In his first three Extrusion Expert webinars, Allan Griff, a consulting engineer and extrusion expert with more than 40 years of experience in plastics extrusion, focused on machinery and auxiliaries. In his fourth, he went inside the barrel and hopper to look at resins.

Entitled, “Do monkey with the formula,” the webinar, which drew more than 700 registrants, sought to empower extruders to be able to make alterations to compounds where needed and better understand how different resins, additives, and compounds interact.

Topics covered included trusting your supplier, processing stability, surface properties, rheology effects, and insights into an array of additives, including antibacterials, antioxidants, and functionalizers. Allan brought lots of info and participants brought lots of questions. Below is a sampling of the queries they had for Allan and his answers. The entire webinar, and the previous events in the series, is archived at plasticstoday.com.

Q: How useful is using two different loads for melt index (MI) to pick up differences in molecular weight (MW) distribution?
Allan Griff: Multiload MIs are very useful to show how materials behave under extrusion conditions, especially at the shear rates in the die lips and screw-to-barrel clearances. Even if the higher load doesn’t reach these lip and clearance rates, the line can be extended up into that range, and materials can be compared there. Three points are even better, as the line will curve and the three points will indicate the curve shape. The only thing better than that is a real viscosity curve, as obtained over a range of stresses.

However, some added information is needed to relate the differences to MW distribution. I suspect that such data exist (e.g., effect of distribution on power law exponent). Since processors need to know flow behavior rather than distribution, the MI information may be enough. Polymer researchers, on the other hand, may find the correlation of distribution variables with viscosity change useful in the designing and production of new polymers and better understanding of old ones.

Q: Are there any additives available to control acetaldehyde (AA) in PET?
AG: PolyOne offers concentrates of an AA scavenger, as does Austrian concentrate manufacturer Gabriel-Chemie. I don’t know the active ingredient, but maybe they will tell you. It must be food-approved, as the need for low AA is based on its effect on the taste of products packed in PET.

Process conditions are important, too. AA content is affected by the moisture content, which degrades the PET at extrusion temperatures and produces AA as a byproduct. Therefore, anything that can keep moisture content very low (e.g., 0.005%), or anything that will reduce the time and/or temperature in the melt phase, will contribute to a low AA content.

You might contact tech-service people at Eastman, who may be more up-to-date on the latest developments than I am. Also, try BASF, who make the PET chain-extender Joncryl, and should be aware of the latest PET chemistry developments. DSM and Kenrich also make additives that will increase molecular weight, but do not react or otherwise minimize the AA.

Q: Can you give an example of an antioxidant?
AG: One of the best known is Ciba’s (now owned by BASF) Irganox series (1010, 1076). Another antioxidant is alpha-tocopherol, better known as Vitamin E. It is OK for food packaging and expensive, but very effective, which means very small proportions and consequent mixing problems. Concentrates are recommended, even for the others, but certainly for this one.

Q: Are there any additives to eliminate unwanted color in PET?
AG: There are several approaches to this. The unwanted color—usually a yellowing—can be masked by another color, typically purple, which seems to have more consumer appeal. In vinyls and styrenics, maybe PET too, they use a dye called Oil of Violet. I don’t know its food status, but it is used for drinking glasses. The problem with all PET additives is the need to take melt temperatures in the 250-300°C range—not all organic compounds can remain stable at such temperatures.

PolyOne has a concentrate called ReBoost for use with recycled PET to minimize yellowing, and I suspect it has a high antioxidant content. The other companies noted above (Gabriel, BASF, Kenrich) may have similar products, or at least the active ingredients. You can also use a UV light absorber if the color arises from outdoor exposure. There are several well-known suppliers of these. Finally, you can choose not use an additive at all, but keep time and temperature in melt phase to a minimum, as noted above for AA reduction.

Q: Can you discuss the use of starch as a degradation promoter or in other thermoplastic applications?
AG: I have only heard of it as a degradation promoter. The starch is attacked by microorganisms, and its breakdown weakens the product and promotes its disintegration into fragments. The starch percentage and type, as well as particle size and product thickness, are important, as the starch particles can’t be too small nor too well encapsulated in the plastic.

This has been extensively studied by state universities in Midwest farm states, as well as the USDA, and you may get more detailed information from them, as well as suppliers of concentrates and ready-to-run compounds. Starch might also serve as a nontoxic additive to reduce gloss.

Q: How do you control melt index?
AG: That depends on what you mean by “control.” If you mean the measurement of the materials you use, you need a melt indexer, which is a capillary rheometer with a fixed opening. You apply the load and temperature that is specified in the test procedure (for LDPE, 190°C and 2.16 kg). On a small extruder, the motor amperes may correlate with melt index (inversely) if run at standardized screw rpm and barrel/die temperatures. Other, more expensive test instruments can provide such data, too.

If you mean how can you change the melt index of the material you use, you can blend it with other resins of higher or lower flow, depending on which direction you want to go. You can also add materials that change viscosity, like processing aids, but that doesn’t really change the melt index of the base resin. With some polymers like PET you can add reactive materials that increase molecule size, which will lower melt index. Finally, you can subject the resin to such high heat that its molecules break down, which will raise melt index, but this is only done rarely and under controlled conditions by the polymer makers.

Q: Since every time a polyolefin is processed, some of the antioxidant is consumed, should antioxidants be added to regrind, especially that which has passed through the process many times?
AG: My answer is yes, as more antioxidant well-mixed into the mass may maintain strength and even reduce discoloration. However, if the product is strong enough to meet its needs without this extra additive, and discoloration is minimal and doesn’t matter, then there is no economic sense to adding more antioxidant, since it will add more cost. Remember that it won’t help strength where the product is weakened by particles of contamination acting as stress concentrators, which is often the case with recycled material. Crosslinking is another chemical way of getting the same effect.

Q: Would adding antioxidant help a situation where all-virgin PP run on a small extruder gives a much better product than on a bigger machine?
AG: First, I must assume it is the same material and the same product dimensions. If not, the problem may be due to the different materials or dimensions. Next, I would like to know what you mean by “better product.” Here are some possibilities:

• It may be uniformly degraded, which is expected when many test specimens show similar and low tensile elongation and/or impact strength. Also, the product may be off color in yellow direction. In this case, more antioxidant may help, as the bigger machine may be running hotter, and the melt may be in there a longer time. Confirm this with observation of melt temperature in the adapter or with an IR device at die exit, and try running cooler in the head/die before opting for more antioxidant, which would increase material costs.

• It may be weaker in some test specimens and not others. This is a sign of contaminants acting as stress concentrators, and these contaminants may be carbonized flakes produced within slow-moving regions in the adapter or even on the screw root. If there are dead areas in the flow path, or the screw(s) have not come out in years and there are coarse or even no screens and disassembly shows such carbon, cleaning the system may solve the problem.

More antioxidant may help, too, as well as a processing aid that coats the metal surfaces. Remember that the dispersion and distribution of the additive is critical, as it controls how much is needed, and thus affects final material cost. This depends on the nature and precision of the devices that feed the additive. Often a concentrate is preferred—even though it costs more per unit weight. The better dispersion may mean less is needed.

If you do use a concentrate, however, make sure you know what the carrier resin is and look at heat-squashed pellets under a microscope to get an idea of particle size. A blend of two antioxidants may work better than the same total quantity of either one (e.g., phenolics plus pentaerythritol-based antioxidants for polyolefins). The additive and concentrate makers can tell you more about what goes best with what.

• Finally, the heat stability may not be the cause of the problem. Cooling rates and methods may be different, and that could make one product stronger than the other. I’d have to know more about the products and the problems to say anything more.

Q: We extrude thick-wall HDPE pipe for a structural part. We think we see “cold flow” of the part for some time after manufacture. How long should we allow the pipe to “relax” before processing further, and can we stabilize the material so that cold flow is minimized?
AG: First of all, we must distinguish between post-extrusion, or residual crystallization, and actual cold flow or creep. Residual crystallization occurs with semicrystalline polymers such as HDPE, where the molecules continually reorganize their positions for some time after the product looks finished. The product is normally not in service, with no applied stress during this time. I worked on such a case with an HDPE sheet, melt index 0.3, where they had to wait up to three days until there was no more shrinkage before cutting and printing.

Creep is a well-known behavior of most all materials, including metals, which are really thermoplastic, and most all plastics. It depends on the forces applied to the material, its temperature, and any restraints applied to it. It is not related to the time elapsed since extrusion, as long as all the residual crystallization was finished.

Returning to your questions: You can follow residual crystallization by measuring a length of pipe and seeing when its shrinkage has stopped. For thick walls, this may take many days.

Also, there may be residual stresses frozen into the pipe, which may relax over a longer period of time, and the rate of relaxation will depend on service temperature and the conditions at which the stresses were frozen in. There are tests for residual shrinkage that heat the pipe to some fixed temperature and see how much its dimensions change. If the changes are significant, the focus should be a more gradual cooling rate and a cooler melt temperature at the die exit. Some additives, such as glass fibers, increase dimensional stability but have their own set of problems (e.g., abrasion, surface treatment, and fiber direction and length).

If the pipe is essentially stress-free but still distorts under load, some of this distortion may be elastic (recovers when load is removed) and some plastic (creep). There is not much that can be done about this in extrusion, except use of glass fibers, perhaps mica or calcium carbonate, or other inorganic fillers.

Q: Do intrinsic viscosity (IV) enhancers work?
AG: There are three ways of enhancing (raising) IV of a PET resin: blending, solid-stating, and chain extension. They all work, but each has its appropriate situations. Blending can be done with high-IV resin sold for sheet extrusion where more melt strength is required. It costs more than lower-IV resin and certainly more than recycled, but the blended result may allow processing of these otherwise-difficult resins.

Solid-stating is essentially “cooking” the solid pellets for several hours at high but sub-melting temperatures. The molecules reorganize themselves and join to form larger molecules, which raises the IV. This is typically done in large volume by resin suppliers and compounders, but in principle can be done by anyone.

Chain extension is done with additives added at processor level, either as the pure reactive chemical or as a concentrate. BASF’s Joncryl is an example. It reacts with the ends of the PET chains so that two or more chains become part of a larger molecule. A very small amount is needed, so it is expedient to use a concentrate. Another reason to use concentrate is the low melting point and thermal instability of the additive. It may be cheaper to use pure powdered additive, but you need a feeder that accurately handles powders, and you may have to mix it, or dilute it, with powdered PET to get more uniform distribution.

Q: Are there additives for PET that can eliminate traces of benzene?
AG: I have heard concerns about benzene as a product of PET breakdown—see Ensobottles.com, where they say benzene can “end up in the beverage from excessive wear of the plastic bottle,” but their bottles don’t have any. Wear? It is chemically possible through the decarboxylation of terephthalic acid, but I wouldn’t expect to find any significant amount. There are plenty of people who will be scared by any connection of a harmful chemical such as benzene with any plastic bottle. —[email protected]