Performance blends based on recycled polymers.
The steady growth in the use of plastic materials in packaging applications has caused, in recent years, an increasing concern about the environment and the problem of solid waste disposal. Post-consumer recycling of plastics has therefore become not only necessary but also increasingly mandatory. The growth in the use of poly(ethylene terephthalate) (PET) in beverage bottles has spurred the initial major interest in the post-consumer plastics recycling. Recycling of PET has thus evolved into a fairly successful industry, with more than 200 Mt/yr capacity in the U.S. alone (1). The cost advantage of recycled PET, at a quality comparable with that of virgin PET, has enabled its reuse in several applications such as fiberfill carpet yarn, films, bottles, and molding compounds.
Post-consumer recycling of other single resins has also become commercially feasible, wherever identification and sortation was possible. Significant amounts of high-density polyethylene (HDPE) from milk and detergent bottles and polypropylene (PP) from car battery cases are currently recovered. Even the PP from barrier packaging is also beginning to be recycled, although it contains some ethylene-vinyl alcohol copolymer (EVOH) and polyethylene.
The trend in plastics recycling is extending also to some engineering resins. For example, recycled polycarbonate (PC) from water bottles and compact disc processing is becoming available for reuse. Similarly, polyamides (PA-6 and 6,6) may also be potentially recycled from carpets and film scrap. With the continued refinements in the quality and cost of recycling, recycled polymers may gain significance as low-cost feed stocks for developing new high performance blend products. Increased usage of recycled polymers in new, value-added products and applications will in no doubt provide further economic incentive for recycling.
In general, high performance properties such as high impact strength, ductility, and solvent and heat resistance are highly desired in a polymer blend to enhance its value and application potential. However, in order to develop such performance blends from recycled polymers one must not only have an acceptable quality in the feedstocks but also have effective methods for the compatibilization and toughening of the blends. In this paper, we will illustrate the use of reactive toughening and compatibilization concepts for developing new recycled PET based blends with elastomeric modifiers and PC. Our primary objective in this study was to evaluate the compositional and morphological parameters that lead to not only a high impact strength but also a high level of thermal embrittlement resistance in these blends. Compatibilization of recycled PET/polyethylene and PA-6/PP blends will also be briefly discussed.
Recycled PET in pellet form was obtained from commercial recycling sources. It had an intrinsic viscosity [[Eta]] = 0.7 dl/g, in phenol-tetrachloroethane and a carboxyl end group content of 0.034 meq/g as determined by potentiometric titration with KOH in benzyl alcohol. The nucleator used for promoting the crystallization of PET was a low molecular weight, sodium ionomer of ethylene acrylic acid copolymer (Aclyn 285 from AlliedSignal; [M.sub.w] [is approximately equal to] 2000).
Recycled polycarbonate was derived either from water bottles (melt index = 3; Polyservices) or industrial scrap (M.I. = 5.8; Star Plastics). Recycled PP (M.I. = 2.5), originating from post-consumer ketchup and syrup bottles, was obtained from Clearview Ltd. of Amsterdam, N.Y., in flake form. It contained 85-90% PP, 5-10% EVOH copolymer as barrier resin and 3-5% of a modified polyethylene as a tie layer material. Recycled HDPE (M.I. = 0.4, Wellman Industries) and LDPE (Dow 640, M.I. = 2) were used as received.
Ethylene-glycidyl methacrylate (92/8) copolymer (E-GMA) of [M.sub.w] = 240,000 and ethylene-ethyl acrylate-glycidyl methacrylate (72/20/8) terpolymer (E-EA-GMA) of [M.sub.w] = 380,000, were obtained from Sumitomo Chemical and Elf Atochem, respectively. Ethylene-ethyl acrylate copolymer (E-EA) was obtained from Union Carbide Corp. Ethylene/propylene rubber (EPR) and maleic anhydride grafted EPR (m-EPR) were obtained from Exxon Co. Poly(methyl methacrylate)-g-poly(butadiene/styrene) (MBS) and poly(butyl acrylate)-g-poly(methyl methacrylate) core/shell type rubbers (Paraloid EXL 3647 & 3330) were obtained from Rohm & Haas Co. PA-6 ([M.sub.n] [is approximately equal to] 20,000) of virgin and recycled types were products of AlliedSignal Inc.
All melt blending experiments were done with thoroughly dried resins extruded under intensive mixing conditions at throughput rates varying from 3 to 10 kg/h depending on the type of extruder used, either: (i) a single-screw extruder (Killion, 25 mm, L/D = 30) equipped with a Maddock mixer and a vacuum vent port, (ii) a conical, counter-rotating twin-screw extruder (Haake, HBI TW-100, L = 330 mm; D = 20-30 mm); or (iii) a co-rotating, twin-screw extruder (Leistritz, 34 mm; L/D = 40) with three feed ports and ten separate mixing zones, which contained some kneading blocks for intensive shear mixing.
The maleation of polypropylene (recycled/virgin grades) was done by extruding PP with 1% maleic anhydride and 0.125% of a peroxide initiator viz., 2,5-dimethyl,2,5(di-t-butyl peroxy) hexyne-3 (Lupersol 130, from Elf Atochem).
All samples were injection molded and tested under standard ASTM conditions. Annealing and heat-aging of test bars were done in hot air ovens at various temperatures and time intervals as specified in the text. To test the annealing effects on the notched Izod impact strength, the notching was done after annealing of the test bars. Crystallinity was measured by WAXS using the method of Statton (2). Morphology was determined by transmission electron microscopy (TEM) using ultrathin cryosections and appropriate staining (Ru[O.sub.4]) as necessary.
RESULTS AND DISCUSSION
Blends Based on Recycled PET
Unreinforced PET is generally not useful as an injection molding resin because of its slow crystallization rate (3) and the tendency to embrittle upon crystallization such as by thermal annealing (4). In contrast, polybutylene terephthlate (PBT) is more widely used for injection molding and extrusion applications because it crystallizes much faster and is more ductile. Under the fast cooling conditions of conventional injection molding and extrusion processes, unmodified PET does not crystallize to any appreciable extent and hence the molded or extruded parts are usually quite amorphous. The rate of crystallization of PET can be improved to some extent by the addition of suitable nucleators such as sodium salts of organic carboxylic acids or sodium ionomers of ethylene-methacrylic (or acrylic acid) copolymers (5). However, commercial injection molding grades of PET generally employ such nucleators only in combination with glass or mineral reinforcement.
Unmodified, recycled PET can be injection molded without difficulty only when relatively low mold temperatures (20-38 [degrees] C) are used. The parts so obtained are amorphous but fairly ductile, as evidenced by high tensile elongation at break. However, annealing at 150 [degrees] C causes significant crystallization of the PET but with severe loss in ductility, viz., a dramatic drop in the tensile elongation at break. Use of an efficient nucleator such as a sodium ionomer of a low molecular weight, ethylene-acrylic acid copolymer causes a significant level of crystallinity to develop in the PET during the fast molding cycle, but this also results in a severe loss of ductility. Regardless of whether induced by annealing or nucleation, the crystallization of PET in unoriented injection molded samples always seems to result in embrittlement, although the modulus and heat resistance (DTUL) improve. In addition, under the conditions of stress concentration, such as when sharply notched, both the amorphous and crystalline PET are quite brittle, exhibiting low notched Izod impact strengths (30-50 J/m). Hence there is a real need for effective toughening of PET, particularly to resist embrittlement after crystallization.
In order to promote the use of untilled recycled PET in injection molding and extrusion applications, we evaluated the reactive toughening and compatibilization approaches to design toughened PET and PET/PC blends that would retain high impact strength and tensile elongation even after prolonged exposure to high temperatures.
Low Modulus/High Impact PET Blends
For some extrusion applications such as wire jacketing and tubing it is desirable to have a high melting, crystalline thermoplastic polymer that would exhibit not only high heat and solvent resistance but also a high level of toughness and low modulus. [TABULAR DATA OMITTED] Although unmodified PET is high melting ([T.sub.m] = 254 [degrees] C), it is too brittle and stiff after crystallization to be used in such applications. In order to overcome this deficiency one must first develop an effective toughening technology. The modulus of the toughened PET can then be lowered further to the desired target by blending with a low modulus polymer or plasticizer.
In our study, various impact modifiers were evaluated for their efficiency in toughening the recycled PET, particularly after annealing at 150 [degrees] C to promote crystallization. Epoxy functionalized ethylene copolymers such as E-GMA copolymers and E-EA-GMA terpolymers were found to be relatively more efficient in toughening the crystalline PET, maintaining high notched Izod impact strength and high elongation at break even after annealing (6). Conventional nonreactive impact modifiers, including controlled particle type core/shell rubbers (MBS and acrylate rubers), were less effective in toughening crystalline PET after annealing, although they improved the toughness of amorphous PET before annealing.
The efficiency of the reactive tougheners was attributed to the melt-phase grafting reaction between the epoxide functionality of the ethylene copolymer and the carboxyl end group of PET. The graft copolymer so formed is believed to play a role not only in compatibilizing the toughener dispersions, but also in controlling the morphology and deformation behavior of the crystalline PET and thus contributing to the thermal embrittlement resistance.
[TABULAR DATA OMITTED]
By blending a corresponding nonreactive ethylene copolymer (which is compatible with the reactive ethylene copolymer), e.g., E-EA copolymer with E-EA-GMA or EP rubber with E-GMA, low modulus PET blends with high retention of impact strength and ductility after annealing were obtained. In fact, the use of a compatible mixture of reactive and nonreactive modifiers (e.g., E-EA-GMA + E-EA) seems to cause a synergistic improvement of impact strength and annealing resistance.
Morphology examination by transmission electron microscopy (TEM) of a high impact, PET/E-EA-GMA/E-EA (60/25/15) ternary blend indicated that the ethylene copolymers were dispersed as predominantly submicron size particles in the PET matrix. The fine particle morphology of the blend was maintained even after thermal annealing. The apparent morphology stability of such blends undoubtedly originates from reactive compatibilization, and is reflected in the observed high retention of ductility even after long-term thermal annealing. Such a high level of retention of ductility with heat-aging is desirable in applications requiring continuous or cumulative exposure to heat, e.g., in automotive under-the-hood applications. However, any undesirable post-mold (or extrusion) shrinkage that might occur as a result of the crystallization of PET should be controlled by the addition of effective nucleators and/or by an on-line annealing treatment.
High Impact PET/PC Blends
Polycarbonate (PC) is a high [T.sub.g] (150 [degrees] C), ductile polymer, but it lacks adequate solvent resistance because of its amorphous nature. PET/PC blends have therefore been developed, because they combine the solvent resistance advantages of PET with the high [T.sub.g] and toughness advantages of PC. There are currently several commercial grades of PET-PC blends that include an impact modifier to improve the impact strength (7, 8).
Binary blends of PET and PC are known to be phase-separated systems exhibiting partial miscibility in the absence of an ester interchange. In blends containing [greater than or equal to] 40% PET, the matrix phase is PET-rich and the dispersed phase is PC-rich (9). This partial miscibility is responsible for the self-compatibilizing nature of the blend, resulting in the good delamination resistance and tensile elongation properties. However, in spite of its good ductility in tensile deformation and drop weight impact behavior, the notched Izod impact strength of a PET/PC (50/50) binary blend is generally quite low ([less than or equal to] 60 J/m). Hence an impact modifier is invariably needed to improve the notch sensitivity.
As discussed earlier, PET is normally in an amorphous state when injection molded. Upon thermal annealing, the amorphous PET matrix phase of the PET/PC blend tends to crystallize and embrittle. Hence, if one wants not only to improve the as-molded impact strength but also to maintain it after thermal annealing, it is important to select a proper impact modifier for PET/PC blends. We extended the reactive toughening and compatibilization techniques to the PET/PC blends using both a recycled PET and a recycled PC in our study.
Blending a PET-reactive impact modifier such as E-EA-GMA in the PET/PC blend was found (10) not only to improve the as-molded impact strength but also to retain it significantly after prolonged annealing. It is believed that because of the selective reaction and the fixation of the impact modifier in the PET matrix, the thermal embrittlement tendency of the PET, and hence that of the blend, is significantly suppressed. The role of the relative location of the impact modifier (PET vs. PC phase) and the effect of reactive vs. nonreactive modifiers on the PET-PC blend's thermal-aging behavior has been reported recently by Akkapeddi et al. (11). The properties obtained for the PET/PC/E-EA-GMA (40/40/20) blends using the recycled PC samples were similar to those obtained by using virgin PC.
Interestingly, it was found that as lithe as 5 wt% of the reactive impact modifier E-EA-GMA was adequate in maintaining the short-term, thermal annealing resistance of the blend, provided the E-EA-GMA was used in combination with a co-miscible, nonreactive impact modifier such as E-EA copolymer. Since the E-EA-GMA and E-EA copolymers are mutually miscible, the mixture presumably acts as a single-phase rubbery dispersion, but interfacially grafted to the PET phase.
Blends Based on Recycled Polyolefins
Compatibilized PET/Polyethylene Blends
Since polyethylenes have lower melting points (90-125 [degrees] C) than PET, we were interested in developing compatibilized blends in which the PET would be the continuous phase and the polyethylene would form stabilized dispersions. Such blends would in principle have higher heat resistance and application potential than those in which the polyethylene forms the continuous phase. However, the type of morphology and its stability will depend on the melt viscosity (and elasticity) difference and the interfacial tension between the resins.
Table 3. Properties of PET/PC (1:1) Blends Containing 20 wt% E-EA-GMA Copolymer as Impact Modifier.
Recycled Recycled Virgin PC PC PC (M.I. = 3) (M.I. = 5.8) (M.I. = 5.5)
Flexural modulus (MPa) 1595 1550 1553 Yield strength (Mpa) 37 37 37 U. elongation (%) 110 130 115 Notched Izod (J/m) 850 840 850 DTUL
Simple blends of PET and polyethylenes are highly immiscible and hence exhibit poor mechanical properties and easy delamination. Previous attempts in the literature for compatibilizing the blends involved the use of styrene-ethylene/butylene-stryrene block copolymers, S-EB-S and maleated S-EB-S (12, 13). In our studies we evaluated E-GMA copolymers as reactive compatibilizers for the PET/HDPE and PET/LDPE blends. We were interested to see if the graft copolymer generated during the melt blending of PET and E-GMA could also compabilize the polyethylene (HDPE or LDPE) phase. On a molecular level, the polyethylene portion of the graft copolymer would be expected to be miscible with the bulk polyethylene phase, and the PET portion of the copolymer would be expected to be miscible in the bulk PET phase.
The E-GMA copolymer itself was found to form fairly miscible blends with the polyolefins (HDPE, LDPE, EPR) without any detectable phase separation. However, owing to the high GMA content (8%) of the copolymer, a high level of coupling reaction took place between the E-GMA and the PET, which caused a substantial increase in the melt viscosity and some entrapment of the E-GMA as dispersed micelles inside PET. In a typical PET/LDPE blend (50/40), addition of 10% E-GMA causes a reasonable level of compatibilization between PET and LDPE, but the increased viscosity of PET resulted in a phase inversion. Accordingly, the morphology of such a blend indicated fine dispersions of PET in an LDPE matrix. In order to prevent phase inversion, it was found necessary to use [greater than] 60% PET in the blend to ensure a PET continuous phase. It is believed that the extent of reaction and phase inversion can be reduced by using an E-GMA copolymer of low GMA comonomer content.
The PET/HDPE blends were also found to be compatibilized by the addition of the E-GMA copolymer, although even in this case, there is a tendency for phase inversion as a result of the grafting reaction (14). The phase inversion tendency could be minimized either by using HDPE of melt index [less than or equal to] 0.4 or PET content of [greater than] 50%. In general, the elongation at break and the notched Izod impact strength of PET/LDPE and PET/HDPE blends improved upon the addition of 10-15% E-GMA as reactive compabilizer, compared with the uncompatibilized blends.
Compatibilized Polyamide 6/Polypropylene Blends
Although polypropylene is used in large volumes in a variety of applications including packaging, until recently post-consumer recycling of PP has not been developed to any significant extent. The use of PP in food packaging, however, seems to be increasing, particularly in retort barrier packaging, in which PP is coextruded with ethylene-vinyl alcohol (EVOH) barrier layer and modified polyethylenes (carboxylated or maleated) as adhesive tie layers. Post-consumer recycled PP from such sources therefore contains 10-15% EVOH, and up to 5% modified polyethylene as contamination. Nevertheless, we evaluated the utility of such a recycled PP for blending with PA-6 using reactive extrusion methods. Compatibilized PA-6/PP blends, with PA-6 as the matrix, are expected to combine the solvent and heat resistance advantages of the PA-6 and the moisture resistance advantages of PP.
Table 4. Key Properties of Recycled PET/Polyethylene Blends.
Tensile Ultimate Notched Strength Elongation Izod (MPa) (%) (J/m)
PET/LDPE (60/40) delamination -- -- PET/LDPE/E-GMA 26 250 80 (60/25/15) PET/LDPE/E-GMA 18 40 140 (50/40/10) PET/HDPE (50/50) 25 2 16 PET/HDPE/E-GMA 23 55 134 (50/40/10)
Owing to the high interfacial tension between the two polymers, simple blending of PA-6 and PP leads to morphologically unstable blends with poor mechanical properties. However, compatibilization can be achieved by modifying the PP first with maleic anhydride grafting (maleation) and then melt blending with the PA-6 (15). A coupling reaction between the anhydride functionality of maleated PP and the amine end group of PA-6 is expected to form a graft copolymer at the interface that can compatibilize the blend.
We have found that recycled PP can be maleated to a similar extent as virgin PP, by extruding with maleic anhydride (1%) in the presence of a peroxide initiator (0.125%). Some decrease in molecular weight occurred, as expected from the thermal and peroxide induced chain scisson. However, the maleation reaction does not seem to be significantly affected by the presence of small amounts of EVOH in the recycled PP. Melt blending of the maleated, recycled PP with nylon 6 in a second extrusion gave a compatibilized nylon 6/PP with mechanical properties similar to those obtained with virgin polypropylenes. It must be noted that in both cases, tensile elongation and notched Izod impact strength are significantly improved compared with the uncompatibilized blend. The slightly lower modulus obtained with recycled PP was caused by the polyethylene and EVOH contaminants.
In summary, our study demonstrated the utility of reactive compatibilization and toughening concepts for designing blend products from recycled polymers exhibiting a combination of desirable performance features such as high toughness, thermalaging resistance, and solvent resistance. Continued progress made in the technology of compatibilization, toughening, processing, and application development of polymer blends is expected to drive the increased usage of recycled polymers.
BuA-C/S = Crosslinked poly(n-butyl acrylate)-g-poly(methyl methacrylate) core shell rubber.
E-AA = Ethylene-acrylic acid copolymer.
E-EA = Ethylene-ethyl acrylate copolymer.
E-EA-GMA = Ethylene-ethyl acrylate-glycidyl methacrylate copolymer.
E-GMA = Ethylene-glycidyl methacrylate copolymer.
EPR = Ethylene-propylene rubber.
m-EPR = Maleated EP rubber.
HDPE = High-density polyethylene.
LDPE = Low-density polyethylene.
MBS = Methacrylate-butadiene-styrene core/shell rubber.
PA-6 = Polyamide 6.
[TABULAR DATA OMITTED]
PC = Polycarbonate.
PET = Poly(ethylene terephthalate).
PP = Polypropylene.
1. J. H. Shut, Plastics Technol., 39(7), 96 (1993).
2. W. O. Statton, J. Appl. Polym. Sci., 7, 803 (1963).
3. M. Gilbert and F. J. Hybart, Polymer, 13, 326 (1892).
4. R. M. Mininni, R. S. Moore, J. R. Flick, and S. E. B. Petrie, J. Macromol. Sci. Phys., B8, 342 (1973).
5. (a) R. Legras, C. Bailley, M. Daumerie, J. Dekoninck, J. Mercier, V. Zichy, and E. Nield, Polymer, 25, 835 (1984). (b) D. Garcia, J. Polym. Sci., Polym. Phys., 22, 2063 (1984).
6. M. K. Akkapeddi, B. Van Buskirk and J. H. Glans, in Advances in Polymer Blends and Alloys Technology, v. 4, K. Finlayson, ed., Technomic Publishing Co. Inc., Lancaster, Pa. (1993).
7. J. Y. Chung, D. Neuray, and M. Wittman, U.S. Pat. 4,554,314 (1985).
8. M. E. J. Dekkers, S. Y. Hobbs, I. Bruker, and V. H. Watkins, Polym. Eng. Sci., 30, 1628 (1990).
9. W. Kim, and C. M. Burns, J. Polym. Sci., Polym. Phys., 28, 1409 (1990).
10. M. K. Akkapeddi and C. D. Mason, Int. Pat. Ap. WO91/15545 (1991).
11. M. K. Akkapeddi, C. D. Mason, and B. Van Buskirk, Am. Chem. Soc., Div. Polym. Chem., Polym. Prepr., 34, 848 (1993).
12. T. D. Traugott, J. W. Barlow, and D. R. Paul, J. Appl. Polym. Sci., 28, 2947 (1983).
13. I. M. Chen and C. M. Shia, SPE ANTEC Tech. Papers, 35, 1802 (1989).
14. M. K. Akkapeddi, B. Van Buskirk, and X. Swamikannu, ACS PMSE Proc., 67, 317 (1992).
15. F. Ide and A. Hasegawa, J. Appl. Polym. Sci., 18, 963 (1974).