2015 plasticized polylactic acid-cellulose nanocomposites prepared using melt-extrusion and liquid...

8/17/2019 2015 Plasticized Polylactic Acid-cellulose Nanocomposites Prepared Using Melt-extrusion and Liquid Feeding, Mech…

http://slidepdf.com/reader/full/2015-plasticized-polylactic-acid-cellulose-nanocomposites-prepared-using-melt-extrusion 1/7

Plasticized polylactic acid/cellulose nanocomposites prepared using

melt-extrusion and liquid feeding: Mechanical, thermal and optical

properties

Natalia Herrera, Aji P. Mathew, Kristiina Oksman ⇑

Composite Center Sweden, Luleå University of Technology, 971 87 Luleå, Sweden

a r t i c l e i n f o

Article history:

Received 20 September 2014Received in revised form 20 November 2014Accepted 23 November 2014Available online 26 November 2014

Keywords:

A. NanocompositesB. Mechanical propertiesB. Thermal propertiesD. Scanning electron microscopy (SEM)E. Extrusion

a b s t r a c t

Plasticized polylactic acid (PLA) and its nanocomposite based on cellulose nanofibers (CNF) and glyceroltriacetate (GTA) were prepared using a co-rotating twin-screw extruder. GTA was used as a plasticizer, aprocessing aid to facilitate nanofiber dispersion and as a liquid medium for their feeding. The optical,thermal and mechanical properties were characterized and the toughening mechanism was studied.The addition of GTA (20%) and CNF (1%) resulted in increased degree of crystallinity and decreased opticaltransparency. Furthermore, these additives showed a positive effect on the elongation at break andtoughness, which increased from 2% to 31% and from 1 to 8 MJ/m3, respectively. The combination of slip-page of the nanofiber–matrix interface and a massive crazing effect as a result of the presence of CNF issuggested for PLA toughening. CNF were expected to restrict the spherulite growth and therefore enhancethe craze nucleation.

2014 Elsevier Ltd. All rights reserved.

1. Introduction

During recent years there has been growing interest in thedevelopment of environmentally friendly materials, as well as aneed to replace many synthetic polymers with biodegradablematerials such as polylactic acid (PLA), especially in packaging.Many reviews have been published on the topic where both indus-trial and research aspects have been discussed [1–6]. Generally,PLA has high strength, high modulus and high optical transpar-ency, comparable with polystyrene. However, PLA is brittle andpresents low thermal stability and moderate gas barrier properties,which are all important properties for packaging applications [6].Additionally, PLA is well known to have low melt strength, whichis a drawback, especially if blow molding or film extrusion are tobe used as processing methods [1]. In order to overcome theseproblems a number of studies have been carried out where plasti-cizers, polymers, layered silicates and other inorganic particleshave been used [7–10]. But to keep PLA composite as green as pos-sible, the additives should also be bio-based and biodegradable.

Cellulose-based fibers are interesting bio-based and biodegrad-able additives and have shown to be able to improve the propertiesof PLA by affecting the crystallinity, mechanical properties and

thermal properties [11,12]. However, when film and packagingapplications are of interest, nanosized cellulose becomes a betteroption to improve the PLA because improved mechanical proper-ties can be reached at low reinforcement level [13,14]. Further-more, nanosized reinforcements generally have lower impact onthe optical properties of the matrix polymer.

Regarding the processing of nanocomposite, most studies havebeen done using solvent casting; however, if industrial manufac-turing of PLA films is the final target, the melt-compounding pro-cess becomes the most interesting technique, since the finalproduct can be easily shaped by blow molding or compressionmolding [13]. However, there are some challenges to overcomewhen using nanosized cellulose in extrusion processes. The mostimportant of these challenges are the difficult feeding of thematerial into the extruder and the tendency of nanosized celluloseto agglomerate when dried. These agglomerates cannot bere-dispersed during the extrusion process.

One possibility for overcoming these problems is to use liquidfeeding during compounding. Liquid feeding involves dispersingthe nanoparticles in a liquid medium, such as water, plasticizerand/or solvents, and then pumping the prepared suspension intothe extruder during compounding. Water and solvents should beremoved during the process [13]. It was reported for the firsttime, in 2006, that extrusion using liquid feeding of cellulosenanocrystals (CNC) is a suitable process to prepare PLA–CNC

http://dx.doi.org/10.1016/j.compscitech.2014.11.012

0266-3538/ 2014 Elsevier Ltd. All rights reserved.

⇑ Corresponding author. Tel.: +46 9 20 49 3371.

E-mail address: [email protected] (K. Oksman).

Composites Science and Technology 106 (2015) 149–155

Contents lists available at ScienceDirect

Composites Science and Technology

j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / c o m p s c i t e c h


mailto:[email protected]


http://www.sciencedirect.com/science/journal/02663538

http://www.elsevier.com/locate/compscitech

http://www.elsevier.com/locate/compscitech

http://www.sciencedirect.com/science/journal/02663538


mailto:[email protected]


http://-/?-

http://-/?-

http://-/?-

http://-/?-

http://crossmark.crossref.org/dialog/?doi=10.1016/j.compscitech.2014.11.012&domain=pdf

http://-/?-



nanocomposites. This initial study showed that the selection of theliquid medium is important, since the medium must not degradePLA during the processing [14]. A similar approach was used withcellulose acetate butyrate (CAB) as matrix, where a mix of plasti-cizer and ethanol was used as liquid medium. The obtained nano-composite exhibited very good mechanical properties and welldispersed nanocrystals, demonstrating that the process was suc-cessful [15].

So far, no publications about the preparation of nanocompositesbased on cellulose nanofibers using extrusion and liquid feedinghave been found. It is noteworthy that the CNF have a tendencyto form more viscous suspensions than the cellulose nanocrystals,making the preparation of cellulose nanocomposites using extru-sion and liquid feeding of CNF more challenging.

Our aim in this study was to use a strategy similar to that usedfor processing CAB–CNC nanocomposites [15] to prepare PLA–CNFnanocomposites with improved toughness. It was of interest to seeif liquid feeding works as well with cellulose nanofiber suspen-sions as previously shown with nanocrystals [14,15]. The optical,thermal and mechanical properties of neat PLA, plasticized PLAsand plasticized PLA–CNF nanocomposite were evaluated. More-over, studies about the interaction between CNF and GTA plasti-cizer and about the fractured surfaces morphology wereperformed to understand the toughening mechanisms.

2. Experimental

2.1. Materials

High molecular weight grade PLA in pellet form supplied byFUTERRO (Escanaffles, Belgium) was used as matrix. Bleachedbanana waste pulp, prepared at Pontifical Bolivarian University,

Medellin, Colombia, was used as starting material for the isolationof CNF. These nanofibers were used as reinforcement for PLA. Glyc-erol triacetate in liquid form (GTA) (P99%, Mw: 218 g/mol) wasused as plasticizer and was purchased from Sigma–Aldrich (Stock-holm, Sweden), and technical acetone from WVR (Stockholm, Swe-den) was used as solvent.

2.2. Processing of cellulose nanocomposites

2.2.1. Preparation of cellulose nanofibers and suspension for liquid

feeding

Cellulose nanofibers were isolated from banana waste pulpusing mechanical fibrillation. As an initial step, the bleached pulp(2 wt%) was dispersed in distilled water using a high shear mixer(L4RT, Silverson, England). The fiber suspension was ground at1000 rpm for 120 min using a super mass colloider, MasukoMKZA10–20J, Japan. A drop of diluted suspension was placed ona freshly cleaved mica plate and dried prior to AFM study. Thenanofiber diameters were measured from AFM height imagesusing the Nanoscope V software and Fig. 1(b) shows diametersvarying between 10 and 60 nm. The suspension preparation for

liquid feeding is illustrated in Fig. 1(a). The prepared aqueousCNF suspension of 2 wt% was concentrated to maximum cellulosecontent of 9.5 wt% by centrifugation. These concentrated CNF werethen diluted with acetone with the ratio 1:5 water to acetone(w:a), since the used plasticizer is not totally soluble in water. Afterthat, a specific amount of GTA plasticizer was added according tothe target composition of the final materials and mixed with mag-netic stirring and finally ultrasonicated. The final composition of the feeding suspension was 1.3 wt% CNF, 25.1 wt% GTA, 12.3 wt%H2O and 61.3 wt% acetone. The visual appearance and an opticalmicroscopy image (Leizt Dialux 20; Wetzlar, Germany) of this

Fig. 1. (a) Schematic representation of the suspension preparation for liquid feeding, (b) height AFM image and fiber diameter measurements from this image and amplitudeAFM image of the isolated CNF and (c) visual appearance and an optical microscopy image of the suspension for liquid feeding.

150 N. Herrera et al. / Composites Science and Technology 106 (2015) 149–155



suspension can be seen in Fig. 1(c). The optical image shows thatthe nanofibers were homogenously distributed in the liquid andGTA plasticizer was dissolved in the w:a mixture without phaseseparation. The viscosity of the suspension was measured to14.6 mPa s, which is 15 times higher than the viscosity of water;however, this viscous liquid was pumpable with a peristalticpump. This suspension was fed into the extruder to prepare thePLA nanocomposite, and pure GTA and GTA dissolved in the same

w:a mixture were fed to prepare the control samples.

2.2.2. Extrusion using liquid feeding

Plasticized PLA–CNF nanocomposite was prepared by com-pounding extrusion followed by compression molding accordingto a slightly modified procedure reported earlier by Oksman andco-workers [14]. A co-rotating twin-screw extruder (ZSK-18MEGALab, Coperion W&P, Stuttgart, Germany) with a gravimetricfeeder (K-tron, Niederlenz, Switzerland) for the neat PLA and aperistaltic pump (PD 5001 Heidolph, Schwabach, Germany) forthe liquid feeding of CNF and plasticizer were used. The total mate-rial throughput was 3 kg/h, the feeding rate of the PLA was 2.3–2.4 kg/h and the pumping rates of GTA, GTAw:a and GTAw:a/CNF

were 0.6, 2.3 and 2.4 kg/h, respectively. Neat PLA and two plasti-cized PLA materials were also prepared as control samples to studythe effect of the plasticizer and the solvents. Both the PLA and thecorresponding liquid for each material were fed at the main port of the extruder. The extrusion was carried out with a screw speed of 300 rpm and temperature profile ranged from 170 C to 200 Cfrom the feeding zone to the flat profile die. Atmospheric ventingat zone 2 and 4 was used to remove the water and acetone vaporand vacuum venting at the end of the extruder was used to evacu-ate the entrapped air and solvent.

For characterization of the material properties, films of com-pounded materials were prepared by compression molding usingan LPC-300 Fontijne Grotnes press (Vlaardingen, Netherlands).About 5–6 g of the material was placed between metal plates

and then in the hot press at 190 C for 3.5 min without pressure,and then for 30 s under a pressure of 2.5 MPa. The plates were

removed from the mold and cooled until they reached room tem-perature (20 min), and then the film was removed. The thicknessof the films was 0.15 mm and they were kept under ambient con-ditions for 1 day before testing. Fig. 2 illustrates the extrusion pro-cess with liquid feeding and Table 1 shows the composition of the

prepared materials.

3. Characterization

3.1. Differential scanning calorimetry

The thermal properties of the prepared films were investigatedusing a Mettler Toledo DSC822e (Schwerzenbach, Switzerland) dif-ferential scanning calorimeter. All DSC scans were performed on5 mg of material placed in a hermetic aluminum pan and wererun under N2 atmosphere from 30 C to 200 C at a heating rateof 10 C/min. The glass transition temperature (T g ), cold crystalliza-tion temperature (T cc ), cold crystallization enthalpy (DH cc ), meltingtemperature (T m) and the melting enthalpy (DH m) of the materials

were determined from the thermograms. The degree of crystallin-ity ( X c ) of the materials was estimated using the following equation

X c ¼ DH m DH cc

DH m

100

w ð1Þ

where DH m and DH cc are the enthalpies of melting and cold crystal-lization, respectively; DH m is the enthalpy of melting for 100% crys-talline PLA material (93 J/g) [7,11] and w is the weight fraction of PLA in the sample. The pre-melt crystallization was subtracted fromthe melting enthalpy where necessary.

3.2. Optical properties

Light transmittance of the materials was measured using a Per-kin Elmer UV/Vis Spectrometer Lambda 2S (Überlingen, Germany).The scan was carried out from 200 nm to 800 nm with a scan speedof 240 nm/min.

3.3. Tensile testing

The mechanical properties of the prepared materials were mea-sured using a conventional Shimadzu AG-X universal testingmachine with a load cell of 1 kN, a gauge length of 20 mm and acrosshead speed of 2 mm/min. Strips of 5 mm 50 mm were pre-pared from the films. Seven samples of each material were testedand the average of the values was reported. The tensile strengthand the elongation at break were obtained directly from testing

results, while the tensile modulus and the work of fracture werecalculated from the stress–strain curves.

Fig. 2. Schematic image of extrusion process with liquid feeding.

Table 1

Compositions of the prepared materials (wt%).

Materials PLA CNF GTA

PLA 100 – –PLA–GTA 80 – 20a

PLA–GTAw:a 80 – 20a,b

PLA–GTAw:a-CNF 79 1a,c 20a,c

a Fed as liquid.b 1.7 kg/h w:a removed as vapor during extrusion.c GTA and CNF content in the suspension were 25.1 wt% and 1.3 wt% respectively,

and 1.8 kg/h w:a removed as vapor during extrusion.

N. Herrera et al. / Composites Science and Technology 106 (2015) 149–155 151

http://-/?-

http://-/?-

http://-/?-



3.4. Microscopy

The morphology of the fractured surfaces of tensile tested filmswas studied using a JEOL JSM-6460LV scanning electron micro-scope (SEM) (Tokyo, Japan). Samples were sputter coated with athin layer of gold to avoid charging.

4. Results and discussion

4.1. Extrusion of nanocomposites using liquid feeding

We have demonstrated that PLA–CNF nanocomposites can beprepared by extrusion using liquid feeding of CNF. The vapor gen-erated due to the evaporation of 73.6 wt% of liquid phase (w:a) wasevacuated during the process through the atmospheric ventingports and vacuum venting at the end. Parameters such as percent-age of liquid phase, liquid feeding rate, ratio w:a were controlled tobe able to carried out the process. The main challenges in liquidfeeding of CNF is the high viscosity of cellulose nanofibers evenin very low concentrations, also a specific extruder is needed. Aco-rotating twin screw extruder is the best choice of equipmentbecause its excellent degassing properties. We have also seen thatthe solvent content need to be kept as low as possible and boilingof liquid phase in the feeding port need to be avoided.

4.2. Thermal properties

DSC was used to study the crystallization behavior of the semi-crystalline PLA and how the plasticizer used and nanofibers affectit. Fig. 3 shows the DSC thermograms for neat PLA, plasticized PLAsand the PLA–CNF nanocomposite in temperature range of 30–200 C. Thermal properties as well as the degree of crystallinity( X c ) of the materials are seen in Table 2. Neat PLA shows glass tran-sition peak, cold crystallization peak and melting peak in the heat-

ing scan in Fig. 3(a) with respective values of T g = 62.8 C,T cc = 99.0 C and T m = 174.5 C, while the plasticized PLAs and PLAnanocomposite only show the Tm peak. Addition of 20 wt% of GTA to neat PLA resulted in a slight decrease in the Tm, from174.5 C to 169.1 C; which is expected to be due to the increasedmolecular chain mobility. For the same reason, a clear increase inthe crystallinity of the neat PLA is seen, from 23% to 60%. Thisincreased crystallinity might explain why the glass transition peakwas not detectable on the DSC thermograms for the plasticizedPLAs or for the PLA nanocomposite. When the degree of crystallin-ity is increased in a polymer, the intensity of the glass transition isdecreased together with a broader transition temperature range,

resulting in a minor transition which becomes difficult to detectusing DSC [16].

It is known that molecular chain mobility is increased in thepresence of plasticizer, and that can also enhance the crystalliza-tion process [7]. Further, Mathew and Oksman [11] and Suryaneg-ara et al. [17] reported that reinforcements such asmicrocrystalline cellulose, cellulose fibers, wood flour and micro-fibrillated cellulose can act as nucleating agents for crystallizationof PLA. In this study, the used CNF were also seen to act as nucle-ation agent, as is seen in the DSC cooling scans at 20 C/min inFig. 3(b). PLA–CNF nanocomposite exhibits a clear melt crystalliza-tion peak, indicating that the crystallization is at 89.5 C due to thepresence of CNF, while plasticized PLAs did not show a melt crys-tallization peak. In contrast to that, the further addition of 1 wt% of CNF to the plasticized PLA generated a slight drop in the crystallin-ity from approx. 60% to 55%. Halász and Csóka [18] reported similarresults for plasticized PLA with 10 wt% of poly(ethylene glycol) andits cellulose composites. In this current work, the decrease in thedegree of crystallinity in the PLA nanocomposite is likely to bedue to a restriction phenomenon in the bulk crystallization processcaused by the presence of CNF and this phenomenon seems to bemore predominant than the nucleating effect of those CNF [11].

4.3. Optical transparency and visual appearance

If the size of individual nanoparticles is smaller than the wave-

length of visible light, transparency can be used as an indicator of the dispersion of the nanoparticles in a matrix [13]. The UV-spec-troscopy in Fig. 4 shows that the light transmittance of the neatPLA, plasticized PLAs and PLA–CNF nanocomposite in the centerof the visible light spectrum (550 nm) is 84%, 33% and 20%, respec-tively, demonstrating that the GTA plasticizer decreases the lighttransmittance of neat PLA approx. 50%. This is expected, becausethe plasticized PLA is more crystalline than the neat PLA, as previ-ously shown with DSC. However, further addition of CNF decreasedthe light transmittance by only 13% (from 33% to 20%), indicatinggood dispersion of the CNF. It can also be seen that the transmit-tance of the PLA nanocomposite in the UV region (below 400 nm)

Fig. 3. DSC curves for the prepared materials (a) heating at 10 C/min and (b) cooling at 20 C/min.

Table 2

DSC data and degree of crystallinity of the prepared materials.

Materials T g (C) T cc (C) DH cc

(J/g)T m (C) DH m

(J/g) X c (%)

PLA 62.8 99.0 23.6 174.5 45.5 23.6PLA–GTA – – – 169.1 44.5 59.8PLA–GTAw:a – – – 167.8 44.2 59.5PLA–GTAw:a-CNF – – – 168.1 40.1 54.5




reached values below 10%, indicating that this nanocompositematerial could be used in applications such as food packagingand agricultural bags, where UV protection is needed. Fig. 4 alsoshows that the neat PLA film is very transparent, while plasticizedPLAs and PLA nanocomposite films look translucent. The PLA nano-

composite film kept a transparency close to that of the plasticizedPLA, as seen in the UV-spectroscopy results. No visible agglomera-tions were observed in the PLA nanocomposite.

4.4. Mechanical properties

The mechanical properties of the prepared materials arereported in Table 3 and the representative stress–strain curvesfor the samples are shown in Fig. 5. The results show that the addi-tion of 20 wt% of GTA had a large effect on the mechanical proper-ties of the neat PLA. The elongation at break as well as the work of fracture increased, while the tensile modulus and the tensilestrength decreased (see Table 3). The addition of 20 wt% pureGTA to neat PLA raised the elongation at break by 678% and the

work of fracture by 389%. However, the elongation at break andthe toughness of the neat PLA were more affected when w:a mix-ture was used in feeding of GTA, resulting in increases by 965% and467%, respectively. These results suggest that the solvent mightfacilitate the mixing of the plasticizer and likely play a role in plas-ticizing the neat PLA. Furthermore, the addition of only 1 wt% of CNF increased the elongation at break and the work of fractureby 27% and 57%, respectively, compared to PLA–GTAw:a. Otherstudies have also shown an increase in the elongation at break withthe addition of nanoparticles; for example, Jiang et al. [9] reportedthat the elongation at break of PLA increased by 33% with the addi-tion of 5 wt% and 7.5 wt% of CaCO3 nanoparticles and alsoincreased by 433% with the addition of 2.5 wt% of montmorillonitenanoparticles. Li et al. [10] also reported an increase in the elonga-

tion of the PLA from 7.9% to 209.7% with the addition of 1 wt% of organically modified rectorite.

Usually, the addition of reinforcements in polymers decreasesthe elongation at break and the work of fracture [17,19]. In thisstudy the addition of CNF increased both elongation at break aswell as work of fracture. This may be due to the plasticization of

the nanofibers, as reported in other studies [20], together withthe slippery effect of the nanofibers in the PLA matrix if the plasti-cizer is located on the fiber surface, as reported in previous studies[14], and/or a massive crazing effect, which could be induced bythe presence of celluloses nanofibers. The latter hypothesis arosefrom the analysis of the visual appearance of the tensile testedsamples shown in Fig. 5. Here, it is evident that the plasticized PLAsand the PLA–CNF nanocomposite show stress whitening withoutnecking after the test, the whitening being stronger in the PLA–GTAw:a-CNF sample. It is known that stress whitening of polymersis attributed to formation of voids and/or intrinsic crazing [21] andthat massive crazing is a toughening mechanism of polymers [9].

For better understanding of the first hypothesis, a study aboutthe interaction between the plasticizer and the cellulose nanofibers

was performed. Plasticized CNF networks (also called cellulosenanopapers) with different amounts of GTA were prepared andthe mechanical properties were studied. The tensile tests showedthat the addition of plasticizer up to 20 wt% decreased the tensilestrength of the cellulose nanopaper without increasing the elonga-tion at break. This denotes that the plasticizer was present amongthe fibers of the nanopaper, but neither made the cellulose nanof-ibers more extensible nor created a slippery effect. However,increasing the added plasticizer to 57 wt% tripled the elongationat break of the cellulose nanopaper and strongly decreased the ten-sile strength, indicating that the slipping effect between the cellu-lose nanofibers occurred only in presence of high amounts of plasticizer (probably over 50 wt%). From this study about the inter-action between CNF and GTA, it was concluded that the presence of

a high amount of plasticizer coating the nanofibers created a slip-page between them, with no plasticizing effect on the own cellu-lose nanofibers. It was assumed that the cellulose nanofibersadopt a similar behavior inside the nanocomposite, where a highamount of plasticizer was coating the fibers and thus creating aslippage at the nanofiber/matrix interface. The high amount of plasticizer surrounding the cellulose nanofibers in the PLA nano-composite could be explained by a possible migration of GTA fromthe crystalline regions to the amorphous regions, since plasticizersare more miscible with amorphous phases [7,22].

4.5. Microstructure of fracture surfaces

Fig. 6 shows SEM images of the fractured surfaces from tensile

test of the neat PLA, plasticized PLAs and PLA nanocomposite.The fractured surfaces of the samples were analyzed in order to

Fig. 4. Transmittance spectra and visual appearance of the prepared films.

Table 3

Mechanical properties of the prepared materials.

Materials Tensilestrength

(MPa)

Tensilemodulus

(GPa)

Elongationat break

(%)

Work of fracture

(MJ/m3)

PLA 62.8 ± 3.7a 3.8 ± 0.1a 2.3 ± 0.3a 0.9 ± 0.1a

PLA–GTA 27.8 ± 0.8b,c 1.2 ± 0.1b 17.9 ± 2.3b 4.4 ± 0.7b

PLA–GTAw:a 26.2 ± 1.0b 0.8 ± 0.1c 24.5 ± 10.9b,c 5.1 ± 1.8b

PLA–GTAw:a-CNF 28.8 ± 0.4c 0.8 ± 0.0c 31.1 ± 8.3c 8.0 ± 2.4c

Same superscript letters within the same column are not significantly different at5% significance level based on the ANOVA and Tukey-HSD multiple comparison test.

Fig. 5. Stress–strain curves of the prepared materials and the photographic imagesof the respective pre-tested and post-tested samples.


http://-/?-

http://-/?-

http://-/?-



visualize the effect of the plasticizer and the nanofibers on the frac-ture behavior as well as the effect of the plasticizer on the distribu-tion and the dispersion of the CNF in the matrix. From the images,it is seen that the fracture of the neat PLA ( Fig. 6(a)) is brittle asexpected and in agreement with the tensile test results, where neatPLA exhibited an elongation at break of only 2.3%. When compar-

ing the fracture of the neat PLA film (Fig. 6(a)) to the plasticizedPLAs, both with and without w:a (Fig. 6(b)) and (c)), it is possibleto see in both cases a combination of flat areas and areas with lar-ger plastic deformation. This type of fracture appears when somecrazes have been formed in the material and then break downalong the interface between the crazed and uncrazed material[21]. This indicated that the density of crazes was higher in theplasticized PLAs than in neat PLA; therefore, the fracture was lessbrittle. This is supported by evidence of stress whitening of theplasticized PLAs, as shown in Fig. 5, and is in agreement with thetensile test results, where plasticized PLAs showed higher elonga-tion at break than neat PLA. On the other hand, any difference onthe fracture surfaces between both plasticized PLAs, with andwithout w:a; was observed (Fig. 6(b) and (c)), indicating that the

solvents did not have a notable effect on the fracture surface mor-phology. Finally, the comparison of the PLA–CNF nanocomposite

fracture surface with the plasticized PLAs clearly shows that thefracture surface of the nanocomposite is formed for numerousfibrils, caused by larger amount of crazes, indicating a massivecrazing effect which resulted in a ductile fracture. This is in accor-dance with the strongest whitening effect that was noticed on thetensile test sample of the PLA–CNF nanocomposite (Fig. 5) togetherwith the highest elongation at break (31.1%) of the same material.Craze formation is affected by the degree of crystallinity; when itdecreases, the amount of amorphous areas where craze nucleationcan take place first is increased. Moreover, it is also affected by thespherulite size; a fine spherulite structure favors the craze nucle-ation, while a coarse spherulite structure prevents the formationof crazes [21]. In this study, DSC results showed that plasticizedPLA–CNF nanocomposite has slight lower crystallinity than plasti-cized PLAs. In addition, it was expected that this material has smal-ler spherulite size than the plasticized PLAs, since the presence of CNF may reduce the space for spherulite growth [11].

5. Conclusions

We have shown that PLA–CNF nanocomposites can be success-fully prepared with a co-rotating twin-screw extruder when liquidfeeding of cellulose nanofibers together with a plasticizer is used.To be able to carry out the extrusion process, parameters such asamount of liquid phase and liquid feed rate should be controlledand effective venting and vacuum ports are needed. Resultsshowed that liquid feeding of 20 wt% GTA to neat PLA resulted inplasticized PLA with increased degree of crystallinity, decreasedoptical transparency, improved elongation at break from 2.3% to25% and improved work of fracture from 0.9 to 5.1 MJ/m3. Furtheraddition of a small amount (1 wt%) of CNF resulted in a plasticizedPLA–CNF nanocomposite with lower degree of crystallinity in com-parison with the plasticized PLA and slightly lower optical trans-parency, indicating that the CNF were well dispersed. Thepresence of CNF showed a positive effect on the elongation atbreak, as well as on the work of fracture of the plasticized PLA, with

a further increase of these values from 25% to 31% and from 5.1 to8.0 MJ/m3, respectively. Hence, the effect of CNF almost doublesthe effect of the plasticizer alone. It is believed that the higherelongation at break and the higher work of fracture of the plasti-cized PLA–CNF nanocomposite in comparison with neat PLA aswell as with the plasticized PLAs is due to the combination of theslippage of the nanofiber–matrix interface and the massive crazingphenomenon as a result of the presence of CNF. It was expectedthat CNF restrict the spherulite growth and therefore enhancethe nucleation of crazes, resulting in a massive crazing. Furtherstudy about the crystallization process, spherulite size and spher-ulite growth rates could be carried out for better understanding.

Acknowledgements

The authors gratefully acknowledge the European Commissionfor financial support under Grant Agreement No. 280786,FP7NMP-2011-ECLIPSE. The Pontifical Bolivarian University (UPB)Colombia is acknowledged for providing the purified/bleached cel-lulose pulp from banana rachis waste. The authors are also gratefulfor the University of Basque Country, Spain especially Prof. JalelLabidi and Asier M. Salaberria for their help with the DSCmeasurements.

References

[1] Lim LT, Auras R, Rubino M. Processing technologies for poly(lactic acid). ProgPolym Sci 2008;33:820–52.

[2] Madhavan NK, Nimisha NR, Rojan JP. An overview of the recent developmentsin polylactide (PLA) research. Bioresour Technol 2010;101:8493–501.

Fig. 6. Fractured surfaces morphology of the tensile tested films.


http://refhub.elsevier.com/S0266-3538(14)00408-4/h0005










[3] Mekonnen T, Mussone P, Khalil H, Bressler D. Progress in bio-based plasticsand plasticizing modifications. J Mater Chem A 2013;1:13379–98 .

[4] Rasal RM, Janorkar AV, Hirt DE. Poly(lactic acid) modifications. Prog Polym Sci2010;35:338–56.

[5] Yu L, Chen L. Polymeric materials from renewable resources. In: Yu L, editor.Biodegradable polymer blends and composites from renewable resources. New

Jersey, USA: John Wiley & Sons; 2009. p. 1–15 [chap. 1].[6] Soulestin J, Prashantha K, Lacrampe MF, Krawczak P. Bioplastics based

nanocomposites for packaging applications. In: Pilla S, editor. Handbook of bioplastics and biocomposites engineering applications, Part II. New

Jersey: Wiley Scrivener Publishing; 2011. p. 77–120 [chap. 4].[7] Ljungberg N, Wesslén B. Preparation and properties of plasticized poly(lactic

acid) films. Biomacromolecules 2005;6:1789–96.[8] Coltelli M, Maggiore ID, Bertoldo M, Signori F, Bronco S, Ciardelli F. Poly(lactic

acid) properties as a consequence of poly(butylene adipate-co-terephthalate)blending and acetyl tributyl citrate plasticization. J Appl Polym Sci2008;110:1250–62.

[9] Jiang L, Zhang J, Wolcott MP. Comparison of polylactide/nano-sized calciumcarbonate and polylactide/montmorillonite composites: reinforcing effectsand toughening mechanisms. Polymer 2007;48:7632–44.

[10] Li B, Dong F, Wang X, Yang J, Wang D, Wang Y. Organically modified rectoritetoughened poly(lactic acid): nanostructures, crystallization and mechanicalproperties. Eur Polym J 2009;45:2996–3003.

[11] Mathew AP, Oksman K, Sain M. The effect of morphology and chemicalcharacteristics of cellulose reinforcements on the crystallinity of polylacticacid. J Appl Polym Sci 2006;101:300–10.

[12] Oksman K, Skrifvars M, Selin J. Natural fibres as reinforcement in polylacticacid (PLA) composites. Compos Sci Technol 2003;63:1317–24.

[13] Oksman K, Mathew AP. Melt compounding process of cellulosenanocomposites. In: Oksman K, Mathew AP, Bismarck A, Rojas O, Sain M,editors. Handbook of green materials, vol. 2. Singapore: World ScientificPublishing; 2014. p. 53–68 [chap. 5].

[14] Oksman K, Mathew AP, Bondeson D, Kvien I. Manufacturing process of cellulose whiskers/polylactic acid nanocomposites. Compos Sci Technol2006;66:2776–84.

[15] Bondeson D, Syre P, Niska KO. All cellulose nanocomposites produced byextrusion. J Biobased Mater Biol 2001;1:367–71.

[16] Chartoff RP. Thermoplastic polymers. In: Turi EA, editor. Thermal

characterization of polymeric materials. San Diego, California: AcademicPress; 1981. p. 484–730 [chap. 3].

[17] Suryanegara L, Nakagaito AN, Yano H. The effect of crystallization of PLA on thethermal and mechanical properties of microfibrillated cellulose-reinforced PLAcomposites. Compos Sci Technol 2009;69:1187–92.

[18] Halász K, Csóka L. Plasticized biodegradable poly(lactic acid) based compositescontaining cellulose in micro- and nanosize. J Eng 2013:1–9.

[19] Jonoobi M, Harun J, Mathew AP, Oksman K. Mechanical properties of cellulosenanofiber (CNF) reinforced polylactic acid (PLA) prepared by twin screwextrusion. Compos Sci Technol 2010;70:1742–7.

[20] Majhi S, Nayak S, Mohanty S, Unnikrishnan L. Mechanical and fracturebehavior of banana fiber reinforced polylactic acid biocomposites. Int J PlastTechnol 2010;14:57–75.

[21] Friedrich K. Crazes and shear bands in semi-crystalline thermoplastics. In:Crazing in polymers. Springer-Verlag; 1983.

[22] Piorkowska E, Kulinski Z, Galeski A, Masirek R. Plasticization of semicrystallinepoly(l-lactide) with poly(propylene glycol). Polymer 2006;47:7178–88.


























































































http://-/?-




http://-/?-




http://-/?-





















2015 plasticized polylactic acid-cellulose nanocomposites prepared using melt-extrusion and liquid...

Documents