Effects of rubber curing ingredients and phenolic-resin on mechanical, thermal, and morphological...

Effects of Rubber Curing Ingredients and Phenolic-Resin on Mechanical, Thermal, and MorphologicalCharacteristics of Rubber/Phenolic-Resin Blends

Babak Derakhshandeh, Akbar Shojaei, Morteza Faghihi

Department of Chemical and Petroleum Engineering, Sharif University of Technology, Tehran 11365-9465, Iran

Received 12 September 2007; accepted 26 December 2007DOI 10.1002/app.28034Published online 12 March 2008 in Wiley InterScience (www.interscience.wiley.com).

ABSTRACT: This article examines the physical and me-chanical characteristics of mixtures of two different syn-thetic rubbers, namely styrene-butadiene rubber (SBR) andnitril-butadiene rubber (NBR), with novolac type phenolic-resin (PH). According to Taguchi experimental designmethod, it is shown that the addition of PH increases thecrosslinking density of rubber phase probably due to itscurative effects. Thermal analysis of the blends indicatesthat, contrary to NBR/PH blend, thermal stability of SBR/PH blend is dependent on sulfur content due to predomi-nant polysulfidic crosslinks formed in SBR. Slight shift inglass-transition temperature (Tg) of pure SBR and NBRvulcanizates by the addition of PH suggests that bothSBR/PH and NBR/PH are incompatible blends with a par-

tially soluble PH in the rubber phase. Two-phase morphol-ogy of the mixtures is also evidenced by scanning electronmicroscopy. Correlation of the rubber/PH modulus versusPH concentration by Halpin-Tsai model shows a deviationfrom the model. Presence of PH in the rubber phase isthought to vary the mechanical properties of the rubberphase by changing both the crosslinking density and rigid-ity of the molecular network of the rubber, leading tomisuse of modulus of pure rubber in Halpin-Tsai equa-tion. � 2008 Wiley Periodicals, Inc. J Appl Polym Sci 108: 3808–3821, 2008

Key words: morphology; blending; rubber; phenolic-resin;crosslinking

INTRODUCTION

Phenolic-resin (PH) has been the most common poly-meric binder that is widely used in the compositefriction materials.1–4 This may be due to its low cost,relatively high heat resistance, i.e., degradation tem-perature of above 4508C,5 and suitable processability.However, the widespread use of PH is restricted byits shortcomings such as brittleness and toxicity.3,6

Thus, it needs to be modified adequately to meet therequired properties for specific applications. Forinstance, PH may be modified with tough polymerssuch as epoxy resin, silicon oil, and silicon rubber toimprove its flexibility.7,8 Additionally, blends of asuitable rubber and PH have also found applicationsin friction materials, particularly in frictional brakingsystem of railroad vehicles,9–11 to achieve a poly-meric binder with combined properties of the rubbermaterial and the resin. The content of the resin insuch blends could be reached up to 40 wt %. Therubber component of the blend could be natural orsynthetic, mainly nitril-butadiene rubber (NBR) andstyrene-butadiene rubber (SBR).9–11 Presence of rub-

ber in the blend makes the friction material conform-able and compressible, which are essential in appli-cations such as railroad friction materials to domi-nate the thermal damages of the friction couple.12,13

Literature shows a limited number of studiesdevoted to the rubber/PH systems.14–16 In addition,the patents on this subject have provided a few tech-nical details due to proprietary reasons.9–11 Kosfeldand Borowitz14,15 studied dynamic mechanical andviscoelastic properties of blends of PH with NBRand SBR with full range of the resin content, i.e., 0–100 vol %. There exists only one recent article on therubber/PH blends to be used specifically as binderof friction material in which the role of heating onthe properties of NBR and PH mixtures has beeninvestigated.16

It is well known that the mechanical and thermalproperties of a rubber vulcanizate are significantlydominated by the rubber curing ingredients includ-ing sulfur, accelerator, and activator.17 It has beenalso claimed that increasing concentration of zinc ox-ide (activator of sulfur vulcanizing system) increasesthe wear rate of friction materials containing theSBR/PH as polymer matrix.11 Therefore, it isexpected that the type and composition of rubbercuring ingredients contribute to the final propertiesof rubber/PH systems as well. The role of rubbercuring ingredients on the properties of NBR andSBR vulcanizates has been extensively studied long

Correspondence to: A. Shojaei ([email protected]).Contract grant sponsor: Railway Research Center of

Iran.

Journal of Applied Polymer Science, Vol. 108, 3808–3821 (2008)VVC 2008 Wiley Periodicals, Inc.

time ago. However, such study for blends of the rub-ber and PH has not been addressed earlier.

Investigation of compositional parameters in rub-ber/PH blends by one-factor-at-a-time method, i.e.,varying only one parameter while keeping all othersconstants, is very complicated task and requires per-forming a large number of experiments. To reducethe size of experiments, several experimental designmethods have been presented.18 Among them,Taguchi’s method has found great successes indesigning and optimizing the engineering systems,so that the literature indicates a number of studiesin this regard.19–23 This method has also been usedsuccessfully to optimize the fillers composition ofthe friction materials.24 Generally, the Taguchimethod can be utilized to obtain various informa-tion about the system under investigation such asaverage factor effect, relative influence of factors(contribution of factor to a result), possible interac-tion between factors (influence of levels of one fac-tor on the effect of another) and optimum condition.These can be achieved by laying out the experi-ments based on appropriate orthogonal array andperforming appropriate statistical calculations to theresults.25

This article deals with the mixtures of two differ-ent synthetic rubbers consisting of SBR and NBRwith novolac type PH, which can be used as organicbinder of the composite friction materials particu-larly for railroad brakes. Special attention is given toacquire further understanding concerning the influ-ence of the constituents of the blend including therubber vulcanizing ingredients and the resin content.To reduce the size of experiments, Taguchi’s experi-mental design method is utilized to investigate thecompositional parameters of the blends. The mor-phological characteristics of the blend are also exam-ined by using dynamic and static mechanical analy-sis and scanning electron microscopy (SEM).

EXPERIMENTAL PROCEDURE AND ANALYSIS

Materials

The polymer materials used in this study includeSBR (SBR1502, styrene content of 23%, density of0.97 g/cm3; BIPC, Iran), NBR (Europrene N 33.45,acrylonitril content of 33%, density of 0.98 g/cm3,Mooney viscosity 45, Enichem, Italy) and PH(Novolac IP502, hexamehylenetetramine content 10wt %, density of 1.28 g/cm3; Rezitan, Iran). The cur-ing ingredients of both SBR and NBR consist of sul-fur, dibenzothiazyl disulfide (MBTS), zinc oxide(ZnO), and stearic acid which are of commercialgrades supplied from local companies. Industrialgrades of toluene and acetone are used as solventsin swelling experiments.

Design of experiments

Taguchi method

In this study, experimental design method based onthe Taguchi approach is utilized to investigate indi-vidually the influence of each component, known asfactor, on the properties of the rubber/PH mixtures.The Taguchi technique helps us to investigate all ofthese factors simultaneously by laying out smallnumber of experiments according to an appropriateorthogonal array, depending on the number offactors and their levels.

As will be discussed later in this study, mixing ofboth NBR and SBR with PH leads to a two-phaseblend, containing cured PH phase dispersed withinthe vulcanized rubber matrix. Properties of the rub-ber/PH blends are expected to be a function ofproperties of individual phases, i.e., rubber and resinphases. On the other hand, physical and mechanicalproperties of rubber phase are dominated by micro-structure of the crosslinks network, namely cross-linking density and polysulfidity of the sulfur link-ages produced between rubber chains. In addition tosulfur content, other curing ingredients such as ac-celerator and activator can also play a noticeable roleon the crosslinks network of the rubber phase.Accordingly, five factors including concentrations ofPH, sulfur, MBTS, ZnO, and stearic acid in terms ofphr (part per hundred parts by weight of rubber)are selected in this study. Since all of these factorscan be changed independently, they are known ascontrol factor not noise factor. To take into consider-ation any possible nonlinear influence of the factorsupon the results, each factor is studied at four levelswithin a defined range including sulfur in 2–10 phr,MBTS in 1–5 phr, ZnO in 2–10 phr, stearic acid in 1–5 phr, and PH in 0–40 phr. For 5 four-level factorswithout considering the outer array for noise factor,modified L16 orthogonal array, L16 (45) or M16, issuggested by the Taguchi methodology.25 Accordingto M16 array, 16 compounds are required for eachrubber/PH blend, as detailed in Table I, to examinethe total compositional parameters. As indicated,each composition is designated by a number rangingfrom 1 to 16. For instance, SBR-8 is relevant to ablend containing SBR whose composition is given inTable I with code 8.

Analysis of the results

All the 16 compounds for each rubber/PH mixtureare subjected to different physical and mechanicaltests, and the results are used to extract the role ofeach ingredient on the properties of the blend. Theprimary goal of using Taguchi method in this studyis to determine (1) the trend of influence of thefactors and (2) contribution of each factor on the

EFFECTS OF RUBBER CURING INGREDIENTS AND PHENOLIC-RESIN 3809

Journal of Applied Polymer Science DOI 10.1002/app

properties of the blends by performing a few num-ber of experiments. The trend of influence of a factorcan be determined by calculating the average effectof a factor at a level. This latter one is obtained bysimple statistical calculation as follows25:

Ai ¼Pn

j¼1 yji

n(1)

where Ai is the average effect of factor A at level i, yji

represents the j-th observation (result) of factor A atlevel i, and n stands for total number of observationsfor factor A. By plotting the average factor effectagainst the corresponding factor level, an averageplot is obtained from which the trend of influence ofa factor on the result is extracted. The relative influ-ence of factors (contribution) on the result can alsobe obtained by somewhat more rigorous statisticalcalculations known as analysis of variance(ANOVA), see the details in Ref. 25. In this study,calculations of both ANOVA and average factor-level effect are performed by using QT4 software.25

Preparation of compounds

The compounds are prepared by a laboratory scaletwo-roll mill. The rubbers are first masticated at tem-perature around 508C for 5 min. Then the masticatedrubbers are mixed further with PH and other ingre-dients for around additional 25 min. The obtainedcompounds are cured into a sheet using compressionmolding at 1508C for 1 h under a pressure of 3.5MPa in accordance to ASTM D-3182. The samples

required for mechanical and physical tests are cutfrom the cured sheets.

Thermal analysis

Differential scanning calorimeter (DSC, Pyris 1, Per-kin–Elmer) is used to investigate the thermal eventsoccurred during the heating of the cured com-pounds. To do this, DSC analysis is performed at aheating rate of 208C/min under the nitrogen withflow rate of 200 mL/min. Figure 1 shows the DSCthermogram of typical compounds. As seen, there isan endothermic peak in the graphs. This peak corre-sponds to the maximum rate of mass loss of thecompounds, and in this study, it is taken as theirthermal stability.

Mechanical testing

The stress–strain (r–e) properties of the cured dumb-bell-shaped specimens are determined at room tem-perature according to ASTM D412 by using an Ins-tron tensile testing machine, operated at a crossheadspeed of 60 mm/min. The tensile strength and elon-gation at break are extracted directly from the r-ecurves, while the young’s modulus is calculatedfrom the initial slope of the curve, namely slope ofthe curve within strain range of 0–2%. Dynamic me-chanical properties of the mixtures are determinedby a dynamic mechanical thermal analyzer (DMTA-Triton, Tritic Technology, England) at a frequency of1 Hz and temperature range from 2150 to 1808Cwith a heating rate of 38C/min. The DMTA recordsthe variation of storage modulus (E0) and loss tan-gent (tan d) with temperature.

TABLE IExperimental Layout of M16 Orthogonal Array

According to Taguchi’s Method

Compound no. Sulfur MBTS ZnOStearicacid PH

1 2 1 2 1 02 2 2.25 4.5 2.25 12.53 2 3.5 7 3.5 254 2 4.75 9.5 4.75 37.55 4.5 1 7 2.25 37.56 4.5 2.25 9.5 1 257 4.5 3.5 2 4.75 12.58 4.5 4.75 4.5 3.5 09 7 1 9.5 3.5 12.510 7 2.25 7 4.75 011 7 3.5 4.5 1 37.512 7 4.75 2 2.25 2513 9.5 1 4.5 4.75 2514 9.5 2.25 2 3.5 37.515 9.5 3.5 9.5 2.25 016 9.5 4.75 7 1 12.5

The levels of ingredients are given in terms of phr.

Figure 1 DSC thermograms of typical compounds. [Colorfigure can be viewed in the online issue, which is availableat www.interscience.wiley.com.]

3810 DERAKHSHANDEH, SHOJAEI, AND FAGHIHI


Morphology

The morphology of the blends is characterized bymeans of a Philips XL30 SEM. To do this, the sam-ples are fractured in liquid nitrogen, and then thefractured cross sections and surface of sampleswhich are sputter-coated with gold layer are used toinvestigate the phase morphology.

Swelling

The vulcanized samples are first subjected to extrac-tion process using acetone for around 120 h to removethe residual soluble components after vulcanization.The extracted samples are vacuum-dried under 608Cfor 2 h and then the dried specimens are weighed. Af-ter drying the samples, they are dipped into tolueneat room temperature for 120 h. The swollen samplesare removed from the solvents and wiped off usingtissue and weighed. The swelling coefficient, Q, isobtained using the following relation26:

Q ¼ mS �mD

mD3

1

ql(2)

where mS and mD are the weights of swollen and thedry extracted sample, respectively, and ql is the den-sity of the swelling solvent.

Crosslinking density of the rubber phase can becalculated by the Flory-Rehner relation expressed asfollows26:

MC ¼qrVS

mr2 � m1=3r

� �

½lnð1� mrÞ þ mr þ vm2r �(3)

where mr denotes the equilibrium volume fraction ofthe rubber in swollen vulcanized rubber which is

estimated here based on the swelling measurements,qr the density of the rubber, VS the molar volume ofsolvent, and, v the polymer solvent interaction pa-rameter. In eq. (3), MC is the average molecularweight between two crosslinks per primary rubberchains which is an indication of the crosslinkingdensity. The lower the MC value, the higher cross-linking density is achieved. The interaction parame-ter reported for SBR-toluene system is 0.39,26 whilefor NBR-toluene it is calculated using the followingexpression27:

v ¼ 0:35þ VS

RTðdS � drÞ2 (4)

where dr and dS are the solubility parameters of therubber and solvent, respectively. Taking dr 5 9.4(cal/cm3)0.5 for NBR and dS 5 8.9 (cal/cm3)0.5 fortoluene,27 interaction parameter for NBR-toluene sys-tems becomes 0.46. The measured average values ofdensity for SBR and NBR vulcanizates are � 1 g/cm3. The density and molar volume of toluene at208C are 0.867 g/cm3 and 106.3 cm3/mol, respec-tively.28

RESULTS AND DISCUSSION

Compositional effects on the properties of therubber/PH blends

Table II summarizes the results of experimentalmeasurements performed for 16 compounds of eachrubber/PH blends indicated in Table I. These resultsare used to estimate the influence of each componenton the swelling, thermal and mechanical propertiesof the blends based on Taguchi’s analysis. It shouldbe pointed out that all of the plots obtained in this

TABLE IIExperimental Results for SBR/PH and NBR/PH Blends

Compoundno.

Elongation (%) Modulus (MPa) Swelling coefficient (mL/g)

NBR SBR NBR SBR NBR SBR

1 653 6 35 766 6 29 1.9 6 0.2 1.8 6 1 2.3 6 0.5 4.4 6 0.12 346 6 14 215 6 62 2.5 6 0.4 3.4 6 1.1 2.3 6 0.1 3.8 6 0.13 256 6 42 139 6 27 4.4 6 0.5 5.9 6 1.1 2 6 0.1 3.8 6 0.14 257 6 23 224 6 83 5.8 6 1.7 6.1 6 0.9 1.8 6 0.15 151 6 18 145 6 29 5.8 6 1.1 6.1 6 1.4 1.6 6 0.1 3.5 6 0.056 194 6 15 97 6 14 5.8 6 1.1 5.3 6 1.4 1.3 6 0.1 3 6 0.17 252 6 13 143 6 17 2.5 6 0.6 3.3 6 0.4 1.5 6 0.2 3.4 6 0.058 262 6 20 223 6 30 1.4 6 0.7 2.2 6 0.2 1.6 6 0.2 3.2 6 0.039 158 6 14 142 6 15 3.4 6 0.4 4.9 6 0.6 1.3 6 0.1 2.5 6 0.07

10 201 6 25 152 6 30 2.4 6 0.1 2.3 6.05 1.5 6 0.2 2.4 6 0.0511 93 6 14 102 6 11 11 6 1.7 8.4 6 1.4 1.5 6 0.9 2.9 6 0.0712 141 6 16 122 6 19 7.4 6 0.5 5.5 6 1.1 1.6 6 0.2 2.6 6 0.0913 124 6 18 104 6 18 7.7 6 1.9 6 6 1.3 1.4 6 0.3 2.3 6 0.0814 124 6 19 105 6 3 11.6 6 2.2 9.3 6 1.2 1.3 6 0.1 2 6 0.0615 183 6 14 120 6 21 3.9 6 0.7 3.3 6 0.5 1.5 6 0.2 1.9 6 0.0716 158 6 19 113 6 19 4.7 6 0.9 3.8 6 1 1.3 6 0.2 1.7 6 0.09



study based on statistical calculations represent onlythe trend of influence of each factor upon the result,not the exact value of the result at a given level of afactor.

Swelling characteristics

Knowledge of the swelling behavior of crosslinkedpolymer can provide further insight into the cross-links microstructure, particularly for two-phase sys-tems. Figure 2 shows the variations of the swellingcoefficient against concentration of rubber curingingredients and PH for both SBR/PH and NBR/PHblends. From Figure 2(a), it is observed that sulfur isthe most important parameter influencing the swel-ling coefficient of the SBR/PH blends. As the swel-ling coefficient of the rubber phase is dominated bythe crosslinking density,27–29 this result suggests thatincreasing the sulfur content within the range of 2–10 phr increases strongly the crosslinking density ofthe SBR. Comparing Figure 2(a,b), it is found thatthe extent of swelling and variation of swelling coef-ficient of NBR/PH are less than those of SBR/PH,but the sulfur has still higher influence with respectto the other curing ingredients. This observation sug-gests that the crosslinking density of the NBR doesnot vary significantly as much as the SBR.

Figure 2(c) shows that the swelling coefficientdecreases by increasing the resin content. On theother hand, the maximum extent of reduction inswelling coefficient of both rubbers within the resincontent of 0–40 phr is approximately the same, i.e.,extent of reduction is about 0.5 mL/g for both rub-bers. This behavior can be explained by the fact thatthe rubber component is responsible for swelling ofthe rubber/PH blends. Because, PH is a highlycrosslinked thermoset and can not be swelled any-more. This explanation is confirmed by performingseparately swelling experiment for cured PH so as tono swelling effect appears. In fact, the swelling ofPH is negligible and it acts as rigid filler. So incorpo-ration of PH decreases the content of rubber compo-nent in the blend, leading to reduction of the swel-ling coefficient of the blends, accordingly.

Table III presents the contribution of the compo-nents of the blend (relative importance of the factors)on the swelling coefficient. It is found that sulfur isthe only dominant parameters on the swelling coeffi-cient of the SBR/PH blends. However, for the NBR/PH, sulfur and PH have almost the similar impor-tance. This could be attributed to lower extent ofswelling of the NBR vulcanizate due to its highercrosslink density, as discussed earlier. On the otherhand, contribution of other curing ingredients,namely MBTS, ZnO, and strearic acid in SBR/PHblend is negligible, while these factors slightly affectthe NBR/PH blend. Slight influence of accelerator

Figure 2 Average effect of components on the swellingcoefficient of blends obtained by Taguchi’s analysis; (a)effect of curing ingredients on SBR/PH, (b) effect of curingingredients on NBR/PH blends, and (c) effect of PH onSBR and NBR blends. [Color figure can be viewed in theonline issue, which is available at www.interscience.wiley.com.]



and activator is thought to be due to their effects onvulcanization process during development of thecrosslinks as reported in the literature.17,29,30

As mentioned earlier, swelling of the rubber/PHblends is relevant to the rubber phase. Therefore,crosslinking density of the rubber phase in the blendcan be calculated by the Flory-Rehner equation, eq.(3), based on the swelling data. Figure 3 illustratesthe variation of MC with the concentration of rubbercuring ingredients. It is found that MC for SBR isgreater than that of the NBR, indicating its lowercrosslinking density as compared with NBR; and itreduces mainly by increasing the sulfur content. ForSBR, MC reduces almost from 8000 to 2000 g/molwhen the sulfur content varies within 2–10 phr. Vari-ation of MC for NBR within the same range of sulfurcontent is between 1500 and 3000 g/mol. Thesevalues of MC for both rubbers stand within theexpected range as described in the literature.26 Thehigher value of MC for SBR compared to NBR couldbe attributed to its chemical structure. In SBR, pres-ence of such large side group as benzene causes amolecular spatial hindrance to produce more sulfurlinkages between two polymer chains, resulting inhigher value of MC. According to this explanation, itshould be expected to form more polysulfidic cross-links in SBR, i.e., longer sulfur linkages, betweentwo polymeric chains to overcome the hindranceeffect. This characteristic will be examined later inthis work using the thermal behavior of the blends.Figure 3 also shows that the value of MC approxi-mately levels off after 4 phr sulfur content for NBR.A possible explanation for this behavior could bethat the molecular weight between two crosslinksafter 4 phr sulfur is small enough, i.e., around1500 g/mol, and two neighbor crosslinks are soclose, so as to beyond that new crosslinks could notbe formed easily.

Figure 3(c) shows that crosslinking density of therubber phase increases slightly by increasing theresin content. This observation may be explained bydissolution of small amount of the resin in the rub-ber phase. This partial solubility of the resin will be

examined further by the dynamic and static mechan-ical experiments carried out in this study. It seemsthat the dissolved PH participates in the crosslinksnetwork developed during vulcanization of the rub-bers in addition to the sulfur linkages.17,30 Thiscrosslink could be made by reaction of resin andrubber chains through methylene bridging, particu-larly in presence of hexamethylenetetramine.31

Thermal properties

Thermal stability of the blends obtained by DSCthermograms is employed in QT4 software for per-forming the statistical analysis, and the results arepresented in Figure 4 and Table III. Overall degrada-tion temperature of the blends, shown in Figure 4,stands within 370–3858C for different compounds.This temperature range is similar to the degradationtemperature of the neat rubber vulcanizate. Theseresults suggest that the degradation temperature ofthe rubber/resin blend is dominated by the rubbercomponent. Camino et al.16 reported that the weightloss of NBR, degradation of NBR obtained by ther-mogravimetric analysis, begins at about 3508C,which is in agreement with the results obtained inthis study. They16 also came into the same conclu-sion about the mixture of NBR and novolac PH, thatthe degradation temperature is dominated by NBR.

The results illustrated in Figure 4-c indicate thatthe resin content has a negligible effect on the degra-dation temperature of the rubber/resin blends. It isshown that the rubber curing ingredients has aminor effect on the degradation temperature of theNBR blends. However, for SBR blends, the sulfurcontent is found to be the most effective parametersinfluencing the degradation temperature. As shownin Figure 4(a), degradation temperature decreasesalmost from 385 to 3728C by increasing the sulfurcontent from 2 to 10 phr. This is probably attribut-able to the structure of the sulfur linkage producedin the SBR network which could be mainly polysulfi-dic. This explanation is in agreement with the swel-ling characteristics of the SBR compounds in which

TABLE IIIContribution (in percent) of Ingredients to the Physical and Mechanical Properties of

Rubber/PH Blends Obtained by QT4 software

Factor

Contribution tothe swellingcoefficient (%)

Contribution tothe modulus (%)

Contribution tothe thermalstability (%)

NBR SBR NBR SBR NBR SBR

PH 35 9.6 68 84 18 3.7Sulfur 47 84 23 8 22 86MBTS 5 2.6 1.7 2.7 4 4ZnO 8 3 4.8 0.8 30 2.8Stearic acid 5 0.8 2.5 4.5 26 3.5



Figure 3 Average effect of components on the Mc; (a)effect of curing ingredients on SBR/PH, (b) effect of curingingredients on NBR/PH blends, and (c) effect of PH onSBR and NBR blends. [Color figure can be viewed in theonline issue, which is available at www.interscience.wiley.com.]

Figure 4 Average effect of components on the thermalstability of blends obtained by Taguchi’s analysis; (a) effectof curing agents on SBR/PH, (b) effect of curing agents onNBR/PH blends, and (c) effect of PH on SBR and NBRblends. [Color figure can be viewed in the online issue,which is available at www.interscience.wiley.com.]



the polysulfidic linkages are thought to be dominantdue to hindrance effect of benzene side group. Thehigher the sulfur content, the more polysulfidic link-ages are produced. It is well known that polysulfidicbonds have low thermal stability, leading to the ther-mal decomposition of the vulcanized rubber at lowertemperatures.29,30

Mechanical properties

Figures 5–7 depict the variation of mechanical prop-erties of the blends with concentration of the rubbercuring ingredients and the resin content. The resultsshow that all of these components influence the me-chanical properties. However, from Table III forYoung’s modulus, it can be found that the contribu-tion of PH is much more pronounced with respect toother components, namely rubber curing ingredients.In addition, from Figure 7, it is shown that by

increasing the resin content, the modulus increasesand elongation at break decreases continuously.Since the PH has very high modulus and low elon-gation at break when compared with both SBR andNBR, see Table IV, these effects of PH are com-pletely expected. We will show in the next sectionthat addition of PH into the rubber material leads toa two-phase system in which the PH componentforms a dispersed phase. This resembles a compositematerial filled with rigid filler, resulting in highermodulus and lower elongation. Moreover, presenceof PH in the rubber phase due to the partial solubil-ity of the resin, as pointed out in the swelling char-acteristics, is thought to increase the stiffness of therubber phase due to the rigidity of the PH chainsand its contribution to the crosslinking density of therubber phase.

Among the curing ingredients, the sulfur contentis expectedly shown to be the significant parameter

Figure 5 Average effect of rubber curing ingredients on the mechanical properties of SBR/PH blends predicted based onTaguchi’s analysis; (a) elongation at break and (b) Young’s modulus. [Color figure can be viewed in the online issue,which is available at www.interscience.wiley.com.]

Figure 6 Average effect of rubber curing ingredients on the mechanical properties of NBR/PH blends predicted basedon Taguchi’s analysis; (a) elongation at break and (b) Young’s modulus. [Color figure can be viewed in the online issue,which is available at www.interscience.wiley.com.]



influencing the mechanical properties. This behaviorcan be associated to the alteration of the networkstructure of the vulcanized rubber. As mentioned inthe preceding section, higher crosslinking density forboth rubbers is achieved by increasing the sulfurcontent. It is well known that the higher crosslinkingdensity restricts the mobility of the chain segmentsfurther, leading to improved mechanical behavior ofthe vulcanizates.26 As observed in the swellingbehavior of both rubbers, enhancement of crosslink-ing density of SBR is much more pronounced whencompared with NBR (see Fig. 3), while the variationof modulus for NBR is slightly greater than that ofSBR. In agreement with the swelling data and ther-mal stability of the SBR/PH blend, longer sulfurlinkage of SBR vulcanizates could be possibly re-sponsible for this behavior. Because, longer cross-links increase the mobility of chain segments of themacromolecules and moderates the role of higherenhancement of crosslinking density.

From Table III, it is shown that MBTS, ZnO, andstearic acid also have minor influence on Young’smodulus of the blend. As mentioned in the swellingdata, this behavior is due to their slight effect uponthe structure of the crosslinks.

Morphological investigation

Dynamic mechanical analysis

The dynamic mechanical analysis is performed forthe blends of both rubbers with varying amount ofPH, ranging from 0 to 23 vol %. In all of theseblends, the concentration of rubber curing ingre-dients is kept constant, i.e., sulfur 2 phr, MBTS 2phr, stearic acid 2 phr, and zinc oxide 5 phr. Figure8 illustrates the storage modulus (E0) of the blendsversus temperature in the range of 2150 to 1508C.

As expected, the storage modulus of the blendsabove the transition temperature, where a sharpdecrease in E0 is observed, increases by increasingthe PH concentration. This is attributed to the higherstiffness of the PH chains.

Plots of tan d versus temperature for the blendsare given in Figure 9. It is observed that there is asharp peak at 218.68C for pure NBR vulcanizate(NBR-1) and at 238.98C for pure SBR vulcanizate(SBR-1). Indeed, these temperatures indicate theglass transition temperature (Tg) of the rubbers,which are consistent with the reported values in theliterature.14,15,32 Additionally, from Figure 9, the Tg

of pure PH is found to be 1388C. It is shown that themaximum value of tan d (tan dmax) for pure rubbervulcanizate decreases from 1.2 to 0.45 for SBR/PHand from 1.19 to 0.54 for NBR/PH blends containing23% by volume of PH. It is well known that thehigher mechanical losses which are related to highenergy input required for the segmental motion ofthe macromolecular chains, lead to greater value oftan dmax. For the rubber/PH blends studied here,presence of hard PH segments may possibly restrictthe movement of the soft rubber segments, resultingin lower value of tan dmax.

The plots of tan d, shown in Figure 9, demonstratethat addition of PH to the rubbers shifts slightly the

Figure 7 Average effect of resin content on the mechanical properties of the blends predicted based on Taguchi’s analy-sis; (a) elongation at break and (b) Young’s modulus. [Color figure can be viewed in the online issue, which is available atwww.interscience.wiley.com.]

TABLE IVYoung’s Modulus of Phenolic Resin and Rubbers at

Two Levels of Sulfur Content

Material Modulus (MPa) Elongation at break (%)

PH 5,100 6 250 1SBR (2 phr sulfur) 1.82 6 0.15 320.6 6 80SBR (7 phr sulfur) 2.75 6 0.2 122.75 6 55NBR (2 phr sulfur) 1.87 6 0.1 347.5 6 32NBR (7 phr sulfur) 2.9 6 0.2 141.75 6 40



tan dmax of both rubbers, namely less than 58C. Onthe other hand, no separate peak due to the PH com-ponent is observed in the blends. Such an observa-tion has been reported by Samui et al.32 for NBR/PH blends as well. Inward shift of Tg can be an indi-cation of solubility of the PH in the rubber phase.26

The extent of solubility can be estimated by the rule

of mixture, i.e., Tg ¼P2

i¼1 Tgi/i.26 According to this

rule and Tg values of the blends, shown in Figure 9,volume fraction of dissolved PH in the rubber phaseis found to be around 0.03 for a mixture containing23 vol % PH. This shows that only small amount ofPH is dissolved in the rubber phase, and mixing ofthe rubbers with PH leads essentially to an incom-patible blend and a two-phase system. The incom-patibility is expected, because PH is highly polarmaterial due to ��OH group, while NBR has a fewand SBR has no polar group. The slight dissolutionof PH in the rubber phase observed by DMTA sup-ports the enhancement of crosslinking density of therubber vulcanizate due to the presence of PH in rub-ber phase as discussed in swelling characteristics of

the blends. Higher shift in Tg of SBR when com-pared with NBR may be associated to slightly higherdissolution of PH in SBR

Halpin-Tsai relationship for rubber/PH blends

It is well known that the mechanical properties of apolymer blend are governed by microstructure ofthe dispersed phase and properties of individualphases.33,34 Hence, as a further evidence to the micro-structural behavior of the rubber/PH blends, modu-lus of these blends is investigated by varying theamount of PH in the blend ranging between 0 and 25vol %, while concentration of the rubber curing ingre-dients is kept constant, i.e., sulfur 2 phr, MBTS 2 phr,stearic acid 2 phr and zinc oxide 5 phr. To take intoconsideration the role of sulfur content on the mor-phological behavior, variation of blend modulus withthe PH concentration is also investigated at higher sul-fur content, i.e., 7 phr, while the concentration of othercuring ingredients is the same as earlier.

Figure 9 Plots of tan d versus temperature; (a) SBR/PH and (b) NBR/PH blends.

Figure 8 Plots of storage modulus versus temperature; (a) SBR/PH and (b) NBR/PH blends.



Table IV presents the modulus and elongation atbreak of PH and pure rubber vulcanizates at twolevels of sulfur content obtained from static tensiletests. It is expectedly realized that increasing the sul-fur content increases the modulus of both SBR andNBR vulcanizates approximately up to 60%. Further-more, it is found that the modulus ratio of the resinto the rubber is so great, i.e., greater than 2000, so asto the dispersed resin can be regarded as rigid filler.

Halpin-Tsai equation is one of the most versatileand widely used semiempirical equations for thepolymeric composites and blends. In this study, thismodel is used to correlate the mechanical behaviorof the rubber/PH blends. Halpin-Tsai equation inwhich the rubber forms the continuous phase isexpressed as follows33:

Eb

Er¼ 1þ AB/p

1� Bw/p

(5)

where subscripts b, r, and p stand for blend, rubber,and PH, respectively, E represents the Young’s mod-ulus, / is the volume fraction, and A, B, and w arethe model parameters which may be given as fol-lows33:

A ¼ kE � 1

B ¼ ðEp=ErÞ�1

ðEp=ErÞþA � 1

w ¼ 1þ 1�/m

/2m

/p

8><>:

(6)

in which /m is the maximum packing fraction of thedispersed phase which can be considered to beabout 0.63,33 and kE is the generalized Einstein coef-ficient. To calculate the modulus of the blend by eq.(5), Einstein’s coefficient kE should be known. Theo-retical value of kE is 2.5 which is given by the origi-nal Einstein equation expressed as follows33:

Eb

Er¼ 1þ kE/p: (7)

This equation is valid for composite systems havingonly spherical rigid particles in very low concentra-tions. In actual situations, morphology of the dis-persed phase may deviate from the ideal case.Therefore, the theoretical value of kE may be inap-propriate in such cases. In this study, the value of kEis determined from experimental data based on theEinstein equation from the slope of Eb=Er versus /p.As Einstein’s equation is valid for a very low con-centration, kE is extracted from a set of blends withlow concentrations of PH, namely up to 3 vol %. kEvalues obtained in this way are 6.16 and 8.4 forSBR/PH with 2 and 7 phr sulfur, respectively, and6.24 and 8.36 for NBR/PH with 2 and 7 phr sulfur,

respectively. In all cases, regression coefficient is soclose to one, namely greater than 0.99, indicating thelinearity of the curve in this range of PH concentra-tion. It is observed that the empirical values of kEare higher than that of the theoretical one. This prob-ably could be due to agglomeration of the resin par-ticles within the rubber matrix. It has also beenreported that the agglomeration of particles as wellas state of agglomeration increase the suspensionviscosity and Einstein’s coefficient in comparisonwith the completely dispersed systems.33 For a sus-pension viscosity including agglomeration of particlewith random packing, a value of 6.76 has beenreported for kE.

35 It is also found that the values ofkE are the same for both SBR and NBR compoundsat the same sulfur content; however, it increases byincreasing the sulfur content from around 6.3 to 8.3.It seems that the sulfur content increases the affinityof the particles to form more agglomeration, whichcould be a consequence of decreasing the surfaceenergy of the rubber phase. Such an observation hasbeen addressed in the literature for blend of ethyl-ene–propylene–diene rubber/propylene that a reduc-tion in surface energy has been observed at highextent of crosslinking density.36 This has been attrib-uted to the reduction of the flexibility of bothinduced and permanent dipoles attached to the rub-ber segments within the vulcanized network.36

Modulus of the rubber/PH blends

Figure 10 shows the modulus of the blends as afunction of volume fraction of PH obtained by theexperimental measurements and estimated valuesusing Halpin-Tsai equation. It is observed that thevariation of the modulus with resin content for bothrubbers can be well described by Halpin-Tsai equa-tion at low resin content. After a certain value ofresin content, � 15 vol %, moduli indicate a devia-tion from eq. (5) and the results show a sharpincrease, particularly at higher resin content. Thisobservation suggests much more contribution of thePH upon the modulus of the blends at higher vol-ume fraction. There could be some possibilities forthis behavior. One possibility could be the alterationof the phase morphology of the blend and convert-ing to cocontinuous morphology. Because, in suchcases, the PH can play considerable role on themodulus of the blends.34,35 However, this explana-tion is somewhat questionable, because the volumefraction of 0.15 seems to be too low for the rubber/PH blend to become cocontinuous. Only partialcocontinuity due to the tendency of the PH-dis-persed phase for agglomeration may be a reasonableexplanation for the mechanical behavior of theblend mentioned earlier.



Another possibility for deviation from eq. (5)could be the change of the mechanical properties ofthe rubber phase as a consequence of the addition ofthe resin. This explanation is most probable than themorphology inversion mentioned earlier and is com-pletely consistent with swelling characteristic anddynamic mechanical behavior of the blends. As men-tioned, addition of the resin to the rubber phaseincreases the crosslinking density of the rubber vul-canizates, resulting in higher modulus of the rubberphase than that of pure rubber vulcanizates. On theother hand, presence of PH in the rubber phase dueto its partial solubility, as examined by dynamic me-chanical tests, increases the rigidity of the molecularchains, leading to higher stiffness of the rubberphase in the blends. These evidences indicate thatthe modulus of the rubber phase in the blend is dif-ferent than that of the pure rubber vulcanizates usedin the Halpin-Tsai equation. Consequently, the mis-use of the modulus of the rubber in Halpin-Tsai

equation could be responsible for the severe devia-tion from the model prediction.

From Figure 10, it is shown that effect of PH onenhancement of modulus of SBR/PH blend at highvolume fraction of PH is much more pronouncedthan that of NBR/PH blend. This can be attributedto slightly higher solubility of the PH in SBR, asillustrated in dynamic mechanical tests, resulting inhigher enhancement of modulus of SBR. Althoughthe difference between solubility of PH in SBR andNBR is too small, but even very small amount of PHcan have observable influence on the modulus of therubber phase due to very high modulus of PH andits curative effect.

SEM observations

The blends of both rubbers with PH at different vol-ume fractions are depicted in Figures 11 and 12.The SEM micrographs show dispersion of the PH

Figure 11 SEM micrographs of NBR/PH blends containing different volume percents of PH; (a) 8% and (b) 23%.

Figure 10 Modulus of blends as a function of resin content; (a) SBR/PH (b) NBR/PH.



particles within the rubber matrix. The dispersedphase is almost spherical and indicates a good ad-hesion with the matrix, because no interfacialdebonding is observed in the SEM micrographs ofthe fractured surfaces. However, it can be shownthat the PH particles tend to form agglomerationeven at low volume fraction of PH. This observationis in complete agreement with the higher value ofkE mentioned earlier in the mechanical properties ofthe blends. It is also seen that the matrix-dispersedtype of microstructure is dominant, although a com-pact dispersed morphology is observed at highvolume fraction of PH.

CONCLUSIONS

Thermal, mechanical, and microstructural character-istics of NBR/PH and SBR/PH blends are investi-gated in this study. It is shown that the glass-transi-tion of both rubbers is shifted slightly at variouscompositions of rubber/PH mixtures, showing par-tial solubility of the PH in the rubber phase. How-ever, these mixtures are incompatible blends and, asevidenced by SEM micrographs, exhibit two-phasemicrostructures in which rigid spherical particles ofPH are embedded in a soft rubber matrix. Theresults obtained based on Taguchi’s analysis showthat sulfur has a dominant effect on the crosslinksstructure of the vulcananized rubber, although othercuring ingredients including accelerator and activa-tor influence the crosslinks network more or less.Moreover, it is shown that crosslinking density ofthe rubber phase in the blend is also affected by PH.Soluble part of the resin phase, as supported bydynamic mechanical analysis, is shown to be respon-sible for alteration of crosslinking density of the rub-ber phase due to curative role of PH. On the otherhand, presence of the rigid PH chains in the cross-

link network of the rubber restricts the molecularmotion and makes the rubber phase to be stifferthan that of pure rubber vulcanizates. Deviation ofthe blend modulus versus PH content from Halpin-Tsai model is known as a further evidence for therole of PH on the rubber phase.

Crosslinking density of the SBR is found to belower than that of NBR at a certain curing agent.This is attributed to the hindrance effect of benzeneside group of SBR which prevents more crosslinks tobe formed in the polymeric chains. Additionally,thermal stability of SBR blends decreases noticeablyby increasing the sulfur content when comparedwith that of NBR/PH. This is probably due to theformation of more polysulfidic crosslinks in SBRvulcanizates.

The authors are grateful to Machine Lent Tehran Co., dueto the donation of all materials used in this study.

References

1. Jacko, M. G.; Rhee, S. K. In Encyclopedia of Composite Materi-als and Components; Grayson, M., Ed.; Wiley, 1983, 144.

2. Bijwe, J. Polym Compos 1997, 18, 378.

3. Chan, D.; Stachowiak, G. W. Proc Inst Mech Eng Part D: JAutomobile Eng 2004, 218, 953.

4. Jacko, M. G.; Tsang, P. H. S.; Rhee, S. K. Wear 1984, 100, 503.

5. Yensik, M. A. U.S. Pat. 5,529,666 (1996).

6. Knop, A.; Pilato, L. A. Phenolic Resins, Chemistry, Applica-tions and Performance; Springer-Verlag: Berlin, 1985.

7. Tsang, P. H.; Coyle, J. P.; Liu, T.; VanderPoorte, J. G. U.S. Pat.4,537,823 (1985).

8. Kane, J. F.; Mowrer, N. R. U.S. Pat. 5,736,619 (1998).

9. Adelmann, J. C. U.S. Pat. 3,959,194 (1976).

10. Littlefield, J. B. U.S. Pat. 4,313,869 (1982).

11. Eschen, F. W. U.S. Pat. 3,832,325 (1974).

12. Gibson, P. A. In Seminar Publication on Railway Traction andBraking; Institute of Mechanical Engineers, 1996; p 75.

13. Haddadi, E.; Abbasi, F.; Shojaei, A. J Appl Polym Sci 2005, 95,1181.

Figure 12 SEM micrographs of SBR/PH blends containing different volume percents of PH; (a) 8% and (b) 21%.



14. Borowitz, J.; Kosfeld, R. Rheol Acta 1981, 20, 370 (in German).15. Kosfeld, R.; Borowitz, J. Rheol Acta 1979, 18, 51–61 (in

German).16. Camino, G.; Alba, E.; Buonfico, P.; Vikoulov, V. J App Polym

Sci 2001, 82, 1346.17. Coran, A. Y. J Appl Polym Sci 2003, 87, 24.18. Montgomery, D. C. Design and Analysis of Experiments;

Wiley: New York, 1997.19. Liu, S. J.; Chen, C. F. Plast Rubber Compos 2000, 29, 411.20. Abbasi, F.; Mirzadeh, H.; Katbab, A. A. J Appl Polym Sci 2002,

85, 1825.21. Kim, S. J; Kim, K. S.; Jang, H. J Mater Process Tech 2003, 136,

202.22. Chen, D.-C.; Chen, C.-F. J Mater Process Tech 2006, 177, 104.23. Yang, C.; Hung, S. W. Int J Adv Manuf Tech 2004, 24, 353.24. Kim, Y. H.; Lee, J. J.; Nisitani, H. Int J Mod Phys B 2003, 17,

1845.25. Roy, R. K. Design of Experiments Using the Taguchi

Approach; Wiley: New York, 2001.26. Sperling, L. H. Introduction to Physical Polymer Science, 4th

ed; Wiley-Interscience: New York, 2006.

27. Rodriguez, F. Principles of Polymer Systems, 2nd ed;McGraw-Hill International Book: London, 1985.

28. Brandrup, J.; Immergut, E. H.; Grulke, E. A. Polymer Hand-book; Wiley-Interscience: New York, 1999.

29. Morrell, S. H. In Rubber Technology and Manufacture; Blow,C. M., Ed.; Second Butterworth & Co Ltd: London, 1975; Chap-ter 5.

30. Hofmann, W. Rubber Technology Handbook; Hanser Publish-ers: Germany, 1988.

31. Odian, G. Principles of Polymerization; Wiley: New York,1981.

32. Samui, A. B.; Suryavanshi, U. G.; Patri, M.; Chakraborty, B. C.;Deb, P. C. J Appl Polym Sci 1998, 68, 255.

33. Nielsen, L. E.; Landel, R. F. Mechanical Properties of Polymersand Composites, 2nd ed; Marcel Dekker: New York, 1994.

34. Nielsen, L. E. Predicting the Properties of Mixtures: MixtureRules in Science and Engineering; Marcel Dekker: New York,1978.

35. Lewis, T. B.; Nielsen, L. E. Trans Soc Rheol 1968, 12, 421.36. Goharpey, F.; Katbab, A. A.; Nazockdast, H. J Appl Polym Sci

2001, 81, 2531.



Date post:	16-Nov-2023
Category:	Documents
Upload:	independent
View:	1 times
Download:	0 times

Effects of rubber curing ingredients and phenolic-resin on mechanical, thermal, and morphological...

Documents