A study on the use of phenoxy resins as compatibilizers of polyamide 6 (PA6) and polybutylene terephthalate (PBT).

INTRODUCTION

In the past few decades, the rate of introduction of new polymers in the commercial market has declined as a result of usually unfavorable economics. instead, polymer blending and alloying has become a favorite alternative for design and development of finished lightweight materials from commercially available polymer components (1). The focus of development of new lightweight plastic materials--often blends--is largely dictated by the automobile industry, which has been one of the largest consumers of polymers, with polymer-made components accounting for approximately 11% of the body weight and 25% of all the parts in a typical automobile (2).

An effort to design polymer blends through random selection of polymer pairs does not always guarantee success. Most polymer pairs are immiscible and produce incompatible blends with properties inferior to those of the component polymers. Consequently, extensive research has been conducted in the recent past in developing compatibilization strategies, in search of compatible polymer pairs, or finding immiscible blends with tailor-made compatibilizer packages (1, 3-5).

The present study was motivated by the increasing environmental awareness on the recycling of mixed plastic wastes of the end-of-life vehicles (ELVs), in view of the abundant use of PA6 and PBT in automotive industries. The European Commission has proposed regulations with set targets on reusability of plastic components on ELVs by the year 2015 and suggested various technologies for recycling of plastic wastes (6). Nevertheless, the current technologies for recycling of mixed plastic wastes from automobiles have not been fully developed due mainly to difficulties associated with sorting of an array of polymers differing in densities, chemical type, and often melting or processing temperatures. In addition, a consensus on industrial standards on what is acceptable and how the recycled polymers can be reused is still lacking.

The objective of the present study was to examine the merits of blending methodology as a means of recycling of plastic wastes. In this regard, we investigated the potential of phenoxy resins as compatibilizers of immiscible blends of polyamide-6 (PA6) and polybutylene terephthlate (PBT)--two engineering thermoplastic polymers with large volume consumption in automotive industry. Apart from large volume usage, the consideration of PA6 and PBT for the study also stemmed from the fact that a companion upstream density sorting operation can be easily developed to utilize the high values of specific gravity (~1.1) of these thermoplastic polymers so as to separate them from the polyolefins with specific gravity of 0.9-0.95, for example, through density segregation in water.

The routes of alloying and compounding for recycling of mixed plastic wastes can be easy and cost effective, as the existing compounding lines can be used without major modifications, provided a set of cheap compatibilizers is also available. However, commercially available compatibilizers can be expensive and in some cases may not exist at all, for example, in view of an array of different polymer types, e.g., polyesters, polyamides, polyolefins, ABS, acrylics, etc. encountered in mixed plastic wastes.

The phenoxy resins were chosen in this study for they are produced in large quantities commercially at lower costs, mainly due to substantial advances in the reactive extrusion technology (7). The phenoxy resins are derived as reaction products of diglycidyl ether of bisphenol A (DGEBA) with either bisphenol A (BPA) or monoethanol amine (MEA) and are thermoplastics in nature, leading to easy implementation in any existing blending or alloying scheme. They are known to be miscible and reactive with polyesters and acrylic resins through ester exchange reactions between the pendent hydroxyl groups of phenoxy and the ester linkages in the molecules of polyesters or acrylics (8-11). On the other hand, phenoxy resins form two-phase materials when blended with PA6; nevertheless, these blends showed improvement in the mechanical properties leading to a belief that some form of energetic interactions exists between these materials (12). Some recent studies reported the use of phenoxy resins as blend compatibilizers of Immiscible blends, for example in the blending of polysulfones with polyesters (13) and liquid crystalline polyesters with PET (14, 15). The use of non-conventional chemical compounds as compatibilizers is gaining interests in the literature as can be seen from some recent studies, whereby low molecular weight thermosetting epoxies have been used as compatibilizers of engineering polymers, such as PBT-polyphenylene ether (16) or PBT-polyarnide 6,6 (17, 18).

EXPERIMENTAL

Materials

PA6 was supplied by DuPont (Wilmington, DE) in the form of extrusion grade pellets, Zytel[R] 7301 NCO10, and PET was obtained from GE Plastics (Pittsfield, MA) as Valox[R] 315. The Dow Chemical Company (Midland, MI) supplied two amorphous phenoxy resins for use as potential compatibilizers. One phenoxy resin, polyhydroxy ether (PHE), derived from the reactions between DGEBA and bisphenol A, exhibited a glass transition ([T.sub.g]) temperature of 98[degrees]C, while the other phenoxy resin, polyhydroxyaminoether (PHAE) available as BLOX[R] 200 was a polymerization product of DGEBA and MEA and showed a [T.sub.g] of 70[degrees]C. Both phenoxy resins contain hydroxyl end groups (19). The details of synthesis and characterization methods for phenoxy resins are available elsewhere (7, 20). The chemical structures of phenoxy resins are shown in Fig. 1, and Table 1 presents some of the properties of the polymer components used in the study.

Two solvents, trifluoroacetic acid (TFA) and tetrahydrofuran (THF), were obtained from Aldrich for extraction studies. TFA dissolved PA6, PET, and PHAE, while PHE was found to be soluble only in THF. Some swelling of PHE was observed in TEA.

Mixing

The amounts of PA6 and PBT in all blends were maintained respectively at 70 and 30 parts by weight after a recent study (21) reported that this composition produces the best properties in this immiscible blend system. The amount of phenoxy resin was varied between 0, 5, 10, and 30 parts by weight.

The binary blends of phenoxy resins separately with PA6 and PBT were prepared in a batch mixer, the Brabender Plasticorder, with a rotor speed of 40 rpm and a total mixing time of 10 minutes. The binary blends containing PHAE and PHE resins were prepared respectively at 240[degrees]C and 255[degrees]C. A lower blending temperature in the case of PHAE was chosen to avoid thermal degradation at higher temperatures. However, the ternary blends of PHAE with PA6 (major) and PBT showed no signs of thermal degradation, such as abrupt increase of mixing torque or materials turning into powders, when mixed at 250[degrees]C for 10 minutes.

A 30-mm Japan Steel Works co-rotating twin-screw extruder was used to prepare the ternary blends of PA6, PET, and phenoxy resins with a feed rate of 2.5 kg/h and a screw speed of 70 rpm, giving a mean residence time of approximately 3 minutes. The component materials were vacuum dried at 100[degrees]C for 24 hours and tumble-mixed before blending. In the case of PHAE, the temperature profile in the extruder was set at 200, 220, 230, 245, and 240[degrees]C respectively between the feed zone and the die, which was changed to 200, 230, 240, 250, 255, and 250[degrees]C for PHE.

Molding

A Wabash [R] compression press was utilized to make 25-mm round discs for rheometry. The materials were molded at 250[degrees]C using a cycle time of 5 minutes. The sample specimens for the tensile and impact tests were injection molded using a BOY [R]-15 injection molding press equipped with air-cooled mold with the temperature setting at 225, 250, and 255[degrees]C in the three zones (feed to nozzle) of the molding machine. All blend materials were dried at 100[degrees]C for 24 hours in a vacuum oven before molding.

Thermal Analysis

All DSC measurements were carried out using a TA Instruments 2910 Thermal Analyzer system at a heating rate of 20[degrees]C/min from -20[degrees]C to 250[degrees]C.

FT-IR Analysis

A Genesis [R] Series FTIR spectrophotometer from ATI Mattson was utilized to obtain IR spectra of the various materials. The binary blends of PA6 and PBT with PHAE were dissolved in TFA and thin films were cast on KBr discs. In other cases, powdered samples were generated through filing of solid polymer blocks, mixed with KBr powder, and formed in to disks for analysis.

Morphology

The blend morphology was probed using Hitachi S-2150 scanning electron microscope (SEM); injection and compression molded sample specimens of respectively ternary and binary blends were used for the purpose. The specimens for morphological study were prepared by cold fracture in liquid nitrogen and etching in oxygen plasma for 7 minutes. The stability of the dispersed phase was inferred from the inspection of SEM micrographs of specimens annealed at 250[degrees]C for 1 hour in a vacuum oven.

Rheology

The rheological properties of the component polymers, except PHAE, and their blends were evaluated using Rheometrics ARES [R] rheometer. All measurements were carried out at 250[degrees] under nitrogen environment using a 25-mm parallel plate setup with a gap of 1.2 mm. Steady rates of shear in the range of 0.01 to 10 [s.sup.-1] were used in all the cases. An elapsed time of approximately 15 minutes was needed in each experiment for preheating of the samples to the desired temperature, for adjustment of the gap, and for removal of the excess materials from the test chamber.

Mechanical Testing

The tensile properties of the blends and the component polymers were evaluated according to ASTM D638 method using Instron tensile tester, while notched Izod impact strength was determined according to ASTM D256 method. The sample specimens were dried in vacuum oven at room temperature for 24 hours before evaluation of the mechanical properties. The relative humidity of the test room was observed to be 50% during all measurements.

RESULTS AND DISCUSSION

PBT-Phenoxy and PA6-Phenoxy Reactions

The Interactions between the phenoxy resins and PBT or PA6 were studied using the binary blends. The first indication of possible reactivity between PBT and the phenoxy resins was observed from the continued increase of torque values during mixing of these components in the Brabender Plasticorder, as presented in Figs. 2 and 3. In addition, the blends thus produced appeared transparent in the molten state, indicating complete miscibility between the phenoxy resins and PBT at the blending temperature. The binary blends of both phenoxy resins with PA6, on the other hand, appeared opaque in the molten state, indicating a two-phase nature at the blending temperature. The mixing torque for PA6-PHE system did not show an increase with mixing time (Fig. 4), thus ruling out the possibility of appreciable chemical reactions between the components.

In the case of PHAE resin, however, the mixing torque increased in two stages during blending with PBT, first after about 200 seconds of mixing and then again after about 450 seconds (Fig. 3). In addition, the mixing torque reached a plateau with the continuation of mixing to about 720 seconds. At that stage, the materials turned into semi-solid masses, and on continued mixing they turned into powders. While the first stage of torque increase after about 200 seconds can be attributed to the reactions between PBT and PHAE, the rapid increase of torque beyond 450 seconds and the materials turning into powdery mass can be explained by the formation of crosslinked products by the molecules of PHAE resin (7). The poor thermal stability of PHAE resin manifested again during mixing with PA6 (Fig. 5)--although no appreciable chemical reactions were expected-and in rheological measurements involving neat PHAE resin and its binary blends with PBT. It was clear from the mixing studies that PHAE resins must not be expos ed to heat at 240[degrees]C for longer than about 450 seconds.

The possibilities of chemical reactions between PBT and the phenoxy resins and the interactions between the phenoxy resins and PA6 were further analyzed using FT-IR spectroscopy and by extraction in solvents. For this purpose, a blend containing 70 parts of PBT and 30 parts of PHE (by weight), mixed at 255[degrees]C for 10 minutes in the Brabender Plasticorder, was considered as an example and the extraction study was carried out with TFA and THF using the following procedure:

(a) A small quantity (1.218 g) of the blend was added to 25 ml of TFA and a turbid solution was obtained with some fine particles of solids suspended in It.

(b) The turbid solution turned clear with the addition of 75 ml THF and a precipitate formed. The precipitate was retained in a ceramic filter and the liquid was evaporated to recover the solids dissolved in it.

(c) The precipitate was further washed with an excess amount of THF and the filtrate was saved for recovery of the dissolved materials.

(d) The precipitate and the solids recovered from the evaporation of the liquids were analyzed by DSC and FT-IR.

It was found that the precipitate accounted for approximately 85 wt% of the starting blend, which contained 70 wt% of PBT. Therefore, a part of the PHE content of the starting material ended up in the precipitate, purportedly through the formation of grafted copolymer with PBT, as already indicated by other investigators (10, 14, 15). A possible mechanism of the formation of these copolymers is presented in Fig. 6. The molecules of grafted copolymers are formed through the ester exchange reactions between the ester groups of PBT chains and the hydroxyl groups of the phenoxy resins. Figure 7 presents FT-IR diagrams of the materials contained in the precipitate and those recovered from the liquids in the extraction study. It is seen from the relative peak heights at 1717 [cm.sup.-1] (C = O stretching of PBT) and 1609 [cm.sup.-1] (C-O-C of PHE) in Fig. 7 that the precipitate is rich in PBT. As PHE was found to be insoluble in TFA, the precipitate must be constituted of any unreacted PBT and grafted copolymers of PBT and PHE formed through ester exchange reactions. The precipitate shows a [T.sub.g] of 56[degrees]C and a melting point of 222[degrees]C, as shown in Fig. 8, which corroborate the findings from FT-IR that grafted copolymers were formed. On the other hand, the solids recovered from the liquid extract showed the opposite behavior, i.e., it is richer in PHE content, e.g., from the comparison of peak heights at 728 [cm.sup.-1] corresponding to -[CH.sub.2]- stretching of PBT and at 828 [cm.sup.-1] corresponding to aromatic -CH stretching of PHE. However, the DSC trace of the solids recovered from the liquids shows a weak glass transition at 4700; also, the melting peak corresponding to PBT is absent. A similar analysis revealed that approximately 77% of the starting materials mixed at 255[degrees]C for 5 minutes ended up as precipitate.

The results of soxhlet extraction for PA6-PHE blends are presented in Table 2, which show that only a small amount of PA6 is retained after the first extraction with TFA. In view of this observation and a stable mixing torque during blending (Fig. 4), it can be inferred that PA6 and PHE did not react with each other and a small amount of PA6 in the residue can be thought of as inclusions inside the PHE-phase domains, which could not be extracted by TFA.

Since PBT, PA6, and PHAE were soluble in a common solvent TFA, the binary blends containing PHAE were dissolved in TFA for preparation of thin films for FT-IR analysis on KBr disc. It is observed in Fig. 9 that the spectra for PBT-PHAE system almost remained unchanged after the mixing operation except in the re-Won of wavenumber 1600-1800 [cm.sup.-1]. A careful examination of this region (Fig. 9a) shows that the peak corresponding to wavenumber 1720 [cm.sup.-1], which is also the characteristic peak of the carbonyl group stretching, was found to be shifting to lower wavenumber 1680 [cm.sup.-1] upon mixing. Similar results were also obtained by Eguiazabal et al. (22) for blends of PET and polyarylate, and it can be argued that this shift in carbonyl peak is due to hydrogen bonding type of interactions (23). However, careful examination of the spectra, especially in the region of 3500-3000 [cm.sup.-1], attributed to hydrogen bonding (Fig. 9b), revealed no shifts upon mixing. Therefore, hydrogen bonding involvin g the carbonyl groups of PBT can be ruled out, and the shift in carbonyl stretching can be attributed to the formation of copolymers of PBT and PHAE. Similar shifts, however, were not observed in Fig. 7 for PBT-PHE blends.

The PA6-PHAE blends also show a shift in the carbonyl stretching frequency to lower values as presented in Fig. 10a. However, this change is associated with a concurrent shift in the hydroxyl group frequency in the vicinity of 3300 [cm.sup.-1] as well as 3000 [cm.sup.-1] (Fig. 10b), lending support to the formation of hydrogen bonding. Nevertheless, the strength of hydrogen bonding, as indicated from the magnitude of the shift in the region of 3000-3500 [cm.sup.-1], was not sufficient to cause miscibility between the blend components. Consequently, blends appeared opaque at the blending temperature and retained their two-phase nature in the solid state, as evident from the SEM micrographs presented in Fig. 11.

Rheology

The shear viscosity of PA6-PBT blends with and without PHE and PHAE phenoxy resins at 250[degrees]C is presented in Figs. 12 and 13. It is observed that the values of shear viscosity of blends increased with phenoxy resin content for both types of phenoxy resins. In addition, the zero shear viscosity of all the ternary blends with phenoxy resins was found to be higher than those of PA6-PBT blends and the component polymers. Shear viscosity data for PHAE resin was not obtained, as an approximate elapsed time of 15 minutes, needed for stabilization of the temperature in the rheometer, was long enough for PHAE resin to degrade. The values of zero shear viscosity of the blends containing PHAE resin were found to be considerably higher than those of the blends containing the PHE resin; the higher reactivity of PHAE towards PBT is the primary reason.

Morphology

The results of morphological characterization by SEM are presented in Figs. 14 and 15. As expected, the domain sizes of the dispersed PBT-phase in the compatibilized blends are found to be much smaller compared to the uncompatibilized blends. The mean size of the dispersed PBT-phase is found to be less than 1 [mu]m and between 2 and 4 [mu]m, respectively, for compatibilized and uncompatibilized blends. In addition, the dispersed phase size distribution is much narrower in the compatibilized blends. However, the typical size of the dispersed PET domains remained insensitive to the amount of phenoxy resins used. Also in the case of blends with 30 phr of PHE, there Is an apparent growth in the phase size, which can result from the high viscosity of the dispersed phase, in this case a mixture of PET and the graft copolymer of PET and PHE. The concentration of the copolymers is thought to be much higher because of the reactions between PBT and PHE, which were present in a 50:50 weight ratio. The ternary blen ds of PHAE were observed to contain dispersed domains of much smaller size, although the reasons were not thoroughly investigated in this study.

The morphology of the ternary blends was found to be stable to annealing, except in the case of blend containing 5 phr of PHE resin. In the latter case, the PBT-phase domains, although they grew upon annealing, were found to be much smaller compared to those in the annealed uncompatibilized blend.

In the light of the absence of reactivity between PA6 and PHE or PA6 and PHAE, the phase stability seen in Figs. 14 and 15 can be interpreted possibly in terms of other interactive forces, such as hydrogen bonding between the molecules of PA6 and PHAE (as inferred from Fig. 10b) or PHE, even if these materials form two-phase structures as shown in Fig. 11. An FT-IR study on uncompatibilized PA6-PBT blend did not reveal any shift in the zones of carbonyl stretching, thereby ruling out the possibility that hydrogen bonding was already present in uncompatibilized blends.

Mechanical Properties

The values of the tensile modulus and the yield stress of the blends, presented in Figs. 16 and 17, respectively, showed an appreciable increase with the increase of PHE or PHAE content in the blends. The ternary blends containing PHE showed a weak influence of the PHE content beyond 5 phr, while the blends containing PHAE resins showed gradual increase in the properties with the increase of PHAE content. However, it is important to note that both the values of the tensile modulus and the yield stress of the blends containing phenoxy resins are higher than those of the component polymers, i.e., PET and PA6. This means that these compatibilized blends can offer strength equal to or better than those of the component polymers as well as their uncompatibilized blends. This can be considered a positive attribute of such blending schemes for recycling of mixed plastic wastes containing PA6 and PBT.

Notched Izod impact properties of the ternary blends and the component polymers are presented in Fig. 18. The blends containing PHE do not show much improvement over the uncompatibilized blends except at 30 phr loading of PHE. while a substantial improvement in the impact strength was obtained in ternary blends containing PHAE. Such behavior can be attributed to the morphology of blends containing PHE and PHAE resins, for example, smaller size of PBT domains were obtained in PHAE-compatibilized blends than those compatibilized by PHE resin.

CONCLUSIONS

It is shown that an incompatible blend system of PA6 and PBT can be easily compatibilized using two thermoplastic phenoxy resins. The compatibility in these blends was derived from in situ formed graft copolymers of PBT and the phenoxy resins, which also formed energetic interactions with the PA6-phase through hydrogen bonding. The stability of morphologies to annealing and much improved tensile and impact properties of the ternary blends of PA6, PBT, and PHE/PHAE resins provide indications that blending can be a feasible route for recycling of polyesters and polyamides. The phenoxy resins, with an average price in the same range as polyesters, can be considered as cheap compatibilizers for immiscible polymer blends of polyesters and polyamides and for recycling of mixed plastic wastes containing polyesters and polyamides.

[FIGURE 2 OMITTED]

[FIGURE 3 OMITTED]

[FIGURE 4 OMITTED]

[FIGURE 5 OMITTED]

[FIGURE 7 OMITTED]

[FIGURE 8 OMITTED]

[FIGURE 9 OMITTED]

[FIGURE 10 OMITTED]

[FIGURE 12 OMITTED]

[FIGURE 13 OMITTED]

[FIGURE 16 OMITTED]

[FIGURE 17 OMITTED]

[FIGURE 18 OMITTED]

Table 1

Properties of the Polymers Used in the Study.

Property                                  Unit   PA6      PBT     PHE

Density                                   g/cc  1.13     1.31    1.18
Glass transition temp. ([T.sub.g])  [degrees]C    47    40-45      98
Melting point ([T.sub.m])           [degrees]C   220  220-250      --
Molecular weight ([M.sub.w])                           26,500  50,000

Property                            BLOX[R]200

Density                                1.20
Glass transition temp. ([T.sub.g])       70
Melting point ([T.sub.m])                --
Molecular weight ([M.sub.w])         45,000

Table 2

Results of Sepration Study of PA6-PHE Blends. Extraction Was Carried Out
With TFA.

                                  Weight  Weight
                           Total  of PA6  of PHE
                            (g)    (g)     (g)

Before extraction          3.049  2.133   0.916
After extraction with TFA  1.127  0.211   0.916
% Retention                  37     10      100

ACKNOWLEDGEMENT

Financial assistance from The Dow Chemical Company in the form of a fellowship to D. Dharaiya is gratefully acknowledged.

REFERENCES

(1.) L. A. Utracki, Commercial Polymer Blends, Chapman & Hall, London (1998).

(2.) J. Maxwell, Plastics in Automotive Industry, Woodhead Publishing, Cambridge (1994).

(3.) M. Xanthos and S. S. Dagli, Polym, Eng. Sci., 31, 929 (1991).

(4.) M. J. Folkes and P. S. Hopes, eds., Polymer Blends and Alloys, Blackie Academic and Professional, London (1993).

(5.) S. Datta and D. J. Lohse, Polymeric Compatibilizers: Uses and Benefits in Polymer Blends, Hanser Publishers, New York (1996).

(6.) K. Bellmann and A. Khare, Technovattion, 19, 721-734 (1999).

(7.) H. Silvis and J. E. White, Polym. News, 23, 6-10 (1998).

(8.) L. M. Robeson and A. B. Furtek, J. AppL Polym. Sci. 23, 645-659 (1979).

(9.) R. W. Seymour and B. E. Zehner, J. Polym. Set, Polym. Phys. Ed., 18, 2299-2301 (1980).

(10.) J. Eguiazabal and J. Nazabal, J. Mater. Sci, 25, 1522-1528 (1990).

(11.) R. Erro, M. Gatzelumendi, and J. Nazabal, J. Appl. Polym. Sci, 45, 339-348 (1992).

(12.) J. Egulaxabal, J. Nazabal, and G. Gurrica-Echevarria, J. Appl. Polym. Sci., 72, 1113-1124(1999).

(13.) L. Mascia and L. Martin, High Perform. Polym., 8. 119-131 (1996).

(14.) A. Bruggeman and A. Tinnemans, J. Appl. polym. Sci. 71, 1107-1123 (1999).

(15.) A Bruggeman and A. Tinnemans, J. Appl. Polym. Sci., 71, 1125-1 131 (1999).

(16.) S. C. Jana, N. Patel, and D. Dharaiya, Polymer, 42, 8681-8693 (2001).

(17.) C. C. Huang and F. C. Chang. Polymer, 38, 2 135-2141 (1997).

(18.) C. C. Huang and F. C. Chang, Polymer. 38, 4287-4293 (1997).

(19.) A. Shafi. The Dow Chemical Company, Freeport, Texas (private communications).

(20.) H. Silvis, J. Bernnan, and J. White. U.S. Patent. 4,647,648 (1987).

(21.) F. C. Chang and K. Chiou, J. Polym. Sci., Part B Polym. Phy., 38, 23-33 (2000).

(22.) J. Eguizabal, M. Cortazar, J. Iruin, and G. Ucar, Polymer. 27, 2013-2018 (1986).

(23.) D R. Paul, W. H. Jo. and C. A. Cruz, J. Polym. Sci., Part B Polym. Phys., 27. 1057-1076 (1989).

SADHAN C. JANA *, and ASJAD SHAFI **

* corresponding author at [e-mail]

** The Dow Chemical Company Freerport, TX 77541