Improvement of Blown Film Extrusion of Poly(Lactic Acid): Structure-Processing-Properties Relationships.
Poly(laclic acid) (PLA) is a promising and environ-mentally friendly polymer for uses in applications requir-ing short-life spans such as biomedical, packaging, andagricultural fields (1). PLA possesses many physicalcharacteristics that renders it suitable as a replacement for commodity polymers (2). The subject is vast, and therefore, it is a daunting task to summarize the remarkably rich and mutiifaceted area of litis field. Besides being biocompatible and available from renewable resources. PLA also demonstrates a high strength, thermoplastic process-ability, good grease, and oil resistance and is an excellent aroma barrier (3-6). However, the brittle behavior of this polyester reduces its use lo very limited applications. PLA also presents low kinetics of crystallization (7). To overcome these drawbacks, much work has been focused on modifying the polymer, mainly through branching reactions, polymer blending, and charge filling (8-10). Some of these studies have involved blending PLA with plasticizers or nucleanis for biomedical applications (11-13) as well as for film applications (14-18).
The first section of this introduction is dedicated to the blown film extrusion process, which is envisioned as an important processing technique lo produce most PLA thin films. The literature regarding the blown processing of PLA is very poor. In blown processing, the molten polymer is extruded al a constant flow rale through an annular die. The film is then defonned axially by the tension applied by the lake-up device and circumfcrentially by the introduction of air inside the polymer lube. For polymers, in general, a stable bubble is required not only for the continuous operation of the process but also for the production of an acceptable film wilh homogeneous thickness and exempt of defect (19). Any restrictions to achieving stable operating conditions limit the rate of production and, because of process/physical property inlerac tions, restrict the attainable range of physical properties (20).
Instabilities with regard to film blowing have previously been reported by Ast (21), Han and Park (22). Han et al. (22, 23) observed that lowering the extrusion temperaiure could stabilize the blown bubble for high-density polyethylene (PE) and low-density PE (LDPE). White and coworkers (24-26) also investigated the blown extrusion of these polymers. By comparing ihe bubble stability of different types of PE. Minoshima and White (25) concluded ihat long-chain branched PEs are the most stahle. The most unstable are PEs with narrower molecular weight distributions (MWD) (27). However, there are no articles dedicated to the blown extrusion processing of PLA and the improvement of its stability window.
Furthemore, synthesizing linear PLA from conventional processes, such as azeotropic polycondensalion or ring-opening polymerization from laclide, does not confer on il the ability to be processed by blown extrusion. When it comes to processing. PLA presents a poor melt strength thai prevents it from being used in specific processes such as film blowing, sheet foaming, or biorienled film extrusion. This specificity, in addition to melt degradation problems (28, 29) arising from PLA's polyester nature, puts studies of this material apart from those of "classical" poly mers. To enlarge its processing window and consequently its range of applications, PLA needs to be melt-strengthened. Some interesting articles have been dedicated lo the reactive exlrusion of PLA (9). The following section briefly describes such chemical modifications and insights in the observed rheological improvements.
PLA possess two distinct reactive functional end-chains, i.e., hydroxyl and carboxyl. With the help of specific reactants. PLA can be converted to a fully hydroxy-lated or carboxylated polyester. Such selective fundionalization is performed to increase the reactivity of specific chain extenders or branehers known to preferentially read wilh hydroxyl or carboxyl groups. Di et al. used the hydroxylation route by sequentially reacting 1,4-butanediol (BD) with PLA before adding 1,4-bulane diisocyanate as a chain extender (30). In the same way, Kylma and Scppala reacted a Pol yCapro lac lone (PLA-PCL) copolymer with BD and then performed chain extension with hexamethylene diisocyanate, which resulted in a (polyester-urethane) elastomer (31). Furthermore. Zhou et al. obtained a dicarboxylaled poly-l-lactic-acid (PLLA) as a result of the coupling of PLLA oligomers with succinic anhydride. A diepoxide based on the diglycidyl ether of bisphenol A was then used as a chain extender (9). Without any carboxylation step, Mihai et al. (32) and Pilla et al. (33, 34) used a commercial master batch containing multifunctional epoxide (Cesa[R] -Extend from Clariant) as a branching agent to enhance the loam ability of PLA. Dorgan et al. synthesized four- and six-branched PLAs by initialing polymerization with tetrol and hexol. respectively (8).
All of these reactions increased PLA's molecular weight. Control of the extension/branching ratio was possible by varying the contenl and ratio of additives, as well as the manner in which they were introduced (i.e.. through either a successive or a simultaneous reaction). Electron beam irradiation was also carried out on PLA blended with glycidyl meihacrylate (GMA). The results showed an efficient branching effect while the same irra dialion on pure PLA degraded ihe polymer (35).
The use of an epoxy-functionalized chain extender was recently reported to give rise to branched structures, significantly improving material properties while widening the processing window of PLA (32-34). The coupling reaction between PLA and epoxy groups was clearly evidenced by Zhou et al. (9). Mihai et al (32) demonstrated that the presence of crystalline nuclei formed before the die exit seemed to play a significant role in foam nuclealion influenced by the chain branching. In a manner similar to when the molar mass is increased, long-chain branching (LCB) accents shear sensitivity while zero shear viscosity and elongalional viscosities are improved as demonstrated by Al-Ilry et al. (29) and Corre et al. (36, 37). The broader relaxation spectra induced by chain branching has been found to influence the elongation melt properties. Thus, a higher entanglement level was assumed to hinder chain orientation during mechanical stretching. Such experiments have been conducted by Mihai et al. on linear and branched PLAs reacted wilh various amounts of a multifunctional epoxide during reactive extrusions (32). It is obvious that the elongation viscosity would increase when going from linear to highly branched PLAs. When the chain extender contenl was increased, this ratio gradually diverged. The major finding from these experiments was the relevant strain hardening that appeared for branched PLAs. This resistance to stretching became stronger as the chain extender contenl was increased.
Conversely, the kinetics of crystallization of neat PLA are also very slow. Alter being processed. PLA parts show a crystalline amouni close to zero. For most industrial applications, PLA products must have a developed crystalline phase to ensure acceptable thennal behavior for handling, transportation, and distribution. Numerous Studies have been dedicated to achieving this purpose, for instance by employing stereocomplexes for crystallization. This was a promising method (38, 39), but it is nol yet industrially viable. Li et al. (40) reported TALC as a good nucleating agent for PLA. The authors have also shown that the use of plasticizers such as polyethylene glycol (PEG) or citrate esters can enhance the kinetics of crystallization for PLA (15, 18). Some other studies (40, 41) have reported on the use of aliphatic amides as good nucleating agents for PLA. Li and Huneault (42) showed that trans-crystallization occurred at the interface between amide and PLA. Nevertheless, these investigations indicated that the crystallization half time was not drastically reduced and thus that the enhancement of the crystallization kinetics of PLA remains a challenging domain in both academic and industrial research.
As briefly summarized above, significant progress over the past decade has been made when il comes to understanding chain extension phenomenon and improving PLA crystallization. However, to the best of our knowledge, few research efforts have been devoted to fundamental and experimental aspects of this subject, and no articles have been dedicated to evaluating the effects of PLA modification on the stability during blown extrusion. Despite the interesting nature of this kind of research, it is of no help when attempting to comprehend either the chain extension or the branching and nuclealion effect on the stability process. The effect of these modifications on the induced crystallization during blown extrusion of neat and modified PLA is still an open issue. Crystallization development of the films will also drastically influence their mechanical and barrier properties.
This study has aimed to obtain a better understanding of the effects that chain extension/branching has on flow stability during blown extrusion of various grades of modified PLA. In addition, the rheological shear-induced crystallization during processing and its influence on the mechanical and ihertnomechanical properties of the obtained blown films with various amouni of mullifunc-lionalized epoxy. nucleating agents, and plasticizer have also been investigated.
Materials and Methods
PLA from Ingeo[R] NatureWorks[R] grades 2002D, 3051D, 4032D, and 4060D were used. An epoxy-funclionalized, Food and Drug Administration-approved and commercially available chain extender under the name Joneryl[R] ADR4368 was kindly provided by BASF[R] (an average functionality of 9 was calculated). TALC (powder Less than 10 [mu]m),N,N'-ethylenbis (stearamide), and PEG ([M.sub.w] = 1450 g.[mol.sup.-1]) were purchased from Sigma-Aldrich. TALC and N,N'-ethylenbis (stearamide) were used as nucleating agents and PEG as the plasticizer. N,N'-ethylenbis (stearamide) is here after referred to as EBSA.
Small Amplitude Oscillatory Shear (SAOS) Rhcology in the Linear Viscoelastic Regime. Melt rheology under oscillatory shear was investigated with an ARES rheometer from rheometrics and the stress-controlled rheometer DHR2 from TA instmments. A parallel-plate geometry (O = 25 mm, gap = 1 mm) was selected, and the experiments were performed under dynamic time and frequency sweeps in the linear viscoelastic regime at angular frequencies ranging from 0.1 to 100 [rad.s.sup.-1]. The investigations were carried out at various temperatures under a continuous nitrogen purge.
Capillary Flow Rheology. A pressure-controlled CEAST capillary rheometer (Smart Rheo 2000 twin) has been used to investigate the melt flow properties. Experiments were carried out at 180[degrees]C with three 180[degrees] entry angle dies having 1-mm diameters and respective length/diameter ratios of 30, 20. and 5.
Elongation Rheology. A "sentmanal extensional rheometer (SER)" coupled to the stress-controlled rheomcler DHR2 from TA instruments was employed to study the transient elongation properties of the neat and modified PLA. Rectangular plates (60 mm * 10 mm * 2 mm) were clamped between two counter-rotating pairs of upper/lower belts. The temperature in the healed chamber was raised to 160[degrees]C before starting ihe experiments and kept under nitrogen to avoid degradation. Because the sample's cross-section becomes reduced during stretching, a video recorder was used to monitor this variation. The sample's section should decrease exponentially with ihe time when the strain rate is kepi constant. For all the rale strain deformations with the SER geometry, we experimentally validated lhal the strain rale was constant. thanks to a camera installed in the oven enabling to measure the real dimensions of the sample. Because clamping is made between iwo moving bells on each side of ihe sample, some slippage can occur and give rise to nonho-mogeneous strain. Monitoring of ihe section rendered ii possible to check for a constant strain rate and to determine the real strain rate applied to the polymer.
Size Exclusion Chromatography
Size exclusion chromatography (SEC) analyses were perfonned in THF (0.5 [ml.min.sup.-1]) at room temperature, using a VARIAN Prostar equipped with a RHEODYNE injector. The latter was made of two Mixed-PL gel columns (G4000 HXL-G1000 HXL) having a porosity range of about 50-100,000 A, and a RI-101 refractive index de-tector. Monodisperse polystyrene (309-944,000 g.[mol.sup.-1]) was used as calibration standards. The PLA solutions were prepared al 5 mg.[ml.sup.-1] in the tetrahydrofuran (THF) and approximately 50 [mu]l of sample was injected into the chromatograph after first having passed through a 0.4-[mu]m filler.
Differential Scanning Calorintetry (DSC)
The thennal properties of the neat and modified PLA and the processed films were characterized using a differential scanning calorimeter model Q10 from TA Instruments. The DSC cell was constantly purged with nitrogen at a flow rate of 50 ml.[min.sup.-1]. Two following procedures were utilized:
1. Nonisothermal Crystallization It was performed under heating/cooling at rates ranging from 5[degrees]C.[min.sup.-1] to 50[degrees]C.[min.sup.-1].
2. Isothennal Crystallization It was carried oul at temperatures ranging from 85[degrees]C to 125[degrees]C. First, the material's thermal history was erased with an isothennal treatment lasting 5 min at 200[degrees]C. The crystallization temperature was then reached rapidly and maintained until ihe crystallization was completed. All data were processed according to Avrami's model to extract the crystallization halt' lime for each characterized sample.
Determination of the Samples' crystallinity
Differential scanning calorimetry was also perfonned to evaluate ihe crystallinity amounts after processing, using a firsi heating ramp at 10[degrees]C.[min.sup.-1] from 25[degrees]C to 200[degrees]C.
Reactive Extrusion Processing
A PRISM TTW co-rotating twin-screw extruder wilh a screw diameter of 16 mm (Thermo Electron Polylab System Rhecord RC400P) was used. The twin screw extruder had a clam shell barrel design with a lengih:diametcr ralio of 25:1. A special twin screw profile was specifically chosen for ihis study (Fig. 1). Before compounding, both PLA pellets and Joncryl were dried under vacuum at 60[degrees]C and 30[degrees]C. respectively, for 24 h to remove mois-lure. Tun processing strategies were performed:
1. Extrusion using a strand die lor neal polymers and ihe modified PLA blended with various amount of Joncryl. The compounding steps before blowing extnision were carried out under nitrogen with a screw speed sci at 40 rpm (corresponding to 3 min of residence lime). The lemperaiure profile in ihe various zones was fixed at 140[degrees]C. 190[degrees]C. 200[degrees]C, and 190[degrees]C. The correlation between the twin screw speed and ihe residence time has been established elsewhere (Lamnawar and Maazouz (43), Al-Itry ei al. (29)). After melt blending, each strand was quenched in a cold water balh and granulated. This was followed by a 12-h drying stage performed at 40[degrees]C under vacuum, after which the granules of blended materials were molded into disks with a 25-mm diameter and a thickness of about 1.7 mm. These disks were subsequently dried in a vacuum oven overnight before rheological analysis and SEC investigations.
i. For ihe sake of clariiy. only the results of ihe modified PLA on the end of the reaction were studied here, for which the viscoelastic parameters (complex viscosity and storage modulus) remain unchanged during lime. The reaction stabilization times for all the investigated PLA (grade 4032D) samples are reported elsewhere (29). The results demonstrated that for modified PLA with less than 0.79% of multifunctionalized epoxy, 3 min was quite sufficient to obtain a toial conversion wilh a melt temperature of 190[degrees]C. The [M.sub.w] and the viscoelastic properties were greater lhan in other modified PLAs obtained at longer times with the occurrence of PLA chain degradation (a [beta]-C-H hydrogen transfer of PLA as opposed to a hydro-lytie one). With a higher Joncryl amount, Al-Itry et al. (29) showed that the reaction required a longer time (Greater than10 min) to stabilize according to the competition between (he former phenomenon and the chain/branching reaction.
ii. The average residence times were detennined by the way of (i) calculation using the twin screw speed and the selected profile and (ii) using a colored PLA maslerbateh.
iii. Prior to the blown extrusion, rheological and micro-structural studies of the compounds, the samples with 0.7 and 1% of Joncryl were regranulaicd and passed once again through the twin screw extruder (at lower speed and under nitrogen) until obtaining a stable torque and reproducible viscoelastic properties and molecular weight.
2. Blown extrusion (Fig. 2) was performed using an annular 25-mm diameter blow die coupled to a melt pump and twin screw extruder to process neat PLA as well as blended formulations of PLA and chain extender (Joncryl) into a tubular shape. The temperature profile was maintained as in step (1) while the temperature of the tubular die was changed from 150[degrees] to 180[degrees]C. By blowing air through the die head. the tube became inllaied into a thin tubular bubble at which time il was Batted. The blown film was then flattened between nip rolls and pulled up by the winder. Different fonns of observed instabilities were defined, and the effects of the film blowing parameters were illustrated. These parameters include the freeze line height (FLH); the blow-up ratio (BUR), which is defined as the ratio of the final bubble diameter to the die diameter; and the lake-up ratio (TUR) or draw ratio, which is the ralio of the take-up velocity to the extruded malerial veloeily at the die exit. The BUR and TUR are Iwo key parameters of the film Mowing process. Increasing the values of these parameters is explicitly desirable in commercial film production. By varying the BUR. screw speed, air pressure, and winder speed, films wilh Ihieknesses varying between 10 to 150 um and with various degrees of orientation could be obtained.
The main purpose of this article was to study the effect of muliifunctionalized epoxy with nucleating agents and plasticizer in a blend of modified PLA. Table 1 summarizes the compositions of the various poIymer/Joncryl blends.
TABLE 1. Description of prepared PLA-based formulations. PLAJc05 PLA 4032D +0.5% wt Joncryl[R] ADR 4368 PLAJc07 PLA 4032D +0.7% wt. Joncryl[R] ADR 4368 PLAJc1 PLA 4032D + 1% wt. Joncryl[R] ADR 4368 PLAJcTem PLA 4032D + 0,5% wt. Joncryl[R] ADR 4368 + 10% wt. PEG + 1% wt. EBSA + 1% wt. TALC
WAXS and FTIR. Wide-angle X-ray scattering (WAXS) and Fourier transform infrared (FTIR) spectroscopy were also performed to establish the samples' crystalline organization. The WAXS measurements were perfonned at room temperature on a Brucker DK Advance dillraclometer. using as the X-ray source a Cu target ([lambda]. = 1.54 A) operating at 45 mA and 33 kW. The FTIR measurements were carried out at room temperature on a Nicolet iS10 spectrometer (Thermo Scientific) in attenuated total reflectance (ATR) mode using a diamond en Mai.
Thermomechanical analysis was done on processed films using a Rheometrics RSA II solid analyzer in the linear viscoelastic regime at a frequency of 1 Hz. A 2[degrees]C.[min.sup.-1] ramp was applied from 30[degrees]C to 120[degrees]C.
Mechanical properties were evaluated using an MTS[R] tensile lest equipment on various blown films obtained under identical experimental conditions (BUR. TUR. temperatures, etc.) with a crosshead speed of 10 mm * [min.sup.-1].
RESULTS AND DISCUSSION
Linear Viscoelastic and Capillary Flow Properties. Linear viscoelastic properties of the: four gr ades of neat PLA have been compa red in terms of storage modulus and dynamic viscosi ty modulus vers us angular frequency at 180[degrees]C (Fig. 3). The chosen polymers exhibi ted rathe r simi lar rheological propert ies in term s of e lasticity al both low and high angula r freque ncies. They a lso man ifested a cwtonian behavior at low angular frequency and a pseudoplastic behavior beyond 5 rad.[s.sup.-1]. The neal NatureWorks[R] 4060D PLA presented a lower viscosity than the other PLA grades (4032D, 3051D, and 2002D. A Cole-Cole diagram for these four grades at 180[degrees]C is portrayed in Fig. 4.
The main results are listed in Table 2 wilh the zero shear viscosity and average weight relaxation times that have been deduced. The higher relaxation times and zero shear viscosities observed for NatureWorks[R] 2002D and 3051D compared with NatureWorks 4032D and 406OD corroborated the higher molecular weight also found by SEC measurements (Table 3).
TABLE 2. Zero shear viscosity and relaxation lime of ihe ileal PLA used. 2002D 3051D 4032D 4060D ([(tau].sub.w] (ms) 59 42 37 26 ([(eta].sub.0] (Pa.s) 5700 4504) 3500 2400
Investigation of the Stability Processing tor Neat PLA.
The PLA blown films have weaker melt strength when compared wilh polyolefins. making more difficult the formation of a stable bubble during extrusion blowing. Moreover, the PLA's specific density of around 1.24 g.[cm.sup.-3]; is much higher than that of polyolefins (0.910.96 g.[cm.sup.-3]).
The stability of the process was characterized by first establishing stability charts. For this purpose, the TUR and the BUR were varied while being on the lookout for bubble stability or insiabilities. The melt velocity was maintained constant during our study to have unifonn PLA chain extension. In the case of BUR. its value could be changed through control of the air flow injected into the die.
Various bubble defects can be observed and have been reported and classified in the literature (44-47). The most commonly found defects are draw resonance, helical instability, frost line oscillation, bubble sag. bubble flutter, bubble breathing, and bubble tears. Figure 5 illustrates the various defects observed for PLA. The most common instabilities that were observed in this study were bubble breathing (Fig. 5d) and bubble dancing (Fig. 5e). Funhennorc. the process was considered stable for a given BUR-TUR couple when no defect appeared during a significant time (at least 5 min). The repeatability and reproducibility of the measurements during the process were also studied for each BUR and TUR couple.
As reported in the study by Ghaneh-Fard et al. (48), three main instabilities and their combinations were observed during processing and ihese have also been described by oiher aulhors (22-26). They were (1) axisymmetric periodic varia-lion of the bubble diameter, (2) helical motions of the bubble, and (3) variation in the position of the solidification line. At low BUR. a bubble instability similar to the one shown in Fig. 5a was observed when the bubble was inflated with a small amount of air. An FLH instability could be observed at different operating conditions for various grades of PLA and formulations. Depending on the FLH fluctuations, the pressure inside the bubble could oscillate leading to changes in bubble diameter. Under similar conditions, Minoshima and White (25) observed fluctuations in bubble tension for PP and LDPE. while Fleissner (19) noticed significant film thickness variations. A helical instability could be observed at high BUR caused by the development of a helical motion between the nip and the rolls, as illustrated in Fig. 5b.
It is important to note that less siablc blown films were obtained at 180[degrees]C for neal PLAs, making it necessary to decrease the die lemperaiure to 150[degrees]C to enable the fonnation of a few stable bubbles during processing.
The processing charts (Figure 6) established show that PLA 4032D possessed poor blowing abilities. Despite the use of low BUR and TUR ratios, it was nol possible to obtain a stable blown film during our experiments. Both PLA 2002D and PLA 4060D had a good blowing ability at low TUR ralio. wilh various BUR ratios obtained for each TUR. At higher TUR values, however, the blown film bubbles became unstable.
PLA 3051D exhibited a good blowing ability at high TUR provided lhal the BUR values were kept low. The obtained data are summarized in Figure 7 wilh the number of stable points for given TUR (7a) and BUR (7b) ranges. These two figures allow for a more precise comparison between the various PLA grades used. As clearly illustrated. PLA 2002D was the most suitable grade for blown extrusion processing but only if done at low TUR. This primary study also clearly showed the poor process-ability of PLA 4032D by blown extrusion compared with the other PLA grades tested. Because of this. PLA 4032D was selected as the base material for investigating various ways to improve blown film extrusion.
Because blown extrusion of neat PLA film is limited in its operating conditions, establishing a proper correlation between melt shear and exiensional rheological properties was very important to get a better handle of this process. Our challenge was to make this process stable at I80(C)C with higher BUR and TUR ratios. This was done by blending PLA with the appropriate additives. The following sections focus on how PLA modifications can become a key parameter in obtaining a stable process, thanks to changes in the melt rheological properties.
Rheology and Processing of Mollified PLA 4032D by Reactive Extrusion
For a simplified reading and comprehension, only the results of modified PLA upon the end of the reaction and for which the viscoelastic parameters (complex viscosity and storage modulus) remained unchanged overtime are presented here. Earlier studies by Corre et al. (36) and especially Al-Itry et al. (29) have been dedicated to gaming a fundamental understanding of the mechanisms governing both thermal degradation and chain extension reactions of PLA during processing. The reactive extrusion of polymers was performed with various amounts of chain extension/branching agent containing nine GMA fund ions (i.e., Joncryl[R]). The incorporation of this multifunctional oligomer led to an improvement in thermal stability. SEC and intrinsic viscosity measurements of the modified PLA confirmed the increase of viscosity, and molecular weight related to the formation of extended and branched chains. This increase became more pronounced as the concentration of Joncryl[R] was increased. The viscoelastic properties as well as the storage modulus, the dynamic viscosity, and the activation energy were assessed and related to the molecular structure of the modified polymers. Hence, the mechanisms of degradation, chain extension with GMA functions, anil their competition have been proposed. The main objective of the work of Al-ltry et al. (29) was to investigate how the reaction extrusion experiments progressed when changing the residence lime and the amouni of chain extender. Furthermore, the optimal reaction times were evaluated by monitoring the experimental in silu torque stabilization versus time, which has also been detailed in (he recenl works of Liu et al. (49) miRafla et at (50).
Based on the study by Al-Itry et al. (29), the epoxide groups of Joncryl seem capable of reacting with both hydroxy] and earboxyi groups of the PLA chains. The work differentiaied between the reaction of carboxylic acid and hydroxy] groups with epoxide, because epoxide groups are known to react differently with -COOH and -OH groups (the carboxylic acid/epoxide and the hydroxyl/epoxide reaction constants are 18 and 1.2, respectively) (51, 52). In the case of polyesters, glycidyl csierificalion of carboxylic acid end groups precedes hydroxyl end group etherification. This latter reaction competes with etherification of secondary hydroxyl groups and main chain trans eslerification. The resultant couplings involve epoxy ring-opening reactions and the creation of covaleni bonds via hydroxyl side group formation. The resultant (polymer-GMA functions) system represents a complex set of concurrent reactions due to the degradation/chain extension/branching balance.
The real reaction mechanism between PLA chains and multifunciionalized oxirane were also pointed out in the work of Al-ltry et al. (29). The contents of carboxylic acid are given and quantified using the "titration method." and the experimental results were corroborated by FTIR, SEC. and solution viscometry measurements.
We have also plotted the evolution of the relative storage modulus [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] versus lime for modified polymers at various temperatures ranging from 180[degrees]C to 200[degrees]C. From this were detennined the rheokinelic properties as well as the conversion and end-reaction lime. In this study, only the shear and elongation properties for stable-modified PLA are discussed.
Small Amplitude Oscillatory Shear and Capillan How Properties. Three modified PLA 4032D blends with 0.5 wt%, 0.7 wt% and 1 wt% of Joncryl have been investigated using small amplitude oscillatory shear rheology at 180[degrees]C (Fig. 8) and capillary rheometry at 180[degrees]C (Fig. 9). These results were compared with those of neal PLA 4032D for property improvements.
According to Fig. 8. the elastic modulus G' shows an overall increase in value with increased Joncryl content, which can be explained by the chain extension reaction and subsequent branching of PLA chains. This increase in elastic pmpcrties led to an improvement in melt strength, which in turn enhanced the rcsisiance of the blown bubble.
For the mollified polymers, the G' values increased steadily over the whole frequency range. An augmentation of the chain extension/branching level resulted in higher G' values especially at low angular frequencies, and the storage modulus became less shear-sensitive when the Joncryl content was increased, ihus revealing a cohesive network viscosity. Furthermore, the increase might be due to the more entangled structure, providing an improved melt elasticity. where chain interactions played a more significant role. In other words, the larger elasticity at low frequencies of the modified PLA as opposed to the unmodified polymers was attributed to more entanglements due to the presence of long-chain branches [55J. However, as the frequency increased, the number of entanglements decreased due to the more shear-thinning character of the modified polymers.
Moreover, the slope of the (G' vs. [omega]) curves decreased drastically with the incorporation of Joncryl and consequently the G'("w) converged at the same value of a plaleau at higher frequency regardless of the amount of Joncryl. Thus, all the curves of the storage miHlulus tended toward the same plateau, which confirmed that the modified PLAs had the same entanglement molar mass Me. Because it was demonstrated that [M.sub.w] increased for the modified PLA with Joncryl. we can say that the number of entanglements per chain Z (54) also increased with the Joncryl contenl, based on the following equation
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]
where [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] and [rho] (g .[cm.sup.-3]) are the plaleau modulus and the density, respectively.
Figure 9 confirms the Cox Merz validation hased on the way the dynamic viscosity modulus versus frequency curves can be superposed to the shear viscosity versus shear rate curves obtained by capillary rheometry using Bagley and Rahinovitch corrections.
It can be clearly observed that the addition of inulti-lunctionalized epoxide (Joncryl[R]) had a direct influence on the rheological behavior of PLA. Results show that the shear viscosity values increased with the Joncryl content, indicating the presence of chain extension and/or a branching phenomenon. The longer and heavier chains created more entanglements, thus giving rise to a higher viscosity. One can also notice that neat PLA 40321) exhibited a Newtonian plaleau at low shear rale, whereas PLA/Joncry[R] blends presented either a shear thinning or a pseudoplastic behavior depending on the shear rate and amount of multifunctional epoxide. The increase in viscosity corroborated the higher [M.sub.w] obtained from SEC (Table 3. example of PLA 4032 + 0.5 J). The increase in Joncryl[R] content also influenced the transition from a Newtonian to as hear thinning behavior. This was more pronounced at 1 and 0.7 wt%. than at 0.5 and 0 wt%., respectively. The polydispersity was also influenced by the Joncryl[R] content, as confirmed by SEC measurements. The effects were clearly demonstrated here and corroborated the results oblained by Gwo-Geng et li (55).
TABLE 3. SEC resulls of neal and processed PLA without and with multifunclionali/ed epoxy. Grade [M.sub.w] [M.sub.n] [l.sub.p] (g/mol) (g/mol i 2002D 202.980 118,400 1.71 305 ID 165.030 97.590 1.69 4032D 100.000 57.000 1.75 4060D 94.370 53.910 1.75 4032D, extruded 92.000 46.900 1.96 al 200[degrees]C 4032D + 0.5% wt. 154,000 36.500 4.22 Joncryl[R], extruded at 200[degrees]C
Elongation Properties of Modified PLA. Rheology was used as a tool to evaluate the meli properties of the chain extended materials. The polymer structure, e.g., the chain length. MWD, branching, directly influenced the melt response during mechanical solicitalions. Dynamic rheology provided information about the viscoelastic shear dependence and chain relaxation profile of polymer in the molten state. Because capillary extrusion involves high shear rates, this last rheological experiment can reveal process limitations in high throughput applications. Furthermore, elongation experiments display the melt behavior during mechanical stretching.
The melt strength is often reported as the elastic behavior of a polymer under shear. The previous section highlights the kind of melt reinforcement provided by the addition of Joncry[R]. In terms of processability, the melt resistance during stretching is more commonly employed. A correlation between these two specific properties has been proposed by Cogswell et al. who established a relationship between the elongation viscosity and die pressure during capillary extrusion experiments, which were governed by melt elasticity. Determination of the elongation flow behavior of molten polymers is still an open issue as it remains difficult to reach very large strains and strain rates wilh existing elongation rheometers. This is why indirect methods for deiermining the elongation viscosity have been proposed, such as convergent flow analysis (CFA). For the last two decades, data obtained by different direcl and/or indirect methods have been frequently compared but contradictions still exisi and the problem of determining the elongation viscosity is therefore nol totally solved.
Indirect Miethod of Determining the Elongation Viscosity-Cogswell Elongation Viscosity. the indirect method widely used for determining the elongation viscosity is by isothermal melt spinning because it is suitable for a large variety of fluids. Cogswell developed an analytical analysis to calculate the "elongation viscosity" from capillary rheometer data. He considered lhal the pressure drop at the die entry was the shear and elongation contributions.
The presence of vortexes renders it possible to treat the flow in a way similar to what is used for slighlly converging pipes. Assuming that the viscosity under simple shear can he described by a power-law relationship over a limited stress range and that the elongation viscosity is stress independent, these equations can be solved for an infinite set of very short tapes yielding analytical expressions for elongation viscosity and rates:
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]
Here, e is the elongation rale ([s.sup.-1]), [gamma] is the shear rate (rad.[s.sub.C] [DELTA]Pe is the capillar) drop pressure (Pa), n is the pseudoplastic coefficient, and [lambda] is the shear viscosity obtained on Bagley correction (Pa.s).
The advantage of this analytic analysis is that it provides indications about elongation viscosity data, which is difficult to measure, lis limitation is thal it is a simplified and approximate theory (56). The method can provide data at significantly higher strain rales (100 [s.sup.-1]) than other methods, but because of the numerous theoretical assumptions and of some surprising literature results, certain contradictions still remain to this day. For instance, in 1971. Cogswell compared results from a steady-stale experiment with a melt tensile rheometer to one wilh a CFA (57). Although the assumptions made by Cogswell in his theory might be questionable, some inconsistencies between authors can also find their origin in unreliable experimental values of the elongation viscosity.
For example, the experimental daia obtained through isothennal fiber-spinning are questionable as the elongation rales were nol constant along the thread-line and steady slate was seldom achieved, Therfore, the values obtained with an elongation viscometer are much more reliable, but the level of strain rates reached is unfortunately much smaller than the one obtained with convergent flows and results are nol easily comparable.
In his original study, Cogswell considered a power-law equation for the viscosity where the power law index n is assumed to be constant regardless of the shear rate. From the shear How curves, n obviously changes with sheaf rate and the power law index therefore has to be laken as a variable: in the calculations. The graph in Fig. 10 charactcrizes the elongation viscocity versus elong.uion rate after Bagley correction. for the four blends studied (with 0 wt%, 0.5 wt%, and 0.7 wt% or Joncryl[R]. The experiments were established for a die with 180[degrees] entry angle with various L/D nuios to allow for Bagley corrections.
It can be observed that the: elongation viscosity increased with the Joncryl[R] content--a direct result of the reaction between the PLA chains and the multifunctionalized cpoxide. Figure 11 shows the evolution of the clongaiion viscosity versus the Joncryl[R] content for a fixed elongation rate (30 [s.sub.-1]). The drastically increased elongalion viscosity confirmed the higher melt strengthening properties of modified PLA compared with the neat polymer.
Uniaxial Experiments. In this Study, a uniaxial elongation rheometer was, used to evaluate the mechanical behavior of molten samples under stretching. Figure 12 displays the elongation viscosities obtained at 160[degrees]C for strain rates ramgomg from around 0.1 up 1 [s.sup.-1]. The main observation is that chain-extended PLAs exhibiled a strain -hardcning behavior under lhe stretching conditions investigated. Conversely, neat PLA maint a linear response over time at all investigates strain rates. This is in a good agreement with the typical response observed for a linear polymer under stretching.
It is known that the elongational viscosily increases when going from linear to highly branched PLAs. Thc linear PLA, where the polymer chains arc free from any branch points, maintain a linear response over time, and the chains are not prc\cnted from slipping over c:lch oIhc:r. In chain-branched polymers. the ch:lin slipping is perturbed by Ihe presence of side branches elllangled \\'ilh the main ch:lins. caw..ing slr;lin h:trdening. In this ca~e. the stretching force or strc~s reache~ a pseuJosleady state before incre;lsing:lg:lin due to the resist~mce caused hy the br;mches inh:nningling with the: main ch:lins of the polymer. A critical sirain for the onscl of strain hardening is shifted to a lower value when the amount of the chain extender increased.
We observed that Trouion's ratio between steady-shear and elongalional viscosities was slighlK greater than the expected value of 3 for the linear sample. When the chain extender contenl was increased, this ratio gradually diverged. These results corroborate those from the literature. The other major finding from these experiments was the relevant sirain hardening lhal appeared in branched PLAs. This resistance to stretching became more significant as the CE content was increased.
The critical strain ([[epsilon].sub.SH.sub.onset])at the onset of sirain hardening was shifted to a lower value when the amouni of Joncryl was increased. Thus, the onset of sirain hardening appeared at a Hencky sirain of l for PLA_1% Joncryl[R]. A Hencky strain of 2 was required for the strain hardening onset of PLA_0.5% Joncryl[R] at a strain rale of around 0.1 [s.sup.-1]. An increase in the relaxation lime spectra with chain exlension could explain this earlier strain hardening. Thus, widening the melt time response would reduce the chain orientation dynamics during stretching, leading to a reinforcement of the mechanical response. Siadler et al. (58) worked on bimodal LDPE and assigned the observed strain hardening to both a wide MWD and a LCB contenl (58). In the present case, this phenomenon was attributed to a MWD enlargement caused by chain extension without more specific inlbnnaiion about the chain extension topology (LCB, short-chain branching, star branching, etc.).
Investigation of Chain Branching!Extension Balance. Recently, Trinkle et al. (59) proposed a rheological method for characterizing polymer melts and chain topology, which is based on the reduced Van-Gurp-Palmen plot (rVGP) (59). An rVGP plot is made up of the loss angle [delta]([delta]=[tan.sup.-1][MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]) against the logarithm of reduced modulus [G.sub.red], and the latter is defined as the complex modulus (G*) divided by the plateau modulus [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (obtained when [delta] [vector] 0). The idea behind this model is to bring out the fingerprints that are characteristic of a certain topology (long-chain. H-shaped, star-shaped, or comb-shaped branched structure), whereby low levels of LCB can be delected and quantified in the linear viscoelastic behavior. The shape of the Van-Gurp-Palmen plot was found to I K unaffecied by the molar mass. Rather it depended on the polydispersily that stretched the curves along the abscissa at low [G.sub.red].
As illustrated in Fig. 13, a transition existed between the neal polymers and their modified counterparts. In the case of PLA_0, the curves derived from rheological daia exhibited the classical shape of a linear polymer where the phase angle displayed a low [G.sub.red], a plaleau at [delta] = 90[degrees] before decreasing rapidly for [G.sub.red] values above 0.01.
For the modified PLA samples, the Van-Gurp-Palmen plots showed a linear decrease in phase angle values with increasing [G.sub.red], without the presence of a plaleau at 90[degrees] similar to PLA_0. When [G.sub.red], reached the 0.01 to 0.04 region, the decrease in phase angle value became more pronounced. This kind of flat profile at low [G.sub.red] characterized the presence of LCB as defined by various authors [59-61]. In addition, the data showed that PLA 0.5 contained the largest amouni of LCB among all the samples.
An increase in phase angle was observed for PLA with a Joncryl content above 0.5 wt%, indicating a decrease in the LCB amount. Because modified PLA did nol present a similar irend to the polymers with known topology (symmetric star, asymmetric star. H-shaped, and comb-shaped branched structure), it was concluded ihat they exhibited a behavior typical for mixtures of linear and randomly branched polymers. They were believed to have complex structures in which were found distributions of backbone lengths, branch lengths, branch point locations, antl branching complexity. The obtained data based on the rVGP corroborated the linear viseoelasticily and the elongation properties, confirming the high elastic behavior of modified PLA. This elastic behavior is a key factor and should enable its processability in blown exlrusion.
Investigation of the Processing Stability of Modified PLA. As demonstrated previously. PLA presents a lower melt strength, which limits its ability for blown exlrusion. To improve ils stability, attempts were made to increase PLA's melt strength properties and its shear and elongalional properties and have studied the corresponding effects.
Figure 14 illustrates the processing stability chart obtained for modified PLA with 0 wt%, 0.25 wt%, and 0.5 wt% Joncryl content. Because it was impossible to obtain a stable bubble for PLA at 180[degrees]C, the die lemperature was set to 150[degrees]C for this particular case only. The challenge of ihis work was to obtain a stable blown film at 180[degrees]C through an increase in elongalional viscosity, achieved by varying the mullifunciionalizcd epoxide content.
Figure 15 illustrates the higher BUR obtained at a higher TUR in the situations where 0.25 wt% and 0.5 wt% of Joncryl[R] was added. The resulting blown films were stable even afler changing the TUR. Stable zones were oblained through chain extension/branching of PLA. It was also observed lhal PLA films were quite stiff and had a much lower elongation than PLA + Joncry[R] Figure 16 shows the evolution of BUR versus Joncryl[R] conlenl at high TUR (3). As can be seen, the BUR values reached and exceeded a 5:1 value for a Joncryl[R] content of 0.5 wt% and above.
Blown Extrusion of Chain-Extended PLA With Added Nucieants and Plasticizer
The impaci on blown processing properties of chain extension/branching coupled to nuclealion and plasticization has also been investigated. A previous study (62) proposed a formulation that improved the crystallization kinetics. This formulation was composed of 1 wt% of EBSA. 1 wt% of TALC, and 10 wt% of PEG, which were blended to PLA 4032D with 0.5 wt% of Joncryl. This blend of EBSA. TALC, and PEG is hereafter referred to as the ternary system (abbreviated as Tern.). Figure 17 shows the evolution of the dynamic viscosity versus the angular frequency for PLA + 0.5 wt% Joncryl[R] + ternary system. When compared with PLA and PLA05Jc, the slight decrease in dynamic viscosily that was observed at low frequency was associated to the plasticizer. However, no significant effect was observed at higher angular frequencies.
Stability charts were established for the processed blends and can be found in Fig. 18. When compared with a PLA/Joncryl[R] blend, the lemary system provided to a clear enlargement of the processing window as bubbles were obtained at higher BUR ratios. This is also illustrated in Fig. 19. Of most interest is the significant increase in the maximal value for stable BUR ratios, going from 7 to 12 at high TUR. It was believed that even higher BUR ratios were attainable but this could not be verified as such experiments exceeded the current limitations of our processing equipment. This surprising behavior was in agreement with the effects of chain-extension branching and (he role of nucleating agents and plasiicizers. The blown bubbles also showed a translucent aspect, which was associated with a semi-crystalline structure later characterized using DSC, WAXS, and FTIR.
Microsiructure and Mechanical Properties of the Stahle Blown Minis. Improvements of the crystallization kinetic of PLA have been widely reported in the literature. TALC has been found by various aulhors to be an effective nucleating agent for PLA (42, 63). and PEG and citrate esters have also been reported to be good plasticizers for PLA by promoting molecular mobility and thus crystalline structure growth. It should be noted that although plaslicizcrs lower the glass transition temperature, fully crystallized PLA with plasticizers has a better thermal stability than neat PLA (42). Others authors have reported that aliphatic amides are effective nucleating agents (40, 41) for PLA. Unfortunately these results cannot he used to achieve a crystalline phase in PLA during processing. This is due to the crystallization half-time still being too long and even higher when using these nucleating agents and plasticizers. However, through a simple association of three additives (TALC, PEG, and aliphatic amide), we were able to significantly improve the kinetics of crystallization for PLA (62).
As one of the objectives ol this work was to improve the crystallization behavior of blown films upon processing, a thorough DSC characterization on the blown films was performed. Figure 20 illustrates the difference between the First and the second healing runs of the PLA-based hlown film. It can be clearly seen that the processed blown film had a melting behavior of a crystalline structure when compared with a neal granulate for which the lower kinetics of crystallization of PLA hindered the formation of sphcrulitic structures during processing.
Figure 21 portrays the heat How versus lemperaiure of three films obtained at various BUR and TUR ratios. The heat flow curves of the PLAJcTem films showed a cold crystallization peak at around 70[degrees]C. which did not appear for the neal PLA. The crystalliniiy of our sample was calculated using Eq. (3). and the results are given in Table 4. The [DELTA][H.sub.m] [infinity] value used for PLA was 93 J.[g.sub.-1] (64).
TABLE 4. Bl K and I I K values and thermal analysis results of films characterized in Fig. 21. Film TUK BUR [DELTA][H.sub.ee] [DELTA][H.sub.m] (J/g) (J/g) 1 1.3 4.6 15.9 31.8 2 2 5.6 16 34.4 3 5.3 9.2 14.7 28.9 Film ([(chi].sub.e] (%) 1 17 2 20 3 15
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]
where [[chi].sub.e] is a percentage of crystallinity in the sample. [DELTA][H.sub.m] is a melt enthalpy (J.[g.sub.-1], [DELTA][H.sub.cc] is a cold crystallization enthalpy (J.[g.sub.-1], and [DELTA][H.sub.m] [infinity] is a melt enthalpy of a spherulile of infinite size (J.[g.sub.-1]).
To investigate the crystallization kinetics, blown films of various PLA formulations were lesied. and the samples are listed in Table 5. The formulation with the ternary system was based on iwo nucleating agents (TALC and EBSA) and a plasticizer (PEG). This system has in a previous study already proven efficient on NatureWorks[R] PLA grade 2002D. The challenge in this study was to validate its effect on the crystallization kinetics of Nature-Works PLA grade 4032D upon blown extrusion processing.
TABLE 5. Description ol lesied lormulaiions lor the investigation of the kinelic of crystallization. ID Description K0 Neat PLA 4032D-blended in extruder K1 PLA 4032D + 10% wt. PEG K2 PLA 4032D + 1% wt. Tale + 1 % wt. EBSA K3 PLA 4032D + 1% wt. Tale + 1 % wt. EBSA + 10% wt. PEG K4 PLA 4032D + 1% wt. Tale + 1 % wt. EBSA + 109 wt. PEG + 0.59 wl. Joncryl[R]
Li et al. (40) and Young Nam et al. (41) have shown lhal typical transcrysiallization occurs at the interface between PLA and the aliphatic amide, wilh PLA crystallizing epilaxially on the amide. They also demonstrated lhal TALC is an effective nucleating agenl for PLA. A coupling effect of PEG with TALC has also been shown by Li et al. (42). with PEG enabling an increase in crystallinity even at high cooling rates.
Isothermal DSC analyses were perfonned as a first step with data derived from the Avrami model (65, 66). The cryslallization half-lime at various isothennal crystallization temperatures is plotted in Fig. 22. which clearly demonstrates that the crystallization half-lime for PLA was improved by the proper use of the nucleating agentsand plasticizers, By adding the ternary blend, the cryslallization half-time lor PLA drastically dropped from more than 2000 s down to 40 s. The presence of Joncryl[R] (K4) did not significantly impact the cristallization of PLA/Tern (K3), which is in agreementwith previous observalionsmade by our group. The addition of the ternary blend enabled PLA to develop a crystalline phase during processing even at the high cooling rate of the laboratory scale blown extrusion process. The use of PEG also allowed for the development of a crystalline phase at lower temperalures. which was appropriate for the polymer cooling conditions used during processing.
To conclude the investigation on the crystallization kinetics of neal PLA and the PLA-based optimized system. DSC heating/cooling ramps were performed on various samples. The results are given in Fig. 23. The samples were first heated at 200[degrees]C for 3 min to erase the polymer's thermal his tory, after which they were cooled and reheated at the selected temperature. The results clearly demonstra ted lh.1I our samples were able to achieve a full crystaliza rion at a cooling rate up 50[degrees]C.min [ .sup.-1], which the maximum rate ach ievable with our DSC instrument. The melt en thalp [gamma] of the crystalized samples during heating was 38 J.[g.sup.-1], a value close to the ma ximum crysta lline ability of our material. No [T.sub.g] was observed, confirming our hypothesis of a fully developped crystalline phase.
Figures 24 and 25 show the FTIR and WAXS spectra of blown film, of neat PLA and PLAJcTent. The 921 [cm.sub.-1] band observed in the Fn R spectra (Fig. 24b) was assigned to a [10.sup.3] helix-sensitive ba nd from the crysta lline phase of PLA deve loped during processing. Previous studie (67, 68) have also auributed the 955 [cm.sup.-1] band to an arno rphouv state. The intensity of this band decreased whe n that of the 921 [cm.sup.-1] band increased. which was in concurrence with our resuh. Also WAXS (Fig. 25) demonstrated tha t our blenJs were able to develop a crystalline phase when being processed, as clearly seen by the sharp 20 peak found at 16.5[degrees].
By using a ternary system with a chain extension/branching reaction, it was possible to achieve higher crystallization kinetics for PLA. Consequently crystallization occurred during the blowing process at higher BUR and TUR ratios (e.g., stretching and orientation of chains). These results are of high enough interest to warrant a thorough analysis of the microstruciural properties of the blown-processed films, which will be the subject of a distinct article.
The Ihermomechanical properties of the PLA and the optimized PLA-based blends were also invesiigated (Fig. 26). This was done by comparing the storage elastic modulus E* and tan [delta] of the various films obtained under identical processing conditions. PLAJcTem exhibited a higher elastic modulus in the rubbery state (Tilde[10.sup.8] Pa) lhan PLA ([10.sup.7] Pa), confirming the presence of a crystalline structure that was induced during processing and testing. Of particular interest is the sharp increase in E' found at Tilde100[degrees]C for neat PLA that can be associated to the cold cryslallization lemperaiure. This phenomenon was less pronounced and occurred at a lower temperature (70[degrees]C) for PLAJcTern. thus confirming the DSC results. The [alpha] transition for PLAJcTem was also lower when compared with neal polymers, as was the peak amplitude of tan i). which decreased from 2.5 to less than 1.5. Work is in progress to optimize the amouni of added adjuvant in the films, but this is dependent on the final application.
Finally, the mechanical properties of PLA and the optimized blends were characterized (Fig. 27 and Table 6). The blend of PLA. Joncryl,[R] and the ternary system showed a ductile behavior when compared with neal PLA, which had a fragile character. The elongalion at break ([[epsilon].sub.B]) reached a value of 360%, and the Young modulus was lowered from 2.5 GPa for neal PLA to less than 1.8 GPa for the blend. Thus, this new malerial was more flexible and less rigid, making it more suitable for packaging and films applications.
TABLE 6. Average values of tensile lesi (Fig. 27) of PLA films and PLA. Joncryl . and ternary system films. E [sigma] ([(epsilon].sub.y] ([(sigma].sub.b] (MPa) (MPa) (%) (MPa) PLA 2540 54.6 3.1 (0,1) 47.3 (1.8) (210) (2.1) PLAJcTem 1770 432 (1.2) 4.5 (0,3) 43.3 (1,4) (23) ([(epsilon].sub.b] (%) PLA 27 (15) PLAJcTem 362 (31)
This work focused on the improvement of bolh the processing stability and microstruciural properties of blown films based on PLA. As demonstrated herein, the extrusion blowing of various grades of neat PLA is Habited. Tims, a proper correlation between melt shear and exiensional rheological properties is very important to gel a betler handle of ihis process. Our challenge was to render ihis process stable at higher melting lemperaiure with the highest possible BUR and TUR ratios, by blending PLA with appropriate additives.
It was clearly observed that the addition of a multifunctional i zed epoxide (Joncryl[R]) had a direct influence on the rheological behavior of PLA. Results show that the shear viscosity increased with the Joncryl content independently of the shear rale, indicating the occurrence of chain extension and/or hranching mechanisms. One can also notice that neat PLA 4032D exhibiled a Newtonian plateau at low shear rale, while the PLA/Joncryl[R] blends presented either a shear-thinning or a pseudoplastic behavior depending on the shear rate and amount of multifunctional epoxide. This improvement in shear and elongalion viscosity was corroborated by the higher masses obtained from SEC.
The increase in Joncryl[R] conlenl also influenced the transition from a Newtonian to a shear thinning behav ior. Shear and exiensional rheology were used as a tool to evaluate the melt properties of the chain-extended materials. Il was shown that the polymer structure directly influenced the melt response during mechanical stress and that both viscoelastic and elongation properties are very sensitive to the chain extension and branching of PLA. I he chain extension and branching balance was also studied based on the Van-Gurp Palmen curves, and the obtained results showed that they exhibited a behavior typical of a mixture of linear and randomly branched polymers. Thus, the materials were believed to have complex structures comprising distributions of backbone lengths, branch lengths, branch point locations, and branching complexity.
Furthermore, stability charts were established for the processed blends. When compared with a PLA/Joncryl[R] blend, the lemary system led to a clear enlargement of the processing window as bubbles were obiained at higher IH R raiios. Of most interest was the significant increase in the maximal value for stable BUR ratios, going from 7 to 12 at high TUR. Higher maximal BUR ratio could be possibly attained but could nol be verified as the limitations of our equipment were reached. This surprising behavior was in agreemenl w ith the effects of chain-extension hranching and the role of nucleating agents and plasticizers.
The blown bubbles also showed a translucent aspect that can be associated wilh a semi-cry slal line siructurc. Thus, they were characterized with DSC. WAXS. and FTIR. By using a ternary system with chain extension/ branching reaction, it was possible to achieve higher cryslallization kinetics for PLA. This in turn enabled crystallization during the blowing process at higher BUR and TUR ratios (e.g., stretching and chains orientation). The obtained results are of high enough interest to warrant a thorough analysis of the microstruciural properties of such blown-processed films in a distinct article.
The thermomechanical properties of the PLA and the optimized PLA blends were also investigated, and confirmed the presence of a cry stalline structure induced during processing. Mechanical properties of the optimized PLA-based blend revealed it to be a more flexible and ductile malerial lhan its neal and brittle counterpart.
This study was carried out within the framework of the AMOPLA (Processability of PLA for industrial plastic converters) FEDER project funded by Region Rhone Alpes (France). We also wish to thank BASF[R] for providing the Joncryl[R] additives. We express our appreciation to the reviewers for their constructive and meticulous assessment of this work, and we finally acknowledge S. Vaudreuil (PhD Eng.) for his help, the Masters students L. Foulon, M. Gois-nard, F. Ravel, and J. Gomilschag for their collaboration.
(1.) L.T. Lim. R. Auras, and M. Ruhino. Prog. Polym. Sci.,33, 820 (20081.
(2.) M. Vert. G. Sehwarch. and J. Coudane, J. Macromol Sci Pure Appl Chem,A32. 787 (1995).
(3.) J. Lunt. Polym. Degrad. Stab..59. 145 (1998).
(4.) R.H. Wchrcnberg. Mater. Eng.,94. 63 (1981).
(5.) R.E. Drumright. P R. Gruber. and D.E. Henlon. Adv. Mater..12. 1X41 (2000).
(6.) E.S. Lipinsky and R.G. Sinclair. Chem. Eng. Prog..82. 26 (1986).
(7.) F.Dc Sanhs. R. Paniani. and Ci. TiUimanlio, I hermoclnm Acta..522. 128(2011).
(8.) J.R. Dorgan. J.S. Williams, and D.N. Lewis. J. Rheol.,43. 1141 (1999).
(9.) Z.F. Zhou. G.Q. Huang. W.B. Xu. and F.M. Ren. Express Polymer Lett.,I. 734 (2007).
(10.) LJ. Paladc. H.J. Lchcrmcicr. and J.R. Dorgan. Macromole-cides.34. 1384 (2001).
(11.) E.W. Henry. T.H. Chili. E. Nyilas. T.M. Brushait. P. Dik kics. R.L. Sidman. Exp. Neurol..90. 652 (1985).
(12.) S. Yollcs. T.D. Lcafc. and F.J. Meyer. 7. Pharm. Sci.,64. 115 (1975).
(13.) S. Yollcs. T. Lcsl. L. Ward, and F. Bocttner. Ahstr. Paper. Am. Chem. Sm. Natl. Meet..Supplement I. 85 (1976).
(14.) N. Ljunghcrg and B. Wcsslen. J. Appl. Polym. Sci..86. 1227 (2002).
(15.) N. Ljunghcrg and B. Wcsslen. Polymer. 44. 7679 (2003).
(16.) N. Ljunghcrg and B. Wcsslen. Biomat romolctules. 6. 1789 (2005).
(17.) O. Martin and L. Avcrous. Polymer.42. 6209 (20011.
(18.) L.V. Labrcet|ue. R.A. Kumar. V. Dave. R.A. Gross, and S.P. McCarthy. J. Appl. Polym. Sci..66. 1507 (1997).
(19.) M. FIcissncr. Int. Polym. Process..2. 229 (1988).
(20.) P.A. Sweeney. G.A. Campbell, and F.A. Feeney. Int. Polym. Process..7. 229 (1992).
(21.) W. Ast. Kunstoffe. 64. 146 (1974).
(22.) CD. Han and J.Y. Park. 19, 3291 (1975).
(23.) CD. Han and R. Shelly, hid Eng Chem Fund..16. 49 (1977).
(24.) T. Kanai and J.L. While. Polym. Eng. Sci..24. 1185 (1984).
(25.) W. Mmoshima and J.L. While, J. Non-Newtonian Fluid Meeh.,19. 275 (1986).
(26.) J.L. While and H. Yamane. Pure Appl. Chem..59. 193 (1987).
(27.) J. Laffarguc, Y. Dcmay, and J.F Agassant. Int. Polym. Process..25. 356 (2010).
(28.) A.L. Arraiza. J.R. Sarasua. J. Verdu. and X. Colin. Int. Polym. Prtness.. 22, 389 (2007).
(29.) R. Al-llry, K. Lanmawar. and A. Maazouz. Polym. Degrad. Stab .97. 1898 (2012).
(30.) Y.W. Di, S. lannaec. E. Di Maio. and L. Nieolais. Macromol. Mater. Eng..290. 1083 (2005).
(31.) J. Kylma and J.V. Scppala. Macromolecules.30. 2876 (1997).
(32.) M. Mihai. M.A. Huncault. and B.D. Favis. Polym. Eng. Sci..50. 629 (2010).
(33.) S. Pilla. S. G. Kim. G.K. Aucr. S.Q. Gong, and C.B. Park. Polym Eng. Sci..49, 1653 (2009).
(34.) S. Pilla. A. Kramschusier. L. Yang. J. Lcc, S. Gong. L.S. Turng. Mater Sci Eng C.29. 1258 (2009).
(35.) B.Y. Shin. D.H. Han. and R. Narayan. J Polvm. Environ..18. 558 (2010).
(36.) Y.M. Corre. J. Duehct. J. Reignier. and A. Maazouz. Rheol. Acta.,50. 613 (2011).
(37.) Y.M. Corre. A. Maazouz. J. Duehel. J. Rcigntcr. J Supercritical Fluids.58.177 ( 2011 i.
(38.) H. Xu. C. Q. Tcng, and M.H. Yu. Polymer,47. 3922 (2006).
(39.) K.S. Anderson and M.A. Hillmycr. Polymer,47, 2030 (2006).
(40.) J Li. D.K. Chen. B.Z. Gui. M.H. Gu. and J. Ren. Polvmer Bulletin.67. 775 (2011).
(41.) J.Y. Nam ct al.. Polymer.47. 1340 (2006).
(42.) H.B. Li and M.A. Huncault. Polymer.48. 6855 (2007).
(43.) K. Lamnawar and A. Maazouz. Rheologica. Acta.,47. 383 (2008).
(44.) J. Laftargue et al.. Int. Polym. Protess..17. 347 (2002).
(45.) Y.L. Ycow. J. Fluid Meclt..75. 577 (1976).
(46.) V. Sidiropoulos and J. Vlaehopoulos, Polym. Eng. Sci.. 40. 1611 (2000).
(47.) V. Sidiropoulos and J. Vlaehopoulos. J Rem/ Plast. Compos..21. 629 (2002).
(48.) A. Ghaneh-Fard. P.J Carrcau. and P.G. Laflcur. AlChE J,42. 1388 (1996).
(49.) Liu. L. Lou. W. Yua. R. Liao. R. Li. and C Zhou. Polymer..51. 5186 (2010).
(50.) P Raffa. M-B. Colcclli. S Savi. S. Bianehi. and V. Caslcl vctro. React Fund. Polym .72. 50 (2012).
(51.) N. Poisson. A. Maazouz. H. Sautercau. M. Taha. X. Gam-bert. J Appl. Polymer Sci..69. 2487 (1998).
(52.) L. Bceu. H. Sauicreau. A. Maazouz. J.F. Gerard. M. Pabon. C. Piehol, Polymers Adv. Tech..6. 316 (1995).
(53.) H.T. Scycd. J.C Pierre, and A. Abdcllah. Polym. Eng. Sci.. 50.191 (2010).
(54.) H. Munslcdl. Soft Matter.7. 2273 (2011 >.
(55.) G.G. Lin. H.H. Shih. P C. Chai. and S.J. Hsu. Polym. Eng Sci..42. 2213 (2002).
(56.) J.Z. Liang and J.N. Ness. Polym. Test..16. 403 (1997).
(57.) F.N. Cogswell. Polym. Eng Sci..12. 64 (1972).
(58.) FJ. Siadlcr, J. Kaschta. H. Munslcdl. F Becker, and M Huhaek, Rheol. Acta..48. 479 (2009).
(59.) S. Trinklc. P. Waller, and C Friedrich. Rheol. Acta.. 41. 103 (2002).
(60.) J.Y. Liu. L. Lou. W. Yu. R. Liao. R. Li. C. Zhou. Polymer. 51.5186 12010).
(61.) S.H. Tabatahaci. P.J. Carrcau. and A. Ajji. Polym. Eng. Sci.,50. 191 (2010).
(62.) A. Maazouz. K. Lamnawar. B. Mallei. Patent: C08L67/00; C08J5/10. FR294I702 (Al) (2010).
(63.) Y. Yurycv, P. Wood Adams. M.C. Heuzey. C. Dubois, and J. Brisson. Polymer.49. 2306 (2008).
(64.) E.W. Fischer. HJ. Stcrzcl. and G. Wcgncr. Kolloid-Zeits-chrift and Zeitschnft Fur Polymere.251. 980 (1973).
(65.) M. Avrami. J. Chem. Phys..7. 1103 (1939).
(66.) A. Choudhury. Mater. Sci. Eng. A-Struct.,491. 492 (2008).
(67.) J.M. Zhang. H. Tsuji. I. Noda. and Y. Ozaki. J Phys. Chem..B108. 11514 (20O4).
(68.) J. Zhang. [gamma] Duan. H. Sato. H. Tsuji. I. Noda. S. Yan. and Y. Ozaki. Macromolecules.38. 8012 (2005).
Correspondence to: Abderrahim Maunuz; e:mail: Abderrahim.email@example.com
Published online in Wiley Online Library (wileyonlinelibrary.com).
[c] 2013 Society of Plastics Engineers
Benoit Mallet, (1), (2) Khalid Lamnawar, (1), (3) Abderrahim Maazouz (1), (2), (4)
(1) Universite de Lyon, INSA Lyon, 20 Avenue Albert Einstein, France
(2) Ingenierie des Materiaux Polymeres, IMP UMR CNRS #5223, INSA de Lyon, France
(3) Laboratoire de Mecanique des Contacts et des Structures, LaMCoS UMR CNRSU5259, INSA-Lyon, France
(4) Hassan II Academy of Science and Technology, Rabat, Morocco
|Printer friendly Cite/link Email Feedback|
|Author:||Mallet, Benoit; Lamnawar, Khalid; Maazouz, Abderrahim|
|Publication:||Polymer Engineering and Science|
|Date:||Apr 1, 2014|
|Previous Article:||Modeling the Rheology of SR1500 and LY556 Epoxies Under Manufacturer's Recommended Cure Cycles after Viscosimetry and Rheometry Characterization.|
|Next Article:||Morphology and Structure of Nylon-6 Crystallized in Epoxy Resin Matrix.|