blends of epoxidized natural rubber and thermoplastics

In: Rubber: Types, Properties and Uses Editor: Gabriel A. Popa

ISBN: 978-1-61761-464-4 © 2011 Nova Science Publishers, Inc.

Chapter 6

BLENDS OF EPOXIDIZED NATURAL RUBBER AND THERMOPLASTICS C. H. Chan *, H. W. Kammer, L. H. Sim, M. K. Harun Faculty of Applied Sciences, Universiti Teknologi MARA, 40450 Shah Alam, Malaysia

ABSTRACT Epoxidized natural rubber (ENR) is chemically modified natural rubber. When blended with crystalline thermoplastic polymers, immiscible systems result. Properties of those systems are ruled by the relative amount of the constituents and the distribution of phases in the alloy. Thermal behavior governs development of morphologies by a delicate balance between rate of crystallization and mobility of the elastomer phase. In blends where the crystalline constituent is in excess, this component crystallizes out from the melt comprising a dispersion of ENR domains. At the opposite side of the composition scale, ENR being in excess, crystalline domains will develop in the ENR matrix. In the chapter, we focus dominantly on three aspects: thermal properties and morphology of elastomer thermoplastic blends. It embraces melting and crystallization behavior in those blends and their influence on morphology development. We follow also the question: Do these blend systems become candidates for polymer electrolytes when one adds salt? In what way is the salt distributed between the blend components? These are crucial fundamental problems and it seems worthwhile to elucidate them. Melt reactions with (hydroxy alkanoate) polymers as well as grafting of ENR chains offer possibilities of manipulating blend morphologies. This might be favorable also for the electrical properties of the salt-polymer blend solutions. Keywords: epoxidized natural rubber, semicrystalline polymer, thermal properties, morphologies, ionic conductivity, melt reaction

* Corresponding author: Email: [email protected] and [email protected]

306

C. H. Chan, H. W. Kammer, L. H. Sim et al.

INTRODUCTION TO BLENDS OF EPOXIDIZED NATURAL RUBBER Natural rubber (NR) is an elastomer that is originally derived from colloidal suspension (latex) found in the sap of some plants. However, only Hevea brasiliensis tree is of commercial importance [1]. The purified form of natural rubber is called poly(isoprene), which can also be produced synthetically. One major difference between natural and synthetic rubber is the presence of proteineous substances in natural rubber. Synthetic dry rubber and latex products do not contain proteins. In what follows, the discussion is limited to NR [poly(isoprene)]. The rubbery molecular chains in NR are highly stereoregular conforming to a cis configuration at the carbon-carbon double bonds in the backbone chains. It has no highly polar substituent. Thus, its intermolecular attraction is largely limited to van der Waals forces. It appears as long and flexible chain; with weak intermolecular forces and occasional crosslinking. At room temperature, NR is amorphous under the quiescent condition, but it is highly crystallizable upon stretching. NR has attained great attention as biomaterial for displaying building tackiness, high strength and resilience, good tear resistance and excellent dynamic properties [2]. However, NR may not be able to compete with other specialty synthetic elastomers (e.g. butyls and nitriles) with regard to properties such as gas permeability and oil resistance. Besides, the performance of NR deteriorates drastically by sunlight, ozone and oxygen due to its high level of unsaturation [3]. Hydrohalogenation, halogenation, oxidation, epoxidation, ozonolysis, hydrogenation, carbine addition, cyclization etc. are the chemical modifications of NR that have been adopted conveniently in regulating the physical and mechanical properties [4]. Epoxidized natural rubber (ENR) is produced by epoxidation of NR where the double bond of the isoprene unit reacts with peracid and forms epoxy group during the epoxidation process (c.f. Figure 1). The peracid is produced in-situ by mixing formic acid (HCOOH) and hydrogen peroxide (H2O2). ENR retains many of the properties of NR with additional interesting properties, such as enhanced oil resistance, polarity, elasticity, abrasion resistance; and decreased gas permeability [5]. Epoxidation of NR have been done since 1922 [6], but commercial values and potential applications of ENR were only realized in the1980s. Epoxidation can be carried out in solution or latex form but only the latter is of commercial value. The properties of ENR can be regulated via the control of the epoxidation level on the macromolecular chain. The level of epoxidation can be varied by adding different concentrations of H2O2 during the epoxidation process [7]. Generally, ENR of any desired level of epoxidation from 1 to 60 mol% can be produced using this preparation method. The glass transition temperature (Tg) of ENR increases by 1 oC with an increase of every 1 mol% of epoxidation level. Hence, with increasing epoxidation level, the macromolecular chain of ENR turns from soft to stiff. In addition, the polarity of the ENR increases with higher epoxidation level. At high epoxidation level, ENR becomes more resistant to hydrocarbons but its resistance to polar solvents decreases. This provides a plausible chance to increase the interfacial adhesion or to enable reactive blending with another functionalized or polar polymer via regulation of epoxidation level of ENR. The increase in polarity of ENR results in changes in the adhesive characteristic as compared to that of the NR. Therefore, ENR can bond with polar polymers, e.g. poly(vinly chloride) (PVC). Higher epoxidation of ENR results in a substantial decrease in air permeability. The air permeability of ENR with 50

Blends of Epoxidized Natural Rubber and Thermoplastics

307

mol% of epoxidation level is similar to that of butyl rubber. Thus, it can be used in applications requiring low air permeability. The demand for new properties for ENR can be satisfied by changing the maromolecular architecture and/or be extended by blending with existing polymers. Polymer blends allow combining the useful properties of different parent polymers to be done through physical rather than chemical means. Blending of polymers turns out to be one of the fastest growing branches in polymer technology because of the ease in controlling the physical properties by compositional change. It is a quick and economical alternative as well as a popular industrial practice as compared to direct synthesis in producing specialized polymer systems. Physical properties which are not attainable from a single polymer can be achieved by this route. The understanding of physical and mechanical properties of blends has revealed new principles, refined earlier fundamental concepts and provided more opportunities for research and practical problem solving. From the point of view of phase behavior, one has to distinguish miscible and immiscible blends in terms of thermodynamics. Miscible blends are molecularly dispersed systems, while immiscible macromolecular system is the co-existence of phases of the neat constituents. Therefore, mechanical, transport and optical properties of immiscible blends depend critically on the morphology of the dispersion and the interfacial adhesion. Most polymers are immiscible from the thermodynamic standpoint since the entropic contribution to the free energy of mixing is negligible. Immiscible polymers may be described as “compatible” when two-phase blends do not exhibit gross symptoms of polymer segregation and may have good adhesion between the constituents. The mechanical properties of such blends are an average of those of the two components. An immiscible polymer blend is termed “incompatible” when it is heterogeneous both on microscopic and on macroscopic scales. ENR/thermoplastic blends form immiscible systems in most cases. For immiscible polymer blends, consisting of amorphous and semicrystalline constituents, one expects linear variation of properties with blend composition. However, morphology formation in those blends is strongly coined by thermal behavior of the components. Properties are ruled by composition of the blends and interaction between the constituents, which can be reflected in the thermal properties of the blends. Thermal properties influence development of morphologies that in turn strongly affect the mechanical behavior of the blends. Table 1 summarizes selected ENR/thermoplastic blends and reason(s) for the blending. The purpose for the blending in these cases points towards two directions: i) toughening the matrix of thermoplastic with dispersed phase of ENR and ii) increase the strength of ENR matrix with dispersed thermoplastic. The compatibility of the blends in some cases can be easily enhanced due to the polarity of ENR and polar or functionalized thermoplastic.

Figure 1. Epoxidation of NR to form ENR

308

C. H. Chan, H. W. Kammer, L. H. Sim et al. Table 1. ENR/thermoplastic blends.

Blends ENR/nylon 6

ENR/poly(ethyleneco-acrylic acid) (PEA) containing 6 wt% acrylic acid

ENR/ethylene vinyl acetate copolymer (EVA) ENR/PVC

ENR/chlorosulphona ted poly(ethylene) (Hypalon) ENR/poly(propylene) (PP)

ENR/PP 60/40 vulcanized with peroxides

Reason(s) for blending • Nylon 6 has poor impact resistance and is brittle at low temperature. • Nylon 6 is toughened by blending up to 40 wt% of ENR. Impact strength increases by six fold for toughened nylon 6. • PEA possesses high tensile strength, good adhesion to metals and good flow properties at relatively high temperature. ENR has low tensile strength. • Reactive blends show improvement on strength properties of ENR as well as its processing characteristic; flexibility and the impact resistance properties of PEA shall be improved. • EVA offers good ozone resistance, weather resistance and good mechanical properties. • Tensile properties of these compatible blends increase with the increase of EVA content. • PVC impart high tensile strength and good chemical resistance, whereas ENR may act as plasticizer to PVC. • These compatible blends are melt processable. With addition of ENR into the PVC matrix, the yielding and ductile behavior of the blends become prominent with increasing ENR content. • Compatibility of ENR and poly(ethylene) (PE) can be enhanced by blending ENR with this functionalized PE. ENR with higher epoxidation level shows better compatibilization with Hypalon. • PP has been used for medical applications due to good thermal and chemical properties. High-energy irradiation may be used to sterilize medical supplies. Irradiation tends to degrade PP. This deleterious effect of irradiation may be reduced by blending PP with ENR. • Optimum dosage of electron beam irradiation on the blends is expected to induce crosslinking in the rubber phase with acceptable level of degradation of PP. This crosslinking increases the tensile properties and decreases the elongation percentage of the blends. • Peroxides may result in crosslinking of ENR and degradation of PP. • Mechanical properties for the blends can be enhanced by careful selection of peroxide in order to have the proper balance between the degree of crosslinking of ENR and the degradation of PP-phase. • However, peroxide has the tendency to form smelly byproducts. (e.g. acetopheneone)

Ref. 8

9, 10

11

12, 13, 14

15

16

17


309

ENR with 50 mole% of epoxy content (average molecular weight, M = 7·105 g mol -1) was blended with poly(ethylene terephthalate) (PET) (M = 1.8·104 g mol -1) [18], poly(3hydroxybutyrate-co-3-hydroxyvalerate) with 12 mol% of hydroxyvalerate content (PHBV) (M = 2.4·105 g mol-1) [19] and poly(ethylene oxide) (PEO) (M = 3·105 g mol -1) [20]. These blends as well as other ENR/thermoplastics blends are discussed extensively in this chapter. Morphology becomes one of the crucial factors that affects the mechanical behavior of the immiscible blends with or even without cohesive force between the components. A strong mechanical resistance may arise from the interlocked array of the phases. The introduction of thermoplastic into the matrix of the elastomer may enhance the mechanical properties (e.g. tensile strength) of elastomer or on the other hand, incorporation of elastomer into the matrix of thermoplastic shall introduce elasticity to the blends. In other words, mechanical properties of these blends strongly depend on existing morphologies. Phase structure may be regulated by the compositions of the blends and the thermal procedures imposed on the blends. An understanding on the thermal properties of those systems is important for efficient adjustment of structure-property relationships. In the chapter, we would highlight three aspects: i) Melting and crystallization behavior for those blends and their influence on morphology development for thermoplastic elastomer blends. ii) The use of thermoplastic elastomer blends as polymer electrolytes when one adds salt. The distribution of salt between the blend components and its correlation to the ionic conductivity of the materials are elucidated. iii) Compatibilization of thermoplastic elastomer blends as well as grafting of thermoplastic onto the ENR backbone, which offer the possibilities to manipulate the blend morphologies are discussed subsequently. Furthermore, the compatibilized thermoplastic elastomer blends with added salt might be favorable for electrical properties. 360 350

Tg / K

340 280

260

240

220

200 0.0

0.2

0.4

0.6

0.8

1.0

Weight fraction of ENR in the blends Figure 2. Glass transition temperatures of ENR/thermoplastic blends [18-20] □ – PET, × - PHBV, ▲ – ENR, ○ – PEO

310


MISCIBLILITY OF ENR/THERMOPLASTIC BLENDS Thermal characterization of polymer blends is a well known method to determine their miscibility. The most commonly used experimental methods are the calorimetric determination of Tg and equilibrium melting temperature (Tmo). It is important to note that, in order to have comparable results on Tg for thermoplastic elastomer blends, the blends has to be heated up to sufficiently high temperature for extended time. Normally, the annealing temperature is above the Tg of the elastomer as well as the melting temperature (Tm) of the thermoplastic. This annealing procedure is aimed to erase all the thermal histories imposed on the blends during the preparatory steps, storage etc.. Afterwards, the blends have to be quenched to temperature below the Tgs of the parent polymers. Quantity Tg is normally determined during the reheating cycle. Nevertheless, it can also be extracted from the first heating cycle provided that all the samples are prepared in exactly the same manner during the preparatory steps. However, the former approach is preferable. ENR was blended with different crystallizable polymers by solution casting method. PET [18], PHBV [19] and PEO [20] were selected to be the second component to be blended with ENR. All of them exhibit appreciable crystallinity. These blends form immiscible systems. Quantities Tg measured in the second heating cycle for ENR, PET, PHBV and PEO are 251, 354, 273 and 219 K, respectively. Blends were prepared by solution casting method. The Tgs of the thermoplastic ENR blends are depicted in Figure 2. Two glass transition temperatures, corresponding to that of the neat constituents, were found for the binary blends of ENR with PET, PHBV as well as PEO. This reflects immiscibility of the constituents for the blends. The immiscibility of the ENR in blends with PET, PHBV as well as PEO also supported by calorimetric determination of quantity Tmo. Equilibrium melting temperature of a polymer is defined as the melting point of an assembly of large crystals, with negligible surface effects, in equilibrium with the polymer liquid. Low molecular substances, for example water exhibits sharp transition between solid and liquid phases at 273 K under standard condition. We may say, water has the Tmo = 273 K. In contrast to low-molecular substances, liquidation (melting) and solidification (crystallization) of polymers cannot be observed in equilibrium. This is because the crystallization is extremely low near and below the Tmo for the polymer due to the crystal nucleation is greatly inhibited at the proximity of Tmo. The rate of crystallization at temperature near to the Tmo for semicrystalline polymer is nucleation rather than diffusion controlled. Hence, crystallization of a polymer can only proceed in a temperature below Tmo. Quantity Tmo of a polymer can be determined experimentally by step-wise annealing procedure after Hoffman-Weeks [21]. Under this procedure, crystallization and melting of polymers proceed under non-equilibrium conditions but near to equilibrium (c.f. Figure 3). The sample is isothermally crystallized in a range of crystallization temperatures (Tc). Half time of crystallization (t0.5) for the polymer is determined. Subsequently, the sample is allowed to crystallize again at the same range of Tcs for equivalent period of time until complete crystallization and the corresponding Tms are obtained from the peak of the endotherms from the DSC traces. As a general practice, five half-times of crystallization (5t0.5) are adopted where crystallization of polymer should proceed in completion. Again, the use of 5t0.5 serves to impose equivalent thermal history to all the samples at all Tcs. Half time


311

of crystallization (t0.5) is the time taken for 50 % of the crystallinity of the crystallizable component to develop. Figure 3 shows a linear plot of the experimental Tm versus Tc of neat PEO, normally known as the Hoffman-Weeks plot. Extrapolation of the Hoffman-Weeks plot to intersect with the theoretical linear curve of Tm = Tc yields the value of Tmo. Analogous results are obtained for PET, PHBV and all others blends of ENR. In all cases, linear variation of the melting temperature with its respective crystallization temperature is noted with a range of undercooling (ΔTc = Tmo – Tc) in the proximity of Tmo. Table 2 shows constancy in equilibrium melting points over wide ranges of respective blend compositions. One expects constancy in Tmo for the crystallizable component with various blend compositions for immiscible thermoplastic elastomer blends. This is in agreement with the results shown in Figure 2. 355 o

Tm o

Tm

Tm / K

350

Deviation from the equilibrium

345

340

Tm = Tc 335 315

320

330

340

350

Tc / K

Figure 3. Hoffman-Weeks plot for neat PEO [20]

Table 2. Equilibrium melting temperature for crystallizable components in ENR blends after Hoffman-Weeks Weight fraction of ENR in the blends 0 0.3 0.5 0.6 0.7

PET 543 544 544 543

Tmo / K PHBV PEO 468 351 467 352 467 350 461 347 467

312


Quantities Tg measured in the first heating cycle by DSC for ENR (M = 5·103 g mol -1) and PVC (M = 1.4·105 g mol-1) are 253 and 348 K, respectively [22]. Blends were prepared by solution casting method. Figure 4 shows quantities Tg by DSC for the blends. PVC is a proton-donating polymer, which may interact with the oxirane ring of ENR and hence the blends are compatibilized via the higher polar interaction between the blend components. The existence of a single composition-dependent Tg that corresponds closely to Fox equation  1 W 1 − WENR   is observed. This result is in agreement with the single = ENR +  Tg, blend Tg, ENR Tg, PVC   composition-dependent Tg for the melt mixed ENR/PVC samples by dynamic mechanical analysis (DMA) [23] (MPVC = 4.5·104 to 5.2·104 g mol -1) and by DSC [24] (c.f. Figure 4). This suggests that the blends are compatible using different preparatory and analytical methods. 380

Tg / K

360

320

280

240 0.0

0.2

0.4

0.6

0.8

1.0

WENR

Figure 4. Glass transition temperatures of ENR/PVC blends. ■ – solution casting sample by DSC [22], ○ – melt mixed sample by DMA [23] and × - melt mixed sample by DSC [24]. The dashed and the dotted curves represent the Fox equation for the solution casting samples (by DSC) and melt mixed samples (by DMA), respectively

Table 3. Tensile properties of ENR/PVC blends by universal testing machine [22]. WENR 0 0.1 0.2 0.3 0.5 0.7 1 N mm-2 = 1 MPa

Tensile strength/ N mm-2 38.5 27.5 20.0 14.0 10.8 7.5

Yield strength/ N mm-2 31.3 25.4 17.2 11.9 8.1

Modulus/ N mm-2 2100 1300 1000 800 230 150

Elongation at break / % 6 20 30 50 75 100


313

PVC is expected to impart high tensile strength while ENR may contribute to the elasticity for the blends. Table 3 shows the tensile properties of the solution casting samples of ENR/PVC [22]. PVC is a brittle material with high tensile strength (38.5 N mm-2) and high modulus of elasticity (2100 N mm-2). With the addition of ENR into the matrix, the yielding and the ductile behavior of the blends become prominent with increasing ENR content.

CRYSTALLINITY OF CRYSTALLIZABLE COMPONENT IN ENR BLENDS For semicrystalline/amorphous polymer blends, the crystallization behavior of the crystalline component in the blends is expected to have close approximation as in the neat crystalline polymer. This is due to the crystallization of crystalline polymer in the blends which takes place within the domains of nearly neat crystalline component, is largely unaffected by the presence of the amorphous polymer. This in principle is true for the crystalline polymer either as the matrix or the dispersed phase in the blends. However, in some cases, experimental evidences indicate that although crystalline and amorphous phases are physically separated, and yet they can still exert a profound influence on each other. The crystallinity (X*) of semicrystalline polymer can be easily estimated from the calorimetric determination of melting enthalpy (ΔHm). The sample is isothermally crystallized at the preselected Tc for an equivalent period of time until complete crystallization and is heated to a temperature above its melting point. The corresponding ΔHm is extracted from the area under the endotherm from the heating cycle of the DSC trace. Crystallinity of semicrystalline polymer in semicrystalline/amorphous blends is defined by the ratio ∆H m

(1 − Wamorphous )⋅ ∆H ref

with ΔHref being the melting enthalpy of 100 % crystalline polymer (it is

also frequently termed as reference enthalpy) and W amorphous denotes mass fraction of amorphous polymer in the blends. The reference enthalpy is normally determined independently from other techniques. 1.0

X

*

0.8 0.6

0.4

0.2 0.0

0.2

0.4

0.6

0.8

1.0

weight fraction of ENR

Figure 5. Crystallinity of crystalline polymer in ENR blends as a function of weight fraction of ENR [18-20]; dotted curves represent constancy of the crystallinity in the blends × – PET crystallized at Tc= 224 oC; ▲- PHBV at Tc=112 oC and ○– PEO at Tc = 49 oC. ● – PEO during the first heating cycle [28]

314


The reference enthalpies were reported to be 188.3, 109 and 117.6 J g-1 for PEO [25], PHBV [26] and PET [27], respectively. Figure 5 shows crystallinity as a function of blend composition. It reveals that the amorphous component does not influence the crystallinity of PET [18], PHBV [19] and PEO [20] in the blends over a wide range of ENR content. In other words, crystallinity of the crystallizable constituents stays constant in the blends to a good approximation. Only at high ENR content, some scattering of crystallinity occurs which may be due to phase inversion (c.f. Figures 7 and 8). In another study by Noor and co-workers (2009), the crystallinity of PEO (M = 6·106 g mol-1) in the solution casting blends characterized by DSC during the first heating cycle stays constant at 0.89 after addition of 20 wt% of ENR to PEO [28].

KINETICS OF ISOTHERMAL CRYSTALLIZATION OF SEMICRYSTALLINE POLYMER IN ENR BLENDS Immiscible semicrystalline/amorphous blends are discussed in the following sections. In general, when the crystallizable component forms the matrix phase in the blends where the phases are physically separated, the crystallization kinetics for the semiscrystalline polymer in the blends will not differ that much from the neat semicrystalline polymer. In other words, the general principles governing the crystallization behavior of the neat semicrystalline polymer, in most cases, remain valid for immiscible blends in which the crystallizable component forms the continuous phase. On the other hand, the crystallization behavior for the dispersed phase of semiscrystalline polymer in the blends, the discrete droplets in an amorphous matrix can be dramatically affected as compared to that of the neat semiscrystalline polymer. The overall rate of isothermal crystallization can be monitored by thermal analysis through the evolution of heat of crystallization by DSC. The sample is isothermally crystallized at the preselected Tc until complete crystallization. The half time of crystallization (t0.5) for the polymer is estimated from the area of the exotherm at Tc = const, where it is the time taken for 50 % of the crystallinity of the crystallizable component to develop. The rate of crystallization of crystallizable component in ENR blends can be easily characterized by the experimentally determined reciprocal half time [(t0.5)-1]. Table 3 shows the range of isothermal crystallization temperature and undercoolings (ΔTc) for the semicrystalline polymers. Quantity ΔTc = Tmo - Tc. It is interesting to note that PHBV needs higher undercoolings to get an appreciable rate of crystallization. In other words, only when the crystallization temperature PHBV is far away from equilibrium temperature, the necessary driving force for crystallization exists and it crystallizes with a sufficiently high rate. The lowest undercoolings for an acceptable rate of crystallization is noted for PEO. PHBV is a thermoplastic polyester produced by bacterial fermentation. Unlike conventional thermoplastics, there is no catalyst residue in PHBV. Self-seeding nucleation is observed in a lot of studies for neat PHBV. Hence, higher undercoolings to drive the crystallization process is needed. Whereas for synthetic thermoplastic such as PET and PEO, impurities such as catalyst are rather difficult to be removed completely even the neat polymers are purified. The tiny amount of impurities may lead to heterogeneous nucleation and hence lower undercoolings may be sufficient to drive the crystallization process.


315

2

-1

(t0.5) / min

-1

1

0.1

0.02 20

60 80

40

90

o

(Tm - Tc) / K

Figure 6. Reciprocal half-time of isothermal crystallization versus undercoolings ■ – PET, ▲ - PHBV, and ● – PEO. Opened markers correspond to the respective 50/50 ENR blends

Table 3. Range of isothermal crystallization temperatures and undercoolings (ΔTc) for the semicrystalline polymers Polymer PET PHBV PEO

Range of Tc / oC 214 - 228 105 - 112 44 - 54

ΔTc / K 56 – 42 90 – 83 34 – 24

The rates of crystallization [(t0.5)-1] of neat PET, PHBV and PEO as well as their corresponding 50/50 ENR blends decrease exponentially with decreasing undercoolings or in other words with ascending Tc as shown in Figure 6. It means that the half times of crystallization of neat semicrystalline polymers or the crystallizable component in ENR blends increase with increasing Tc. At higher Tc (lower ∆Tc), fewer nuclei are available to nucleate the molten phase of the semicrystalline polymer, thus leading to the formation of fewer but larger spherulites. Besides, the rates of crystallization decrease drastically and cause the formation of more perfect spherulites which melt at higher Tm (c.f. Figure 3). To compare the rates at ∆Tc = const, quantities (t0.5)-1 of PET and PEO do not change markedly with blend composition at least in the range where crystallizable component is in excess [mass fraction of ENR (WENR) ≤ 0.4]. Decreasing rates of crystallization of PET and PEO with ascending content of ENR (WENR > 0.4) are observed, which may be due to the influence of the borderlines that slightly reduces the rate. Decreasing rates of crystallization of PHBV with ascending ENR content is observed [19]. This may be due to the intraspherulitic entrapment of the dispersed phase of ENR by the growing PHBV spherulite leading to the slowing down of the rate of crystallization (c.f. Figure 7 c).

316


Figure 7. Morphology of (a) neat PHBV after annealing at Ta = 234 oC for 1 min and then isothermal crystallized at Tc = 50 oC for 1 day, micrograph was captured at temperature of the POM (TPOM) = 50 o C; (b) ENR/PHBV 50/50 at Ta = 175 oC for 1 min and TPOM = 175 oC; ENR/PHBV blends at Ta = 175 o C for 1 min and then at Tc = 120 oC for 60 min (TPOM = 120 oC): (c) 10/90, (d) 40/60 and (e) 70/30

MORPHOLOGIES OF THERMOPLASTICS/ENR BLENDS Spherulites form during the crystallization of semicrytsalline polymer. Spherulites are spherical aggregates of lamellae. Polarizing optical microscopy is one of the optical techniques commonly used to monitor the growth of a spherulite as a function of time as well as the morphologies of blends comprising at least one crystallizable component. The presence of the amorphous component for immiscible semicrystalline/amorphous blends can interrupt the normal crystallization process, although the two phases are separated by interphase of various thicknesses. This results in changes in the crystallization kinetics, the growth rate of spherulite and the spherulitic morphology as compared to that of the neat semicrystalline polymer. In general, nucleation density attributed to the nucleating activity at the polymer/polymer interphase plays an important role in regulating the final morphology of immiscible blends.


317

The growth rate of spherulite and the diffusion rate of the amorphous component are important parameters, which determine whether the amorphous materials are segregated in large domains in the inter-spherulitic regions or occulated as small domains within the spherulites. The composition of the blends is another important factor that strongly influences the final morphologies. High concentration of amorphous component in the blends can markedly disturb the growth of spherulites leading to the formation of deformed spherulites or coarse texture with open structures and irregular outlines. In this chapter, the discussions on the techniques adopted to characterize the morphology of the blends are limited to polarizing optical microscopy (POM) and scanning electron microscopy (SEM). We start with the inspection under POM. In general, when neat PHBV is annealed around 10 – 20 oC above its apparent melting temperature (Tm) for a sufficient length of time, it forms fibrillar fine texture of spherulites after isothermal crystallization at Tc around 120 oC (figure not shown) [19]. However, when the annealing temperature (Ta) is 234 o C and annealing time (ta) is 1 min, the growth rate of the spherulites decreases markedly and crystallization of spherulites can be observed at Tc = 50 oC. Ring-banded PHBV spherulites can be detected in this case as depicted in Figure 7 (a). PHBV is prone to random chain scission at temperature slightly above its melting temperature (c.f. Figure 16). Shorter PHBV chains after the sample was heated up at sufficiently high Ta may lead to the formation of the ring-banded structure. Two phases can be clearly seen in the molten stage at Ta =175 oC for the immiscible ENR/PHBV blends in Figure 7 b). However, the crystallization of PHBV at Tc = 120 oC as the matrix in the blends is not without any disruption from ENR though they are physically separated. The growing front of the PHBV spherulites can not reject ENR efficiently, thus results in fine intrasperulitic dispersion of ENR in PHBV even with ENR content as low as 10 wt% in the blends [c.f. Figure 7 (c)]. This is the attribution of decrease in rate of crystallization of PHBV in the immiscible blends with ascending ENR content. Fine dispersion of ENR coalesces to bigger islands at 40 wt% of ENR [c.f. Figure 7 (d)] which disperse intra- as well as interspherulitically. Phase inversion can be seen at mass fraction of ENR ≥ 60wt% as indicated in Figure 7 (e). In the case of PEO and PET in the blends with low content of ENR, the dispersed phase of ENR diffuses fast enough away from the growing front of the spherulites and be pushed along until complete crystallization [e.f. Figures 8 (a) and (c)]. The dispersed phase is then found mainly in the intersphrulitic regions. Hence, the rate of crystallization of PEO and PET in the blends does not change markedly with blend composition at least in the range where crystallizable component is in excess. Phase inversion occurs in the ENR/PET 50/50 blend. The texture of PEO spherulites becomes increasingly irregular and coarser at relatively high ENR content. At 60 wt% of ENR [c.f. Figure 8 (b)], the slowly growing front of PEO spherulites cannot reject the ENR phase, but by-passes it resulting in the formation of a quite irregular co-continuous morphology. For solution casting ENR/PVC blends [22], the fractured surface of the tensile specimen was sputter coated with gold immediately after testing and the micrograph was taken using SEM. The tensile fractured surface of PVC shows that the sample undergoes a brittle mode of failure where undeformed surface and closely placed line pattern can be observed. When ENR is added to PVC, a shift from the brittle fractured surface to a ductile fractured surface (parabolic and wavy fractured front) is noted. This means that the PVC matrix becomes more flexible and the ductile nature of the blends increases with ENR content. The fractorgraph of

318


the blends are homogenous. This again suggests the compatibility of the components in the blends

(a)

(b)

(c)

Figure 8. Morphology of thermoplastics in blends with ENR, ENR/PEO blends at Ta = 80 oC for 1 min and then at Tc = 49 oC for 60 min (TPOM = 49 oC): (a) 40/60 and (b) 60/40; (c) ENR/PET 30/70 at Ta = 280 oC for 1 min and then at Tc = 224 oC for 60 min (TPOM = 224 oC)

For the melt blended ENR/PVC 30/70 and 50/50 blends [23], the cryogenically fractured blends were etched with concentrated nitric acid at 50 oC for 50 hr to remove the ENR phase. The phase structure of the washed and dried samples was made using SEM. Macro level domain structure of ENR in the matrix of PVC is not seen in the micrographs. However, micro level domains of ENR can be observed for these compatible blends.

BLENDS OF ENR AND THERMOPLASTICS WITH ADDITION OF LITHIUM SALT Polymer electrolytes formed by polymer-salt solutions attracted a great deal of scientific interest since 1970s. It is related to the hope of applying these systems in new generations of highly efficient batteries. Therefore, superior properties of polymer electrolytes such as high ionic conductivity, wide electrochemical window, good dimensional stability, sufficient thermal stability, good mechanical stability and high electrochemical stability are expected. Preferably used constituents are PEO and lithium (Li) salts. However, solid solutions of poly(vinyl acetate) (PVAc), poly(vinylidene fluoride) etc. added with Li salts have also been proven to be promising candidates as solid polymer electrolytes. There are several identified factors governing the coordination between the salt and the polymer. High concentration and high polarizability of sequential polar solvating groups along the polymer chains; and low lattice energy of the salt as well as the polymers may


319

enhance the solubility of the salt in the polymer host. Efforts for the improvements of polymer-salt solutions pointed preferably towards two closely related directions, enhancement of both carrier density and mobility. Most of the binary systems of polymer and salt do not exhibit sufficient conductivity (∼10-4 - 10-3 S cm-1) for applications. Hence, not only homopolymers have been used as the polymer host, but also polymer blends, graft copolymer, block copolymer, polymer composite etc. for the efforts to enhance the conductivity of solid polymer electrolytes. Thermoplastic elastomer blends as the polymer host will be presented in this section. We note here that isotropic dispersion of Li salts in different phases of the blends may not be the desired morphology. Instead, hierarchical morphology with percolation pathways that offers maximal ionic conductivity shall be the interesting focus here. However, not much effort has been devoted to elucidate this problem systematically except some hints are given in ref. [20]. We start the discussion of solid polymer electrolytes for ENR/PEO blends added with Li salts and afterwards followed by ENR/PVC blends added with Li salts. PEO and other polyethers can form ionic complexes with metal salts of low lattice energy. Li salts are the most studied salts because of their excellent solubility in PEO. The presence of the lone pair electron in the ether oxygen of PEO enables it to coordinate with high concentration of Li salts [29]. PEO, a semicrystalline polymer, shows low conductivity (∼10-10 S cm-1) at room temperature [30]. For solid polymer electrolyte, it is widely accepted that the crystalline phases do not contribute to conductivity and the ion dynamics is essentially governed by the segmental motion of the amorphous phases in PEO. Mixtures of PEO and the Li salt undergo liquid-solid demixing while cooling down from temperature above the melting temperature of PEO. At low content of Li salt in the system, the mixture decays into pure crystalline PEO and an amorphous mixture of the PEO + Li salt. As mentioned earlier, blends of ENR/PEO are immiscible. Hence, it is expected that the properties of PEO that are governing the ionic conductivity as well as the elasticity of ENR in the blends can be retained, respectively. The aim to insert polymer blends as polymer matrix in polymer electrolytes with addition of Li salt is for improving both the mechanical properties as well as the ionic conductivity of the principle polymer (PEO in this case). Both PEO and ENR possess oxygen in their respective chemical structures, which may be able to coordinate with the Li salt added as demonstrated in Figure 9. The question now is: in what way does the salt distribute between the blend components?

Figure 9. Schematic representations of the possible coordination formed between the salt and the polymers. Li+ ion coordinates with (a) ether oxygen of PEO and (b) oxirane group of ENR

320


Blends of ENR/PEO added with LiClO4 were prepared by solution casting method [20]. Dried polymer films were analyzed after the samples were dried in the oven at 50 oC for 24 hr and further dried under vacuum at 50 oC for 48 hr. Figure 10 shows the glass transition temperatures (Tg) after the DSC analysis of ENR/PEO blends as a function of weight fraction of ENR with weight fraction of LiClO4 (Wsalt) at 0.107. Quantities Tg of neat ENR and neat PEO are observed at 254 and 219 K, respectively. Coordination of Li salt to the ether oxygen in PEO and oxirane group in ENR leads to the increase in the Tg of the two polymers indicating the formation of stiffer polymer chains. Hence, we propose, the change in Tg (ΔT g = Tg,polymer+salt –Tg,polymer) is a function of dissolved salt concentration. This may give some hints about the distribution of the salt in the two polymer phases of the blends where ΔTg,PEO is approximately ∝ mass of the salt in the PEO phase. Similar approximation is applied to ΔTg,ENR. Two Tgs, corresponding to those of the neat constituents, are found in blends with and without salt. Glass transition temperature studies reveal immiscibility of the polymers for the blends with and without salt over the entire composition range. Quantity of ΔTg as reported in Table 4 as well as in Figure 10 for neat PEO and ENR after addition of Li salt at Wsalt = 0.107 is approximately at 4 – 5 K. It means that Li salt dissolves in both neat PEO and neat ENR. In the blends, situation may change because Li salt may have the choice to dissolve preferably in one of the components or in both the components equivalently. The distribution coefficient (KD) can be defined as below: KD =

mass of Li in PEO ∆Tg,PEO ≈ mass of Li in ENR ∆Tg,ENR

(1)

270

∆Tg,ENR 260

Tg / K

250

240

230

∆Tg,PEO

220 ENR dispersed phase

PEO dispersed phase

210 0.0

0.2

0.4

0.6

0.8

1.0


Figure 10. Glass transition temperatures of PEO and ENR in ENR/PEO blends and blends with Li salt at weight fraction = 0.107. ● – PEO and ■ – ENR. Opened markers correspond to the respective saltcontaining blends. Solid arrows represent the shift of the glass transition temperature of respective component in 40/60 blends after addition of Li salt

321

Blends of Epoxidized Natural Rubber and Thermoplastics Table 4. The change of glass transition temperature for the polymer in blends of ENR/PEO after addition of Li salt at Wsalt = 0.107 Weight fraction of ENR in the blends 0 0.4 0.6 1

ΔTg,PEO / K 5 20 15

ΔTg,ENR / K

KD

5 4 4

4 3.8

When quantity KD > 1, it points towards preferential solubility of salt in PEO instead of ENR. Now, let us look at the middle range of the composition for the blends (WENR = 0.4 to 0.6). Quantities ΔTg,PEO ≥ 15 K, whereas ΔTg,ENR ≈ 4 K are noted and thus quantity KD > 1. This suggests that the dissolution of salt favors PEO slightly more than ENR in the middle range of the composition at a certain range of salt concentration. PEO is the component which governs the ionic conductivity of the blends after the addition of Li salt. Hence, the morphology corresponds to the formation of percolating conductivity pathways is extraordinarily important in this case to enhance the ionic conductivity. Morphologies of dispersed phase of ENR or co-continuous for the ENR/PEO blends with the addition of Li salt are expected to display higher ionic conductivity due to the existence of the percolating conductivity pathways. On the other hand, the percolating conductivity pathways may be interrupted with the dispersed phase of PEO in the blends resulting in a dramatic decline in conductivity. Figure 11 presents as-prepared morphologies under POM at 30 oC for solution casting samples after drying under vacuum. Dispersed phase of ENR in the matrix of PEO can be observed for ENR/PEO 10/90 to 50/50. For the ENR/PEO 40/60 blend, the dispersion of less quantity but bigger size of ENR droplets in the PEO matrix are detected. Whereas for the ENR/PEO 50/50 blend, high density of smaller size ENR droplets are observed in the PEO matrix. Phase inversion or dispersed phase of PEO in the matrix of ENR is noticed when the ENR content is more than 60 wt%. The amorphous regions of PEO and ENR cannot be distinguished under the inspection of POM for blends added with Li salt. Hence, the morphologies of as-prepared samples for blends at 30 oC are assumed to be closely related to the amorphous regions in the blends with salt. As mentioned before, for ionic conductivity, the crystalline regions of PEO are relatively unimportant. The continuity of the amorphous regions of PEO (the percolation pathways) in blends with salt is vital for the enhancement of ionic conductivity. The influence of morphology on the ionic conductivity at 30 oC for the asprepared sample is discussed. As shown in Figure 12, the ionic conductivities (σ) of PEO and ENR at 30 oC measured using ac-impedance spectroscopy are approximately 10 -10 and 10-11 S cm-1, respectively. LiClO4 at Wsalt = 0.107 was added to blends as well as to the neat polymers. Quantities σ for salt-containing PEO and ENR increase to 10-6 (increase by 4 orders of magnitude as compared to neat polymer) and 10-9 (3 orders of magnitude) S cm-1, respectively. Enhancement in the ionic conductivities of polymer-salt mixtures are to be directed towards the increasing of the carrier density (n) or the ion mobility (µ) or both (since σ = n·q·µ). The carrier density is constant in this case. Hence, the mobility of the charge carrier in PEO is relatively higher than in ENR.

322


(a)

(b)

Figure 11. Morphology under POM of ENR/PEO blends for as-prepared sample at 30 oC and TPOM = 30 o C. ENR/PEO blends (a) 40/60 and (b) 50/50

In general, the ionic conductivity of the salt-containing blends is higher for compositions with dispersed phase of ENR as compared to that of the compositions with dispersed phase of PEO as depicted in Figure 12. Salt-containing 40/60 blends exhibits the highest conductivity with 1.6 ·10-5 S cm-1. The results of ionic conductivity suggest the availability of percolating conductivity pathways that are formed by amorphous PEO and dissolved Li salt for the compositions with dispersed phase of ENR. These pathways exist in neat PEO, but in this case, the Li salt is less concentrated in neat PEO as compared to that of the 40/60 blends. This is because Li salt is dominantly dissolved in PEO but only to a negligible or minor extent in ENR as derived from Tg data. The percolating pathways in 50/50 blend become narrower and beyond 40 wt% of ENR, a dramatic decline in conductivity is observed. Above 40 wt% of ENR, the morphology is transferred progressively into dispersion of PEO in ENR. It means that the percolating conductivity pathways are interrupted. Detailed discussion of the distribution of salt in different phases for the blends was reported in ref. [20]. 10-4 10-5

σ / S cm-1

10-6 10-7 10-8 10-9 10-10 ENR dispersed phase

10-11 0.0

0.2

0.4

PEO dispersed phase

0.6

0.8

1.0


Figure 12. Ionic conductivity as a function of weight fraction of ENR at 30 oC. • - ENR/PEO blends and ○ - ENR/PEO blends with Li salt at Wsalt = 0.107


323

Salt concentration dependence of ionic conductivity for the ENR/PEO blends added with LiClO4 [20] and LiCF3SO3 [28, 31-33], ENR/PVC blends added with LiClO4 [34, 35] is depicted in Figures 13 and 14. All systems were prepared by solution casting method. They were initially dried either at room temperature or in an oven at approximately 50 oC and later dried under vacuum around 50 oC for one or two day(s). We note here that, comparable results for the different systems can be obtained provided similar preparatory steps were followed strictly for all the systems. However, the systems mentioned above were not prepared in exactly the same way. Hence, the solubility of the salts in different phases of the above systems can only be discussed qualitatively. 10-3

10-3

10-4

10-5

10-5

10-6

10-6

-1

10-4

σ / S cm

σ / S cm

-1

(a)

10-7 10-8

10-7 10-8

10-9

10-9

10-10

10-10

10-11 0.01

10-11 0.01

0.1

Weight fraction of LiClO4

(b)

0.1

Weight fraction of LiCF3SO3

Figure 13. Ionic conductivity for the blends of ENR/PEO as a function of weight fraction of Li salt at room temperature (∼30 oC). : • - PEO [20, 33], ■ – ENR [20, 32]; ENR/PEO blends: Δ – 20/80 [28], × 30/70 [31] and ○ 40/60 [20]. Dotted curves represent the regression curves after Eq. (2). 10-5

10-6

σ / S cm-1

10-7

10-8

10-9

10-10

10-11 0.01

0.1

Weight fraction of LiClO4

Figure 14. Ionic conductivity for the blends of ENR/PVC as a function of weight fraction of Li salt at room temperature (∼30 oC). : • - PVC [35], ■ – ENR [20], □ – ENR/PVC 70/30 blend [34]. Dotted curves represent the regression curves after Eq. (2).

324


Quantities Tg for ENR, PEO and PVC are approximately at 251, 219 and 353 K. After the addition of the salt, for an example, Wsalt = 0.107, the ΔTg is normally around 5 – 20 K. The conductivity measurement was carried out at around 303 K. If the Li salt dissolves in PEO and ENR, the Li salt is still relatively mobile to conduct electricity at 303 K, whereas the Li salt is frozen in PVC at 303 K because PVC is in glassy state. Hence, PVC may not be a good candidate for solid polymer electrolyte if the salt preferably dissolves in PVC for the blends. The salt which dissolves in PVC may not be able to contribute to the enhancement of conductivity. In solid polymer electrolyte of a salt in a polymer, salt molecules, especially cations, are solvated by chain segments. Therefore, properties of these systems are strongly governed by the interactions between the salt molecules and the chain segments. We adopt here the approach of investigating the concentration dependence of ionic conductivity presented in Ref. [20]. In the range of low salt concentration, the conductivity is given by

σ = const (Wsalt )x

(2)

Exponent x describes the extent of correlations between salt molecules and segments and it has to be determined experimentally. The exponent x for various systems is shown in Table 5. The exponent x for PEO is higher than ENR. Hence, for the blends of ENR/PEO added with Li salts, the solubility of Li salt is relatively higher in the PEO phase. The ionic conductivity for the salt-containing blends is higher for compositions with dispersed phase of ENR as compared to compositions with dispersed phase of PEO added either with LiClO4 or LiCF3SO3. We observe higher value of exponent x for 20/80, 30/70, 40/60 blends (added with LiClO4 or LiCF3SO3) with ENR dispersed phase as compared to that of the neat PEO which governs the conductivity in this case. For the blends of ENR/PVC, not much information was provided for the morphology, quantity Tg and other related parameters for the blends from the literatures [34, 35]. The solubility of the salt in different phases of the blends cannot be discussed convincingly. Nevertheless, the conductivity of the salt-containing blend with PVC dispersed phase is still higher than the neat ENR added with salt. In this case, ENR phase governs the conductivity.


325

Table 5. Exponent x after Eq. (2) for different ENR-based blends added with Li salts Salt LiClO4

Polymer x ENR 1.68 PEO 2.93 ENR/PEO 40/60 2.95 LiCF3SO3 ENR 2.24 PEO 2.89 ENR/PEO 20/80 2.92 ENR/PEO 30/70 3.15 LiClO4 ENR 1.68 PVC 1.55 ENR/PVC 70/30 2.32 These thermoplastic elastomer systems with suitable morphologies may become potential candidates for the application as solid polymer electrolytes. Dispersed phase of ENR in PEO for the salt-containing blends (hierarchical morphology with percolation pathways) shall be attended. Due to the preferential solubility of the Li salt in PEO as compared to that of ENR, the concentration of Li salt in PEO becomes more concentrated in blends as compared to the neat PEO. The conductivity of the salt-containing blends shall be higher than that of the saltcontaining PEO with the addition of an equivalent amount of salt. On top of this, the insertion of dispersed phase of elastomer into the thermoplastic may form “high-impact” thermoplastic which is analogue to the case of high-impact poly(styrene) [dispersion of poly(butadiene) in poly(styrene)] provided that suitable compatibilization between the components can be established. In the next section, the possibility in stabilizing the immiscible thermoplastic/ENR blends is highlighted via different compatibilization routes.

COMPATIBILIZATION OF ENR/THERMOPLASTIC BLENDS Most of the polymer blends consist of thermodynamically immiscible components. Immiscible polymer blends are often preferred over the miscible types because they may combine some of the favorable characteristics of each blend component. However, the properties of an immiscible blend cannot be easily predicted from the properties of the individual component as in the case of miscible blends. In some cases, immiscible blends have inferior mechanical properties relative to their individual components due to weak interfacial adhesion. These deficiencies encountered in immiscible polymer blends can be approached via a compatibilization strategy. This strategy may facilitate the controlling and the stabilizing of the phase morphology, thus, improving the interfacial adhesion between the phases of the immiscible blend. A compatibilized blend is characterized by a finely dispersed phase, good adhesion between the blend phases and strong resistance to phase coalescence. Three methods have been applied to stabilize the morphologies of immiscible polymer blends. The first method is cold quenching of non-equilibrium morphologies produced during the mixing of blend components at sufficiently high temperatures. The major difficulties here are to achieve and to stabilize sub-micrometer structures and relaxation phenomenon can often be observed for the non-equilibrium morphologies.

326


Secondly, adding compatibilizing agents (or they are also known as compatibilizers, interfacial agents, adhesion promoters and emulsifiers) may help to control the coalescence of the dispersed phase and improve the adhesion between the two phases (the dispersed and the matrix phases). For example, compatibilization can be achieved by the addition of selfassembling block or graft copolymers with structuring tendencies (e.g. linear or graft macromolecules consisting of at least two monomer sequences covalently linked together) to the immiscible blends. The copolymer segments can be identical (but are not necessarily) to the blend components. Poly(A-co-B) copolymers, with carefully controlled molecular weight and composition, are expected to self-assemble when mixed with the poly(A)/poly(B) blends in forming thermodynamically stable structures at the micro- to nanometer scale under favorable conditions. When the copolymer segment(s) is(are) not identical to the blend component(s), then, the segment(s) is(are) miscible or at least has sufficient chemical interaction with the blend component(s). Poly(A-co-C) can compatibilize poly(A)/poly(B) immiscible blend, provided that segment C is miscible or exhibits sufficient interactions with poly(B). Poly(A)/poly(B) blend can be compatibilized using poly(C-co-D) provided that C and D segments are miscible or have sufficient chemical interactions with poly(A) and poly(B), respectively. The major challenges here are the added copolymers which are more often preferentially dispersed as micelles in the homopolymer phases and their transport to interfaces, very often, is an issue. Besides, the even more important problem here is that suitable copolymers which act as compatibilizing agent cannot be prepared for many pairs of poly(A) and poly(B). The third method is through the replacement of premade poly(A-co-B) copolymers by interchain copolymer formed in-situ at the interface of different phases in the blends of poly(A)/poly(B). A heterogeneous reaction takes place across a melt-phase boundary in the blends. Thus, interfacial adhesion is promoted by lowering the interfacial tension between the two polymer phases. Reactive blending of two functionalized homopolymers can generate copolymers. The copolymers may improve the thermodynamic stability of the morphologies.

Figure 14. Schematic representations of the possible reaction between the functionalized polymers. Pairs of reactive groups : (a) epoxy/aromatic or aliphatic ―NH2 and (b) epoxy/―COOH


327

Figure 15. Proposed degradation mechanism for thermal degradation of HB unit in PHBV via random scission process

In this chapter, studies on the compatibilization of immiscible blends for ENR/thermoplastic via reactive blending are highlighted. In the reactive blending process, two polymers with complementary reactive groups are mixed together and react at high temperature in the molten state. Copolymers form in-situ during the thermal and/or catalyst induced melt reaction. The most commonly used pairs of reactive groups in reactive compatibilization of ENR/thermoplastic are epoxy/―COOH and, epoxy/aromatic or aliphatic ―NH2 as shown in Figure 14. We start the discussion with reactive blends of ENR/poly(hydroxyalkanotes) (PHA). The longer chains of ENR (the backbone) have reactive groups (the epoxy group) along their length, whereas the shorter polyester chains (the grafts, e.g. PHA in this case) are only reactive at terminal ends after random chain scission process. This will result in grafted copolymers consist of different lengths of side chains [36, 37]. At moderately high temperatures, it is widely accepted that poly(hydroxyalkanotes) (PHA) decomposes through a random scission process. Figure 15 shows the proposed degradation mechanism for the thermal degradation of hydroxybutyrate (HB) unit in PHBV via random scission process. In this study, at temperature above the melting points of PHBV (M = 2.38·105 g mol -1) and PHB (M = 2.68·105 g mol-1), thermal decomposition of PHBV and PHB via random scission process may take place leading to shorter chains with terminal carboxyl groups. Figure 16 demonstrates the reduction in weight-average molecular mass (Mw) estimated by using gel permeation chromatograph (GPC) for PHBV and PHB as a function of annealing temperature (Ta). The PHBV and PHB were heated at Ta for only 1 min. PHBV and PHB degrade relatively fast to shorter chains at higher annealing temperature is noted. The terminal carboxyl groups for PHBV and PHB after random scission process may trigger chemical reactions with the reactive sites in ENR at relatively high temperature as shown in Figure 17. As expected, the physical properties of theses blends are strongly controlled by existing morphologies and the extent of interfacial adhesion. The presence of the chemical bonds may stabilize morphologies and improve the interfacial adhesion.

328

C. H. Chan, H. W. Kammer, L. H. Sim et al. 280x103 240x103

Mw / g mol

-1

200x103 160x103 120x103 80x103 40x103 0 160

180

200

220

240

o

Annealing temperature / C

Figure 16. Weight average molecular mass for PHA after annealing at relatively high temperature for 1 min [38]. ● – PHBV and □ – PHB

Figure 17. Proposed reaction mechanism for melt reaction of ENR and PHBV or PHB

Figure 18 shows the DSC traces that reflect the melt reaction of solution casting samples for ENR/PHBV 50/50 blends at different Tas. At temperature above around 220 oC, one observes in the blends an exothermic reaction enthalpy. Half time of the melt reaction (t0.5,rxn) for the blends is estimated from the area of the exotherm at Ta = const, where it is the time taken for 50 % of the material reacted. The rate of melt reaction for ENR blends can be easily characterized by the experimentally determined reciprocal half time [(t0.5,rxn)-1]. No exotherm at temperature above around 220 oC for neat ENR and PHBV can be observed. Melt reaction temperature range for the solution casting samples of ENR/PHBV and ENR/PHB blends are 220 – 234 oC and 184 – 199 oC, respectively.


329

Figure 18. DSC traces for ENR/PHBV 50/50 when the samples were isothermally annealed at various Tas until complete melt reaction. Ta: (a) 220, (b) 230 and (c) 240 oC. Curves are displaced for better identification

The rate of reaction [(t0.5,rxn) -1] and the enthalpy of the reaction (ΔHrxn) for both of the ENR/PHBV and ENR/PHB blends increases with annealing temperature for different blend compositions as depicted in Figure 19. The molecular masses of the monomer units of ENR, PHBV and PHB are assumed to be the same. Hence, weight fraction (W) ≈ volume fraction ( φ ). The dependencies of quantity ΔHrxn on weight fraction of ENR at different Tas are shown in Figure 20. The dependence of quantity ΔHrxn on blend composition at Ta = const might be approximated by a function type presented in Eq. (3). The results for the exponent α are tabulated in Table 6.

∆H rxn ∝ (φENR )

α 3

(1 − φENR )

(3)

330

C. H. Chan, H. W. Kammer, L. H. Sim et al. 0.30

240

(a)

(b)

0.25

∆Hrxn / J g-1

(t0.5,rxn)-1 / min-1

200 0.20

0.15

160

120 0.10

0.05

80 180

190

200 220

230

240

190

180

Annealing temperature / oC

230

200 220

240

o

Annealing temperature / C

Figure 19. (a) Rate of reaction and (b) enthalpy of melt reaction as a function of annealing temperature. Circle: ENR/PHBV blends [36] and square: ENR/PHB blends [37]; opened marker: 50/50 and solid markers: 70/30.

250

250

(a)

(b) 200

∆Hrxn / J g-1

∆Hrxn / J g-1

200

150

100

50

150

100

50

0

0 0.0

0.2

0.4

0.6


0.8

1.0

0.0

0.2

0.4

0.6

0.8

1.0


Figure 20. Enthalpy of melt reaction as a function of weight fraction of ENR for (a) ENR/PHBV blends [36] and (b) ENR/PHB blends [37]. Annealing temperature: ● – 220, ○ – 234, ▲ – 184 and Δ – 199 oC. Dashed curves represent regression functions for solid markers after Eq. (3).

Table 6. Exponent α after Eq. (3) for ENR in blends with PHBV and PHB. Blends ENR/PHBV ENR/PHB

Annealing temperature / oC 220 234 184 199

α 0.75 1.3 2.85 2.76

331


Figure 21 shows the rate of reaction as a function of volume fraction of PHBV and PHB. The rate decreases according to the power law for the blend compositions where PHBV or PHB forms the dispersed phase as shown in Eq. (4):

(t0.5,rxn )−1 ∝ (φPHBV or PHB )β

(4)

Under perfect conditions, where PHBV or PHB forms monodispersed spheres in the

matrix of ENR, (t0.5, rxn )−1 ∝ (φPHBV or PHB )−1 3 can be approximated [36]. The experimental results for the exponent β are tabulated in Table 7. The results reveal that PHBV and PHB do not form uniform spheres in the ENR matrix. Elongated island of PHBV in the matrix of ENR (c.f. Figure 7 e) is expected. However, with increasing annealing temperature, the dispersion of PHBV or PHB in the ENR matrix becomes more homogenous (β → -1/3). 0.3

(a)

(t0.5,rxn)-1 / min-1

(t0.5,rxn)-1 / min-1

0.3

0.1

(b)

0.1

PHB dispersed phase

PHBV dispersed phase 0.05

0.05 0.1

1

volume fraction of PHBV

0.1

1

volume fraction of PHB

Figure 21. Rate of reaction as a function of volume fraction of (a) PHBV [36] and (b) PHB [37]. Annealing temperature: ● – 220, □ – 228, × – 234, ▲ – 184, Δ – 190 and ■ – 199 oC

Table 7. Exponent β after Eq. (4) for ENR in blends with PHBV and PHB Blends ENR/PHBV

ENR/PHB

Annealing temperature / oC 220 228 234 184 190 199

β -0.46 -0.41 -0.39 -0.41 -0.40 -0.36

Figure 22 demonstrates the enhancement of the compatibility of the ENR/PHBV blends after melt reaction by DSC analysis for quantity Tg. Before melt reaction [Ta = 234 oC for annealing time (ta) = 1 min], two Tgs which correspond to those of the neat constituents were found for the blends. Meanwhile, the Tgs of ENR and PHBV in the blends are relatively

332


constant throughout the entire composition range. Immiscibility of the components for the blends over the entire composition range is suggested. The blends were annealed at Ta = 234 o C for ta = 1t0.5,rxn. Again, the use of 1t0.5,rxn serves to impose equivalent thermal history to all the samples after melt reaction. After melt reaction, existence of a single compositiondependent Tg that corresponds closely to Fox equation 

W 1 − WENR 1 = ENR +  Tg,blend Tg,ENR Tg,PHBV 

  is obtained.  

275

270

Tg / K

265

260

255

250 0.0

0.2

0.4

0.6

0.8

1.0


Figure 22. Glass transition temperatures for ENR/PHBV blends before and after melt reaction at Ta = 234 oC. Before melt reaction (ta = 1 min): ● – ENR and ▲ – PHBV. □ – for blends after melt reaction (ta = 1t0.5,rxn). The dashed curve represents the Fox equation with Tg,PHBV = 271 K and Tg,ENR = 255 K 275

270

Tg / K

265

260

255

250 0

2

4

6

8

10

annealing time / min

Figure 23. Glass transition temperatures for ENR/PHB 50/50 blends at Ta = 190 oC as a function of annealing time. Tg: ● – PHB and ○ – ENR. Tg calculated after Fox equation with Tg,PHB = 274 K and Tg,ENR = 255 K is represented by + [37]

The blend of ENR/PHB 50/50 was annealed at Ta = 190 oC. Two Tgs are observed when the sample was annealed for 1 min. The two Tgs shift inwardly when the annealing time increases to 4 min and finally merge into one Tg at 8 min as depicted in Figure 23. The Tg of


333

the 50/50 (260 K) after melt reaction at 8 min is relatively close to the estimated Tg after Fox equation (264 K). This observation indicates a progressive enhancement of compatibility for this blend with the increase in reaction duration. Grafting of PHBV chains onto the ENR backbone shall alter the crystallization kinetic as well as the morphology of the blends after melt reaction. Figure 24 shows selected examples of fully developed spherulitic morphologies before melt reaction (Ta = 234 oC and ta = 1 min) and after completion in melt reaction at Ta = 234 oC for 4t0.5,rxn. Afterwards, samples were allowed to isothermal crystallize at Tc = 50 oC until completion and micrographs were taken using POM. ENR/PHBV 20/80 shows volume-filling spherulites before and after melt reaction. However, much larger size of the spherulites are shown after melt reaction due to the marked reduction in the rate of crystallization at Tc = 50 oC. This observation is generally true for all other compositions. The crystallization rate of PHBV is significantly reduced in blends after melt reaction. Random dispersion of PHBV spherulites in the matrix for the ENR/PHBV 60/40 blends can be seen (c.f. Figure 24 c). We note here that, the morphology and the crystallization of the PHBV can be finely adjusted by controlling the melt reaction. The compatibility of the blends can be enhanced after the melt reaction.

(a)

(b)

(c) Figure 24. Morphology under POM of ENR/PHBV blends crystallized at Tc = 50 oC until completion after the samples were annealed at Ta = 234 oC and TPOM = 50 oC. (a) ENR/PHBV 20/80 before melt reaction at Tc = 50 oC after 4 hr. Melt reacted ENR/PHBV at Tc = 50 oC after 1 day: (b) 20/80 and (c) 40/60

334

C. H. Chan, H. W. Kammer, L. H. Sim et al. Table 8. Specific melting enthalpy by DSC, notched Izod impact strength, storage modulus by DMA and rubber particle size for PHB/ENR toughened blends [39] Sample

Impact Storage strength modulus at 23 WPHB Wplasticizer WENR WMPB o /J m-1 C/GPa 0.90 0.10 90.4 23 1.6 0.54 0.06 0.30 89.4 25 0.5 0.54 0.06 0.30 0.10 79.6 124 0.8 a The PHB in this study has 10% proprietary plasticizer mixed in it as supplied. a

ΔH/WPHB /J g-1

Rubber particle size/μm 20 - 25 1-5

Samples of PHB/ENR with added maleated polybutadiene rubber (MPB) (M = 3100 g mol -1) with 6 maleic anhydride groups per chain were melt compounded by micro twin screw extruder with injection molder system at 175 oC and 200 rpm for 3 min [39]. PHB/ENR blends are not compatibilized after melt compounding, where the specific melting enthalpy and impact strength for PHB and PHB in blends with ENR remain unchanged as indicated in Table 8. This implies that ENR is not interacting (or compatible) with PHB under the experimental conditions. Roughly 440% increase in the notched Izod impact strength and the reduction of specific melting enthalpy can be detected for the blend with the addition of the compatibilizer (MPB in this case) as compared to that of the blend without MPB. At 23 oC, the storage modulus of PHB decreases by 68 % upon addition of ENR but only decreases by 50 % upon addition of ENR and MPB. The dispersion of ENR in PHB is enhanced with smaller rubber particle size after the addition of MPB (c.f. Table 8). This reactive and compatibilized blend shows better impact resistance and only 50% loss in storage modulus as compared to neat PHB. The interesting question here is: what will be the variation in conductivity for the compatibilized polyether/ENR blends added with Li salt after thermal or catalyst induced melt reaction? Architectures of the graft polymers, modification of micro and nanophase structures, changes of physical-chemical and thermal properties of the blends after the melt reaction will definitely impose an impact on the conductivity. The insertion of dispersed phase of elastomer into the thermoplastic coupled with suitable compatibilization between the components may lead to hierarchical morphology with continuous percolation pathways. This may offer maximal electrical conductivity for the system. Extensive study needs to be carried out in order to give a satisfactory answer to this important question.

CONCLUSIONS •

Thermal properties and morphologies of the immiscible systems for ENR blended with PET, PHBV and PEO are ruled by the relative amount of the constituents and the distribution of phases in the blends. Thermal behavior governs the development of morphologies for the systems. Blends of ENR/PVC are compatible, with addition of ENR into the PVC matrix, the yielding and the ductile behavior of the blends are enhanced with increasing ENR content.

Blends of Epoxidized Natural Rubber and Thermoplastics •

•

335

The salt-containing blends with dispersed phase of ENR in PEO (hierarchical morphology with continuous percolation pathways) are potential candidates for their application as solid polymer electrolytes. This is due to the preferential solubility of the Li salt in PEO as compared to that of ENR. Salt-containing blends with dispersed phase of PVC in ENR show enhancement of conductivity as compared to that of the neat parent polymers with the addition of salt. The compatibility of the ENR/PHBV and ENR/PHB blends can be enhanced after the melt reaction. Besides, the morphology and the crystallization of the PHBV can be finely adjusted by controlling the melt reaction. Melt compounding ENR/PHB blend with MPB compatibilizer exhibits higher impact strength. This offers the possibility to compatibilize polyether and ENR added with Li salt in a similar approach. The compatibilized blends may have enhanced mechanical properties without the conductivity of the salt-containing blends.

REFERENCES [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21]

Brown, H. Rubber: Its Source, Cultivation, and Preparation; John Murray, London, 1918. Semegen, ST. In Rubber Technology; Morton, M. (Ed.); Van Nostrand Reinhold Company, New York, 1987, Chap. 6, 169. Tobing, SD. Rubber World, 1988, 197, 33 – 37. Senyek, ML. In Encyclopedia of Polymer Science and Engineering, 2nd Ed; Mark, HF; et al. (Ed); John Wiley and Sons, New York, 1989, Vol. 8. Baker, CSL; Gelling, IR. Samsuri, AJ. Nat. Rubb. Res., 1986, 1, 135 -144. Pummere, D; Burkard, PA. Ber. Dtsch. Chem. Ges., 1922, 55, 3458 – 3472. Derouet, D; Mulder-Houdayer, S; Brosse, JC. J. Nat. Rubb. Res., 2006, 9, 1. Tanrattanakul, V; Sungthong, N; Raksa, P. Polym. Testing., 2008, 27, 794-800. Mohanty, S; Roy, S; Santra, RN; Nando, GB. J. Appl. Polym. Sci., 1995, 58, 19471957. Mohanty, S; Nando, GB; Vijayan, K; Neelakanthan, NR. Polym., 1996, 37, 5387-5394. Mohamad, Z; Ismail, H; Chantara Thevy, R. J. Appl. Polym. Sci., 2006, 99, 1504-1515. Margaritis, AG; Kalfoglou, NK. Polym., 1987, 28, 497-502. Margaritis, AG; Kallitsis, JK; Kalfoglou, NK. Polym., 1987, 28, 2122-2129. Varughese, KT; Nando, GB; De, PP; Sanyal, SK. J. Mat. Sci., 1988, 23, 3903-3909. Mukhopadhyay, S; De, SK. Polym., 1991, 32, 1223-1229. Senna, MMH; Abdel-Fattah, AA; Abdel-Monem, YK. Nucl. Instrum. Methods Phys. Res., B. 2008, 266, 2599-2606. Thitithammawong, A; Nakason, C; Sahakara, K; Noordermeer, JWM. Eur. Polym. J., 2007, 43, 4008-4018. Sulaiman, SF; Chan, CH; Harun, MK. Mat. Res. Inno., 2009, 13, 225-208. Chan, CH; Kammer, HW. Polym. Bull., 2009, 63, 673-686. Chan, CH; Kammer, HW. J. Appl. Polym. Sci., 2008, 110, 424-432. Hoffman, JD; Weeks, JJ. J. Res. Nat. Bur. Stand.-A, Phys. Chem., 1962, 66A, 13-28.

336


[22] Radhakrishnan Nair, MN; Thomas, GV; Gopinathan Nair, MR. Polym. Bull., 2007, 59, 413-426. [23] Perera, MCS; Ishiaku, US; Ishak, ZAM. Polym. Degrad. Stab., 2000, 68, 393-402. [24] Varughese, KT; Nando, GB., De, PP; De, SK. J. Mat. Sci., 1988, 23, 3894-3902. [25] Cimmino, S; Di Pace, E; Martuscelli, E; Silvestre, C. Makromol. Chem., 1990, 191, 2447-2454. [26] Scandola, M; Focarete, ML; Adamus, G; Sikorska, W; Baranowska, I; Swierczek, S; Gnatowaski, M; Kowalczuk, M; Jedlinski, Z. Macromolecules, 1997, 30, 2568-2574. [27] Groeninckx, G; Reynaers, H; Berghmans, H; Smets, G. J. Polym. Sci., Polym. Phys., Ed. 1980, 18, 1311-1324. [28] Noor, SAM; Ahmad, A; Rahman, MYA; Talib, IA. J. Appl. Polym. Sci., 2009, 113, 855-859. [29] Abdul Nasir, NH; Chan, CH; Kammer, HW; Sim, LH; Yahya, MZA. Macromol. Symp., 2010, 290, 46-55. [30] Florjańczyk, Z; Marcinek, M; Wieczorek, W; Langwarld, N. Polish J. Chem., 2004, 78, 1279-1304. [31] Noor, SAM; Ahmad, A; Talib, IA; Rahman, MYA. Ionics, 2010, 16, 161-170. [32] Latif, F; Aziz, M; Katun, N; Ali, AMM; Yahya, MZA. J. Power Sources, 2006, 159, 1401-1404. [33] Ramesh, S; Tai, FY; Chia, JS. Spectrochimica Acta Part A, 2008, 69, 670-675. [34] Rahman, MYA; Ahmad, A; Wahad, SA. Ionics, 2009, 15, 221-225. [35] Ahmad, A; Rahman, MYA; Ali, MLM; Hashim, H; Kalam, FA. Ionics, 2007, 13, 6770. [36] Chan, CH; Ismail, J; Kammer, HW. Polym. Degrad. Stab., 2004, 85, 947-955. [37] Lee, HK; Ismail, J; Kammer, HK; Bakar, MA. J. Appl. Polym. Sci., 2005, 95, 113-129. [38] Chan, CH; Kummerlöwe, C; Kammer, HW. Macrom. Chem. Phys., 2004, 205, 664675. [39] Parulekar, Y; Mohanty, AK. Green Chem., 2006, 8, 206-213.