By Jan H. Schut
Three unusual new reactive thermoplastic technologies are being introduced at two Society of Plastics Engineers (www.4spe.org) conferences in September. Two make fiber composites by wetting fibers out with liquid monomer, then polymerizing during extrusion or injection molding. The third is a shape-memory elastomer made of sulfonated EPDM plus about a third of an unexpected ingredient—soap—as a tunable temperature trigger for shape recovery. They’re presented in papers at the Thermoplastic Elastomers TOPCON 2012 held September 10-12 in Akron, Ohio (www.4spe.org/conferences/tpe-topcon%C2%AE-2012) and at the Automotive Composites Conference & Exhibition held September 11-13 in Troy, Mich. (http://speautomotive.com/comp.htm).
REACTIVE EXTRUSION MEETS DLFT
Remember the ring-shaped oligomer CBT (cyclic butylene terephthalate), developed by GE Plastics (www.sabic-ip.com) in the ‘80s and spun off over a decade ago as Cyclics? CBT’s ring shape with no chain entanglement gives it very low viscosity. Cyclics Inc. (www.cyclics.com) in Schenectady, N.Y. sells pelletized CBT as a blending material to lower melt viscosity, Tg and HDT for better wetting and higher loading of mineral fillers and fibers and better dispersion of nano materials in applications like master batches and filled compounds.
But CBT can also be made into new PBT polymers and copolymers with different properties from commercial PBT made by polycondensation. (Commercial PBTs include Sabic Innovative Plastics’ Valox, DuPont’s Crastin, and BASF’s Ultradur.) Cyclics patented and announced reactive extrusion of PBT ten years ago (U.S. Pat. # 6436549), but never commercialized it because of cost. Cyclics made a developmental PBT elastomer in 2007, called XL101, a copolymer of CBT and caprolactone for rotational molding of high barrier marine gas tanks. XL101, which was reacted and polymerized in the rotomold, met gas permeation tests, but was too expensive for the application.
Now Cyclics wants to make PBT reactively in a twin-screw extruder. Last year Cyclics partnered with the National Research Council of Canada in Boucherville, Que. (www.nrc-cnrc.gc.ca) to run the first reactive extrusion trials making CBT into thermoplastic glass-filled PBT. The trials were done on a 70-mm-diameter twin-screw extruder at the Magna-NRC Composites Centre of Excellence in Concord, Ont. , and combined reactive extrusion with direct long fiber molding thermoplastic extrusion (DLFT), which is believed to be a first.
Three papers at the two SPE conferences present different aspects of this R&D. Cyclics’ CEO James Mihalich presents “Reactive Extrusion of Polyester Elastomers” at the TPE TopCon and “In Situ Polymerization of Reinforced Thermoplastics” at the Automotive Composites Conference. Victor Bravo, research officer at the NRCC, presents “DLFT Experiments with Cyclic Butylene Terephthalate” at the Automotive Composites Conference.
Here’s how reactive extrusion of long-fiber PBT works. Pelletized CBT is fed into a co-rotating intermeshing twin-screw extruder, where it melts and becomes watery. Multiple long glass rovings are added and wetted out. “Glass fiber can be added in very long lengths,” NRCC’s Bravo explains, “because you don’t put in the same degree of mechanical work” as in conventional thermoplastic compounding. Screw design affects fiber length.
The catalyst is present in the pellet and triggered by temperature. After the fiber is wetted out, when the mixture reaches 225 degrees C, CBT polymerizes into linear chains of PBT. “The challenge is getting residence time and operating temperatures right,” says NRCC’s Bravo. “Temperature determines the rate of polymerization. The potential problem is incomplete polymerization, which would produce short chains, low molecular weight and poor mechanical properties.” Co-reactants like hydroxyl groups or esters can be used to bond PBT directly to fibers or fillers or make copolymers. Additives could also cross-link PBT into a thermoset, but this wasn’t part of the NRCC’s R&D work.
The extrudate was then compression molded in line using two different test molds, a specimen mold for a 2 kg part and an actual automotive mold for a 15.5 kg part. Long-glass PBT was tested with different glass loadings of up to 40 weight %. The target is semi-structural automotive parts for weight reduction like door carriers and front end carriers.
GE’s original technology made CBT by breaking PBT polymer (made by polycondensation) into oligomers. Cyclics’ 5 million lb/year plant for CBT in Schwarzheide, Germany, makes the oligomers from monomers, not from PBT (U.S. Pat. # 6855798), which is less expensive. Cyclics’ Mihalich also developed a new patent-pending process to make CBT reactively by toll compounding, which reportedly could cut the cost of CBT in half.
REACTIVE MOLDING OF HIGH-GLASS THERMOPLASTIC NYLON
The other new composite technology polymerizes nylon with a glass-fiber prepreg in an injection mold. It’s reported for the first time in the U.S. at the Automotive Composites Conference where Joachim Kragl, director of advanced injection molding at Engel Machinery Inc., York, Pa. (www.engelglobal.com) describes how “Organomelt & In Situ Polymerization Provide New Opportunities for Injection Molding of Composite Structures.”
Thermoplastic nylon RIM isn’t new. Monsanto offered its Nyrim process in the 1980s and later sold it to DSM, which also combined it with glass fiber. What is believed to be new in Engel’s technology is combining nylon RIM with glass fiber prepregs. The development was introduced last June at an Engel symposium in St. Valentin, Austria.
Instead of taking a composite fabric of thermoplastic and glass fiber, thermoforming it into a shape, moving it to an injection molding machine, and molding over it, Engel’s new process puts a less expensive dry fiber structure into the mold cavity and injects caprolactam monomer, catalyst and activator, and polymerizes it into thermoplastic nylon 6 in the mold. Engel combines caprolactam plus catalyst (a metal salt) with caprolactam plus activator (isocyanate) in a mixing head similar to those used for thermoset reaction injection molding (RIM). The mixture is injected at 120 degrees C into a heated mold, which requires special seals because the caprolactam is so watery.
“Theoretically polymerization starts in the mixing head, but there is time to fill out the mold while the monomers are still very low viscosity for good fiber wetting,” Engel’s Kragl explains. Molding time is a little longer than standard injection molding, in the range of fast epoxy curing. Thermoplastic glass-filled parts are melt reprocessable, where thermoset parts aren’t. In-mold polymerizing of a thermoplastic also gives more design freedom than a thermoset because a thermoplastic composite part can be over-molded and welded.
Engel has worked on the development for about a year and a half in partnership with Lanxess (www.lanxess.com), ZF Friedrichshafen AG (www.zf.com), Trumpf Group (www.trumpf.com), PD-Interglas Technologies AG (www.pd-fibreglass.com), and Fraunhofer ICT (www.ict.fraunhofer.de). The process was tested on a mold for an insert for a brake pedal about a foot long by 2 inches wide and molded on a prototype 120-ton Engel injection molding machine.
SOAP TRIGGERS NEW SHAPE MEMORY POLYMERS
At the TPE TOPCON conference, Robert Weiss, professor of polymer engineering at the University of Akron in Ohio (www.uakkron.edu) and a Fellow of the SPE, also presents a new reactively made thermoplastic material: “Shape Memory Elastomers Based on Fatty Acid/Ionomer Compounds.” U. Akron’s Weiss first reported the shape memory potential of cross-linked blends of sulfonated EPDM ionomer and low molecular weight fatty acid salt in 2008 in a paper with the American Chemical Society. But at that time the material could only recover about 90% of its original dimensions. Now Weiss reports that slight vulcanization of the EPDM matrix introduces covalent cross links that strengthen the permanent network, so shape recovery is 100% and repeatable.
Weiss’s developmental shape memory polymers are made by reacting sulfonated EPDM ([poly(ethylene-propylene-5ethylidene-2-norbornene)] ionomer, which is acidic, with fatty acids—surfactants, or essentially soap—which are basic. The melt temperature of the surfactant triggers shape recovery. Zinc/sulfonate in the SEPDM ionomer makes the permanent physical cross-links, while very small zinc stearate crystals (0.5 mm) interact with the Zn-SEPDM (zinc salt of sulfonated EPDM) and make the temporary physical cross links. The temporary network is based on polar interactions between the metal salts in the ionomer and the low molecular weight fatty acid below the softening point of the crystals.
More significant, though, is that Weiss is now applying this shape memory trigger across a broad range of temperatures, determined by the melt temperature of different fatty acids. “Different surfactants offer a range of trigger temperatures from 30 to 130 degrees C,” U. Akron’s Weiss explains. He is working on both high and low temperature shape memory elastomers, including a shape-memory fiber for a bio medical application with a 40 degree C trigger and a high temperature sulfonated PEEK ionomer/fatty acid blend for an aerospace application. Shape memory PEEK could open exciting possibilities approaching the properties of polyimide, but with better processability.
Shape memory polymers are usually physically cross-linked glassy polymers like epoxy or cross-linked semicrystalline elastomers like polyurethane, which have a permanent cross-linked structure and then are stretched or deformed into a temporary shape. When the deformed material reaches a trigger condition, typically temperature, it returns to its original shape. Shape memory polymers can also be triggered by light, moisture or electricity and are used in applications like medical devices, sensors, switches, and artificial muscles. A patent application (U.S. Pat. Applic. 20080287582) on Weiss’s system has been dropped because of a previous expired Exxon patent on sulfonated EPDM elastomers from the ‘70s, but the expired Exxon patent didn’t reference shape memory.