**4.4.1 Hydrolytic degradation of LDPE/TPS blends**

It is well known that acid hydrolysis of starch involves the random cleavage of glycoside bonds producing from oligosaccharides fractions to glucose units (Leach, 1984). In order to quantitatively determine the extent of continuity of TPS blends, samples were exposed to hydrolytic extraction. Figure 10 shows the percent continuity of starch as a function of TPS content for PE1/TPS40 and PE2/TPS40 blends. In both cases there is a monotonic increase in continuity as the concentration of TPS increases. At concentration of 43% or lower, blend

The relative elongation at break (b/b0) in the machine direction of PE1/TPS blends is shown in Figure 9a. The results are excellent and demonstrate that at high glycerol contents (36% and 40%), the blends have an b comparable to the virgin polyethylene (b0) even at 53 wt% TPS. The b values of PE1 blends drop with the addition of TPS29. If these data are compared with the morphology results from the previous section, it is clear that the high <sup>b</sup> for blends with TPS36 and TPS40 is closely related to the ability to deform the TPS phase.

In St-Pierre's work (St. Pierre et al., 1997), PE/TPS blends presented a maximum in the b at around 10 wt% TPS followed by a dramatic drop at 22 wt%. In this work, the improved extrusion process and the controlled deformation of the TPS phase yields an important improvement in the b of PE/TPS blends as a function of composition, as observed in Figure 9a. Such an improvement in b is also, in part, due to a highly effective removal of water by venting before blending with polyethylene. In St-Pierre's process, TPS was blended with LDPE and then passed through the venting section. At low concentration, TPS was probably encapsulated into a LDPE matrix, which impeded proper water removal. The presence of water at the blending temperature (150C) can lead to the formation of bubbles in the extrudate, which weakens the final product (Verhoogt et al., 1995). In the present system, water was completely devolatilized from TPS before mixing with polyethylene (Favis et al.,

The relative Young's modulus (E/E0) is demonstrated in Figure 9b. Once again the results are excellent. The E can be maintained at high levels even at high loadings of TPS36 and TPS40. At lower levels of glycerol (TPS29) the E of the blend can be seen to even exceed that of the neat polyethylene. These are unusual results considering the high levels of immiscibility between PE and TPS. The results also indicate the potential of tailoring the mechanical properties of the blend through an appropriate glycerol content. This unexpected result can be explained by good interfacial contact. Leclair and Favis found that the compression exerted by a crystalline matrix (HDPE), during crystallization, on an amorphous dispersed phase (PC) can result in good interfacial contact and a higher elastic modulus (Leclair and Favis, 1996). They also observed that this effect had a positive influence on the modulus only when the contraction took place on a smooth, non-

It is well known that acid hydrolysis of starch involves the random cleavage of glycoside bonds producing from oligosaccharides fractions to glucose units (Leach, 1984). In order to quantitatively determine the extent of continuity of TPS blends, samples were exposed to hydrolytic extraction. Figure 10 shows the percent continuity of starch as a function of TPS content for PE1/TPS40 and PE2/TPS40 blends. In both cases there is a monotonic increase in continuity as the concentration of TPS increases. At concentration of 43% or lower, blend

**4.3 Mechanical properties 4.3.1 Elongation at break (b)** 

2003).

**4.3.2 Young's modulus** 

deformable surface.

**4.4 Connectivity of TPS particles** 

**4.4.1 Hydrolytic degradation of LDPE/TPS blends** 

morphology plays an important role on percent continuity of LDPE/TPS40 blends. Blends depicting elongated particles show higher percent continuity at comparative concentrations than those displaying spherical morphology. For instance, PE1/TPS40 blends containing 32% TPS40 have 66% continuity while PE2/TPS40 blends composing of 31% TPS40 have only 38% continuity. Above 50% TPS40, at almost 95% continuity, blend morphology does not make any significant difference. At 62 wt% TPS40 the percent continuity of starch domains reaches 100% and the starch phase could be completely extracted. This is indicative of the full connectivity of starch particles through the entirely sample (Figure 10). The use of hydrolytic degradation as previous technique to biodegradation studies could be an important tool to predict enzymatic and bacterial biodegradation.

Fig. 10. Accessibility of starch domains LDPE/TPS40 blends exposed in solution of HCl 6N for 72 hours.

#### **4.4.2 Enzymatic degradation of LDPE/TPS40 blends**

Numerous studies have been done to investigate the enzymatic hydrolysis of starch-based materials. These works involve blends system with synthetic polymers like LDPE (Danjaji, 2002), ethylene vinyl acetate (EVA) (Simons & Thomas, 1995; Araujo et al., 2004) and polycaprolactone (PCL) (Seretoudi et al., 2002). The kinetic of enzymatic degradation of TPS40 and LDPE/TPS40 blends is shown in Figure 11. Amylase from the enzymatic cocktail triggers the cleavage of 1-4 acetal link while glucoamylase attacks the 1-6 links of

Melt Blending with Thermoplastic Starch 17

and low molecular fractions become soluble, but the specimen shape remains intact. Enzymatic hydrolysis of insoluble polymers is known to be affected by the mode of interaction between the enzymes and the polymeric chains and typically involves four steps: (i) enzyme diffusion from the bulk solution to the solid surface, (ii) enzyme adsorption on the substrate, resulting in the formation of enzyme-substrate complex, (iii) catalysis of the hydrolysis reaction, and (iv) diffusion of the hydrolyzed fraction from the solid substrate to the solution (Azevedo et al., 2003). Blends with high loadings of TPS40 show an enzymatic degradation rate as fast as that of the raw TPS40 during the first 3 hours of exposure. This is probably due to the large amount of TPS40 observed on the surface of LDPE/TPS40 blends. Similarly, blends containing about 30% of TPS40 have less starch available on the surface and, consequently, the initial enzymatic degradation rate is slower than the others. As the soluble degradation products of TPS40 diffuse out of the sample, the number of active enzyme units available for starch degradation decreases resulting in a reduction of degradation rate. TPS40 is completely degraded in 36 hours, whereas PE1/TPS40 having 62% and 32% TPS40 and PE2/TPS40 compounded with 31% TPS40 reach their maximum degradation in 72 hours. Conversely, the 69:31 PE2/TPS40 stabilizes at a short period of about 20 hours, whereas the 68:32 PE1/TPS40 blends reaches its plateau at 48 hours. This is likely due to the connectivity of PE2/TPS40 (69:31) blend of starch from the surface;

Weight loss as a function of time is the most useful method employed to monitor biodegradation (Swanson et al., 2003; Bikiaris et al., 1997b). Figure 12 shows the weight loss of LDPE/TPS40 blends exposed to activated sludge as a function of degradation time. As expected, raw PE1 remains unchanged after 45 days. On the contrary, raw TPS40 is completely consumed within 21 days of exposure. For the LDPE/TPS40 blends, the maximum biodegradation extent is observed at times longer than the raw TPS40. If TPS40 particles are present only on the surface, and not interconnected with particles inside the LDPE/TPS40 blends, then it could be expected that starch domains would be completely biodegraded like the raw TPS40. Percent continuity observed in Figure 10 shows that TPS40 particles are interconnected one to another. At TPS40 concentration of about 30%, interconnection increases when the morphology of starch domains changes from spherical (PE2/TPS40 blend) to fiber-like particles (PE1/TPS40 blend). The extent of biodegradation of TPS40 at 45 days of extraction for PE1/TPS40 blends at 62%, 32% of TPS40 and PE2/TPS40 (69:31) were 92%, 39% and 22%, respectively. However, when the maximum biological extraction is compared with the maximum enzymatic degradation, important

Kinetic of biodegradation of TPS40 and LDPE/TPS40 blends shows two stages (Table 1). In all cases, there is a fast weight loss during the first 1.5 days, followed by another stage where biodegradation rate decreases progressively. The fast stage could be related to the combined effect of biodegradation and diffusion of glycerol and low molecular starch fractions out of the sample. Diffusion of water soluble components can be accelerated by starch swelling, as observed in raw TPS40, during the first 6 hr. Weight loss during this period is almost 4 times faster than the following 30 hr. In the case of LDPE/TPS40 blends,

therefore the path of the enzyme is less obstructive.

difference is noticeable, especially in blends with ca. 30 wt% TPS40.

**4.4.3 Microbial biodegradation** 

amylopectin (Chaplin & Kenedy, 1986), which results in starch solubilization and, consequently, weight loss. The extent of enzymatic degradation of starch is depended on TPS40 concentration. As expected, raw TPS40 is completely degraded during the first 36 hours. Blends of PE1/TPS40 having 62% and 32% and PE2/TPS40 (69:31) result in weight losses of TPS40 of 97%, 65% and 32%, respectively at 72 hours. Therefore, weight loss percent is related to the total amount of TPS40 in the blends. Percolation theory is concerned with the connectivity of one component (in our case, TPS40) randomly dispersed in another (Peanaski et al., 1991). Peanansky showed that below an apparent percolation threshold of 30% by volume (40 wt%) of granular starch, only small amounts were accessible for removal. Granular starches are compact particles, such as those observed in the PE2/TPS40 blends. Fiber-like particles observed in PE1/TPS40 blends could be responsible for a lower apparent percolation threshold in this system and, consequently, higher enzymatic degradation values (Li et al., 2005). Extent of enzymatic degradation of LDPE/TPS40 blends is very similar to that obtained by acid hydrolysis.

Fig. 11. Enzymatic degradation kinetic expressed as weight loss for raw TPS40 (), PE1/TPS40 blends: () 62 wt % TPS40, () 32 wt % TPS40 and PE2/TPS40 blends with () 31 wt % TPS40 as a function of incubation time.

On the other hand, TPS40 enzymatic degradation rate is depended on starch concentration and the accessibility of starch domains as is in the case of LDPE/TPS40 blends. TPS40 is almost insoluble in cold water. When TPS40 is exposed to cold water, it swells and glycerol

amylopectin (Chaplin & Kenedy, 1986), which results in starch solubilization and, consequently, weight loss. The extent of enzymatic degradation of starch is depended on TPS40 concentration. As expected, raw TPS40 is completely degraded during the first 36 hours. Blends of PE1/TPS40 having 62% and 32% and PE2/TPS40 (69:31) result in weight losses of TPS40 of 97%, 65% and 32%, respectively at 72 hours. Therefore, weight loss percent is related to the total amount of TPS40 in the blends. Percolation theory is concerned with the connectivity of one component (in our case, TPS40) randomly dispersed in another (Peanaski et al., 1991). Peanansky showed that below an apparent percolation threshold of 30% by volume (40 wt%) of granular starch, only small amounts were accessible for removal. Granular starches are compact particles, such as those observed in the PE2/TPS40 blends. Fiber-like particles observed in PE1/TPS40 blends could be responsible for a lower apparent percolation threshold in this system and, consequently, higher enzymatic degradation values (Li et al., 2005). Extent of enzymatic degradation of LDPE/TPS40 blends

0 12 24 36 48 60 72

PE1/TPS40 blends: () 62 wt % TPS40, () 32 wt % TPS40 and PE2/TPS40 blends with ()

On the other hand, TPS40 enzymatic degradation rate is depended on starch concentration and the accessibility of starch domains as is in the case of LDPE/TPS40 blends. TPS40 is almost insoluble in cold water. When TPS40 is exposed to cold water, it swells and glycerol

Fig. 11. Enzymatic degradation kinetic expressed as weight loss for raw TPS40 (),

Time (hours)

TPS

 PE1/TPS (38:62) PE1/TPS (68:32) PE2/TPS (69:31)

is very similar to that obtained by acid hydrolysis.

0

31 wt % TPS40 as a function of incubation time.

10

20

30

40

50

Weight Loss of TPS (

*%*)

60

70

80

90

100

and low molecular fractions become soluble, but the specimen shape remains intact. Enzymatic hydrolysis of insoluble polymers is known to be affected by the mode of interaction between the enzymes and the polymeric chains and typically involves four steps: (i) enzyme diffusion from the bulk solution to the solid surface, (ii) enzyme adsorption on the substrate, resulting in the formation of enzyme-substrate complex, (iii) catalysis of the hydrolysis reaction, and (iv) diffusion of the hydrolyzed fraction from the solid substrate to the solution (Azevedo et al., 2003). Blends with high loadings of TPS40 show an enzymatic degradation rate as fast as that of the raw TPS40 during the first 3 hours of exposure. This is probably due to the large amount of TPS40 observed on the surface of LDPE/TPS40 blends. Similarly, blends containing about 30% of TPS40 have less starch available on the surface and, consequently, the initial enzymatic degradation rate is slower than the others. As the soluble degradation products of TPS40 diffuse out of the sample, the number of active enzyme units available for starch degradation decreases resulting in a reduction of degradation rate. TPS40 is completely degraded in 36 hours, whereas PE1/TPS40 having 62% and 32% TPS40 and PE2/TPS40 compounded with 31% TPS40 reach their maximum degradation in 72 hours. Conversely, the 69:31 PE2/TPS40 stabilizes at a short period of about 20 hours, whereas the 68:32 PE1/TPS40 blends reaches its plateau at 48 hours. This is likely due to the connectivity of PE2/TPS40 (69:31) blend of starch from the surface; therefore the path of the enzyme is less obstructive.

#### **4.4.3 Microbial biodegradation**

Weight loss as a function of time is the most useful method employed to monitor biodegradation (Swanson et al., 2003; Bikiaris et al., 1997b). Figure 12 shows the weight loss of LDPE/TPS40 blends exposed to activated sludge as a function of degradation time. As expected, raw PE1 remains unchanged after 45 days. On the contrary, raw TPS40 is completely consumed within 21 days of exposure. For the LDPE/TPS40 blends, the maximum biodegradation extent is observed at times longer than the raw TPS40. If TPS40 particles are present only on the surface, and not interconnected with particles inside the LDPE/TPS40 blends, then it could be expected that starch domains would be completely biodegraded like the raw TPS40. Percent continuity observed in Figure 10 shows that TPS40 particles are interconnected one to another. At TPS40 concentration of about 30%, interconnection increases when the morphology of starch domains changes from spherical (PE2/TPS40 blend) to fiber-like particles (PE1/TPS40 blend). The extent of biodegradation of TPS40 at 45 days of extraction for PE1/TPS40 blends at 62%, 32% of TPS40 and PE2/TPS40 (69:31) were 92%, 39% and 22%, respectively. However, when the maximum biological extraction is compared with the maximum enzymatic degradation, important difference is noticeable, especially in blends with ca. 30 wt% TPS40.

Kinetic of biodegradation of TPS40 and LDPE/TPS40 blends shows two stages (Table 1). In all cases, there is a fast weight loss during the first 1.5 days, followed by another stage where biodegradation rate decreases progressively. The fast stage could be related to the combined effect of biodegradation and diffusion of glycerol and low molecular starch fractions out of the sample. Diffusion of water soluble components can be accelerated by starch swelling, as observed in raw TPS40, during the first 6 hr. Weight loss during this period is almost 4 times faster than the following 30 hr. In the case of LDPE/TPS40 blends,

Melt Blending with Thermoplastic Starch 19

Time (days) TPS40 PE1/TPS40 (38:62) PE1/TPS40 (68:32) PE2/TPS40 (69:31) 0.25 81.6 27.6 5.9 6.4 0.5 85.8 14.4 11.6 5.2 0.75 22.6 16.9 7.7 5.2 1.5 26.0 19.1 8.2 5.2 3 6.3 5.0 2.7 1.5 7 2.1 4.7 1.5 1.1 14 1.5 2.6 0.8 0.5 21 0.6 1.1 0.6 0.4 30 0.3 1.0 0.5 0.2 Table 1. Biodegradation rate for TPS40 and LDPE/TPS40 blends as a function of exposure

The analysis of thermal properties of water-free TPS materials prepared in a TSE showed that granular starch was completely disrupted and that TPS shows a thermal transition below room temperature corresponding to the glass transition temperature and this Tg is dependent on glycerol content. As was observed for the thermal properties, the rheological properties were also highly dependent on glycerol content. of TPS36 at shear rate ~ 130 s-1 decreases by 20% when the glycerol content is increased from 36 to 40%. In the same way, G' and G" also decrease as glycerol content increases. However, a particularly dramatic variation is observed when the glycerol content is varied from 29 to 33%. These latter results suggest a phase inversion from a hard elastic network matrix to a soft amorphous one. The glycerol plasticization threshold thus occurs at a content of approximately 30%. This result concerning a critical plasticization threshold is very important for morphology control

The PE/TPS blends prepared using the one-step process demonstrated levels of ductility and modulus similar to the virgin polyethylene even at very high loadings of TPS without the addition of any interfacial modifier. The excellent properties are a combination of both the melt blending process and a sophisticated morphology control. Through a control of the glycerol content and thermoplastic starch volume fraction, the above process can result in morphological structures, which run the full range of those observed in classical blends of synthetic thermoplastics. Spherical, fiber-like and co-continuous morphologies are observed. Control of the glycerol content of the starch allows one to control the properties of starch from that of a solid filler through to that of a highly deformable thermoplastic material. A wide range of potential properties can be exploited for this type

This material has the added benefit of containing large quantities of a renewable resource and hence represents a more sustainable alternative to pure synthetic polymers. Since the starch can be fully interconnected through morphology control, it is also completely accessible for biodegradation as opposed to the case of starch particles dispersed in a

dC/dt (g/l.days)

time in activated sludge.

**5. Conclusions** 

strategies.

of material.

synthetic polymer matrix.

starch swelling is limited by polyethylene matrix, which results in longer diffusion time. Decrease of biodegradation rate observed after 3 days could be explained by the lower degradability of TPS40 domains that remain in the material.

Fig. 12. Bacterial biodegradation kinetic expressed as weight loss for TPS40 (), PE1/TPS40 blends with: () 62 wt% TPS40, () 32 wt% TPS40 and PE2/TPS40 blends () 31 wt% TPS40 during exposure in activated sludge.

From comparison of the three degradation techniques, it can be inferred that some phenomenon is taking place during the bacterial degradation of LDPE/TPS40 blends. Weight losses for acid hydrolysis and biodegradation were 100% and 92%, 66 and 39%, and 38% and 22%, respectively for PE1/TPS40 (38:62), PE1/TPS40 (68:32), and PE2/TPS40 (69:31). In the case of PE1/TPS40 (38:62), the difference can be neglected due to the possibility of bacterial waste accumulation inside polyethylene cavities. At around 30% TPS40, however, differences are more prominent. This could be related to other phenomena. Micrographs of the surface of PE1/TPS40 and PE2/TPS40 blends (reported elsewhere) show that pores on PE1 matrix left after TPS40 extraction are below 1 m, while those observed on PE2 ranged between 3 to 10 m (Tena-Salcido et al., 2008). On the other hand, different microorganisms have a length between 0.4 and 14 m and width of 0.2 to 12 m (Gibbon, 1997). In the case of blends having about 30% TPS40, it is possible that microorganisms or their colonies can restrict starch diffusion by obstructing the polyethylene pores to result in a significant reduction of the final extent of biodegradation.

starch swelling is limited by polyethylene matrix, which results in longer diffusion time. Decrease of biodegradation rate observed after 3 days could be explained by the lower

0 7 14 21 28 35 42 49

Time (days)

Fig. 12. Bacterial biodegradation kinetic expressed as weight loss for TPS40 (), PE1/TPS40 blends with: () 62 wt% TPS40, () 32 wt% TPS40 and PE2/TPS40 blends () 31 wt%

From comparison of the three degradation techniques, it can be inferred that some phenomenon is taking place during the bacterial degradation of LDPE/TPS40 blends. Weight losses for acid hydrolysis and biodegradation were 100% and 92%, 66 and 39%, and 38% and 22%, respectively for PE1/TPS40 (38:62), PE1/TPS40 (68:32), and PE2/TPS40 (69:31). In the case of PE1/TPS40 (38:62), the difference can be neglected due to the possibility of bacterial waste accumulation inside polyethylene cavities. At around 30% TPS40, however, differences are more prominent. This could be related to other phenomena. Micrographs of the surface of PE1/TPS40 and PE2/TPS40 blends (reported elsewhere) show that pores on PE1 matrix left after TPS40 extraction are below 1 m, while those observed on PE2 ranged between 3 to 10 m (Tena-Salcido et al., 2008). On the other hand, different microorganisms have a length between 0.4 and 14 m and width of 0.2 to 12 m (Gibbon, 1997). In the case of blends having about 30% TPS40, it is possible that microorganisms or their colonies can restrict starch diffusion by obstructing the polyethylene pores to result in

 TPS 100*%* PE1/TPS (38:62) PE1/TPS (68:32) PE2/TPS (69:31) PE1 100*%*

degradability of TPS40 domains that remain in the material.

0

TPS40 during exposure in activated sludge.

a significant reduction of the final extent of biodegradation.

10

20

30

40

50

Weight Loss of TPS (*%* )

60

70

80

90

100


Table 1. Biodegradation rate for TPS40 and LDPE/TPS40 blends as a function of exposure time in activated sludge.
