COMPOSITES

www.elsevier.com/locate/compscitech

Composites Science and Technology 67 (2007) 413–421

SCIENCE ANDTECHNOLOGY

Preparation and characterization of melt-extruded thermoplasticstarch/clay nanocomposites

Katherine Dean, Long Yu *, Dong Yang Wu

CSIRO Manufacturing and Infrastructure Technology, 37 Graham Road, Highett, 3190 Vic., Australia

Received 20 July 2006; received in revised form 1 September 2006; accepted 5 September 2006Available online 24 October 2006

Abstract

A series of gelatinized starch–clay nanocomposites which exhibit intercalated and exfoliated structures have been developed. Variousnanoclay dispersions were prepared (either by standard mixing or through the use of ultrasonics) prior to their combination with a highamylose content starch using high-speed mixing and extrusion technology. Intercalated and exfoliated type structures were observed inthe sheet extruded nanocomposites using X-ray diffraction and transmission electron microscopy (TEM). Due to the hydrophilic natureof the gelatinized starch nanocomposite a novel preparatory technique was developed to produce nano scale sections for TEM. A rangeof plasticiser levels were used in conjunction with different unmodified nanoclays (sodium montmorillonite (Na-MMT) and fluorohect-orite (Na-FHT)) having different cationic exchange capacities. It was shown that an optimum level of both plasticiser and nanoclayexisted to produce a gelatinized starch film with the highest levels of exfoliation, resulting in superior properties. The use of ultrasonicswas only advantageous in terms of clay dispersions at medium clay concentrations in the Na-MMT nanocomposites and higher clayconcentrations in the Na-FHT, most probably due to the difference in cationic exchange capacity; however when the level of clay, waterand starch was optimised an exfoliated structure was produced via standard mixing which exhibited comparable improvements inmechanical properties to ultrasonically treated samples.� 2006 Elsevier Ltd. All rights reserved.

Keywords: Thermoplastic starch; Nanocomposite; E. Extrusion; Montmorillonite; A. Nanostructure

1. Introduction

Throughout the world there is an increased awareness ofthe issue of plastics waste, in particular from packagingapplications and as such industries, and governments aredirecting resources into developing environmentallyfriendly materials in the world. So far, many of the materi-als developed are far from acceptable in their ability tomeet all the application requirements. Starch is one suchalternative. It is produced in plants and is a mixture of lin-ear amylose (poly-a-1,4-D-glucopyranoside) and branchedamylopectin (poly-a-1,4-D-glucopyranoside and a-1,6-D-glucopyranoside). The ratio of amylose to amylopectin var-ies with the source. Amylose is the minor component

0266-3538/$ - see front matter � 2006 Elsevier Ltd. All rights reserved.

doi:10.1016/j.compscitech.2006.09.003

* Corresponding author.E-mail address: long.yu@csiro.au (L. Yu).

(approximately 20%) of the starches and forms the amor-phous regions whereas the short branching chains of theamylopectin are predominantly responsible for the crystal-line properties of the starches [1–3]. More recently granularstarch has been processed using traditional extrusion tech-nology to produce thermoplastic starch using a range oftemperatures (generally from 100 to 200 �C), pressuresand plasticisers (more commonly water and/or glycerol)[4–10].

A number of researchers have investigated methods toimprove the properties of starch-based materials (includ-ing solvent cast, blended and extruded thermoplastictypes) using nanocomposites [11–25]. Papers publishedpreviously by Wilhelm et al. [11] discussing solvent castingmethods and by McGlashan and Halley [12], Avella et al.[13] and Kalambur [14,15] for starch nanocomposite andstarch–polyester nanocomposites, all report relatively

414 K. Dean et al. / Composites Science and Technology 67 (2007) 413–421

good dispersion of nanoclays leading to improved pro-cessing and mechanical properties. Park and Li et al.[16,17] reported the preparation and properties ofstarch/clay nanocomposites using a number of differentmontmorillonite clays (both unmodified and modified).Using a batch mixing process and high levels of plasticis-ers Park et al. and Li et al. [16,17] produced intercalatedstarch/clay samples which exhibited improved mechanicalproperties (in particular modulus) and improved barrierproperties. Huang et al. [18] also investigated the proper-ties of thermoplastic starch/clay nanocomposites using areduced level of plasticiser and found an improvementin tensile mechanical properties. In both cases [16–18] atwo step procedure was used in which the granulated ther-moplastic starch was melt blended with the clay. Chenand Evans [20] also observed an increase in modulususing both solvent casting and melt extrusion in thermo-plastic starch nanocomposites with good clay dispersion.Finally in one of the more comprehensive works, Fischeret al. [22] produced a series of thermoplastic starch/claynanocomposites via melt extrusion and solvent castingmethods, Fischer [22] showed that with high plasticisercontent exfoliated type structures could be formed (asseen by TEM and XRD).

The present work seeks to map the effects of clay, plas-ticisers and dispersion methods on dispersion of nanoclaysand the overall mechanical performance of extruded ther-moplastic starch nanocomposites, a field in which the pre-vious literature review shows only a few key publicationsreporting varying degrees of successful dispersion ofnanoclays and property improvement.

2. Experimental

2.1. Materials

Two types of clays have been used in this study, onebased on naturally occurring sodium montmorillonite clay(Cloisite Na+ – supplied by Southern Clay Products) – Na-MMT and the other a synthetic fluoromica (SomasifME100 – supplied by Uni-Coop Japan) – Na-FHT. Waterwas used as the plasticiser. The matrix polymer used in thisstudy was a high amylose cornstarch.

2.2. Formation of thermoplastic starch nanocomposites

The various nanoclays were either (1) dry blended(DB) with starch in a high speed mixer (HSM); (2) con-ventionally dispersed (CD) in water using conventionalmixers prior to blending with starch in a HSM or (3)dispersed in water using ultrasonics (U) prior to blendingwith starch in a HSM . In the conventionally mixedcases a standard bench mixer was used to dispersenanoclays (1500 rpm process time 1 h). In the ultrasoni-cally treated cases a Branson sonifier (model 250 cell dis-ruptor) 200 W with a maximum mechanical vibrationfrequency of 20 kHz was used (process time 1 h) to aid

with intercalation and exfoliation of the clay and plasti-cisers. A standard 0.5 in. diameter flat horn tip was usedat approximately 40% output. The water (in the case ofthe dry blended formulation) or water/clay dispersionswere then added drop-wise to the starch in a high-speedmixer, prior to extrusion. A Theysohn co-rotating twinscrew extruder, with diameter 30 mm and L/D 40, wasused to process the starch nanocomposities (Table 1 forlist of formulations) using a profile producing a melttemperature of 110 �C. The starch nanocomposite mate-rial was extruded directly into sheet form via a 0.5 mmdie attached to the extruder. The extrusion of each for-mulation was duplicated to ensure reproducibility.

3. Characterization

X-ray diffraction (XRD) was used to monitor the d0 0 1

spacing corresponding to the interlayer spacing of the clay.The XRD measurements were performed on the starchnanocomposite sheet samples using a Bruker D8 Diffrac-tometer operating at 40 kV, 40 mA, Cu Ka radiationmonochromatised with a graphite sample monochromator.A diffractogram was recorded between 2h angles of 2� and25�. The d1 0 0 at 20� 2h was used as an internal standard inmany cases to standardize diffractograms in relation to thepercentage of clay in each system.

The starch nanocomposites were imaged using a Jeol100S TEM using an accelerating voltage of 100 keV atmagnifications of 25000–100000 times to study dispersionsof clay particles. The samples were stabilized by exposureto osmium tetroxide (OsO4) vapour for 2 h, followed byimmersion in uncured epoxy resin for 12 h (to allow diffu-sion of epoxy into the samples) and subsequent curing andmicrotoming at room temperature.

Mechanical testing was undertaken as outlined inASTM 638 using dog-bone shaped samples which werecut from the thermoplastic starch sheet immediately afterextrusion. A United Tensile Testing instrument STM10with a 1 kN load cell in tensile mode was used to measuremodulus, tensile strength and elongation for the thermo-plastic starch sheet. A test speed of 10 mm/min was usedand a minimum of 10 samples were tested for each formu-lation. Baseline samples (not containing nanoclays) weretested at the beginning of each set of samples to ensurereproducibility. Samples were conditioned at 22.5 �C and65% relative humidity for 14 days until a relatively stableamount of retrogradation was achieved.

4. Results and discussion

4.1. Dispersion of nanoclays using ultrasonics

The neat Na-MMT (as received from the manufacturer)had a d0 0 1 spacing (or intergallery separation) of 12.2 A.When the Na-MMT was initially dispersed in water usinga high ratio of Na-MMT to water (1:4 wt:wt) the interlayerspacing of the clay increased 15.6 A corresponding to

Table 1Starch nanocomposites formulations which were produced via meltextrusion (for two series of clays – Na-MMT and Na-FHT)

Sample Water (wt%) Clay (wt%)

Low water contentLow water–high clay(clay:water 1:4 LWHC)

13 3.2

Low water–medium clay(clay:water 1:5 LWMC)

13 2.6

Low water–low clay(clay:water 1:13 LWLC)

13 1

Medium water contentMedium water–medium clay(clay:water 1:9 MWMC)

18 2

High water contentWater–high clay(clay:water 1:6 HWHC)

20 3.2

High water–low clay(clay:water 1:13 HWLC)

20 1.5

Table 3Na-FHT:H2O (distilled) d0 0 1 spacing (A) for different ratios of clay andwater at different sonication times

System XRD peaks

Neat Na-FHT 12.6 and 10.8Water:Na-FHT: 4:1 sonified for 30 min 15.4 and broad

amorphous peak

K. Dean et al. / Composites Science and Technology 67 (2007) 413–421 415

approximately two layers of water molecules in the Na-MMT clay interlayers. As the sonication time wasincreased a peak at 35 A was observed and some broaden-ing of the d0 0 1 peak at around 15.6 A was seen. The extentof the increase in intergallery separation was limited by thehigh viscosity of the pastes making complete dispersion viasonication difficult (Table 2). With the further addition ofwater (1:5 wt:wt Na-MMT:water) a greater amount ofintercalation was observed, with broadening of the XRDpeak (originally at 10 A) to 35–40 A after 15 min of sonica-tion. Further decreasing the Na-MMT to water ratio (1:10wt:wt) increased the dispersion of the clay and enabled thesonication treatment to be more efficient primarily due tothe lower viscosity of the dispersion.

Similar results were observed for the dispersions of Na-FHT in water (Table 3). The neat Na-FHT (as received

Table 2Na-MMT:H2O (distilled) d0 0 1 spacing (A) for different ratios of clay andwater at different sonication times

Sonication time 1:4 1:5 1:10

0 min 15.7 15.6 15.5

5 min 15.4 15.6 15.510 min 15.6 32.6 33

Some broadeningaround 32–35 A

15.7 15.5

15 min 15.7 32.6 33Some broadeningaround 32–35 A

15.62 15.6

20 min 15.7 32 45–33 broadpeak

15.5129 15.7

25 min 15.5 32 4615.58 15.3

30 min Some broadeningup to 60 A

– 62.2

15.6 15.4

from the manufacturer) exhibited a d0 0 1 spacing of10.8 A and another peak at 12.6 A. This peak at 12.6 Amay be associated with bound water, as when the claywas dried under vacuum at 100 �C for 12 h this peak disap-peared. A single layer of water molecules expands the inter-layer distance to 12.6 A, a second water layer to 15.2 A andso on. When dispersed in water up to a ratio of 10 partswater to 1 part clay, both the Na-MMT and Na-FHT clayshad a peak at around 15.2–15.6 A corresponding to thepresence of two layers of water molecules in the interlayergallery.

4.2. Formation of thermoplastic starch nanocomposites

The various starch/clay/water formulations (Table 1)were processed using the three techniques of dry blending(DB), conventional dispersion (CD) and ultrasonic disper-sion (U) followed by extrusion and the resulting XRD datahas been tabulated in Table 4 for Na-MMT and Table 5 forNa-FHT.

The key interaction between unmodified layered silicatesand water is the ion–dipole interaction between the sodiumion (located in the intergallery space of the clay) and thedipole of the water molecule (Fig. 1), this kind of mecha-nism has been implied by other authors [11,26]. The hydro-xyl groups of the starch could also interact directly with thesodium ion of the clay (as the water molecules do) or withthe edge hydroxyl groups of the clay [18] making a verycompatible system.

Figs. 2 and 3 illustrate some key XRD data for theNa-MMT nanocomposites. With a high to medium Na-MMT concentration (2.6–3.2 wt%) and low water con-centration (13 wt%) some intercalation was observed inthe systems (intergallery expansion up to 20 A (seeFig. 2)), however the method of dispersion did not havea significant effect and in some cases dry blendingappeared more successful than the use of ultrasonics.As the level of Na-MMT was reduced down to 1 wt%the intensity of peaks decreased significantly giving astrong indication that the clay platelets were exfoliated.As the level of water in the thermoplastic starch nano-composite was increased a greater amount of Na-MMTwas able to be exfoliated into the system without theuse of ultrasonics (Table 4).

Similar results were observed in the Na-FHT thermo-plastic starch nanocomposites (Table 5). Complete exfolia-tion was not clearly observed using this type of clay for anumber of reasons. As Na-FHT is a synthetic fluorohector-ite, it has a much more uniform structure as compared with

Table 4XRD peaks and relative intensities for Na-MMT series of nanocomposites

Sample Water (%) Clay (%) XRD peaks (A) and relative intensity ( )a fordifferent dispersion methods

Low water contentNa-MMT 0 100 12.2 A (1.0) (Dry)

Low water–high clay (clay:water 1:4 LWHC) 13 3.2 11.9 (0.56), 20.0 (0.12) (Ultrasonic)12.2 (0.17), 20.0 (0.1) (Dry blended)– (Dispersed)

Low water–medium clay (clay:water 1:5 LWMC) 13 2.6 11.9 (0.68), 20.0 (0.17) (Ultrasonic)– (Dry blended)– (Dispersed)

Low water–low clay (clay:water 1:13 LWLC) 13 1 11.9 (0.1), 20.0 (0.15) (Ultrasonic)11.9 (0.5) (Dry blended)11.9 (0.35) broad peak (Dispersed)

Medium water contentMedium water–medium clay (clay:water 1:9 MWMC) 18 2 20.0 (0.05) (Ultrasonic)

12.4 (0.5) broad peak (Dry blended)12.4 (0.05) (Dispersed)

High water contentHigh water–high clay (clay:water 1:6 HWHC) 20 3.2 12.2 (0.85) sharp peak (Ultrasonic)

12.2 (0.19) (Dry blended)– (Dispersed)

High water–low clay (clay:water 1:13 HWLC) 20 1.5 24.3 (0.18) (Ultrasonic)17.5 (0.17) (Dry blended)17.5 (0.06) (Dispersed)

a Relative intensity is expressed in brackets as a fraction of intensity of the neat Na-MMT (800 counts).

Table 5XRD peaks and relative intensities for Na-FHT series of nanocomposites

Sample Water (%) Clay (%) XRD peaks (A) and relative intensity ( )a fordifferent dispersion methods

Low water contentNa-FHT 0 100 10.8 (1.0), 12.6 (0.05) Dry

Low water–high clay (clay:water 1:4 LWHC) 13 3.2 12.1(0.49) (Ultrasonic)12.2 (0.6) (Dry blended)– (Dispersed)

Low water–medium clay (clay:water 1:5 LWMC) 13 2.6 12.1 (0.45) (Ultrasonic)– (Dry blended)– (Dispersed)

Low water–low clay (clay:water 1:13 LWLC) 13 1 12.1 (0.02) (Ultrasonic)12.1 (0.08) (Dry blended)12.1 (0.07) (Dispersed)

Medium water contentMedium water–medium clay (clay:water 1:9 MWMC) 18 2 12.2 (0.36),14.7 (0.07) (Ultrasonic)

12.2 (0.46), 14.7 (0.10) (Dry blended)12.2 (0.24), 14.7 (0.06) (Dispersed)

High water contentHigh water–high clay (clay:water 1:6 HWHC) 20 3.2 12.2 (0.47) (Ultrasonic)

12.2 (0.27) (Dry blended)– (Dispersed)

High water–low clay (clay:water 1:13 HWLC) 20 1.5 12.1 (0.27) (Ultrasonic)12.2 (0.23) (Dry blended)12.2 (0.16) (Dispersed)

a Relative intensity is expressed in brackets as a fraction of intensity of the main peak for neat Na-FHT (11 000 counts).

416 K. Dean et al. / Composites Science and Technology 67 (2007) 413–421

O

H

H RO

H

M+

Fig. 1. Ion–dipole interaction between clay intergallery metal ion andwater molecule.

151050

1000

2000

3000

4000

5000

H%31 ,yalc %12

)cinosartlu( O

H%31 ,yalc %12

)desrepsid lanoitnevnoc( O

H%31 ,yalc %12

)dednelb yrd( O

H %31 ,yalc %2.32

)cinosartlu( O

TMM-aN taen

H%31 ,yalc %6.22

)cinosartlu( O

H %31 ,yalc %2.32

)dednelb yrd( 0

coun

ts

2˚

Fig. 2. XRD traces for the series of low water content (13 wt%) Na-MMT/ gelatinized starch nanocomposites with various clay loadings anddispersion methods.

151050

500

1000

1500

2000

2500

3000

3500

4000

H %81 ,yalc %22

)cinosartlu( O

H %81 ,yalc %22

)desrepsid lanoitnevnoc( O

t Na-MMTaen

H %81 ,yalc %22

)dednelb yrd( O

coun

ts

2˚

Fig. 3. XRD traces for the series of medium water content (18 wt%) Na-MMT/gelatinized starch nanocomposites with 2 wt% clay.

K. Dean et al. / Composites Science and Technology 67 (2007) 413–421 417

the naturally occurring Na-MMT; as a consequence of thisorder the d0 0 1 diffraction peak is much sharper and moreintense in the Na-FHT (11000 counts c/f 800 for Na-MMT). So although there was a significant decrease inintensity of the d0 0 1 peak corresponding to the intergalleryspacing of the clay it did not disappear completely (TEMimages of the Na-FHT) samples do show what appearsto be close to complete exfoliation (Fig. 4a).

Other TEM images are shown in Fig. 4; these includethe medium water content (18 wt%), 2 wt% Na-MMTwith ultrasonic dispersion used (see Fig. 4b) showing anexfoliated type structure. As the clay content wasincreased to 3.2 wt% some agglomeration of clay plateletswas observed (although dispersion of platelets was stillgood), even in systems containing higher amounts ofwater (up to 20 wt%) (Fig. 4c). It was concluded fromthe XRD and TEM studies that a fine balance existedbetween intercalated and exfoliated structures and clayconcentrations.

4.3. Mechanical properties

Three-dimensional plots of the weight percentage ofclay, the weight percentage of water and the dispersionmethod used (Figs. 5–7Figs. 9–11) help to illustrate theeffects that each has on modulus, yield strength and breakelongation.

The most significant improvement in modulus for theNa-MMT is observed in the systems containing higherlevels of clay (2.6–3.2 wt%) for both dry blended andultrasonically treated samples, this is not unexpected asthe modulus tends to relate to the intrinsic properties ofthe components rather than dispersion and interfacialinteractions per se. Significant increases in modulus areeven seen in conventional clay–polymer nanocompositescontaining tactoids – often in sacrifice of tensile strength,elongation and toughness [27]. At higher clay concentra-tions conventional mixing was shown to be a little lessefficient than dry blending predominantly due to viscosityissues of the clay/water. Similar improvements wereobserved for yield strength (see Fig. 6), in which the addi-tion of a higher level of Na-MMT led to a greaterimprovement in strength. Once again at this high levelof clay (3.2 wt%) the ultrasonically treated samplesappeared to give better values at higher water loading(20 wt%) and the dry blended samples at low water load-ing (13 wt%) the latter may have also been attributed topoor mixing in the high viscosity water/clay systems lead-ing to poor clay dispersion.

The elongation at break measured for the Na-MMTthermoplastic starch samples was shown to decrease com-pared to the neat thermoplastic starch samples, this hasbeen observed by a number of other authors in otherpolymer systems [28,29]. The starch thermoplastic nano-composite systems did however differ from more conven-tional polymer nanocomposites in that the trend betweenclay content and elongation at break was not linear. Theelongation at break peaked at around 2 wt% clay. Thismay be explained by a number of different mechanisms.(1) Once melt extruded the thermoplastic starch beginsto undergo retrogradation (or re-crystallization), this pro-cess causes embrittlement and reduces the break elonga-tion. The inclusion of nanoclays may disrupt this re-crystallization process. (2) The addition of nanoclaysand their strong interaction with water molecules

418 K. Dean et al. / Composites Science and Technology 67 (2007) 413–421

(through ion–dipole interactions) may help retain mois-ture in the samples, leading to more plasticised materialwith a greater elongation at break. Fig. 8 illustrates therelationship between yield strength and elongation atbreak and gives some indication of the toughness of thethermoplastic starch nanocomposite materials. Withinerror the 3.2 wt% dry blended nanocomposite containing13 wt% water and the 3.2 wt% ultrasonically treatednanocomposite containing 20 wt% water show the bestratio of strength to elongation; as suggested previously(in XRD analysis) the dry blending of high clay concen-tration samples with low water content appeared to bethe most successful and the ultrasonic treatment of sam-ples containing high clay concentrations was only success-ful when a high level of water was also used. In systemscontaining higher levels of clays and lower levels of water,the main method of dispersion in the final thermoplasticstarch sheet materials was the extrusion process itself, dis-persion methods for the clay in water prior to the extru-sion step did not appear to have a significant effect.

Fig. 4. (a) TEM image of low water content (13 wt%) Na-FHT gelatinized staimage of medium water content(18 wt%) Na-MMT/gelatinized starch nanocomhigh water content (20 wt%) Na-MMT/gelatinized starch nanocomposites wit

Simliar results were also observed in the Na-FHT ther-moplastic starch nanocomposite samples where the mostsignificant improvement in modulus is observed in the sys-tems containing higher levels of clay (2.6–3.2 wt%) for bothdry blended and ultrasonically treated samples (see Fig. 9).Similar trends were also observed in the yield strength (seeFig. 10). Unlike the Na-MMT nanocomposites the yieldelongation did decrease with increasing clay content indi-cating the Na-FHT was not as capable of disrupting re-crystallization or did not bind the water molecules astightly into the matrix, this may be related to the cationicexchange capacity (CEC) of the clay. The CEC is definedby [30]:

CEC ¼ 1000n�M

ð1Þ

where n is the mean layer charge of the clay and �M is themean molecular weight for one formula unit.

The (CEC) of Na-MMT is 92 mequiv./100 g whereasthe CEC of Na-FHT is only 70–80 mequiv./100 g. The

rch nanocomposites 1 wt% clay (conventional dispersed 30 min); (b) TEMposites with 2 wt% clay (ultrasonically treated 2 h), and (c) TEM image of

h 3.2 wt% clay (ultrasonically treated 2 h).

0.00 5.

1.01 5.

2 0.2 5.

3 0.3.5 0

1000

2000

3000

4000

5000

6000

13wt%water U

18wt%water U

20wt%water U

13wt%water DB

18wt%water DB

20wt%water DB

13wt%water CB

18wt%water CB

20wt%water CB

Mod

ulus

(MP

a)

wt%clay

Fig. 5. Tensile modulus (MPa) of Na-MMT/gelatinized starch nanocom-posites for various clay loading, water loading and dispersion methods.

0.00 5.

1 0.1 5.

2 0.2.5

3 0.3.5 0

10

20

30

40

50

60

13wt%water U

18wt%water U

20wt%water U

13wt%water DB

18wt%water DB

20wt%water DB

13wt%water CB

18wt%water CB

20wt%water CB

Yi

led

Str

netgh

(MP

a)

w%t

c al y

Fig. 6. Yield strength (MPa) for Na-MMT/gelatinized starch nanocom-posites for various clay loading, water loading and dispersion methods.

0

1

2

3012

3

4

5

6

7

8

9

13wt%water U

18wt%water U

20wt%water U

13wt%water DB

18wt%water DB

20wt%water DB

13wt%water CB

18wt%water CB

20wt%water CB

bre

kael

onag

tion

(%)

wt %clay

Fig. 7. Elongation at break (%) for Na-MMT/gelatinized starch nano-composites for various clay loading, water loading and dispersionmethods.

01987654321

25

30

35

40

45

50

55

60

H%02 ,yalc%12

DC O H%02 ,yalc%12

BD OH %02 ,yalc%5.1

2BD O

H %02 ,yalc %2.32

BD OH %31 ,yalc%2.3

2U O

H %81 ,yalc %22

DC O

H %81 ,yalc %22

U O

yiel

d st

reng

th (

MP

a)

)%( noitagnole kaerb

H %31 ,yalc%2.32

BD O

H %02 ,yalc %2.32

U O

H %81 ,yalc %22

BD O

H %312O

H %812OH %02

2O

H %31 ,yalc %6.22

U O

H %02 ,yalc%5.12

U O

H%02 ,yalc%5.12

DC O

H%02 ,yalc%12

U O

Fig. 8. Yield strength versus break elongation for Na-MMT/gelatinizedstarch nanocomposites for various clay loading, water loadings anddispersion methods (U, clay dispersed using ultrasonics; DB, clay andstarch dry blended; CD, clay dispersed using conventional mixing).

K. Dean et al. / Composites Science and Technology 67 (2007) 413–421 419

larger CEC relates to the concentration of Na+ ions andpotentially the water retention properties (provided theinteraction between a water molecule and the interlayerion is one of the most dominant). This difference mayaccount for the variations in elongation to break trendsobserved where the higher CEC of the Na-MMT allows

for a stronger retention of water in the system, leadingto a more stable plasticised material. The Na-MMT nano-composite samples showed generally higher break elonga-tions compared to the Na-FHT nanocomposite samples(compare Figs. 8 and 12).

Fig. 12 illustrates the relationship between yieldstrength and elongation at break and gives some indica-tion of the toughness of the Na-FHT thermoplastic starchnanocomposite materials. Within error the 2.6 and3.2 wt% ultrasonically treated nanocomposite containing13 wt% water show the best ratio of strength toelongation.

0.00 5.

1 0.1 5.

2 0.2.5

3 0.3.5

0

10

20

30

40

50

60

13wt%water U

18wt%water U

20wt%water U

13wt%water DB

18wt%water DB

20wt%water DB

13wt%water CB

18wt%water CB

20wt%water CB

Yi

led

Str

entgh

(MP

)a

tw%

lc ay

Fig. 10. Yield strength (MPa) for Na-FHT gelatinized starch nanocom-posites for various clay loading, water loadings and dispersion methods.

0.00 5.

1 0.1 5.

2 0.2.5

3 0.3.5 0

1

2

3

4

5

6

7

8

9

13wt%water U

18wt%water U

20wt%water U

13wt%water DB

18wt%water DB

20wt%water DB

13wt%water CB

18wt%water CB

20wt%water CB

brea

kel

onag

tion

(%)

tw%

c al y

Fig. 11. Break elongation (%) for Na-FHT gelatinized starch nanocom-posites for various clay loading, water loadings and dispersion methods.

01987654321

25

30

35

40

45

50

55

60

65

H %31 ,yalc%1 2 BD,D,U O

H %02 ,yalc%5.1 2 BD O

H %02 ,yalc%5.1 2 U O

H %02 ,yalc%5.1 2 DC O

H %81 ,yalc%2 2 BD OH %81 ,yalc%2 2 DC O

H %81 ,yalc%2 2 U O

H %02 ,yalc%2.3 2 BD O

H %02 ,yalc%2.3 2 U O

H %81 ,yalc%2.3 2 U O

H %31 ,yalc%2.3 2 U O

H %81 2OH %02 2O

yiel

d st

reng

th (

MP

a)

)%( noitagnole kaerb

H %31 2 O

H %31 ,yalc%6.2 2 U O

Fig. 12. Yield strength versus break elongation for Na-FHT /gelatinizedstarch nanocomposites for various clay loading, water loadings anddispersion methods (U, clay dispersed using ultrasonics; DB, clay andstarch dry blended; CD, clay dispersed using conventional mixing).

0.00 5.

1 0.1 5.

2 0.2.5

3 0.3.5 0

1000

2000

3000

4000

5000

6000

13wt%water U

18wt%water U

20wt%water U

13wt%water DB

18wt%water DB

20wt%water DB

13wt%water CB

18wt%water CB

20wt%water CB

Mod

ulus

M(P

a)tw%

lc ay

Fig. 9. Tensile modulus (MPa) for Na-FHT gelatinized starch nanocom-posites for various clay loading, water loadings and dispersion methods.

420 K. Dean et al. / Composites Science and Technology 67 (2007) 413–421

5. Conclusion

We have developed a greater understanding of the rela-tionship between structure and properties in a series of gel-atinzed starch–clay nanocomposites. XRD and TEM haveshown that both intercalated and exfoliated structures canbe produced depending on levels of nanoclays and plasti-ciser, types of nanoclays, dispersion methodologies, and

processing conditions. Three different mixing regimes werestudied. These included: (A) dry blending of componentsprior to extrusion; (B) conventional mixing of clays in solu-tion prior to addition to the starch and extrusion; and (C)ultrasonic treatment of clays in solution prior to the addi-tion of starch and extrusion. It was shown that an optimumlevel of both plasticiser and nanoclay existed for each clay(which had some dependence on cationic exchange capac-ity) to produce a gelatinized starch film with the highestlevels of exfoliation and best improvement in mechanicalproperties. The use of ultrasonics was only advantageousin terms of clay dispersion at medium clay concentrationsin the Na-MMT samples and at higher clay concentrations

K. Dean et al. / Composites Science and Technology 67 (2007) 413–421 421

in the Na-FHT due to the difference in CEC. Significantimprovements in modulus, yield strength and break elon-gation were observed. When the level of clay, water andstarch was optimised an exfoliated structure was producedvia standard mixing which exhibited comparable improve-ments in mechanical properties to ultrasonically treatedsamples.

Acknowledgements

The authors wish to acknowledge Elizabeth Goodall forcompleting some of the XRD experimental work for thisproject and the Monash University Micro-Imaging Groupfor the use of their TEM facilities.

References

[1] Mohanty A, Misra M, Hinrichsen G. Macromol Mater Eng2000;276–277(1):1–24.

[2] Yu L, Christie G. Carbohydr Polym 2001;46(2):179–84.[3] Yu L, Christov V, Christie G, Beh H, Smyth R, Gray J et al.

Mechanical properties and microstructures of PLA/thermoplasticstarch blends. In: Proceedings of the IUPAC world polymer congress,Brisbane, Qld, Australia; 1998. p. 420.

[4] Yu L, Christie G. J Mater Sci 2005;40(1):111–6.[5] Van Soest J, Vliegenthart J. Trends Biotechnol 1997;15(6):208–13.[6] Van Soest J, Bezemer R, De Wit D, Vliegenthart J. Ind Crops Prod

1996;5(1):1–9.[7] Van Soest J, Hulleman S, De Wit D, Vliegenthart J. Carbohydr

Polym 1996;29(3):225–32.

[8] Van Soest J, Hulleman S, De Wit D, Vliegenthart J. Ind Crops Prod1996;5(1):11–22.

[9] Van Soest J, Benes K, De Wit D, Vliegenthart J. Polymer1996;37(16):3543–52.

[10] De Graaf R, Karman A, Janssen L. Starch/StaRke 2003;55(2):80–6.[11] Wilhelm H, Sierakowski M, Souza G, Wypych F. Carbohydr Polym

2003;52(2):101–10.[12] McGlashan S, Halley P. J Polym Int 2003;52(11):1767–73.[13] Avella M, De Vlieger J, Errico M, Fischer S, Vacca P, Volpe M. Food

Chem 2005;93(3):467–74.[14] Kalambur S, Rizvi S. Polym Int 2004;53(10):1413–6.[15] Kalambur S, Rizvi S. J Appl Polym Sci 2005;96(4):1072–82.[16] Park H-M, Lee W-K, Park C-Y, Cho W-J, Ha C-S. J Mater Sci

2003;38(5):909–15.[17] Park H, Li X, Jin C, Park C, Cho W, Ha C. Macromol Mater Eng

2002;287(8):553–8.[18] Huang M, Yu J, Ma X. Polymer 2004;45(20):7017–23.[19] Huang M, Yu J, Ma X, Jin P. Polymer 2005;46(9):3157–62.[20] Chen B, Evans J. Carbohydr Polym 2005;61(4):455–63.[21] Chiou B, Yee E, Glenn G, Orts W. Carbohydr Polym

2005;59(4):467–75.[22] Fischer H, Fischer, S. WO 01/68762 A1.[23] Fischer H. Mater Sci Eng: C 2003;23(6–8):763–72.[24] Pandey J, Singh R. Starch/StaRke 2005;57(1):8–15.[25] Wilhelm H, Sierakowski M, Souza G, Wypych F. Polym Int

2003;52(6):1035–44.[26] Kozak M, Domka L. J Phys Chem Solids 2004;65(2–3):441–5.[27] Lebaron P, Wang Z, Pinnavaia T. J Appl Clay Sci 1999;15

(1–2):11–29.[28] Moad G, Dean K, Edmond L, Kukaleva N, Li G, Mayadunne R,

et al. Macromol Symp 2006;233(1):170–9.[29] Sinha Ray S, Okamoto M. Prog Polym Sci 2003;28(11):1539–641.[30] Reichert P, Kressler J, Thomann R, Mullhaupt R, Stoppelmann G.

Acta Polym 1999;49:116–23.