performance of rubberwood fiber–thermoplastic natural rubber composites

14

Click here to load reader

Upload: s

Post on 30-Mar-2017

217 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Performance of Rubberwood Fiber–Thermoplastic Natural Rubber Composites

Performance of Rubberwood Fiber–ThermoplasticNatural Rubber Composites

J. K. Sameni,* S. H. Ahmad, and S. Zakaria

School of Applied Physics, Faculty of Science and Technology,

Universiti Kebangsaan Malaysia, Selangor Darul Ehsan, Malaysia

ABSTRACT

Rubberwood fiber–thermoplastic natural rubber (RWF–TPNR) com-

posites were prepared and evaluated in this study. The RWFs were

supplied in the form of thermomechanical pulp (TMP) from the Hume

Fiberbord Company in Malaysia. The fibers were screened to 0.5–1 mm

sizes and blended with TPNR and maleic anhydride–grafted poly-

propylene (MAPP) as a compatibilizer in the internal mixer of Brabender

Plasticorder machine. The mechanical, thermal properties and mor-

phology of the composites were investigated by tensile, impact, flexural,

thermogravimetric analysis (TGA), differential scanning calorimetry

(DSC), and scanning electron microscope (SEM). The results showed that

the tensile strength and Young’s modulus increased, while tensile strain

and impact strength decreased with increasing of fiber loading.

Thermogravimetric analysis showed an increasing degradation tempera-

ture of the samples with RWF. On the other hand, SEM micrographs

139

DOI: 10.1081/PPT-120016340 0360-2559 (Print); 1525-6111 (Online)

Copyright q 2003 by Marcel Dekker, Inc. www.dekker.com

*Correspondence: J. K. Sameni, School of Applied Physics, Faculty of Science and

Technology, Universiti Kebangsaan Malaysia, 43600 Bangi, Selangor Darul Ehsan,

Malaysia; E-mail: [email protected].

POLYMER–PLASTICS TECHNOLOGY AND ENGINEERING

Vol. 42, No. 1, pp. 139–152, 2003

MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016

©2003 Marcel Dekker, Inc. All rights reserved. This material may not be used or reproduced in any form without the express written permission of Marcel Dekker, Inc.

Page 2: Performance of Rubberwood Fiber–Thermoplastic Natural Rubber Composites

showed good adherence between fibers and matrix with the presence of

MAPP.

Key Words: Composite; Rubberwood fiber; TPNR; MAPP; Mechanical

properties.

INTRODUCTION

It is well-known that the performance (i.e., the mechanical properties) of

composites depends on the properties of the individual components and their

interfacial compatibility. Natural fibers, one component of the investigated

composites, are strongly polar because of hydroxyl groups and CZOZC links

in their structure. The potential of natural fiber-based composites using

cellulose, wood, jute, kenaf, hemp, sisal, pineapple, or coir as reinforcing

fibers in a thermoplastics matrix has received considerable attention among

scientists all over the world[1 – 3] for their excellent specific properties. Some of

them have already been used as industrial products.[4]

Natural fibers are compatible with polar, acidic, or basic, rather than with

nonpolar polymers. Because of this inherently poor compatibility between the

hydrophilic cellulose fibers and typical hydrophobic commodity thermo-

plastics, such as polyolefines, a pretreatment of the fiber surface,[5] of the

matrix polymer,[6] or the incorporation of surface modifiers[7 – 10] is generally

required. Gauthier et al.[11] pointed out that the compatibilizers used in the

cellulosic fiber/polyolefin composites must possess a function highly reactive

with the OH groups of the cellulose.

Proper selection of additives is necessary to improve the interaction

and adhesion between the fiber and matrix phases. Maleic anhydride–

grafted polypropylene (MAPP) has been reported to function efficiently for

lignocellulosic/PP system.[12] Earlier results suggested that the amount of

MA grafted and the molecular weight were both important parameters to

determine the efficiency of the additive.[13]

It should be noted that the incorporation of MAPP into composites also

exhibits beneficial effects in enhancing properties such as interfacial bonding.

In general there are two approaches commonly used to enhance the chemical

or physical interactions between the reinforcements and polymer matrix. The

first method involves the modification of the matrix structure via chemical

reactions; the second route uses coupling agents to modify the chemical nature

of the fiber surface.[14]

Sameni, Ahmad, and Zakaria140

MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016

©2003 Marcel Dekker, Inc. All rights reserved. This material may not be used or reproduced in any form without the express written permission of Marcel Dekker, Inc.

Page 3: Performance of Rubberwood Fiber–Thermoplastic Natural Rubber Composites

In the present study, the mechanical and thermal properties of composites

obtained from different rubberwood fiber (RWF) fraction and the effect of

coupling agent have been reported.

METHODOLOGY

Materials

Rubberwood fibers were obtained from Hume Fiberbord Sdn. Bhd. in

Malaysia and separated into lengths of about 0.5–1 mm. The fibers were dried

to remove any moisture. The PP was supplied by Dow Chemicals (with

density 0.913 g/cm3) and natural rubber (NR) by Guthrie (M) Sdn. Bhd.,

Malaysia. Liquid natural rubber (LNR) was produced in a chemistry

laboratory, Universiti Kebangsaan, Malaysia. MAPP (from the Aldrich

Company) was used as a coupling agent to improve the compatibility and

adhesion between the fibers and matrix.

Sample Preparation and Testing

Blending of PP:NR:LNR with a composition ratio of 60:30:10 was

prepared in a laboratory internal mixer (Brabender) at 1758C. Thermoplastic

natural rubber (TPNR) and rubberwood fiber with different fractions (0, 10,

20, and 30%) were mixed in Brabender. The TPNR was allowed to melt in the

mixer for 3 min before the rubberwood fiber was charged. In the case of the

addition of compatibilizer, the MAPP was added into the mixer 2 min after

introducing TPNR. MAPP and TPNR were mixed for 1 min before RWF was

added into the mixer.

Mechanical Measurements

Tensile measurements were made on dumbbell-shaped specimens with

thickness of 1 mm. The tensile properties were measured at room temperature

on an Instron Universal Tester (model Micro 350) according to ASTM D412.

The gauge was kept at 50 mm with a crosshead speed of 50 mm min21. From

the stress strain curves, tensile strength, Young’s modulus, elongation at

break, and maximum strain were determined by instrument software. Five tests

were run for each blend sample. Flexural tests were carried out according to

ASTM D790 on the same machine. The distance between the spans was

Performance of RWF–TPNR Composites 141

MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016

©2003 Marcel Dekker, Inc. All rights reserved. This material may not be used or reproduced in any form without the express written permission of Marcel Dekker, Inc.

Page 4: Performance of Rubberwood Fiber–Thermoplastic Natural Rubber Composites

100 mm, and the strain rate (compression speed) was 5 mm min21.

Five samples were tested for each composition, and average values were

reported. Notched Izod impact resistance test specimens were prepared from

3-mm thick samples. An Izod impact test was performed using a 2 hammer on

a Universal Fractoscope (model CEAST 6546). Five specimens were tested

for each sample. The test method conducted was based on ASTM D 256-93.

Thermal Analysis

The thermal behavior of composite was determined using a Mettler

Toledo thermal analyzer model TGA/SDTA581e. The temperature was

scanned from 308C to 5508C with a heating rate of 108C min21. The change in

weight with temperature was processed to display a weight vs. temperature

plot.

DSC (differential scanning calorimetric) measurements were made on a

Mettler Toledo thermal analyzer model DSC882e. Small amounts (about

6 mg) of sample were placed in aluminum cups that were subsequently

crimped hermetically. The sample and reference, which consisted of an empty

aluminum cup similarly crimped, were heated to 2008C. The DSC melting

curves were obtained as heat flow vs. temperature plots by heating from

21008C to 2008C at a rate of 108C min21.

Morphological Examination

The fractured surfaces of the composites from the tensile test were

examined by scanning electron micrographs (Philips SEM XL 30). The

fracture ends of the specimens were mounted on an aluminum stub and coated

with thin layer gold to avoid electrostatic charging during examination.

RESULTS AND DISCUSSION

Mechanical Properties

The effect of fiber loading upon the tensile properties of composites was

determined for a range of fiber contents from 0% to 30% (by weight). The

surfaces of cellulose fibers are rendered more hydrophobic by MAPP, and it

would be expected that enhanced compatibility between fiber and matrix. This

will improved the wetting of the fiber by the TPNR, then lead to a stronger

Sameni, Ahmad, and Zakaria142

MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016

©2003 Marcel Dekker, Inc. All rights reserved. This material may not be used or reproduced in any form without the express written permission of Marcel Dekker, Inc.

Page 5: Performance of Rubberwood Fiber–Thermoplastic Natural Rubber Composites

interfacial bond. It will therefore be anticipated that more efficient stress

transfer would occur between the fibers as a load is applied. The possible

chemical reaction is depicted in Fig. 1.

The effects of fiber loading and coupling agent on the mechanical

properties are shown in Fig. 2. A higher value for the tensile strength of the

composite would thus be expected for reinforcement with MAPP compared to

unmodified fibers. The results are presented in Fig. 2(a) showed that higher

tensile strength was obtained when using MAPP. Furthermore, it was apparent

that the tensile strength increased slightly because of the fiber content of the

composite.

The effects of fiber loading and coupling agent on the Young’s

modulus properties are illustrated in Fig. 2(b). The Young’s modulus of

composites with MAPP was higher than the composites without MAPP. It

was obvious that the stiffness of composites increased with increasing

fiber loading.

The relationship between the elongation at break and fiber content of

composites is shown in Fig. 2(c). A decreased in elongation at break was

exhibited with composites reinforced with fiber and coupling agent. The effect

of MAPP on the maximum strain decreased with increasing compatibilizer

content and fiber loading (Fig. 2(d)).

The effects of the fiber loading and coupling agent on the flexural

stiffness (MOE) are shown in Figs. 3(a) and 3(b), respectively. Studies of

composites formed of jute/PP revealed that there was a linear relationship

between the flexural modulus and fiber loading.[4] The results of MOE of

RWF/TPNR composites show a linear increase with rubberwood fiber

loading. Addition of MAPP affected the improvement of stiffness of

RWF/TPNR composites. Therefore, RWF/TPNR composites were shown

to exhibit modulus of elasticity lower than RWF/TPNR/MAPP composites.

Figure 1. The reaction of cellulose with MAPP.

Performance of RWF–TPNR Composites 143

MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016

©2003 Marcel Dekker, Inc. All rights reserved. This material may not be used or reproduced in any form without the express written permission of Marcel Dekker, Inc.

Page 6: Performance of Rubberwood Fiber–Thermoplastic Natural Rubber Composites

The relationship between fiber loading and impact strength demonstrated

an approximately linear decrease (Fig. 4(a)). Tjong et al.[15] also reported that

the impact resistance of whisker-reinforced polypropylene composite showed

a decrease with fiber loading. At a fiber loading of 10%, the composites

without MAPP had higher impact strengths than those obtained from the

composites with MAPP, but no significant difference was observed at higher

fiber loading. The impact strength of a composite is influenced by several

factors, including the toughness of the reinforcement material, the nature of

the interfacial region, and the frictional force involved in pulling out the fibers

from the matrix.[16] The effect of a coupling agent on the impact strength is

illustrated in Fig. 4(b). Impact energy decreased with an increasing MAPP

content.

Thermal Behavior

Thermogravimetry (TG) result of RWF, TPNR and RWF/TPNR

composites are shown in Fig. 5. Degradation temperature and maximum

rate of weight loss for each component derived from respective

thermogravimetric curves as presented in Table 1.

Figure 2. Comparison of tensile properties of RWF/TPNR composites (B) with and

(A) without MAPP.

Sameni, Ahmad, and Zakaria144

MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016

©2003 Marcel Dekker, Inc. All rights reserved. This material may not be used or reproduced in any form without the express written permission of Marcel Dekker, Inc.

Page 7: Performance of Rubberwood Fiber–Thermoplastic Natural Rubber Composites

The degradation of natural fibers is a crucial aspect in the

development of natural fiber composites and thus has a bearing on the

curing temperature in thermoplastics composites.[17] It has been shown

previously that, pyrolysis of hemicellulose, lignin, and cellulose occurred

at 200, 220 and 2508C, respectively.[18] In this study, volatilization of

RWF began at 2508C, and a maximum weight loss of 28%/min was

recorded at 3458C.

The initial decomposition temperatures (Td) of RWF/TPNR composites,

depending on the percentage of fiber loading, were lower than matrix TPNR

(see Table 1). The maximum rates of weight loss for TPNR were

correspondingly much lower than those for RWF/TPNR composites (around

10%/min as compared with 25%/min).

The DSC thermogram of RWF showed an endothermic peak at about

698C that might have been related to some extractive latex (Fig. 6).

Two exotherm peaks around 3388C and 4708C were attributed to

decomposition and oxidative combustion, respectively.

The effect of fiber on the melting point (Tm) was not significant (Fig. 7).

With introduction of crosslinks, the melting temperature increased. In this

Figure 3. Effect of fiber loading and coupling agent on the modulus of elasticity.

Performance of RWF–TPNR Composites 145

MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016

©2003 Marcel Dekker, Inc. All rights reserved. This material may not be used or reproduced in any form without the express written permission of Marcel Dekker, Inc.

Page 8: Performance of Rubberwood Fiber–Thermoplastic Natural Rubber Composites

experiment no significant changes were observed in the melting point;

therefore, it can be explained that bonding does not occur between

rubberwood and matrix.

The crystallinity (Xcom) of composites were determined using the

following relationship:

Xcomð% crystallinityÞ ¼ DHf=DH8f £ 100 ð1Þ

where a value of DH8f ¼ 190 J g21 was taken for 100% crystalline PP

homopolymer. Xcom, which is calculated using this equation, however, gives

only the overall crystallinity of the composites based on the total weight of

composites including noncrystalline fractions, and it is not the true

crystallinity of the PP phase. Therefore, to investigate the crystallinity for

PP phase (Xpp) the PP fraction was normalized to Eq. (1) formula as follows:

Xpp ¼ ðXcomÞ=WfPP ð2Þ

Figure 4. Effect of fiber loading and coupling agent on the impact strength.

Sameni, Ahmad, and Zakaria146

MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016

©2003 Marcel Dekker, Inc. All rights reserved. This material may not be used or reproduced in any form without the express written permission of Marcel Dekker, Inc.

Page 9: Performance of Rubberwood Fiber–Thermoplastic Natural Rubber Composites

Figure 5. TGA thermograms of RWF, TPNR, and RWF/TPNR composites.

Table 1. TG and DSC data.

Sample

Decomposition

temp.

(Td) (8C)

Max.

rate of

wt loss

%/min.

(Temp.)

Glass

transition

Tg (8C)

Melting

point

Tm (8C)

DHf(com)

(J g21)

Xcom

(%)

XPP

(%)

TPNR

(PP/NR)

340 10

(3808C)

264.7 168.8 53.3 28.0 46.6

TPNR þ

RWF10%

320 15

(3708C)

265.2 167.5 48.0 25.3 46.8

TPNR þ

RWF20%

300 21

(3608C)

264.7 167.5 44.3 23.3 48.5

TPNR þ

RWF30%

275 25

(3558C)

264.0 167.0 39.7 21.0 50.0

RWF 250 28

(3458C)

— — — — —

Natural

rubber

290 40

(3808C)

263.5 — — — —

Polypropylene 400 62

(4608C)

— 168.0 75.8 — 40.0

Performance of RWF–TPNR Composites 147

MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016

©2003 Marcel Dekker, Inc. All rights reserved. This material may not be used or reproduced in any form without the express written permission of Marcel Dekker, Inc.

Page 10: Performance of Rubberwood Fiber–Thermoplastic Natural Rubber Composites

where WfPP is the weight fraction of PP in the composites. Increasing

crystallization rate of PP can be explained by the assumption that the RWF

acts as nucleating agent.[20] These results indicate clearly that the

incorporation of RWF in the investigated TPNR types slightly affect

Figure 7. DSC thermograms of TPNR and TPNR/RWF.

Figure 6. DSC thermogram of rubberwood fiber.

Sameni, Ahmad, and Zakaria148

MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016

©2003 Marcel Dekker, Inc. All rights reserved. This material may not be used or reproduced in any form without the express written permission of Marcel Dekker, Inc.

Page 11: Performance of Rubberwood Fiber–Thermoplastic Natural Rubber Composites

the crystallization of PP. Therefore, the percentages of crystallinity of the PP

component changed with fiber content.

Table 1 also shows that the value of DHf(com) and Xcom decreased with

increasing RWF loading. This is due to decreasing of PP concentration at high

fiber loading.

Morphology

Scanning electron microscope micrographs with different magnifications

of the fractured surfaces are shown in Fig. 8. It is well-known that the

composites with satisfactory mechanical behavior depend on good dispersion

Figure 8. SEM micrograph of the tensile fractured surface of TPNR composites

without MAPP. (a) 250 £ (b) 500 £ .

Figure 9. SEM micrograph of the tensile fractured surface of TPNR composites with

MAPP. (a) 250 £ (b) 500 £ .

Performance of RWF–TPNR Composites 149

MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016

©2003 Marcel Dekker, Inc. All rights reserved. This material may not be used or reproduced in any form without the express written permission of Marcel Dekker, Inc.

Page 12: Performance of Rubberwood Fiber–Thermoplastic Natural Rubber Composites

and wetting of the filler in the matrix.[21] In the case of the RWF/TPNR

composite, RWF didn’t have the ability to form good filler–matrix interactions

with the nonpolar TPNR. Figures 8(a) and 8(b) show some indentation and

pullout in the fractured surface of composites with different magnifications.

These explained the lack of adhesion between the fibers and polymer matrix,

which resulted in poor tensile strength. Figures 9(a) and 9(b) show less fiber

pullout and bonding in the composites after addition of MAPP.

CONCLUSION

The Young’s modulus and tensile strength increased with increasing fiber

content. On the other hand, maximum strain and elongation at break decreased

with increasing fiber content. This research showed that MAPP as a coupling

agent was able to significantly improve the tensile strength and MOE of the

RWF-filled composites. The treatment with MAPP resulted in a decrease in

tensile strain and impact strength of the composites. Scanning electron

microscope micrographs of fractured surface of the composites showed that

there was good adhesion between the fiber and the matrix. Thermogravimetric

analysis revealed that the present of rubberwood fiber improved the

degradation temperature of the composite. In addition, DSC thermograms

showed an increasing percentage of crystallinity resulting from a

reinforcement effect and interaction in the interphase area of fiber and matrix.

ACKNOWLEDGMENT

The financial support of the Ministry of Science and Technology, under

the project IRPA 09-02-02-0074, is gratefully acknowledged.

REFERENCES

1. Mitra, B.M.; Basak, R.K.; Sharkar, M. Studies on jute-reinforced

composites, its limitations, and some solutions through chemical

modifications of fibers. J. Appl. Polym. Sci. 1998, 67, 1093.

2. Pavithran, C.; Mukherjee, P.S.; Brahamkumar, M. Impact properties of

natural fiber composites. J. Reinf. Plast. Compos. 1991, 10, 91.

3. Karmaker, A.C; Youngquist, J.A. Injection molding of polypropylene

reinforced with short jute fibers. J. Appl. Polym. Sci. 1996, 62, 1147.

Sameni, Ahmad, and Zakaria150

MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016

©2003 Marcel Dekker, Inc. All rights reserved. This material may not be used or reproduced in any form without the express written permission of Marcel Dekker, Inc.

Page 13: Performance of Rubberwood Fiber–Thermoplastic Natural Rubber Composites

4. Rana, A.K.; Mitra, B.C.; Banerjee, A.N. Short fiber-reinforced

polypropylene composites: dynamic mechanical study. J. Appl. Polym.

Sci. 1999, 71, 531.

5. Kokta, B.V.; Chen, R.; Danneault, C.; Valade, J.L. Use of wood fibers in

thermoplastic composites. Polym. Compos. 1983, 4, 229.

6. Dong, S.; Sapieha, S.; Schreiber, H.P. Mechanical properties of corona-

modified cellulose/polyethylene composites. Polym. Eng. Sci. 1993, 33

(6), 1734.

7. Dalvag, H.; Klason, C.; Stromvall, H.E. The efficiency of cellulosic filler

in common thermoplastics. Part II. Filling with processing aids and

coupling agents. Int. J. Polym. Mater. 1985, 11, 9.

8. Woodhams, R.T.; Thomas, G.; Rodgers, D.K. Wood fibers as reinforcing

fillers for polyolefins. Polym. Eng. Sci. 1984, 24, 1166.

9. Sapieha, S.; Allard, P.; Zang, Y.H. Dicumyl peroxide modified

cellulose/LLDPE composites. J. Appl. Polym. Sci. 1990, 41, 2039.

10. Bataille, P.; Richard, L.; Sapieha, S. Effect of cellulose in polypropylene

composite. Polym. Compos. 1989, 10, 103.

11. Gouthier, R.; Joly, C.; Coupas, A.C.; Gouthier, H.; Escoubes, M. Interface

in polyolefin/cellulose fiber modified for composite application. Polym.

Compos. 1998, 19, 287.

12. Rowell, R.M.; Young, R.A.; Rowell, J.K. Agro-fiber thermoplastic

composites. In Paper and Composites from Agro-Based Resources;

Rowell, R.M., Young, R.A., Rowell, J.K., Eds.; Lewis Publishers: New

Jersey, 1996; 377–401.

13. Felix, J.M.; Gatenholm, P.; Schreiber, H.P. Controlled interaction in

cellulose polymer composites. I. Effect of mechanical properties. Polym.

Compos. 1993, 14, 449.

14. Tjong, S.C.; Xu, Y.; Meng, Y.Z. Composites based on maleated

polypropylene and methyl cellulosic fiber: mechanical and thermal

properties. J. Appl. Polym. Sci. 1999, 72, 1647.

15. Tjong, S.C.; Meng, Y.Z. Morphology and performance of potassium

titanate whisker reinforced polypropylene composites. J. Appl. Polym.

Sci. 1998, 70, 431.

16. Hill, C.A.S.; Abdul Khalil, H.P.S. Effect of fiber treatments on

mechanical properties of coir or oil palm fiber reinforced polyester

composites. J. Appl. Polym. Sci. 2000, 78, 1685.

17. Yap, M.G.S.; Que, Y.T.; Chia, L.H.L.; Chan, H.S.O. Thermal properties of

tropical wood-polymer composites. J. Appl. Polym. Sci. 1991, 43, 2057.

18. Nguyen, T.; Zavarin, E.; Barall, E.M. J. Macromol. Sci. Rev. Macromol.

Chem. 1981, C20, 1.

Performance of RWF–TPNR Composites 151

MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016

©2003 Marcel Dekker, Inc. All rights reserved. This material may not be used or reproduced in any form without the express written permission of Marcel Dekker, Inc.

Page 14: Performance of Rubberwood Fiber–Thermoplastic Natural Rubber Composites

19. Nabi Saheb, D.; Jog, J.P. Natural fiber polymer composites: a review.

Adv. polym. Technol. 1999, 18 (4), 351–363.

20. Quillin, D.T.; Caulfield, D.F.; Koutsky, J.A. Crystallinity in the

polypropylene/cellulose system. II. Crystallization kinetics. J. Appl.

Polym. Sci. 1994, 52, 605.

21. Rozman, H.D.; Kon, B.K.; Abusamah, A.; Kumar, R.N.; Mohd. Ishak,

Z.A. Rubber-high density polyethylene composites: effect of filler size

and coupling agents on mechanical properties. J. Appl. Polym. Sci. 1998,

69, 1993.

Sameni, Ahmad, and Zakaria152

MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016

©2003 Marcel Dekker, Inc. All rights reserved. This material may not be used or reproduced in any form without the express written permission of Marcel Dekker, Inc.