Printer Friendly

Compatibilization of starch-polyester blends using reactive extrusion.


Plastics obtained from petrochemical sources generally have lifetimes of hundreds of years when buried in typical solid waste sites, and thus have posed significant environmental threats [1]. These plastics have contributed 17-25% of the volume of waste being land filled [2]. Hence, recently, much research has concentrated on the development of environmental-friendly biodegradable plastics obtained from various natural renewable resources including starch, cellulose, poly(hydroxyalkanoates), and poly(lactic acid).

Three main classes of biodegradable polymers have been recognized. The first class consists of synthetic polymers sourced from nonrenewable resources with vulnerable groups susceptible to hydrolysis attack by microbes. The second class of materials is composed of naturally-occurring bacterial polymers such as polyhydroxybuterates (PHB), polyhydroxyvalerates (PHV), and polyhydroxybuterates-co-valerates (PHB/V). Although polyhydroxyalkanoates are truly renewable and biodegradable, the determination of processing parameters and high production costs are the concerning factors [1]. The third class of materials is biodegradable polymers from renewable resources such as starch, cellulose, and polylactic acid. Starch-based thermoplastics are inexpensive, but generally possess high viscosities and poor melt properties that make them difficult to process. As a result, products made from starch are often brittle and water-sensitive [3]. To alleviate these problems, starch polymers are often blended with more high performance synthetic polymers. Of course, in the case of starch-polyester blend systems one must also factor in the higher cost of the polyester, and aim to increase the percentage of the lower cost starch in the blend while maintaining the desired mechanical properties and performance of the polymer product. Thus to achieve this cost-performance optimization, much must be understood and optimized about the individual polymers and the blending process.


Recent NMR studies suggest that there are at least three distinct components in wheat starch granules [4]. They are as follows: (i) highly crystalline regions formed from double-helical starch chains, (ii) solid-like regions formed from amylose-lipid inclusions complexes, and (iii) completely amorphous regions associated with branching regions of the amylopectin components of starch and possibly the lipidfree amylose. Starch granules essentially consists of linear poly-(1 [right arrow] 4)-[alpha]-D-glucose, amylose, and the branched molecule, amylopectin, where the linear chains are connected through (1 [right arrow] 6)-[alpha] linkages. When starch granules are observed under polarized light, a characteristic dark cross is seen, which has led to regard the granules as distorted sphaerocrystals [5]. The proportion of amylopectin ranges from 95% in the low amylose or waxy starches through 70-75% in normal starches to 30% in some high amylose starches [6, 7].

Hydroxyl groups cause starch to generally behave as an alcohol during chemical reactions. This property of starch is important when considering reactive blending of starch with synthetic polymers. The presence of such large number of -OH groups affords starch hydrophilic properties and therefore adds affinity for moisture and dispersability in water [8]. However, hydrophilicity is undesirable in many plastic packaging applications and hence it is a major limitation in using starch as a homopolymer.


Most synthetic polymers are incompatible with starch and it is difficult to add starch to synthetic polymer systems in large quantities without dramatically reducing performance properties [9]. To enhance the compatibility of starch and synthetic polymers, a reactive functional group such as maleic anhydride (MA) can be added to the synthetic polymer [10]. In these blends, the interfacial tension is lowered to generate a small phase size and strong interfacial adhesion to transmit applied force effectively between the component phases [11]. Thus, compatibilization through in situ formation of compatibilizer in polymer blends has become increasingly important and an alternative to replace the method of adding block or graft copolymers separately [12]. Synthetic polymers having functional groups such as carboxylic acid, anhydride, epoxy. urethane, or oxazoline can react with the -OH group in starch to form a blend with stable morphology. These reactive groups form a chemical or physical bond with hydroxyl or carboxyl groups in natural polymers such as starch [13].

The role of a compatibilizer is to control the properties of multiphase blends not simply by converting immiscible blends into fully miscible blends, but by controlling the size of the phase domains of immiscible blends. Effective compatibilizers must be located at the interface between the phase domains of the immiscible blend. Most importantly, the degree of compatibilization in a particular system depends on the reactivity of the compatibilizer used. It has been found that a compatibilizer is most effective when its sections are of higher molecular weight than the corresponding blend components [14]. Several theories have been proposed to explain the role of compatibilizers, of which two mechanisms are considered plausible. The first mechanism is thermodynamic in nature in that the compatibilizer reduces the interfacial tension between the phases. The second mechanism is kinetic in nature in that the presence of the compatibilizer at the interface reduces the agglomeration of domains by steric stabilization. However it is often unclear which of these mechanisms dominates the reduction in the particle size [14], and hence, the mechanism of compatibilization is still a debatable issue.

Several researchers have showed that dissimilar blends could be compatibilized using different techniques. Bacon and Farmer [15] showed that MA reacts with natural rubber in presence of benzoyl peroxide and results in the addition of MA to the double bonds and to the [alpha]-methylene groups of the polymer. Alder et al. [16] have confirmed the addition reaction of MA with rubber in solid phase and in solution. Mani, Bhattacharya, and Tang [12] have used initiators such as dicumyl peroxide (DCP), benzoyl peroxide, and di-tert-butyl peroxide (DBP) for grafting biodegradable polybutylene succinate (PBS) with starch in presence of MA as a crosslinking agent. Sailaja and Chanda [17] have used MA-grafted polyethylene (PE) for preparing PE-starch blends in presence of benzoyl peroxide as initiator. Ceric ammonium nitrate (CAN) is another initiator commonly used during reactive grafting. Luftor et al. [18] studied the kinetics of graft polymerization of acrylonitrile onto sago starch using CAN as initiator. These works suggested the use of other crosslinking agents such as phthalic anhydride and dodecyl succinic anhydride in the presence of initiators to facilitate grafting reactions.

Of all the methods used to achieve compatibilization in starch-polyester blends, reactive extrusion (REX) is a well-utilized technique as it combines the traditionally separated chemical processes (polymer synthesis and/or modification) and extrusion (melting, blending, structuring, devolatilization, and shaping) into a single process carried out in the extruder [19]. Using REX, modified polymer can be obtained in a ready-to-use form at the die.

Unmodified starch-based thermoplastics generally have higher viscosities and poor melt properties than traditional synthetic polymers that make them difficult to process. Also, starch and synthetic polymers are generally thermodynamically dissimilar in nature, and hence are incompatible unless a compatibilizer is used. In REX of crosslinking dissimilar polymers. free-radical initiation plays a predominant role. However in the extruder, mechanochemistry, by itself, is not powerful enough for free-radical generation in such grafting reactions. Hence, in this study, we have attempted to graft MA to a biodegradable polyester in presence of a free-radical initiator during stage-one of extrusion, and further crosslinked MA-grafted polyester to starch during a second extrusion stage. We used different temperature profiles during the REX process. Should the compatibilized blends exhibit improved mechanical properties, determination of interfacial tension and controlling the interfacial properties of such blends will lead us on the path to developing well-controlled biodegradable blends with tailorable mechanical properties.


EnPol[R], a biodegradable thermoplastic polyester, was obtained from IRe Chemical, Korea. Two types of starch with different levels of amylopectin-to-amylose ratios were used in a preset blend ratio in all formulations. The first was a low-amylose common waxy wheat starch and the other a high-amylose. chemically-modified maize starch (hydroxypropylated starch). The second has a lower gelatinization temperature than the first and is able to withstand higher processing temperatures. Both starches were sourced from Penford (Australia). These starches were blended together with plasticizers in accordance with the patents [20, 21] held by the Cooperative Research Centre for International Food Manufacture and Packaging Science and a spin-off company. Plantic[TM] Technologies. Melbourne, Australia. Maleic anhydride (MA) (98.06%), obtained from ICN Biomedicals. USA. was used as a crosslinking agent, and Dicumyl peroxide (DCP. 99%), obtained from Sigma--Aldrich, was used as a free-radical initiator. Inert nitrogen gas was used to prevent the effect of moisture on grafting during the first stage of extrusion.

Blend Preparation

A batch of Starch Patent Formulation (SPF) was prepared using a granulator. The composition of SPF to EnPol was maintained at 40:60 (wt%) in all the blends. The different blends prepared using a laboratory scale PRISM corotating twin-screw extruder (length to diameter ratio of 40:1 and screw diameter of 16 mm) are shown in Table 1.

Grafting Procedure

The grafting reactions were carried out in a laboratory scale PRISM co-rotating twin-screw extruder with a barrel length to diameter ratio of 40:1, a screw diameter of 16 mm and eight heating zones. The composition of SPF and the polyester was maintained at 40:60 (wt%) in all the blends.

The first step was the preparation of maleated polyester. From initial trials we found that the color of the extrudate varied as the composition of MA was increased (0.5% MA, white; 1% MA, light-pink; 1.5% MA, dark-pink; 2% MA, brown; 3% MA, brownish-black; 5% MA. black). Maleated polyesters with MA concentration of more than 2% were found to be unsuitable for compounding with starch. Therefore, after optimizing the process, three different compositions of MA (0.5, 1, 1.5%) and four different compositions of DCP (0.3, 0.5, 0.8, 1.2%) were selected. The temperature profiles for stage-1 and stage-2 of extrusion are shown in Table 2. EnPol was dried under vacuum for 24 h prior to the day of extrusion. MA and DCP were used in their powder form. A mixture of polyester, MA, and DCP was introduced using a mechanical feeder at a feed rate of 0.38 kg/h. Screw speed of 45 rpm was used here. A continuous flow of nitrogen was maintained with the help of an inlet device during the first stage of extrusion. The torque and die pressure were monitored using a torque-meter and pressure-transducer, respectively. The temperature between mixing and transportation zones was maintained at 180[degrees]C to facilitate peroxide-initiated free-radical generation. The extrudate obtained from the first stage of extrusion was pelletized using a pelletizer and stored in a humidifier at 60% RH until further use.

The second step was compounding of the pelletized MA-grafted polyester with SPF. Here, a screw speed of 70 rpm and a feed rate of 0.45 kg/h were used. The temperature between mixing and transportation zones was maintained at 145[degrees]C. The extruded strands were stored in a humidifier at 60% RH.


Tensile Testing

The strands of different blends were subjected to tensile testing using an Instron Universal tensile testing machine (model 5584). A 5 KN load cell was used for determining important properties such as stress at break, strain at break, Young's modulus, and stress and strain at maximum load. A grip separation of 50 mm, crosshead speed of 5 mm/min, and a sensitivity factor of 20% were adopted. All testings were done according to the procedure outlined in ASTM test method D-638.

Dynamic Mechanical and Thermal Analysis

A Dynamic Mechanical and Thermal Analysis (DMTA) instrument (model IV) from Rheometric Scientific. Piscataway, NJ was used to study the effect of crosslinking and phase separation on the thermal properties of starch-EnPol blends. A temperature sweep test was conducted with a temperature range of 30-85[degrees]C, a frequency of 1 Hz, soak time of 10 sec, and a fixed strain of 0.1%. The extruded samples were sample-pressed to a size of 100 X 10 X 1 mm and subjected to DMTA.

X-ray Photoelectron Spectroscopy

Compression-molded films (thickness 1 mm) of six different blends were subjected to X-ray photoelectron spectroscopy (XPS) using a PHI model 560 XPS/SAM/SIMS I multitechnique surface analysis system incorporating a model 25-270 AR electron energy analyzer. Mg Ka X-rays (1253.6 eV) generated using 400 W (15 kV. 27 mA) was used to produce photoelectrons from the sample surface. Survey (wide) scans, at analyzer pass energy of 100 eV, over 1000 eV were taken, followed by multiplex (narrow) scans of C Is. O Is, and Si 2p energy levels at high resolution using a pass energy of 25 eV. Curve fittings of C (Is) and O (Is) were carved out using Resident PHI V6.0 curve-fitting software to establish the relative percentage of different functional groups.

Differential Scanning Calorimetry

A TA Instrument modulated DSC (TA2920) was used to determine the thermal transitions of starch-polyester blends. The sample size was 10-15 mg, with heating and cooling rates of 10 and 20[degrees]C [min.sup.-1], respectively.

Optical Microscopy (OM)

Analysis of the morphology of the blends was performed using an optical microscope. An optical microscope Olympus BH-2 model and a lens of the type Splan 10PL were used. A scale of 10 [micro]m was adopted for all the samples. The blends were pressed into films of thickness of 10 [micro]m using a sample press device. The thin sections were stained with iodine/KI solution before obtaining images under the microscope at a magnification of X 10. This method was used to differentiate starch from EnPol in the optical micrographs.


Before examining the study's main results, it is instructive to discuss hypothesized reaction mechanism. The proposed crosslinking mechanism for our starch-polyester system is described by the scheme below.

Consider the structure of EnPol.


Consider b = 4. There are three steps involved in the crosslinking of starch to the polyester. They are initiation, propagation, and termination.


Step 1.

Formation of EnPol microradical


Step 3.

The EnPol microradical undergoes [beta]-scission under high temperature conditions (>150[degrees]C).


Step 4.

R'O* could further react with {2} to form another microradical.



Step 5.

The free radical associated chain attaches itself to MA (double-bond cleavage)


Step 6.

Microradical {3} formed in step 4 could participate in the grafting process, resulting in the formation of a stable complex shown below.


Step 7.

During stage 2 of extrusion where starch is compounded with MA-grafted EnPol, C-6 atom of starch [12] reacts with the anhydride group of MA to form starch-MA-polyester complex.


The following starch-MA-polyester complex can also be expected.



Step 8.

The reaction could be terminated in two ways; combination (coupling) or disproportionation. The possible termination steps are given below.

Scheme 1(a) of Coupling


Scheme 1(b) of Coupling:


Scheme 2(a) of Disproportionation:


Scheme 2(b) of Disproportionation:


It has been found that polystyrene terminates predominantly by combination, whereas poly(methyl methacrylate) terminates by disproportionation at polymerization temper atures higher than 60[degrees]C, and partly by each mechanism at lower temperatures [22]. Major studies on termination reaction schemes have been completed with linear-chain polymers, but no published research is available on the termination mechanism of starch-polyester crosslinking.


Twin Screw Extrusion

The twin-screw extruder here was used as a reactive extruder to combine peroxide-initiated grafting reaction and conventional extrusion into a single process. The screw configuration, torque, rotational speed, and mass flow rate are the important terms in determining Specific Mechanical Energy (SME). which is given as

SME = [M.sub.d][omega]/m. (13)

Where, SME = Specific mechanical energy (J/kg),

[M.sub.d] = Torque (Nm)

[omega] = Rotational speed of screw ([s.sup.-1]),

m = Mass flow rate (kg/s).

The greater the torque in the extruder the higher the bulk viscosity of the system. Thus, for crosslinked systems, the SME required should be more. The calculated SMEs for different blends are given in Table 3.

During blending of starch with polyesters in presence of a crosslinking agent and initiator, the anhydride functional group could react with -OH group of starch to form ester linkages [23]. Hence, in this reactive extrusion (REX), inter-crosslinked polymer chains are expected, and as a result, the torque generated and hence the SME of a compatibilized blend should be higher than the SME of uncompatibilized blends. It is evident from Table 3 that the SME of each compatibilized blend is higher than the SME of the uncompatibilized blend (40S, 60PEst). It can be observed that the SME variation in blends containing 1% MA is greater than in other blends. In case of blends containing 0.5% and 1.5% MA, the variation in SME was negligible. This indicates that blends containing 1% MA had higher microradical generation and better crosslinking than in other blends. This could have increased the bulk viscosity and SME of system.

Tensile Testing Analysis

Three important parameters were considered for tensile testing analysis, namely tensile stress at maximum load, stress at break, and Young's modulus (tangent 5%). The effect of the concentration of MA and DCP on these parameters was determined and is shown in Table 4.

The blends at 0.5% MA exhibited better tensile properties in terms of Young's modulus and stress at break, than the unmodified blend, but showed little or no improvement in elongation at break. The blends with 1% MA had moderately higher Young's modulus and stress at break values than the unmodified blend. They also showed much higher elongation at break than the uncompatibilized blend and indeed the other reactively extruded blends. All blends containing 1.5% MA had lower elongation at break than the uncompatibilized blend. The Young's modulus and stress at break were increased over the uncompatibilized system. In comparing between the compatibilized blends, blends containing 0.5% MA and 1.5% MA had higher Young's modulus than the blends containing 1% MA. However the blends at 1% MA had greatly improved elongation at break values than 0.5% and 1.5% MA blends.

Mani et al. [23] observed that starch blends that encountered higher specific mechanical energy had high tensile strength. Our results indicated that only few starch blends with higher specific mechanical energy had better tensile strength. It is hence important to consider the effects of interfacial properties and maleation on the tensile properties of starch-polyester blends.

Bhattacharya et al. [24] noticed that the addition of compatibilizers (styrene MA copolymer and ethylene-propylene-g-maleic anhydride copolymer) had a profound effect on the tensile properties of the starch blend (60% starch, 40% compatibilizer). However, those blends exhibited poor elongation at break. Avella et al. [25] have shown that an increase in the composition of starch and precompatibilizer decreases both tensile strength and elongation at break, but increases Young's modulus. In synthetic polymer blends, the addition of a second phase to the polymer matrix usually diminishes the elongation properties at break [26] and in many cases, when 20% of the dispersed minor phase has been added highly deformable matrix materials are transformed into brittle materials [27]. The elongation at break in synthetic polymer blends is therefore considered to be highly sensitive to the state of the interface. However, we observed a notable increase in elongation at break and overall decrease in Young's modulus in blends containing 1% MA. Krishnan and Narayan [28] have described in their patent that the hydroxyl groups of plasticizers and starch molecules could interact with compatibilizers, promoting interfacial adhesion. They also indicate an acute possibility of plasticizers acting as stretching agents. If it is assumed that maleation did not occur, then starch plasticizers should have increased elongation at break and interfacial adhesion in all blends, irrespective of the composition of the crosslinking agent and the initiator. The observation that only few maleated starch blends showed improved mechanical properties indicates the importance of optimizing the composition of the crosslinking agent and the initiator. In particular, the blend (40S, 60PEst, 1MA, 0.8DCP) showed the highest elongation at break (257%). This could have resulted from the strong interfacial adhesion due to crosslinking of the two phases, and the greater ability of this interface to withstand higher extension to break. However, blends containing 1.5% MA had relatively higher Young's modulus and lower elongation at break. Ideally an optimized compatibilized blend here is a compromise between desired mechanical properties and starch composition.

Dynamic Mechanical and Thermal Analysis

Dynamic Mechanical and Thermal Analysis (DMTA) experiments are used to investigate the mechanical behavior of materials, and to obtain information about the relaxation mechanisms that may be correlated with the dynamics and the microstructure of the material [29]. The effect of temperature on storage modulus (E'), loss modulus (E"), and loss factor (tan [delta]) at a fixed strain rate (0.1%) was studied here. The storage modulus E' is related with the mechanical energy stored during each load cycle and per unit volume. Loss modulus E" signifies the dissipation of energy as heat during the deformation. The loss factor tan [delta] is equal to E"/E' and is thus sensitive to balance of the dissipated and stored energy of the system and is useful to detect thermomechanical relaxations [30].


E' for the uncompatibilized blend (40S, 60PEst) was found to decrease gradually with increase in temperature, indicating a stiffness loss. Here, a definite glass transition was observed for E" and tan [delta] curves between 35 and 50[degrees]C, which should be due to starch. The graphs for E' and tan [delta] are shown in Figure 1.

Analysis of compatibilized blends

Graphs of storage modulus versus temperature of different blends containing 0.5, 1, 1.5% MA, and various compositions of DCP are shown in Figures 2-4.

Although there appears to be no specific trend, the blend (40S, 60PEst, 0.5MA, 0.8DCP) exhibited the highest storage modulus (4.64 X [10.sup.8]Pa) at 30[degrees]C, followed by blends containing 0.5, 0.3, and 1.2% DCP (Fig. 2). It was also observed that at higher temperatures (>80[degrees]C), blends containing higher compositions of DCP lost their stiffness more quickly than the blends containing lower compositions of DCP. For the blend (40S, 60PEst, 0.5MA, 0.3DCP), two distinct peaks were observed at 55 and 68[degrees]C. The first peak at 55[degrees]C is due to the addition of the interface modifier (compatibilizer). The other peak at 68[degrees]C could be due to the molecular motions within the starch phase. At higher concentrations of DCP, these peaks were found to disappear. When compared with the transition peak in the uncompatibilized blend, it is found that with the addition of compatibilizer the transition peak in compatibilized blends shifted towards higher temperatures. It can be observed from Figure 3 that in all the blends containing 1% MA, a drastic decrease in storage modulus (E') is observed. This indicates that as the concentration of MA in the blend is increased from 0.5 to 1%, there is a sharp decrease in stiffness. This is in agreement with the tensile testing results. In all the blends containing 1% MA, a transition peak is observed between 70[degrees]C and 85[degrees]C, which is a shift by about 10[degrees]C when compared with the transition peak in the blend (40S, 60PEst, 0.5MA, 0.3DCP). This indicates that the crosslinked points could have been obstructing the conformational mobility of the segments of EnPol, resulting in further shift of the transition peak. A series of transitions observed in the blends containing 0.3, 0.5, and 0.8% DCP could be due to the side groups and segmental motions of the crosslinked chains.



It is noticeable that all the blends containing 1.5% MA have storage modulus between that of blends containing 1 and 0.5% MA. These blends are found to exhibit similar tensile properties to the blends containing 0.5% MA. Most of these blends exhibit a transition between 50 and 60[degrees]C, which could again be due to the presence of interface modifier. These blends also showed transition peak between 60 and 70[degrees]C, signifying the molecular motion within starch.

X-ray Photoelectron Spectroscopy

Although X-ray Photoelectron Spectroscopy (XPS) is a surface analysis technique, it still gives information about the structure of crosslinked polymers. Compression molded films of six blends [(40S, 60PEst), (40S, 60PEst, 0.5MA, 0.5DCP), (40S, 60PEst, 1MA), (40S, 60PEst, 1MA, 0.8DCP), (40S, 60PEst, 1.5MA), (40S, 60PEst, 1.5MA, 0.5DCP)] were washed using hexane to remove the surface impurities and then examined with XPS. In the "survey scans" of all the graphs, Si (2p) and Si (2s) peaks were observed in the range of 102 and 106 eV. Peaks at binding energies of 102 and 103.4 eV correspond to Si and methyl (C[H.sub.3]) group, respectively. Hence, it is assumed that the active peaks between 102 and 106 eV were the presence of poly(dimethylsiloxane) [PDMS], which could have been present in the polyester in the form of crystal impurity. In the survey scan, a graph of electron count Vs binding energy (eV) was plotted. In all the graphs, distinct peaks corresponding to C (1s) and O (1s) were observed in the range 250-300 eV and 500-550 eV, respectively.


Then a high resolution scan (called multiplex) of the C (1s) and O (1s) energy levels of the mentioned blends was carried out at a pass energy of 25 eV. It was observed that the ratio of C/O of the blends was approximately 3. According to the proposed structure, the -OH group of C-6 of starch attaches itself to the C=O group of MA, resulting in the formation of a carboxylic group (-COOH). We determined the compositions of -CO and -COO groups in different blends using curve-fitting software "Resident PHI V6.0". All the data was charge-corrected at 101.8 eV using Si (2p) as reference for PDMS. The results of the curvefitting are shown in Figure 5.


The curve fitting results of C (1s) energy levels of different blends are shown in Table 5.

For six different blends, the factor [A.sub.1]/[A.sub.2] (Area of -COO/Area of -CO) was determined based on the area of respective peaks in the curve-fitted graphs. It is to be noted here that XPS does not detect H. It was observed that the composition of -COO groups in the blend (40S, 60PEst, 1MA, 0.8DCP) was higher than the composition of -COO groups in other blends. The -COO peak in the curve-fitted graph is due to the presence of -COO groups of EnPol and -COOH groups, which are formed due to crosslinking (refer to mechanism). As crosslinking increases, the percentage of -COOH groups in the compatibilized blends also increases, which is due to the transfer of -OH group of C-6 of starch to the C=O group of MA to form -COOH group. The number of C=O groups in the compatibilized blends varies with the degree of crosslinking. For example, even in compatibilized blends, the presence of unreacted MA increases the number of C=O groups. Also, in the blends that did not contain DCP, XPS detected more C=O groups, indicating that there is a possibility of MA remaining in unreacted state or partially reacted state.

Differential Scanning Calorimetry

Differential scanning calorimetry was used to evaluate the thermal transition of the blends. To eliminate thermal history, the samples were equilibrated at 30[degrees]C, heated to 220[degrees]C at 10[degrees]C [min.sup.-1], cooled to -70[degrees]C at 20[degrees]C [min.sup.-1], maintained under isothermal conditions for 5 min, and heated to 220[degrees]C [min.sup.-1] at 10[degrees]C [min.sup.-1].

Thermal transitions such as melting and crystallization are of high importance in polymer processing techniques. For the base blend containing 60% polyester and 40% starch based plastic, the melting endotherm occurred at 98.8[degrees]C, while the crystallization exotherm occurred at 63.8[degrees]C (Table 6). For compatibilized starch-polyester blends, no significant change in the endotherm (melting) temperature was observed, but the exotherm (crystallization) temperatures of all the compatibilized blends were lower than the crystallization temperature of the base blend (refer to Table 6). Hence, crosslinking agents do have significant effect on the crystallization temperature of starch-polyester blends. Because of [beta]-scission of the polyester the chains could form crosslinked regions, resulting in the restriction of polymer chain mobility, which could significantly reduce the degree of crystallinity. However, it is interesting to note that the crystallization temperatures of the two blends at 1.0% MA addition {(40S, 60PEst, 1MA, 0.8DCP) and (40S, 60PEst, 1MA, 1.2DCP)} were higher than the crystallization temperatures of other compatibilized blends. In fact, the greatest increase in exotherm temperature was found in (40S, 60PEst, 1MA, 1.2DCP). The presence of optimum concentration of the compatibilizer could have prevented appreciable micelle formation and hence reduced the interfacial energy of the blend. As a result, the degree of crystallization could have increased in both phases and at the interface, further leading to an increase in the number of nucleation sites. In the two blends mentioned earlier, the compositions of MA and DCP could hence be in the proximity of their respective optimum concentrations. The crystallization temperatures of other compatibilized blends were low, indicating that the compositions of the crosslinking agent and the initiator in these blends are not optimized. It could also be observed that the crystallization temperature for the uncompatibilized blend (40S, 60PEst) is quite high (63.8[degrees]C), and yet the blend exhibited poor mechanical properties. The absence of crosslinking agent could promote micelle formation at the interface of the hydrophobic polyester and the hydrophilic starch, thus contributing towards poor mechanical properties. At the same time, the two phases could crystallize in their domains, resulting in an increase in the crystallization temperature of the blend.

Optical Microscopy

The optical micrographs of the uncompatibilized blend and blends that did not contain DCP are shown in Figure 6.


In the uncompatibilized blend (Fig. 6a), due to interfacial tension between the starch and the polyester phases, the starch phase is not homogeneously distributed in the polymer matrix. In the blend {40S, 60PEst, 1MA}, the starch phase could have remained as a separate phase (Fig. 6b), whereas in the other blend (40S, 60PEst, 1.5MA), the starch phases are finer than the phases of the former blend. This could be due to hydrolysis of starch by MA that is present in excess concentration, leading to the break down of the starch phase (Fig. 6c).


Analysis of Blends Containing 0.5% MA.

The optical micrographs of the blends containing 0.5% MA and different compositions of DCP are shown in Figure 7.

It could be observed in the micrographs (Fig. 7) that the starch phase is homogeneously distributed and has formed a cocontinuous phase with the polyester phase. These blends exhibited better mechanical properties than the uncompatibilized blend.

Analysis of Blends Containing 1% MA.

The optical micrographs of blends containing 1% MA and different compositions of DCP are shown in Figure 8.

It can be observed that in the blends with higher DCP content ((40S. 60PEst. 1MA. 0:8DCP) and (40S, 60PEst. 1MA. 1.2DCP)) (Figure 8c and d), starch is evenly dispersed throughout the polymer matrix. This suggests optimum compositions of MA and DCP helps generate polyester microradicals, promoting efficient crosslinking whereupon more and more starch reacts with the compatibilizer. This reduces the interfacial tension between the two dissimilar phases and promotes adhesion.


Analysis of Blends Containing 1.5% MA.

The micrographs of the blends containing 1.5% MA and different compositions of DCP are shown in Figure 9.

It can be observed from micrographs of blends containing 1.5% MA and different compositions of DCP that starch phases are concentrated in certain regions of the polymer matrix, and that the cocontinuous phases exhibited at 1% MA, have now been lost. In these blends with higher composition of DCP (Fig. 9b, c, and d), numerous microradicals generated could react among themselves (disproportionation or coupling) and terminate the reaction. By this, the crosslinking of the two phases is hindered forcing some starch phase to remain in unreacted state.



It is apparent that the mechanical and thermal properties of the compatibilized starch based polymer-polyester polymers are affected by the addition of the compatibilization system. In summary it appears that

* All blends (0.5, 1.0, and 1.5 wt% MA) showed an improvement in Young's modulus and stress at break. Blends with 1.0% MA showed larger improvements in elongation at break.

* DMTA studies revealed that the blends containing 1% MA had lower stiffness (E') than blends containing 0.5% MA and 1.5% MA.

* XPS analysis indicated the presence of more -COO groups in the blend with 1% MA, thus supporting the proposed compatibilized structure for the blend.

* The DSC results indicated that the crystallization temperatures were reduced with increasing compatibilizers because of an increased hindrance to crystallization. Crystallization temperatures of the two blends at 1.0% MA addition {(40S, 60PEst. 1MA. 0.8DCP) and (40S. 60PEst, 1MA. 1.2DCP)} had higher than the crystallization temperatures than that of other compatibilized blends.

* Optical micrographs of compatibilized blends with 1% MA revealed uniform distribution of starch in the polymer matrix, indicating compatibilization between starch and polyester phases

Thus optical, DSC, and XPS tests indicates that a reactively compatibilized structure is optimized for the 1.0% MA samples. Interestingly the 1.0% MA samples showed lower elastic modulus or stiffness during DMTA tests, yet showed high young's modulus, stress at break and elongation at break during tensile testing. This is possibly due to the fact the DMTA tests are linear deformation tests and the tensile tests induce nonlinear deformation. That is, the linear DMTA tests determine a lower linear elastic modulus under small deformations, possibly due to the plasticizing addition of the MA to the system as previously noted [30], and the MA induced crosslinks have little effect. However, under nonlinear tensile testing the deformation is larger and more destructive, and the compatibilized network structure is able to reduce the breaking of tie layers commonly described in basic polymer fracture studies. It is also interesting that the Young's modulus, stress at break and elongation at break all increase, unlike normal composite systems, which sacrifice elongation [extensibility] for increases in strength.


The main conclusions from this work are summarized below. First, all blends (0.5, 1.0, and 1.5 wt% MA) showed an improvement in Young's modulus and stress at break. Blends with 1.0% MA showed larger improvements in elongation at break. DMTA studies revealed that the blends containing 1% MA had lower stiffness than blends containing 0.5% MA and 1.5% MA. The compatibilized blends exhibited several transition peaks due to the presence of interface modifier, molecular motions within starch phase, and side groups and segmental motions of the crosslinked chains. XPS analysis indicated the presence of more -COO groups in the compatibilized blend with 1% MA, thus supporting the proposed structure for the blend. The DSC results indicated that the crystallization temperatures were reduced with increasing compatibilizers because of an increased hindrance to crystallization. Crystallization temperatures of the two blends at 1.0% MA addition {(40S, 60PEst, 1MA, 0.8DCP) and (40S, 60PEst, 1MA, 1.2DCP)} had higher crystallization temperatures than that of other compatibilized blends. We believe that the presence of optimum concentrations of MA and DCP in these blends reduces appreciable micelle formation and increases the number of nucleation sites resulting in an increase in crystallization temperature and interfacial adhesion. Optical micrographs of compatibilized blends with 1% MA revealed uniform distribution of starch in the polymer matrix, indicating compatibilization between starch and polyester phases.

In this study, compatibilization of starch and biodegradable polyester has been achieved using REX, and thus it offers a new direction for enhancing the properties of low cost-base renewable biodegradable polymers. Also, for compatibilized blends the study of interfacial tension (subject of future work) is a vital aspect with regards to improving the properties of biopolymers. This ability to tailor biodegradable polymer morphology and properties is crucial if low cost biodegradable polymers are ever to be fully optimized for appropriate performance properties.


1. R. Chandra and R. Rustgi, Polym. Degrad. Stab., 56, 185 (1997).

2. W.M. Doane, J. Polym. Mater., 11, 229 (1994).

3. S.A. McGlashan and P.J. Halley, "Preparation and Characterisation of Biodegradable Starch-Based Nanocomposite Packaging Materials," in Polymer symposium. University of Queensland (2001).

4. P.J. Frazier, P. Richmond, and A.M. Donald, Starch Structure and Functionality, Royal Society of Chemistry, Cambridge, UK (1997).

5. T. Galliard. Starch: Properties and Potential, Wiley: New York (1987).

6. R.F. Tester, J.B. South, W.R. Morrison, and R.P. Ellis, J. Cereal Sci., 13. 113 (1991).

7. R.F. Tester, W.R. Morrison, R.P. Ellis, J.R. Piggott, and G.R. Batts, J. Cereal Sci., 22. 63 (1995).

8. A. Imberty, A. Buleon, V. Tran, and S. Perez, Starch, 43, 375 (1991).

9. D.H.S. Ramkumar, M. Bhattacharya, and D. Zhang, Polym. Network Blends, 7, 51 (1997).

10. D.H.S. Ramkumar, Z. Yang, and M. Bhattacharya, Polym. Network Blends, 7, 31 (1997).

11. J.W. Barlow and D.R. Paul, Polym. Eng. Sci., 24, 525 (1984).

12. R. Mani, M. Bhattacharya, and J. Tang, J. Polym. Sci. Pol. Chem., 37, 1693 (1999).

13. J. John, J. Tang, Z.H. Yang, and M. Bhattacharya, J. Polym. Sci. Pol. Chem., 35, 1139 (1997).

14. D.J. Lohse and S. Datta, Polymeric materials encyclopaedia. CRC Press. Boca Raton, FL (1996).

15. R.G. Bacon and E.H. Farmer, Rubber chem. Technol., 12, 200 (1939).

16. N.G. Gaylord, Maleic Anhydride Grafting. CRC press. New York (1996).

17. R.R.N. Sailaja and M. Chanda, J. Polym. Mater., 17, 165 (2000).

18. M.R. Lutfor, M.Z.A. Rahman, S. Sidik, A. Mansor, J. Haron, and W. Yunus, Des. Monomers Polym., 4, 253 (2001).

19. J.A. Covas, J.F. Agassant, A.C. Diogo, J. Vlachopoulos, and K. Walters, Rheological Fundamentals of Polymer Processing, Kluwer Academic. Boston (1995).

20. F.S. Buehler, J. Casanova, H. Ernst, and H.J. Schultze, Process and apparatus for producing starch melt and products obtainable by this process. U.S. Patent 5,316,578.

21. F.S. Buehler, E. Schmid, and H.J. Schultze, Starch/polymer mixture, process for the preparation of thereof, and products obtainable therefrom, U.S. Patent 5,346.936.

22. F. Billmeyer, Textbook of Polymer Science, 3rd Edition, Wiley-Interscience, New York (1984).

23. R. Mani and M. Bhattacharya, Eur. Polym. J., 34, 1467 (1998).

24. M. Bhattacharya, U. Vaidya, D. Zhang, and R. Narayan, J. Appl. Polym Sci., 57, 539 (1995).

25. M. Avella, M.E. Errico, P. Laurienzo, E. Martuscelli, M. Raimo, and R. Rimedio, Polymer, 41, 3875 (2000).

26. N. St-Pierre, B.D. Favis, and B.A. Ramsay, Polymer, 38, 647 (1997).

27. D.R. Paul, C.E. Vision, and C.E. Locke, Polym. Eng. Sci., 12. 157 (1972).

28. M. Krishnan and R. Narayan, Biodegradable multicomponent polymeric materials based on unmodified starch-like polysaccharides, U.S. Patent 5,500,465.

29. D. Demirgoz, C. Elvira, and J.F. Mano, Polym Degrad, Stabil., 70, 161 (2000).

30. U.R. Vaidya, M. Bhattacharya, and D. Zhang, Polymer, 36. 1179 (1995).

R.B. Maliger, S.A. McGlashan, P.J. Halley, L.G. Matthew

Centre for High Performance Polymers (CHPP), Division of Chemical Engineering, University of Queensland, St. Lucia, Brisbane, Queensland-4072, Australia

Correspondence to: Peter J. Halley; e-mail:

Contract grant sponsor: Plantic[TM] Technologies, Melbourne.
TABLE 1. Blends prepared using twin-screw extruder.

S (wt%) PEst (wt%) MA (wt%) DCP (wt%)

40 60 0.0 0.0
40 60 0.5 0.3
40 60 0.5 0.5
40 60 0.5 0.8
40 60 0.5 1.2
40 60 1.0 0.0
40 60 1.0 0.3
40 60 1.0 0.5
40 60 1.0 0.8
40 60 1.0 1.2
40 60 1.5 0.0
40 60 1.5 0.3
40 60 1.5 0.5
40 60 1.5 0.8
40 60 1.5 1.2

TABLE 2. Temperature ([degrees]C) profiles for stage-1 and stage-2 of

 Stage-1 Stage-2

Die 100 100
Zone9 120 120
Zone8 160 160
Zone7 180 180
Zone6 180 180
Zone5 180 180
Zone6 180 180
Zone5 180 180
Zone4 180 180
Zone3 180 180
Zone2 150 150
Zone1 120 120

TABLE 3. Specific mechanical energy of different blends.

BLEND SME (kj/kg)

40S, 60PEst 152.98
40S, 60PEst, 0.5MA 198.53
40S, 60PEst, 0.5MA, 0.3DCP 535.23
40S, 60PEst, 0.5MA, 0.5DCP 560.42
40S, 60PEst, 0.5MA, 0.8DCP 579.31
40S, 60PEst, 0.5MA, 1.2DCP 540.81
40S, 60PEst, 1MA 214.35
40S, 60PEst, 1MA, 0.3DCP 271.66
40S, 60PEst, 1MA, 0.5DCP 353.23
40S, 60PEst, 1MA, 0.8DCP 413.23
40S, 60PEst, 1MA, 1.2DCP 598.97
40S, 60PEst, 1.5MA 348.94
40S, 60PEst, 1.5MA, 0.3DCP 453.37
40S, 60PEst, 1.5MA, 0.5DCP 454.46
40S, 60PEst, 1.5MA, 0.8DCP 452.53
40S, 60PEst, 1.5MA, 1.2DCP 458.9

TABLE 4. Tensile properties of different blends.

 Tensile stress at
Blend maximum load (Mpa) Strain at break (%)

40S, 60PEst 14.68 [+ or -] 0.95 78.68 [+ or -] 9.58
40S, 60PEst, 0.5MA, 0DCP 15.70 [+ or -] 1.22 80.82 [+ or -] 4.38
40S, 60PEst, 0.5MA, 0.3DCP 20.11 [+ or -] 0.9 75.18 [+ or -] 8.74
40S, 60PEst, 0.5MA, 0.5DCP 19.89 [+ or -] 1.78 78.68 [+ or -] 11.52
40S, 60PEst, 0.5MA, 0.8DCP 18.63 [+ or -] 1.75 92.38 [+ or -] -8.65
40S, 60PEst, 0.5MA, 1.2DCP 17.55 [+ or -] 1.55 50.24 [+ or -] 10.19
40S, 60PEst, 1MA 12.54 [+ or -] 4.42 31.06 [+ or -] 2.42
40S, 60PEst, 1MA, 0.3DCP 10.5 [+ or -] 3.33 150.12 [+ or -] 15.15
40S, 60PEst, 1MA, 0.5DCP 15.71 [+ or -] -0.96 111.24 [+ or -] 12.98
40S, 60PEst, 1MA, 0.8DCP 18.92 [+ or -] 1.04 257.35 [+ or -] 37.69
40S, 60PEst, 1MA, 1.2DCP 16.05 [+ or -] 1.71 165.73 [+ or -] 38.61
40S, 60PEst, 1.5MA, 17.61 [+ or -] 0.54 35.34 [+ or -] 3.62
40S, 60PEst, 1.5MA, 0.3DCP 19.8 [+ or -] 2.78 40.86 [+ or -] 4.81
40S, 60PEst, 1.5MA, 0.5DCP 18.27 [+ or -] 1.56 36.06 [+ or -] 5.51
40S, 60PEst, 1.5MA, 0.8DCP 19.35 [+ or -] 1.61 44.08 [+ or -] 7.85
40S, 60PEst, 1.5MA, 1.2DCP 18.98 [+ or -] 1.58 43.02 [+ or -] 4.96

Blend Young's modulus (5% Tangent) (MPa)

40S, 60PEst 36.77 [+ or -] 6.02
40S, 60PEst, 0.5MA, 0DCP 38.57 [+ or -] 12.48
40S, 60PEst, 0.5MA, 0.3DCP 86.87 [+ or -] 13.22
40S, 60PEst, 0.5MA, 0.5DCP 79.87 [+ or -] 1.10
40S, 60PEst, 0.5MA, 0.8DCP 62.87 [+ or -] 15.55
40S, 60PEst, 0.5MA, 1.2DCP 48.38 [+ or -] 19.83
40S, 60PEst, 1MA 51.74 [+ or -] 21.79
40S, 60PEst, 1MA, 0.3DCP 30.88 [+ or -] 9.94
40S, 60PEst, 1MA, 0.5DCP 43.07 [+ or -] 9.09
40S, 60PEst, 1MA, 0.8DCP 51.59 [+ or -] 6.89
40S, 60PEst, 1MA, 1.2DCP 44.07 [+ or -] 5.88
40S, 60PEst, 1.5MA 82.64 [+ or -] 20.33
40S, 60PEst, 1.5MA, 0.3DCP 88.31 [+ or -] 11.63
40S, 60PEst, 1.5MA, 0.5DCP 80.86 [+ or -] 5.18
40S, 60PEst, 1.5MA, 0.8DCP 74.78 [+ or -] 27.36
40S, 60PEst, 1.5MA, 1.2DCP 79.99 [+ or -] 20.28

TABLE 5. Curve-fitting results of C (1s) energy levels of different

 Area of -COO Area of -CO
Blend groups ([A.sub.1]) groups (A2) A1/A2

40S, 60PEst 983 1372 0.72
40S, 60PEst, 0.5MA, 0.5DCP 880 1074 0.82
40S, 60PEst, 1MA 1246 1508 0.83
40S, 60PEst, 1MA, 0.8DCP 1362 1125 1.21
40S, 60PEst, 1.5MA 872 1354 0.64
40S, 60PEst, 1.5MA, 0.8DCP 789 977 0.81

TABLE 6. Thermal transitions for starch/polyester uncompatibilized and
compatibilized blends.

 Exotherm T
Blends Endotherm T ([degrees]C) ([degrees]C)

40S, 60PEst 98.8 63.8
40S, 60PEst, 0.5MA, 0.3DCP 99.1 49.6
40S, 60PEst, 0.5MA, 0.5DCP 98.4 45.9
40S, 60PEst, 0.5MA, 0.8DCP 98.4 51.6
40S, 60PEst, 0.5MA, 1.2DCP 97.1 52.6
40S, 60PEst, 1MA 99.9 48.4
40S, 60PEst, 1MA, 0.3DCP 98.5 50.7
40S, 60PEst, 1MA, 0.5DCP 98.2 48.4
40S, 60PEst, 1MA, 0.8DCP 97.5 59.2
40S, 60PEst, 1MA, 1.2DCP 97.5 64.8
40S, 60PEst, 1.5MA 99.0 50.2
40S, 60PEst, 1.5MA, 0.3DCP 98.9 55.9
40S, 60PEst, 1.5MA, 0.5DCP 98.1 45.3
40S, 60PEst, 1.5MA, 0.8DCP 98.1 44.8
40S, 60PEst, 1.5MA, 1.2DCP 98.2 45.1
COPYRIGHT 2006 Society of Plastics Engineers, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2006 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Maliger, R.B.; McGlashan, S.A.; Halley, P.J.; Matthew, L.G.
Publication:Polymer Engineering and Science
Geographic Code:8AUST
Date:Mar 1, 2006
Previous Article:Flash*.
Next Article:Magnetorheological elastomers based on isobutylene-isoprene rubber.

Terms of use | Privacy policy | Copyright © 2021 Farlex, Inc. | Feedback | For webmasters