦C. By adding 1% (w/w) of talc, the crystalliza­

tion half time of PLA was decreased from a few hours to

8 min. In contrast, the MMT tested was less effective as a

nucleating agent; the lowest half­time achieved was 30 min

[61].

Unlike quiescent crystallization discussed above, strain­

induced crystallization occurs when the polymer is

mechanically orientated. This phenomenon is prevalent

during the production of oriented PLA films, stretch blow

molding of bottles, thermoforming of containers, and fiber

spinning. As expected, the proportion of d­ and l­isomers

has an effect on the strain­induced crystallinity during

the mechanical orientation. As shown in Fig. 8, the per­

cent crystallinity of amorphous PLA sheet increases with

increasing draw ratio. Moreover, the crystallinity decreases

as the stereoisomeric purity of the polymer decreases [38].

The amount of crystallinity attained through orientation

also depends on the mode of stretching (sequential ver­

826 L.­T. Lim et al. / Progress in Polymer Science 33 (2008) 820–852

Fig. 9. Comparison of zero­shear viscosity values versus molecular weight

for poly(85% l­co­15% d­lactide) at 85 and 100 ◦C as reported by Witkze

[48], and PLLA at 180 ◦C as reported by Dorgan et al.[12].

sus simultaneous), strain rate, temperature, and annealing

conditions [38,62,63]. More discussions on this topic will

be presented in Section 7.2.

5. Rheological properties

Melt rheological properties of PLA have a profound

effect on how the polymer flows during the conver­

sion process. Since the PLA rheological properties are

highly dependent on temperature, molecular weight and

shear rate, they must be taken into consideration during

tooling design, process optimization, and process model­

ing/simulation. Melt viscosities of high­molecular­weight

PLA are in the order of 5000–10,000 P (500–1000 Pa s)

at shear rates of 10–50 s−1. These polymer grades are

equivalent to Mw ∼100,000 Da for injection molding to

∼300,000 Da for film cast extrusion applications [4]. The

melts of high molecular weight PLA behave like a pseu­

doplastic, non­Newtonian fluid. In contrast, low molecular

weight PLA (∼40,000 Da) shows Newtonian­like behavior

at shear rates typical of film extrusion [64]. Under iden­

tical processing conditions, semicrystalline PLA tends to

possess higher shear viscosity than its amorphous counter­

part. Moreover, as shear rates increase, the viscosities of the

melt decrease considerably, i.e., the polymer melt exhibits

shear­thinning behavior [65].

Viscoelastic properties of polymer melts can be charac­

terized by zero­shear viscosity, �0, and recoverable shear

compliance JOe . Both of these parameters can be obtained

from dynamic experiments by determining the dynamic

moduli at the limit of low frequency [48]. The product of

these two values (�0 × JOe ) gives the average relaxation time

required for final stress equilibration time in the liquid �0.

The value of �0 is strongly affected by the molecular weight,

which is typically described empirically by the power law

equation. Cooper­White and Mackay reported that the �0

of PLLA melt showed dependence on Mw to the 4.0 power

instead of the theoretical value of 3.4 [64]. In comparison,

Dorgan et al. reported a power index of 4.6 [66]. Fig. 9 shows

the relationship between �0 and Mw for PLLA (100:0) at

180 ◦C [12], and 15% d­lactide PDLA at 85 and 100 ◦C [48].

Witkze showed that the temperature effect on �0 for 15%

d­lactide PLA can be described by [48]:

�0 = n0,ref

(

Mw

100, 000

)a

exp

(

Ea

R

(

1

T(K)−

1

373

))

(5)

where a = 3.38±0.13, the activation energy of flow

Ea = 190 kJ/mol, �0,ref = 89,400±9300 Pa s, R is the gas

constant 8.314 J/K mol, and T is the temperature in K.

Witzke further showed that �0 can be correlated with

the isomer composition by fitting to the well­known

Williams–Landel–Ferry equation (WLF) [48]:

�0 = (a1 + a2Wmeso + a3Wl­mer)

(

Mw

100, 000

)3.38

×exp

(

−C1(T(C)− 100)

C2 + (T(C)− 100)

)

(6)

where Wmeso and Wl­mer are the initial weight frac­

tions for meso­lactide and l­lactide, respectively,

a1 =−13,000, a2 =−142,000, a3 = 112,000, C1 = 15.6±1.6,

and C2 = 110±11 ◦C; a1, a2, a3, and C1 do not have units;

and T(C) is the testing temperature in ◦C. Eq. (6) can

be used to predict �0 of amorphous polylactides with

l­monomer composition higher than 50% between Tg

and Tg + 100 ◦C. The equation predicts that �0 increases

with increasing l­monomer and decrease as meso­lactide

content increases [48].

The rheological properties of PLA can be modified by the

introduction of branching into the polymer chain architec­

ture. Many routes, such as multifunctional polymerization

initiators, hydroxycyclic ester initiators, multicyclic ester,

and crosslinking via free radical addition have been used

to introduce branching in PLA [12,67–69]. Lehermeier

and Dorgan blended PLA with 5% d­isomer with varying

proportions of branched PLA produced through peroxide

initiated crosslinking of linear PLA by reactive extrusion

[67]. They observed that �0 of the blends deviated con­

siderably from the log additive rule and attributed this to

the effect of free volume. Lehermeier and Dorgan showed

that tris(nonylphenyl) phosphite was effective for stabi­

lizing the viscosity of PLA during the thermorheological

time sweep experiment of branched PLA polymers [67]. In

another study from the same research group, the stabiliz­

ing effect of tris(nonylphenyl) phosphate was elucidated

by using the time­temperature superposition technique,

showing that this compound greatly facilitated the ther­

morheological experiments by prevented the confounding

effect from degradation reactions [69,70].

Carreau–Yasuda model (Eq. (7)) has been used to model

the viscosity and shear rate relationship of linear PLA and

linear­branched PLA blends [69]:

� = C1[1+ (C2 ̇)C3 ](C4−1/C3) (7)

where � is the viscosity, is the shear rate, and C1, C2, C3

and C4 are material dependent parameters. The constants

for the model are summarized in Table 3. C1 determines

�0 which decreases with increasing linear content. C2 is

the relaxation time approximately corresponded to the

reciprocal of frequency for the onset of shear thinning.

C3 determined the shear thinning which increased with

L.­T. Lim et al. / Progress in Polymer Science 33 (2008) 820–852 827

Table 3

Carreau–Yasuda model parameters for Eq. (7)

Blend, % Carreau–Yasuda parameters

C1 (Pa s) C2 (s) C3 C4

0 10,303 0.01022 0.3572 −0.0340

20 8,418 0.00664 0.3612 −0.0731

40 6,409 0.01364 0.4523 0.0523

60 5,647 0.00513 0.4356 −0.1002

80 4,683 0.00450 0.4754 −0.1108

100 3,824 0.01122 0.7283 0.0889

Adapted from Lehermeir and Dorgan [69].

increasing linear content, i.e., branched PLA shear thinned

stronger than the linear material [69]. The increase of both

�0 and shear thinning with the addition of branching is

also reported by other studies on PLA polymers with star

polymer chain architectures [12,66].

Palade et al. studied the extensional viscosities of high l­

content PLA (100,000–120,000Mw). They showed that PLA

can be drawn to large Hencky strains without breaking. The

polymer also exhibited strain­hardening behaviors during

the deformation [70], which is an important characteris­

tic for processing operations, such as fiber spinning, film

casting, and film blowing. Yamane et al. reported that the

addition of PDLA to PLLA enhanced the strain hardening

properties of the resulting blends even at very low PDLA

contents (<5 wt%). They also reported that low Mw PDLA

affected the shear rheology of the blends much more sig­

nificantly than high Mw PDLA [71]. This may provide an

effective avenue for modifying the spinning behavior of the

PLA.

Although solution viscosity of PLA in solvent does not

provide direct relevance to the processing of molten PLA

polymer, this property is frequently evaluated to deter­

mine the molecular weight of resins and processed parts

to ensure that they are within the required specifications.

The relationship between viscosity and molecular weight

of PLA dissolved in dilute solution is commonly modeled

using the Mark–Houwink equation:

[�] = K ×Mav (8)

where [�] is the intrinsic viscosity, K and a are constants,

and Mv is the experimental viscosity average molecular

weight. The Mark–Houwink equation is dependent on the

type of PLA, the solvent used, and the temperature of the

solution. Table 4 summarizes the Mark–Houwink parame­

ters for different compositions of PLA polymers in different

solvent solutions.

6. Thermal degradation

One of the drawbacks of processing PLA in the molten

state is its tendency to undergo thermal degradation, which

is related both to the process temperature and the resi­

dence time in the extruder and hot runner [72]. By and

large, thermal degradation of PLA can be attributed to:

(a) hydrolysis by trace amounts of water, (b) zipper­like

depolymerization, (c) oxidative, random main­chain scis­

sion, (d) intermolecular transesterification to monomer

and oligomeric esters, and (e) intramolecular transesteri­

fication resulting in formation of monomer and oligomer

lactides of low Mw [73]. Kopinke et al. proposed that above

200 ◦C, PLA can degrade through intra­ and intermolecu­

lar ester exchange, cis­elimination, radical and concerted

non­radical reactions [41], resulting in the formation of CO,

CO2, acetaldehyde and methylketene. In contrast, McNeill

and Leiper proposed that thermal degradation of PLA is a

non­radical, “backbiting” ester interchange reaction involv­

ing the ­OH chain ends [74]. Depending on the point in

the backbone at which the reaction occurs, the product

can be a lactide molecule, an oligomeric ring, or acetalde­

hyde plus carbon monoxide (Fig. 10). Similar degradation

mechanisms were reported by Kopinke et al. [41]. At tem­

peratures in excess of 270 ◦C, homolysis of the polymer

backbone can occur. The formation of acetaldehyde is

expected to increase with increasing process temperature

due to the increased rate of the degradation reactions. In

the 230–440 ◦C temperature range explored by McNeill and

Leiper [74], acetaldehyde is formed in highest proportion at

230 ◦C and a marked decrease is observed at 440 ◦C, which

is believed to be caused by the thermal degradation of

acetaldehyde, involving a complex chain reaction to form

methane and carbon monoxide at the elevated tempera­

ture. McNeill and Leiper also proposed that the formation

of butane­2,3­dione, another byproduct detected, is likely

caused by the radical combination of acetyl radicals from

Table 4

Mark–Houwink constants PLA in selected solvents

Polymer types Equations Conditions

(1) PLLA [�] = 5.45× 10−4 M0.73v 25 ◦C in chloroform [59,189]

(2) PDLLA [�] = 1.29× 10−5 M0.82v 25 ◦C in chloroform [190]

(3) PDLLA [�] = 2.21× 10−4 M0.77v 25 ◦C in chlofoform [59,189]

(4) Linear PLLA [�] = 4.41× 10−4 M0.72v 25 ◦C in chloroform [190]

(5) “Star” PLLA (six arms) [�] = 2.04× 10−4 M0.77v 25 ◦C in chloroform [190]

(6) PDLLA [�]¦ = 2.59× 10−4 M0.689v 35 ◦C in THFa [191]

(7) PDLLA [�] = 5.50× 10−4 M0.639v 31.15 ◦C in THF [191]

(8) PLLA (amorphous) [�] = 6.40× 10−4 M0.68v 30 ◦C in THF [192]

(9) PLLA (amorphous/semi­crystalline) [�]¦ = 8.50× 10−4 M0.66v 30 ◦C in THF [192]

(10) PLLA (semi­crystalline) [�] = 1.00× 10−3 M0.65v 30 ◦C in THF [192]

(11) PDLLA [�] = 2.27× 10−4 M0.75v (one point method) 30 ◦C in benzene [193], Tuan–Fuoss viscometer

(12) PDLLA [�] = 1.58× 10−4 M0.78v 25 ◦C in ethyl acetate [194]

Adapted from Garlotta [4].a THF: tetrahydrofuran.

828 L.­T. Lim et al. / Progress in Polymer Science 33 (2008) 820–852

Fig. 10. Thermal degradation of PLA. Adapted from McNeill and Leiper [74] by permission of Elsevier B.V.

the chain reaction [74]. Although acetaldehyde is consid­

ered to be non­toxic and it is naturally present in many

foods, the acetaldehyde generated during melt processing

of PLA must be minimized, especially if the converted PLA

(e.g., container, bottle, and films) are to be used for food

packaging. The migration of acetaldehyde into the con­

tained food can result in off­flavor which may impact the

organoleptic properties and consumer acceptance of the

product [75–77].

From the production point of view, the formation of

lactide due to depolymerization is undesirable. Besides

reducing PLA melt viscosity and elasticity, the volatile lac­

tide formed can result in fuming and/or fouling of the

processing equipment such as chilled rollers, molds and

tooling surfaces [78]. The latter is characterized by the

gradual building up of a layer of lactide on the equipment

surfaces, commonly known as plate out. To overcome this

problem, the temperature of the equipment is generally

elevated to reduce the tendency of condensation of lactide.

Taubner and Shishoo showed that the moisture content

of resin, temperature, and residence time of PLA melt

during extrusion are important contributors to molecular

weight drop of the polymer during extrusion [72]. Pro­

cessing of dried PLLA with initial Mn of 40,000 g/mol in

a twin­screw extruder at 210 ◦C caused the Mn to drop to

33,600 and 30,200 g/mol, when screw rotation speeds of

120 and 20 rpm were used, respectively. Using the same

120 and 20 rpm screw speeds but processing at 240 ◦C,

the Mn values decreased dramatically to 25,600 and

13,600 g/mol, respectively. In contrast, Mn for extruded

articles produced from wet resins (equilibrated at 20 ◦C

65% RH to give 0.3%, w/w, moisture content) were 18,400

and 12,000 g/mol, respectively. These results highlighted

the importance of minimizing the residence time and

process temperature during PLA extrusion. From a resin

formulation point of view, the residual polymerizing

catalysts present in the resin are also known to catalyze

the reverse depolymerization and hydrolysis reactions

[48,79]. This may partially explain the large variation of

molecular weight drop for melt processed PLA reported

in the literature. For instance, Witzke, Gogolewski et al.

and Perego et al. reported molecular weight losses for

injection molded PLA parts of 5–52%, 50–88% and 14–40%,

respectively [48,59,80]. To stabilize the polymer during

melt processing, the removal or deactivation of the residual

catalyst is important to minimize the molecular weight

loss which will impact the mechanical properties of the

PLA parts. Strategies to improve the melt stability of PLA

can be found in patent publications [79,81,82]. Due to

the different processes and technologies used, the melt

stability of PLA polymer may be different from supplier to

supplier. Injection molded PLA made from properly dried

good quality PLA resins and optimal processes should

exhibit molecular weight loss of 10% or less [83].

7. Processing of PLA

7.1. Drying and extrusion

Prior to melting processing of PLA, the polymer must be

dried sufficiently to prevent excessive hydrolysis (molec­

ular weight drop) which can compromise the physical

properties of the polymer. Typically the polymer is dried to

less than 100 ppm (0.01%, w/w). Natureworks LLC, one of

the main suppliers for PLA polymers, recommended that

resins should be dried to 250 ppm (0.025%, w/w) mois­

ture content or below before extrusion. Processes that

have long residence times or high temperature approaching

240 ◦C should dry resins below 50 ppm to achieve maxi­

mum retention of molecular weight [84,85]. Drying of PLA

L.­T. Lim et al. / Progress in Polymer Science 33 (2008) 820–852 829

Table 5

Drying half times for PLA pellets under−40 ◦C dew point and air flow rate

of 0.016 m3/(min kg) [108]

Drying temperature (◦C) Drying half time (h)

Amorphous pellets

40 4.0

Crystalline pellets

40 4.3

50 3.9

60 3.3

70 2.1

80 1.3

100 0.6

takes place in the temperature range of 80–100 ◦C. The

required drying time is dependent on the drying temper­

ature (Table 5). Commercial grade PLA resin pellets are

usually crystallized, which permits drying at higher tem­

peratures to reduce the required drying time. In contrast,

amorphous pellets must be dried below the Tg (∼60 ◦C) to

prevent the resin pellets from sticking together, which can

bridge and plug the dryer. It is noteworthy that because

PLA degrades at elevated temperatures and high relative

humidity, the resins should be protected from hot and

humid environments. Henton et al. reported that amor­

phous PLA can dramatically reduce its Mw in less than a

month when exposed to 60 ◦C and 80% RH (Fig. 11) [44].

To achieve an effective drying, the dew point of the

drying air should be−40 ◦C or lower. Drying of PLA is com­

monly achieved using a closed loop dual­bed regenerative

desiccant­type dryer. In this type of dryer, the resin pellets

are contained in a hopper that is purged with dry air at

elevated temperature. The dry air is generated by the des­

iccant bed. During the operation, one desiccant bed is in

the process air stream which removes moisture from the

resin, while the other stand­by bed is being regenerated

(Fig. 12). The hot air from the process stream removes the

moisture from the resin in the hopper. The air is then circu­

lated back to the dryer where it is cooled and the moisture

is stripped by the desiccant. The air is reheated before it is

channeled back to the hopper. When the dew point of the

Fig. 11. Plots of molecular weight loss of PLA versus time under different

environment conditions. Curves are based on the original data published

by Henton et al. [44].

Fig. 12. Typical closed loop dual­bed regenerative desiccant­type dryer

for drying PLA before extrusion.

process air is greater than the set point, the desiccant goes

into the regeneration cycle where the desiccant is heated

to desorb the moisture from the desiccant and vent it to the

atmosphere. Meanwhile, the process air is directed to the

stand­by desiccant which was previously dried.

Extrusion is the most important technique for contin­

uously melt processing of PLA. The plasticizing extruder

can be part of the forming machine systems for injection

molding, blow molding, film blowing and melt spinning.

Fig. 13 shows a schematic representation of the major com­

ponents of an extruder for an injection molding machine. A

typical screw consists of three sections: (1) feed section –

acts as an auger which receives the polymer pellets and

conveys the polymer into the screw; (2) transition sec­

tion (also known as compression or melting sections) –

flight depth decreases gradually, which compresses the pel­

lets to enhance the friction and contact with the barrel. In

order to segregate the molten polymer pool from the pellet

unmelted pellets, various barrier flight designs have been

adopted; (3) metering section – characterized by a constant

and shallow flight depth, which acts as a pump to meter

accurately the required quantity of molten polymer. The

l/d ratio, which is the ratio of flight length of the screw

to its outer diameter, determines the shear and residence

time of the melt. Screws with large l/d ratio provide greater

shear heating, better mixing, and longer melt residence

time in the extruder. Commercial grade PLA resins can typ­

ically be processed using a conventional extruder equipped

with a general purpose screw of l/d ratio of 24–30. Extruder

screws for processing PET, which are typically low­shear for

gentle mixing to minimize resin degradation and acetalde­

hyde generation, are also suitable for processing PLA resin

[14]. Another important screw parameter is the compres­

sion ratio, which is the ratio of the flight depth in the feed

section to the flight depth in the metering section. The

greater the compression ratio a screw possesses, the greater

the shear heating it provides. The recommended compres­

sion ratio for PLA processing is in the range of 2–3 [86].

830 L.­T. Lim et al. / Progress in Polymer Science 33 (2008) 820–852

Fig. 13. Typical geometries of a screw for single­screw extruder.

During the plasticizing process, PLA resin pellets are

fed from a hopper near the end of a barrel. The screw,

driven by an electric or hydraulic motor, rotates and trans­

ports the material towards the other end of the barrel. The

heat required for melting is provided by the heater bands

wrapped around the barrel. As the screw rotates, the flights

shear and push the polymer against the wall of the barrel

which also provides frictional heat for melting the poly­

mer. The combined thermal energy from the heater and

frictional heat due to friction between the plastic and the

screw and barrel, provide sufficient heat to raise the PLA

polymer above its melting point (170–180 ◦C) by the time

it reaches the end of the barrel. To ensure that all the crys­

talline phases are melted and to achieve an optimal melt

viscosity for processing, the heater set point is usually set

at 200–210 ◦C.

7.2. Injection molding

Injection molding is the most widely used converting

process for thermoplastic articles, especially for those that

are complex in shape and require high dimensional preci­

sion. All injection molding machines have an extruder for

plasticizing the polymer melt. Unlike a standard extruder,

the extruder unit for injection molding machine is designed

such the screw can reciprocate within the barrel to pro­

vide enough injection pressure to deliver the polymer melt

into the mold cavities (Fig. 14). Most injection molding

machines for PLA are based on the reciprocating screw

extruder, although two­stage systems, which integrate a

shooting pot and extruder in a single machine, have also

been deployed for injection molding of preforms for PLA

bottles. The two­stage system consists an in­line extruder

integrated to a shooting pot. The extruder plasticizes and

feeds the melt into the shooting pot under relatively low

injection pressure, from which the melt is injected into the

hot runner under high pressure by a plunger in the shooting

pot. While the reciprocating machine must stop the screw

during the injection and packing phases, the screw for the

two­stage machine can rotate during the majority of the

cycle. The two­stage system presents some advantages over

its reciprocating counterpart, including shorter cycle time,

small screw motor drive, more consistent melt quality, and

more consistent shot size [87].

A typical cycle for an injection molding machine is pre­

sented in Fig. 15. The beginning of mold close is usually

taken as the start of an injection molding cycle. Immedi­

ately after the molds clamp up, the nozzle opens and the

screw moves forward, injecting the polymer melt into the

mold cavity. To compensate for the material shrinkage dur­

ing cooling in the mold, the screw is maintained in the

forward position by a holding pressure. At the end of the

holding phase, the nozzle is shut off and the screw begins to

recover, while the part continues to be cooled in the mold.

During the recovery phase, the screw rotates and conveys

the polymer forward along the screw. At the same time,

the screw is allowed to slide backward within the barrel

against a controlled back pressure exerted on the screw

by a hydraulic cylinder. To ensure that the part is dimen­

sionally stable enough to withstand the opening stroke the

Fig. 14. Major components of an injection molding machine showing the extruder (reciprocal screw) and clamp units.

L.­T. Lim et al. / Progress in Polymer Science 33 (2008) 820–852 831

Fig. 15. Typical cycle for an injection molding process.

molds, sufficient cooling time must be given. In the mold­

ing cycle, heat removal takes place predominantly during

the fill, hold and cool phases, although mold opening phase

also contributes to partial cooling since one side of the part

(core­contacting side) is still being cooled prior to ejection.

Cycle time is an important process parameter which is

often minimized to maximize the production throughput.

To reduce the cycle time, it is quite common to transfer

the partially cooled injection molded article to a post­mold

cooling device, to provide an extended cooling of the part

outside the molds, either by direct contact on a chilled sur­

face and/or by forced air. From Fig. 15, it is also evident

that minimizing the duration for non­process events, such

as mold opening, part ejection and mold closing is also

important for reducing the cycle time. Lowering mold tem­

perature can also increase the heat extraction rate from

the polymer. Nevertheless, the propensity of lactide con­

densation on the cold tooling surfaces, which can affect

the surface finish and weight of the molded articles, limits

the minimal temperature that can be used during injec­

tion molding of PLA to 25–30 ◦C. The use of molds with

polished surfaces, in conjunction with an increased injec­

tion speed during fill, can also reduce the deposition of the

lactide layers.

The fill, hold and cool events that take place during

injection molding have an important implication on the

shrinkage of the injection molded articles. This effect can

be best elucidated using a pressure–volume–temperature

(PVT) diagram. Fig. 16 shows PVT diagrams for PLA from two

references [88,89]. The different profiles shown here are

likely due to the different grades of PLA used. During injec­

tion molding, the polymer is first subjected to isothermal

injection of the polymer melt into the mold cavity, dur­

ing which the pressure increases as the polymer is being

injected and packed to the holding pressure (trace ab in

Fig. 16). The polymer then undergoes isobaric cooling in

the holding phase (trace bc), followed by isochoric cool­

ing. When the polymer cools below the freezing point, the

gate freezes and the pressure in the mold cavities drops

to one atmospheric pressure (trace cd). In the last cool­

ing phase, the article continues to cool isobarically to room

temperature (trace de). The change in specific volume dur­

ing the final isobaric cooling (trace de) dictates the extent

of part shrinkage. The hold pressure and temperature play

an important role in determining how much the molded

article shrinks.

The PVT relationship can be modeled mathematically,

such as by using the modified two­domain Tait model

[90–92]. This model is often used for numerical simula­

tion of injection molding processes involving finite element

analysis for predicting the shrinkage behavior of injection

Fig. 16. PVT plots for PLA based on the data from Sato et al. and Natureworks LLC [88,89]. The continuous lines represent the fitted results based on the

two­domain modified Tait model (Eq. (8)).

832 L.­T. Lim et al. / Progress in Polymer Science 33 (2008) 820–852

Fig. 17. Effects of temperature and time on the aging of injection molded 4% d­lactide PDLA specimens. (A) DSC curves of PLA aged at room temperature

for various aging times. (B) DSC curves of PLA annealed for 24 h at different temperatures. Plots are created based on the data from Cai et al. [47].

molded articles. The modified two­domain Tait PVT model

takes the form:

v(T, p) = V0(T)

[

1− C ln

(

1+p

B(T)

)]

+ Vt(T, p) (9)

where v(T, p) is specific volume at temperature T and pres­

sure P; V0 is specific volume at zero gauge pressure and C is

a constant, 0.0894. When the temperature of the material

is greater than the transition temperature, V0(T) and B(T)

are determined by b1m, b2m, b3m, b4m and b5 as follows:

V0 = b1m + b2m(T − b5) (10)

B(T) = b3m exp[−b4m(T − b5)] (11)

In contrast, when the material temperature is lower than

the transition temperature, V0(T) and B(T) are determined

by b1s, b2s, b3s, b4s and b5 as follows:

V0 = b1s + b2s(T − b5) (12)

B(T) = b3s exp[−b4s(T − b5)] (13)

Because the transition temperature, Ttrans(P), is often pres­

sure dependent, it is often correlated with pressure and

the transition temperature at zero gauge pressure (b5) as

follows:

Ttrans(P) = b5 + b6p (14)

For non­amorphous materials, an additional transition

function is required:

Vt(T, p) = b7 exp[b8(T − b5)− b9p] (15)

The estimated parameter values for the modified two­

domain Tait model are shown in the inserts in Fig. 16.

In general, injection molded PLA articles are relatively

brittle. The brittleness of PLA has been attributed to the

rapid physical aging of the polymer since ambient tem­

perature is only about 25 ◦C below the Tg [37,46,48]. The

aging of PLA can be evaluated by studying the Tg region of

a DSC scan. By measuring the development of endother­

mic enthalpy relaxation 1Hrel using DSC on injection

molded samples made from PLA (96% l­lactide), Cai et al.

showed that 1Hrel increased with increasing aging time

[47] (Fig. 17). They also showed that as the aging temper­

ature increased towards the Tg, the rate of physical aging

also became faster. However, when the aging temperature

went above the Tg (60 ◦C), the excess enthalpy relaxation

was reduced, indicating that physical aging was no longer

taking place when the aging temperature was above Tg [47].

Celli and Scandola observed a similar aging trend for PLLA

using DSC and a dynamic mechanical analyzer [46]. They

observed that the extent of aging increased with decreas­

ing molecular weight (i.e., 1Hrel increased with decreasing

molecular weight), which was attributed to the increased

chain terminals that possess higher motional freedom than

the internal chain segments [46]. The physical implication

of aging was elucidated by Witzke, who reported that injec­

tion molded articles tested immediately after quenching

to very cold temperatures exhibited a much larger exten­

sion to break. However, when the molded specimens were

aged at room temperature for 3–8 h, they became very

brittle [48]. This phenomenon was attributed to the reduc­

tion of free volume of the polymer due to rapid relaxation

towards the equilibrium amorphous state. Aging below Tg is

exclusively related to the amorphous phase of the polymer;

accordingly, increasing the crystallinity of the polymer (e.g.,

by adjusting d­isomer composition or the use of nucleat­

ing agents) will reduce the aging effect. Furthermore, the

crystallites formed also act like physical crosslinks to retard

the polymer chain mobility. However, amorphous injection

L.­T. Lim et al. / Progress in Polymer Science 33 (2008) 820–852 833

molded articles which are intended for further process­

ing (e.g., preforms for stretch blow molding), the storage

conditions prior to subsequent processing may need to

be controlled. Moreover, process parameters such as mold

temperature, packing pressure, cooling rate, and post­mold

cooling treatment are expected to influence the PLA aging

behavior as well.

7.3. Stretch blow molding

Due to the recent consumers’ heightened environmen­

tal awareness, there is a sustained interest from the food

industry to replace the existing non­biodegradable ther­

moplastics with PLA for certain beverage products. To date,

PLA bottles are predominantly used for beverages which are

not sensitive to oxygen (e.g., flat water beverages, pasteur­

ized milk). While barrier properties of PLA bottles may be

improved by various technologies (multilayer structures,

external coating, internal plasma deposition, oxygen scav­

enger), their implementation is currently limited due to

higher production costs.

The production of PLA bottles is based on injection

stretch blow molding (ISBM) technique. This process pro­

duces biaxial orientated PLA bottle with much improved

physical and barrier properties compared to injec­

tion molded amorphous PLA. The molecular orientation

induced during the ISBM process decreases the effect of

aging by stabilizing the polymer free volume [48]. The

crystallites produced during strain­induced crystallization

also reduce the aging effect since they can act as physical

crosslinks to stabilize the amorphous phase, thereby reduc­

ing its brittleness. Similar effects have been reported for

semicrystalline PET [93]. The ISBM process for PLA bottles

is depicted in Fig. 18. It involves first the formation of pre­

form (also known as parison) using an injection molding

machine. The preform is then transferred to a blow mold­

ing machine where it is stretched in the axial direction and

blown in the hoop direction to achieve biaxial orientation

of the polymer. In the blow molding machine, the preform

is heated in front of several banks of infrared heater to tem­

peratures (85–110 ◦C) suitable for blow molding (Fig. 18a).

Different power settings are usually applied to the infrared

heaters to give a temperature profile optimal for stretch­

ing the preform into bottle with uniform wall thickness

distribution. Frequently, reheat additives, such as carbon

black dispersed in a liquid carrier, are added to the resin

in the extruder to increase its infrared energy absorption.

PLA preforms have a tendency to shrink after reheat, espe­

Fig. 18. Injection stretch blow molding (ISBM) of PLA bottle.

834 L.­T. Lim et al. / Progress in Polymer Science 33 (2008) 820–852

cially regions near the neck and the end cap where the

residual injection molding stresses are the greatest. This

may be moderated through proper preform design, with

gradual transition regions. When the preform has attained

the optimal temperature, it is transferred to the blow mold

(Fig. 18b). The blow nozzle is lowered to seal the preform

finish, while the stretch rod travels towards the preform,

at a typical speed of 1–1.5 m/s, and stretches the preform

to the base cup (Fig. 18c–e). During the preblow phase

(Fig. 18d and e), compressed air of 0.5–2.0 MPa is admitted

to the preform through the blow nozzle to partially inflate

the preform to prevent it from touching the stretch rod dur­

ing the axial stretching. When the stretch rod arrives at the

base cup and pins the preform to the mold base, the air pres­

sure ramps up to 3.8–4.0 MPa to fully inflate the preform.

This forces the inflated preform to take the shape of the

blow mold and to imprint the surface details of the bottles

(Fig. 18f and g). The high blow pressure is maintained for

several seconds to allow the bottle to cool down sufficiently

before discharging the bottle.

The aforementioned process is known as the two­stage

process. In contrast, the one­stage process entails the injec­

tion and blow molding of the preform within the same

machine equipped with both injection and blow mold­

ing units. In this process, the injection molded preform

is partially cooled down to 100–120 ◦C and then stretch

blown in the blow molding station. Fig. 19 summarizes the

thermal history of PLA from resin pellets to bottle for the

two processes. As shown, PLA preform made in the one­

stage process does not go through the aging process during

which the polymer tends to embrittle. Thus, PLA preforms

intended for one­ and two­stage processes may need to be

designed and processed differently. The neck finish of the

preform is highly amorphous and is quite brittle. Therefore,

the neck finish must be designed such that the side wall is

thick enough to prevent the neck from blowing out or crack­

ing due to the compression load from the blow nozzle. The

blow mold temperature for PLA is typically set at around

35 ◦C. Because the base of the bottle tends to be quite thick,

the residual heat can cause the base to roll out after the

bottle is ejected from the blow mold. This problem can be

overcome by incorporating radial ribs to reinforce the base

and/or chilling the base mold insert to a temperature lower

than the mold halves [94].

Similarly to PET, PLA exhibits strain­hardening when

stretched to high strain. This self­leveling phenomenon is

desirable for blow molding of preforms to achieve optimal

bottle side wall orientation and minimize wall thickness

variation. Since strain­hardening occurs only when the PLA

is stretched beyond its natural stretch ratio, the preform

must be designed to match the target bottle size and shape,

such that optimal stretch ratios are achieved during blow

molding (Fig. 20). Preforms that are under­stretched will

result in bottles with excessive wall thickness variation,

weak mechanical properties and poor aesthetic appeal (e.g.,

lens defect below the support ledge region). In contrast,

overstretched bottles can also result in stress whitening

due to the formation of micro­cracks on the bottle surfaces

that diffract light. Typical commercial grade PLA resins for

bottle applications require preform axial stretch ratios of

2.8–3.2 and hoop stretch ratios of 2–3, with the desirable

planar stretch ratio of 8–11 [94,95]. It is noteworthy that the

ultimate amount of crystallinity after stretching decreases

with the decreasing stereoisomeric purity of the polymer

[38]. Accordingly, the optimal stretch ratios depend on the

grade of PLA used.

Preform designs are often proprietary, and therefore

there is a lack of information in the open literature. An opti­

mal preform design should meet the minimum required

stretching which is above the natural stretch ratio, by vary­

ing the shape, diameter, length, blend radius, and transition

features, to meet the part weight requirement. Depending

Fig. 19. Thermal history of PLA polymer during one­ and two­stage PLA bottle manufacturing.

L.­T. Lim et al. / Progress in Polymer Science 33 (2008) 820–852 835

Fig. 20. Schematic representation of PLA preform (left) and bottle (right), showing their key features and main stretch ratios used for preform design.

on the shape of the bottle, subtle but critical features such

as transition shape (reverse versus standard taper), step

changes, and pinch points on the core and cavity may also

be incorporated in the preform design. Since the stretching

behavior of PLA is similar to PET but not entirely the same,

conversion of materials using existing PET preform designs

may be feasible, although design modifications are often

required to achieve an optimized bottle.

7.4. Cast film and sheet

PLA with l­lactide contents of 92–98% have been

successfully extruded using conventional extruders. The

production of PLA film and sheet is practically identical; the

main difference between them is their stiffness and flexi­

bility due to the difference in their thicknesses. Typically,

films are ≤0.076 mm (0.003 in.) in thickness, while sheets

are typically≥0.25 mm (0.01 in.). In cast film extrusion, the

molten PLA is extruded through a sheet die and quenched

on polished chrome rollers that are cooled with circulating

water. Due to the thermal sensitivity of PLA, the use of exter­

nal deckles on the die should be avoided since the degraded

resin behind the deckles can lead to edge instability. Usually

the die gap is set to 10% or 25–50 mm (1–2 mils) greater than

the target sheet thickness [84]. Ljungberg et al. extruded

neat PLA in a Haake Rheomex 254 extruder with a Rheo­

cord 90 drive unit (Karlsruhe, Germany) [96]. The 19.3 mm

diameter screw has a compression ratio of 2:1 and l/d ratio

of 25. In this study, the temperatures for the feeding zone,

the barrel and the die were 160, 180, and 175 ◦C, respec­

tively [96]. Similar extruder temperature profiles were used

by Gruber et al. [79].

Sheet and film forming can be achieved on a three­roll

stack. Because of the low melt strength of PLA, horizontal

roll stacks configuration is preferred. To avoid the con­

densation of lactide monomers and slippage of web on

the rollers, relatively high roller temperatures (25–50 ◦C)

are usually used. Lactide monomer buildup around the

die could be further prevented by using an exhaust sys­

tem. Nevertheless, extreme high temperatures should be

avoided as the web will stick to the rollers, resulting in

poor quality sheet. To reduce the chance of trapping air

and reduce film or sheet defects, one resin supplier recom­

mended that the die be positioned as close as possible to

the entrance nip and slightly higher than the nip to accom­

modate the slight drooping of the molten PLA web [84].

To cast PLA film, Ljungberg et al. used a 200­mm fishtail

die with a 300–400 mm split gap and a casting air gap of

15 mm [96]. Generally, hydraulic rolls stands, capable of

producing pressure around 800–900 lbs/linear inch of die is

required to prevent floating of the rolls which would result

in uneven PLA surfaces, edge instability, and neck­in [84].

Good contact between the web and rolls is also important

to minimize lactide buildup. Casting of PLA film usually

requires edge pinning (electrostatic or low pressure air) to

eliminate streaking, reduce neck­in, and improve edge sta­

bility [97]. Slitting and web handling of PLA is similar to

PS. Edge trimming of PLA should be carried out with rotary

shear knives since razor knives may yield rough edges and

web breaks. Winding of the PLA web should be done with

good tension control in order to obtain a consistent gauge.

Similar to PP, PET and PS films, the physical properties

of PLA films can be enhanced through orientation. Uniaxial

orientation of PLA is achieved in conventional machine­

direction orientation (MDO) rolls. Since PLA tends to neck in

during drawing, nipped rolls are usually required. Through

mechanical drawing, it is possible to improve thermal and

impact resistance of the PLA film or sheet to a level simi­

lar as oriented polystyrene (OPS), oriented polypropylene

(OPP) or polyester. An oriented PLA film can be obtained

by stretching it to two to ten times its original length at

60–80 ◦C [51], which is much lower compared with OPP

and PET. Typical drawing temperatures for PLA films in the

machine (MD) and transverse directions (TD) are presented

836 L.­T. Lim et al. / Progress in Polymer Science 33 (2008) 820–852

Table 6

Recommended drawing conditions in the machine and transverse direc­

tion for PLA [97]

Section Temperature range (◦C)

Machine direction preheat 45–65

Slow draw 55–70

Fast draw 70–75

Annealing 45–55

Transverse direction

Preheat 65–70

Draw 70–85

Annealing 125–140

in Table 6. In generally, for 98% l­lactide PLA, machine

direction orientation of 2–3× is expected, while transverse

stretch ratios of 2–4× may be used. At higher d­lactide

contents, the machine and transverse stretch ratio can be

increased. Fig. 21 shows a typical extrusion cast line for

producing biaxially oriented PLA film.

The orientation in PLLA films depends on the draw

rate, temperature and ratio. High strain rate, low tem­

perature and high stretch ratio favor strain­induced

crystallization during orientation. Taking the competitive

crystallization and relaxation effects into consideration,

Lee at al. concluded that the optimal drawing tem­

perature to obtain highly oriented PLLA films (Mw of

190,000 g/mol) is about 80 ◦C [63]. In contrast, Gruber et al.

used somewhat lower temperatures for biaxial orientation

of 100,000–150,000Mn PLA polymer with 10–20% meso­

lactide content (65–72 and 20 ◦C for preheat and cooling

rolls, respectively, for MD stretching; 63–70 ◦C and circu­

lated ambient air cooling for TD drawing) [79]. Ou and

Cakmak prepared biaxially oriented PLA films by stretching

cast PLA in both MD and TD to different ratios, followed by

annealing these films at elevated temperatures to induce

crystallinity and dimensional stability [62]. Their wide

angle X­ray (WAXS) results showed that the development

of crystalline order and orientation were dependent on

the mode of orientation. They observed that simultaneous

biaxial stretching of PLA film resulted in poor crystalline

order, while sequential stretching promoted a greater crys­

talline order [62]. Hence, the properties of PLA films are

expected to change depending on the stretching sequence

used during the orientation process.

PLA has excellent optical properties and high modulus.

However, it has low elongation, tear and burst strengths.

To overcome these shortcomings, PLA is often coextruded

with other polymers to form multilayer structures to

enhance its properties. For instance, to reduce electrostatic

buildup, Rosenbaum et al. disclosed methods for forming

biaxially oriented multilayer films made of one PLA­based

layer and two outer layers consist of PLA and glycerol fatty

acid esters to achieve films with antistatic surfaces [98]. The

extruder temperatures used ranged from 170 to 200 ◦C with

the take off roll set at 60 ◦C. The biaxial orientation took

place sequentially, first at 68 ◦C in the machine direction

by rollers running at different speeds, followed by trans­

verse direction stretching using a tenter frame at 88 ◦C.

Stretch ratios were 2.0 and 5.5 for machine and transverse

direction stretching, respectively. To impart dimensional

stability to the film, heat­setting was conducted at 75 ◦C.

Noda et al. disclosed a method of coextruding multi­

layer laminate film consisting of polyhydroxyalkanoate

(PHA) copolymer (copolymer of 3­hydroxybutyrate with 3­

hydroxyhexanoate) and PLA to impart softness to the PLA,

and at the same time reduce the tackiness of the PHA. By

preventing the web from sticking to itself or the processing

equipment, the speed of production and product quality

can be improved [99].

PLA films tend to have higher surface energy than

untreated polyolefin films. Gruber et al. reported surface

energy of about 44 dynes/cm for pure PLA films [79].

The surface energy values for 98% l­lactide and 94% l­

lactide films were reported as 42 and 34–38 dynes/cm,

respectively [100]. Higher surface energy will provide more

satisfactory printing properties without surface treatment.

If higher surface energy is needed for downstream process­

ing, the surface can be treated by corona discharge.

7.5. Extrusion blown film

In extrusion blow film process, molten PLA is extruded

to form a tube using an annular die. By blowing air through

the die head, the tube is inflated into a thin tubular bub­

ble and cooled. The tube is then flattened in the nip rolls

and taken up by the winder (Fig. 22). The ratio of bubble

diameter to the die diameter is called the blow­up­ratio

(BUR). BUR ratios of 2:1–4:1 with the die temperature

of 190–200 ◦C have been used for extrusion blowing of

PLA films [101,102]. By varying the BUR, screw speed, air

pressure, and winder speed, films of different thicknesses

(∼10–150 mm) and degree of orientation can be achieved.

Fig. 21. Biaxial oriented extrusion cast film machine.

L.­T. Lim et al. / Progress in Polymer Science 33 (2008) 820–852 837

Fig. 22. Extrusion blown film line.

PLA has a specific density of about 1.24 g/cm3 which is

much higher than polyolefins (0.91–0.96 g/cm3). While PLA

may be processed in extruders designed for polyolefins,

if the extruder is already operating at close to maximum

power of the screw drive, the extruder may not have enough

power to process PLA due to the substantial higher den­

sity for PLA [103]. Compared to polyolefins, PLA has weaker

melt strength, and therefore, the formation of a stable bub­

ble during extrusion blowing is more difficult. As a result,

extrusion blowing of PLA film often requires the use of addi­

tives, such as viscosity enhancers to strengthen its melt

strength. These additives protect the polymer from degra­

dation and/or couple polymer chains to attenuate overall

loss of molecular weight and viscosity of the polymer melt.

The formulation of coupling agents is often proprietary. One

commercially available coupling agent for PLA is made up

of copolymer of styrene, methyl methacrylate and glycidyl

methacrylate [102]. Sodergard et al. disclosed a method to

stabilize PLA and enhance its melt strength by adding an

organic peroxy compound (e.g., tert­butylperoxybenzoate,

dibenzoylperoxide, tert­butylperoxyacetate) during melt

processing, wherein the peroxide is added in about 0.01–3%

by weight of PLA [101].

Since PLA films are quite stiff and have much lower elon­

gation than polyolefins, collapsing of bubble in the nips

rolls tends to produce wrinkles which tend to permanently

remain in the film due to the high dead­fold properties of

PLA. This problem can be overcome by incorporating fillers

into PLA during extrusion. To reduce the adhesion between

films, Hiltunen et al. blended PLA with triacetin plasticizer

(glycerol triacetate), together with various anti­adhesion

agents, such as talc, TiO2 and CaCO3. They claimed that the

bursting strengths of the resulting blown films were better

than typical polyethylene and PP films [104]. Slip additives

(e.g., oleamide, stearamide, N,N′­ethylene bisstearamide,

oleyl palmitamide) have also been added to reduce the

coefficient of friction between overlapping films [102]. Typ­

ically, slip additive of less than 0.5–1.0% by polymer weight

is used, as excessive amounts will compromise the abil­

ity of print inks, stickers to adhere to the film surface. To

avoid the use of copolymerization techniques, blending, or

plasticizers, Tweed et al. developed a method to obtain PLA

blown films by elevating the viscosity of PLA through suc­

cessive steps in a polymer cooling unit or by internal cooling

of the die mandrel using air or liquid fluid to control the

temperature of the die [102]. Mitsui Chemicals successfully

developed PLA­based films by copolymerization technol­

ogy, and it is commercializing it as one of the LACEA brand

resins [51].

7.6. Thermoforming

Thermoforming is commonly used for forming pack­

aging containers that do not have complicated features.

PLA polymers have been successfully thermoformed into

disposable cups, single­use food trays, lids, and blister pack­

aging.

Fig. 23 shows the typical steps for thermoforming of

PLA container. In this process, PLA sheet is heated to soften

the polymer, forced either pneumatic and/or mechanically

against the mold, allowed to cool, removed from the mold,

and then trimmed. Heating of PLA sheet for thermoform­

ing is generally achieved by infrared red (IR) radiation

from heater elements. Each polymer has an optimum IR

absorbance frequency in the IR region. Therefore, the heater

element should be set at the temperature at which the

majority of energy is absorbed by the polymer. For PS, the

ideal wavelength is 3.2–3.7 mm [105]. Values for PLA have

838 L.­T. Lim et al. / Progress in Polymer Science 33 (2008) 820–852

Fig. 23. Main steps for thermoforming process.

not been reported in the literature. In general, the thermo­

forming temperatures for PLA are much lower than other

conventional thermoformed plastics (e.g., PET, PS, and PP)

in the range of 80–110 ◦C when the sheet enters the mold

[106,107].

Typically, aluminum molds are used for thermoform­

ing PLA containers. Molds, trim tools and ovens designed

for thermoforming of PET, high impact polystyrene (HIPS)

and OPS can be used for forming PLA containers. However,

molds for thermoforming of PP may not be used inter­

changeably for PLA, since PP shrinks more considerably

than PLA during cooling. For a given part thickness, cool­

ing times required for PLA containers in the mold tend to be

higher than PET and PS containers due to the lower thermal

conductivity and Tg for PLA polymers. Table 7 compares the

thermal properties of PLA, PS and PET.

Orientation increases toughness of PLA containers.

Regions of PLA articles that are highly drawn are less brit­

tle as compared to flanges and lips that received minimal

orientation. Extruded sheet prior to thermoforming is rela­

tively brittle at room temperature. To ensure smooth travel

of the web and to prevent web breakage, a tight radius

should be avoided in the unwind stations and skeleton

rewind stations. A minimum rewind radius of 25 cm is

recommended [108]. If PLA sheet needs to be trimmed

before thermoforming, it should be heated to tempera­

tures near 90 ◦C to prevent cracking. Storage conditions

for the sheet stock need to be controlled as well. As a

guide, PLA should not be exposed to temperature above

40 ◦C or to RH above 50% as the sheet will block and resist

unwinding due to its low heat deflection temperature. After

thermoforming, precaution should be taken to store PLA

below 40 ◦C since Mw breakdown can accelerate when it

is exposed to elevated temperature (Fig. 11). A compari­

Table 7

Thermal properties of PLA, PS, and PET [106]

PLA PS PET

Thermal conductivity (×10−4 cal. cm−1 s−1 ◦C−1) 2.9 4.3 5.7

Heat capacity (cal. g−1 ◦C−1) 0.39 0.54 0.44

Glass transition temperature (◦C) 55 105 75

Thermal expansion coefficient (×10−6 ◦C−1) 70 70 70

son of the mechanical, physical and barrier properties of

thermoformed PLA, PS, and PET containers showed that

PLA containers outperform PET and PS at lower temper­

atures [109]. Moreover, the use of 40–50% PLA regrind did

not significantly change in the container performance [9].

7.7. Foaming

Due to their biocompatibility and large surface area,

PLA foams have a niche in tissue engineering and medical

implant applications [110–112]. Foaming of PLA is generally

carried out by dissolving a blowing agent in the PLA matrix.

The solubility of the blowing agent is then reduced rapidly

by producing thermodynamic instability in the structure

(e.g., temperature increase or pressure decrease), to induce

nucleation of the bubbles. To stabilize the bubbles, the foam

cells are vitrified when the temperature is reduced below

the Tg of the polymer [113,114].

Various foaming strategies have been adopted to reduce

PLA density and improve foam mechanical properties. Di

et al. used 1,4­butanediol (BD) and 1,4­butane diisocyanate

(BDI) as chain extenders to increase the molecular weight of

PLA so that its viscoelastic properties are more optimal for

foaming. They produced modified PLA samples by sequen­

tially adding different ratios of BD and BDI in a Haake melt

mixer operating at 170 ◦C and mixer speed of 60 rpm under

a nitrogen atmosphere. Tin(II) 2­ethylhexanoate was added

as a catalyst at 0.05 wt% of PLA. They found that the chain­

extender modified PLA produced foams with reduced cell

size, increased cell density and lowered bulk foam density

as compared to the neat PLA foam control [113]. Mikos et

al. prepared PLLA membranes with and without sodium

chloride, sodium tartrate, and sodium citrate by solvent­

casting techniques [114]. The PLLA and PLLA/salt composite

membranes were foamed by heating them at 195 ◦C (15 ◦C

higher than Tm) for 90 min and then quenched in liquid

nitrogen for 15 min. They were able to produce membranes

with porosity as high as 93% with a desired surface/volume

ratio depending on the salt used. Ajioka et al. disclosed in

a patent the method of manufacturing PLA foams suitable

for use as disposable food trays, cups, thermal insulators,

and cushioning materials [115]. Their approach involves

L.­T. Lim et al. / Progress in Polymer Science 33 (2008) 820–852 839

mixing various proportions of PLLA and PDLA together

with 0.5% talc (w/w) in an extruder at 200 ◦C. An expand­

ing agent, either dichlorodifluoromethane or butane was

charged under pressure into the extruder. The mixture was

cooled to 140 ◦C and extruded through a slit die to give sheet

foam. An alternate method adopted by these inventors

involved mixing and heating azodicarbonamides powder

(a food additive) with PLA resins using an extruder, in

which the azodicarbonamide decomposed, thereby releas­

ing nitrogen gas to induce the formation of bubbles [115].

Another patent described a method for injection molding

of PLA foams by adding 15–25 wt% of solvent to PLA dur­

ing extrusion [116]. Solvents reported to be suitable here

were methyl formate, ethyl formate, methyl acetate, propyl

acetate, dioxane and methyl ethyl ketone.

Loose­fill packaging materials provide cushioning, pro­

tection, and stabilization of packaged goods during

shipping. Over the past decade, the use of expanded PS

foams for loose­fill packaging has declined due to the

replacement with the environmentally more benign starch­

based expanded foams. To overcome the hydrophilic nature

of starch, these biobased foams are often blended with

petroleum polymers. Recently, PLA has been use for blend­

ing with starch during foaming. Guan et al. extruded

PLA and acetylated starches, along with 5% talc (w/w)

and various amounts of ethanol in a co­rotating conical

twin­screw extruder (180 rpm screw speed; 160 ◦C barrel

and die temperatures) [117]. The blends were conditioned

for 24 h at 25 ◦C before extrusion. The authors observed

that ethanol functioned as an effective blowing agent to

expand the foams, and it acted as a solubilizing agent to

depolymerize PLA and starch to form homogenous dough.

They observed that increasing the PLA content caused an

increased foam expansion [117]. In an attempt to improve

the physical/mechanical properties, and moisture resis­

tance of starch­based foams, Preechawong et al. prepared

starch–PLLA hybrid foams by baking PLA–starch mixtures

in a hot mold at 220 ◦C for 2 min [118]. They found that the

addition of PLA improved the ultimate tensile strength and

elongation at break, as compared to the starch foam, with

a concomitant increased resistance to water absorption.

Besides adjusting the cellular morphology of PLA

through the use of different blowing agents and pro­

cess optimization (pressure and temperature adjustments),

much recent research activity has been directed at affect­

ing the size and formation of the cellular structure by

dispersing MMT nanoclay particles in the bulk material

[119]. Ray and Okamoto foamed pure PLA and PLA com­

posite with organically modified MMT in a batch system

below Tm. These authors found that foamed PLA reinforced

with MMT produced homogeneous cells with closed cell

structure with a diameter of 2.6 mm; however, in the case

of neat PLA the cells were non­uniform with average cell

size of 230 mm [23]. They concluded that the MMT parti­

cles acted as nucleating agents for the cell formation. Ema

et al. also incorporated nanoclay in PLA for batch foam­

ing using supercritical carbon dioxide as a foaming agent.

These researchers noted that cells of ∼200 nm were local­

ized along the surfaces of the dispersed nanoclay and drew

a similar conclusion that the dispersed nanoclay acted as

nucleating sites for the cell formation [120]. Di et al. foamed

neat PLA and PLA reinforced with two commercial organ­

ically treated MMT samples. They found that the foam of

PLA/MMT exhibited a nicely interconnected, energetically

stable closed­cell structure with pentagonal and hexagonal

faces [121]. In contrast, pure PLA foam showed a relatively

large cell size (∼230 mm). The size of the cells decreased as

the amount of MMT increased, and it leveled off at higher

MMT concentrations. Di et al. concluded that it is possible

to foam PLA with different cellular structures by controlling

the amount of organoclay content [121].

Recently, supercritical CO2 has attracted consider­

able research attention as an environmentally friendly

solvent for many processes because it is inexpensive, non­

flammable and can be easily removed from the products.

A liquid becomes a supercritical fluid when both pressure

and temperature are above the critical point. In this state,

the substance possesses a combined gas­like viscosity and

liquid­like density, making it an excellent solvent for var­

ious applications. For CO2, its solubility and diffusivity in

many polymers tend to increase considerably when it exists

in the critical state, thereby facilitating the plasticization

of many polymers and enabling the forming process to be

conducted at lower temperatures [122]. Moreover, the crit­

ical conditions of CO2 at 31.1 ◦C and 7.38 MPa are readily

attainable within the safety of commercial and laboratory

setups [123,124]. The supercritical CO2 approach is based

on the larger Tg depression effect of supercritical CO2 on

polymers, which keeps the polymer in the liquid state at rel­

atively low temperatures. The sudden reduction in pressure

leads to the formation of CO2 nuclei which grow sponta­

neously. Meanwhile, as the pressure decreases, the Tg for

the polymer also elevates and eventually rises above the

foaming temperature, at which point the cellular structure

is locked in place to produce a cellular network. Fujiwara

et al. investigated the effect of d­ and l­lactide composi­

tions (1.0–28.5% d­lactide contents) on PLA foaming using

supercritical CO2 [125]. The samples were heated to 50 ◦C in

the reactor chamber followed by adding CO2. The pressure

of the vessel was ramped to 69 and 414 bar, respectively,

for amorphous and crystalline samples. The temperature

was increased stepwise while the pressure was kept con­

stant. The expansion of the polymer was monitored during

heating using a linear variable differential transducer. Fuji­

wara et al. reported that the average pore diameters for the

porous structures were 5.4 and 3.3 mm for 1 and 4.2% d­

lactide polymers, respectively, suggesting that the cellular

morphology was crystallinity dependent. In contrast, under

the same conditions employed, porous structures were not

detected for amorphous PLA samples with 10 and 28.5% d­

lactide content. Moreover, they also reported an increased

linear swelling of the polymer with decreasing crystallinity

and the porous supercritical CO2 treated PLA samples

have higher crystallinity than the as received polymers

[125]. Porosity of supercritical CO2 treated foams not only

depends on the pressure and temperature, but also on the

rate of its release from the polymer (Fig. 24) [123]. Mooney

et al. showed that the formation of PLA foams can also take

place below the critical conditions by exposing the poly­

mer to a pressure of 55 bar at 20–23 ◦C for 72 h, followed by

rapid depressurization to atmospheric pressure in 10–15 s

[126]. Pores of 10–100 mm were observed, although solid

840 L.­T. Lim et al. / Progress in Polymer Science 33 (2008) 820–852

Fig. 24. SEM micrographs of PDLLA polymer after processing in supercritical CO2 at 240 bar, 35 ◦C with (a and b) 12 min venting and (c and d) 60 min venting

[123]. Reproduced from Quirk et al. by permission from Elsevier B.V.

skin layers were present on the sample surface due to the

rapid diffusion of the dissolved gas from the surfaces [126].

Matuana investigated the effect of microcellular structures

on the mechanical properties of foamed PLA by using a

two­stage CO2 foaming process. The process first involved

saturating PLA samples with CO2 in a pressure chamber

at 5.5 MPa and room temperatures for 2 days. The CO2­

saturated samples were then removed from the chamber

and heated above Tg in a glycerin bath [127]. Compared with

the unfoamed PLA, Matuana reported a twofold increase in

impact resistance, up to twofold increase in strain at break,

and up to fourfold increase in toughness for the foamed

samples. The increased impact strength was attributed to

the presence of small bubbles which inhibited the crack

propagation by blunting the crack tip and increasing the

amount of energy needed to propagate the crack [127].

The low temperature supercritical CO2 processes reviewed

above are expected to find their place in the manufactur­

ing of structural foams for which mechanical properties are

critical, since thermal and hydrolytic degradation encoun­

tered in the typical thermal processes can be avoided.

7.8. Fiber spinning

High water vapor transmission rate of PLA often pre­

cludes its use in applications where moisture barrier

is critical. However, this property can be leveraged for

fabricating fibers used in garments (e.g., shirts, dresses,

underwear, shoes, etc.) to improve their “breathability”.

While PLA fibers are not as wettable as cotton, they exhibit

much greater water vapor transmission than polyester or

nylon fibers.

The manufacturing of PLA fiber is carried out either by

dry or melt spinning processes. Commercially PLA fibers

are generally produced using the melt spinning tech­

nique [128–136]. Here, PLA fibers are typically melt­spun

at approximately 185–240 ◦C through spinnerets with l/d

ratios of 2–10. The processing temperature is compara­

ble to polyolefins [137,138]. The melting temperature used

also depends on the optical purity of the polymer used.

Lower processing temperatures can be used for lower opti­

cal purity polymer (i.e., greater d­isomer content), which

can help reduce the thermal and hydrolytic degradation.

Similar to the injection molding process, fiber­grade PLA

needs to be dried to less than 0.005% (w/w) moisture before

melting in the extruder to minimize molecular weight

drop.

In a two­stage melt spinning process, the polymer is first

heated above its melting point and extruded through the

spinneret. The solidification of the extrudate is achieved

by cooling in the air and the take up roller. In the second

stage, the fiber undergoes hot drawing, where the filament

is pulled down by a take­up roll with a specific speed to

achieve fiber orientation, which is important to increase the

tenacity and stiffness of the fibers. PLA can be melt spun in a

high­speed spinning process with take­up velocity of up to

5000 m/min and a draw ratio of up to six [130]. The degree

of crystallinity of the fiber increases with spinning veloc­

ity due to stress­induced crystallization. By adjusting the

draw ratio, a wide range of mechanical properties can be

achieved for the fiber (Fig. 25). While high speed spinning

provides high fiber manufacturing throughput, manufac­

turing of fiber of uniform diameter is not a trivial task under

high speed condition due to process instability caused by

the draw resonance known as spinline neck­like deforma­

tion, which is characterized by an abrupt attenuation of the

spin­line cross­sectional area. The phenomenon is related

to the spinline flow­induced crystallization [131,139].

L.­T. Lim et al. / Progress in Polymer Science 33 (2008) 820–852 841

Fig. 25. Stress–strain curves for high­speed spun PLA fiber (l­PLA with 8% d­isomer), subjected to various (a) drawing speed and (b) draw ratios. The

extruder used in both processes was 18 mm screw at 185 ◦C. The diameter of spinneret hole was 0.3 mm with a length of 0.6 mm. For the process in (b), the

filaments were taken up at 200 m/min by the first godet and drawn by the second godet spun at 800 and 1200 m/min. The first and second godets were set

at 65 and 110 ◦C, respectively. Curves are reproduced based on the data published by Schmack et al. [130] by permission of John Wiley & Sons, Inc.

Another approach to form PLA fiber is based on the dry

spinning process, which involves dissolving the polymer

in a solvent (typically chloroform, toluene or a mixture of

the two solvents) and extruding the polymer solution in air

or inert gas. Evaporation of the solvent causes the extruded

filaments to solidify [136,140–143]. Although the melt spin­

ning process is relatively straightforward, the process tends

to induce molecular weight drop of the PLA polymer due to

thermal­induced hydrolytic degradation during the melt­

ing step in the presence of residual water. In contrast,

the dry spinning technique is quite effective in preserv­

ing the molecular weight of the polymer due to the lower

processing temperature used. In general, solution­spun

fibers are superior to melt­spun fibers from the stand­

point of mechanical properties. This is attributed to the

lower chain entanglement of polymer molecules in the

solution state as compared to the melt state. By transfer­

ring this dilute entanglement network to the solid state

in the spinning process, the as­spun fibers tend to exhibit

high draw ratios. For instance, Penning et al. reported that

dry spinning followed by hot drawing resulted in low crys­

tallinity fibers having a tensile strength of 1 GPa, whereas

fibers prepared from melt spinning followed by hot draw­

ing have much considerably lower strengths, ranging from

0.19 for completely amorphous copolymer to 0.53 GPa for

PLA homopolymer [141]. The major drawback to solution

spinning is the use of organic solvents, which can pose

environmental problems.

One important parameter that affects the fiber mor­

phology during dry solution­spinning is the composition

of the solvent. Instead of using one solvent, several studies

have successfully manipulated the morphology and physi­

cal properties of PLA fibers by using multi or binary solvent

systems. Postema et al. showed that PLA fibers spun from

a solvent made up of 40/60 chloroform/toluene exhibited

the highest tensile strength of 2.3 GPa after hot draw­

ing among other chloroform/toluene proportions tested

[142,143]. Similar observations were reported by Leenslag

and Pennings [140]. They postulated that under the 40/60

chloroform/toluene condition, the PLA adopted an inter­

rupted helical conformation. Solidification of PLA from

the solution during solvent evaporation caused the heli­

cal aggregates to form crystalline junctions that hampered

re­entangling of polymer chains, leading to crystalline

polymer with good draw­ability [140]. In this binary sys­

tem which contained a good PLA solvent (chloroform) and

a poor PLA solvent (toluene), the greater tendency for chlo­

roform to evaporate during spinning causes an increased

PLA concentration with a concomitant decrease in sol­

vent power, resulting in the formation of polymer­poor

and polymer­rich phases. The polymer­rich phase under­

went rapid solidification, thereby generating porous fiber

structures [142,143].

The wet process is similar to the dry process except

that the polymer solution is spun into a bath containing

coagulating solution which causes the polymer filament

to solidify. With this approach, PLA is normally dissolved

in chloroform which is then extruded into a toluene or

methanol bath [136]. Tables 8 and 9 compare the pro­

cess parameters between the melt and solution spinning.

Compared to polyester, Lyocell, cotton, and viscose fibers,

PLA fiber tends to be more extensible. When PLA stretches

beyond several percent of extension, it yields, stretches

quite easily, and then ruptures at relatively high elonga­

tion values. As shown in Fig. 26, the tenacity behavior of

PLA is akin to wool.

One variant of dry spinning is known as electrospinning

which utilizes an electrostatic force to draw fiber. Recently,

several research groups have successfully spun PLA fibers of

submicron diameter using this approach. The electrospin­

ning process is discussed in the next section.

7.9. Electrospinning of ultrafine fibers

Electrospinning is a technique for producing fibers

that are much smaller in diameter than those produced

using the conventional techniques. Electrospun fibers typi­

cally have diameters range from micrometer to nanometer

(Fig. 27). Similar to the conventional dry spinning process,

electrospinning requires the solubilization of the polymer

in a solvent. However, unlike drying spinning which relies

on mechanical extrusion, the electrospinning process uses

842 L.­T. Lim et al. / Progress in Polymer Science 33 (2008) 820–852

Ta

ble

8

Pro

pert

ies

an

dco

nd

itio

ns

for

melt

­sp

un

PLA

fib

ers

Refe

ren

ces

d­C

on

ten

t

(%)

Init

ial

MW

(×10

3)

Fin

al

MW

(×10

3)

Co

llect

ion

speed

(m/m

in)

Sp

inn

ere

td

/l

(mm

)

Die

tem

pera

ture

(◦C

)

Dra

wra

tio

Dra

wte

mp

era

ture

(◦C

)

Cry

stall

init

y

(%)

Ten

sile

stre

nth

(MP

a)

Mo

du

lus

(GP

a)

Fib

er

dia

mete

r

(mm

)

Pen

nin

g[1

41

]0

–15

28

0–

60

0–

10

.25

/–15

0–

210

7–

96

0–

10

0–

18

5–

53

00

.5–

0.9

3–

Cic

ero

[12

9]

411

1–

131

67

–7

9–

––

4–

8–

<3

56

0–

40

00

.5–

3.1

Fam

bri

[12

8]

03

30

110

1.8

–10

1.0

/–2

40

7–

21

16

03

0–

38

87

09

.24

8–

10

6

Eli

ng

[13

6]

018

0–

26

0–

0.2

5–

0.3

51

/10

18

58

–2

511

0–

48

0–

50

06

–7

–

Sch

mak

et

al.

[13

0]

42

07

18

08

00

–5

00

00

.3/0

.618

54

–6

65

–11

00

–2

419

8–

45

03

.1–

6.3

<5

00

Cic

ero

[13

2]

29

9–

10

96

2–

71

–2

.16

–1

–8

–4

3–

50

10

0–

35

01

–3

30

0

Mezg

han

i[1

35

]0

.7212

16

310

00

0.7

6/4

23

3–

–<

43

90

–3

80

3.6

–6

.012

–7

3

Yu

an

[13

4]

02

63

–4

95

10

5–

217

3.2

1.0

/–210

–2

40

4.7

–5

.912

017

–2

34

2–

10

31.2

–5

.411

0–

36

0

Ta

ble

9

Pro

pert

ies

of

solu

tio

n­s

pu

nP

LA

fib

ers

Refe

ren

ces

d­C

on

ten

t

(%)

Init

ial

MW

(×10

3)

So

lven

tC

oll

ect

ion

speed

(m/m

in)

Sp

inn

ere

td

/l

(mm

)

So

luti

on

tem

pera

ture

(◦C

)

Dra

wra

tio

Dra

w

tem

pera

ture

(◦C

)

Cry

stall

init

y

(%)

Ten

sile

stre

ng

th

(MP

a)

Mo

du

lus

(GP

a)

Fib

er

dia

mete

r

(mm

)

Po

stem

a[1

42

]0

910

4%

,w/v

,ch

loro

form

/to

luen

e(4

0/6

0)

10

–18

20

.25

/23

60

10

–18

19

0–

90

0–

14

00

––

Po

stem

a[1

43

]0

910

4%

,w/v

,ch

loro

form

/to

luen

e(4

0/6

0)

30

.25

/23

9–

60

12

–14

19

0–

110

0–

22

00

–17

–2

8

Pen

nin

g[1

41

]5

–5

–6

%,w

/v,c

hlo

rofo

rm/t

olu

en

e3

0.2

5/–

60

–11

0–

95

09

.2–

Eli

ng

[13

6]

03

00

–5

00

6–

12

%,w

/w,t

olu

en

e0

.02

5–

0.0

35

1/–

110

4–

26

18

0–

20

1–

28

0–

10

00

7–

10

–

Leen

slag

[14

0]

09

00

4%

,w/v

,40

/60

,v/v

,ch

loro

form

/to

luen

e0

.012

–0

.017

0.2

5/2

36

02

02

04

53

210

09

–16

6.6

L.­T. Lim et al. / Progress in Polymer Science 33 (2008) 820–852 843

Fig. 26. Tenacity–stretch ratio curves of PLA as compared to other com­

mon textile fibers. The curves are based on the data originally published

by Farrington et al. [137] and Rajkhowa et al. [187].

electrostatic force to spin the solution into fibers. Due to

their small diameter, electrospun fibers possess very large

area, making them an ideal material for applications such as

medical tissue scaffold, wound dressing, carrier for drugs,

protective fabrics, high performance filter media, filler for

nanocomposite materials, etc. [144].

A typical laboratory electrospinner is made up of four

main components: (1) a high voltage DC supply; (2) a spin­

neret, charged by a DC power supply; (3) an infusion or

peristaltic pump to deliver polymer solution to the spin­

neret; and (4) a metal fiber collector which also acts as a

counter electrode (Fig. 28). To increase throughput, mul­

Fig. 27. Comparison of surface area and diameter for various fibers.

Adapted from Ko [188].

tiple spinnerets have been used in conjunction with a

conveyor belt to achieve a continuous process [145–147].

Most of the setups reported in the literature involve apply­

ing a positive electrode to the spinneret and grounding the

counter electrode, although it is also possible to spin fiber

by reversing the polarity. The basis of electrospinning is

to charge the polymer solution in the spinneret tip with

a high voltage such that the induced charges cause the

polymer solution to eject and travel towards the ground

(or oppositely charged) collector. When the polymer solu­

tion is charged, the induced electrostatic repulsion works

against the surface tension of the solution, causing the poly­

mer solution to elongate and form a characteristic feature

known as a Taylor cone (Fig. 28). When the voltage reaches

a critical level (typically in the order of 10–20 kV), the elec­

trostatic repulsion overcomes the surface tension of the

solution, causing the polymer to eject towards the collector.

As the polymer jet takes flight in the air, the solvent vapor­

izes rapidly, producing a continuous fiber which deposits

on the collector [144,148,149]. By allowing the fiber to spin

for some time, a nonwoven fibrous mat is formed on the

collector. The morphology of the fiber can be affected by

using collector of different configurations. For instance, by

translating the collector plate in the X–Y plane, a large area

of fibrous mat consists of randomly laid fiber can be formed.

To achieve a fibrous membrane with oriented fibers, a rotat­

ing drum, disc, or parallel electrodes may be used (Fig. 28)

[150].

PLA has been successfully electrospun into fibers,

primarily for tissue engineering and biomedical appli­

cations. For instance, a number of studies showed that

scaffolds for the regeneration of cardiac, neural, bone

and blood vessel tissues can be fabricated from electro­

spun PLA fiber through post­spinning orientation and/or

using rotating target collectors [151–154]. PLA has been

electrospun into different forms of ultrafine fiber and

used as carriers for bioactive agents, including antibiotics

[155], anticancer drugs [156,157], and antibacterial silver

nanoparticles [158]. Other composite PLA fibers contain­

ing nano­components such as nanoclay (montmorillonite,

MMT) and TiO2 nanoparticles have also been successfully

produced using the electrospinning technique [157,159].

By and large, PLA solutions with lower polymer con­

centrations favor the formation of small diameter fibers.

But these fibers are less consistent in morphology and

tend coexhibit beads along their length (Fig. 29A and B).

These defects can be overcome by incorporating an organic

or inorganic salt, such as pyridinium formiate, Kh2PO4,

NaH2PO4, or NaCl, in the fiber­forming solution to enhance

its electrical conductivity [160,161] (Fig. 29C and D). The

solvent used will also affect the surface morphology of the

fiber. Using dichloromethane as a solvent and a PLLA poly­

mer solution of 5%, Bognitzki et al. observed that the surface

of the PLLA fibers exhibited regular pore structures, which

were attributed to rapid phase separation of the solvent

[162]. Typical solvents used for PLA, along with the pro­

cess conditions used for electrospinning, are summarized

in Table 10.

Recently, an innovative technique was reported for elec­

trospinning PLA in the melt phase using a CO2­laser [163].

This technique relies on using a laser beam to melt the PLA

844 L.­T. Lim et al. / Progress in Polymer Science 33 (2008) 820–852

Fig. 28. Typical setup for electrospinning, showing the main components. Selected collector configurations are shown here: stationary plate, rotating

drum/disc, and parallel electrodes.

polymer locally, thus minimizing thermal degradation due

to prolonged heat exposure, which has been observed pre­

viously for melt­electrospinning for other polymers based

on the heat conduction melting approach. Moreover, the

laser technique also eliminates the use of solvent, making

the process more environmentally benign. This technique

is capable of producing fibers with diameters smaller than

1 mm [163].

While research on electrospinning has exploded in the

past decade for fibers spun from various polymers, the

commercial production of ultrafine PLA fibers using this

technique has not been forthcoming due to the low pro­

duction throughput, the requirement for the use of specific

solvents, and variation in fiber diameter. Based on the

present state­the­art of electrospinning, this technique is

likely to find uses for products containing PLA­nanofibers

in pharmaceutical and biomedical applications.

7.10. PLA blends with other polymers

In many film applications, such as grocery and garbage

bags, bursting strength, elongation and tear strength are

important properties. As discussed in Sections 7.4 and 7.3,

these properties can be improved to a certain extent by

mechanical drawing, such as biaxial orientation and stretch

blow molding. However, for other PLA parts where the

use of mechanical orientation is not feasible (e.g., injection

molded articles), blending of PLA with other polymers is a

useful strategy to impart flexibility and toughness. Another

motive for blending PLA with other polymers is to reduce

the material cost since the cost of PLA is relatively higher

compared to other petroleum plastics.

Various polymers have been used for improving the

properties of PLA, including elastomers [164], thermoplas­

tic starch [165], poly(ethylene glycol) (PEG) [104,165,166],

triacetin and tributyl citrate [96,104,167], and PHA [99,168].

Lee and McCarthy modified the toughness of 4% d­PLA

with poly(3­hydroxyloctanoate) (PHO). In order to over­

come the process difficulties due to the large difference

in melt viscosity between the two polymers, they modi­

fied the PHO with hexamethylene diisocyanate (HMDI) by

reacting the hydroxyl group of PHO to form urethane link­

ages. HMDI of 2–5.5% (w/w) was reacted with PHO using

a counter­rotating screw at 40 rpm for 2 min at 100 ◦C.

The modified PHO was melt­blended with PLA at 40 rpm

for 3 min at 175 ◦C and then compression molded [168].

Noda et al. prepared plastic films by melt compounding

various proportions of PLA, PHA copolymer (copolymer of

3­hydroxybutyrate with 21 mol% of 3­hydroxyhexanoate),

and poly(ethylene oxide). These authors claimed that the

inherent tackiness of PHA polymer and brittleness of PLA

can be overcome by this approach [99]. Recently, based

on viscosity and gel permeation chromatography mea­

surements, Conrad et al. reported that PLA/PHA blends

degraded more rapidly as compare to the neat PLA [169]. In

another patent, Randall et al. blended PLA with epoxidized

natural rubber in the presence of a compatibilizing agent

(maleic anhydride/polybutadiene copolymer or maleic

anhydride/polybutadiene/PS copolymer at 1–2 wt% relative

of the epoxidized rubber). They observed an increased ulti­

mate elongation from 72.6 to 295%, and an increase of Izod

impact value from 2.08 to 3.8 ft lbs/in. for injection molded

test bars, as compared to the non­compatibilized control

[164]. The blends were produced by extruding the poly­

mers through a twin­screw extruder (150–175 ◦C extruder

zone temperatures; 250 rpm), cooling in a water trough,

chopping the extruded strand into pellets, drying to remove

water, and then injection molding into test bars. Randell et

L.­T. Lim et al. / Progress in Polymer Science 33 (2008) 820–852 845

Fig. 29. Scanning electron micrographs of electrospun PLA fibers prepared from dichloromethane with and without pyridinium formiate (PF), a salt that

increases the conductivity of the polymer solution: (A) 5% (w/w) PLA without additive; (B) 1% (w/w) PLA without additive; (C) 5% (w/w) PLA with 0.8%

(w/w) PF; (D) 0.1% (w/w) PLA with 0.8% (w/w) PF. Micrographs are adapted from [160] with permission from e­Polymers Foundation.

al. noted that reprocessing the polymer blends, i.e., pass­

ing them through the extruder more than once, tended

to increase the impact resistance of the resulting molded

articles [164].

Averous attempted to blend thermoplastic starch, PEG,

glycerol and oligomeric lactic acid by using a single­screw

extruder equipped with a conical­shaped shear element

[165]. Glycerol was found to be the least efficient plas­

ticizer, while oligomeric lactic acid and PEG provided

substantial increases in elongation. Affinity of PLA and

thermoplastic starch was poor, leading to blends that

possessed much weaker mechanical properties as com­

pared to the individual polymers [165]. Moura reported

that the tensile strength, elongation, and damping for

PLA blended with starch particles increased with aver­

age particle size of the starch granules, but declined

when the granules were greater than 45 mm [170]. More­

over, crystallinity increased as the particle size decreased.

The use of methylenediphenyl diisocyanate as a coupling

agent dramatically improved the mechanical properties

of the composite [170]. Sheth et al. melt­blended PLA

and PEG using a counter­rotating twin­screw extruder at

120–180 ◦C. They reported that PEG can form miscible to

partially miscible blends with PLA, depending on the blend

concentration. Below 50% PEG, the plasticized PLA sam­

ples have high elongation with a concomitant reduction

in modulus values. However, above 50% PEG content, the

blend crystallinity increases, resulting in an increased mod­

ulus and a considerable decrease in the elongation at break

[166]. Although PEG is effective in decreasing the stiffness

of PLA, the use of low molecular weight plasticizer has a

disadvantage in that it has a tendency to migrate in the PLA

matrix. For instance, Hiltunen et al. reported that although

PEG plasticized PLA can be extrusion blown to form film

with a reasonable tensile strength, PEG tended to migrate

out of the film after a few days or weeks [104]. Ljungberg

and Wesslen reported that the Tg of PLA can be effectively

depressed by blending with triacetine and tributyl citrate

up to about 25% (w/w), above which phase separation of

plasticizer tended to occur. They also observed that phase

separation of plasticizer accelerated at elevated tempera­

ture (50 ◦C) due to the increased crystallinity of the PLA

phase [167]. During storage of triacetine and tributyl cit­

rate plasticized PLA film, the migration of the plasticizer

846 L.­T. Lim et al. / Progress in Polymer Science 33 (2008) 820–852

Ta

ble

10

Ty

pic

al

pro

cess

con

dit

ion

san

dso

lven

tu

sed

for

ele

ctro

spin

nin

go

fP

LA

Refe

ren

ces

PLA

gra

de

Po

lym

er

solu

tio

nSp

inn

ere

td

iam

ete

r

(mm

)

Sp

inn

ere

t­ta

rget

dis

tan

ce(c

m)

Vo

ltage

(kV

)So

luti

on

feed

rate

(mL/h

)

Co

llect

or

typ

eFib

er

dia

mete

r

(mm

)

Kim

et

al.

[15

4]

PLA

20

%(w

/v)

inch

loro

form

–15

25

0.1

Ro

tati

ng

man

dre

l1

–2

Vaz

et

al.

[15

3]

PLA

14

%(w

/v)

in15

:3m

ixtu

reo

f

CH

Cl 3

:dim

eth

ylf

orm

am

ide

–2

013

1.5

Ro

tati

ng

man

dre

l0

.8–

3

Zo

ng

et

al.

[151

]5

%d

­lact

ide

10

%(w

/w)

in1,1

,1,3

,3,3

­

hex

afl

uo

ro­2

­pro

pan

ol

–15

30

6Sta

tio

nary

pla

te0

.9–

1

Zo

ng

et

al.

[16

1]

5%d

­lact

ide

20

–3

5%

(w/w

)in

1.5

:1(w

/w)

meth

yle

ne

chlo

rid

e:

dim

eth

ylf

orm

am

ide

ble

nd

0.7

15

20

­30

1.2

Ro

tati

ng

dru

m0

.2–

1

Yan

get

al.

[15

2]

PLLA

1–

5%

(w/w

)in

70

:30

dic

hlo

rom

eth

an

e/N

,N­

dim

eth

ylf

orm

am

ide

ble

nd

0.7

–1.2

10

12

1.0

Sta

tio

nary

alu

min

um

pla

tean

dro

tati

ng

dis

c

0.1

–3

Ken

aw

yet

al.

[15

5]

PLLA

14

%(w

/v)

inch

loro

form

–3

015

18

–21

Sta

inle

ssst

eel

sheet

on

rota

tin

gd

rum

3–

6

Xu

et

al.

[15

6]

PLLA

5.5

–6

%(w

/w)

inch

loro

form

0.4

18

45

–5

0.4

3.0

–4

.2Sta

tio

nary

0.3

–1

Jun

et

al.

[16

0]

PLLA

1–

5%

(w/w

)d

ich

loro

meth

an

e–

14

40

1.3

Sta

tio

nary

0.3

–2

.4

Bo

gn

itzk

iet

al.

[16

2]

PLLA

5%

(w/w

)in

dic

hlo

rom

eth

an

e0

.52

03

5–

45

–Sta

tio

nary

∼1

Ogata

et

al.

[16

3]

1.4

–1.8

%d

­lact

ide

Melt

­ele

ctro

spin

nin

g—

no

solv

en

tu

sed

–1

–5

16

–41

–R

ota

tin

gd

isk

0.7

–2

to the film surfaces and the concomitant increased crys­

tallinity of the bulk material have been reported to cause

difficulty in subsequent heat welding [96].

7.11. Compounding of PLA composites

While PLA has modulus and tensile strength compara­

ble to petroleum­based polymers, PLA is brittle and has

relatively much lower thermal and impact resistance prop­

erties. Moreover, the cost of PLA is higher than that for

conventional thermoplastics. To overcome these limita­

tions, other polymers, minerals and biobased materials

have been incorporated in PLA to produce composites

with enhanced properties [171,172]. In particular, bio­fillers

derived from renewable resources (e.g., natural fibers,

starches, proteins) have attracted a great deal of interest

for the reinforcement of PLA due to their sustainable supply

and environmentally benign production.

The extent of reinforcement in PLA composites is largely

dependent on how well the filler material disperses in

the PLA matrix, and the nature of interfacial interaction

between the filler and the PLA phase. Typically, a conven­

tional single screw extruder does not provide sufficient

dispersive mixing to break up the additive agglomerates.

Typically, a twin­screw compounder is needed to achieve

a better dispersion of filler particles. Ouchi et al. disclosed

a method of forming PLA nanocomposites which provides

a rapid crystallization rate to improve heat resistance,

moldability and mold release properties. The approach

involves dispersing organically modified layered clay min­

eral, along with a low molecular weight compound that

has one or more amide groups. One of the examples given

involves compounding 3 wt% MMT and 1 wt% ethylene­bis­

12­hydroxystearic acid amide with PLA in a twin­screw

compounder at 220 ◦C at 300 rpm [173]. Mathew et al.

investigated the mechanical properties of PLA compos­

ites containing microcrystalline cellulose, wood flour and

wood pulp. A twin­screw compounder was used to disperse

the biomaterials in PLA at up to 25 wt% cellulose load­

ing [27]. The compounding conditions used were: screw

speed 250 rpm; heater temperature 170–200 ◦C; and vac­

uum vented extruder. The injection molding was conducted

at 200 ◦C at an injection rate of 60 mm/s, with a pack pres­

sure of 400 bars. The mold temperature used was 50 ◦C,

and the cooling time was 15 s. The studies showed that the

microcrystalline cellulose, which exists as aggregates did

not separate into nano­whiskers during the compounding

process. While good dispersion of cellulose reinforcements

in the PLA matrix was observed, these authors attributed

the reduced tensile strength and elongation at break to

the poor interfacial adhesion between the phases [27].

Using the same compounding technique, Oksman et al.

compared the compression of molded PLA and PP compos­

ites containing flax fibers at 30 and 40 wt% loading. These

authors reported that the tensile strength for PLA­flax

fiber composites is about 50% better than PP­flax com­

posites, which are used today in many automotive panels

[24].

Huda et al. evaluated the mechanical and thermo­

physical properties of PLA composites containing chopped

glass fiber or recycled newspaper cellulose fiber producing

L.­T. Lim et al. / Progress in Polymer Science 33 (2008) 820–852 847

by a twin­screw extruder and an injection molding system

[26]. The dried cellulose fibers and PLA resin, mixed at 70

and 30% (w/w) ratios were compounded in the twin­screw

extruder with an l/d ratio of 30. The extruder tempera­

ture was set at 183 ◦C and the screw speed was 100 rpm.

The extrudate was palletized and dried prior to injection

molding. The extruder temperature was 183–185 ◦C. A rel­

atively long cooling time of 50 s was used to bring the

injection molded part to 65 ◦C in the mold. The study

showed that both mechanical and thermo­physical prop­

erties of the recycled newspaper composite compared

favorably with glass fiber reinforced PLA composite, sug­

gesting that these cellulose fiber reinforced composites

have a potential to replace glass composite in applications

where very high load bearing capabilities are not needed

[26].

The effects of cooling on the crystallization behavior of

PLLA­cellulose composites were investigated by Mathew et

al. [28]. The 25% (w/w) composites were prepared in a vac­

uum vented twin­screw extruder operated at 250 rpm and

170–200 ◦C. The resulting composites were pelletized and

subsequently injection molded at 200 ◦C into a 50 ◦C mold

at the injection speed of 60 mm/s. The Tg values for the

molded PLLA, microcrystalline cellulose, cellulose fiber and

wood flour composites, were 54.1, 56.6, 57.5 and 58.3 ◦C,

while the crystallinity values were 19, 45, 35 and 45%,

respectively. These authors attributed the delayed polymer

relaxation for the composites to the restriction of polymer

chain mobility due to the increased crystallinity in the pres­

ence of fillers. The presence of cellulose fillers has a similar

effect of inducing crystallinity in PLLA polymer during slow

cooling (2 ◦C/min) from the melt (absence of glass transi­

tion around 54–58 ◦C regions for traces III and V in Fig. 30).

Shibata et al. evaluated the effect of dispersing abaca

fiber (Manila hemp) in several biodegradable polyesters,

including PLA. The biofiber was incorporated up to 20%

(w/w) in a twin­screw compounder at 190 ◦C for 5 min

at a screw speed of 50 rpm. These authors observed an

increased flexural moduli for poly(butylenes succinate)

and polyestercarbonate/PLA blends as the fiber content

increased. In contrast, for the PLA­abaca fiber composite, a

minimal increase in flexural strength was observed which

is attributed to inherent high flexural strength of neat PLA

[25].

7.12. PLA nanocomposites

Inorganic or organic nanoparticles have been incor­

porated to enhance the mechanical, barrier and thermal

properties of PLA. Unlike micro­ and macro­scaled particles

(e.g., talc, glass fiber, carbon particles, etc.), nanopar­

ticles can improve material properties at much lower

added quantities (2–8%, w/w). Over the past few years,

various nanomaterials have been investigated for rein­

forcing PLA, including layered silicates, carbon nanotube,

hydroxyapaite, layered titanate, aluminum hydroxide, etc.

[21–23,119,154,174–179]. Among the nanomaterials inves­

tigated, layered silicate clays have been studied in the

greatest detail by researchers from both academia and

industry. The heightened interest for these nanofillers

can be attributed to their ability to dramatically improve

material properties of the nanocomposite structures as

compared with the pure PLA, including improved mechan­

ical and flexural properties, elevated heat distortion

temperature, enhanced barrier properties and accelerated

biodegradation [21,171,175,180]. This section will focus

mainly on reviewing the literature on the processing

aspects and material properties of MMT nanocomposites.

MMT belongs to a family of clays known as smectite

with crystal structure made up of two fused silica tetra­

hedral sheets sandwiched with an edge­shared octahedral

sheet of either aluminum or magnesium hydroxide. The

thickness of a single layer is about 1 nm, while the lat­

eral dimension of the crystals can range from 30 nm to

several microns or greater. The crystal layers are stacked

regularly to provide Van der Waals gaps, known as gal­

Fig. 30. DSC curves of PLA composite materials obtained from different heating­cooling cycles: (a) neat PLA, (b) microcrystalline cellulose, (c) cellulose

fibers, and (d) wood flour. (I) Heating from 30 to 200 ◦C at 10 ◦C/min; (II) cooling from melt at 20 ◦C/min (fast cooling); (III) heating from 30 to 200 ◦C at

10 ◦C/min, after fast cooling; (IV) cooling from melt at 2 ◦C/min (slow cooling); (V) heating from 30 to 200 ◦C at 10 ◦C/min, after slow cooling. The curves are

based on the original data published by Mathew et al. [28] by permission from Elsevier B.V.

848 L.­T. Lim et al. / Progress in Polymer Science 33 (2008) 820–852

leries [119,181]. The silicate surface of MMT is relatively

more hydrophilic than PLA. Therefore, it must be organi­

cally modified to compatibilize and facilitate its dispersion

in PLA. One useful characteristic of MMT is that the cations

in the galleries, typically Na+, Li+, Ca2+, Fe2+ and Mg2+ can

be substituted readily through ion exchange with organic

cations, by treating the clay with surfactants including

primary, secondary, tertiary or quaternary alkylammo­

nium or alkylphosphonium cations [119]. Another unique

property of MMT is related to its ability to delaminate

and disperse in a polymer to give individual imperme­

able platelets of about 1 nm thick. By virtue of their very

high aspect ratio, large interfacial area and nano­thickness,

the dispersed platelets lead to many characteristics that

are important for enhancing the performance of poly­

mers.

Pluta evaluated the effect of organoclay and compound­

ing conditions on polylactide/organoclay nanocomposites

(4.1% d­lactide PLLA) [20]. The PLA pellets were melt­

blended at 180 ◦C with Cloisite® 30B (3%, w/w; modified

with methyl­bis(2­hydroxyethyl) tallowalkyl ammonium

cations) in a counter­rotating mixer at 50 rpm screw speed

for 6–30 min. The study showed that compounding of

PLA/organoclay composite resulted in 14–32% molecular

weight drop, when the composite mix was compounded

for 6–30 min, even if a dry nitrogen atmosphere was

used during the melt processing. Similar reduction in

PLA molecular weight was reported by Ray and Okamoto

on their study of various organically modified layered

silicates, which was attributed to the elevated shear

mixing of PLA in the presence of the silicate, and the

presence of hydroxyl groups in the modified salt, both

of which are capable of causing hydrolysis or transes­

terification reactions in PLA when exposed to elevated

processing temperature [182]. In this study, the nanocom­

posites were first prepared by dry­mixing of PLA resin

with organoclay and extruding in twin­screw extruder

(screw speed = 100 rpm, feed rate = 120 g/min) to form

nanocomposite strands. The strand were pelletized and

dried under vacuum at 60 ◦C for 48 h, pressed into

0.7–2 mm sheet with 1.5 MPa at 190 ◦C for 3 min, and then

quenched and annealed at 110 ◦C for 1.5 h before testing

[182].

The properties of MMT nanocomposites are highly

dependent on how well the clay disperses in the polymer

matrix. In general, clay dispersion can be distinguished in

three modes: (1) when the clay particles are not delam­

inated, the resulting materials tend to exhibit similar

properties as conventional microcomposites. The unsep­

arated MMT layers surrounded by the polymer are often

referred to as tactoids; (2) when the polymer chains are

inserted into the galleries of the swollen silicate layers,

the clay is known as intercalated, leading to decreased

polymer chain mobility and resulting in material reinforce­

ment; and (3) when the clay is completely delaminated

and homogeneously dispersed in the continuous polymer

matrix, the layered silicates are termed exfoliated nanocom­

posites, giving rise to the maximal potential for physical

properties enhancement. The exfoliation of MMT is largely

dependent on the chemical compatibility of the clay and

polymer matrix, as well as the process conditions used to

disperse the silicate layers of the clay. By far, exfoliation

of MMT in PLA can be best achieved by using a twin­

screw compounder. Dennis et al. reported that in order

to exfoliate MMT effectively, the clay particles need to

be sheared and fractured to form smaller stacks of tac­

toid platelets (∼100 nm in thickness). Once platelets of

shorter stack heights are formed, further delamination of

the platelets was primarily driven by the diffusion of the

polymer chains into the clay galleries, which is highly

dependent on the chemical compatibility of the poly­

mer and the organoclay surface [183]. Therefore, unless

the compatibility of the clay and polymer is improved

through chemical treatment, increasing the shear inten­

sity alone will only improve the distributive homogeneity

of the particles in the polymer, but not the delamina­

tion of the clay particles. These authors concluded that

the residence time under low and mild shearing condi­

tions is required to allow polymer chains to penetrate the

clay galleries and peel the platelets apart [183]. Unfortu­

nately, under the typical extrusion conditions, increasing

the residence time will also cause unwanted thermal

degradation, leading to molecular weight drop which

diminishes the performance­enhancing effect of nanopar­

ticles [184].

Various forms of modified MMT have been used to

enhance the material properties of PLA. Chang et al.

incorporated hexadecylamine, dodecyltrimethyl ammo­

nium bromide, and quaternary ammonium salt modified

MMT clays in PLA films at 0–10% (w/w) MMT levels, based

on wet casting method, using N,N′­dimethylacetamide as

a solvent [185]. These authors observed that there is an

optimal loading of organoclay (∼3–4%, w/w) for achiev­

ing the greatest improvement in mechanical properties.

At higher MMT contents, the nanoparticles tended to

agglomerate, leading to mechanical weakening. In contrast,

within the 0–10% MMT loading levels, oxygen transmis­

sion rates decreased with increasing clay content which

was attributed to the increased tortuous diffusion path

of oxygen molecules through the impermeable crystalline

platelets [185]. Ogata et al. used a similar film casting

methodology to evaluate the impact of organophilic MMT

(0–10%, w/w) on PLA, except that chloroform was used

as a solvent. These authors reported that only microcom­

posites were resulted and intercalated structures were

not achieved [18]. It is unknown whether the formation

of microcomposites was due to incomplete exfoliation

of MMT in the chloroform solution or caused by the

agglomeration of MMT during the film drying step. Nev­

ertheless, these studies do highlight the importance of

solvent selection which can affect the exfoliation of the

layered silicates.

Thellen et al. studied acetyltriethyl citrate plasticized

MMT­PLA composite films formed using compounding and

extrusion blowing processes [186]. They observed about

50% improvement in oxygen and water barrier properties,

and 20% increase in modulus as compared to the neat

PLA films. Although thermal properties were not affected

by the addition of MMT, the thermal stability improved

marginally.

In summary, MMT exhibits a strong potential for

enhancing the material properties of PLA. The promising

L.­T. Lim et al. / Progress in Polymer Science 33 (2008) 820–852 849

outlook for MMT is mainly lies in its abundant sources,

benign characteristics to the environment, and low cost.

8. Conclusion: prospects of PLA polymers

PLA is a highly versatile biodegradable polymer which

can be tailor­made into different resin grades for process­

ing into a wide spectrum of products. More importantly,

the polymer can be processed using the conventional

production infrastructure with minimal equipment mod­

ification. New technologies for processing PLA, such as

using supercritical processes for foaming and electrospin­

ning for producing nanofibers, will further expand the

use of this polymer. From the environmental viewpoint,

the compostable characteristics of PLA are well suited

for many applications where recycling, reuse and recov­

ery of products are not feasible. Since the raw materials

for PLA is based on agricultural feedstock, the increased

demand for PLA resins will create a positive impact on

the global agricultural economy. Nevertheless, there are a

number of areas which still need to be improved, espe­

cially in applications where PLA is intended to be used

as a substitution for existing thermoplastics. For instance,

in food products where high barrier protection is impor­

tant, replacement of PET by PLA packaging may not be

feasible, since the barrier properties of PLA are not in

par with PET. The brittleness of PLA may also prevent its

use in applications where toughness and impact resis­

tance are critical. The fact that PLA is biodegradable may

in some cases result in unpredicted performance if the

polymer is exposed to uncontrollable abusive temperature

and humidity conditions. Aging studies which take a wider

perspective covering different environmental exposure and

process conditions will be useful. Some of these challenges

are expected to overcome through blending PLA with other

polymers, formation of micro­ and nanocomposites, coat­

ing with high barrier materials, and polymer modification.

Research and development in these areas may open up

new opportunities for PLA for use as high performance

biodegradable materials.

To date, PLA is relatively more expensive than most

of the petroleum based polymers. Nonetheless, increas­

ing oil prices and the implementation of environmental

policies from the government, such as “green taxes” in

countries like Germany or Japan, and mandatory use of

compostable polymers for packaging by some large corpo­

rations, will create a push to expand the use of PLA. As the

uses for PLA continue to increase, the demand for the agri­

cultural feedstock for PLA production (mainly corn today),

will increase as well. To overcome the potential competi­

tive issues of raw materials with human and livestock food

supply chains, innovations involving the use of alternative

starch and sugar sources, including biomass and other low

value byproduct wastes, are expected to take place.

Acknowledgements

The authors are grateful to Susan E. Selke for value com­

ments and suggestions, and Ana Cristina Vega Lugo for her

assistance in preparing the manuscript.

References

[1] EPA. Municipal solid waste in the United States, 2005 facts andfigures. Washington, DC: Environmental Protection Agency; 2005.

[2] Datta R, Henry M. Lactic acid: recent advances in products, pro­cesses and technologies—a review. J Chem Technol Biotechnol2006;81:1119–29.

[3] Auras R, Harte B, Selke S. An overview of polylactides as packagingmaterials. Macromol Biosci 2004;4:835–64.

[4] Garlotta D. A literature review of poly(lactic acid). J Polym Environ2001;9:63–84.

[5] Conn RE, Kolstad JJ, Borzelleca JF, Dixler DS, Filer Jr LJ, LaDu BN, etal. Safety assessment of polylactide (PLA) for use as a food­contactpolymer. Food Chem Toxicol 1995;33:273–83.

[6] Weber CJ, Haugaard V, Festersen R, Bertelsen G. Production andapplications of biobased packaging materials for the food industry.Food Addit Contam 2002;19:172–7.

[7] Auras R, Harte B, Selke S, Hernandez R. Mechanical, physical, andbarrier properties of poly(lactide) films. J Plastic Film Sheeting2003;19:123–35.

[8] Auras R, Harte B, Selke S. Effect of water on the oxygen barrier prop­erties of poly(ethylene terephthalate) and polylactide films. J ApplPolym Sci 2004;92:1790–803.

[9] Auras RA, Singh SP, Singh JJ. Evaluation of oriented poly(lactide)polymers vs. existing PET and oriented PS for fresh food servicecontainers. Packag Technol Sci 2005;18:207–16.

[10] Lehermeier H, Dorgan J, Way JD. Gas permeation properties ofpoly(lactic acid). J Membr Sci 2001;190:243–51.

[11] Auras R, Harte R, Selke S. Sorption of ethyl acetate and d­limonenein poly(lactide) polymers. J Sci Food Agric 2006;86:648–56.

[12] Dorgan JR, Lehermeier H, Mang M. Thermal and rheological prop­erties of commercial­grade poly(lactic acid)s. J Polym Environ2000;8:1–9.

[13] Siebott V. PLA—the future of rigid packaging? Bioplastics Mag2007;2(2):28–9.

[14] Anonymous. Making preforms for PLA bottles. Bioplastics Mag2006;1(2):16–8.

[15] Kimura K, Horikoshi Y. Bio­based polymers. Fujitsu Sci Technol J2005;41:173–80.

[16] Serizawa S, Inoue K, Iji M. Kenaf­fiber­reinforced poly(lactic acid)used for electronic products. J Appl Polym Sci 2006;100:618–24.

[17] Anonymous. Complete mobile phone housing made of PLA rein­forced with Kenaf fibres. Bioplastics Mag 2006;1(1):18–9.

[18] Ogata N, Jimenez G, Kawai H, Ogihara T. Structure and ther­mal/mechanical properties of poly(l­lactide)­clay blend. J Polym SciPart B: Polym Phys 1997;35:389–96.

[19] Paul M­A, Delcourt C, Alexandre M, Degee P, Monteverde F,Dubois P. Polylactide/montmorillonite nanocomposites: studyof the hydrolytic degradation. Polym Degrad Stab 2005;87:535–42.

[20] Pluta M. Melt compounding of polylactide/organoclay: structureand properties of nanocomposites. J Polym Sci 2006;B44:3392–405.

[21] Ray SS, Bousmina M. Biodegradable polymers and their layered sili­cate nanocomposites: in greening the 21st century materials world.Prog Mater Sci 2005;50:962–1079.

[22] Ray SS, Yamada K, Okamoto M, Ogami A, Ueda K. New poly­lactide/layered silicate nanocomposites. 3. High­performancebiodegradable materials. Chem Mater 2003;15:456–1465.

[23] Ray SS, Okamoto M. New polylactide/layered silicate nanocompos­ites, 6a melt rheology and foam processing. Macromol Mater Eng2003;288:936–44.

[24] Oksman K, Skrifvars M, Selin JF. Natural fibres as reinforce­ment in polylactic acid (PLA) composites. Compos Sci Technol2003;63:1317–24.

[25] Shibata M, Ozawa K, Teramoto N, Yosomiya R, Takeishi H. Biocom­posites made from short abaca fiber and biodegradable polyesters.Macromol Mater Eng 2003;288:35–43.

[26] Huda MS, Drzal LT, Mohanty AK, Misra M. Chopped glass andrecycled newspapers as reinforcement fibers in injection moldedpoly(lactic acid) (PLA) composites: a comparative study. ComposSci Technol 2006;66:1813–24.

[27] Mathew AP, Oksman K, Sain M. Mechanical properties of biodegrad­able composites from polylactic acid (PLA) and microcrystallinecellulose (MCC). J Appl Polym Sci 2005;97:2014–25.

[28] Mathew AP, Oksman K, Sain M. The effect of morphology and chem­ical characteristics of cellulose reinforcements on the crystallinityof polylactic acid. J Appl Polym Sci 2006;101:300–10.

[29] Ikada Y, Tsuji H. Biodegradable polyesters for medical and ecologicalapplications. Macromol Rapid Commun 2000;21:117–32.

850 L.­T. Lim et al. / Progress in Polymer Science 33 (2008) 820–852

[30] Urayama H, Moon SI, Kimura Y. Microstructure and thermal prop­erties of polylactides with different l­ and d­unit sequences:importance of the helical nature of the l­sequenced segments.Macromol Mater Eng 2003;288:137–43.

[31] Tsuji H, Okino R, Daimon H, Fujie K. Water vapor permeability ofpoly(lactide)s: effects of molecular characteristics and crystallinity.J Appl Polym Sci 2005;99:2245–52.

[32] Sarasua JR, Arraiza AL, Balerdi P, Maiza I. Crystallinity and mechani­cal properties of optically pure polylactides and their blends. PolymEng Sci 2005;45:745–53.

[33] Bigg DM. Effect of copolymer ratio on the crystallinity and prop­erties of polylactic acid copolymers. Annu Techn Conf—Soc PlasticsEng 1996:2028–39.

[34] Tsuji H, Ikada Y. Crystallization from the melt of PLA with dif­ferent optical purities and their blends. Macromol Chem Phys1996;197:3483–99.

[35] Hutchinson M, Dorgan J, Knauss D, Hait S. Optical properties ofpolylactides. J Polym Environ 2006;14:119–24.

[36] Pyda M, Bopp RC, Wunderlich B. Heat capacity of poly(lactic acid).J Chem Thermodynam 2004;36:731–42.

[37] Dorgan JR, Jansen J, Clayton MP. Melt rheology of variable l­contentpoly(lactic acid). J Rheol 2005;49:607–19.

[38] Drumright RE, Gruber PR, Henton DE. Polylactic acid technology.Adv Mater 2000;12:1841–6.

[39] Tsukegi T, Motoyama T, Shirai Y. Racemization behavior of l,l­lactideduring heating. Polym Degrad Stab 2007;92:552–9.

[40] Fan Y, Nishida H, Shirai Y, Endo T. Control of racemization for feed­stock recycling of PLLA. Green Chem 2003;5:575–9.

[41] Kopinke FD, Remmler M, Mackenzie K, Moder M, Wachsen O. Ther­mal decomposition of biodegradable polyesters. II. Poly(lactic acid).Polym Degrad Stab 1996;53:329–42.

[42] Khabbaz F, Karlsson S, Albertsson A­C. Py­GC/MS an effec­tive technique to characterizing of degradation mechanism ofpoly(l­lactide) in the different environment. J Appl Polym Sci2000;78:2369–78.

[43] Westphal C, Perrot C, Karlsson S. Py­GC/MS as a means to predictdegree of degradation by giving microstructural changes modelledon LDPE and PLA. Polym Degrad Stab 2001;73:281–7.

[44] Henton DE, Gruber P, Lunt J, Randall J. Polylactic acid technology.In: Mohanty AK, Misra M, Drzal LT, editors. Natural fibers, biopoly­mers, and biocomposites. Boca Raton, FL: Taylor & Francis; 2005. p.527–77.

[45] Jamshidi K, Hyon S­H, Okada Y. Thermal characterization of poly­lactides. Polymer 1988;29:2229–34.

[46] Celli A, Scandola M. Thermal properties and physical ageing ofpoly(l­lactic acid). Polymer 1992;33:2699–703.

[47] Cai H, Dave V, Gross RA, McCarthy P. Effects of physical aging,crystallinity, and orientation on the enzymatic degradation ofpoly(lactic acid). J Polym Sci 1996;B34:2701–8.

[48] Witzke DR. Introduction to properties, engineering, and prospectsof polylactide polymers. PhD thesis. East Lansing, MI: MichiganState University; 1997.

[49] Hartmann MH. High molecular weight polylactic acid polymers.In: Kaplan DL, editor. Biopolymers from renewable resources.Berlin/Heidelberg: Springer­Verlag; 1998. p. 367–411.

[50] Kister G, Cassanas G, Vert M, Pauvert B, Terol A. Vibrational analysisof poly(l­lactic acid). J Raman Spectrosc 1995;26:307–11.

[51] Kawashima N, Ogawa S, Obuchi S, Matsuo M, Yagi T. Polylactic acid“LACEA”. In: Doi Y, Steinbuchel A, editors. Biopolymers polyestersIII applications and commercial products. Weinheim: Wiley–VCHVerlag GmbH; 2002. p. 251–74.

[52] Hartmann MH. Advances in the commercialization of poly­(lactic acid). Polym Prepr (Am Chem Soc, Div Polym Chem)1999;40(1):570–1.

[53] Kister G, Cassanas G, Vert M. Effects of morphology, conformationand configuration on the IR and Raman spectra of various poly(lacticacid)s. Polymer 1998;39:267–73.

[54] Kishore K, Vasanthakumari R. Nucleation parameters for poly­mer crystallization from non­isothermal thermal analysis. ColloidPolym Sci 1988;266:999–1002.

[55] Kishore K, Vasanthakumari R, Pennings AJ. Isothermal meltingbehavior of poly(l­lactic acid). J Polym Sci Part B: Polym Phys1984;22:537–42.

[56] Kolstad JJ. Crystallization kinetics of poly(l­lactide­co­meso­lactide). J Appl Polym Sci 1996;62:1079–91.

[57] Vasanthakumari R, Pennings AJ. Crystallization kinetics of poly(l­lactic acid). Polymer 1983;24:175–8.

[58] Kalb B, Pennings AJ. General crystallization behavior of poly(l­lacticacid). Polymer 1980;21:607–12.

[59] Perego G, Cella GD, Bastioli C. Effect of molecular weight andcrystallinity on poly(lactic acid) mechanical properties. Polymer1996;59:37–43.

[60] Yasuniwa M, Tsubakihara S, Sugimoto Y, Nakafuku C. Thermal anal­ysis of the double­melting behavior of poly(l­lactic acid). J PolymSci Part B: Polym Phys 2004;42:25–32.

[61] Li H, Huneault MA. Nucleation and crystallization of PLA. In: ANTEC2007. 2007. p. 2615–8.

[62] Ou X, Cakmak M. X­ray studies of structural development dur­ing sequential and simultaneous biaxial of polylactic acid film. In:ANTEC 2003. 2003. p. 1701–5.

[63] Lee JK, Lee KH, Jin BS. Structural development and biodegradabilityof uniaxially stretched poly(l­lactide). Eur Polym J 2001;37:907–14.

[64] Cooper­White JJ, Mackay ME. Rheological properties ofpoly(lactides). Effect of molecular weight and temperatureon the viscoelasticity of poly(l­lactic acid). J Polym Sci Part B:Polym Phys 1999;37:1803–14.

[65] Fang Q, Hanna MA. Rheological properties of amorphousand semicrystalline polylactic acid polymers. Ind Crops Prod1999;10:47–53.

[66] Dorgan JR, Williams JS, Lewis DN. Melt rheology of poly(lacticacid): entanglement and chain architecture effects. J Rheol1999;43:1141–55.

[67] Lehermeier HJ, Dorgan JR. Poly(lactic acid) properties and prospectof an environmentally benign plastic: melt rheology of linear andbranched blends. In: Fourteenth symposium on thermophysicalproperties. 2000.

[68] Zhang W, Zheng S. Synthesis and characterization of dendritic starpoly(l­lactide)s. Polym Bull 2007;58:767–75.

[69] Lehermeier HJ, Dorgan JR. Melt rheology of poly(lactic acid):consequences of blending chain architectures. Polym Eng Sci2001;41:2172–84.

[70] Palade L­I, Lehermeier h J, Dorgan JR. Melt rheology of high l­content poly(lactic acid). Macromolecules 2001;34:1384–90.

[71] Yamane H, Sasai K, Takano M. Poly(d­lactic acid) as a rheologicalmodifier of poly(l­lactic acid): shear and biaxial extensional flowbehavior. J Rheol 2004;48:599–609.

[72] Taubner V, Shishoo R. Influence of processing parameters on thedegradation of poly(l­lactide) during extrusion. J Appl Polym Sci2001;79:2128–35.

[73] Sodergard A, Stold M. Properties of lactic acid based poly­mers and their correlation with composition. Prog Mater Sci2002;27:1123–63.

[74] McNeill IC, Leiper HA. Degradation studies of some polyesters andpolycarbonates. 1. Polylactide: degradation under isothermal con­ditions, thermal degradation mechanism and photolysis of thepolymer. Polym Degrad Stab 1985;11:309–26.

[75] Bashir A, Al­Uraini A­A, Jamjoom M, Al­Khalid A, Al­Hafez M, Ali S.Acetaldehyde generation in poly(ethylene terephthalate) resins forwater bottles. J Macromol Sci 2002;A39:1407–33.

[76] Villain F, Coudane J, Vert M. Titration of aldehydes present inpoly(ethylene terephthalate). J Appl Polym Sci 1994;52:55–60.

[77] Sugaya N, Nakagawa T, Sajurai K, Morita M, Onodera S. Analysisof aldehydes in water by headspace­GC/MS. J Health Sci 2001;47:21–7.

[78] Kolstad JJ, Witzke DR, Hartmann MH, Hall ES, Nangeroni J. Lacticacid residue containing polymer composition and product havingimproved stability, and method for preparation and use thereof. USPatent 6,353,086B1; 2002.

[79] Gruber PR, Kolstad JJ, Ryan CM, Hall ES, Conn RSE. Melt­stable amor­phous lactide polymer film and process for manufacturing thereof.US Patent 5,484,881; 1996.

[80] Gogolewski S, Jovanovic M, Perren SM. The effect of melt­processing on the degradation of selected polyhydroxyacids:polylactides, polyhydroxybutyrate, and polyhydroxybutyrate­co­valerates. Polym Degrad Stab 1993;40:313–22.

[81] Gruber PR, Jeffrey SP, Kolstad JJ, Conn RSE, Ryan CM. Hydroxyl­terminated lactide polymer composition. US Patent 5,446,123;1995.

[82] Suizu H, Takagi M, Ajioka M, Yamaguchi A. Purification process ofaluphatic polyester. US Patent 5,496,923; 1996.

[83] Natureworks. Personal communications. Minnetonka, MN: Nature­works LLC; 2007.

[84] Natureworks. Sheet extrusion processing guide. Minnetonka, MN:Natureworks LLC; 2005.

[85] Natureworks. Natureworks PLA injection molding guide for 3051D.Minnetonka, MN: Natureworks LLC; 2006.

[86] Natureworks. Natureworks polymer technical data sheets—PLApolymer 7032D. Minnetonka, MN: Natureworks LLC; 2005.

L.­T. Lim et al. / Progress in Polymer Science 33 (2008) 820–852 851

[87] Rees H. Understanding injection molding technology. Cincinnati:Hanser/Gardner Publications, Inc.; 1995.

[88] Natureworks. Moldflow material testing report—MAT2238 Nature­works PLA. Minnetonka, MN: Natureworks LLC; 2004.

[89] Sato Y, Inohara K, Takishima S, Masuoka H, Imaizumi M, YamamotoH, et al. Pressure–volume–temperature behavior of polylactide,poly(butylene succinate), and poly(butylene succinate­co­adipate).Polym Eng Sci 2000;40:2602–9.

[90] Chen X. Mathematical modeling of the in­mold coating process forinjection molded thermoplastic parts. PhD thesis. The Ohio StateUniversity; 2003.

[91] Vilela Pontes AJ. Shrinkage and ejection forces in injection mouldedproducts. PhD thesis. Universidade do Minho; 2002.

[92] Capt L. The pressure–volume–temperature behavior and the effectof pressure on crystallization kinetics of polyethylene resins. MScthesis. McGill University; 1999.

[93] Mukherjee S, Jabarin SA. Aging characteristics of orientedpoly(ethylene terephthalate). Polym Eng Sci 1995;35:1145–54.

[94] Cink K, Bopp RC, Sikkema K. Injection stretch blow molding processusing polylactide resins. WO Patent 2006/002409; 2006.

[95] Natureworks. Natureworks PLA ISBM bottle guide. Minnetonka,MN: Natureworks LLC; 2005.

[96] Ljungberg N, Andersson T, Wesslen B. Film extrusion and film weld­ability of poly(lactic acid) plasticized with triacetine and tributylcitrate. J Appl Polym Sci 2003;88:3239–47.

[97] Natureworks. Natureworks PLA processing guide for biaxially ori­ented film. Minnetonka, MN: Natureworks LLC; 2006.

[98] Rosenbaum S, Rosenbaum M, Rosenbaum M, Hade P, Busch DD.Polylactic acid (PLA) film having good antistatic properties. USPatent 7,144,634; 2006.

[99] Noda I, Bond EB, Melik DH. Polyhydroxyalkanoate copolymer andpolylactic acid polymer compositions for laminates and films. USPatent 6,808,795; 2004.

[100] Auras R. Investigation of polylactide as packaging material. PhDthesis. East Lansing: Michigan State University; 2004.

[101] Sodergard A, Selin J­F, Niemi M, Johansson C­J, Meinander K. Pro­cessable poly(hydroxy acids). US Patent 6,559,244B1; 2003.

[102] Tweed EC, Stephens HM, Riegert TE. Polylactic acid blown filmand method of manufacturing same. US Patent Application2006/0045940A1; 2006.

[103] Natureworks. Production of Natureworks polylactide films onblown film equipment designed for producing low densitypolyethylene film. Minnetonka, MN: Natureworks LLC; 2003.

[104] Hiltunen E, Selin J­F, Skog M. Polyactide films. US Patent 6,117,928;2000.

[105] Selke SEM, Culter JD, Hernandez RJ. Plastics packaging, properties,processing, applications and regulations. 2nd ed. Cincinnati, OH:Hanser; 2004.

[106] Natureworks. Processing guide for thermoforming articles. Nature­works LLC: Minnetonka, MN; 2005.

[107] Bosiers L, Engelmann S. Thermoformed packaging made of PLA.Kunststoffe Plast Europe 2003;12:21–4.

[108] Natureworks. Natureworks polymer processing guides—cryst­allizing and drying of PLA. Natureworks LLC: Minnetonka, MN;2007.

[109] Auras R, Singh JJ, Singh SP. Performance evaluation of PLA existingPET and PS containers. J Test Eval 2006;34:530–6.

[110] Mathieu LM, Mueller TL, Bourban P­E, Pioletti DP, Muller R,Manson J­AE. Architecture and properties of anisotropic poly­mer composite scaffolds for bone tissue engineering. Biomaterials2006;27:905–16.

[111] Maquet V, Martin D, Malgrange B, Franzen R, Schoenen J, Moo­nen G, et al. Peripheral nerve regeneration using bioresorbablemacroporous polylactide scaffolds. J Biomed Mater Res 2000;52:639–51.

[112] Busby W, Cameron NR, Jahoda CAB. Tissue engineering matrixes byemulsion templating. Polym Int 2002;51:871–81.

[113] Di Y, Iannace S, Di Maio E, Nicolais L. Reactively modified poly(lacticacid): properties and foam processing. Macromol Mater Eng2005;290:1083–90.

[114] Mikos AG, Thorsen AJ, Czerwonka LA, Bao Y, Langer R, Winslow DN,et al. Preparation and characterization of poly(l­lactic acid) foams.Polymer 1994;35:1068–77.

[115] Ajioka M, Enomoto K, Yamaguchi A, Suzuki K, Watanabe T, KitaharaY. Degradable foam and use of same. US Patent 5,447,962; 1995.

[116] Sterzel H­J. Production of foamed polylactide injection moldings ofhigh strength and rigidity. US Patent 5,422,053; 1995.

[117] Guan J, Eskridge KM, Hanna MA. Acetylated starch–polylactic acidloose­fill packaging materials. Ind Crops Prod 2005;22:109–23.

[118] Preechawong D, Peesan M, Supaphol P, Rujiravanit R. Preparationand characterization of starch/poly(l­lactic acid) hybrid foams. Car­bohydr Polym 2005;59:329–37.

[119] Ray SS, Okamoto M. Polymer/layered silicate nanocompos­ites: a review from preparation to processing. Prog Mater Sci2003;28:1539–641.

[120] Ema Y, Ikeya M, Okamoto M. Foam processing and cellular structureof polylactide­based nanocomposites. Polymer 2006;47:5350–9.

[121] Di Y, Iannace S, Di Maio E, Nicolais L. Poly(lactic acid)/organoclaynanocomposites: thermal, rheological properties and foam pro­cessing. J Polym Sci Part B: Polym Phys 2005;43:689–98.

[122] Nalawade S, Picchioni F, Janssen LPBM. Supercritical carbon dioxideas a green solvent for processing polymer melts: processing aspectsand applications. Prog Polym Sci 2006;31:19–43.

[123] Quirk RA, France RM, Shakesheff KM, Howdle SM. Supercritical fluidtechnologies and tissue engineering scaffolds. Curr Opin Solid StateMater Sci 2004;8:313–21.

[124] Xu Q, Pang M, Peng Q, Li J, Jiang Y. Application of supercriticalcarbon dioxide in the preparation of biodegradable polylactidemembranes. J Appl Polym Sci 2004;94:2158–63.

[125] Fujiwara T, Yamaoka T, Kimura Y, Wynne KJ. Poly(lactide) swellingand melting behavior in supercritical carbon dioxide and post­venting porous material. Biomacromolecules 2005;6:2370–3.

[126] Mooney DJ, Baldwin DF, Suh NP, Vacanti JP, Langer R. Novel approachto fabricate porous sponges of poly(d,l­lactic­co­glycolic acid) with­out the use of organic solvents. Biomaterials 1996;17:1417–22.

[127] Matuana LM. Solid state microcellular foamed poly(lactic acid):morphology and property characterization. Bioresource Technol2008;99:3643–50.

[128] Fambri L, Pegoretti A, Fenner R, Incardona SD, Migliaresi C.Biodegradable fibres of poly(l­lactic acid) produced by melt spin­ning. Polymer 1997;38:79–85.

[129] Cicero JA, Dorgan JR, Garrett J, Runt J, Lin JS. Effects of moleculararchitecture on two­step, melt­spun poly(lactic acid) fibers. J ApplPolym Sci 2002;86:2839–46.

[130] Schmack G, Tandler B, Vogel R, Beyreuther R, Jacobsen S, Fritz H­G. Biodegradable fibers of poly(l­lactide) produced by high­speedmelt spinning and spin drawing. J Appl Polym Sci 1999;73:2785–97.

[131] Kohler WH, Shrikhande P, McHugh AJ. Modeling melt spinning ofPLA fibers. J Macromol Sci B 2005;44:185–202.

[132] Cicero JA, Dorgan JR. Physical properties and fiber morphology ofpoly(lactic acid) obtained from continuous two­step melt spinning.J Polym Environ 2002;9:1–10.

[133] Cicero JA, Dorgan JR, Janzen J, Garrett J, Runt J, Lin JS. Supramolecularmorphology of two­step, melt­spun poly(lactic acid) fibers. J ApplPolym Sci 2002;86:2828–38.

[134] Yuan X, Mak AFT, Kwok KW, Yung BKO, Yao K. Characterization ofpoly(l­lactic acid) fibers produced by melt spinning. J Appl PolymSci 2001;81:251–60.

[135] Mezghani K, Spruiell JE. High speed melt spinning of poly(l­lacticacid) filaments. J Polym Sci 1998;B36:1005–12.

[136] Eling B, Gogolewski S, Pennings AJ. Biodegradable materials ofpoly(l­lactic acid). 1. Melt­spun and solution­spun fibres. Polymer1982;23:1587–93.

[137] Farrington DW, Davies JL, Blackburn RS. Poly(lactic acid) fibers. In:Blackburn RS, editor. Biodegradable and sustainable fibers. Cam­bridge, England: Woodhead Publishing Limited; 2005.

[138] Agrawal AK, Bhalla R. Advances in the production of poly(lactic acid)fibers—a review. J Macromol Sci C 2003;43:479–503.

[139] Shin DM, Lee JS, Jung HW, Hyun JC. High­speed fiber spinningprocess with spinline flow­induced crystallization and neck­likedeformation. Rheol Acta 2006;45:575–82.

[140] Leenslag JW, Pennings AJ. High­strength poly(l­lactide) fibresby a dry­spinning/hot­drawing process. Polymer 1987;28:1695–702.

[141] Penning JP, Dijkstra H, Pennings AJ. Preparation and proper­ties of absorbable fibres from l­lactide copolymers. Polymer1993;34:942–51.

[142] Postema AR, Luiten AH, Oostra H, Pennings AJ. High­strength poly(l­lactide) fibers by a dry­spinning/hot­drawing process. II. Influenceof the extrusion speed and winding speed on the dry­spinning pro­cess. J Appl Polym Sci 1990;39:1275–88.

[143] Postema AR, Luiten AH, Pennings AJ. High­strength poly(l­lactide)fibers by a dry­spinning/hot­drawing process. I. Influence of theambient temperature on the dry­spinning process. J Appl Polym Sci1990;39:1265–74.

[144] Huang Z­M, Zhang Y­Z, Kotaki M, Ramakrishna S. A review onpolymer nanofibers by electrospinning and their applications innanocomposites. Compos Sci Technol 2003;63:2223–53.

852 L.­T. Lim et al. / Progress in Polymer Science 33 (2008) 820–852

[145] Chu B, Hsiao BS, Fang D. Apparatus and methods for electrospinningpolymeric fibers and membranes. US Patent 6,713,011; 2004.

[146] Gogins MA, Weik TM. Air filter assembly for filtering an air streamto remove particulate matter entrained in the stream. US Patent6,716,274; 2004.

[147] Lee WS, Jo SM, Go SG, Chun SW. Apparatus of polymer web byelectrospinning process. US Patent 6,616,435; 2003.

[148] Li D, Xia Y. Electrospinning of nanofibers: reinventing the wheel?Adv Mater 2004;16:1151–70.

[149] Ramakrishna S, Fujihara K, Teo W­E, Lim T­C, Ma Z. An introductionto electrospinning and nanofibers. World Scientific; 2005.

[150] Teo WE, Ramakrishna S. A review on electrospinning design andnanofiber assemblies. Nanotechnology 2006;17:R89–106.

[151] Zong X, Bien H, Chung C­Y, Yin L, Fang D, Hsiao BS, et al. Electrospunfine­textured scaffolds for heart tissue constructs. Biomaterials2005;26:5330–8.

[152] Yang F, Murugan R, Wang S, Ramakrishna S. Electrospinning ofnano/micro scale poly(l­lactic acid) aligned fibers and their poten­tial in neural tissue engineering. Biomaterials 2005;26:2603–10.

[153] Vaz CM, van Tuijl S, Bouten CVC, Baaijens FPT. Design of scaffoldsfor blood vessel tissue engineering using a multi­layering electro­spinning technique. Acta Biomater 2005;1:575–82.

[154] Kim H­W, Lee H­H, Knowles JC. Electrospinning biomedicalnanocomposite fibers of hydroxyapaite/poly(lactic acid) for boneregeneration. J Biomed Mater Res 2006;79A:643–9.

[155] Kenawy E­R, Bowlin GL, Mansfield K, Layman J, Simpson DG, SandersEH, et al. Release of tetracycline hydrochloride from electrospunpoly(ethylene­co­vinylacetate), poly(lactic acid), and a blend. J Con­trol Release 2002;81:57–64.

[156] Xu X, Yang L, Xu X, Wang X, Chen X, Liang Q, et al. Ultrafine med­icated fibers electrospun from W/O emulsions. J Control Release2005;108:33–42.

[157] Song M, Pan C, Li J, Wang X, Gu Z. Electrochemical study onsynergistic effect of the blending of nano TiO2 and PLA polymeron the interaction of antitumor drug with DNA. Electroanalysis2006;18:1995–2000.

[158] Xu X, Yang Q, Wang Y, Yu H, Chen X, Jing X. Biodegradableelectrospun poly(l­lactide) fibers containing antibacterial silvernanoparticles. Eur Polym J 2006;42:2081–7.

[159] Lee YH, Lee JH, An I­G, Kim C, Lee DS, Lee YK, et al. Electrospundual­porosity structure and biodegradation morphology of Mont­morillonite reinforced PLLA nanocomposite scaffolds. Biomaterials2005;26:3165–72.

[160] Jun Z, Hou H, Schaper A, Wendorff JH, Greiner A. Poly­l­lactidenanofibers by electrospinning—influence of solution viscosity andelectrical conductivity on fiber diameter and fiber morphology. e­Polymers 2003 [No. 009].

[161] Zong X, Kim K, Fang D, Ran S, Hsiao BS, Chu B. Structure and processrelationship of electrospun bioabsorbable nanofiber membranes.Polymer 2002;43:4403–12.

[162] Bognitzki M, Czado W, Frese T, Schaper A, Hellwig M, SteinhartM, et al. Nanostructured fibers via electrospinning. Adv Mater2001;13:70–2.

[163] Ogata N, Yamaguchi S, Shimada N, Lu G, Iwata T, Nakane K, et al.Poly(lactide) nanofibers produced by a melt­electrospinning sys­tem with a laser melting device. J Appl Polym Sci 2007;104:1640–5.

[164] Randall JR, Ryan CM, Lunt J, Hartman MH, Hall ES, Kolstad JJ, etal. Impact modified melt­stable lactide polymer compositions andprocesses for manufacture thereof. US Patent 6,495,631; 2002.

[165] Averous OM. Poly(lactic acid): plasticization and properties ofbiodegradable multiphase systems. Polymer 2001;42:6209–19.

[166] Sheth M, Kumar RA, Dave V, Gross RA, McCarthy SP. Biodegradablepolymer blends of poly(lactic acid) and poly(ethylene glycol). J ApplPolym Sci 1997;66:1495–505.

[167] Ljungberg N, Wesslen B. The effects of plasticizers on the dynamicmechanical and thermal properties of poly(lactic acid). J Appl PolymSci 2002;86:1227–34.

[168] Lee J, McCarthy S. Improving the toughness of poly(lactic acid)through co­continuous, immiscible, biodegradable blends withPHA. In: ANTEC 2007. 2007. p. 1569–72.

[169] Conrad JD, Eickhoff JP, Harrison GM. Rheological and degradationproperties of renewable resource polymer blends. In: ANTEC 2007.2007. p. 2624–7.

[170] Moura RA. The effect of physical aging, starch particle size, andstarch oxidation on thermal–mechanical properties of poly(lacticacid)/starch composites. PhD thesis. Manhattan, KS: Kansas StateUniversity; 2006.

[171] Yu L, Dean K, Li L. Polymer blends and composites from renewableresources. Prog Polym Sci 2006;31:576–602.

[172] Mohanty AK, Misra M, Hinrichsen G. Biofibers, biodegradablepolymers and biocomposites—an overview. Macromol Mater Eng2000;276/277:1–24.

[173] Ouchi M, Okamoto H, Nakano M, Takeuchi H, Usuki A, Kanamori T,et al. Polylactic acid composite material and moded body. US Patent7,084,192; 2006.

[174] Chen G­X, Kim H­S, Park BH, Yoon J­S. Controlled functionalizationof multiwalled carbon nanotubes with various molecular­weightpoly(l­lactic acid). J Phys Chem B 2005;109:22237–43.

[175] Singh S, Ray SS. Polylactide based nanostructured biomaterials andtheir applications. J Nanosci Nanotechnol 2007;7:2596–615.

[176] Kuan C­F, Chen C­H, Kuan H­C, Lin K­C. Mechanical, electrical andthermal properties of MWCNT/poly(lactic acid) composites. In:ANTEC 2007. 2007. p. 2250–4.

[177] Nishida H, Fan Y, Mori T, Oyagi N, Shirai Y, Endo T. Feedstockrecycling of flame­resisting poly(lactic acid)/aluminum hydroxidecomposite to l,l­lactide. Ind Eng Chem Res 2005;44:1433–7.

[178] Hiroi R, Ray SS, Okamoto M, Shiroi T. Organically modifiedlayered titanate—a new nanofiller to improve the perfor­mance of biodegradable polylactide. Macromol Rapid Commun2004;25:1359–64.

[179] Ray SS, Yamada K, Okamoto M, Ueda K. Polylactide­layered sil­icate nanocomposite—a novel biodegradable material. Nano Lett2002;2:1093–6.

[180] Pandey JK, Reddy KR, Kumar AP, Singh RP. An overview on thedegradability of polymer nanocomposites. Polym Degrad Stab2005;88:234–50.

[181] Zeng QH, Yu AB, Lu GQ, Paul DR. Clay­based polymernanocomposites—research and commercial development. JNanosci Nanotechnol 2005;5:1574–92.

[182] Ray SS, Okamoto M. Biodegradable polylactide and its nanocom­posites: opening a new dimension for plastics and composites.Macromol Rapid Commun 2003;24:815–40.

[183] Dennis HR, Hunter DL, Chang D, Kim S, White JL, Cho JW, etal. Effect of melt processing conditions on the extent of exfo­liation in organoclay­based nanocomposites. Polymer 2001;42:9513–22.

[184] Altpeter H, Bevis MJ, Grijpma DW, Feijen J. Non­conventional injec­tion molding of poly(lactide) and poly(«­caprolactone) intendedfor orthodedic applications. J Mater Sci: Mater M 2004;15:175–84.

[185] Chang J­H, An YU, Sur GS. Poly(lactic acid) nanocomposites withvarious organoclays. I. Thermomechanical properties, morphology,and gas permeability. J Polym Sci Part B: Polym Phys 2003;41:94–103.

[186] Thellen C, Orroth C, Froio D, Ziegler D, Lucciarini J, Farrell R, et al.Influence of montmorillonite layered silicate on plasticized poly(l­lactide) blown films. Polymer 2005;46:11716–27.

[187] Rajkhowa R, Gupta VB, Kothari VK. Tensile stress–strain and recov­ery behavior of Indian silk fibers and their structural dependence. JAppl Polym Sci 2000;77:2418–29.

[188] Ko FK. Nanofiber technology—bridging the gap between nano andmacro World. Antalya, Turkey: NATO—Advance Study Institute(ASI), Nanoengineered Nanofibrous Materials; September 1–12,2003.

[189] Tsuji H, Ikada Y. Blends of isotactic and atactic poly(lactide)s.2. Molecular­weight effects of atactic component on crystalliza­tion and morphology of equimolar blends from the melt. Polymer1996;37:595–602.

[190] Doi Y, Fukuda K. Biodegradable plastics and polymers. In: Doi Y,Fukuda K, editors. Proceedings of the third international scientificworkshop on biodegradable plastics and polymers. Amsterdam:Elsevier Science; 1993. p. 464–9.

[191] Van Dijk JAPP, Smit JAM, Kohn FE, Feijen J. Characterization ofpoly(d,l­lactic acid) by gel permeation chromatography. J Polym Sci:Polym Chem 1983;21:197–208.

[192] Spinu M, Jackson C, Keating MY, Gardner KH. Material design inpoly(lactic acid) systems: block copolymers, star homo­ and copoly­mers, and stereocomplexes. J Macromol Sci A 1996;33:1497–530.

[193] Gupta MC, Deshmukh VG. Thermal oxidative degradation of poly­lactic acid. Part II. molecular weight and electronic spectra duringisothermal heating. Colloid Polym Sci 1982;260:514–7.

[194] Xu K, Kozluca A, Denkbas EB, Piskin E. Synthesis and characteriza­tion of PDLLA homopolymers with different molecular weights. JAppl Polym Sci 1996;59:561–3.