Sustainable polymer composites: immiscible blends prepared by extrusion of poly(trimethylene terephthalate) and polyamide6,10 with high bio-based content

Abstract

Components for binary polymer blends were sought to produce an immiscible blend of improved renewable character and with good structural properties. The poly(trimethylene terephthalate) and polyamide6,10 system was selected based on the molecular structure of the molecules and the bio-based origin of the feedstocks. A preliminary study of three compositions in this system demonstrated the similar thermal properties of the two polymers as measured by differential scanning calorimetry (DSC) and the ability of these polymers to be processed together in conventional extrusion equipment to produce blends with micrometer-scale domains. Dispersed phases were observed by electron microscopy near the end members. Available viscosity data and the appearance of columnar blends at the 50/50 composition suggest the possibility of co-continuous blends in close proximity to this composition.

Keywords:

• bio-based polymers
• polymer blends
• sustainable material development
• sustainability
• green engineering

1. Introduction

Polymers based on renewable feedstocks are of increasing interest due to the sustainable nature of these raw materials and the need to reduce our ever increasing reliance on petroleum.

The blending of immiscible polymers via specialised melt processing either with or without compatibilisers is an important and productive approach to the development of composites with targeted properties from existing polymer components. This approach reduces the need to engineer new polymers from the synthesis stage (Reignier and Favis 2000 cited Giancola and Lehman 2009), a long and costly development effort usually limited to well-funded research efforts at large corporations. Successful immiscible polymer blending has produced composites with either rule of mixture properties or synergistic properties in excess of the linear relationship (Giancola and Lehman 2010).

The goal of this study was to make a preliminary investigation into the use of immiscible polymer blending to improve the bio-based content of a relatively new polymer, poly(trimethylene terephthalate) (PTT), while maintaining or increasing the engineering and processing properties. Overall, PTT is a polymer with high elastic modulus (Table 1) but with a limited yield strain and a tendency towards brittle fracture. Thus, PTT is an excellent load-bearing polymer but without ductility or high toughness. polyamide6,10 (PA6,10), on the other hand, has 10–20% lower tensile and flexural modulus compared to PTT and about 25% lower yield stress, yet it exhibits nearly 65 times more strain at break and approximately twice the impact strength of PTT. By combining these two materials, we seek an engineering polymer with good load-bearing capacity as well as the high yield strain and impact strength that constitute a commercially valuable engineering polymer.

Table 1 Selected mechanical properties of the component polymers.

Property	PTT	PA6,10
Yield stress (MPa)	64	47
Yield strain (%)	4
Strain at break (%)	4	258
Tensile modulus (MPa)	2600	2315
Flexural modulus (MPa)	2400	1900
Notched charpy impact strength (kJ/m^2)	3.0	5.6

Empirical viscosity–volume fraction relationships were used to determine the co-continuous region of the blended polymers so that interconnected morphologies could be developed where mechanical grafting (Joshi et al. 2006) provides good load transfer even in the absence of chemical bonding (Thirtha et al. 2006 cited Giancola and Lehman 2010). Highly co-continuous structures are useful as tissue engineering scaffolds, for electrical conductivity and dielectric media, and in drug delivery devices (Zhang et al. 2007). The PTT/PA6,10 blends are envisioned to find application as structural materials in numerous areas of application, such as transportation, aerospace, home appliances, and various consumer and industrial products.

2. Methodology

The thermoplastic polyester PTT is a comparatively new polymer with properties similar to the poly(ethylene terephthalate) (PET), the well-known polyester which has been used commercially for decades in fabrics, carpeting, apparel and packaging. PTT is used in similar applications but contrasts in that PTT exhibits superior elastic recovery compared to PET (Ward and Wilding 1976 cited Kurian 2005). One of the most prominent features of PTT is that it possesses partial renewable content, arising from the fact that it is derived by reacting terephthalic acid with 1,3propanediol (trimethylene glycol). The diol part of 1,3propanediol is derived from corn sugar and is the contributing factor to make PTT a bio-based polymer with a renewable fraction of approximately 36% (Kurian 2005 cited Bhatia and Kurian 2008).

A good matching polymer for blending with PTT will have bulk engineering properties that enhance PTT, functional pendant groups that are just partially compatible with PTT, a high bio-based content and viscosity/temperature characteristics that allow the two polymers to be processed together. After reviewing numerous polymers, the polyamide family was selected and PA6,10 in particular was picked as a good candidate for immiscible polymer blending with PTT. The polyamide polymers exhibit good engineering properties, such as superior toughness, strength and abrasion resistance, (Hu et al. 2006) which will complement PTT.

PA6,10 currently finds uses in monofilament form in applications such as bristles and brushes (Hu et al. 2006). The molecular configuration of PA6,10 provides an ideal level of compatibility with the PTT molecule. The goal for good immiscible blend processing is to have two polymers with sufficient cohesive energy density differences, as measured by solubility parameter, to make the polymers immiscible, but with sufficient similarity in cohesive energy density and dipole character to enable a modest level of compatibility in the interfacial regions between domains. The molecular structures of PTT and PA6,10 (Figure 1(a),(b)) illustrate the similarities and differences. PTT is a thermoplastic aromatic polyester with 11 carbon atoms in the repeat unit, whereas PA6,10 has 16 carbon atoms in a longer linear backbone. However, both molecules have four high-electronegativity atoms (nitrogen or oxygen), two in the backbone and two as pendant carbonyl oxygen. This combination of different backbone repeating units but with a similar distribution of high electronegativity atoms in the backbone and pendant groups provides a level of dipole–dipole interaction and compatibility that enhances interfacial secondary bonding in the blends. From a more macroscopic perspective, the solubility parameter of PTT and PA6,10 calculated from the Hoy group contribution method (Billmeyer 1984, Jang et al. 2005, Gupta et al. 2011) are 35.4 and 21.3 MPa^1/2, respectively. The polymers are suited for immiscible blending since their solubility parameters differ by 14.1 MPa^1/2, greater than the generally accepted Δδ>10 MPa^1/2 (Forster et al. 2001 cited Gupta et al. 2011) needed to create immiscibility, but not so great as to preclude secondary bonding in the interfacial regions. Ideally, one should compare solubility parameters measured at the processing temperatures of the blend. However, solubility parameters are typically measured near room temperature and scant literature exists (King 1995, Papadopoulou and Kalfoglou 1997, Dritsas et al. 2009) regarding their elevated temperature values or activation energies. The available data show that values decrease with increasing temperature for all common polymers. Considering that the values for PTT and PA6,10 must also both decrease with temperature, the difference between the two values is expected to change little due to the similarity in the dipole character of the polymers. In any event, experience in our laboratory has shown that published room temperature solubility parameters are an effective means for predicting the immiscibility of binary polymer blends at melt temperatures.

Figure 1 (a) The structure of PTT. (b) The structure of PA6,10.

With regard to producing an immiscible blend with PTT that has increased bio-based content, the companion polymer must have greater bio-based content. PA6,10 is a good choice in this regard since approximately 60% of PA6,10 can originate from renewable sources. The polymer is synthesised by reacting hexamethylenediamine with sebacic acid (dicarboxylic acid). The renewable portion of this polymer is attributed to sebacic acid which is a derivative of castor oil. Castor oil is a by-product of the castor seed, extracted from the seed of Ricinus communis (Kirk-Othmer 1979 cited Ogunniyi 2006). Castor plants grow prolifically in tropical and subtropical climates and their toxicity as a food substance makes the plants readily available for non-food uses (Ogunniyi 2006).

Melt processing of thermoplastic polymers to produce a two-component immiscible polymer blend requires specific viscosity temperature characteristics. Although this study sought to simply demonstrate that the two selected polymers can be co-processed by extrusion to produce an immiscible polymer blend with domains at the micrometer level, the general guiding principles of co-continuous immiscible blend processing were followed. Such principles include empirical relationships that relate the viscosity (η) of each polymer at the processing temperature to the necessary volume fraction (Φ) (Kurian 2005). The Jordhamo relation (Equation (1)) is one such relationship and was used in this study.

This relationship indicates that co-continuity occurs when the volume fraction ratio of the blend is approximately equal to the viscosity ratio at the shear rate and temperature of the extrusion processing. Co-continuity provides several important benefits to immiscible polymer blends. The intricate intertwined structure provides a useful morphology for various functional composites, such as the unique channels and cavities required in biomaterials for implants and scaffolding, and for structural applications where the intimate micro-scale contact between the polymers promotes good load transfer. Some investigators have used various compatibilisers to promote interfacial adhesion (Raquez et al. 2008), whereas our group prefers to rely on mechanical grafting, the intimate contact and load transfer that occur when immiscible blends are melt processed under conditions of high shear and extended residence time. While both approaches produce good results, the avoidance of compatibilisers circumvents the addition of extra components that may degrade some properties or be an issue in biomaterial applications where regulatory approval is required.

In a comprehensive assessment of any immiscible blend system, a full rheology work-up is needed spanning all shear rates and temperatures. In this preliminary study, we approximated values based on supplier data and differential scanning calorimetry (DSC) analyses of melting points described in the materials characterisation section.

3. Experimental

3.1 Materials selection and characterisation

The PTT and PA6,10 used in this study were obtained in pellet form from commercial suppliers as indicated in Table 2. PTT is a semicrystalline polymer and the pellets are opaque with a density of 1.33 g/cm³. PA6,10 is also semicrystalline and the pellets are translucent with a density of 1.08 g/cm³.

Table 2 Materials selection.

Material	Product	Supplier	Density (g/cm^3)
Poly(trimethylene terephthalate)	Sorona® 3001	DuPont	1.33
Polyamide6,10	ZYTRSLC3090	DuPont	1.08

Although some thermal data are available from the suppliers for these commercial products, we sought to confirm the properties via differential scanning calorimetry (DSC). Approximately 10 mg specimens of each polymer were run on a TA Instruments DSC Q1000 from ambient to 275°C at a heating rate of 10°C/min. After a brief hold at 275°C, additional enthalpy data were collected on cooling at 5°C/min. The analyses were replicated three times and averaged, although only miniscule variability was observed between runs. Results are shown in Figure 2(a),(b). The PTT polymer melted on heating at 225°C and on cooling at 166°C, a difference of 59°C, a comparatively large value indicating that a substantial amount of undercooling is required to overcome the kinetic barrier to crystallisation. The comparatively low amount of crystallisation that occurs during cooling, approximately 27% of the crystallinity revealed during melting as measured by integrated enthalpy ratios, again suggests a relatively viscous material at this temperature. The PA6,10 melting point, as measured during heating is 218°C and upon cooling is 183°C, a somewhat smaller difference of 35°C. From enthalpy ratios, 47% of the pellet crystallinity is recovered during cooling. The lesser difference in temperatures and the greater crystallisation of the polyamide suggests a lesser kinetic barrier to crystallisation compared to the PTT and possibly a lower viscosity at the melting point. Regardless of the exact behaviour, this simple differential scanning calorimetry (DSC) analysis suggests that both polymers have similar melting points (PTT heat/cool mean = 196°C; PA6,10 heat/cool mean = 201°C) and can likely be processed satisfactorily together as an immiscible polymer blend at temperatures nominally above the melting point of both polymers. Melting temperatures (heating) from supplier data for PTT and PA6,10 are 228 and 218°C, respectively, in good agreement with the 225 and 218°C values measured in this study.

Figure 2 (a) DSC trace of PTT. (b) DSC trace of PA6,10.

3.2 Formulation and blending

Several simple blend compositions were formulated to confirm that these polymers could be melt processed together and to demonstrate the degree of immiscibility in this system. Furthermore, we sought to identify the degree of dispersion and domain size near the end members and at the 50/50 composition. Batch formulations are given in (Table 3) both in volume and weight percentages. Both the PTT and PA6,10 pellets were dried under vacuum at 120 and 80°C, respectively, for 12 h. The batch formulations were weighed and mixed to a homogeneous state and either immediately melt processed or stored in a hermetically sealed glass jar held at 80°C until just before the composition was extruded.

Table 3 Formulations used in extrusion.

Extruded blends
Percent by volume (Weight)			Extruded thermal profile (°C)
Blend	PTT	Polyamide6,10	Solids/trans/metering/die
1	20(23.5)	80(76.6)	215/230/250/245
2	50(55.2)	50(44.8)	230/250/255/235
3	80(83.1)	20(16.9)	200/230/250/245

3.3 Extrusion

The pelletised blends were processed using a Brabender Intellitorque 0.75″ 30:1 single screw extruder with four controlled heating zones over all processing zones; the conveying zone, the transition zone, the metering zone and the die. Temperatures for these zones varied considerably and were adjusted empirically to enable good extrusion (Table 3). Most critical was adjustment of the solids conveying zone temperature to enable good wetting of the extruder wall to provide sufficient shear stress for extrusion. Exit temperatures at the die were controlled to give good extrudate surface texture without melt fracture or excess fluidity. After processing, specimens were cryo-fractured in preparation for SEM analysis.

3.4 Image analysis

Cold-fractured specimens from the various formulations were analysed using the Zeiss Sigma FESEM, Carl Zeiss Microscopy, Thornwood, New York, USA. All specimens were coated with 5 nm of iridium to prevent charging. Images were collected at a working distance of 14 mm and an accelerating voltage of 5 kV. SEM analysis was important in observing the microstructure present to determine the immiscible morphology generated by the extrusion process.

4. Discussion

PTT and PA6,10 are well suited as immiscible blend components. The polymers have complimentary properties that together provide a superior engineering polymer property profile. Furthermore, the polymers have similar melting points (PTT heat = 225°C; PA6,10 heat = 218°C) and their molecular structures and solubility parameters suggest that the immiscible polymer interfaces will be partially bonded by secondary chemical bonds.

In this preliminary study of this novel system, three compositions were processed and the extruder operation was adjusted empirically to produce good extrudates. In future work, rigorous rheological characterisation will guide composition and processing parameters as a region of co-continuity is sought. Nonetheless, the three compositions prepared in this study processed well and the morphology of the three blends shows promising immiscible behaviour as illustrated in the electron micrographs. The 20/80 and 80/20 blend compositions (Figures 3 and 4) illustrated strong dispersed domain behaviour, as expected since the thermal properties of the components are not extremely different. The domain size of the dispersed phase is approximately one micrometer, indicating a good level of dispersion under the processing conditions used. The cold fracture image reveals hemispherical depressions and full dispersed spheres on the surface, showing that the interphase boundary is only weakly bonded. The 50/50 blend (Figure 5) shows a somewhat oriented structure with columnar, fibrous, fine domains of one polymer in a matrix of the second. The size of these columnar elements varies in the range of approximately 2–5 μ, consistent with typical processing morphologies of immiscible polymer blends. Although this is not a co-continuous structure, our experience with systems of this type suggests that columnar structures such as that shown in Figure 5 are often close, typically within 10% points, of the co-continuous composition.

Figure 3 FESEM image of single extruded 80% PA6,10 and 20% PTT blend by volume.

Figure 4 FESEM image of single extruded 20% PA6,10 and 80% PTT blend by volume.

Figure 5 FESEM image of single extruded 50% PA6,10 and 50% PTT blend by volume.

In the examination of the 50/50 blend cold fracture surface, the polymers are present in equal quantities and thus the identity of each phase is not clear without further analysis. However, since the appearance of a columnar semi-dispersed phase precedes the appearance of the co-continuous region, the phase inversion composition for this system will be at a higher volume fraction of the columnar phase polymer. Using the Jordhamo relationship (Equation (1)) for guidance and inserting available viscosity data (T = 260°C, γ(s^− 1) = 100; η = 565 Pa s for PTT and 900 Pa s for PA6,10) (Kurian 2005), the co-continuous region is predicted to be near 60% PA6,10. Thus, if the Jordhamo relationship is correct, the columnar fibres in Figure 5 are PA6,10 and the matrix is PTT. Both PTT and PA6,10 are thought to have viscosities that sharply decrease with temperature under these processing conditions and future efforts in rheological characterisation will more fully elucidate this behaviour and determine the quantitative applicability of the Jordhamo relationship to this system.

5. Summary and conclusions

PTT and PA6,10 were thermally co-processed in a single screw extruder to form micrometer-scale immiscible blends. Dispersed blends were observed near each endpoint in this binary system and a columnar-type blend was observed at the 50/50 volume fraction composition, suggesting that co-continuous blend structures of 2–5 μm domain size are achievable near this composition. The ability to process these polymers together is significant since blends of these polymers appear compatible based on solubility parameter and the electronegativity of nitrogen and oxygen atoms in the backbone and pendant groups of each polymer. Most importantly, blends in this system offer the opportunity for highly renewable materials via bio-based fractions approaching 50% and the potential for maintaining or improving the mechanical properties of the blend over that of the component polymers.

Acknowledgements

The authors wish to acknowledge the funding and support of the AMIPP Advanced Polymer Center at Rutgers University that made this study possible.