-
EUROPEAN
European Polymer Journal 43 (2007) 2819–2835
www.elsevier.com/locate/europolj
POLYMERJOURNAL
LO
GY
Macromolecular Nanotechnology
Polylactide/montmorillonite nanocomposites:Structure,
dielectric, viscoelastic and thermal properties
M. Pluta a,*, J.K. Jeszka a, G. Boiteux b
a Department of Polymer Physics, Centre of Molecular and
Macromolecular Studies, Polish Academy of Sciences,
Sienkiewicza 112, 90-363 Lodz, Polandb Universite Lyon 1,
Laboratoire des Materiaux Polymeres et des Biomateriaux, CNRS URA
507, F-69622 Villeurbanne, France
Received 21 December 2006; received in revised form 21 March
2007; accepted 9 April 2007Available online 27 April 2007
MA
CR
OM
OL
EC
UL
AR
NA
NO
TE
CH
NO
Abstract
Polylactide-based systems composed of an organoclay (Cloisite�
30B) and/or a compatibilizer (Exxelor VA1803) pre-pared by melt
blending were investigated. Two types of not compatibilized
nanocomposites containing 3 wt% or 10 wt% ofthe organoclay were
studied to reveal the effect of the filler concentration on the
nanostructure and physical properties ofsuch systems. The 3
wt%-nanocomposite was also additionally compatibilized in order to
improve the nanoclay dispersion.Neat polylactide and polylactide
with the compatibilizer processed in similar conditions were used
as reference samples.The X-ray investigations showed the presence
of exfoliated nanostructure in 3 wt%-nanocomposite.
Compatibilizationof such system noticeably enhanced the degree of
exfoliation of the organoclay. Viscoelastic spectra (DMTA) showedan
increase of the storage and loss moduli with the increase of the
organoclay content and dispersion. Dielectric propertiesof the
nanocomposites show a weak influence of the nanoclay on segmental
(aS) and local (b)-relaxations in PLA, exceptfor the highest
nanoclay content. Above Tg a strong increase of dc conductivity
related to ionic species in the clay isobserved. It gives rise also
to the Maxwell–Wagner–Sillars interfacial polarization and both
real and imaginary parts ofe strongly increase. In the temperature
dependence of low frequency dielectric constant and mechanical
moduli(at 1 Hz) an additional maximum around 80–90 �C is observed
due to cold crystallization of PLA.� 2007 Elsevier Ltd. All rights
reserved.
Keywords: Polylactide; Nanocomposites; Compatibilization;
Thermal properties; Viscoelastic properties; Dielectric
properties
1. Introduction
Polylactide (PLA) is one of the most importantbiodegradable and
biocompatible polymers in a
0014-3057/$ - see front matter � 2007 Elsevier Ltd. All rights
reserveddoi:10.1016/j.eurpolymj.2007.04.009
* Corresponding author. Tel.: +48 (42) 6803237; fax: +48
(42)6847126.
E-mail address: [email protected] (M. Pluta).URL:
http://www.cbmm.lodz.pl (M. Pluta).
group of degradable plastics. It can be derived fromrenewable
resources, it is friendly for the environ-ment and exhibits
interesting physical properties,which can be further modulated by
filling withselected additives by simple blending in the
moltenstate. All these features make PLA attractive alter-native
for synthetic plastic materials of petrochemi-cal origin which
degrade slowly (even a few hundredyears) [1]. Therefore, PLA is the
subject of growing
.
mailto:[email protected]://www.cbmm.lodz.pl
-
2820 M. Pluta et al. / European Polymer Journal 43 (2007)
2819–2835
MA
CR
OM
OL
EC
UL
AR
NA
NO
TE
CH
NO
LO
GY
scientific and practical interest in the last years. Inthe
literature one can find numerous papers con-cerning polylactide
filled with layered silicates of dif-ferent nature and properties
[2–11]. If thesecomponents are compatible, a true
nanocompositesystem can be formed. Such nanocomposites
exhibitimproved physical properties, comparing to those ofthe
unfilled polymer matrix: mechanical strength,barrier properties,
thermal resistance and dimen-sional stability, even at low filler
concentration(1–5 wt%). However, dielectric properties of
poly-lactide-based nanocomposites containing layeredsilicate have
not been explored yet.
Polylactide (Scheme 1) is a type A polyester(there is a
component of the dipole moment parallelto the chain). In amorphous,
racemic samples threerelaxation maxima are observed in the
dielectricspectra: at ca. �80 �C (b-relaxation process,
localmotions), at ca. 60 �C (aS-relaxation, dynamic
glasstransition) and at ca. 85 �C (aN-relaxation, normalmode) (all
at 1 Hz) [12–17]. In the samples whichcrystallise the normal mode
(aN) is suppressed[15,17]. In the investigations of dynamic
mechanicalproperties the maximum related to crystalline frac-tion
is usually referred to as the a-relaxation processand consequently
the b-process corresponds to aS(and Tg) and the c-process
corresponds to the b-relaxation (local movements).
In our previous studies on PLA nanocomposites[5,11,18] we
concentrated on optimization of thepreparation of the
nanocomposites (nanoclaydelamination), their mechanical properties
and mor-phology. This paper is aimed at the study of thephysical
properties of PLA-layered silicate nano-composites with a special
attention being paid tothe dielectric properties and their
relationship withmechanical and thermal properties. Dielectric
spec-troscopy covers broader range of frequency thandynamic
mechanical measurements and is sensitiveto movements of the
elements of the polymer chainwhich possess dipole moments. It is
thereforeappropriate to study polylactide. Dielectric mea-surements
were used to study molecular dynamicsin polymer nanocomposites
based on polyiso-prene (PI) [19], polyamide, polystyrene (PS)
[20],
Scheme 1. Molecular structure of PLA.
poly(methyl methacrylate) (PMMA) [21], and poly-propylene (PP)
[22,23] but only in few cases fullpotential of this technique was
used. In the case ofPS it was claimed that organoclay decreases
dielec-tric permittivity [20,21]. In PP grafted with
maleicanhydrite an increase of dielectric permittivity,attributed
to Maxwell–Wagner–Sillars interfacialpolarization, was observed
[23].
The nanocomposites investigated in this studycontained
organomodified montmorillonite(organoclay), Cloisite� 30B, selected
basing onthe previous investigations’ [11,18]. The effect ofthe
organoclay concentration was studied for thenanocomposites
containing 3 wt% or 10 wt% oforganoclay. To improve the filler
dispersion a com-patibilizer – elastomeric ethylene copolymer
func-tionalized with maleic anhydride was also used.Reference
samples, PLA and PLA with the com-patibilizer, were investigated as
well. The systemswere characterized using various techniques:
broad-band dielectric spectroscopy (BDS), melt rheology,X-ray
diffraction (XRD), dynamic periodical defor-mation (DMTA),
differential scanning calorimetry(DSC).
2. Experimental
2.1. Materials
Polylactide was kindly provided by Cargill-Dow.It contains 95.9%
of L-lactide and 4.1% D-lactide(residual lactide content is 0.1%).
Its melt flow indexis 6.7 g/10 min (210 �C @ 2.16 kg). Organically
trea-ted montmorillonite Cloisite� 30B from SouthernClay Products
(Gonzales, TX) was used as a filler.This organoclay contains
methyl-bis(2-hydroxy-ethyl) tallowalkyl ammonium cations (Scheme
2,where T denotes tallow consisting of �65% C18,�30% C16 and �5%
C14.) Its content is 29.2 wt%(as determined with TGA, in N2 at 20
�C/min).The organic modifier increases the interlayer dis-tance
(i.e. the gallery thickness) to 1.84 nm. Exxe-lorTM VA1803
(ExxonMobil Chemical) was usedfor compatibilization. Exxelor
modifiers (amor-phous maleic anhydrite functionalised
elastomeric
Scheme 2. Chemical formula of the organomodifier.
-
M. Pluta et al. / European Polymer Journal 43 (2007) 2819–2835
2821
MA
CR
OM
OL
EC
UL
AR
NA
NO
TE
CH
NO
LO
GY
ethylene copolymers) are typically designed for useas impact
modifiers, compatibilizers, couplingagents and adhesion promoters
[24].
2.2. Sample preparation
After drying at 105 �C for 4 h under reducedpressure, PLA was
melt-blended with other compo-nents in a counterrotating internal
mixer (Brab-ender OHG, Duisburg, Germany). Unfilled PLAwas also
melt-processed to have a reference mate-rial. Melt processing was
carried out at the rotationspeed of 50 rpm for 20 min, in a dry
nitrogen atmo-sphere to prevent thermo-oxidative degradation ofPLA.
The temperature was set at 180 �C, however,it increased by about 10
�C during blending due toshearing processes. Concentration of the
Cloisite�
30B was 3 wt% or 10 wt% (inorganic fraction).Two polymer systems
compatibilized with 3 wt%Exxelor VA1803, 3 wt% Cloisite� 30B
containingnanocomposite, and polylactide were also studied.Sample
abbreviations, composition and molecularweight of PLA matrices are
given in Table 1. Spec-imens for the investigations were prepared
by com-pression molding (185 �C) to the thickness of100 lm for the
dielectric studies and to 0.5 mm forstructural characterization
using other techniques.The specimens were melt-quenched to obtain
amor-phous PLA.
2.3. Characterization
Molecular weight was determined for the neatPLA and all
melt-processed samples, after removingadditives, by size exclusion
chromatography (SEC)method in methylene chloride as described in
[18].The X-ray diffraction (XRD) technique was usedto characterize
nanocomposites on the nanostruc-ture level. The measurements were
performed in
Table 1Description of the investigated samples
Systems Compositions Mw Mw/Mn
PLA Processed PLA 94400 1.48N3M 97 wt%PLA + 3 wt%Cloisite� 30B
73700 1.25N10M 90 wt%PLA + 10 wt%Cloisite� 30B 89000 2.00PLA-C 97
wt%PLA + 3 wt%Exxelor
VA1803118000 1.68
N3M-C 94 wt%PLA + 3 wt%Cloisite�
30B + 3 wt%Exxelor VA180372000 1.25
Molecular masses of PLA determined using SEC in
methylenechloride (neat polylactide has Mw = 126000, Mw/Mn =
1.48).
the transmission mode (coupled h/2h) in the 1.2�to 8� range of
2h. A wide-angle goniometer was cou-pled to a sealed-tube source of
filtered CuKa radia-tion, operating at 50 kV and 30 mA
(PhilipsPW3830). The slit system enabled collection of dif-fracted
beam with the divergence angle of less than0.05�.
Rheological properties were studied with anadvanced
research-grade rheometer (ARES; Rheo-metric Scientific) using a
parallel plate geometry(diameter 25 mm). This technique is
sensitive to fol-low the dispersion of the filler particles in
polymermatrix [18,25]. The compression molded sampleswere placed
between hot plates and stabilized at170 �C for about 5 min before
the measurement.Dynamic frequency sweep was performed in theregion
of linear viscoelastic response (LVR), witha strain 0.8% for the
nanocomposites and 2% forsamples not containing montmorillonite,
startingfrom high frequency, 512 rad/s down to 0.02 rad/s.The LVR
region was experimentally establishedbecause it is dependent on the
sample compositionand degree of dispersion of the nanoclay
[18].Experimental data were related to the actual gapvalue (�0.9
mm).
Thermal properties of the samples were investi-gated with a DSC
2920, TA Instruments, undernitrogen atmosphere. The crystallization
behaviorof the PLA matrix from the initially glassy, amor-phous
state was investigated at a heating rate of3 �C/min, following the
melt-quenching scan.
Dynamic mechanical properties of 0.5 mm thicksamples were
measured with an MkIII DMTAapparatus (Rheometric Scientific, Inc.)
in a dual-cantiliver bending mode. The dynamic storage andloss
moduli (E 0 and E00) were determined at a con-stant frequency of 1
Hz as a function of temperaturefrom �90 �C to 140 �C at a heating
rate of 2 �C/min.
The dielectric properties were studied in parallelplate
geometry. The samples were provided withevaporated circular Al
electrodes (30 mm in diame-ter). Before Al deposition the specimens
were driedovernight in vacuum at 40 �C (i.e. below Tg to avoidthe
cold crystallization of the initially amorphousPLA matrix). The
complex dielectric function wasmeasured at constant temperatures in
the frequencyrange from 10�1 Hz to 106 Hz by a NovocontrolConcept
40 a-analyzer interfaced to the sample bya broadband dielectric
converter (BDC, Novocon-trol) or a lock-in amplifier (Stanford
Research810). The temperature of the sample was varied
-
Fig. 1. X-ray diffractograms recorded for the nanocompositesN3M,
N10M and N3M-C. Reference diffractograms of theircomponents:
unfilled PLA, Cloisite� 30B and 3 wt% ExxelorVA1803 nanocomposite
(NEx3M) are shown for comparison.The diffractograms are vertically
shifted for clarity.
2822 M. Pluta et al. / European Polymer Journal 43 (2007)
2819–2835
MA
CR
OM
OL
EC
UL
AR
NA
NO
TE
CH
NO
LO
GY
from �100 �C to +120 �C in steps of 3 �C and wascontrolled with
a stability of DT = 0.1 �C (Novo-control Quatro system controller
BDS 1330). Aver-age heating rate was ca. 0.5 �C/min.
3. Results and discussion
3.1. Size exclusion chromatography (SEC)
The melt processing leads to some decrease ofmolar weight (Mw)
due to degradation of the poly-lactide (Table 1) although
protective nitrogen atmo-sphere has been used. This is due to
susceptibility ofPLA to degradation in the molten state
observedalso by other groups [26,27]. It can be seen thatthe Mw of
the unfilled PLA is decreased by �25%,while that of the
compatibilized PLA (PLA-C) onlyby �6% as compared to the neat PLA.
It suggestssome stabilizing effect of the compatibilizer towardsPLA
degradation or some chemical interactionsbetween these components.
The melt processing inthe presence of the nanoclay contributes to
furtherdecrease of the Mw, even if the compatibilizer wasused.
3.2. XRD analysis
Fig. 1 shows X-ray diffractograms of the nano-composites N3M,
N10M and N3M-C. Diffracto-grams of unfilled PLA, of Cloisite� 30B
powderand of Exxelor VA1803 filled with 3 wt% Cloisite�
30B (sample NEx3M) are also shown for compari-son. The unfilled
PLA sample shows typical back-ground scattering with an intensity
increasingbelow 1.2�. The diffractogram for Cloisite� 30Bhas a
distinct maximum around 2h = 4.7�(d001 = 1.8 nm). This maximum is
not observed forthe nanocomposites. The nanocomposites revealonly
very small bulge around 2h � 2.5� (d001 =3.5 nm). This feature is
characteristic of a good dis-persion of the organoclay, achieved by
an intercala-tion followed by tactoids formation and
thenexfoliation of the nanoplatelets in the PLA matrix.These
features were observed by TEM for systemssimilar to N3M, discussed
in [18]. Compatibilizationof 3 wt% nanocomposite (sample N3M-C)
results ina further increase of the organoclay dispersion asthe
bulge in the diffractogram is decreased and thescattering at the
lowest 2h is stronger than that forthe not compatibilized
counterpart (N3M). In orderto investigate the interaction between
the organo-clay and compatibilizer, an additional nanocompos-
ite (Exxelor VA1803 + 3 wt% Cloisite� 30B) wasalso prepared.
Diffractogram for this system, in con-trast to the other ones,
reveals a strong intensityincrease at low 2h (typical for the
exfoliated struc-tures) and a small diffraction maximum at2h �
6.0�, i.e. larger than for the Cloisite� 30B(2h = 4.7�). It
corresponds to the decreased inter-layer thickness (to about 1.5
nm). This indicates acollapse of some fraction of the nanoplatelets
inthe assembled regions. Diffractogram of the
unfilledcompatibilizer (not shown in Fig. 1) is similar tothat of
the unfilled PLA sample.
The shape of the diffractogram obtained for the10%-nanocomposite
(sample N10M) indicates alsoa high degree of the organoclay
dispersion. How-ever, this diffractogram has a slightly larger
bulgearound 2.5� than the less filled N3M sample. Thiscan mean that
the content of the organoclay in theN10M nanocomposite is too high
to obtain homo-geneous dispersion of the organoclay and full
sepa-ration of the silicate nanoplateles in the PLAmatrix. Probably
in this nanocomposite system con-centration of the organoclay is
above a percolationlevel and some montmotrillonite network is
formed.
3.3. Rheological properties
Polylactide filled with the organoclay and/orcompatibilizer
exhibits different rheological proper-ties as compared with neat
PLA. This is illustrated
-
M. Pluta et al. / European Polymer Journal 43 (2007) 2819–2835
2823
LO
GY
in Fig. 2 showing dependencies of log G 0, log G00 andlog g* vs.
log x for the considered samples. Theunfilled PLA is featured by a
typical increase ofthe G 0 and G00 with deformation frequency
x(Fig. 2a). At lower x the G 0 is smaller than G00, indi-cating a
classical liquid-like behavior of the moltenpolymer. The difference
between G 0 and G00
decreases gradually with raising deformation fre-quency.
Finally, the curves cross at x � 115 rad/s,and above this frequency
G 0 is higher than G00
revealing domination of the solid response. It isworth to
notice, that at the highest frequencies allthe samples exhibit
comparable values of G 0 andcomparable values of G00, with G 0 >
G00. This indi-cates that at these frequencies the response of
thesamples is weakly affected by the nanostructureand determined
mainly by the properties of thepolymer matrix. Similar relations at
high frequen-
Fig. 2. Comparison of frequency dependencies of the storage
modulusPLA, compatibilized PLA (PLA-C), nanocomposites (N3M and
N10M)the same for all the samples to illustrate/highlight changes
in the rheol
cies have been found for other nanocomposites(e.g. [28]).
Polylactide with the compatibilizer (sample PLA-C) exhibits
rheological behavior similar to that ofunfilled PLA. For the former
sample the differencebetween G 0 and G00 is somewhat smaller, and
thecrossover point appears at slightly lower x �100 rad/s (Fig.
2b). The complex viscosity of theunfilled PLA and the
compatibilized sample exhibita Newtonian behavior (g* is
independent of fre-quency) with values about 4000 Pa s and 5000Pa
s, respectively, then, above 10 rad/s a shear-thin-ning takes
place. This behavior is typical forhomopolymers.
Samples containing the organoclay – nanocom-posites N3M, N3M-C
and N10M, exhibit morepronounced changes of the rheological
parameters.One can see that the higher organoclay dispersion
(G 0), loss modulus (G00) and complex viscosity (g*) of the
unfilledand compatibilized nanocomposite (N3M-C). The vertical
scale isogical behavior.
MA
CR
OM
OL
EC
UL
AR
NA
NO
TE
CH
NO
-
Fig. 3. DSC heating thermograms recorded for the samples at
aheating rate of 3 �C/min. The scans were performed directly
aftermelt-quenching at a cooling rate of 3 �C/min. The curves
arenormalized to the mass of PLA.
2824 M. Pluta et al. / European Polymer Journal 43 (2007)
2819–2835
MA
CR
OM
OL
EC
UL
AR
NA
NO
TE
CH
NO
LO
GY
(samples N3M and N3M-C) and the higher organo-clay concentration
(samples N3M and N10M) thelarger increase of G 0 and G00 at low
frequencies isobserved (compare Fig. 2c with Fig. 2e and d withFig.
2e, respectively). Moreover, for the 3 wt%-nanocomposites (N3M and
N3M-C) two crossoverpoints are observed: x1 � 0.02 rad/s and x2
�180 rad/s for N3M; x1 � 50 rad/s and x2 � 150rad/s for N3M-C,
respectively (arrows in Fig. 2.Between x1 and x2 the behavior of
the samples isliquid-like (G 0 < G00). Below x1, G 0 exceeds G00
andshows a plateau which indicates the so called‘‘pseudo-solid-like
behavior’’ of the material atlow deformation frequencies [28].
The plateau region is larger for the nanocompos-ite N3M-C than
for the nanocomposite N3M due tobetter dispersion of the
organoclay. The dispersiondependent plateau/pseudo-solid-like
response wasalso observed for similar PLA/organoclay systemsin
which exfoliation degree was increased by prolon-gation of the
compounding time [18]. For the nano-composite N10M with higher
organoclay contentthe plateau region is enlarged which leads to
therelation G 0 > G00 (‘‘pseudo-solid-like’’ response) inthe
whole frequency range.
Viscosity g* of all the nanocomposites (N3M,N3M-C, N10M) does
not show a Newtonian behav-ior in the deformation frequency range
studied. Theviscosity decreases with increasing frequency, show-ing
a shear-thinning response, the more pronouncedthe higher is
dispersion and/or concentration of theorganoclay (the initial g*
slope is: �1.02 for N10M,�0.98 for N3M-C and �0.95 for N3M, while
it isclose to zero for PLA and PLA-C samples). Theincrease of the
moduli and of the viscosity, seen dis-tinctly at lower frequency
range, reflects a reinforce-ment of the molten polymer matrix by
the filler. Thereinforcing effect results from the
interactionsbetween the components due to hydrogen bondingof
hydroxyl groups in the organic ‘‘surfactant’’ inthe organoclay and
carbonyl groups of PLA chainsegments. The interactions are the
stronger the lar-ger is the interface area, thus being related to
higherdispersion and/or concentration of the organoclayin the PLA
matrix. Similar concentration/dispersiondependent rheological
spectra were observed forpolystyrene-clay nanocomposites [29].
3.4. Calorimetric characterization
Fig. 3 shows DSC heating thermograms of sam-ples of initially
amorphous PLA and of the nano-
composites. The measurements were performedimmediately after
melt-quenching scans, so the sam-ples have the same thermal history
without agingcycle. The samples reveal the following thermalevents
with increasing temperature: glass–rubbertransition (at Tg), cold
crystallization process (char-acterized by Tcc and DHcc) and the
melting processwith two components (characterized by Tm1, DHm1and
Tm2, DHm2). Tcc and DHcc denote crystalliza-tion temperature and
enthalpy, respectively. Tm1,and Tm2, DHm1 and DHm2 denote the
temperaturesof the melting peaks and the corresponding
meltingenthalpies. Comparing thermograms and calorimet-ric
parameters collected in Table 2 one can see thatfilling of PLA
modifies, more or less, the individualthermal processes. Tg is
somewhat decreased, from57.7 �C for PLA to 56.9 �C and to 53.0 �C
for thenanocomposites N3M and N10M, respectively. ThisTg decrease
can be explained by a plasticizing effectof the clay surfactant
(�30% in the organoclay).Compatibilized PLA (sample PLA-C)
exhibitTg = 58.6 �C, higher than unfilled PLA (57.7 �C),suggesting
some reinforcement of the polylactidematrix by the compatibilizer.
The compatibilizednanocomposite N3M-C shows slightly lowerTg = 57.1
�C than PLA (DTg = �0.6 �C) indicatingsome balancing effect of the
compatibilizer and thenanoclay surfactant.
The influence of Cloisite� 30B content on thecold
crystallization of the PLA matrix is discernible.The Tcc decreases
from 104.6 �C for unfilled PLA to98.3 �C for N3M and to 92.0 �C for
N10M, while
-
Table 2Calorimetric parameters of the investigated materials
(heating rate 3 �C/min)
Sample Tg (�C) Cold crystallization Melting
Tcc (�C) DHcc (J/g) Tm1 (�C) DHm1 (J/g) Tm2 (�C) DHm2 (J/g)
DHmTotal (J/g)
PLA 57.7 104.6 29.1 148.8 12.8 156.8 16.9 29.0N3M 56.9 98.3 29.2
147.3 8.1 155.4 21.5 29.6N10M 53.0 92.0 31.4 – – 154.0 33.4
33.4PLA-C 58.6 105.2 29.0 149.4 14.7 156.8 14.7 29.4N3M-C 57.1
107.8 30.3 149.6 15.6 156.3 15.9 31.5
M. Pluta et al. / European Polymer Journal 43 (2007) 2819–2835
2825
MA
CR
OM
OL
EC
UL
AR
NA
NO
TE
CH
NO
LO
GY
the crystallization enthalpies are comparable(�30 J/g) (Table
2). The decrease of the cold crystal-lization temperature can be
ascribed to a nucleatingeffect of the filler. The concentration
dependentnucleating effect has also been observed in
similarcompositions of PLA and Cloisite� 30B [10]. More-over, the
organoclay surfactant can also contributeto lowering of the cold
crystallization temperature,because of its plasticizing effect. The
effect of plast-icizers on the crystallization of PLA is discussed
indetail in other papers [30–32]. Compatibilized sam-ples PLA-C and
N3M-C reveal the cold crystalliza-tion at higher temperatures than
the PLA sample,with an exothermic peaks at 105.2 �C and107.8 �C,
respectively. The crystallization enthalpyis equal to �30 J/g –
like those of the other samplesdiscussed (Table 2). These results
indicate that thecompatibilizer reduces mobility of PLA
macromol-ecules. This effect is stronger in the
compatibilizednanocomposite (sample N3M-C) due to
improveddispersion of the organoclay (accordingly to X-rayand
rheological data). The influence of the organo-clay dispersion on
crystallization of PLA was dis-cussed in detail in [18].
Melting of the samples is reflected by an endo-therm with two
components, at Tm1 (around148.5 ± 1.2 �C) and at Tm2 (around 155.5
±1.3 �C). Two melting peaks reflect the process ofmelting of
crystallites having different sizes and/orperfection of ordering.
In the case of non-compati-bilized samples (PLA, N3M and N10M) the
sizeand temperature position of the low temperaturemelting
component decreases with the organoclaycontent, which corresponds
to a decrease the coldcrystallization temperature (Tcc) of these
samples.For the 10 wt% nanocomposite (N10M) the meltingcomponent at
Tm1 is seen as a low temperatureshoulder of the main melting peak
at Tm2. Thismeans, the lower is Tcc the lower is the contributionof
the melting process at Tm1, while that at Tm2relatively increases.
For the N10M sample the
total melting enthalpy DHmTotal � 33.0 J/g, i.e. it ishigher
than the crystallization enthalpy DHcc �31.4 J/g (Table 2). This is
connected with the factthat smaller, less perfect crystallites,
formed at lowerTcc undergo during the measurement a
furtherrecrystallization at higher temperature, above
Tcc,contributing to the melting enthalphy at Tm2.
Therecrystallization process was also revealed by anon-reversing
signal of the TMDSC (not shownhere). Compatibilized samples, PLA
and the nano-composite N3M-C, are featured by a slightly largerlow
temperature melting component (DHcc �14.6 J/g) which appears at a
higher Tm1 � 149.5 �C,than for the unfilled PLA (DHcc = 12.8 J/g,
Tm1 �148.8 �C). This is due to the cold crystallization athigher
temperature in the case of the compatibilizedsamples (Fig. 3).
Therefore more stable crystallitesare formed, which, give raise
only to one meltingprocess on further heating, without
recrystallizationcontribution (as indicated by the non-reversing
sig-nal of the TMDSC – not shown here).
It is interesting to note that the crystallizationenthalpy
(DHcc) of PLA is similar for all the samples(within an experimental
error) and it is close to thetotal melting enthalpy (DHmTotal),
(Table 2). Thisindicates that the additives used (organoclay,
com-patibilizer) does not affect significantly the
finalcrystallinity of the PLA matrix developed duringheating, but
have some influence on the crystal sizeand perfection (as Tcc
changes).
3.5. Viscoelastic properties
Fig. 4a shows the storage modulus (E 0) of the dis-cussed
samples determined by dynamic mechanicalthermal analysis (DMTA) at
1 Hz, plotted vs. tem-perature. For all the samples the following
charac-teristic E 0 changes with raising temperature areobserved: a
gradual decrease in the region �90 �Cto 50 �C, a rapid drop below
55–60 �C due to theglass–rubber transition (Tg), an increase in the
cold
-
Fig. 4. Temperature dependencies of E 0 (a) and E00 (b)
recordedfor the investigated systems of different composition.
2826 M. Pluta et al. / European Polymer Journal 43 (2007)
2819–2835
MA
CR
OM
OL
EC
UL
AR
NA
NO
TE
CH
NO
LO
GY
crystallization range (around 100 �C) owing to rein-forcing by
the crystallites being formed, and then adecrease as a result of
the pre-melting process. It isworth to mention that below Tg E
0 slightlyincreases, before dropping down. This effectdecreases
with filling and it was observed also forother amorphous PLA
samples [33]. An increaseof E 0 prior to Tg can be ascribed to a
relaxation ofinternal stresses, frozen in during melt-quenchingof
the material.
Filling PLA with the organoclay and/or the com-patibilizer,
increases E 0 of the system in the wholetemperature range. Below
the glass transition tem-perature E 0 increases in the order PLA
< PLA-C � N3M-C < N3M < N10M. It means, that E 0increases
with the organoclay content, but it is alsoenhanced by the
compatibilizer. This indicates thatthe compatibilizer (which is an
elastomer with Tg
around �57 �C) being dispersed in the PLA matrixreinforces it
slightly ðE0PLA-C > E0PLAÞ. This suggestssome interaction of
both components. The presenceof the organoclay in the
compatibilized system(N3M-C) results also in the increase of E 0.
The max-imum around 100 �C in E 0 (and E00) correspond tothe cold
crystallization, revealed also by the DSCscans (Fig. 3).
The dependencies of E00 on temperature areshown in Fig. 4b. The
filled systems show theincreased mechanical loss in comparison with
theunfilled PLA. A small maximum is observed around�60 �C, probably
related to local movements isreferred to as a c-relaxation in
mechanical measure-ments and a b-relaxation in the dielectric
conven-tion. The maximum in E00 about 60 �C correspondsto segmental
relaxation (mechanical b-relaxationor aS in the dielectric
convention), related to theglass–rubber transition in PLA. The
maximum ofthe E00 around 100 �C (a-process according to
themechanical convention) reflects an increase of themechanical
loss due to the cold crystallization. Itsintensity increases with
the increasing organoclaycontent.
3.6. Broadband dielectric spectroscopy
Fig. 5 shows frequency dependencies of the imag-inary component
of the complex dielectric functione00 for initially amorphous
samples of unfilled PLAand the compatibilized nanocomposite
(N3M-C)at selected temperatures. The maximum, clearlyobserved above
60 �C for PLA and also for thenanocomposite, is ascribed to the
aS-relaxation[12–17]. Above 80 �C the crystallization of
initiallyamorphous PLA begins, which leads to a ca. two-fold
decrease of the aS maximum at 90 �C and101 �C (observed also by
other groups [14,17]).The temperature range in which PLA
crystallizationtakes place in dielectric measurements is
differentfrom that observed in DSC (Fig. 3) because theheating rate
in BBDS experiment is much lower(compare also Figs. 9a and 10a
below).
The most important difference in the dielectricspectrum of the
nanocomposite, as compared withthe unfilled PLA, is a strong
increase of e00 at lowfrequencies, the more important the higher is
thetemperature. Its frequency dependence and compar-ison with the
results obtained for pure Closite� 30Bsample, show that is should
be ascribed to directcurrent (dc) conductivity and interfacial
polariza-tion (Maxwell–Wagner–Sillars effect – MWS), most
-
Fig. 5. Dielectric spectra of PLA (a) and of the
nanocompositeN3M-C (b) at several temperatures around and above
Tg.
Fig. 6. Comparison of e00 spectra of the investigated systems
at80 �C. Solid lines represent fits using HN functions for
aS-relaxation and MWS process and rDC � x�0.85. Thin solid
andbroken lines show fit components for PLA and
N3M,respectively.
M. Pluta et al. / European Polymer Journal 43 (2007) 2819–2835
2827
MA
CR
OM
OL
EC
UL
AR
NA
NO
TE
CH
NO
LO
GY
probably related to the presence of ionic speciesused to
intercalate the montmorillonite during itsorganomodification (vide
infra). Similar behaviorwas observed for the non-compatibilized
nanocom-posite (N3M).
The frequency dependence of the imaginary partof dielectric
permittivity (e*) in the investigated sys-tems can be fitted using
a superposition of Haviri-lak–Negami (HN) empirical functions
[34]
e� � e1 ¼De
ð1þ ðixsHNÞaÞcð1Þ
where e1 is the real part of e* for frequencies muchhigher than
the maximum frequency of a given pro-cess, De is relaxation
strength, x denote frequency,sHN is so called Havirilak–Negami
relaxation timeand exponents a and c represent broadening of
thedistribution of the relaxation times.
A contribution of the dc conductivity and elec-trode
polarization can be taken into account assum-ing its frequency
dependence in a usual form [34]
e00 ¼ rdce0xm
where e0 denotes the dielectric permittivity of freespace and
rdc – the dc conductivity (at frequencyapproaching zero). The
exponent m should be equalto 1 in the case of pure electronic
conductivity butusually is smaller than one because electrode
andinterfacial polarization effects also come into play.In our
systems at 80 �C m was equal to 0.8 (alsofor pressed nanoclay
pellets) and increased withtemperature up to 0.95 at 120 �C.
In Fig. 6 e00 spectra vs. frequency for all the sam-ples at 80
�C (>Tg) are compared. This temperatureallows to observe all
relevant processes in practi-cally still amorphous samples. The
lines goingthrough the experimental points represent fittingresults
and thin lines show the components contrib-uting to the fits for
PLA and N3M. The componentsof other fittings were omitted for
clarity. It can beseen that the e00 spectra of PLA and PLA-C can
bewell fitted using a superposition of two main pro-cesses:
aS-relaxation with the maximum around104 Hz and rdc contribution at
low frequencies.An additional weak process around 3 Hz may
berelated to the normal mode (aN) (which shouldappear around this
temperature [12,16]). Its smallintensity may be caused by
crystallization of PLAwhich begins below 80 �C (lower heating rate
as
-
Fig. 7. Comparison of the tand spectra of PLA, PLA-C and
thenanocomposites (N3M, N3M-C and N10M) recorded at �70 �C(a) and
tand spectra of the nanocomposites (N3M, N10M) atvarious
temperatures (b) Corresponding spectra of montmoril-lonite at �10
�C, �40 �C and �70 �C (divided by 10) are shownfor comparison
(lines with small symbols).
2828 M. Pluta et al. / European Polymer Journal 43 (2007)
2819–2835
MA
CR
OM
OL
EC
UL
AR
NA
NO
TE
CH
NO
LO
GY
compared with DSC) and appearing crystallites hin-der chain
relaxation.
The intensity of the aS process is weakly depen-dent on sample
composition, the contribution ofrdc increases in the
nanocomposites. In the nano-composites, in spite of the high
conductivity, wecan clearly distinguish an additional strong
maxi-mum in e00. This maximum is observed also in
otherrepresentations which reduce the effect of dc con-ductivity –
modulus representation and de 0/d(log f)(not shown). It can be
assigned to the Maxwell–Wagner–Sillars effect as e 0 also strongly
increasesat low frequencies in this temperature range (videinfra,
Figs. 9 and 10). MWS process was alsoreported in polymer-silicate
nanocomposites basedon polyisoprene [19] and PP [23]. The fitting
param-eters are collected in Table 3.
The distribution of the aS-process is similar asfound by other
authors (a-coefficient equal to 0.55[12] or 0.66–0.76 [12,16] and c
= 0.65–0.8 [16]). Inthe case of the nanocomposites, especially
N10M,the fitting of the aS-process is not so reliable
becauseslightly different sets of parameters for the overlap-ping
processes give similar errors.
The relaxation time of the maximum assigned tothe
Maxwell–Wagner–Sillars polarization decreaseswith increasing
conductivity of the nanocompositesas expected. Surprisingly its
intensity slightlydecreases for higher organoclay loading, whichcan
indicate somewhat worse dispersion and separa-tion of the filler
nanoplatelets in the polymer matrix.DC conductivity of the
nanocomposites is roughlyproportional to the organoclay content and
hasthe same frequency dependence as that of neat Clo-site� 30B
sample (pressed discs) at low frequencies(rdc = 1.5 · 10�8 Ohm/m,
detailed data not shown).
The influence of montmorillonite on tand at alow temperature
(�70 �C) is shown in Fig. 7a. Atthis temperature the b-relaxation
maximum is wellseen. This process is attributed to local
movementsof small chain elements (carboxyl groups). The spec-
Table 3H–N parameters and rdc used for fitting of e00 spectra
presented in Fig
aS
a c s De
PLA 0.62 0.65 2.4 · 10�4 2.55PLA-C 0.62 0.65 2.8 · 10�4 2.2N3M
0.62 0.56 2.8 · 10�4 2.55N3M-C 0.62 0.56 2.8 · 10�4 2.5N10M 0.57
0.45 2.8 · 10�4 3.1
tra of the nanocomposites with 3% of organoclay(N3M and N3M-C)
are similar to those of theunfilled PLA and PLA-C, respectively.
Theyappear, however, at somewhat higher frequencieswhich indicates
that the presence of the additives
. 6 using Eq. (1)
MWS rdc
a c s De
0.56 1 0.23 0.25 3.35 · 10�12
0.62 1 0.9 0.23 1.2 · 10�11
0.66 1 0.15 3.25 3.25 · 10�10
0.66 1 0.15 3.3 3.2 · 10�10
0.7 1 0.02 2.5 1.3 · 10�9
-
M. Pluta et al. / European Polymer Journal 43 (2007) 2819–2835
2829
only slightly increases local mobility of PLA chains.The
b-relaxation is highly distributed (a coefficientin Eq. (1) equal
to ca. 0.3) and symmetric (c = 1).
By contrast the 10 wt%-nanocomposite (N10M)has a different tand
spectrum. It shows considerablyenhanced dielectric loss with a
pronounced tandmaxima at frequency regions below 2 Hz and above106
Hz which mask much weaker b-relaxation peakwith maximum around
102–103 Hz. The position ofthe low frequency tand maximum is
practically tem-perature-independent, while the high frequency
one
Fig. 8. The temperature dependence of the r
moves, for both N10M and N3M, to higher fre-quencies as the
temperature is increased (Fig. 7a).
These strong maxima should be related to thepresence of low
molecular weight species (organicnanoclay modifier and possibly
strongly bound,not removed H2O). The remarkable differencebetween
N3M and N10M and Closite� 30B suggestthat a particular percolative
filler nanostructure isformed in the N10M nanocomposite. The fact
thatneither of these relaxation processes is observed inneat
Closite� 30B (Fig. 7b) proves good dispersion
elaxation times of a- and b-relaxation.
MA
CR
OM
OL
EC
UL
AR
NA
NO
TE
CH
NO
LO
GY
-
Fig. 9. Temperature dependence of dielectric permittivity e 0
(a)and tand (b) at 1 kHz for PLA and for the
nanocomposites(initially amorphous – 1st run and crystalline
samples 2nd run).
2830 M. Pluta et al. / European Polymer Journal 43 (2007)
2819–2835
MA
CR
OM
OL
EC
UL
AR
NA
NO
TE
CH
NO
LO
GY
of the nanofiller in spite of its high concentration,
inagreement with XRD data (cf. Fig. 1). The percola-tion paths
and/or nanostructure are considerablyaffected by heating and
crystallization as thespectrum of the crystalline N10M sample
(secondrun – not shown) is qualitatively different from thatof the
amorphous sample (but also different fromthat of unfilled,
semicrystalline PLA).
Fig. 8 shows the temperature dependence of therelaxation times
for the aS and b processes deter-mined from frequencies of maximum
loss
s ¼ 12pfmax
For the b-maximum the temperature dependenceobeys the Arrhenius
law with an activation energy9.9 kcal/mol (41.5 · 103 J/K/mol) in
reasonableagreement with the activation energy 10.5 kcal/molgiven
in [12]. It can also be seen that neither theorganoclay nor the
compatibilizer have significanteffect on the activation energy,
only on the pre-exponential factor is decreased.
The temperature dependence of the aS-relaxationfor PLA and the
nanocomposites obeys Vogel–Ful-cher–Tammann law
s ¼ s0 expB
T � T 0
� �
with s0 = 13, B = 800 K and T0 = 302 K in reason-able agreement
with previous results s0 = 12.5–14.3[12] or 11.6–12.6 [16], B =
520–760 K [12] or 259–289 K [12] and 452–564 [16] and T0 = 277–285
[16](for materials of different molecular weight).
In the case of the composites (especially N10M)the relaxation
times at higher temperatures cannotbe unambiguously determined,
because of overlap-ping of Maxwell–Wagner–Sillars maximum and
dcconductivity.
Fig. 9 shows a comparison of e 0 and tand vs. tem-perature above
0 �C for unfilled PLA and the nano-composites at frequency 1 kHz.
The spectra forinitially amorphous and for semicrystalline
samples(the second run on the same sample after heating to120 �C
and annealing for 1 h) are presented to showthe effect of
crystallisation. The maximum of e 0
around 80 �C, observed in the initially amorphoussamples,
appears as a result of the cold crystalliza-tion and it is
discussed in more detail below. It isnot observed in the
semicrystalline samples in whichcold crystallization does not occur
any more. Amonotonic increase of e 0 above 50 �C in the
nano-composites is due to the Maxwell–Wagner effect
related with increasing dc conductivity or possiblyalso with
electrode polarization.
Tand shows maxima for both crystalline andamorphous samples. In
crystalline samples theirintensity is decreased and it is shifted
by ca. 5 �Cto higher temperatures. This is due to limited
move-ments of the chains partially incorporated in thecrystallites,
in agreement with [15,17]. The maxi-mum for PLA with the
compatibilizer (PLA-C) isalso slightly shifted to higher
temperatures. Crystal-lization does not have a significant effect
on anincrease of tand at higher temperatures due to
dcconductivity.
Fig. 10 shows a comparison of the temperaturedependence of e 0
and tand at fixed frequency 1 Hz(the same as used in the mechanical
measurements).At lower frequency the aS-relaxation is observed
at
-
Fig. 10. Temperature dependence of dielectric permittivity(a)
and tand (b) at 1 Hz for PLA and the nanocomposites.
Fig. 11. Temperature dependence of tand for PLA-C (a) andN3M-C
(b) at different frequencies.
M. Pluta et al. / European Polymer Journal 43 (2007) 2819–2835
2831
MA
CR
OM
OL
EC
UL
AR
NA
NO
TE
CH
NO
LO
GY
lower temperatures (c.f. Fig. 9) and it is clearly seenthat the
dielectric spectra of the amorphous samplesin the range 50–90 �C
are influenced by the coldcrystallization. The increase at high
temperaturesis more important at 1 Hz (please mind the logarith-mic
scale) which confirms that it is related mostly tothe
Maxwell–Wagner–Sillars effect and dc conduc-tivity. The
compatibilizer does not show any signif-icant influence on the
dielectric properties (noadditional relaxation process is observed)
but itspresence slows down the crystallization in agree-ment with
the results obtained by DSC (Fig. 3)and DMTA (Fig. 4).
The plot of tand vs. temperature (Fig. 9b) showsclearly the
maximum corresponding to the aS-relax-ation. Its position �60 �C is
closer to Tg determined
from the DSC data (Table 2), because of lower fre-quency. Its
intensity also decreases after crystalliza-tion. The second
maximum, which is not observedin the crystalline samples (data not
shown for clar-ity) is a result of fast crystallization (the
relaxationstrength decreases). Its maximum temperature isslightly
lower as compared with the mechanical data(cf. Fig. 4b).
It can be seen that the presence of the organoclayresults in a
strong increase of the dielectric loss at1 Hz already above the
room temperature. The aS-relaxation is in the same position (around
60 �C)but it is much broader. This effect is observed alsoin the
mechanical relaxation. It should be relatedin part to the
organoclay itself, which exhibits a
-
Table 4Comparison of the influence of the nanofiller and
compatibilizeron the maximum temperature of segmental relaxation
(dynamicglass transition) as determined using dynamic mechanical
anal-ysis and dielectric spectroscopy and Tg determined by DSC (all
in�C)
Sample DMA(tand at1 Hz)
BBDS(tand at1 Hz)
BBDS(tand at1 kHz)
DSC(inflectionpoint)
PLA 60.3 61.2 74.3 57.7N3M 58.9 60.4 73.2 56.9N10M 57.3 58.5
72.0 53.0PLA-C 59.0 61.2 74.3 58.6N3M-C 59.3 59.7 73.5 57.1
2832 M. Pluta et al. / European Polymer Journal 43 (2007)
2819–2835
MA
CR
OM
OL
EC
UL
AR
NA
NO
TE
CH
NO
LO
GY
strong loss in this temperature range. However, theobserved
spectrum is not a simple superposition.Dipolar species present in
the organoclay (amine)are now in a different medium (as compared
withneat Closite� 30B) and their relaxation is controlledto some
extend by the mobility of adjacent polymerchains. The maximum
attributed to the crystalliza-tion of PLA around 83–90 �C is also
observed inthe nanocomposites and it is much stronger (mindthe
logarithmic scale), as the crystallizationdecreases also ion
mobility and thus rdc). It shouldalso be noted that the influence
of the compatibilizeron the position of this maximum is similar as
in thecase of PLA. These observations confirm an interac-tion
between the compatibilizer and the polymermatrix.
In Fig. 11 one can see an evolution of the temper-ature
dependence of tan d for PLA-C and the nano-composite at different
frequencies. The maximumrelated to the aS-process shifts towards
higher tem-peratures as the frequency increases while the max-imum
related to crystallization is in the sameposition. Thus, both
maxima merge together for fre-quencies higher than 100 Hz.
4. Discussion
Various experimental techniques used enabled usto address the
problem of the influence of the nano-filler on physical properties
of the polymer matrix,in particular on chain relaxation. Mechanical
prop-erties and viscosity are sensitive primarily to largescale
chain movements while dielectric spectroscopyprovide information
mainly on local dipolar groupsfluctuations. The presence of rigid
silicate planes inthe nanocomposite must restrict movements of
thesurrounding polymer chains. The better is the dis-persion of the
clay (delamination) and the higheris the clay content the stronger
effect should beobserved.
The most significant influence is expected for thelarge scale
movements – on the rheological proper-ties in the terminal region
and on the normal mode(aN relaxation). We indeed observed
significant stiff-ening of the composite as compared with
neatpolymer matrix in the dynamic mechanical measure-ments (Fig. 4)
but the most evident effect is thesolid-like behavior of the
complex viscosity at lowfrequencies (Fig. 2) (in spite of some
lowering ofthe molecular mass of PLA – Table 1). It was notpossible
to observe a shift of the aN relaxation inthe dielectric spectra in
our systems because the
crystallization of PLA takes place in the same tem-perature
range (above 80 �C), in which this modewas observed [12] and it is
suppressed. In the nano-composites there is also a strong MWS
maximumobserved in this temperature range.
The smaller is the scale of molecular motions theless evident is
the effect of space confinement. Wemust also remember that low
molecular weightorganic molecules with long alkyl chains, presenton
the clay sheets, increase local mobility at theclay/polymer
interface. Thus, in the case of segmen-tal relaxation related to
the glass transition (termedas in discussions of dielectric
properties and b inmechanical properties) the situation is not so
obvi-ous. We observed a small effect of the filler on
thetemperature of this relaxation as measured usingdifferent
methods. The data are collected in Table4. It can be seen that the
values obtained for differ-ent materials do not differ much. DSC
gives thevalues lower by ca. 2 �C because they correspondto the
frequency of ca. 10�2 Hz. Mechanical anddielectric measurements at
1 Hz give similar results.In the case of BBDS an error of the
results obtainedfor the nanocomposites (especially for N10M) at1 Hz
is of the order of 1–2 �C because of the strongeffects related to
conductivity at low frequencieswhich make fitting ambiguous.
However, the sametrend is observed at higher frequencies (where
inturn some error might be due to an overlap withPLA
crystallization). In summary we can concludethat the presence of
the nanofiller shifts the segmen-tal relaxation to lower
temperatures by ca. 1 �C at3% filler content and by ca. 3 �C at 10
wt% content.The effect of the compatibilizer is small and
notobvious as different methods give contradictoryresults.
-
M. Pluta et al. / European Polymer Journal 43 (2007) 2819–2835
2833
MA
CR
OM
OL
EC
UL
AR
NA
NO
TE
CH
NO
LO
GY
Space confinement can increase or decrease chainmobility and Tg,
depending on interactions with thesolid surface as discussed
extensively e.g. in [34] andreferences therein. However, neither
the experi-ments on confinement in nanopores nor those
onspin-coated thin films, strictly correspond to the sit-uation in
the investigated layered nanocomposites.It should be noted that
taking into account specificsurface of the clay (750 m2/g) and its
content, evenat perfect delamination an average distance betweenthe
nanoplates can be estimated to be at most ca70 nm or 20 nm for N3M
and N10M respectively.It is much bigger than estimated radius of
gyration,so a small effect is not surprising, especially takinginto
account that in the real system the dispersionand delamination are
not perfect. Increased hin-drance of the chain segment motions
should alsoresult in a broadening of the relaxation maximumwhich
indeed was also observed in the temperaturedependencies (Fig. 11)
and as an increase of Havir-ilak–Negami exponents (a coefficient
equal to ca.0.6 for PLA and PLA-C, 0.55 for N3M and N3M-C and 0.53
for N10M. The effect for the nanocom-posites is, however, again
obscured by the conduc-tivity, so it is difficult to separate the
symmetricand asymmetric broadening effects.
The short range local relaxations giving rise tothe b-process in
the dielectric spectra are notexpected to be influenced by the
nanofiller andindeed only a small shift was observed. The
mostsignificant is the effect of the compatibilizer inPLA (but not
in N3M-C where it is probablylocated at the nanofiller/PLA
interface). The b-pro-cess is practically not visible in mechanical
relaxa-tion except for the systems with the compatibilizer.It is
not clear if a small maximum observed at�61 �C and �66 in PLA-C and
N3M-C respectivelyis due to Tg of the compatibilizer or to
b-process inPLA. Evidently in the dielectric spectra it is due
tob-relaxation in all cases, as evidenced by the temper-ature
dependencies of the relaxation times (Fig. 8b).At low temperatures
the most striking is the strongincrease of tand in N10M while in
N3M no signifi-cant difference as compared with PLA is found.
Weattribute it to a percolation of the nanoclay network(responsible
also for the stiffening observed in rheo-logical measurements –
Fig. 2), which facilitates dif-fusion of low molecular weigh ions
on longerdistances along the filler/PLA interface.
As expected for the two-phase system a maxi-mum due to the
Maxwell–Wagner effect wasobserved in the dielectric spectra of the
nanocom-
posites. Similar effect was observed also in nano-composites
based on other polymers: polyisoprene[19] polyamide-6 [35,36]
poly(propylene-graft-maleic anhydride) [23] relation of its
parameters todelamination/exfoliation is however not obvious.
5. Conclusions
The X-ray investigations exhibited exfoliatednanostructure in 3
wt%-nanocomposite. Compati-bilization noticeably enhanced the
degree of exfolia-tion of the organoclay due to combined
interactionsof the organoclay surfactant with polylactide chainsand
maleic anhydrite groups of the compatibilizer.In the 10
wt%-nanocomposite mixed – intercalatedand exfoliated nanostructures
were detected due tohigh concentration of the filler. Rheological
andmechanical properties suggest that a sort of silicatenetwork was
formed.
Using different techniques we were able to studythe effect of
the nanoclay on relaxation processes ina broad range of temperature
and frequency. Rheo-logical measurements, which are sensitive to
themovements of whole chains (terminal processes)show a significant
increase of viscosity and rein-forcement of the molten PLA matrix
due to thepresence of the dispersed organoclay. This effectwas
enhanced by the compatibilization andincreased with the organoclay
content. Rheologicalproperties of the unfilled PLA and
compatibilizedPLA were comparable indicating negligible influ-ence
of the compatibilizer on PLA in the moltenstate.
Viscoelastic spectra (DMTA data) showed agradual increase of the
storage and loss moduli withthe increase of the organoclay content
andimproved dispersion. The mechanical relaxationprocesses (a, b
and c) were affected to some extendby the material composition and
nanostructure. Thenanostructure affected the cold crystallization
pro-cess of PLA matrix as detected by DMTA, DSCand BBDS. In the
systems with the compatibilizercrystallization of PLA occurred at
highertemperature.
The dielectric properties of all the nanocompos-ites in the high
temperature range are dominatedby the processes related to ionic
species present inthe nanoclay – dc conductivity,
Maxwell–Wagner–Sillars effect and electrode polarization. At
hightemperatures both real and imaginary parts of e*
strongly increase, especially at low frequencies. Anadditional
maximum of e00 at around 80–90 �C
-
2834 M. Pluta et al. / European Polymer Journal 43 (2007)
2819–2835
MA
CR
OM
OL
EC
UL
AR
NA
NO
TE
CH
NO
LO
GY
resulting from the cold crystallization is observed.Dielectric
studies show a relatively weak effect ofthe nanofiller on aS and b
relaxation processes inPLA matrix, except for the highest
organoclay con-tent (N10M system), most probably because theaverage
distance between PLA segments and claysheets is too big to observe
space confinementeffects.
Acknowledgments
This work was partially supported by the Minis-try of Science
and Information Society Technolo-gies (Poland) through the Center
of Molecular andMacromolecular Studies, PAS, under Grant
No.PBZ–KBN-070/T09/2001, 2003–2006.
The author are indebted to Dr. I. Stevenson(LMBP) helping with
dielectric measurements andgratefully acknowledges the Cargill-Dow
Polymers,LLC, for supplying PLA.
References
[1] Sinclair RG. The case for polylactic acid as a
commoditypackaging plastic. J Macromol Sci Pure Appl
Chem1996;A33(5):585–97.
[2] Ogata N, Jimenez G, Kawai H, Ogihara T. Structure
andthermal/mechanical properties of poly(L-lactide) blend. JPolym
Sci, Part B: Polym Phys 1997;35(2):389–96.
[3] LeBaron PC, Wang Z, Pinnavaia TJ. Polymer-layeredsilicate
nanocomposites: an overview. Appl Clay Sci1999;15(1–2):11–29.
[4] Alexandre M, Dubois P. Polymer-layered silicate
nanocom-posites: preparation, properties and uses of a new class
ofmaterials. Mater Sci Eng R-Rep 2000;28(1–2):1–63.
[5] Pluta M, Galeski A, Alexandre M, Paul MA, Dubois
P.Polylactide/montmorillonite nanocomposites and micro-composites
prepared by melt blending: Structure andsome physical properties. J
Appl Polym Sci 2002;86(6):1497–506.
[6] Ray SS, Yamada K, Okamoto M, Fujimoto Y, Ogami A,Ueda K. New
polylactide/layered silicate nanocomposites. 5.Designing of
materials with desired properties. Polymer2003;44(21):6633–46.
[7] Ray SS, Okamoto M. New polylactide/layered
silicatenanocomposites. 6. Melt rheology and foam
processing.Macromol Mater Eng 2003;288(12):936–44.
[8] Ray SS, Okamoto M. Polymer/layered silicate nanocom-posites:
a review from preparation to processing. ProgPolym Sci
2003;28(11):1539–41.
[9] Ray SS, Bousmina M. Biodegradable polymers and theirlayered
silicate nanocomposites: In greening the 21st centurymaterials
world. Prog Mater Sci 2005;50(8):962–1079.
[10] Di YW, Iannace S, Di Maio E, Nicolais L. Poly(lactic
acid)/organoclay nanocomposites: Thermal, rheological propertiesand
foam processing. J Polym Sci, Part B: Polym
Phys2005;43(6):689–98.
[11] Pluta M, Paul MA, Alexandre M, Dubois P.
Plasticizedpolylactide/clay nanocomposites. I. The role of filler
contentand its surface organo-modification on the
physico-chemicalproperties. J Polym Sci, Part B: Polym
Phys2006;44(2):299–311.
[12] Mierzwa M, Floudas G, Dorgan J, Knauss D, Wegner J.Local
and global dynamics of polylactides. A dielectricspectroscopy
study. J Non-Cryst Solids 2002;307:296–303.
[13] Henry F, Costa LC, Devassine M. The evolution ofpoly(lactic
acid) degradability by dielectric spectroscopymeasurements. Eur
Polym J 2005;41(9):2122–6.
[14] Mijovic J, Sy JW. Molecular dynamics during
crystallizationof poly(L-lactic acid) as studied by broad-band
dielectricrelaxation spectroscopy. Macromolecules
2002;35(16):6370–6.
[15] Ren JD, Adachi K. Dielectric relaxation in blends
ofamorphous poly(DL-lactic acid) and semicrystalline poly(L-lactic
acid). Macromolecules 2003;36(14):5180–6.
[16] Ren JD, Urakawa O, Adachi K. Dielectric and
viscoelasticstudies of segmental and normal mode relaxations
inundiluted poly(D,L-lactic acid). Macromolecules
2003;36(1):210–9.
[17] Fitz BD, Andjelic S. Real-time monitoring of
segmentaldynamics during crystallization of poly(L(-)-lactide)
bysimultaneous DRS/SALS technique. Polymer 2003;44(10):3031–6.
[18] Pluta M. Melt compounding of
polylactide/organoclay:Structure and properties of nanocomposites.
J Polym Sci,Part B: Polym Phys 2006;44(23):3392–405.
[19] Mijovic J, Lee HK, Kenny J, Mays J. Dynamics in
polymer-silicate nanocomposites as studied by dielectric
relaxationspectroscopy and dynamic mechanical spectroscopy.
Mac-romolecules 2006;39(6):2172–82.
[20] Wang HW, Chang KC, Chu HC, Liou SJ, Yeh JM.Significant
decreased dielectric constant and loss of polysty-rene-clay
nanocomposite materials by using long-chainintercalation agent. J
Appl Polym Sci 2004;92(4):2402–10.
[21] Wang HW, Shieh CF, Chang KC, Chu HC. Synthesis
anddielectric properties of poly(methyl methacrylate)-clay
nano-composite materials. J Appl Polym Sci 2005;97(6):2175–81.
[22] Kim DH, Cho KS, Mitsumata T, Ahn KH, Lee SJ.Microstructural
evolution of electrically activated polypro-pylene/layered silicate
nanocomposites investigated byin situ synchrotron wide-angle X-ray
scattering and dielec-tric relaxation analysis. Polymer
2006;47(16):5938–45.
[23] Bohning M, Goering H, Fritz A, Brzezinka KW, Turky
G,Schonhals A, et al. Dielectric study of molecular mobility
inpoly(propylene-graft- maleic anhydride)/clay nanocompos-ites.
Macromolecules 2005;38(7):2764–74.
[24] Exxelor-datasheet. , 2006.[25] Wagener R, Reisinger TJG. A
rheological method to
compare the degree of exfoliation of nanocomposites.Polymer
2003;44(23):7513–8.
[26] Degee P, Dubois P, Jerome R. Bulk polymerization oflactides
initiated by aluminium isopropoxide. 3. Thermalstability and
viscoelastic properties. Macromol Chem Phys1997;198(6):1985–95.
[27] Wachsen O, Platkowski K, Reichert KH. Thermal degra-dation
of poly-L-lactide – Studies on kinetics, modelling andmelt
stabilisation. Polym Degrad Stabil 1997;57(1):87–94.
[28] Ren JX, Silva AS, Krishnamoorti R. Linear viscoelasticityof
disordered polystyrene–polyisoprene block copolymer
http://mymatweb.com
-
M. Pluta et al. / European Polymer Journal 43 (2007) 2819–2835
2835
based layered-silicate nanocomposites.
Macromolecules2000;33(10):3739–46.
[29] Zhao J, Morgan AB, Harris JD. Rheological characteriza-tion
of poly styrene-clay nanocomposites to compare thedegree of
exfoliation and dispersion. Polymer 2005;46(20):8641–8660.
[30] Kulinski Z, Piorkowska E. Crystallization, structure
andproperties of plasticized poly(L-lactide). Polymer
2005;46(23):10290–300.
[31] Pluta M. Morphology and properties of polylactide
modifiedby thermal treatment, filling with layered silicates
andplasticization. Polymer 2004;45(24):8239–51.
[32] Jacobsen S, Fritz HG. Plasticizing polylactide – The effect
ofdifferent plasticizers on the mechanical properties. PolymEng Sci
1999;39(7):1303–10.
[33] Pluta M, Galeski A. Plastic deformation of amor-phous
poly(L/DL-lactide). Biomacromolecules, doi:10.1021/bm061229v, in
press.
[34] Kremer F, Schoenhals A. Broadband Dielectric Spectros-copy.
Heidelberg: Springer Verlag; 2002.
[35] Noda N, Lee YH, Bur AJ, Prabhu VM, Snyder CR, RothSC, et
al. Dielectric properties of nylon 6/clay nanocom-posites from on-
line process monitoring and off-linemeasurements. Polymer
2005;46(18):7201–17.
[36] Davis RD, Bur AJ, McBrearty M, Lee YH, Gilman JW,Start PR.
Dielectric spectroscopy during extrusion process-ing of polymer
nanocomposites: a high throughput process-ing/characterization
method to measure layered silicatecontent and exfoliation. Polymer
2004;45(19):6487–93.
MA
CR
OM
OL
EC
UL
AR
NA
NO
TE
CH
NO
LO
GY
http://dx.doi.org/10.1021/bm061229vhttp://dx.doi.org/10.1021/bm061229v
Polylactide/montmorillonite nanocomposites: Structure,
dielectric, viscoelastic and thermal
propertiesIntroductionExperimentalMaterialsSample
preparationCharacterization
Results and discussionSize exclusion chromatography (SEC)XRD
analysisRheological propertiesCalorimetric
characterizationViscoelastic propertiesBroadband dielectric
spectroscopy
DiscussionConclusionsAcknowledgmentsReferences