Mechanical and dynamic mechanical properties of rice husk ash-filled natural rubber compounds

Mechanical and Dynamic Mechanical Properties of RiceHusk Ash–Filled Natural Rubber Compounds

H. M. DA COSTA,1 L. L. Y. VISCONTE,1 R. C. R. NUNES,1 C. R. G. FURTADO2

1 Instituto de Macromoleculas Professora Eloisa Mano, Universidade Federal do Rio de Janeiro, P.O. Box 68525,21945-970 Rio de Janeiro, Brazil

2 DPI/IQ/UERJ, Rua Sao Francisco Xavier 524 Maracana, Rio de Janeiro, Brazil

Received 28 December 2000; accepted 30 April 2001Published online 2 January 2002

ABSTRACT: A laboratory-sized two-roll mill was used to incorporate rice husk ash intonatural rubber (NR). A conventional vulcanization system was used for curing and curestudies were carried out on a Monsanto rheometer. Physical testing of the NR vulca-nizates involved determining tensile and tear resistances and hardness. Swellingbehavior of NR compounds and scanning electron microscopy were used to investigatethe interaction between rice husk ash and natural rubber. Also, dynamical mechanicalthermal analysis was used to assess filler–rubber interactions in terms of storagemodulus (E�) and loss tangent (tan �). For comparison purposes, two commercial fillers,precipitated silica (Zeosil-175) and carbon black (N774), were also used. © 2002 JohnWiley & Sons, Inc. J Appl Polym Sci 83: 2331–2346, 2002

INTRODUCTION

Rice is the basis of peoples’ meals in Asia and iscultivated in substantial amounts in variousAmerican and European countries. The produc-tion of rice generates rice husks and straw. Chem-ical analysis of rice husk (RH) reveals the follow-ing typical composition (referred to as substancefree from loss on ignition): 9% water, 3.5% pro-tein, 0.5% fats, 30–42% cellulose, 14–18% pen-tosan, and 14–30% mineral ash. In view of thecomposition of rice husk, a number of suggestionshave been made for its large-scale use in agricul-ture or industry. These suggestions relate to di-rect use, with or without comminution; chemicaldecomposition of rice husk on an industrial scaleto obtain organic chemical basic materials; com-bustion for obtaining heat; and use of the mineralash residue.1,2

Some studies have shown the possibility of uti-lizing rice husk as an alternative material forgenerating magnesium silicide (Mg2Si) semicon-ductors3 and higher-quality solar grade silicon.4

The formation of SiC whiskers from compacts ofraw rice husks has also been reported.5

A filler derived from rice hulls appears to meetmany of the performance and cost requirementsfor use in rubber compounding. The compositionand structure of rice hulls make them an unusualand intriguing potential raw material from whicha filler for rubber could be prepared. Althoughpredominantly carbonaceous in nature, thesehulls contain a considerable amount of silica in ahydrated amorphous form. The ash content in ricehulls is high enough to yield a filler by burningaway the organic fraction. Chemical analysesshow that this ash primarily is composed of verypure silica and that no metal constituents arepresent in sufficient quantities to cause deleteri-ous aging effects upon rubber vulcanizates. Haxoand Mehta6 described the use of rice husk ash(RHA) as a reinforcing agent for synthetic and

Correspondence to: L. L. Y. Visconte.Journal of Applied Polymer Science, Vol. 83, 2331–2346 (2002)© 2002 John Wiley & Sons, Inc.DOI 10.1002/app.10125

2331

natural rubbers. In their work, they observed thatRHA did not adversely affect either the vulcani-zation characteristics or aging. In addition, RHAfiller is a satisfactory substitute for medium ther-mal (MT) black and, in blends with blacks, it canbe effectively used as a partial replacement forfiner and more reinforcing blacks.

Applications of RHA as filler in polymers havebeen reported by Fuad et al.,7,8 who observed thatincorporating rice husk ash into polypropyleneleads to a significant increase in the flexural mod-ulus of the composites, one that is comparable tothat imparted by commercial fillers such as mica.Ismail et al.9–12 reported the effect of rice huskash used as a filler for epoxidized natural rubber(ENR) compounds and the effect of a multifunc-tional additive (MFA) and silane coupling agentsin natural rubber compounds filled with RHA.Overall, studies show that MFAs can partiallyreplace Si69 without much effect on curing char-acteristics and mechanical properties.

Because rice husk is usually regarded as agri-cultural waste and, therefore, an environmentalhazard, we discuss in this article the interactionbetween rice husk ash and natural rubber interms of physical properties, dynamical mechan-ical thermal analysis (DMTA), scanning electronmicroscopy (SEM), and vulcanizate swelling. Forcomparison purposes, the two most frequentlyused commercial fillers, silica and carbon black,were also investigated.

EXPERIMENTAL

Materials and Compounding

All materials were used as received. Natural rub-ber was supplied by Irwin Industrial e Comercial

Ltda. and raw husk ash by EMBRAPA. Carbonblack (N762) was supplied by Columbian Chemi-cals Brasil S.A. and precipitated silica (Zeosil-175) by Rhodia Brasil Ltda. The antioxidant Ami-nox was supplied by Uniroyal Quımica S.A. Othercompounding ingredients, such as zinc oxide andstearic acid were standard reference materials. Sul-fur was supplied by Vetec Quımica Fina Ltda (RJ)and accelerator by Bann Quımica Ltda (SP), Brazil.

Rice husk ash was milled for 5 h and sieved ona 325-mesh sieve. RHs, when burned in the openair, yielded two types of filler: white rice husk ash(WRHA) and black rice husk ash (BRHA). Theupper layer of the RH mound, when subjected toopen burning, yielded BRHA in the form of acarbonized layer. The inner layer of the mound,being subjected to higher temperatures, yieldedWRHA.6

The formulation shown in Table I was em-ployed to evaluate the curing and physical prop-erties of natural rubber (NR) vulcanizates. Thechemical composition was obtained with induc-tively coupled plasma emission spectroscopy. Par-ticle size distribution was determined by aGLOBAL LAB Image (SP0550) software package.Surface area was calculated with the Brunauer–Emmett–Teller method on an ASAP 2010 accel-erated surface area and porosimetry system.Filler density was measured in a glass pyknome-ter and pH was determined by the ASTM D 1512method. The chemical and physical properties ofBRHA, WRHA, silica, and carbon black are pre-sented in Tables II and III.

Preparation of Mixes, Rheometry, and Preparationof Test Samples

Mixing was carried out on a two-roll mill at aspeed ratio of 1:1.25 at 70°C, according to ASTM

Table I Typical Formulation

Material phr

Natural rubber 100Zinc oxide 3.5Filler 0–50Stearic acid 2.5CBSa 0.8Aminoxb 2.0Sulphur 2.5

a N-cyclohexyl-2-benzothiazole-2-sulphenamide.b Antioxidant, low-temperature reaction product of di-

phenylamine and acetone.

Table II Chemical Composition of RiceHusk Ash

Chemical Composition (%) BRHA WRHA

CaO 0.44 0.75MgO 0.54 0.60Fe2O3 0.10 0.24K2O 1.23 1.28Na2O 0.25 1.17Al2O3 0.24 —MnO 0.15 0.13TiO2 0.01 —P2O5 0.99 0.17SiO2 (silica) 41.62 77.18Loss on ignition (LOI) 25.26 2.36

2332 H. M. DA COSTA ET AL.

D 3182. RH ash was dried at 120°C for 24 himmediately before use. Optimum cure times at150°C were obtained from a Monsanto RheometerTM-100. Mixes were vulcanized in an electricallyheated press at 150°C and 3.0 MPa. Vulcanizateswere conditioned for 24 h before testing. All prop-erties were measured in the direction of the grain.

Physicomechanical Testing of Samples

Stress–strain data were determined according toASTM D 412 on an Instron universal machineusing C-type dumbbell specimens. Other physico-mechanical tests were tear strength (ASTM D624) and hardness (ASTM D 2240).

DMTA

The dynamic mechanical properties of rubber vul-canizates filled with BRHA, WRHA, and commer-cial fillers were determined with DMTA equip-ment (Rheometric Scientific, model MKIII; Plus Vsoftware, version 5.42). The run conditions weresingle-cantilever mode of deformation geometry,known as clamped-bending; temperature rangefrom �80 to �20°C at a heating rate of 2°C/min;strain amplitude of 0.05%; and frequency of 1 Hz.To ensure uniformity of the rubber compound,three test pieces were selected from differentparts of the same compression-molded sheet foreach composition. Subsequently, the storage mod-ulus (E�) and the maximum loss tangent (tan�max) were determined by averaging the threedeterminations.

SEM

Examination of the fracture surface was carriedout on a scanning electron microscope, modelJEOL JSM-5300. The objective was to get an in-sight into the fracture mode in an attempt todraw a picture of the matrix and filler surfaces

and filler dispersion. The fractured ends of thetensile specimens were mounted on aluminumslabs and spatter coated with a thin layer of goldto avoid electrical charging during examination.

Swelling Testing

Small pieces, approximately 2 � 2 � 0.3 cm, werecut from the vulcanized sheets. The dry unswol-len samples were weighed occasionally untilweight variations �0.0001g were achieved. Thespecimens were then immersed in heptane at30°C and maintained in darkness. The swollenspecimens were then taken out periodically, theexcess liquid on the surface was wiped off, and thespecimens were immediately weighed with ut-most care. The swelling degree of the NR vulca-nizates was then calculated.

RESULTS AND DISCUSSION

Effect of BRHA, WRHA, and Commercial Fillers onthe Cure Parameters and Mechanical Properties ofNR Vulcanizates

Table IV summarizes the values of optimum curetime (t90), scorch time (ts2), and minimum andmaximum torque of BRHA-, WRHA-, silica-, andcarbon black–filled NR vulcanizates. Increasedamounts of WRHA, BRHA, and carbon black hadthe same effect both on t90 and ts2: scorch timeslightly increased whereas t90 practically was in-dependent of the filler content. Compared to puregum, the presence of carbon black accelerates thevulcanization process but reduces scorch time. Onthe other hand, WRHA and BRHA ashes do notinterfere with these parameters. With increasingamounts of filler, silica imparts a somewhat dif-ferent behavior as both t90 and scorch time in-crease. This different trend in cure characteristicsmay be attributed to differences in filler proper-ties such as surface area, surface reactivity, par-ticle size, moisture content, and metal content. Ingeneral, a faster cure rate is obtained with fillersthat have low surface area, high moisture con-tent, and high metal oxide content.13 This indeedseems to be the case as, among the fillers inves-tigated in this work, carbon black and WRHAhave the lowest surface areas and, therefore,shorter t90 values. The cure retardation for thesilica vulcanizates can be attributed to a silica–accelerator interaction. This filler reacts with zincoxide and subsequently reduces zinc reactivity,thus slowing the sulfur reaction.14

Table III Physical Properties of Rice HuskAsh, Silica, and Carbon Black

Properties BRHA WRHA SilicaCarbonBlack

Mean particlesize (�m) 2.5 2.2 0.018 0.054

Surface area(m2/g) 109 17 185 30

Density (g/cm3) 1.9 2.0 2.0 1.9pH 9.5 9.4 6.5 6.4

PROPERTIES OF NATURAL RUBBER COMPOUNDS 2333

The marked increase in maximum torque withincreasing filler loadings indicates that the pres-ence of fillers in the rubber matrix reduces themobility of the rubber’s macromolecular chains.The high values of maximum torque for carbonblack and silica composites indicate that there isa high restriction to the molecular motion of themacromolecules, probably caused by the greaterinteraction between these commercial fillers andthe rubber matrix. This seems to be a reflection ofthe particle size of the fillers (Table II).

Table V summarizes the mechanical propertiesof BRHA-, WRHA-, silica-, and carbon black–filled NR vulcanizates. In any of the studied sys-tems, tensile strength increases with increasingfiller content until a maximum level is reached atapproximately 20 phr, as shown in Figure 1. Afurther increase in filler loadings has a deleteri-ous effect on this property. The reduction instrength may be caused by filler particles agglom-erating to form domains that act as foreign par-ticles, or it simply may be the result of physicalcontacts between aggregates.

The reinforcement of elastomers by particulatefillers has been studied in depth in numerousinvestigations and it is generally accepted thatthis phenomenon is, to a large extent, dependenton polymer and filler properties and on process-ing. Generally speaking, the primary filler factorsthat influence elastomer reinforcement15 are par-ticle size, which determines surface area per unitweight and, with that, the solid–elastomer inter-face per cm3 of compound; specific surface activityper cm2 of surface area; shape and structure, asdetermined by void volume under standard pack-ing conditions; and particle porosity, usually re-ferred to as pores of very small size. These prop-erties usually are not mutually independent;therefore their product rather than their sum de-termines final behavior.

As seen in Table V, the continuous addition ofBRHA results in values of tensile strength thatincrease at the beginning, reach a maximum at 20phr, and then start to decrease with additionalamounts of filler, yet still do not show a signifi-cant variation when compared with pure gum.

Table IV Cure Time (t90), Scorch Time (ts2), and Minimum and Maximum Torque of BRHA-, WRHA-,Silica-, and Carbon Black–Filled NR Vulcanizates

FillerLoading

MinimumTorque(dN � m)

MaximumTorque(dN � m)

Optimum Cure Timet90 (min)

Scorch Timets2 (min)

0 8.3 43.2 16.3 6.6WRHA

10 10.0 48.0 14.5 6.020 10.3 48.7 15.4 6.230 10.7 54.5 14.4 6.540 10.8 57.3 14.2 6.650 10.9 58.3 15.1 7.1

BRHA10 8.4 44.7 17.1 6.020 10.5 48.2 17.5 6.330 12.2 52.1 16.3 6.540 13.6 54.3 17.3 6.650 14.2 56.8 16.5 6.7

Carbon black10 10.3 49.2 12.5 4.220 11.9 53.9 13.1 4.330 12.5 57.4 13.4 4.440 13.3 63.5 14.4 4.650 14.9 69.0 13.4 4.8

Silica10 11.6 45.3 16.6 6.420 14.9 48.2 18.3 7.130 17.6 53.2 20.0 8.540 19.5 57.5 22.5 9.250 23.9 68.8 25.1 10.4


Perhaps the particle size of BRHA is responsiblefor this because, according to Fetterman,16 fillerswith sizes on the order of 50 nm or greater areclassified as semi- or nonreinforcing. On the otherhand, WRHA has a positive effect on this prop-erty, at least up to 30 phr. The superior tensilestrength of the WRHA-filled vulcanizate suggeststhat factors other than particle size influence theproperties of NR vulcanizates, factors such as,probably, surface activity and the bonding qualitybetween WRHA and the NR matrix. In addition,the dual nature of the BRHA filler, given by thepresence both of silica and carbon components,with different physical and chemical properties,may also reduce the efficiency of this filler instrengthening the rubber matrix.

From the SEM photomicrographs of BRHA-,WRHA-, silica-, and carbon black–filled NR com-posites shown in Figure 3, it can be seen that thefracture surface of WRHA is more uniform com-pared with that of BRHA. However, the lack ofintensive interactions between WRHA filler and

the NR matrix is evident from these photomicro-graphs when they are compared with the moreuniform fracture surfaces of silica and carbonblack vulcanizates, which are smoother becauseof better filler dispersion. In addition, the largerparticle size and, therefore, smaller surface areaof both ash fillers, which implies poorer filler–rubber interactions, is evident on the fractureplane.

Tear strength data presented a tendency todecrease in BRHA- and WRHA-filled systems, asshown in Table V and Figure 2. Although theyhave similar behaviors, WRHA composites havesuperior values for this property, as comparedwith BRHA, because of their higher silica content.The formulation containing 20 phr of filler, whichgave the best tensile strength results both forBRHA and WRHA, did not perform as satisfacto-rily as did the commercial fillers. Tear strength,like tensile strength, is affected by filler particlesize and surface area. In addition, this property iscontrolled by the nature both of the rubber and

Table V Mechanical Properties of NR Mixed with BRHA, WRHA, Silica, and Carbon Black

Property

TensileStrength

(MPa)Elongation at

Break (%)

Modulus at300%(MPa)

TearStrength(kN/m)

Hardness(Shore A)

Filler loading0 19.5 900 3.5 30.7 35

WRHA10 21.5 820 6.2 29.1 3720 22.7 755 6.8 27.8 3930 20.0 710 7.9 26.8 4240 18.7 690 8.4 23.8 4550 17.8 660 9.0 23.0 47

BRHA10 19.7 855 5.9 26.5 3620 20.2 820 6.0 25.7 3830 18.4 765 6.2 24.2 4140 17.5 720 7.0 21.0 4350 16.0 680 7.2 19.1 45

Carbon black10 23.5 750 6.0 32.7 3920 25.2 675 8.7 35.5 4430 23.0 620 10.4 38.9 4840 22.4 575 11.6 42.5 5250 21.0 530 12.5 48.7 56

Silica10 22.2 810 5.8 33.6 3820 23.5 750 7.8 37.8 4230 22.0 670 9.9 40.5 4740 20.9 620 10.5 47.9 5050 19.5 520 11.8 55.6 52


the filler, as well as by the rate and temperatureof tearing.17 Despite having a silica content com-parable to that of commercial silicas, the WRHA-filled vulcanizates are not as tear-resistant as arethose containing the commercial product. There-fore, the different performances may be the resultof a much more significant influence of the nature

of the surface, particle size, and surface area ofthe filler used.

As for the stiffness properties, the trend ob-served is that which was already expected. Hard-ness and modulus at 300%, both for BRHA- andWRHA-filled vulcanizates, increased with in-creasing filler content, as shown in Table V. This

Figure 1 Effect of filler loading on tensile strength of BRHA-, WRHA-, silica-, andcarbon black–filled NR vulcanizates.

Figure 2 Effect of filler loading on tear strength of BRHA-, WRHA-, silica-, andcarbon black–filled NR vulcanizates.


result was expected because, as more filler parti-cles are introduced into the rubber, the elasticityof the rubber chains is reduced, resulting in morerigid vulcanizates. In general, BRHA- andWRHA-filled vulcanizates have lower modulusand hardness than do the corresponding com-pounds with silica and carbon black. This againmay be caused by differences in filler properties.The inferior stiffness of BRHA- and WRHA-filledvulcanizates may be explained by two factors.First, these fillers have larger particle sizes and,therefore, smaller surface areas. Second, WRHAand, mainly, BRHA fillers show a greater ten-

dency toward aggregation. The SEM photomicro-graphs shown in Figure 3 confirm that, in addi-tion to possessing larger particle size and broadparticle size distribution, the dispersion of WRHAand BRHA fillers in the rubber matrix is notuniform when compared with silica and carbonblack. This poor filler dispersion reduces filler–rubber interactions and consequently decreasesthe ability of these fillers to restrain gross defor-mation of the rubber matrix.

The effect of fillers on the crosslink densities ofvulcanizates is determined by modulus measure-ments or by swelling experiments. The latter

Figure 3 SEM photomicrographs of WRHA-, BRHA-, silica-, and carbon black–filledNR at 20 phr filler loading after tensile fracture (�150).


method yields some interesting results becausethe solid surface of the filler can develop threetypes of interaction with the elastomer: adher-ence at fixed points, nonadherence, and adher-ence with two-dimensional mobility over the sur-face of the filler. The difference in the type ofinteraction becomes clear when a vulcanizate isswollen in solvents of increasing swelling powerand the degrees of swelling are compared withthose for the unfilled vulcanizate. This compari-son assumes that the chemical crosslink densityof the polymer has not been changed by the pres-ence of the filler.15

The degree of swelling usually is expressed interms of the reciprocal of the fraction:

Vr � volume of rubber/

volume of the swollen rubber � solvent gel (1)

Vr is very much dependent on the swellingpower of the solvent (high swelling power meanslow Vr) and the crosslink density. Highercrosslink density means restraint on the networkand thus results in lower swelling, that is, higherVr (in the same solvent).

In a filled vulcanizate the important parameteris

Vrf � volume of rubber in vulcanizate/

volume of swollen rubber � gel (2)

where the numerator refers to the total volumeminus the filler volume and the denominator isthe total swollen volume minus the filler volume.It is of interest to note that the ratio R � VrO/Vrf(VrO being Vr for the unfilled vulcanizate) withincreasing filler loadings.

The Lorenz–Parks model,18 represented by eq.(3), was used to investigate the swelling of filler-reinforced vulcanizates.

Qf /Qg � a � e�z � b (3)

where Q is the amount of solvent imbibed per unitweight of rubber, f and g refer to filled and gummixes, respectively, z is the weight fraction offiller in the polymer, and a and b are constantsthat depend on filler activity. A high a value anda low b value indicate strong polymer attach-ment.19

Lorenz and Parks18 explain the swelling be-havior of a filled stock without assuming that the

presence of filler would enhance the crosslinkingefficiency of the curing agent. The filled stock isthought to be composed of x different zones thatexhibit different swelling abilities. The zonesnearest the filler particles would exhibit thesmallest swelling ability whereas zones suffi-ciently far from the particles would be identicalwith a gum phase crosslinked to the same extent.A filled stock composed of different zones withdifferent swelling abilities would exhibit a swell-ing value proportional to that of an identicallycrosslinked gum stock, provided that the numberand the size of the zones were independent of thenumber of crosslinks introduced by chemicalmeans.

In the present article, the Cunnen–Russelequation,19 given in eq. (4) and originally derivedby Lorenz and Parks,18 was used.

VrO/Vrf � a � e�z � b (4)

where VrO and Vrf are the volume fractions of therubber in unfilled and filled vulcanizates, respec-tively, swollen in a solvent.

Another model used to investigate the swellingof filler-reinforced vulcanizates is the Kraus mod-el,20 given in eq. (5), which examines the conse-quences of the Lorentz and Parks model18 andshows how to calculate quantitatively the effecton the swelling of particles either completely un-bound or completely and permanently bonded tothe polymer.

VrO/Vrf � 1 � �m�/�1 � �� (5)

where

m � 3C�1 � VrO1/3� � VrO � 1 (6)

VrO and Vrf are the volume fractions of rubber inthe solvent-swollen gum and filled vulcanizates,respectively. C is a constant characteristic of thefiller but is independent of the solvent; � is thevolume fraction of filler in the vulcanizate. Theparameter m, obtained from the slope of the plotVrO/Vrf versus �/(1 � �), describes how muchswelling is restricted for a given volume fractionof filler; it is basically a measure of the carbon–polymer interaction during the swelling process.Similarly, VrO is a measure of the polymer–sol-vent interaction, the exact relationship beinggiven by the Flory–Rehner theory.21 If the fillerswells just as much as the surrounding rubber


matrix, then the ratio VrO/Vrf will be equal tounity, even with the increase in filler percentage.However, most fillers do not swell and, if themovements of the matrix are restricted becausepolymer segments are attached to the filler sur-face, the ratio VrO/Vrf will decrease as filler load-ing increases. Remarkably, this ratio increaseswith increasing filler fraction when the filler is ofthe nonadhering type, as pointed out by Kraus.20

The reason for this is that, in this case, a pocketfilled with solvent forms around each filler parti-cle. Because the solvent in the pocket is not taken

into account, the ratio VrO/Vrf increases with thefiller volume fraction.15 Swelling results for vul-canizates containing BRHA, WRHA, and the com-mercial fillers are shown in Figures 4 and 5.

Kraus20 suggests, for the case of carbon black,that swelling is restricted at the carbon surfacebut becomes normal at a distance sufficiently farfrom this surface. Boonstra and Taylor22 suggestthat the restriction of swelling by reinforcing fill-ers, especially carbon black, could be explained bytwo other mechanisms: (1) the filler has a cata-lytic effect on vulcanization so that different

Figure 4 Kraus equation for swelling test of NR vulcanizates containing rice huskash and commercial fillers.

Figure 5 Cunnen–Russel equation for swelling test of NR vulcanizates containingrice husk ash and commercial fillers.


states of crosslinking are obtained depending onwhether one is dealing with filled or unfilled com-positions or (2) additional crosslinks are formedon the filler surface and the carbon black particlesact as giant crosslinks.

In addition, fillers can be separated into twodifferent groups20,22: (1) those whose adhesionforces to the surrounding polymer are weak; theseforces are almost negligible compared with thoseacting at the solvent/filler interface and those be-tween solvent and polymer (swelling pressure)and (2) those whose adhesive forces to the sur-rounding polymer molecules are sufficientlystrong so that these forces are not at all or notcompletely displaced by solvent molecules.

Figure 4 shows the plot of VrO/Vrf versus �/(1� �) according to the Kraus equation.20 It can beobserved that BRHA compositions exhibit devia-tions, in relation to the proposed linear mathe-matical model, that are greater than those for thecommercial fillers. In the case of BRHA, the VrO/Vrf ratio has an ascendant profile with values 1as filler loading increases; this means that thematrix around the filler particles will swell and,because the particle is not affected by the solvent,it will keep its original volume and the swollenelastomer will separate from the particle.20 Theporosity of this filler is another point to be con-sidered. As shown in Table III, BRHA and WRHAhave comparable particle sizes but the surfacearea of BRHA is much larger than that of WRHA,which probably is caused by the presence of pores.Voet23 and Pal24 investigated the reinforcementby silica and suggested that the deviation of lin-earity found in this material seems to indicate theonset of dewetting and void formation; the spacesbetween the particle and the swelling vulcanizatebeing filled with solvent, they no longer fulfill theconditions set forth in the Kraus model. In itsturn, WRHA presents a larger interaction withthe NR matrix because, for this filler, in the load-ing range studied, VrO/Vrf is �1.

The systems that use the commercial fillersshow not only a good fit to the Kraus model butalso a more effective interaction filler/NR matrix(the ratio VrO/Vrf � 1 is an indication), mainly forthe composites containing carbon black. An expla-nation for a deviation from the linear relationshipat higher loadings may be that, because of thehigh viscosity imparted by these high black load-ings, the mechanical breakdown of the polymerduring mixing is much more intense than it is atlow loadings. Therefore, the crosslinking level isnot the same as it is at zero or at low black

loadings, where breakdown does not take place,and a lower Vrf is found.22 This increases theVrO/Vrf ratio. Another explanation for this phe-nomenon, put forward by Kraus,20 points out thatthe presence of carbon affects the accelerator sys-tem so that the VrO of the gum vulcanizate is nolonger representative of the crosslink density at-tained in the presence of black.

Figure 5 shows the plots according to the Cun-nen–Russel equation.19 It is again observed thatBRHA filler does not adhere very well to the NRmatrix (VrO/Vrf 1) whereas carbon black andsilica develop strong interaction (VrO/Vrf � 1).WRHA exhibits the behavior already observed inthe Kraus model.

Effect of BRHA, WRHA, and Commercial Fillers onDynamic Mechanical Properties of NR Compounds

The mechanical properties of filled rubbers usu-ally are described in terms of tensile strength,stiffness, abrasion, and tearing properties. Butfillers, when added to polymer, are also known tocause a considerable change in the dynamic prop-erties, not only of the dynamic moduli, both vis-cous and elastic, but also in their ratios, i.e., theirloss factor, which is related to the portion of en-ergy dissipated during dynamic deformation. Onemethod that has been used to characterize filler–rubber interactions is DMTA. This technique isparticularly useful as a nondestructive test foridentifying the molecular mechanisms of polymermaterials.

Figures 6–9 show the dependence of storagemodulus (E�) on temperature for different concen-trations of WRHA, BRHA, carbon black, and sil-ica in the NR vulcanizates. The storage modulusdecreases with increasing temperature. This ten-dency was proved to be very extreme around thetransition region, being caused by the increasingmobility of polymer chains with increasing tem-perature. In the glassy region (approximately�80 to �55°C), the change in the modulus E� withdifferent filler concentrations seemed to be small.However, this effect gradually became more pro-nounced upon entering the transition and rub-bery regions.

The variation of E� at various WRHA contents(Fig. 6) in the rubbery region is of interest here.E� decreased with increasing WRHA filler con-tent, up to 20 phr, and then increased at higherloadings. In Figure 7 it can be seen that the be-havior of E� for BRHA compounds was the same.However, for the carbon black and silica com-


pounds, E� increased with increasing filler con-tent, up to 20 phr for carbon black and 30 phr forsilica, and then decreased at higher loadings.

The results for WRHA and BRHA may be in-terpreted as the rubber being trapped or caged.With increasing filler concentration and lowstrain amplitude, interaggregate association (fill-er networks formed in the polymer matrix andkept in association by physical forces) is not bro-

ken down by straining so that the rubber trappedin the filler network or agglomerates is at leastpartially “dead,” losing its identity as an elas-tomer and instead behaving as a filler in terms ofstress–stain. Therefore, the effective volume ofthe polymer bearing the stress imposed upon thesample is reduced by filler networking, which re-sults in an increased modulus that is governedprimarily by the filler concentration. The break-

Figure 6 Influence of WRHA content on E� of NR compositions as a function oftemperature.

Figure 7 Influence of BRHA addition on E� of NR vulcanizates as a function oftemperature.


down of the filler network by increasing strainamplitude would release the trapped rubber sothat the modulus would decrease.25,26

Concerning the commercial fillers, Sombatsom-pop27 observed that the carbon black–rubber in-teraction is attributed to many phenomena, in-cluding (1) the formation of more stable, short poly-sulfide linkages (low x value in ROSOSxOSOR�structures), which results in greater mechanicalproperties; (2) the formation of bound rubber,which involves the reaction of carbon black with

free radicals formed by mechanicochemical scis-sion of the polymer; and (3) a reaction of activesurfaces of carbon black formed during com-pounding with the rubber.

It also has been argued that, depending on thestrength of the polymer–filler interaction, physi-cal and/or chemisorption of rubber moleculesmight take place on the filler segments. Depend-ing on the intensity of the filler–polymer interac-tion and the distance from the filler surface, themobility of the polymer segments near the inter-

Figure 8 Influence of carbon black content on E� of NR vulcanizates as a function oftemperature.

Figure 9 Influence of silica content on E� of NR vulcanizates as a function of tem-perature.


face is less than it is in the matrix. The decreasein the modulus of the rubber vulcanizates, whichwas observed when higher filler loadings wereadded, occurred because the packing of carbonblack and silica aggregates had reached a criticalpoint where they were no longer separated by thepolymer matrices. In the case of a highly polarfiller, such as silica, that is incompatible withhydrocarbon rubbers, filler agglomerates maytake place primarily by contact between aggre-

gates, in this particular case perhaps even viahydrogen bonding, to give a rigid construc-tion.25,26

Figures 10–13 show the dependence of thedamping characteristics, measured as the tan-gent of the phase angle (tan �) on temperature, fordifferent concentrations of WRHA, BRHA, carbonblack, and silica in the rubber vulcanizates.

It can be seen that, at sufficiently low temper-atures, tan � is very low because the viscosity of

Figure 10 Effect of WRHA content on tan � of NR composites as a function oftemperature.

Figure 11 Effect of BRHA content on tan � of NR vulcanizates as a function oftemperature.


the rubber is so high and the free volume in thepolymer is so small that the movement of thepolymer segments, and the adjustment of theirrelative positions, can hardly take place duringthe time involved in the normal dynamic experi-ment (frequency). This results in low energy dis-sipation and, therefore, low hysteresis. Underthis condition, the polymer decreases in theglassy state with a very high elastic modu-lus.26–28

When temperature is increased, the move-ments of the polymer segments increase. Whenthe temperature reaches a certain level, the freevolume associated with the polymer increasesmore rapidly than does the volume expansion ofthe molecules, facilitating the segmental motion.From this point, which is known as the glasstransition temperature (Tg), the viscosity of thepolymer decreases very rapidly and the molecularadjustments take place more easily so that the

Figure 12 Effect of carbon black addition on tan � of NR vulcanizates as a function oftemperature.

Figure 13 Effect of silica addition on tan � of NR vulcanizates as a function oftemperature.


elastic modulus is decreased and the energy dis-sipation among polymer molecules will increasewith increasing temperature. This results in highhysteresis.26 However, at temperatures highenough for Brownian motion to become so rapidand viscosity to become so low that, in the poly-meric solid, the thermal energy is comparable tothe potential energy barriers to segment rotation,the molecular adjustment is quick enough to beable to follow the dynamic strain. In this case,long-range contour changes of polymer moleculesmay take place that are associated with high en-tropic elasticity and low resistance to strain. Thematerial falls in a so-called rubbery region that ischaracterized by low modulus and low energy dis-sipation during dynamic deformation.28

It can be seen that the tan � values of rubbercompounds filled with carbon black and silica(Figs. 12–13) were lower than those found in rub-ber gum. The tan �max values of the vulcanizatesdecreased considerably when commercial fillerswere added. The decreased value of tan �max inthe investigated load range suggests that a rela-tively high interaction of the filler and rubber wasobtained. For rubber compounds with WRHA andBRHA fillers (Figs. 10–11), the decrease in tan �values with increased filler content was not asintense. This again may be caused by differencesin filler properties. WRHA and BRHA fillers havelarger particle sizes and, therefore, smaller sur-face areas and show a greater tendency to aggre-gate, as seen in SEM photomicrographs (Fig. 3).

In addition, the results shown in Figures10–13 lead to the conclusion that the loadingeffects at different temperature regions are gov-erned by different mechanisms. It seems that, attemperatures near the tan � peak in the transi-tion zone, the presence of fillers gives a lowerhysteresis for a given energy input. This may beinterpreted as a reduction in the polymer fractionwithin the composite because the polymer per sewould be responsible for the high proportion ofenergy dissipation whereas individual solid fillerparticles in the polymer matrix may not absorbenergy significantly. Although this interpretationis reasonable for hysteresis in the transition zone,it may be not true at high temperatures wherehysteresis is increased when filler is introduced.It has been demonstrated routinely in rubbercompounds that, when the same volume fractionof fillers is incorporated into the same polymersystem, fillers with different morphologies andsurface characteristics have different responses

to the temperature dependency of the hystere-sis.25–28

A further observation of the damping charac-teristics was made by considering the overallwidth of the loss tangent peaks. It was found that,in the presence of fillers, the peaks became clearlybroadened, which indicates greater dynamic loss-es; this was more pronounced around the rubberyregion. The losses could be caused by more me-chanical energy input being converted into heat,which is generated during deformation, and moreenergy losses from the movements of polymerchains in the rubbery state. The Tg values of filledrubber vulcanizates shifted to higher tempera-tures and showed higher values than did those ofthe unfilled compound. The shift of Tg to highertemperatures might be caused by the presence ofcrosslinks, which restrict the mobility of the poly-mer chains, and by a slight increase in the inter-action between fillers and rubber phase, whichoccurs principally in commercial fillers.25–28

CONCLUSION

The main objective of this study was to verify thepossibility of utilizing rice husk ash as alternativefiller for NR. Concerning the mechanical proper-ties, the results show that the tear strengths andhardnesses of BRHA-, WRHA-, carbon black-, andsilica-filled NR compounds are very similar forlow filler loadings; regarding tensile strength,BRHA is not as efficient as the commonly usedfillers carbon black and silica. As for WRHA, inspite of parameters such as surface area and par-ticle size, the vulcanizate with 20 phr of this filler,which gave the best results, showed physicalproperties not much inferior to commercial car-bon black- or silica-filled vulcanizates. Further-more, cure characteristics were not prejudicedwith increased filler loading, in spite of the non-homogeneous nature of WRHA chemical composi-tion. From the analysis of VrO/Vrf as a function offiller loading and from the behavior found by ap-plying the Kraus and Cunnen–Russel models, itcan be concluded that WRHA adheres better tothe rubber matrix than does BRHA. However,although WRHA can have better wetting and dis-persion, this was not enough to cause an enhance-ment in the dynamic mechanical performance ofWRHA as reinforcing filler.

Nevertheless, rice husk ash can still be thoughtof as an alternative filler, depending on the per-formance desired for a given application, because


of its extremely low cost as a by-product of the riceindustry. There is also the possibility of having itbehave as reinforcing filler if it is obtainedthrough controlled burning or is submitted tochemical treatment. However, the most impor-tant reason why these ashes should be given agood use of is the possibility of minimizing envi-ronmental impact by the adequate utilization ofthis residue (1 kg of husks produces 300 g of ash).At present, the most common method of husk ashdisposal is wasteland dumping, which creates anenvironmental hazard through pollution and landdereliction.

REFERENCES

1. Tutsek, A.; Bartha, P. U.S. Patent 4,049,464, Sept.20, 1977.

2. Mehta, P. K. U.S. Patent 5,346,548, Sept. 13, 1994.3. Bose, S.; Acharya, H. N.; Banerjee, A. D. J Mater

Sci 1993, 28, 5461.4. Hunt, L. P.; Dismukes, J. P.; Amick, J. A.; Schei, A.;

Larsen, K. J Eletrochem Soc Solid-State Sci Tech1984, 131, 1683.

5. Krishnarao, R. V.; Godkhindi, M. M.; Mukunda,P. G. J Mater Sci 1994, 29, 2741.

6. Haxo, H. E.; Mehta, P. K. Rubber Chem Technol1975, 48, 271.

7. Fuad, M. Y. A.; Ismail, Z.; Mansor, M. S.; Ishak,Z. A. M.; Omar, A. K. M. Polym J 1995, 27, 1002.

8. Fuad, M. Y. A.; Ismail, Z.; Ishak, Z. A. M.; Omar,A. K. M. Eur Polym J 1995, 31, 885.

9. Ismail, H.; Ishiaku, U. S.; Lu, E. S.; Ishak, Z. A. M.Intern J Polymeric Mater 1997, 38, 275.

10. Ismail, H.; Ishiaku, U. S.; Lu, E. S.; Ishak, Z. A. M.Intern J Polymeric Mater 1998, 39, 45.

11. Ismail, H.; Nasaruddin, M. N.; Rozman, H. D. EurPolym J 1999, 35, 1429.

12. Ismail, H; Nasaruddin, M. N.; Ishiaku, U. S. PolymTesting 1999, 18, 287.

13. Wagner, M. P. Rubber Chem Technol 1976, 49, 703.14. Nasir, M.; Poh, B. T.; Nig, P. S. Eur Polym J 1988,

24, 961.15. Boonstra, B. B. Polymer 1979, 20, 691.16. Fetterman, E. Q. Rubber World 1985, 194, 38.17. Mark, J. E.; Erman, B.; Eirich, F. R. Science and

Technology of Rubber, 2nd ed.; Academic: NewYork, 1994.

18. Lorenz, O.; Parks, C. R. J Polym Sci 1961, 50, 299.19. Cunnen, J. I.; Russel, R. M. J Rubber Res Inst

Malaya 1969, 22, 308.20. Kraus, G. J Appl Polym Sci 1963, 7, 861.21. Flory, P. J.; Rehner, J., Jr. J Chem Phys 1943, 11,

521.22. Boonstra, B. B.; Taylor, G. L. Rubber Chem Tech-

nol 1965, 38, 943.23. Voet, A.; Morawski, J. C.; Donnet, J. B. Rubber

Chem Technol 1977, 50, 342.24. Pal, P. K.; De, S. K. Rubber Chem Technol 1982,

55, 1370.25. Wang, M. J. Rubber Chem Technol 1998, 71, 520.26. Wang, M. J. Rubber Chem Technol 1999, 72, 430.27. Sombatsompop, N. Polym Plast Technol Eng 1998,

37, 1.28. Aklonis, J. J.; Macknight, W. J. Introduction to

Polymer Viscoelasticity, 2nd ed.; John Wiley andSons: New York, 1983.


Mechanical and dynamic mechanical properties of rice husk ash-filled natural rubber compounds

Documents