Printer Friendly

Role of chain dynamics and topological confinements in cold crystallization of PLA-clay nanocomposites.


Polymer nanocomposites have attracted considerable interest from academic and industrial standpoints over the past decades. Nanocomposites show sizeable improvement of physical and mechanical properties compared to conventional composites, achieved by the addition of a small amount of nanoparticles [1-6]. An interphase layer is extended along the phase border of the polymer and filler due to the attractive interaction between them, and a large surface area provided by the dispersion of the nanoparticles [7, 8], The fraction of amorphous chains trapped in the interphase (entangled to the absorbed layer) exhibits restricted dynamics compared to bulk-like chains. The outstanding properties of the nanocomposites are attributed to the formation and development of the interphase [2, 7-9].

The addition of the inorganic phase, also, influences the crystallization process of the matrix [10-12], Essentially, the microstructure of semi-crystalline polymers depends on the crystallization parameters such as the degree of crystallinity, crystallization kinetics, and the morphology of the crystallites [13]. From literature [3, 14, 15], crystallization kinetics can be either enhanced or retarded dramatically by the presence of nanoparticles. The interphase can provide heterogeneous nucleating sites, mostly at low particle content, subsequently increasing the overall crystallization rate. Nevertheless, it was reported that nucleation might be hindered by the shielding effect of the organically modified nanoparticles and, also, that the growth rate is not significantly altered by the presence of nanoparticles [12, 16]. The overall crystallization rate might decrease at higher particle loading due to the topological constraints and the extension of the interphase layer. Diffusion of the crystallizing segments into the growth front is impeded by the confined segmental motion, following the formation of a percolated structure and the spreading of a bridging configuration [3, 10, 11, 17]. Recently, a nonmonotonic trend has been suggested for the self-diffusion coefficient of macromolecules by the increase of nanofillers content [4, 18].

The addition of nanoparticles can also affect the morphology of crystallites and the internal structure of the crystalline domains. Crystal growth is disturbed by the formation of morphological restrictions, at high particle content. At a lower particle loading, however, the crystallizing segments might manoeuvre through the polymer particle interphase [19].

Furthermore, a semi-crystalline polymer can be considered as a composite including a crystalline domain and an amorphous microphase with significantly different physical properties [20], Frequently, the long chains cross the crystalline phase boundary creating a broad interphase between the crystalline and amorphous phases. This interphase is formed by the mobility restricted fraction of the amorphous chains. The rigid amorphous fraction shows different molecular dynamics within the confined environment of the interphase imposed by the strong coupling, not participating in large-amplitude cooperative conformational motion of the bulk-like chains, at the glass transition temperature [21, 22]. Therefore, the rigid amorphous fraction in nanocomposites arises from the polymer-filler interphase and the crystalline-amorphous interphase, with wide conformational potential energy differences [8, 9]. However, variations in the crystallization rate can influence the chain folding mechanism, which in turn alters the fraction of the interlamellar and intralainellar tie molecules [23, 24].

Thermal analysis, including differential scanning calorimetry (DSC) and dynamic mechanical analysis (DMA) was extensively employed, to study the dynamics of polymer chains. These characterization techniques provide some information about molecular motion and conformational changes, rearrangements and reorganization of the metastable phases [9, 25-27]. The morphology of semi-crystalline polymers can be controlled by the exploration of molecular dynamics in the vicinity of the melting temperature.

In this work, we used PLA as a model polymer. This semi-crystalline polyester is recognized as a popular biopolymer produced from renewable sources. Polylactide shows promising physical and mechanical performance compared to conventional petroleum-based polymers, recently turning into an inevitable material in the packaging industry. It, also, has widespread applications in medical science and tissue engineering due to its biocompatibility and biodegradability [28-33],

PLA crystallization has been extensively investigated over the past few years. Three crystalline phases ([alpha], [beta], and [gamma]) have been observed for PLA, depending on the processing conditions and the D-stereoisomer content. [alpha]-Crystals, including two parallel chains with [10.sub.3] helical conformations in an orthorhombic unit cell, can be formed during melt or solution crystallization. Moreover, a disordered [alpha]-form ([alpha]'-crystals) has been, predominantly, obtained during cold crystallization. The crystallization rate of PLA is very high during cold crystallization. The process has been assumed to be sequential [32, 34]. The packing of the crystallizing segments and the intermolecular ordering is preceded by intramolecular conformational ordering [34, 35],

The internal structure of the crystalline domains has been, widely, investigated using small angle X-ray scattering (SAXS) [36-39], The alternate microstructure of the crystalline stacks was determined from variations of the scattering pattern caused by electron density fluctuation, originated in several molecular arrangements and intermolecular ordering along a crystalline domain.

In this study, the structural evolution and chain dynamics were investigated. For that purpose, an annealing process (for different times and at different temperatures), along with subsequent reheating (in different temperature regimes), was employed to ascertain the influence of the addition of organically modified nanoparticles on chain dynamics, and consequently crystallization parameters.

We, also, attempted to illustrate the relationship between devitrification of the rigid amorphous fraction and crystallization kinetics. Moreover, periodicity of the crystalline and amorphous layers was related to the development of rigid amorphous molecules and internal structure of the crystalline domains.



Poly(L,D-lactic acid), used in this study, is a commercial grade of PLA(4043D) with approximately 4.8 mol% D-content, provided by NatureWorks LLC. Cloisite 15A, used as the source of nanoparticles, was purchased from Southern Clay Product Inc. The organoclay was modified through a surface treatment process of montmorillonite, in the presence of quaternary ammonium salts, dimethyl dehydrogenated tallow quaternary ammonium. Hereinafter, the hybrid of PLA/Cloisite 15A is designated PLAAi, where A represents Cloisite 15A and i stands for the particle loading (wt%) of the nanoparticles.

Nanocomposite Preparation

All the materials were dried prior to all processing and measurement steps. An overnight drying procedure was applied at 70[degrees]C under reduced pressure. A master batch was prepared with a high content of the nanoparticles (10 wt%), in order to achieve a better dispersion of the layered stacks. In order to prepare the master batch, the PLA granules, and the nanoparticles were manually mixed by tumbling in a sealed plastic zip-lock bag, and, subsequently, fed into an extruder. The mixture was melt compounded into pellets via a Leistritz corotating twin-screw extruder (TSE),with the four temperature zones tuned at 180, 190, 190, 200[degrees]C, and a screw speed of 150 rpm, providing a flow rate of 3 kg/h. The outgoing melt was fed into a water bath at 20[degrees]C and pelletized. The melt compounding of the nanoparticles has been known as an economic process, which is also more environmentally friendly compared to other methods [40]. Subsequently, the nanocomposites were prepared through the melt mixing of neat PLA granules and the pre-extruded master batch. The nanocomposite with different compositions was fed by a volumetric feeder into the TSE, using the described extrusion procedure. The pure sample (PLA) went through the same procedure as that for the nanocomposites. The pellets were compression molded into thin sheets and discs at 200[degrees]C according to the procedure, which enabled obtaining a smooth surface, under a nitrogen atmosphere. The samples were held for an additional 3 min at the molding temperature after removing the pressure to eliminate any thermal and deformation history. Then, the samples were quenched in an ice-water bath at 0[degrees]C. The quenched samples are designated PLAAi(Q).


Wide angle X-ray diffraction measurements were conducted using a Philips X'PERT with a Cu K[alpha] radiation, operating at 40 kV/50 mA at room temperature. The distance between two adjacent layers in the stack of the nanoparticles were estimated based on the position of the diffraction peak over a diffraction angle (2[theta]) range of 1.5-10[degrees], using Bragg's law (incident beam wavelength, [lambda] = 1.54 [Angstrom]).


The experiments were performed using a parallel-plates geometry (R = 25 m, gap = 1 mm), on an Anton Paar MCR 501 rheometer at 190[degrees]C, under nitrogen atmosphere. All dynamic measurements were conducted within the linear viscoelastic region. A strain sweep experiment was carried out over a strain range of 0.1-100% at frequencies of 0.1 and 10 Hz to determine the linear domain. Thermal stability of the samples during the frequency sweep measurements was verified by a time sweep experiment within the linear region. The frequency sweep experiments were conducted over a frequency range of 0.1-150 rad/s to determine storage and loss moduli as well as complex viscosity.

Thermal Analysis

Nonisothermal studies were performed on a TA-Instruments; Temperature Modulated Differential Scanning Calorimeter, TMDSC Model Q1000. Several physical transitions might overlap during the heating process in a conventional DSC such as melting and recrystallization. Therefore, the temperature modulated DSC is considered a better alternative to discerning the reversing and nonreversing contributions to the heat flow [41-43], Approximately 7 mg samples were weighed and sealed in an aluminum pan and lid. The samples were equilibrated at 30[degrees]C and heated to 200[degrees]C, far above the melting point, in the standard mode at a fairly fast heating rate of 10[degrees]C/min to prevent crystallite reorganization during heating. The temperature modulated measurements were performed at an underlying heating rate of 3[degrees]C/min to allow sufficient modulation during heating. The amplitude and the period of modulation were determined to be 0.7[degrees]C and 60 s, respectively. This enables the sample, to follow the modulated heat flow, without any distortion [43].

The temperature modulated DSC can distinguish the reversing thermodynamic heat capacity and the nonreversing latent heat. The enthalpy changes derived from a variation of temperature and composition are given by Eq. 1.

dH = [(dH/dT).sub.p,n] dt + [([partial derivative][DELTA]/[partial derivative]n).sub.P,T] dn (1)

where [(dH/dT).sub.p,n] is the heat capacity at a constant pressure (p) and composition [(n).[partial derivative][DELTA]H/[partial derivative]n).sub.P,T] is the latent heat [25]. The latent heat will be zero when there is no phase transition (dn=0).The degree of Crystallinity ([X.sub.CR]) is typically calculated as follows, Eq. 2:

[X.sub.CR] = ([DELTA][H.sub.f] - [SIGMA] [DELTA][]/(1 - i)[DELTA][H.sup.0.sub.f] (2)

where [DELTA][H.sub.f], [DELTA][], [DELTA][H.sup.0.sub.f] and i are: enthalpy of fusion, enthalpy of cold crystallization, specific enthalpy of fusion of a perfect crystal (~93.6 J/g), and weight fraction of the nanoparticles, respectively [1].

The isothermal measurements were performed on the same instrument in the standard mode. The quenched samples were heated to the annealing temperatures at a rate of 60[degrees]C/min, and subsequently held for a constant period of time, allowing the trace for all samples to return to the calorimeter baseline.

These annealed samples are designated PLAAi(Qj)[T.sub.a], where j and [T.sub.a] represent the annealing time and temperature, respectively. The relative degree of crystallinity is calculated from the ratio of the heat evolved up to time t, the area under the curve, divided by the total heat of crystallization, the total area, according to Eq. 3:

[X.sub.t] = [X.sub.CR](t)/[X.sub.CR]([infinity]) = [[integral].sup.t.sub.0] (dH/dt)dt/[[integral].sup.[infinity].sub.0](dh/dt)dt (3)

Dynamic mechanical properties of the samples were investigated on a TA-Instruments Q800; a dynamic mechanical analysis instrument (DMA), using a dual-cantilever fixture. Length, width and thickness of the molded bars were 35, 11.75, and (~)1.8 mm, respectively. A multi-strain experiment was performed to find out the linear viscoelastic region at a frequency of 1 Hz, over the strain range of 1-100% at -20[degrees]C. The temperature sweep experiments were carried out over a temperature range of -60 to 120[degrees]C, with a heating rate of 3[degrees]C/min, at a frequency of 1 Hz and a strain of 1%.

Fourier Transform Infrared Spectroscopy (FTIR)

The spectra were obtained using a Perkin-Elmer (spectrum 65) Fourier Transform Infrared Spectrometer in ATR mode. The data was collected at a resolution of 2 [cm.sub.-1] and an accumulation of 16 scans at ambient temperature.


Small angle X-ray scattering patterns were collected using a Bruker AXS Nanostar, equipped with a Microfocus Copper Anode, MONTAL OPTICS and a VANTEC 2000 2D detector, located 107.2 mm away from the sample, and operating at 45 kV/0.65 mA, at room temperature. The distance was calibrated through a Silver Behenate standard, prior the measurements. Collection exposure times were 500 s per sample. The scattered intensity was integrated over a scattering angle (20) range of 0.14-2[degrees], on -20 to +20[degrees] along the stretching plane. The intensity was azimuthally integrated and a 20 of 0.20-3[degrees].


In Fig. 1, the organically modified clay showed a diffraction peak at around 2[theta] = 2.52[degrees], related to the interlayer spacing of 3.5 nm as estimated based on Bragg's law. The diffraction peak occurred at smaller angles in PLAA1.2(Q) (2[theta] = 2.31[degrees], [d.sub.001] = 3.82 nm), PLAA3.6(Q) (2[theta] = 2.23[degrees], [d.sub.001] = 3.95 nm), and PLAA6(Q) (2[theta] = 2.18[degrees], [d.sub.001] =4.04 nm), demonstrating the formation of an intercalated structure in the nanocomposites. The slightly larger interlayer spacing with the increase of the nanoparticle content does not necessarily shows a higher degree of delamination of the nanoparticle stacks. Furthermore, the higher intensity of the diffraction peak is caused by the higher concentration of the diffracting objects in PLAA3.6(Q) and PLAA6(Q), than in PLAA1.2(Q). It is noteworthy that, the second diffraction peak in the spectra observed in Fig. 1, is related to a weak overtone at 2 X (2[theta]), 3 X (2[theta]).

The storage modulus becomes nearly frequency-independent with the increase of nanoparticle loading, i, from 1.2 to 3.6 wt%, within the low-frequency region, as shown in Fig. 2a. The structure was not disturbed by the small-amplitude deformation in the linear viscoelastic region [44]. An amplitude sweep experiment was conducted to determine the linear region and to ensure that the microstructure was not notably affected by oscillatory deformation during the frequency sweep measurements.

From Fig. 2a, it can be observed that a percolation threshold, from now on referred to as the threshold, occurred at a composition between 3.6 and 1.2 wt% in the intercalated nanocomposites. The appearance of the pseudo-solid-like behavior demonstrates the formation of a polymer mediated network structure [45-47]. One can observe, from Fig. 2a that, in the high-frequency region, the pure sample showed a storage modulus higher than those for all the nanocomposites, investigated in this study. This is attributed to an accelerated hydrolytic degradation of polylactide in the presence of nanoparticles at the experiment temperature, since the high-frequency region is known as the matrix-control region [45, 48].

The plateau storage modulus shows a strong interaction between the polymer and the particles, as well as, the morphological restrictions, leading to the extension of the relaxation process [5], The relaxation spectrum (H([tau])) was determined by the software (NLREG), using the storage and the loss modulus data as a function of frequency at 190[degrees]C as illustrated in Fig. 2b.

An incomplete relaxation was triggered over a longer time scale at the experiment temperature due to the topological constraints, resulting from the formation of the polymer-mediated network structure in PLAA3.6 and PLAA6 [47], It has been illustrated that, the trapped chains in an interphase might entangle with molecules from another interphase (bridging and telebridging), rather than with bulk-like chains at higher nanoparticle content, increasing the required time of disentanglement [8, 17], A second peak appeared in the relaxation spectrum of PLAA1.2 over the long time range in Fig. 2b, accompanied with the one observed at a shorter range where the bulk-like chains responded. The secondary transition is related to the development of a restricted fraction of the bulk-like chains with an extended relaxation time.

The cold crystallization temperature ([]) of the quenched samples is shifted to higher temperatures by the addition of nanoparticles at a heating rate of 10[degrees]C/min. Meanwhile for compositions above the percolation threshold, cold crystallization is barely noticeable for the experiment time scale (inversely proportional to the heating rate), as revealed in Fig. 3a.

Attractive interaction along the interphase and the morphological constraints reduced the chain mobility and subsequently, affected the short distance diffusion of the crystallizing segments across the phase boundary. Thus, nucleation and crystallization growth are affected due to an increase of the free enthalpy of activation, as well as, the free energy of crystallization of chain segments. The free enthalpy of activation is influenced by chain mobility. It has been demonstrated that, the free enthalpy of activation becomes zero as the crystallization temperature approaches the glass transition temperature [23]. Surprisingly, the cold crystallization temperature shifted to a lower temperature for PLAA1.2(Q) compared to PLA(Q) following a decrease of the underlying heating rate from 10 to 3[degrees]C/min, as can be seen in Fig. 3b. Therefore, the diffusion rate of crystallizing segments into the growth front manifested a nonmonotonic function of the particle loading.

The structural rearrangements and the re-adjustment of chain conformation occur during the reduced heating rate, caused by the increased time scale of the experiment. Hence, a lower cold crystallization temperature was observed in PLAA1.2(Q), over the longer time scale due to the heterogeneous nucleating role played by nanoparticles at this concentration. The enhanced intramolecular interaction in PLAA1.2 provided the required conformational ordering to create helical structure prior to the formation of nuclei of a critical size. Furthermore, the cold crystallization peak of PLAA1.2(Q) is narrower than that for the other samples, due to the effect of the heterogeneous nucleation.

The broad cold crystallization peak, associated with compositions above the threshold, revealed a considerable contribution of sporadic nucleation, demonstrating a fluctuation of nuclei population during the crystallization process, in contrast to the nearly constant number of nuclei expected in heterogeneous nucleation [23, 49], Moreover, some degree of crystallization is detected during the reheating of PLAA3.6(Q) and PLAA6(Q), over the longer time scale (heating rate of 3[degrees]C/min), as a consequence of the structural evolution and reorganization of the interphase. Thus, competition between the topological confinements and the enhanced intramolecular interaction following the addition of nanoparticles is manifested.

A nonequilibrium glassy state is obtained by quenching of the polymer melt to a temperature far below the glass transition [31, 50, 51]. The quenched samples were annealed for 1 h at [T.sub.a] = 70[degrees]C ([T.sub.a] ~ [T.sub.g] + 10[degrees]C). An FT-IR spectrum investigation of the conformation sensitive bands reveals conformational rearrangements and preordering. Cold crystallization can be viewed as a multi-step process. The intramolecular conformational ordering precedes the formation of the [10.sub.3] helical structure in the crystals [34].

The bands related to helix-coil transition and the disordered amorphous phase emerged at the 921 and 956 [cm.sup.-1] wavenumbers, respectively, in FT-IR spectrum. Peak 921 [cm.sub.-1] arises from the coupling of C-C backbone, and peak 956 [cm.sub.-1] is rooted in the C[H.sub.3] rocking mode [32, 34, 35, 52, 53]. The integrated intensity of peak 1745 [cm.sup.-1] related to C=O stretching vibrations was assigned as reference. The formation of the short-range ordering is observed in Fig. 4 for PLAA1.2(Q1h)70, corresponding to the reduced intensity of the band at 956 [cm.sup.-1], and the appearance of the band at 921 [cm.sup.-1], caused by the enhanced intramolecular conformational ordering. The formation of an ordered structure is not observed for the other samples. The decreased intensity of the band at wavenumber 956 [cm.sup.-1] might be associated with the reduction of the disordered mobile amorphous segments.

Molecular ordering influences the nucleation process and subsequently the overall growth rate. The cold crystallization temperatures, ([]), as a function of nanoparticle content, for two samples, and at two heating rates is shown in Fig. 5a, where it can be seen that, the cold crystallization temperature, in the samples annealed for 1 h at [T.sub.g] + 10[degrees]C, decreases. The restricted diffusion arising from the percolated structures of PLAA3.6(Q1h)70 and PLAA6(Qlh)70 caused the time required for disentanglement of the chains to become longer; however, they showed a lower cold crystallization temperature than PLAA3.6(Q) and PLAA6(Q). It is attributed to the reorganization of the interphase within the annealing process.

The elastic modulus was enhanced above the glass transition temperature following the development of the crystalline domains. The variation of the elastic modulus, obtained from dynamic mechanical analysis, at a heating rate of 3[degrees]C/min, is illustrated in Fig. 5b. The elastic modulus upturn occurred at a lower temperature for PLAA1.2(Q) compared to the other quenched samples (PLAAi(Q)). The same behavior was observed for the samples annealed for 1 h at 70[degrees]C (PLAAi(Qlh)70) while, the modulus enhancement threshold moved to a lower temperature range. Moreover, the crystallization kinetics can be related to the rate of enhancement. A sluggish rate is observed for the quenched samples with a composition above that corresponding to percolation; however, following the annealing process, the rate is accelerated, and accompanied by a shift of the enhancement upturn to a lower temperature range. This is consistent with the behavior observed in Fig. 5a.

The rate of heat evolution of the quenched samples (PLAAi(Q1h)[T.sub.a]), during the annealing process at several temperatures, was obtained. It was found that the fastest crystallization occurs for PLA(Q1h)110 as illustrated in Fig. 6a and b. PLAA3.6(Q1h)[T.sub.a] and PLAA6(Q1h)[T.sub.a] show similar heat evolution and relative crystallinity at the same annealing temperature, as shown in Fig. 6.

A slow increase of the heat flow was observed at the end of the primary crystallization for nanocomposites compared to the pure sample. This slow crystallization reveals the presence of the secondary crystallization. The secondary crystallization develops by nuclei deposition and perpendicular row accomplishment, on a smooth layer formed during the primary one, along the helix axis. The evolution of the crystalline structure continues even after the apparent return to the baseline. It was not possible to observe this directly, due to the limited precision of the instrument. The persistence of the secondary crystallization can be inferred, however, from the difference between the degree of crystallization/heat evolution measured during the isothermal process, and the heat of fusion of subsequent nonisothermal measurements (at a heating rate of 10[degrees]C/min). The faster growth of the relative crystallinity was observed for PLAA1.2(Q1h)130 compared to PLA(Q1h)130, Fig. 6c and d. In contrast during annealing at 110[degrees]C, PLA(Qlh)110 exhibited faster crystallization.

The crystallization half-life and the Avrami parameters were calculated, in order to evaluate the isothermal cold crystallization quantitatively. The Avrami equation is expressed as:

[X.sub.t]/[X.sub.[infinity]] = 1 -exp(-[Zt.sup.[??]]) (4)

where, [X.sub.t] and [X.sub.[infinity]] are the relative fractional extent of crystallinity, at time t and at the end of heat evolution, respectively.Z is the composite rate constant and n is the Avrami exponent, n is often considered to be characteristic of the nucleation type and the crystal growth geometry. The simultaneous or sporadic nature of the nucleation process and the growth dimension of the expanding waves can be related to the value of the rate constant. Therefore, the Avrami equation is typically used to provide kinetic aspects of the crystallization process. Nonetheless, it does not provide a lucid insight into the molecular organization of the crystallites [13, 23, 54].

The crystallization half-life, [t.sub.1/2], is defined as the time required to reach half of the final degree of crystallization calculated by [(ln 2/Z).sup.1/n]. It is assumed that in cold crystallization, the overall crystallization rate G, is proportional to 1/[t.sub.1/2] [33], In Fig. 7a, the overall crystallization rate was observed to be considerably low for PLAA3.6(Q1h)[T.sub.a] and PLAA6(Q1h)[T.sub.a] due to the morphological confinement, as a consequence of the network formation. Thus, the overall crystallization rate does not show any noticeable dependence on particle loading above the composition corresponding to the percolation threshold.

The degree of crystallinity, [X.sub.CR], goes up moderately upon the increase of the annealing temperature, as shown in Fig. 7b. The effect of the addition of nanoparticles on the degree of crystallinity for the crystallization of the nanocomposites at 90[degrees]C manifests itself through the increased conformational free energy. The degree of crystallinity of PLAA3.6(Q1h)[T.sub.a] is similar to that of PLAA6(Q1h)[T.sub.a], within the crystallization period, at different annealing temperatures (90[degrees]C[less than or equal to][T.sub.a][less than or equal to]130[degrees]C).

The mobile amorphous chains participate in a long-range coordinated molecular motion at the glass transition temperature. Essentially, the solid fraction (SF), consisting of the crystalline domains and other glassy matter, is not devitrified at this transition temperature. Therefore, the mobile amorphous fraction ([A.sub.MA]) is proportional to the increase of heat flow, and heat capacity, at [T.sub.g]. The increase in heat flow, [DELTA]H, is related to the solid fraction enthalpy, where the cooperative conformational change starts, in addition to the minor contribution of the vibrational motion. The mobile amorphous fraction was quantified using Eq. 5, where [DELTA][H.sub.rev] is the reversing heat flow variation of the samples, and [DELTA][H.sub.rev,am] is the reversing heat flow variation of a totally amorphous sample, at [T.sub.g].

[X.sub.MA] = [DELTA][H.sub.rev]/[DELTA][] (5)

PLA(Q) is considered to be a totally amorphous sample, in this study. The absence of crystallites was verified beforehand using the DSC and WAXS techniques (data are not shown). Therefore, the mobile amorphous fraction of PLA(Q) is considered equal to 1.

The broadening of the glass transition range is explained by the existence of the restricted-mobility molecules, in the polymerparticle and crystal-amorphous interphases. Furthermore, the rigid fraction of the amorphous phase has no noticeable contribution to the energy dissipation through the cooperative conformational motion which takes place at the glass transition temperature.

The variation of the loss factor can be used to estimate [X.sub.MA], applying Eq. 6, where (tan[delta])/[(tan [delta])] is the relative loss factor and [(tan[delta])] is the loss tangent of a totally amorphous sample (PLA(Q)). The relative loss factor was averaged over a temperature range related to the occurrence of the glass transition. The loss factor variations were shown in Fig. 8.

[X.sub.MA] = (tan [delta])/[(tan [delta])] (6)

It has been demonstrated [9, 24] that, not the entire system can be, adequately, described by the conventional two-phase model, which considers only the amorphous fraction and the crystalline fraction ([X.sub.MA] + [X.sub.CR] <1). The three-phase model introduces another constituent, the so-called rigid amorphous nanophase, as a portion of the solid phase, in addition to the mobile amorphous fraction ([X.sub.RA] = 1 - ([X.sub.MA] + [X.sub.C])).

The fraction of the rigid amorphous chain ([X.sub.RA]) estimated from the DMA data, Fig. 9b, was found to be larger than the ones calculated based on the DSC data, Fig. 9a. Both sets of data, however, illustrate a similar behavior. This discrepancy is, probably, due to the dynamic nature of the loss factor measurement.

Theoretically, the rigid amorphous fraction. [X.sub.RA], is extended by an increase in particle loading. The degree by which it will be extended depends on the surface area of stacks. A larger interphase would be provided by the increase of the nanoparticle content employing the same mixing process, where agglomeration of the stacks might take place above a certain concentration, disturbing the monotonie relationship of the contact surface and particle loading.

A significant development of [X.sub.RA] was found for PLAA1.2(Q1h)70 due to the formation of short-range ordering. A short-range ordering was observed in PLAA1.2(Q1h)70 due to the enhanced intrachain interaction. The predominant effect of enhanced intrachain interaction leads to an unexpected increase of RAF compared to that in other samples annealed at the same conditions. It can, also, be ascribed to the appearance of a cohesive region in PLAA1.2(Q) due to molecular ordering [51, 55].

From Fig. 7a, it can be seen that the overall crystallization rate increases as a function of the annealing temperature up to a certain maximum value. At that point, the trend no longer holds due to the growing entropy barrier of the uncrystallizing segments. The overall crystallization rate is significantly higher for PLAA1.2(Q1h)[T.sub.a] compared to the other nanocomposites, at the same annealing temperature as shown in Fig. 7a. This is particularly true for temperatures above the critical point (120[degrees]C), where the rate becomes even faster than that for PLA(Q1h)[T.sub.a].

Therefore, it is speculated that, the nanoparticles begin to act as nucleating agents, at the annealing temperature above 120[degrees]C, for this particle loading (i = 1.2 wt%), raising the overall crystallization rate. It can be assumed that a portion of the rigid amorphous chains, trapped in the polymer-particle interphase, is devitrified above a critical temperature, increasing the nucleation rate. The glass transition of the rigid amorphous fraction may occur at a temperature higher than [T.sub.g], between [T.sub.g] and [T.sub.m], or even above [T.sub.m] [25], The same behavior was not observed for PLAA3.6(Q1h)120 and PLAA6(Q1h)120 which demonstrates that the devitrification of RAF might occur at a higher temperature in these samples. The appearance of the pseudo-solid-like behavior illustrated the incomplete relaxation in the molten state at the time scale of the experiment, as was shown in Fig. 2b.

The RAF increases considerably by the growing crystalline domains, during annealing at [T.sub.a] (90[degrees]C [less than or equal to] [T.sub.a] [less than or equal to] 130[degrees]C), due to the development of the crystalline-amorphous interphase. It is hard to distinguish the contribution of the rigid amorphous chains trapped in the crystalline-amorphous interphase and the polymernanoparticle interphase. RAF decreased for PLA(Qlh)130 compared to PLA(Q1h)110 while; the former has a higher degree of crystallinity, Fig. 7b. This is interpreted by the higher crystallization rate of PLA(Q1h)110 than for PLA(Q1h)130, as can be noticed in Fig. 7a. A higher rate of crystallization increases the probability that long chains might cross the crystalline phase boundary in the course of cold crystallization [23, 56] and consequently, increase the contribution of interlamellar and intralamellar tie molecules.

Furthermore, RAF increased in PLAA1.2(Q1h)130 compared to PLAA1.2(Q1h)110 as a consequence of the enhancement of the crystallization rate, which in turn is due to the heterogeneous nucleating effect of the nanoparticles over this temperature range. Despite the fraction of the amorphous chains restricted in the polymer-nanoparticle interphase, PLAA3.6(Q1h)[T.sub.a] and PLAA6(Q1h)[T.sub.a] showed lower RAF compared to PLA(Q1h)[T.sub.a] and PLAA1.2(Q1h)[T.sub.a], as illustrated in Fig. 9, due to the lower fraction of the crystalline domains (90[degrees]C [less than or equal to] [T.sub.a] [less than or equal to] 130[degrees]C). This decrease in RAF results from the lower overall crystallization rate due to the topological constraints above the percolation threshold. Consequently, the fraction of the rigid amorphous chains increases with the increase of the overall crystallization rate in this temperature range.

A correlation function was applied to the integrated intensity of SAXS data to investigate the electron density fluctuations at a correlation distance (z).Thus, the correlation function can reveal the changes in the internal structure of crystalline domains, triggered by variations of the long period, which is influenced by the lateral dimension of the folded stem, the extent of the amorphous layer, and the crystal-amorphous interphase [36-39], The one-dimensional correlation function ([[gamma].sub.1](z)) is described by Eq. 7 where, q and l(q) stand for the scattering vector (=4[pi]Sin0/[lambda]) and the integrated intensity, respectively. The scattering invariant (Q) was estimated over the experimentally accessible range of the scattering vector as expressed in Eq. 8. In Eq. 8, [q.sub.1] are a scattering vector related to the first obtained reliable data and [q.sub.2] is assigned to a scattering vector where l(q) levels off. The scattering data were collected within a limited range of the scattering vector. Therefore, the data were extended to a large q range using Porod-Ruland model (I=[K.sub.p][q.sup.-4]exp(-[[sigma].sup.2][q.sup.2])), where [K.sub.p] is the Porod's constant and [sigma] is related to the interphase thickness. The data were, also, fitted to a low q range making use of the Guinier approximation, I = [I.sub.0]exp (- [1/3] [R.sub.g.sup.2][q.sup.2]), where [I.sub.0] and [R.sub.g] are the forward intensity and the radius of gyration, respectively. It is noteworthy that, the scattering spectra of PLAAi(Q1h)[T.sub.a] were subtracted by the spectra of PLAAi(Q) as a background.



The correlation function was estimated for PLA(Q1h)[T.sub.a] and PLAA1.2(Q1h)[T.sub.a] ([T.sub.a] = 110, 130[degrees]C), as illustrated in Fig. 10. The long period (L) was estimated along the correlation distance where, the first maximum of [[gamma].sub.1](z) occurred. Furthermore, the lamellae thickness ([l.sub.c]) was determined by the interception point of the tangent line at the first minimum of the correlation function and the extended linear portion of [[gamma].sub.1](z) over the low correlation distance range. The calculated thickness of the crystalline lamellae ([l.sub.c]) and the amorphous layer sandwiched in between ([l.sub.a] = L - [l.sub.c]) are shown in Table 1.

The lamellae thickness, [l.sub.c], which grew with the increase of the annealing temperature from 110 to 130[degrees]C, showed a slight change with the addition of the nanoparticles in PLAA1.2(Q1)[T.sub.a]. From Table 1, a lower thickness of the amorphous layer can be observed for PLA(Q1h)110 than for PLAA1.2(Q1h)110, as well as, for PLAA1.2(Q1h)130 than for PLA(Q1h)130. This might be attributed to the larger contribution of RAF, developed in those samples at the crystalline-amorphous interphase, and the larger density of the rigid fraction of the amorphous chains compared to the bulk-like chains [37]. The restricted-mobility of the rigid amorphous segments influence the ratio of the population of trans to gauche conformed through the widening of the potential energy difference of the conformers, as a result of the lengthened relaxation process of RAF compared to that of bulk-like chains. This is consistent with the variation of the heat flow at the glass transition temperature (measured by DSC), and also with the influence of the development of the crystalline-amorphous interphase on energy dissipation (estimated by DMA), as revealed in Fig. 9.


Temperature modulated differential scanning calorimetry (TMDSC) and dynamic mechanical analysis (DMA) were used to investigate the molecular dynamics of PLA/Cloisite 15A nanocomposites, during annealing, at different times and temperatures. Rheological measurements were also conducted to study the chain dynamics in the molten state.

A pseudo-solid-like behavior was observed during the investigation of the storage modulus over the linear region accompanied by an incomplete relaxation, indicating the formation of topological constraints in compositions above the percolation threshold. The development of a polymer-mediated network structure leads to a higher free enthalpy of activation, and a higher energy of crystallization of chain segments, as a result of the addition of nanoparticles, which introduce topological constraints and restricted chain mobility. This can be attributed to the reduction of the diffusion rate of the crystallizing segments across the phase boundary, thus hindering the nucleation and growth rates.

On the other hand, in the absence of a percolated structure, the nanoparticles act as heterogeneous nucleating sites, at an extended crystallization time scale (the decrease of the underlying heating rate from 10 to 3[degrees]C/min). The increased time scale causes structural rearrangements of the trapped chains in the polymer-particle interphase and a predominant effect of the enhanced intrachain interactions, upon the addition of the nanoparticles. Therefore, we propose that there are two competing effects at play, the enhanced intrachain interaction, and the restricted mobility. Which effect will be dominant depends on whether devitrification of the rigid amorphous chains takes place, during the allotted time.

The overall growth rate was evaluated by the crystallization half-life, during heat evolution, within the annealing process. A lower overall growth rate was observed for the nanocomposites compared to that of the neat polymer. However, devitrification of the rigid fraction of the amorphous chains, trapped in the polymer-particle interphase, above a critical annealing temperature, was observed to lead to improved kinetics of isothermal crystallization, for compositions below the percolation threshold. The critical temperature is related to the glass transition temperature of the rigid amorphous nanophase. The overall crystallization rate did not show a remarkable dependence on the particle loading during annealing due to the formation of network structures, which leads to a restricted morphology. Thus, it can be concluded that the nanoparticles play a heterogeneous nucleating role, where the rigid amorphous fraction trapped in the polymer-particle interphase shows enough mobility above its glass transition, at the annealing temperature or time scale of nonisothermal crystallization.

A rigid amorphous fraction develops with the addition of nanoparticle. This behavior, however, levels off with the increase of the nanoparticle content above the percolation threshold, due to the formation of aggregates. RAF can, also, be extended following the formation of a cohesive region. The cohesive region is a result of short-range conformational ordering, derived from the enhanced intramolecular interaction.

The fraction of the rigid amorphous chains is proportional to the degree of crystallinity and the overall crystallization rate, with the formation of the crystalline domains. It is concluded that, a higher crystallization rate leads to a larger RAF and a more dense amorphous layer, spread between periodically stacked lamellar structures, for equal crystalline phase content. It was found that the lateral dimension of the folded stem did not change noticeably with the nanoparticle content since the nanoparticles were excluded from the crystalline domains.


[1.] V. Krikorian, and D.J. Pochan, Macromolecules, 37, 17 (2004).

[2.] R.K. Bharadwaj, A.R. Mehrabi, C. Hamilton, C. Trujillo, M. Murga, R. Fan, A. Chavira, and A.K. Thompson, Polymer, 43, 13 (2002).

[3.] E. Di Maio, S. Iannace, L. Sorrentino, and L. Nicolais, Polymer, 45, 26 (2004).

[4.] A. Kutvonen, G. Rossi, and T. Ala-Nissila, Phys. Rev. E, 85, 4 (2012).

[5.] W. Xu, S. Raychowdhury, D.D. Jiang, H. Retsos, and E.P. Giannelis, Small, 4, 5 (2008).

[6.] A. Fujimori, N. Ninomiya, and T. Masuko, Polym. Eng. Sci., 48, 6 (2008).

[7.] D. Ciprari, K. Jacob, and R. Tannenbaum, Macromolecules, 39, 19 (2006).

[8.] Y.Q. Rao and J.M. Pochan, Macromolecules, 40, 2 (2007).

[9.] A. Wurm, M. Ismail, B. Kretzschmar, D. Pospiech, and C. Schick, Macromolecules, 43, 3 (2010).

[10.] T.D. Fornes and D.R. Paul, Polymer, 44, 14 (2003).

[11.] A. Somwangthanaroj, E.C. Lee, and M.J. Solomon, Macromolecules, 36, 7 (2003).

[12.] P.H. Nam, N. Ninomiya, A. Fujimori, and T. Masuko, Polym. Eng. Sci., 46, 1 (2006).

[13.] X.F. Lu and J.N. Hay, Polymer, 42, 23 (2001).

[14.] P. Maiti, P.H. Nam, M. Okamoto, N. Hasegawa, and A. Usuki, Macromolecules, 35, 6 (2002).

[15.] J.Y. Nam, S.S. Ray, and M. Okamoto, Macromolecules, 36, 19 (2003).

[16.] P.H. Nam, A. Fujimori, and T. Masuko, J. Appl. Polym. Sci., 93, 6 (2004).

[17.] J.B. Hooper and K.S. Schweizer, Macromolecules, 39, 15 (2006).

[18.] M.F. Mu, M.E. Seitz, N. Clarke, R.J. Composto, and K.I. Winey, Macromolecules, 44, 2 (2011).

[19.] J. Khan, S.E. Harton, P. Akcora, B.C. Benicewicz, and S.K. Kumar, Macromolecules, 42, 15 (2009).

[20.] J.A. Kornfield, G. Kumaraswamy, and A.M. Issaian, Indus. Eng. Client. Res., 41, 25 (2002).

[21.] E. Zuza, J.M. Ugartemendia, A. Lopez, E. Meaurio, A. Lejardi, and J.R. Sarasua, Polymer, 49, 20 (2008).

[22.] Y.M. Wang and J.F. Mano, Eur. Polym. J., 41, 10 (2005).

[23.] B. Wunderlich, Macromolecular Physics, Academie Press, New York (1976).

[24.] M.L. Di Lorenzo, M.C. Righetti, and B. Wunderlich, Macromolecules, 42, 23 (2009).

[25.] B. Wunderlich, Progr. Polym. Sci., 28, 3 (2003).

[26.] R. Androsch and B. Wunderlich, Polymer, 46, 26 (2005).

[27.] J. Pak, M. Pyda, and B. Wunderlich, Macromolecules, 36, 2 (2003).

[28.] J.M. Zhang, H. Tsuji, I. Noda, and Y. Ozaki, Macromol ecu!es, 37, 17 (2004).

[29.] P. Pan, W. Kai, B. Zhu, T. Dong, and Y. Inoue, Macromolecules, 40, 19 (2007).

[30.] P. Pan, Z. Liang, B. Zhu, T. Dong, and Y. Inoue, Macromolecules, 41, 21 (2008).

[31.] P.J. Pan, B. Zhu, and Y. Inoue, Macromolecules, 40, 26 (2007).

[32.] J.M. Zhang, Y.X. Duan, H. Sato, H. Tsuji, I. Noda, S. Yan, and Y. Ozaki, Macromolecules, 38. 19 (2005).

[33.] T. Kawai, N. Rahman, G. Matsuba, K. Nishida, T. Kanaya, M. Nakano, H. Okamoto, J. Kawada, A. Usuki, N. Honma, K. Nakajima, and M. Matsuda, Macromolecules, 40, 26 (2007).

[34.] B. Na, N.N. Tian, R.H. Lv, ZJ. Li, W.F. Xu, and Q. Fu, Polymer, 51, 2 (2010).

[35.] B. Na, R.H. Lv, S.F. Zou, Z.J. Li, N.N. Tian, and Q. Fu, Macromolecules, 43, 4 (2010).

[36.] E.L. Heeley, T. Gough, DJ. Hughes. W. Bras, J. Rieger, and A.J. Ryan, Polymer, 54, 24 (2013).

[37.] P.D. Hong, W.T. Chuang, W.J. Yeh, and T.L. Lin, Polymer, 43, 25 (2002).

[38.] E.L. Heeley, A.V. Maidens, P.D. Olmsted, W. Bras, LP. Dolbnya, J.P.A. Fairclough, NJ. Terrill, and A.J. Ryan, Macromolecules, 36, 10 (2003).

[39.] E.L. Heeley, T. Gough, D.J. Hughes, W. Bras, J. Rieger, and A.J. Ryan, Macromolecules, 42, 12 (2009).

[40.] V. Ojijo and S. Sinha Ray, Progr. Polym. Sci., 38, 10 (2013).

[41.] M. Salmeron Sanchez, J.L. Gomez Ribelles, F. Hernandez Sanchez, and J.F. Mano, Thermochim. Acta, 430, 1 (2005).

[42.] J. Qiu, Z.G. Wang, L. Yang, J.C. Zhao, Y.H. Niu, and B.S. Hsiao, Polymer, 48, 23 (2007).

[43.] S. Solarski, M. Ferreira, and E. Devaux, Polymer, 46, 25 (2005).

[44.] R.G. Larson, The Structure and Rheology of Complex Fluids, Oxford University Press, New York (1999).

[45.] J. Vermant, S. Ceccia, M.K. Dolgovskij, P.L. Maffettone, and

C. W. Macosko, J. Rheol., 51, 3 (2007).

[46.] MJ. Solomon, A.S. Almusallam, K.F. Seefeldt, A. Somwangthanaroj, and P. Varadan, Macromolecules, 34, 6 (2001).

[47.] K. Wang, S. Liang, J.N. Deng, H. Yang, Q. Zhang, Q. Fu, X. Dong, D.J. Wang, and C.C. Han, Polymer, 47, 20 (2006).

[48.] K. Fukushima. C. Abbate, D. Tabuani, M. Gennari, and G. Camino, Polym. Degrad. Stahil.. 94, 10 (2009).

[49.] L. Jiang, M.P. Wolcott, and J.W. Zhang, Biomacromolecules, 7, 1 (2006).

[50.] Y. Wang, J.L.G. Ribelles, M.S. Sanchez, and J.F. Mano, Macromolecules, 38, 4712 (2005).

[51.] E.A. McGonigle, J.H. Daly, S. Gallagher, S.D. Jenkins, JJ. Liggat, I. Olsson, and R.A. Pethrick, Polymer, 40, 17 (1999).

[52.] V. Krikorian and D.J. Pochan, Macromolecules, 38, 15 (2005).

[53.] S.H. Kang, S.L. Hsu, H.D. Stidham, P.B. Smith, M.A. Leugers, and X.Z. Yang, Macromolecules, 34 13 (2001).

[54.] L.H. Sperling, Introduction to Physical Polymer Science, Wiley, Hoboken, New Jersey (2006).

[55.] R.Y. Qian, L.H. Wu, D.Y. Shen, D.H. Napper, R.A. Mann, and D. F. Sangster, Macromolecules, 26, 11 (1993).

[56.] U.W. Gedde, Polymer Physics, Chapman and Hall, London, UK (1995).

Shahir Karami, Pierre G. Lafleur

Chemical Engineering Department, CREPEC, Ecole Polytechnique de Montreal, C.P 6079, Suce. Centre ville Montreal, Quebec, Canada H3C 3A7

Correspondence to: Pierre Lafleur; e-mail: Contract grant sponsor: CREPEC, Natural Sciences and Engineering Research Council of Canada (NSERC).

DOI 10.1002/pen.24070

Published online in Wiley Online Library (

TABLE 1. Thickness of the crystalline lamellae ([l.sub.c])
and the amorphous layer ([l.sub.a]), calculated based
on the correlation function.

                     ([l.sub.c])   ([l.sub.a])
                        (nm)          (nm)

PLA(Q1h)110             5.72          12.3
PLAA1.2(Q1h)110         5.58          13.83
PLA(Q1h)130             6.42          13.51
PLAA1.2(Q1h)130         6.21          13.14
COPYRIGHT 2015 Society of Plastics Engineers, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2015 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Karami, Shahir; Lafleur, Pierre G.
Publication:Polymer Engineering and Science
Article Type:Report
Date:Jun 1, 2015
Previous Article:Fracture of unsaturated polyester and the limitation of layered silicates.
Next Article:Preparation and characterization of biodegradable polymer blends from poly(3-hydroxybutyrate)/poly(vinyl acetate)-modified corn starch.

Terms of use | Privacy policy | Copyright © 2021 Farlex, Inc. | Feedback | For webmasters