Sugarcane bagasse cellulose/HDPE composites obtained by extrusion

Page 1

Composites Science and Technology 69 (2009) 214–219

Contents lists available at ScienceDirect

Composites Science and Technology

journal homepage: www.elsevier .com/ locate /compsci tech

Sugarcane bagasse cellulose/HDPE composites obtained by extrusion

Daniella Regina Mulinari a,*, Herman J.C. Voorwald a, Maria Odila H. Cioffi a, Maria Lúcia C.P. da Silva b,Tessie Gouvêa da Cruz a, Clodoaldo Saron b

a Faculdade de Engenharia de Guaratinguetá/UNESP, Av Dr. Ariberto Pereira da Cunha, 333, Pedregulho, CEP, 12516-410 Guaratinguetá/SP, Brazilb Escola de Engenharia de Lorena, Universidade de São Paulo, P.O. Box 116, CEP, 12600-000 Lorena/SP, Brazil

a r t i c l e i n f o

Article history:Received 11 April 2008Received in revised form 8 October 2008Accepted 8 October 2008Available online 17 October 2008

Keywords:A. FibersA. Polymer-matrix compositesB. Mechanical propertiesD. Scanning electron microscopy (SEM)E. Extrusion

0266-3538/$ - see front matter � 2008 Elsevier Ltd. Adoi:10.1016/j.compscitech.2008.10.006

* Corresponding author. Tel.: +55 12 31232865; faxE-mail addresses: [email protected],

(D.R. Mulinari).

a b s t r a c t

Natural fibers used in this study were both pre-treated and modified residues from sugarcane bagasse.Polymer of high density polyethylene (HDPE) was employed as matrix in to composites, which were pro-duced by mixing high density polyethylene with cellulose (10%) and Cell/ZrO2�nH2O (10%), using anextruder and hydraulic press. Tensile tests showed that the Cell/ZrO2�nH2O (10%)/HDPE composites pres-ent better tensile strength than cellulose (10%)/HDPE composites. Cellulose agglomerations were respon-sible for poor adhesion between fiber and matrix in cellulose (10%)/HDPE composites. HDPE/naturalfibers composites showed also lower tensile strength in comparison to the polymer. The increase inYoung’s modulus is associated to fibers reinforcement. SEM analysis showed that the cellulose fibersinsertion in the matrix caused an increase of defects, which were reduced when modified cellulose fiberswere used.

� 2008 Elsevier Ltd. All rights reserved.

1. Introduction

In recent years natural fibers composites have found an increas-ing number of applications [1–4]. These composites have shownspecial interest in interior components for automotives such asseat frames, side panel and central consoles [5,6]. Advantages ofnatural fibers are low cost, low density, high specific properties,biodegradable and non-adhesive characteristics. The main disad-vantages of natural fibers are low permissible processing tempera-tures, the tendency to form clumps, and the hydrophilic nature [7].

Some research works were developed using thermoplastic poly-mers/natural fibers showing excellent results for many applica-tions, especially in interior components for automotives.Thermoplastics such as polyethylene (PE) [8,9], polypropylene(PP) [10,11], poly(lactic acid) (PLA) [12] and poly(vinyl chloride)(PVC) [13] have been compounded with natural fibers (such as si-sal, jute, and sugarcane bagasse) to prepare composites.

Sugarcane bagasse is a residue widely produced and containscellulose (46.0%), hemicellulose (24.5%), lignin (19.95%), fat andwaxes (3.5%), ash (2.4%), silica (2.0%) and other elements (1.7%)[14]. Bagasse is a vegetable fiber mainly constituted by cellulose,that is a glucose-polymer with relatively high modulus, oftenfound as fibrillar component of many naturally occurring compos-

ll rights reserved.

: +55 12 [email protected]

ites (wood, sugarcane straw and bagasse) in association with lignin[15,16].

Methods as extrusion, compression and injection molding areused to place together fibers and thermoplastics matrix [17,18].The high density polyethylene (HDPE) is an engineering thermo-plastic used for several industrial applications due to low cost, de-sired mechanical properties and processing facility [8,9,19]. Thecombination of lignocellulosic material with thermoplastic matrixin general presents a considerable problem associated to incom-patibility between the polar and hygroscopic fiber and the non-po-lar and hydrophobic matrix.

Superficial treatments in natural fibers have been used to im-prove the matrix-reinforcement adhesion in composites [2]; how-ever, it is a critical step considering that the process could degradethe strong interfiber hydrogen bonding, which holds the fibers to-gether. Surface modification methods can be physical or chemicalaccording to the fiber surface modification mechanism. Frequently,the methods used for surface modification are bleaching, acetyla-tion and alkali treatment [20,21].

Studies using the presence of hydroxyl groups which are reac-tive and susceptible to chemical reactions have been developedto analyze weakness of the matrix–fiber interface. A chemicalmodification is carried out according to the insertion of non-polargroups on the fibers resulting in a hydrophobic surface compatiblewith thermoplastic matrices [22].

Several techniques have been used to determine possible mod-ifications in lignocellulosic materials such as X-ray diffractometry(XRD) and thermal analysis (TG). Gomes et al. [23] studied the

Page 2

D.R. Mulinari et al. / Composites Science and Technology 69 (2009) 214–219 215

development and effect of alkali treatment on tensile properties ofcuraua fiber green composites and evidenced that appropriate al-kali treatment is a key technology for improving mechanical prop-erties of cellulose-based fiber composites.

Sisal fiber/high density polyethylene (HDPE) composites inter-face was studied and an improved method, which evaluates moreaccurately the natural fiber and polymeric matrices interfacialproperties, was proposed. By this method two types of fiber surfacemodification were conducted: the chemical coupling by silane andthe oxidation process using permanganate. It was observed, in bothcases, an increase of fibers roughness introducing a mechanicalinterlocking with the matrix [9].

Coupling agents were used also by Keener et al. [24] to treatnatural fiber polyolefin composites increasing significantly theinterface strength, and consequently it was observed thatmechanical properties as tensile strength and impact energy in-creased twice and three times, respectively, when compared tonon-coupled blend of wood and polyethylene. Tserki et al. [25]studied the effect of acetylation and propionylation surfacetreatments on natural fibers and observed that the esterificationphenomena decreased the hydrophilic characteristics of thenatural fiber surface beyond the fiber crystalline as a result ofthe increase in the amorphous portion produced by thetreatment.

Shao et al. [26] studied the modification of carboxymethyl cel-lulose (CMC) and polyvinyl alcohol (PVA) with molybdenum oxideand it was observed a formation of highly regularly packed poly-mer/molybdenum oxide. Marques et al. [27] studied titanium diox-ide/cellulose nanocomposites prepared through the titanylsulphate hydrolysis in acidic medium in the presence of cellulosefibers and it was observed that this material can be used as rein-forced fibers in polymer matrix.

In this research work, sugarcane bagasse cellulose/HDPE com-posite and sugarcane bagasse cellulose modified with zirconiumoxychloride (ZrOCl2�8H2O)/HDPE composite were mixed by extru-sion and compression molded. Mechanical and thermal character-ization, X-ray diffraction and surface area measurement wereobtained and compared.

2. Methods

2.1. Isolation crude cellulose from sugarcane bagasse

The sugarcane bagasse was pre-treated with 10% sulfuric acidsolution (reactor of 350 L at 120 �C, 10 min) to isolate the cellulose,followed by centrifugation to separate the rich pentosanes solu-tion. Extracted lignocellulosic fraction was deslignificated with

Fig. 1. X-ray: (a) pure cellulose; (b) Cell/ZrO2�nH2O.

1% sodium hydroxide solution (reactor of 350 L at 100 �C, 1 h) toobtain the crude pulp, which was bleached with sodium chloride.The bleached cellulose (Fig. 1b) was dried in a store at 50 �C for12 h.

2.2. Chemical modification of sugarcane bagasse cellulose withzirconium oxychloride

About 2 g of zirconium oxychloride (ZrOCl2�8H2O) were dis-solved in 100 mL of aqueous hydrochloric acid solution(0.5 mol L�1) and 5 g of cellulose were immersed in this solution.The material was precipitated with ammonium solution (1:3) atpH 10.0, under stirring, filtered under vacuum and exhaustivelywashed with distilled water for the complete removal of chlorideions (negative silver nitrate test). The product was dried at 50 �Cfor 24 h. The resulting material was designated as Cell/ZrO2�nH2O[28].

2.3. Materials Characterization

Sugarcane bagasse cellulose, hydrous zirconium oxide and sug-arcane bagasse cellulose modified with metallic oxide were charac-terized by X-ray diffractometry (XRD), Surface area measurements(BET) and Thermal analysis (TG). X-ray diffractograms were ob-tained in a Rich Seifert diffractometer model ISO-DEBYFEX1001.Conditions used were: radiation CuKa, tension of 30 kV, currentof 40 mA and 0.05 (2h/5 s) scanning from values of 2h it enters10–70� (2h). Surface area measurements were performed in aQuantachrome instrument model NOVA 1000 in nitrogen atmo-sphere. Materials were pre-treated at 50 �C for 3 h. TG curves weregenerated in a Shimadzu instrument model TGA-50. The experi-ments were carried out under continuous nitrogen flow and witha heating rate of 10 �C min�1.

2.4. Preparation of composites

Sugarcane bagasse cellulose and sugarcane bagasse cellulosemodified with zirconium oxychloride (ZrOCl2�8H2O) were mixedwith the polymeric matrix (HDPE) in an extruder screw, marksIMACOM, in which fibers were responsible for 10 wt% in thecomposition. Respective temperatures for the four different pro-cessing zones from the hopper to horizontal die of the extruderwere set as 120/130/140/150 �C and the screw speed rate wasmaintained at 50 rpm. The extruded materials were subse-quently compression molded into samples for mechanicaltesting.

2.5. Compression molding

In an uniaxial press, containing heating elements (inferior andsuperior), a molding tool was placed, containing the materials forcompression (cellulose (10%)/HDPE composite, Cell/ZrO2�nH2O(10%)/HDPE composite and high density polyethylene (HDPE)).Both composites and HDPE plates were obtained at 150 �C during5 min at 5.000 kgf.

Cellulose (10%)/HDPE composite, Cell/ZrO2�nH2O (10%)/HDPEcomposite and high density polyethylene (HDPE) plates were cutin the necessary dimensions for the mechanical tests.

2.6. Tensile tests

Mechanical tests were carried out according to ASTM D-638specification. Five specimens from each composition were testedin an INSTRON universal-testing machine (model-8801),equipped with pneumatic claws at a cross-head speed of4.5 mm min�1.

Page 3

Table 2Results of the thermogravimetric (TG) curves of the materials, with the mass losses(m), in the respective range of temperature (DT) and its respective residues (R).

Material m (%) DT (�C) R (%)

Cellulose 4.7 40–200 2.883.9 200–500

8.6 500–800

Cell/ZrO2�nH2O 7.1 40–200 12.377.7 200–500

2.9 500–800

216 D.R. Mulinari et al. / Composites Science and Technology 69 (2009) 214–219

2.7. Scanning electron microscopy

The specimens submitted to tensile tests were cut and the com-posite intact fracture surface was analyzed in LEO 1450 V scanningelectron microscopy with tungsten filament operating at 20 kV,employing low vacuum technique and secondary electron detector.

3. results and discussion

The incorporation of oxide on sugarcane bagasse cellulose sur-face can be described by the following reaction:

Cellþ ZrOCl2 � 8H2OþHClþ NH4OH

! Cell=ZrO2 � nH2Oþ NH4Cl ð1Þ

Main differences between sugarcane bagasse cellulose and mod-ified cellulose with zirconium oxychloride (ZrOCl2�8H2O) can be ob-served in the X-ray diffractogram technique presented in Fig. 1.

It is possible to observe a major diffraction peak for 2h rangingbetween 22� and 23�, which corresponds to the cellulose (002)crystallographic planes.

The spectrum corresponding to the non-modified sugarcane ba-gasse cellulose fibers shows diffraction peaks at the following 2hangles: 15.7� and 22.8�. For modified sugarcane bagasse cellulosethe same peaks were observed at 15.9� and 22.8�. The presenceof the peaks at 15� and 22� is an evidence of the modification onfiber. These peaks indicate an increase of the interplanar distancein relation to the modified fiber. This behavior occurs due to thegeneration of disorder when fibers are modified. The projectionsubstituting groups along the axis is associated with an increasein the interfibrillar distance [29].

Patterns for both materials are similar; however, non-modifiedfiber is less crystalline than the modified one. Crystallinity index(CI), which is a measurement of the amount of crystalline cellulosewith respect to the global amount of amorphous materials, wasevaluated using Segal empirical method according to followingEq. (2):

CI ð%Þ ¼ I0 0 2 � Iam

Iam� 100 ð2Þ

where I0 0 2 is the maximum intensity of the 002 lattice reflection ofthe cellulose and Iam is the maximum intensity of X-ray scatteringbroad band due to the amorphous part of the sample. Accordingto this method, non-modified and modified fibers presented 47%and 53% of crystallinity, respectively.

The amount of oxide incorporated into cellulose was deter-mined by calcining 0.3 g of sample (Cell/ZrO2�nH2O) at 1073 K inair for 3 h and by weighting the residue. The quantity of oxidewas calculated and results are presented in Table 1.

These data were confirmed from surface area measurements.Comparing the specific superficial areas of the pure cellulose andCell/ZrO2�nH2O material, it was observed an increase. The oxidedeposition is an effective phenomena associated to the increaseof specific surface area, SBET which was from 0 m2 g�1 (see Table1) for pure cellulose to 36 m2 g�1 for the Cell/ZrO2�nH2O material[18].

Thermal analyses (TG) also confirm these results which areindicated in the Table 2, and presents values of mass loss and res-idue in the respective range temperature.

Table 1Quantity of ZrO2�nH2O incorporated on cellulose surface.

Samples ZrO2�nH2O incorporated (wt%) SBET (m2 g�1)

Cellulose – 0Cell/ZrO2�nH2O 3.66 36

Data (Table 2) indicate that the TG curve of the pure cellulosepresents higher mass loss in the range temperature 200–500 �C.The amount of residue in the Cell/ZrO2�nH2O material increasedfrom 2.8% to 12.3% due to the presence oxide on the cellulose sur-face. TG curves are presented in Fig. 2.

Curve A corresponds to sugarcane bagasse cellulose and curve Bto sugarcane bagasse cellulose modified with metallic oxide (Cell/ZrO2�nH2O).

Pure cellulose presents two decomposition phases: the first oneat 300 �C corresponds to degradation temperature and the secondone at 380 �C corresponds to complete temperature decomposition.

Cell/ZrO2�nH2O material shows two different phases comparedto pure cellulose: at 260 and 338 �C. This decrease of the temper-ature compared to pure cellulose is attributed to the presence ofoxide particles on the cellulose surface, indicating a strong interac-tion between hydrous zirconium oxide and cellulose fibers [28].

Cell/ZrO2�nH2O material was obtained under determined condi-tion, appropriated (acid medium at the beginning and basic med-ium at the end) to avoid the degradation of the cellulose fiberbecause it is known that the cellulose degradation occurs at pH10 [30]. Therefore, during the process care was taken in ordernot to reach this pH. Toledo et al. [31] studied the antimony (III)oxide film on a cellulose fiber surface and observed similar results.

Table 3 indicates the mechanical properties of these materialsinvolved in this research, in special the effect of chemical modifica-tion of sugarcane bagasse cellulose with zirconium oxychloride(ZrOCl2�8H2O) reinforced HDPE.

Composite reinforced with non-modified cellulose presentslower average values for tensile strength and elongation at breakcompared to high density polyethylene.

On the other hand, composite reinforced with modified cellu-lose shows significant decrease in elongation at break comparedto high density polyethylene and cellulose (10%)/HDPE composite.

Cell/ZrO2�nH2O/HDPE composite presented higher tensilestrength compared to the cellulose/HDPE composite, but still lowerthan high density polyethylene. An interesting increase in tensile

Fig. 2. TG curves: (A) pure cellulose; (B) Cell/ZrO2�nH2O.

Page 4

Table 3Mechanical properties of the materials obtained by compression molding.

Materials Elongation at break(tensile) (%)

Tensilestrength (MPa)

Tensilemodulus (MPa)

High densitypolyethylene(HDPE)

1.96 ± 0.087 16.7 ± 0.15 850.9 ± 28.2

Cellulose/HDPEcomposite

1.62 ± 0.097 14.4 ± 0.58 880.1 ± 63.5

Cell/ZrO2�nH2O/HDPEcomposite

1.2 ± 0.185 15.6 ± 1.11 1324.2 ± 211.0

Reinforcement in wt%.

D.R. Mulinari et al. / Composites Science and Technology 69 (2009) 214–219 217

modulus occurred as a consequence of the oxide particles on cellu-lose fibers.

Experimental data in Table 3 also showed a poor interaction be-tween fibers and matrix during the mixture process. Compositeswere obtained with homogeneous distribution, but with agglomer-ations in some points caused by inefficient fibers dispersion insidematrix. This agglomeration of reinforcement was responsible bydecrease of the tensile strength compared to the high densitypolyethylene.

Composites reinforced with modified cellulose presented bettertensile strength and adhesion between fiber and matrix than non-modified cellulose, due to the agglomerations decrease. Luz et al.[32] observed similar results by mechanical testing and micro-structural analysis of sugarcane bagasse fibers reinforced polypro-pylene composites.

Fig. 3 shows the presence of pull out in the cellulose/HDPE com-posite fracture surface by the SEM technique.

The fracture of the specimen was caused by the presence ofagglomerates, which is considered as defect in the material. Image

Fig. 3. Scanning electron microscopy of fractured surfaces of composites: (a) cellulose/HDregion between fiber and matrix; (c) Cell/ZrO2�nH2O/HDPE composite; (d) Cell/ZrO2�nH2

matrix.

analyses method was used in this research to quantify the rein-forcement particles (or fibers).

Analyzing Fig. 3a and b it was observed pull out caused by lowadhesion, which acts as stress concentrator. However, it was notobserved the same considering Fig. 3c and d.

The amount of agglomerates in the composites reinforced withmodified cellulose decreased, which resulted, as a consequence,the decrease in ductility. The reduction in the elongation at breakfor this condition compared to the high density polyethylene isassociated to defects generated in the material after fibers inser-tion. However, the addition of modified cellulose in the matrix re-duced these defects, confirming the interfacial improvementbetween fibers and matrix.

The mechanical behavior of composites is strongly dependenton the microstructural distribution of the reinforced particles inmatrix. Particle agglomerates act as defects or stress concentrationsites for crack nucleation. The ability to quantify the size and dis-tribution of particle agglomerates and their effect on crack growthbehavior is extremely important [33].

Tang et al. [34] studied particle distribution characterizationby multi-scalar analysis of area fractions (MSAAF) technique toanalyze the particle spatial distribution of composite samples ob-tained by vacuum hot pressing (VHP). The basis principle ofMSAAF technique is to characterize the spatial heterogeneity incomposite microstructures by obtaining statistical informationabout the variability of reinforced particle area fractions over var-ious length scales by quantitative image analyses methods. Agray-scale image often contains only two levels of significantinformation, namely the foreground level constituting objects ofinterest and the background level against which the foregroundis discriminated [35].

PE composite; (b) cellulose/HDPE composite with higher magnification of the limitO/HDPE composite with higher magnification of the limit region between fiber and

Page 5

Table 4Reinforcement particle area fractions obtained for statistical analyses.

Materials Average

Cellulose/HDPE composite 46.6 ± 7.8Cell/ZrO2�nH2O/HDPE composite 33.2 ± 5.5

218 D.R. Mulinari et al. / Composites Science and Technology 69 (2009) 214–219

In this article 26 images in random regions of each sample (Cel-lulose/HDPE composite and Cell/ZrO2�nH2O/HDPE composite) werecaptured to establish the reinforcement particle area fractions ob-tained by SEM (Fig. 3) utilizing image analyses method.

These area fractions were classified as clear regions (lesser val-ues of gray levels) and less clear regions (higher values of gray lev-els). The clearest regions are associated to higher reinforcementparticle and the least clear regions mentions to lesser reinforce-ment particle.

Table 4 shows results obtained for statistical analyses ofcomposites.

The area fraction obtained for Cell/ZrO2�nH2O/HDPE compositedecreases 13.4%, indicating an increase of filled regions by rein-forcement particle. Data of Table 4 indicate a tendency of nonhomogeneous distribution in the reinforcement particle. Therefore,these results showed the fulfilling regions, confirming that modifi-cation of the sugarcane bagasse cellulose with zirconium oxychlo-ride improved the adhesion between fibers and matrix.

This way, the modification of sugarcane bagasse cellulose andcompression molding were adequate for thermoplastic compositespreparation.

4. Conclusions

The modification of sugarcane bagasse cellulose with zirconiumoxychloride was successfully accomplished and it was verified thateffectively improves the tensile strength compared to non-modi-fied sugarcane bagasse cellulose.

The preparation of composites using non-modified sugarcanebagasse cellulose presented agglomerations. The agglomerationsand/or the non-homogeneity distribution of fibers are defects inthe material that directly interfere in the mechanical properties,harming the production of a high resistance material.

The modification of sugarcane bagasse cellulose reduced thecomposites elongation to 26% compared to non-modified sugar-cane bagasse cellulose; on the other hand the tensile modulus in-creased 50%. Therefore the molding process using extrusion andhydraulic press was appropriate.

Acknowledgements

The authors express their acknowledgements to CAPES for thefinancial support.

Appendix A

BET

surface area measurement Cell/

ZrO2�nH2O

sugarcane bagasse cellulosemodified with hydrous zirconium oxide

Cellulose

sugarcane bagasse cellulose non-modified HDPE high density polyethylene MSAAF multi-scalar analysis of area fractions SEM scanning electron microscopy VHP vacuum hot pressing XRD X-ray diffractometry TG thermal analysis

References

[1] Facca AG, Kortschot MT, Yan N. Predicting the tensile strength of natural fibrereinforced thermoplastics. Compos Sci Technol 2007;67:2454–66.

[2] Torres FG, Cubillas ML. Study of the interfacial properties of natural fiberreinforced polyethylene. Polym Test 2005;24:694–8.

[3] Abdelmouleh M, Boufi S, Belgacem MN, Dufresne A. Short natural-fibrereinforced polyethylene and natural rubber composites: effect of silanecoupling agents and fibres loading. Compos Sci Technol 2007;67:1627–39.

[4] Georgopoulos ST, Tarantili PA, Avgerinos E, Andreipoulos AG, Loukios EG.Thermoplastic polymers reinforced with fibrous agricultural residues. PolymDegrad Stab 2005;90:303–12.

[5] Farag MM. Quantitative methods of materials substitution: application toautomotive components. Mater Design 2008;29:374–80.

[6] Zah R, Hischier R, Leão AL, Braun I. Curauá fibers in the automobile industry – asustainability assessment. J Cleaner Prod 2007;15:1032–40.

[7] Liu CF, Sun RC, Zhang AP, Ren JL, Wang XA, Qin MH, et al. Homogeneousmodification of sugarcane bagasse cellulose with succinic anhydride using aionic liquid as reaction medium. Carbohyd Polym 2007;342:919–26.

[8] Lei Y, Wu Q, Yao F, Xu Y. Preparation and properties of recycled HDPE/naturalfiber composites. Composites A 2007;38:1664–74.

[9] Li Y, Hu C, Yu Y. Interfacial studies of sisal fiber reinforced high densitypolyethylene (HDPE) composites. Composites A 2008;39:570–8.

[10] Sain M, Park SH, Suhara F, Law S. Flame retardant and mechanical properties ofnatural fibre–PP composites containing magnesium hydroxide. Polym DegradStab 2004;83:363–7.

[11] Zampaloni M, Pourboghrat F, Yankovich SA, Rodgers BN, Moore J, Drzal LT,et al. Kenaf natural fiber reinforced polypropylene composites: a discussion onmanufacturing problems and solutions. Composites A 2007;38:1569–80.

[12] Ochi S. Mechanical properties of kenaf fibers and kenaf/PLA composites. MechMater 2008;40:446–52.

[13] Zheng Y, Cao D, Wang D, Chen J. Study on the interface modification of bagassefibre and the mechanical properties of its composite with PVC. Composites A2007;38:20–5.

[14] Sene L, Converti A, Felipi MGA, Zilli M. Sugarcane bagasse as alternativepacking material for biofiltration of benzene polluted gaseous streams: apreliminary study. Bioresource Technol 2002;83:153–7.

[15] Alonso PW, Garzone P, Cornacchia G. Agro-industry sugarcane residuesdisposal: the trends of their conversion into energy carriers in Cuba. WasteManage 2007;27:869–85.

[16] Pandey A, Soccol CR, Nigam P, Soccol VT. Biotechnological potential of agro-industrial residues: I. Sugarcane bagasse. Bioresearch Technol 2000;74:69–80.

[17] Chiellini E, Cinelli P, Corti A, Kenawy ER. Composite films based on wastegelatin: thermal–mechanical properties and biodegradation testing. PolymDegrad Stab 2001;73:549–55.

[18] Curvelo AAS, De Carvalho AJF, Agnelli JAM. Thermoplastics starch–cellulosicfibers composites: preliminary results. Carbohyd Polym 2001;45:183–8.

[19] Albano C, Karam A, Domínguez N, Sánches Y, González J, Aguirre O, et al.Thermal, mechanical, morphological, thermogravimetric, rheological andtoxicological behavior of HDPE/seaweed residues composites. Compos Struct2005;71:282–8.

[20] Liu CF, Suna RC, Qin MH, Zhang AP, Rena JL, Xu F, et al. Chemical modificationof ultrasound-pretreated sugarcane bagasse with maleic anhydride. Ind CropsProd 2007;26:212–9.

[21] Karnitz Jr O, Gurgel LVA, Melo JCP, Botaro V, Melo TMS, Gil RPF, et al.Adsorption of heavy metal ion from aqueous single metal solution bychemically modified sugarcane bagasse. Bioresource Technol 2007;98:1291–7.

[22] Ma X, Yu J, Kennedy JF. Studies on the properties of natural fibers-reinforcedthermoplastic starch composites. Carbohyd Polym 2005;62:19–24.

[23] Gomes A, Matsuo T, Goda K, Ohgi J. Development and effect of alkali treatmenton tensile properties of curaua fiber green composites. Composites A2007;38:1811–20.

[24] Keener TJ, Stuart RK, Brown TK. Maleated coupling agents for natural fibrecomposites. Composites A 2004;35:357–62.

[25] Tserki V, Zafeiropoulos NE, Simon F, Panayiotou C. A study of the effect ofacetylation and propionylation surface treatments on natural fibres.Composites A 2005;36:1110–8.

[26] Shao K, Liao S, Luo H, Wang M. Self-assembly of polymer and molybdenumoxide into lamellar hybrid materials. J Colloid Interf Sci 2008;320:445–51.

[27] Marques PAA, Trindade T, Neto CP. Titanium dioxide/cellulose nanocompositesprepared by a controlled hydrolysis method. Compos Sci Technol2006;66:1038–44.

[28] Mulinari DR, Da Silva MLCP, Da Silva GLJP. Adsorção de íons dicromato noscompósitos celulose/ZrO2�nH2O preparados pelos métodos da precipitaçãoconvencional e em solução homogênea. Quim Nova 2006;29:496–500.

[29] Rodrigues Filho G, Assunção RMN, Vieira JG, Meireles CS, Cerqueira DA, BarudHS, et al. Characterization of methylcellulose produced from sugar canebagasse cellulose: crystallinity and thermal properties. Polym Degrad Stab2007;92:205–10.

[30] Borgo CA, Gushikem Y. Zirconium phosphate dispersed on a cellulose fibersurface: preparation, characterization, and selective adsorption of Li+, Na+ andK+ from aqueous solution. J Colloid Interf Sci 2002;246:343–7.

[31] Toledo EA, Gushikem Y, Castro SC. Antimony (III) oxide film on a cellulose fibersurface: preparation and characterization of the composite. J Colloid Interf Sci2000;225:455–9.

Page 6

D.R. Mulinari et al. / Composites Science and Technology 69 (2009) 214–219 219

[32] Luz SM, Gonçalves AR, Del’Arco Jr AP. Mechanical behavior andmicrostructural analysis of sugarcane bagasse fibers reinforcedpolypropylene composites. Composites A 2007;38:1455–61.

[33] Ganguly P, Poole WJ. Characterization of reinforcement distributioninhomogeneity in metal matrix composites. Mater Sci Eng A2002;332:301–10.

[34] Tang F, Meeks H, Spowart JE, Gnaeupel-Herold T, Prask H, Anderson IE.Consolidation effects on tensile properties of an elemental Al matrixcomposite. Mater Sci Eng A 2004;386:194–204.

[35] Seul M, O’Gorman L, Sammon MJ. Practical algorithms for image analysis –description, example and code. Cambridge: Cambridge University Press;2000.