Dynamic mechanical analysis of binary and ternary … · 2.4. Measurement of dynamic mechanical...

1. Introduction

Thermoplastic elastomers (TPEs) are being used ina variety of applications because of their desirablemechanical properties and recyclability [1, 2].Dynamic mechanical thermal analysis (DMTA) hasproved to be an effective tool in the characterisationstudies of viscoelastic materials.Several researchers [3–9] have investigated effectsof blend ratio and compatibilisation on the dynamicmechanical properties. Al-Malaika and Kong [10]studied the compatibilisation of functionalised EPRwith GMA. Examination of the DMA of the reac-tive and physically compatibilized blends shows asmaller separation between the glass transition tem-

peratures. Rajan et al. [11] have studied thedynamic mechanical properties of PP in the ther-moplastic elastomeric state. α-relaxation peaks ofPEEK and PES and their blends were studied byNandan et al. [12]. Kumar et al. [13] have studiedthe dynamic mechanical properties of EPDM-g-VOS/LLDPE blends with special reference to theeffect of blend ratio. They found that increasing theproportion of LLDPE decreases the Tg value ofblend and there were increase in E' and E '' due toincrease in crystallinity. Karger-Kocsis and Kiss[14] have investigated morphology and dynamicmechanical properties of EPDM/PP blends and PPblock polymers. In these blends as the concentra-

641

*Corresponding author, e-mail: [email protected]© BME-PT and GTE

Dynamic mechanical analysis of binary and ternary polymerblends based on nylon copolymer/EPDM rubber and EPMgrafted maleic anhydride compatibilizer

C. Komalan1, K. E. George2, P. A. S. Kumar3, K. T. Varughese4, S. Thomas5*

1S.N.M College, Maliankara, Kerala-683516, India2Department of PS&RT, Cochin University of Science and Technology, Cochin-22, Kerala, India3NIT Calicut P.O, Calicut, Kerala, India4Central Power Research Institute, Bangalore-560080, Karnataka, India5School of Chemical Science, M.G University, Kottayam-686560, Kerala, India

Received 30 March 2007; accepted in revised form 17 June 2007

Abstract. The dynamic mechanical properties such as storage modulus, loss modulus and damping properties of blends ofnylon copolymer (PA6,66) with ethylene propylene diene (EPDM) rubber was investigated with special reference to theeffect of blend ratio and compatibilisation over a temperature range –100°C to 150°C at different frequencies. The effect ofchange in the composition of the polymer blends on tanδ was studied to understand the extent of polymer miscibility anddamping characteristics. The loss tangent curve of the blends exhibited two transition peaks, corresponding to the glasstransition temperature (Tg) of individual components indicating incompatibility of the blend systems. The morphology ofthe blends has been examined by using scanning electron microscopy. The Arrhenius relationship was used to calculate theactivation energy for the glass transition of the blends. Finally, attempts have been made to compare the experimental datawith theoretical models.

Keywords: polymer blends and alloys, dynamic mechanical analysis, compatibilisation

eXPRESS Polymer Letters Vol.1, No.10 (2007) 641–653Available online at www.expresspolymlett.comDOI: 10.3144/expresspolymlett.2007.88

tion of EPDM increases, E' of the blends decreases.The dynamic mechanical spectrum showed twoseparate damping peaks and therefore has a two-phase morphology indicating that the blend isincompatible. Guo et al. [15] studied the dynamicmechanical properties of PA6/PS blend systems.Gopakumar et al. [16] found that the compatibilisa-tion does not affect the transition temperature of therespective components while several otherresearchers [17, 18] have reported that the incorpo-ration of the compatibilizer leads to obvious shift-ing of Tg’s. These authors are of the opinion that theeffect of compatibilizers on the position of Tg is areflection of the extent of compatibilisation. Molyet al. [19] have studied the effect of compatibilisa-tion on the dynamic mechanical properties ofLLDPE/EVA blends and found that compatibilisa-tion increased the storage modulus of the systemwhich is due to the fine dispersion of EVA domainsin the LLDPE matrix providing an increased inter-facial interaction.In this communication, effects of blend ratio andcompatibilisation on the dynamic mechanical prop-erties of nylon copolymer/EPDM blends wereinvestigated. Viscoelastic properties like storagemodulus, loss modulus and tanδ were investigatedas a function of temperature.The dynamic mechan-ical properties of the new generation of nyloncopolymer/EPDM blends are extremely important,as these data are still limited in the literature. Thedynamic mechanical properties have been corre-lated with the morphology of the blends. Attemptshave also been made to correlate the experimentalviscoelastic data with various theoretical models.

2. Experimental

2.1. Materials

The raw materials used for this work are as follows.Nylon copolymer (PA6,66) (Tufnyl120) of meltingpoint 148°C and density of 1.12 g/mol, with Mn

38 000 was supplied by SRF Ltd (Madras, India).EPDM (KELTAN 720) with E/P ratio 58/35.5 wt%and DCPD content 6.5 wt% with Mn 120 000 wasobtained from DSM (Netherlands).The EPM-g-MA with 1 wt% of MA grafting withE/P ratio 55/45 having density 0.88 g/mol (Roy-altuf 465) was supplied by Uniroyal ChemicalCompany, Germany.

2.2. Blend preparation

Nylon copolymer (PA6,66) pellets were dried invacuum oven at 80°C for 24 h before blending .Theblends were prepared in a Brabender at a tempera-ture of 180°C and a rotor speed of 60 rpm. Nylonwas first melted for 2 minutes and then EPDM wasadded and the mixing was continued for 6 moreminutes. In the case of compatibilized blends, thecompatibilizer EPM-g-MA was mixed with EPDMin a two roll mill at room temperature so as to getEPM-g-MA coated EPDM. Nylon is fed into thepreheated Brabender at 180°C and melted for2 minutes. Then EPM-g-MA coated EPDM isadded and melting is continued for 6 more minutes(Total mixing time fixed was 8 minutes). Test sam-ples were prepared by compression molding in ahydraulic press at 180°C. The volumetric propor-tion of 70/30 nylon/EPDM blends were preparedwith compatibilizer concentrations of 0, 1, 2.5, 5,10 wt% based on the minor phase. Test specimensare compression moulded into 2 mm thick sheets at180°C for 3 minutes in an electrically heatedhydraulic press at a pressure of 10 MPa.

2.3. Designation of the blends

The blends were designated as follows. N0 meansEPDM and N100 means nylon copolymer. N70

means a blend of 70 parts of nylon and 30 parts ofEPDM. The binary blends were designated as N0,N30, N50, N70 and N100 where the subscripts denotethe weight percentage of nylon in the blend. TheEPM-g-MA compatibilized N70 blends were desig-nated as N70x, where x denotes the weight percent-age of compatibilizer in the blend. Various blendsstudied with their compositions are summarized inTable 1.

642

Komalan et al. – eXPRESS Polymer Letters Vol.1, No.10 (2007) 641–653

Table 1. Compositions and codes of the samples used

N – nylon copolymer, N0 – ethylene propylene diene rubber

SystemPA6,66 EPDM EPM-g-MA

[wt%] [vol%] [wt%] [vol%] [wt%] [vol%]N0 – – 100.0 100.0 – –N30 30 27.6 70.0 72.4 – –N50 50 50.0 50.0 .050.0 – –N70 70 71.4 30.0 30.6 – –N100 100 100.0 – – – –N701 70 71.4 29.0 29.6 1.0 0.3N702.5 70 71.4 27.5 28.1 2.5 0.7N705 70 71.4 25.0 25.5 5.0 1.3N7010 70 71.4 20.0 20.4 10.0 2.0

2.4. Measurement of dynamic mechanicalproperties

The dynamic mechanical properties of nylon/EPDM blends were determined using DynamicMechanical Thermal Analyzer (DMTA; Model-2980, supplied by TA Instrument (USA). Theshape of test sample is a rectangular strip of40×12×2 mm3. The dual cantilever mode of defor-mation was used under the test temperature rangefrom –100 to 150°C with a heating rate of 2°C/minunder liquid nitrogen flow. Storage modulus (E ' )loss modulus (E '') and loss tangent (tanδ) of eachsample were recorded in multi-frequency mode at0.1, 1, 10 Hz.

2.5. Morphology

The morphology of the freeze-fractured cross-sec-tions was examined using a scanning electronmicroscope JEOL JSM-820. Observations weremade from etched samples. From N70 and N50

blends EPDM phase was extracted using xylenevapours. From N30 blend, nylon phase is extractedusing formic acid. The samples were dried in an airoven at 120°C for 24 h and preserved in a desicca-tor. The solvent extracted samples were sputtercoated with gold (Au) and SEM observations weremade.

3. Results and discussions

3.1. Effect of blend ratio (uncompatibilizedblends)

Dynamic mechanical analysis helps to study poly-mer/polymer miscibility in polymer blends and alsomeasure the glass transition temperatures (Tg) ofpolymers. Moreover, we can obtain an idea aboutthe storage (dynamic) modulus, loss modulus anddamping behaviour (internal friction). The resultsof dynamic mechanical analysis add informationabout the behaviour of the blends and phase mor-phology. The effects of temperature and blend ratioon the storage modulus (E ') of the samples at a fre-quency of 10 Hz are given in the Figure 1. Thevalue of storage modulus, E ' signifies the stiffnessof the material. All the curves show three distinctregions: a glassy high modulus region where thesegmental mobility is restricted, a transition zonewhere a substantial decrease in the E ' values withincrease of temperature and a rubbery region (the

flow region) where a drastic decay in the moduluswith temperature. The storage modulus curve ofEPDM (N0) shows a typical behaviour of an unvul-canized elastomer. EPDM shows a high modulusbelow its Tg followed by a drastic drop in its magni-tude around –52°C. This drastic decrease in modu-lus with temperature around –52°C indicates dis-tinct transition from glassy to rubbery state. At anyfixed rate of deformation, the temperature at whichE ' starts to decrease rapidly corresponds to theglass transition temperature. But in the case ofnylon, changes in the storage modulus are lesssevere around the glass transition zone because ofits semi-crystalline nature. In the semi-crystallinematerials like nylon, the crystalline chains arearranged in a regular order, and it will remain intactuntil the temperature reaches the melting point(Tm). Only the amorphous part undergoes segmen-tal motion. So, in the case of nylon, the storagemodulus decreases to a smaller extent than EPDMdoes in the transition. It may be noted that nylon has the maximum andEPDM has the minimum E ' values. The E ' valuesof the blends are found to be intermediate betweenthose of pure components depending on the propor-tion of EPDM. At very low temperature, modulusof the blends is high. The blends show transitionand plateau region corresponding to EPDM andnylon. The storage modulus decreased with increasein temperature and finally levels off at high temper-ature. It is seen from figure1 that the value of stor-age modulus decreases with increase in the concen-tration of EPDM content, which is more pro-nounced at high temperatures. On adding more

643


Figure 1. Storage mudulus (E') curves as function of tem-perature for nylon, EPDM and physical blend ofnylon/EPDM

EPDM, the crystallinity is reduced. Therefore themodulus of EPDM rich blends decreases muchfaster at high temperatures. EPDM is having verylow modulus in the rubbery plateau. The two-stepcurves in the figure for the blends are due to two-phase morphology indicating immiscibility. Theabove results are very much in agreement withthose of systems based on blends of PP\EPDMstudied by Papke and Karger-Kocsis [20]. Thechange from dispersed phase morphology in N30 toco-continuous morphology in N50 leads to anincrease in modulus. In co-continuous structuresthe storage modulus-temperature dependencereflects a greater contribution of both components,whereas in dispersed structures, the blend modulusis dominated by the matrix component [21]. Therelationship between storage modulus at a giventemperature can also determine the region of phaseinversion and the existence of co-continuous struc-tures, as shown by Dedecker and Groeninckx [22]for reactively compatibilized PA6/PMMA blends.Variation of loss modulus (E '') with temperature ofnylon, EPDM and nylon/EPDM blends are given inFigure 2. The loss modulus peak corresponds to themaximum heat dissipation per unit deformation.The glass transition temperature, Tg was selected asthe peak position of E '' when plotted as a functionof temperature. From the figure, it is clear that N100

shows a peak around 13.5°C corresponding to theTg of nylon whereas EPDM (N0) shows a peakaround –44°C. Another important observation thatcan be made from the figures is that nylon exhibitsa strong β-relaxation around –57°C. However twodistinct peaks each exactly corresponding to theglass transition temperatures of nylon and EPDM

can be observed in all the blends indicating theincompatibility and immiscibility between thephases. The loss modulus increases with increase inEPDM content. Various researchers [20, 22, 23]used dynamical mechanical investigation to predictthe miscibility of polymeric systems. Generally forincompatible system, the tanδ versus temperaturecurve shows two damping peaks corresponding tothe glass transition temperatures of individual poly-mers [20]. When blend components are compatible,a single peak is found for the combined processes[23]. Broadening of the transition occurs in the caseof partially compatible systems. Shift in the Tg tohigher or lower temperatures as a function of com-position also indicates the partial miscibility. Fig-ure 3 shows the variation of loss tangent (tanδ) withtemperature for the pure components and nylon/EPDM blends. The main relaxation processes in theamorphous region for the nylon and EPDM phasesin the blends can be detected from the dampingcurves presented in Figure 3. As the temperatureincreases, damping goes through a maximum nearTg in the transition region, and then a minimum inthe rubbery region. At Tg, micro-Brownian motionof molecular segments begins where short rangediffusion can take place. The damping is low belowTg because thermal energy is insufficient to causerotational and transnational motions of the seg-ments [24]. As a result of this, the chain segmentsare frozen in. So, below Tg, the molecular slipresulting in viscous flow is low. Above Tg also thedamping is low because molecular segments arevery free to move about and there is little resistancefor their flow. Hence, when the segments are eitherfrozen in or are free to move, damping is low. In the

644


Figure 2. Loss modulus (E'') curves as function of temper-ature for nylon, EPDM and nylon/EPDM blend

Figure 3. Tanδ curves as function of temperature fornylon, EPDM and l blends nylon/EPDM

transition region a part of the segments are free tomove about and the remainder are not so free. Afrozen segment stores energy through deformationand it ultimately releases it as viscous energy whenit becomes free to move. The tanδ curve of EPDMshows a peak at –32°C due to the α-transition aris-ing from the segmental motion. This α-transitioncorresponds to the glass transition temperature (Tg)of EPDM. Nylon shows the glass transition temper-ature at 26°C. Another important observation thatcan be made from the Figures 2, 3 is that nylonexhibits a strong β-relaxation [25] around –55°C. Itis believed that the β damping peak is due to thecarbonyl group of nylon forming hydrogen bonds[26]. The temperature of this peak is influenced bynylon moisture content. Some researchers relatedthis secondary transition to movements involvingcarbonyl groups, which have formed hydrogenbonds with absorbed water [27]. The absence of theβ-relaxation peaks in the blends indicates that addi-tion of EPDM into nylon decreased the wateruptake property of nylon as discussed else where.EPDM has higher damping than nylon because ofits rubbery nature and the flexible rubber chainsrespond rapidly towards a cyclic loading. Since therubber is uncrosslinked, the uncoiling and recoilingprocess on the application and removal of the stressmakes more permanent deformation and therebyregistering the highest loss tangent (tanδmax) values.As can be seen from the figures, all the blend com-positions show two distinct and clearly separatetanδ peaks corresponding to the Tg’s of nylon andEPDM indicating that the blends are incompatible.The Tg values obtained from the loss modulus

curve and tanδ curves are given in Table 2. Asusual, the Tg values obtained from E '' curves arelower than those from tanδ curves. The Tg valuestaken from the tanδ peaks of virgin polymers andblends at different frequencies are given in Table 3.In all cases the Tg values increase with the fre-quency. As the mechanical frequency is increasedthe position of the glass-rubber transition moves toa higher temperature because the polymer chainsneed more energy to respond to the shorter timescale stresses imposed at higher frequencies. Theshift in the Tg values upon the addition of theEPDM especially at higher concentrations ofEPDM may be due to enhanced chain mobility ofnylon due to the plasticizing action of the flexibleEPDM phase. The EPDM has the highest damping.It is accepted that higher the tanδmax the greater themechanical losses. These losses are related to highenergy input required for the motion of the molecu-lar chains of the polymer as the transition is beingapproached [28]. The results also suggested thatincreasing the rubber content caused an increase ofelastic behaviour. A small hump in the tanδ peak atlower temperature (–73°C) of N30 may be due tomotion of the side chain and imperfection.The damping behaviour of the blends increaseswith an increase in the concentration of EPDM rub-ber. The variation of tanδmax of blend as a function

645


Table 2. Tg values of various nylon/EPDM blens (10 Hz)

Sample

Tg [°C] with respect toEPDM Nylon

From tanδδmax

FromE''

From tanδδmax

FromE''

N0 –32 –42 – –N30 –31 –46 31 17,0N50 –37 –41 27 17,0N70 –44 –46 26 16.7N100 – – 26 13.5

Table 3. Tg [°C] values of various nylon/EPDM blends at various frequencies obtained

Frequency[Hz]

N0

N30 N50 N70

N100Due toEPDM

Due tonylon

Due toEPDM

Due tonylon

Due toEPDM

Due tonylon

0.1 –41 –45 17 –47 20 –51 17 131 –38 –39 20 –42 22 –47 21 1910 –32 –31 31 –37 27 –40 26 26

Figure 4. Variation of tanδ of nylon, EPDM andnylon/EPDM blends with temperature

of EPDM content measured at 10 Hz is shown inFigure 4. EPDM shows maximum value of tanδindicating its excellent damping behaviour. In theblends with nylon, the crystalline plastic phase act-ing as physical crosslinks imposes some restrictiontowards cyclic loading and the tanδmax decreaseswith increases in nylon content. The increase in thedamping and tanδmax with increase in EPDM con-tent is due to the reduction in the crystalline volumeof the system on increasing the concentration ofEPDM whose damping is always higher thannylon.The increase in tanδmax of N30 (70% ofEPDM) in Figure 4 can be explained by morpho-logical changes. Examination of the morphology of

the blends by scanning electron microscope (Fig-ure 5) reveals that for blends containing 70%EPDM, the nylon phase is dispersed as sphericalparticles (observed as semi-spherical holes afteretching by hot xylene to remove the EPDM phase)in the continuous EPDM matrix. In N30, the rubberphase has become continuous and its responsetowards a cyclic loading is more effective and thereby exhibiting highest tanδmax. In N70, nylon is foundto form the continuous phase in the blends due to itsincreased concentration. In this contribution of rub-ber towards tanδmax is restricted by the crystallineplastic phase. N50 blend shows a co-continuousmorphology and the rubber particles exist as largedomains and their contribution towards tanδmax isincreased. SEM micrographs of the blends show aclear two phase morphology with the rubber parti-cles being coarsely dispersed in the continuousnylon phase and having clear and sharp interfacialboundaries which may be attributed to high interfa-cial tension indicating poor adhesion at phaseboundaries, and this is a manifestation of theincompatibility of the polymer components in theseblends. It is well known that blends based onimmiscible polymer pairs are characterized bygreat interfacial tension, which makes the disper-sion during the blending operation difficult, andcontributes to unstable morphology and poor adhe-sion [29].The activation energy, E for the glass transition ofthe blends can be calculated from the Arrheniusequation (1):

(1)

where f is the experimental frequency of transition,A, is a constant, E is the activation energy, R is theuniversal gas constant, T is the temperature corre-sponding to the maximum of the tanδ curve inKelvin scale [K]. The plots of log f versus 1/T wereconstructed and the values of the activation ener-gies were calculated from the slope of the plots.This slope corresponds to the activation energy ofthe viscous flow that accompanies the glass transi-tion. The Arrhenius plot corresponding to EPDMtransition in 70/30 nylon/EPDM blend is given inthe Figure 6. Activation energy values are given inthe Table 4. The blends show higher activationenergy than the EPDM. Due to the amorphousnature of the EPDM, it is less sensitive to tempera-

646


RT

EAf 030.2loglog −=

Figure 5. Scanning electron micrograph of nylon copoly-mer/EPDM blends at a magnification of 250times

ture hence the activation energy decreases withincrease in EPDM content. The chain flexibilityincreases with increase in the rubber content. Asthe flexibility increases, the crystallinity decreases.As a result of this the activation energy decreases.

3.1.1. Theoretical analysis of storage modulus

Model studies were carried out to assess the behav-iour of the two-phase blend from the componentproperty data. The various composite models suchas parallel, series, Coran’s and Takayanagi’s havebeen used to predict the viscoelasic behaviour of thebinary blends. The upper bound of the storage mod-ulus is given by the rule of mixtures (Equation (2)):

Eu = E1φ1 + E2φ2 (2)

where Eu is the property of the blend and E1 and E2

are the corresponding properties of the componentsφ1 and φ2 respectively. and represent the volumefraction of components 1 and 2 respectively. Thisequation is applicable to the materials in which thecomponents are arranged parallel to the appliedstress. The applied stress elongates each componentby the same amount. In the lowest lower boundseries model, the blend components are arranged in

series (Reuss prediction) perpendicular to the direc-tion of the applied force. The equation for the seriescombination of the components is given by Equa-tion (3):

(3)

where EL is the moduli of the blend in the seriesmodel. For both these models, there is no relationbetween the morphology and the property.According to Halpin-Tsai model [30, 31], theEquation (4) that relates the morphology of thepolymer blend to the properties is:

(4)

where

(5)

In the Equation (5) the subscript 1 and 2 refer to thecontinuous and dispersed phases respectively. Theconstant Ai is determined by the morphology of thesystem. For elastomer domains, dispersed in a con-tinuous hard matrix, Ai = 0.66 and when the hardmaterial forms the dispersed phase, Ai is 1.5. Thismodel was also useful in determining the propertiesof polymer blends that contained both continuousand discontinuous phases. This model has also beensuccessfully applied by several researchers to sys-tems [32, 33] of polymer composites.In Coran’s model [34], the mechanical propertiesare generally in between the upper bound parallelmodel (EU) and the lower bound series model (EL)(Equation (6)):

E = f(EU–EL) + EL (6)

where ‘f’ can very between zero and unity. Thevalue of ‘f’ is given by Equation (7):

(7)

where n contains the aspects of phase morphology.VH and VS are the volume fractions of hard phaseand soft phase respectively. It can be seen fromFigure 7 that the experimental data are very close tothe Coran’s modal with an n value of 2.2.

647


2

2

1

11

EEEL

φ+φ=

2

21

1

1

φ−φ+

=i

ii

B

BA

E

E

i

i

AE

EE

E

B+

−=

2

1

2

1 1

Figure 6. Arrhenious plot corresponding to EPDM transi-tion in 70/30 nylon/EPDM blends

Table 4. Activation energy for the transition of EPDMphase in nylon/EPDM blends

SampleActivation energy (ΔΔE) [kJ/mol–1]

for EPDM phase transitionN70 423N50 261N30 189N0 51

( )1+= SnH nVVf

The viscoelastic behaviour of the heterogeneouspolymer blend can be predicted using Takayanagimodel [35] in which the concept of percolation isintroduced. According to this model (Equation (8)):

(8)

where E1 is the modulus of the matrix phase, E2 isthe modulus of the dispersed phase and the valuesof λ and φ2 are related to the degree of series-paral-lel coupling.The degree of parallel coupling of the model can beexpressed by Equation (9):

(9)

The curves resulting from the different theoreticalmodels and that of the experimental data for thevariation of storage modulus at 25°C with volumefraction of nylon are given in Figure 7. Coran’smodel (n = 2) fits well with the experimental curveand the Takayanagi model with 45% parallel cou-pling agrees to some extent with the experimentalcurve. The series model is the lowest bound overthe entire composition range.

3.2. Effect of compatibilization

As discussed in the above section, the nyloncopolymer/EPDM blend systems are incompatible.The incorporation of a compatibilizer into animmiscible blend reduces the interfacial energy of

the phases, stabilizes the morphology against coa-lescence and improves the interfacial adhesion. Asa result, systems with improved and reproducibleproperties are obtained.Dynamic mechanical analysis (DMA) is sensitiveto molecular motions and transitions [36] has beenchosen as a tool for characterizing blend compati-bilisation. The objective of the current work is touse DMA to study the glass transition behaviour ofimmiscible polymer blends in an effort to charac-terize blend compatibilisation on the molecularlevel and to make a correlation between phase mor-phology and dynamic mechanical properties.The variation of storage modulus as a function oftemperature for 70/30 nylon/EPDM blends compat-ibilized with different concentrations of EPM-g-MA is shown in Figure 8. The E ' values of compat-ibilized N70 blends are higher than that of uncom-patibilized blend at the same temperature. Theaddition of EPM-g-MA makes the blend technolog-ically compatible to some extent even thoughmolecular level miscibility cannot be achieved.Even by the addition of 1% compatibilizer, the stor-age modulus of the blends shifted towards thelower temperatures, i. e, below the glass transitiontemperature of nylon. The increase in the modulusupon the addition of the compatibilizer is due to theincrease in the interfacial adhesion caused by theemulsifying effect of the copolymer formed by thereaction between EPM-g-MA and nylon. The pres-ence of compatibilizer improves the interfacialadhesion by enhancing the interfacial thickness andthereby facilitating the stress transfer between thecomponents. Thus EPM-g-MA, improved the inter-

648


[ ] ( )100·

1

1parallel%

2

2 ⎥⎦

⎤⎢⎣

⎡λφ−λ−φ=

Figure 8. Effect of compatibilisation on the variation ofstorage modulus as a function of temperature inN70 blends

( )

2

2

1

21 1

1

EE

EEφ+φ−

λ+λ−=

Figure 7. Experimental and theoretical curves of storagemodulus of nylon/EPDM blends as a function ofvolume fraction of nylon at room temperature(28°C)

facial adhesion, consequently, the storage modulusof the blends enhanced. The increased interfacialinteraction is evident from the small and uniformdispersion of EPDM particles upon the addition ofEPM-g-MA (Figure 9). A decrease in E ' by theaddition of 10 wt% EPM-g-MA indicates the for-mation of micelles in the nylon matrix. At higherloading of the compatibilizer, modulus valuedecreases due to the formation of agglomerates ofthe compatibilizer molecule.Now let us consider the effect of compatibilisationon the Tg’s of the blends. Figure 10 depicts thechange of loss modulus (E '') as function of temper-ature for nylon/EPDM blends compatibilized byEPM-g-MA. The E '' values of the compatibilizedsystems are higher than those of N70 uncompatibi-lized blend. The shift in the Tg values of the nylonphase of the compatibilized blends compared to theuncompatibilized blends indicate the strong inter-action between the nylon and EPDM as a result ofreactive compatibilisation. Thus a shifting of Tg

values of the components is taken as an indicationof miscibility enhancement and therefore can beconsidered as an evidence for compatibilisation.

Due to the reaction the mobility of the nylon phasedecreases.There is no change in the Tg values of EPDM phaseof the compatibilized blends in the loss moduluscurve. This is because, at the lower Tg in the com-patibilized blends, there are two types of con-straints imposed on the motions of molecularsegments in the vicinity of the interface. One is dueto the presence of another phase and other is due to

649


Figure 10. Effect of compatibilisation on the variation ofloss modulus as a function of temperature inN70 blends

Figure 9. Scanning electron micrographs of N70 blend containing different levels of EPM-g-MA

the presence of chemical bonds between the twophases. As the blend goes through the lower Tg, onephase retains its glassiness. The constraints imposedby the glassy phase apparently dominates the pres-ence of chemical bonds between phases. Therefore,no difference is observed when one compares thelower Tg behaviour of the compatibilized anduncompatibilized blends. And below Tg, the molec-ular mobility is reduced. So the rate of free volumechanges with temperature is very small. Therefore,when both blend components are below theirrespective Tg, free volume effects are small and E ''is relatively constant with temperature. Since thecompatibilisation reduces interfacial energy andcreates a finer dispersion by increasing the interfa-cial area [37], one can expect an overall volume ofthe interaction zone to increase when copolymer isadded. When the blend goes through the high Tg, allthe molecules are relaxed. In the compatibilizedblends the linked segments do not have the degreeof mobility as the unlinked segments have, so thatthe effective number of segments free to com-pletely relaxed has decreased. So there is a shifttowards lower Tg. The Tg values obtained (Table 5)from E '' vs. temperature is always less than thatobtained from tanδ maximum.DMA has become a classical method for the deter-mination of miscibility because the height and posi-tion of the mechanical damping peaks are affectedremarkably by miscibility, intermolecular interac-tion, interface feature and morphology. The varia-tion of tanδ as a function of temperature of thecompatibilized blends is given in Figure 11. Thecompatibilized blends also show the presence oftwo peaks corresponding to the Tg’s of nylon andEPDM similar to those of the uncompatibilized

blends. This indicated that compatibilisation didnot alter the degree of miscibility. The broadeningof the peaks is an indication of the compatibilisa-tion. The general broadening of the tanδ peaks isassociated with increased molecular mixing [38],which is expected. As the compatibility increases,the interpenetration of the components increases[39]. There is no appreciable shift in the Tg towardsthe average value. This emphasizes that significantmiscibility increase does not occur for the compati-bilized blends. However, domain boundary mixing[40] or interface mixing [41] have been shown tocause an increased temperature dependence of stor-age modulus and increased level of loss modulusbetween Tg for compatibilized relative to uncom-patibilized blends. From the Figure 11 it is clearthat 2.5% compatibilizer loaded blend shows thehighest tanδ than other blends in various tempera-tures. This may be due to the maximum interfacialinteraction provided by the compatibilizer. This isevident from the SEM analysis also. Maximum par-ticle size reduction is obtained for 2.5% of compat-ibilizer (Figure 9).We can conclude that both electron microscopy anddynamical mechanical analysis clearly show thatthe system is still phase separated even in the pres-ence of the compatibilizer and tanδ curve also indi-cates that compatibilizer addition could not makethe system completely miscible. This is in agree-ment with the conclusions made by Paul [42] that iftwo polymers are far from being miscible, then nocopolymer is likely to make a single phase system.In a completely immiscible system, the main role ofthe copolymer is to act as an interfacial agent.

650


Figure 11. Effect of compatibilisation on the variation oftanδ as a function of temperature in N70 blends

Table 5. Effect of compatibilisation on the Tg values ofEPDM and nylon in N70 blends at 10 Hz

Blends

Glass transition temperature [°C]From E'' vs. T curve From tanδδ vs. T curveEPDMphase

Nylonphase

EPDMphase

Nylonphase

N70 –46 16.7 –40 26

1%Compatibiliser –44 –11.0 –41 23

2.5% Compatibiliser –44 –9.3 –42 24



3.3. Degree of entanglement density

Degree of entanglement of the polymer blends canbe obtained from swelling measurements, stress-strain curve and also from dynamic mechanicalanalysis. We can use the storage modulus data fordetermining the entanglement density using theEquation (10):

(10)

where E ' is the storage modulus obtained from theplateau region of E ' versus temperature curve, R isthe universal gas constant and T, is the absolutetemperature. A higher degree of entanglement isexhibited by the compatibilized blends when com-pared to the uncompatibilized blends. This isbecause by the addition of compatibilizer the entan-glement between the homopolymers increases anda better adhesion is achieved as a result of decreasein the interfacial tension. The value for degreeentanglement between nylon and EPDM are givenin the Table 6. From the morphology studies it isclear that immiscible polymers exhibit a coarse dis-persion. By the addition of the compatibilizer theparticle size decreased leading to a better adhesion.Oommen et al. [43] reported that the entanglementdensity increases as a result of compatibilisation. Itis interesting to note that, the entanglement densityincreases up to 2.5 wt% of compatibilizer concen-tration followed by a decrease at higher loading.This indicates clearly that the compatibilizer leavesthe interface after critical micelle concentration(CMC). Our earlier studies revealed that the mini-mum amount of compatibilizer (CMC) required tomodify the interface is 2.5% of EPM-g-MA.

4. Conclusions

1. In uncompatibilized blends, nylon shows themaximum and EPDM shows the minimum stor-age modulus. The storage modulus of blends

was found to be intermediate between the purecomponents.

2. For the uncompatibilized blends, loss modulusand tanδ values did not able to give any informa-tion about the favourable interactions betweennylon and EPDM, as their transition peaks (Tg ofnylon ~26°C, Tg of EPDM ~–32°C and β-relax-ation of nylon ~–57°C) experienced no shift asa function of blend ratio.

3. As the concentration of the EPDM rubberincreases the storage modulus decreases and thedamping and loss modulus increases.

4. The experimental data of uncompatibilizedblends were theoretically modeled and observedthat Coran’s model (n = 2.2) was best suited forthe present system and the Takayanagi modelwith 45% parallel coupling agrees to someextent.

5. The addition of EPM-g-MA as compatibilizerimproved the viscoelastic properties indicatingimproved interaction between the two compo-nents in the compatibilized system.

6. A higher degree of entanglement is exhibited bythe compatibilized blends when compared to theuncompatibilized blends.

7. Morphologies of the blends have got a profoundinfluence on the dynamic mechanical properties.

References[1] Craver C. D., Carraher C.: Applied polymer science-

21st century. Elsevier, New York (2000).[2] Holden G., Legge N. R., Quirk R. Q., Schroeder H. E.:

Thermoplastic elastomers. Hanser-Gardner Publica-tions, Cincinnati (1996).

[3] Yang H., Lai M., Liu W., Sun C., Liu J.: Morphologyand thermal and mechanical Properties of PBT/HIPSand PBT/HIPS-g-GMA blends. Journal of AppliedPolymer Science, 85, 2600–2608 (2002).

[4] John B., Varughese K. T., Oommen Z., Pötschke P.,Thomas S.: Dynamic mechanical behavior of high-density polyethylene/ethylene vinyl acetate copolymerblends: The effects of the blend ratio, reactive compat-ibilization, and dynamic vulcanization. Journal ofApplied Polymer Science, 87, 2083–2099 (2003).

[5] George J.: Compatibilisation of high density polyeth-ylene/nitrile rubber blends. PhD thesis, MahatmaGandhi University (2003).

[6] Ramirez-Vargas E. R., Sandoval-Arellano Z., Hernan-dez-Valdez J. S., Martinez-Colunga J. G., Sanchez-Valdes S.: Compatibility of HDPE/postconsumerHDPE blends using compatibilizing agents. Journal ofApplied Polymer Science, 100, 3696–3706 (2006).

651


RT

EN

6

'=

Table 6. Entanglement Density of 70/30 nylon/EPDMblends at 10 Hz

Temperature[°C]

Compatibiliser[wt%]

Entanglement density[moles/m3]

239

0.0 1231.0 1522.5 2735.0 251

10.0 234

[7] Diaz M. F., Barbosa S. E., Capiati N. J.: Additioncompatibilization of PP/PS blends by tailor-madecopolymers. Polymer Engineering and Science, 46,329–336 (2006).

[8] Mirzazadeh H., Katbab A. A.: PP/EPDM-based ther-moplastic dynamic vulcanizates with organoclay:morphology, mechanical and viscoelastic properties.Polymers for Advanced Technologyes, 17, 975–980(2006).

[9] Yang Y-L., Lin Y., Sell C., Hiver J-M., Bai S-L.:Dynamic mechanical properties and morphology ofhigh-density polyethylene/CaCO3 blends with andwithout an impact modifier. Journal of Applied Poly-mer Science, 103, 3907–3914 (2007).

[10] Al-Malaika S., Kong W.: Reactive processing of poly-mers: effect of in situ compatibilization on character-istics of blends of polyethylene terephthalate andeth-ylene-propylene rubber. Polymer, 46, 209–228 (2005).

[11] Rajan G. S., Vu Y. T., Mark J. E., Myers C. L.: Ther-mal and mechanical properties of polypropylene in thethermoplastic elastomeric state. European PolymerJournal, 40, 63–71 (2004).

[12] Nandan B., Kandpal L. D., Mathur G. N.: Glass transi-tion behaviour of poly(ether ether ketone)/poly(arylether sulphone) blends: dynamic mechanical anddielectric relaxation studies. Polymer, 44, 1267–1279(2003).

[13] Kumar M. S. C., Alagar M., Prabu A. A.: Studies ondynamic mechanical and mechanical properties ofvinyloxyaminosilane grafted ethylene propylene dieneterpolymer/linear low density polyethylene (EPDM-g-VOS/LLDPE) blends. European Polymer Journal, 39,805–816 (2003).

[14] Karger-Kocsis J., Kiss L.: Dynamic mechanical prop-erties and morphology of polypropylene block copoly-mers and polypropylene/elastomer blends. PolymerEngineering and Science, 27, 254–262 (1987).

[15] Guo T-Y., Song M-D., Hao G-J., Zang B-H.: Compat-ibility and mechanical properties of polyamide-6/polystyrene/diblock poly(styrene-b-ethylene oxide)copolymer blends. European Polymer Journal, 37,241–246 (2001).

[16] Gopakumar T. G., Ponrathnum S., Lele A., Rajan C.R., Fradet A.: In situ compatibilization of poly (pheny-lene sulphide)/wholly aromatic thermotropic liquidcrystalline polymer blends by reactive extrusion: mor-phology, thermal and mechanical properties. Polymer,40, 357–364 (1999).

[17] Seo Y.: TLCP ternary blends for in situ composites: Insitu compatibilizer. Journal of Applied Polymer Sci-ence, 70, 1589–1595 (1998).

[18] Tjong S. C., Li R. K. Y., Xie X. L.: Compatibilizingeffect of styrene-maleic anhydride copolymer on theproperties of polyamide-6/liquid crystalline copoly-ester composites. Journal of Applied Polymer Science,77, 1964–1974 (2000).

[19] Moly K. A., Bhagawan S. S., Groeninckx G., ThomasS.: Correlation between the morphology and dynamicmechanical properties of ethylene vinyl acetate/linearlow-density polyethylene blends: Effects of the blendratio and compatibilisation. Journal of Applied Poly-mer Science, 100, 4526–4538 (2006).

[20] Papke N., Karger-Kocsis J.: Thermoplastic elastomersbased on compatibilized poly(ethylene terephthalate)blends: effect of rubber type and dynamic curing.Polymer, 42, 1109–1120 (2001).

[21] Quintens D., Groeninckx G., Guest M., Aerts L.:Mechanical behavior related to the phase morphologyof PC/SAN polymer blends. Polymer Engineering andScience, 30, 1474–1483 (1990).

[22] Dedecker K., Groeninckx G.: Reactive compatibilisa-tion of A/(B/C) polymer blends. Part 2. Analysis ofthe phase inversion region and the co-continuousphase morphology. Polymer, 39, 4993–5000 (1998).

[23] Varughese K. T., Nando G. B., De P. P., De S. K.:Miscible blends from rigid poly (vinyl chloride) andepoxidised NR. Journal of Material Science, 23,3894–3900 (1988).

[24] Ferry J. D.: Viscoelastic properties of polymers. JohnWiley, New York (1980).

[25] Dickie R. A.: Heterogeneous polymer-polymer com-posites. I. Theory of viscoelastic properties and equiv-alent mechanical models. Journal of Applied PolymerScience, 17, 45–63 (1973).

[26] Bell J. P., Murayama T.: Relation between dynamicmechanical properties and dye diffusion behavior inacrylic fibers. Journal of Applied Polymer Science,12, 1795–1799 (1968).

[27] Gadekar R., Kulkarni A., Jog J. P.: Blends of nylonwith polyethylene: Effect of compatibilization onmechanical and dynamic mechanical properties. Jour-nal of Applied Polymer Science, 69, 161–168 (1998).

[28] Sombatsompop N.: Dynamic mechanical properties ofSBR and EPDM Vulcanisates filled with cryogeni-cally pulverized flexible polyurethane foam particles.Journal of Applied Polymer Science, 74, 1129–1139(1999).

[29] Hu G-H., Sun Y-J., Lambla M.: Effects of processingparameters on the in situ compatibilization of poly-propylene and poly(butylene terephthalate) blends byone-step reactive extrusion. Journal of Applied Poly-mer Science, 61, 1039–1047 (1996).

[30] Nielsen L. E.: Morphology and the elastic modulus ofblock polymers and poly blends. Rheologica Acta, 13,82–92 (1974).

[31] Halpin J. C.: Structure-property relations and reliabil-ity concepts. Journal of Composite Materials, 6, 208–2031 (1972).

[32] Kalaprasad G., Joseph K., Thomas S., Pavithran C.:Theoritical modelling of tensile properties of shortsisal fibre reinforced low density polyethylene com-posites. Journal of Material Science, 32, 4261–4267(1997).

652


[33] Nair K. C. M., Divan S. M., Thomas S.: Tensile prop-erties of short sisal fiber reinforced polystyrene com-posites. Journal of Applied Polymer Science, 60,1483–1497 (1996).

[34] Stephens H. L., Bhowmick A. K.: Handbook of elas-tomers: New developments and technology MarcelDekker, New York (1988).

[35] Dickie R. A.: Heterogeneous polymer-polymer com-posites. I. Theory of viscoelastic properties and equiv-alent mechanical models. Journal of Applied PolymerScience, 17, 45–763 (1973).

[36] Murayama T.: Dynamic Mechanical analysis of poly-mer material. Elsevier, New York (1978).

[37] Paul D. R.: Interfacial agents (compactibilizers) forpolymer blends. in ‘Polymer Blends’ (eds.: Paul D. R.,Newman S.) Academic Press INC, New York, Vol 2,35–62 (1978).

[38] Thomas D. A., Sperling C. H.: Interpenetrating poly-mer networks. in ‘Polymer Blends’ (eds.: Paul D. R.,Newman S.) Academic Press INC, New York, Vol 2,1–33 (1978).

[39] Wu S.: Interfacial energy, structure and adhesionbetween polymers. in ‘Polymer Blends’ (eds.: Paul D.R., Newman S.) Academic Press INC, New York,Vol 1, 243–293 (1978).

[40] Hashimoto T., Tsukahara Y., Tachi K., Kawai H.:Structure and properties of tapered block polymers. 4.‘Domain-boundary mixing’ and ‘mixing-in-domain’effects on microdomain morphology and lineardynamic mechanical response. Macromolecules, 16,648–657 (1983).

[41] Feng D., Wilkens G. L., Crivello J. V.: Structure-prop-erty behaviour of free radical synthesized polydi-methylsiloxane-polystyrene multiblock polymers: 1.Effect of the siloxane block length. Polymer, 30,1800–1813 (1989).

[42] Newman S.: Rubber modifications of plastics. in‘Polymer Blends’ (eds.: Paul D. R., Newman S.) Aca-demic Press INC, New York, Vol 2, 63–89 (1978).

[43] Oommen Z., Groninckx G., Thomas S.: Dynamicmechanical and thermal properties of physically com-patibilized natural rubber/poly (methyl methacrylate)blends by the addition of natural rubber-graft-poly(methyl methacrylate). Journal of Polymer SciencePart B: Polymer Physics, 38, 525–536 (2000).

653


Dynamic mechanical analysis of binary and ternary … · 2.4. Measurement of dynamic mechanical...

Documents

Transcript of Dynamic mechanical analysis of binary and ternary … · 2.4. Measurement of dynamic mechanical...