Rigid barrier coextrusion: where it panned out.

Ten years ago, the rush was on to make barrier plastic containers to replace steel and glass. The learning curve was unexpectedly steep. Costs were high, successes few. Now that the noise has died down, it's time to ask how the technology evolved and where it's headed now.

Plastic sheet and blow molded containers with oxygen-barrier layers burst on the technological scene 15 years ago. Excitement ran so high that the topic packed three technical conferences a year. Machinery suppliers still refer to those as "the good old days" of coextrusion. Large metal and glass packaging companies invested big bucks in multilayer sheet and bottle ventures. In sheet, for example, Pittsburgh-based Alcoa (Aluminum Co. of America) and then Metalbox Co. of England (now Carnaud Metalbox SA) invested over $50 million in a joint-venture called Genesis Packaging Systems in Charleroi, Pa., which installed two huge coex lines to make seven-layer barrier sheet for shelf-stable foods.

But these pioneer ventures struggled, and many failed. Genesis became the most expensive debacle of all when it closed in 1990, having made only test-market products, nothing fully commercial. Sheet lines and thermoformers had run only intermittently for a few weeks at a time. Genesis had mastered a complex coextrusion process with a PVDC barrier layer (Dow Plastics' Saran). But PVDC became "unacceptable in the marketplace. None of the people we were working with would accept the scrap factor," says Simon Doyle, Metalbox's director of high-performance plastics. The plant switched to EVOH and then closed. It was later sold to a recycler. Today its fancy coex lines make monolayer recycled PET sheet.

Unfortunately, Genesis' problems were not unique. Other barrier coex dropouts included Central States Can Co.'s 11-layer sheet extrusion and thermoforming operation in Massilon, Ohio, and Composite Container Corp. in Medford, Mass. Central State's sheet line was broken up and sold in pieces; Composite Containers's multilayer sheet lines were sent to South America. Still others, like a seven-layer barrier line installed in 1989 at Hopple Plastics Inc. in Florence, Ky., stayed put but were simply never used for high-barrier sheet.

Barrier coextrusion was not an unmitigated disaster. Many millions of these rigid containers are produced today, and there are even some new applications emerging, though many fewer than forecast. But there's little in the way of really new technology on the horizon. That's partly because the trend in barrier packaging appears to be toward greater simplicity in numbers of layers and materials, rather than greater complexity. What's more, much of the action in rigid coextruded packaging has turned from barrier structures to decorative multicolor or clear-and-pigmented effects, or to sandwich structures with buried recycle layers.

Where barrier coex development remains active is in flexible packaging (a subject for another story) and in non-packaging durable goods, such as refrigerator-liner sheet and blow molded gas tanks. This article will examine all these trends because where the market leads, technical developments in coextrusion are pretty sure to follow. But in order to grasp where coextrusion is going, it's important first to look back at where it has been.

WHAT WENT WRONG?

Processors had "promised food packagers everything," recalls Peter Cloeren, president of Cloeren Co. in Orange, Texas, a leading designer of coex feedblocks and dies. "Then when they |processors~ couldn't deliver, the market lost interest. There's a big difference between successfully making a package in a lab and making it in production. Extrusion is extremely difficult to scale up." Ironically, five years later, by the mid to late '80s, the bugs were worked out. The process was ready, but the market was gone. "Food packagers are still gun-shy," Cloeren adds.

Not all the early rigid barrier attempts failed by any means. Some firms built markets for successful niche products, like six- and seven-layer sheet for aseptic form-fill-sealed portion packs of applesauce and pudding. But there has been little growth even in these applications.

"These markets aren't growing and no one is buying new equipment," says Frank Nissel, president of Welex Inc., Blue Bell, Pa. Several key machinery makers say they haven't sold a new rigid barrier coex line since the salad days of the mid '80s. They estimate there are no more than a dozen machines left, down by at least half from the initial installations.

Food formulations didn't help either. "The lunch bucket market was a victim of the food in the package too," says George M. Schroeder of consultants George O. Schroeder Associates Inc., Appleton, Wis. "These were high-priced packages for small quantities of low-quality food."

Meanwhile, barrier coex machinery changed hands frequently during the epidemic of packaging-company restructurings in the '80s. For example, Continental Can's U.S. barrier coex sheet division in West Chicago, Ill., merged last year with what was left of the sheet division of Ball Corp., in Muncie, Ind., which was whittled down in size and now a division of Alltrista Corp. One of Continental's two seven-layer coex lines went to Muncie, the other went to Australia. In the U.K., Carnaud Metalbox recently consolidated and moved it's barrier coex equipment to a French unit, ONO. Separately, Metalbox is in the process of selling its sheet and thermoforming operations, including barrier, to a Finnish conglomerate, Polar Pack.

GROWING PAINS IN BARRIER BLOW MOLDING

Complex six- and seven-layer blow-molded containers were attempted in the early '80s, soon after the first barrier sheet projects got going. Extra extruders were added to continuous-extrusion blow molders and new multilayer heads were developed. But some machine makers never sold a model of these adapted coex machines. Markets for barrier bottles remained stubbornly limited to squeezable catsup, plus some juice and pesticide containers--not exactly pushing glass off the grocer's shelf.

Of five resin companies that made EVOH a decade ago, only two are left now, a further reflection on the dimmed prospects for rigid barriers. Solvay & Cie in Belgium closed its small European EVOH plant in 1991. DuPont Co., Wilmington, Del., opened a U.S. EVOH plant that same year, then abruptly closed it this spring (see PT, April '93, p. 106). Also in that year, Quantum Chemical Co. in Cincinnati sold its half interest in the only other domestic EVOH producer, Eval Co. of America (Evalca), Lisle, Ill., to its joint-venture partner Kuraray Co. of Japan (PT, Jan. '92, p. 14). That leaves world production of EVOH to Kuraray and Nippon Gohsei of Japan (represented by Nichimen America Inc., N.Y.C., and Morton International Inc., Chicago).

Only recently has there been some new interest in barrier bottles. In the past two years, at least four new barrier coex wheels were installed in the U.S. for high-volume production of PP/EVOH bottles for apple juice and fruit punches. "Coex bottle capacity is once again full, and I'm getting requests to increase output," says Russ Labelle, president of Wilmington Machinery Inc., Wilmington, N.C., who built many early barrier coextrusion wheel blow molders. He estimates there are 35-40 barrier coex blow molding machines in the U.S., including the newer installations.

LAYER DISTRIBUTION PROBLEMS

Ten years ago, barrier-layer distribution wasn't reliable either in sheet or blow molding. "Feedblocks worked impeccably, but coex sheet dies failed to achieve perfect layer uniformity," says sales and marketing v.p. Robert Weeks of Battenfeld Gloucester Engineering Co., Gloucester, Mass. This created problems in ensuring the integrity of barrier layers only a few mils thick. "The packaging customer didn't have confidence in the barrier layer being there in a consistent fashion."

There were other problems, too: Machine controls were rudimentary; resins didn't have consistent melt flows; adhesive and barrier layers sprung pinholes when they stretched in subsequent thermoforming. All that raised the specter of barrier failures, delaminations and product liability, forcing processors to add an expensive safety margin to barrier and adhesive layer thickness. That increased costs because the extra barrier material made web and flash hard to use as regrind. Adding to the economic disadvantage was extremely slow downstream filling and sealing of plastic packages compared with conventional cans and bottles.

Sheet processors had a choice primarily of two high-barrier resins, EVOH or PVDC, each with processing problems. EVOH was moisture sensitive, while PVDC was hard on equipment. It had a nasty habit of decomposing at higher processing temperatures and giving off hydrochloric acid.

Blow molding processors stayed away from PVDC but had the added choice of nylon barrier resins. Nylon on the outside of cosmetic bottles is used for gloss and a barrier to keep expensive perfumes from evaporating. Nylon on the inside of juice bottles prevents flavor scalping by HDPE. But nylon is moisture sensitive and tricky to pinch off when it's inside a container wall, so it requires special processing and molds.

MORE BARRIERS TO COEX

Besides technical difficulties, rigid barrier coextrusions in food packaging faced consumer obstacles. The original six-layer PP/EVOH catsup bottle became a lightning rod for environmentalists because the mixture of materials was allegedly not recyclable. The real issue was probably the lack of infrastructure for reclaiming PP bottles, since more than one research study had shown that PP/EVOH sheet and bottle scrap is readily reprocessable into monolayer structures (PT, May '90, p. 11; Aug. '90, p. 19; Feb. '91 p. 31). Some catsup then switched to a three-layer PET/EVOH bottle without adhesive, so layers could separate when the bottle was ground (PT, May '90, p. 34).

Multilayer sheet attracted less attention as a post-consumer recycling issue. But if web scrap couldn't be reused -- which depended on layer configurations, resin combinations and the amount of barrier and adhesive materials used--that added 40-50% to material cost.

And then there was the problem that U.S. consumers never really understood shelf-stable food. Shoppers were suspicious of prepared food stored in plastic at room temperature. So grocers often had to refrigerate shelf-stable products to get customers to buy them.

Another irritant for early processors was Dow's dominant patent on coextrusions of three or more layers. Dow licensed machine makers to sell sheet lines for three layers or more. Then Dow wanted royalties from processors making multilayer packages. "They were one of the major negative factors. They wanted 2-3% of sales," recalls one disgruntled machine maker. "That was often the entire profit margin on a package." Dow's U.S. patent expired in 1989, but by then the bloom was off barrier packaging anyway.

WHY IT WORKS TODAY

A lot of small improvements came together in the last few years. Advances in computer controls and gravimetric feeding now make it possible to guarantee

barrier and adhesive layers far thinner as a percent of total wall than even two years ago, according to some processors. "Formerly, we had to use an 8-mil barrier target thickness to be sure we had 6 mils," says Jozef Marynissen, managing director of Cobelplast in Belgium. With today's gravimetric controls, extruder operators can simply dial in a layer thickness as a percentage of total thickness. Such controls from Process Control Corp., Atlanta, have been installed for several years controlling thin layers on cups at Solo Cup Co., Chicago.

Closed-loop control of gravimetric feeding and extruder speed are also beginning to be used in barrier coex blow molding. Graham Packaging in York, Pa., uses them to bring EVOH or nylon barrier layers down to as little as 0.5-mil, versus 1.5-mil a few years ago, on six-layer barrier bottles. Coex heads today must use higher pressures to achieve thinner layers--e.g., 4000 psi, vs. 3000 psi previously. Gravimetric controls are also helping in coextrusion blow molding with recycled material. Gravimetric feeding can, for instance, compensate for fluctuations in bulk density of mixtures of regrind and post-consumer reclaim.

KNOW-HOW ADVANCED

Several companies use computer modeling of flow geometry to predict layering behavior. Quantum developed proprietary flow-analysis software for its technical service people to use in optimizing resin choices and processing conditions (PT, Feb. '91, p. 54). Particularly tricky are asymmetrical sheet structures. "Coextruding symmetrical structures allows a processor to match resin viscosities," notes Quantum technical-service specialist Paul Jackson. "However, coextruding asymmetrical structures dictates that material viscosities be quite different to obtain minimal interfacial shear stress between layers." Excessive interfacial shear stress causes coextrusion instability, commonly resulting in poor layer uniformity, weak adhesion, wavy flow appearance, haze, or warpage when the sheet is eventually thermoformed.

Cloeren also designed its own flow-modeling program in-house and uses it to improve equipment's processing efficiency. Cloeren says it has developed a grooved extruder barrel, low-temperature mixing screw, and proprietary die that together can recycle up to 60% of PP/EVOH trim without compatibilizers. Some earlier efforts could use trim with compatibilizers, but only as a separate scrap layer. This meant the cost and complication of an extra extruder and an asymmetrical wall.

Computer simulation also showed Klockner ER-WE-PA in Germany (U.S. offices are in Smyrna, Ga.) how to solve a processing problem involving layers of high- and low-viscosity materials. By inserting one or two layers of graduated blends of the two resins in between layers as a sort of flow compatibilizer, ER-WE-PA found it could coextrude the dissimilar materials with double the throughput, according to R&D manager Dragan Djordjevic.

NEW MATERIALS HELP

New barrier resins allow thinner barrier layers to be used. New EVOH grades from Kuraray have greater heat stability, less water absorption, and greater ability to draw down without rupturing, says Cobelplast's Marynissen. Kuraray's Evalca subsidiary also has a new sheet coextrusion grade with improved thermoformability, and a new processing aid that compatibilizes EVOH and PS multilayer scrap (see Technology Newsfocus).

Desiccant and oxygen-scavenging additives are used to enhance the function of thinner barriers. American National Can Co., Chicago, makes seven-layer barrier microwave soup bowls for Campbell Soup Co., using a desiccant layer attached with adhesive to the moisture-sensitive EVOH. This improves shelf-life over the previous six-layer package. And in Japan, a tiny pack of a ferrous oxygen scavenger enclosed in a shelf-stable food package lets a barrier layer suffice at only 5% of total thickness, instead of 10% without the scavenger. In this country, oxygen-scavenging technology based on proprietary, synthetic molecules is being marketed by Advanced Oxygen Technologies, Inc. (formerly Aquanautics Corp.), Alameda, Calif.

ENVIRONMENTAL STIGMA

Nowadays, when new rigid barrier coextrusion packages are introduced, it's with less fanfare than in the past. "Ten years ago there was a lot of publicity and not much action; now it's the opposite," says Jack Arvin, v.p. of sales and marketing at the Plastic Packaging Div. of Alltrista. One reason for less publicity today is that several packaging companies, both in the U.S. and Europe, have seen multilayer packages killed by negative environmental publicity. "Packaging is a dangerous game," says ER-WE-PA's Djordjevic. "Many people make a beautiful barrier package, but they keep it quite." Cobelplast's Marynissen agrees: "To give big publicity to multilayer can work against you."

The social stigmas against barrier packaging are still there despite the fact that mixed-material scrap is more reusable today. Also as barriers get thinner, they present little if any disruption to current recycling technology. Graham, for example, makes a four-layer HDPE Tropicana orange-juice bottle with a color layer, regrind, adhesive, and thin nylon barrier on the inside. This barrier bottle has a recycling designation "7" on the bottom, for mixed material, but in fact is almost entirely HDPE and can be recycled with HDPE without a problem, Graham says.

Consumers aren't likely to know which multilayer containers present recycling difficulties and which don't. In Europe, for instance, six-layer PP/EVOH bottles are often given a seventh polycarbonate layer for gloss. The PC, however, makes flash impossible to use as regrind. The PC gloss layer was tried in the U.S., but never caught on for environmental reasons.

A FEW NEW BARRIER PACKAGES

Form-fill-seal barrier coex sheet applications are growing slightly. In the past year and a half, Europe has also seen growth in six- and seven-layer sheet with EVOH barrier for controlled-atmosphere food packages. This sheet has a glossy surface of crystal PS, together with HIPS, adhesive, EVOH, adhesive, and finally PE as a chemical and moisture barrier. A seventh modified-PE peel layer may also be used. When this relatively new package is sealed and flushed with a mixture of gases low in oxygen, it can extend shelf life of fresh foods like pasta from one week to six weeks without a thicker barrier. Cobelplast added two new coex lines last year for this market, which Marynissen says is growing 50% a year.

A couple of new rigid barrier coextrusions are notable for how few layers they use. In fact, Harry Lippert, sales manager of Extrusion Dies Inc., Chippewa Falls, Wis., says future barrier coextrusion will see more simple structures of high-tech resins and alloys: "The people emphasizing seven to nine layers are living in the past." A case in point is barrier sheet for clear dairy and deli containers recently introduced by Sandusky Plastics in Sandusky, Ohio. It has only two layers, a clear styrene-butadiene copolymer (Phillips 66's K-resin) and a thin inner layer of PETG as a barrier against both oxygen and fats that would stress-crack the PS.

NON-BARRIER COEX PACKAGING

Less technically challenging, non-barrier coextrusions of sheet and bottles are growing steadily for decorative and recycling purposes. Uses include burying reclaim inside a three-layer sandwich, adding view stripes or glossy coatings, pairing clear tops with pigmented bottoms in clamshell containers, and saving colorant cost by combining pigmented and unpigmented layers. "We hardly build a line without coextrusion these days, but it's not for barrier," says Welex's Nissel.

Sophisticated four-layer coextrusion is used to make PS drink cups with a thin crystal-PS or PC gloss layer, HIPS pigment layer, regrind, and white PS inside. Solo Cup and Sweetheart Plastics Inc. in Somerville, Mass., among others, make these cups. Gloss layers must be super thin or they are too brittle--and expensive, in the case of PC. Gloss layers are now down to 2-2.5% of overall wall thickness from 5% five years or so ago when the product was introduced.

Two- or three-layer coextrusion in PE household-chemical bottles allows post-consumer reclaim to be used up to 100%. Owens-Brockway in Toledo, Ohio, for instance, uses coextrusion to make a 100%-recycled fabric-softener bottle while maintaining ESCR and top-load strength. The patented structure has an inner layer of recycled LDPE, middle of recycled homopolymer HDPE and regrind, and outer layer of pigmented homopolymer PCR.

BARRIER COEX IN DURABLES

To find barrier coextrusion activity today, you may have to look beyond packaging to durable goods, such as three-layer sheet for refrigerator liners made by Maytag's Galesburg Refrigeration Products div. (formerly Admiral Co.), in Galesburg, Ill. This sheet has a glossy outer layer of medium-impact PS, a refrigeration-grade HIPS/regrind core, and a proprietary alloy as a chemical barrier inside. The barrier (from Dow Chemical and BASF) will protect HIPS from attack by the new HCFC blowing agents that will be used to foam the insulation.

Blow-molded plastic gas tanks for cars are an emerging, though not fully established, market for high-barrier coextrusion. The only barrier gas tank in production is still the Nissan Altima's five-layer coextruded tank with nylon barrier at 2.5-3% of wall thickness, made in Smyrna, Tenn., on an accumulator-head blow molder from Japan Steel Works Ltd. (represented in the U.S. by JSW Plastics Machinery Inc., Santa Fe Springs, Calif.).

Also getting into coextruded fuel tanks is Ford Motor Co., which is testing HDPE/EVOH gas tanks with both five and six layers, the sixth being a regrind layer between the adhesive and PE. Ford's Milan, Mich., plant has a new multilayer, single-head continuous-extrusion blow molding machine from Krupp Kautex, Edison, N.J., and a second dual-head machine on order. New barrier materials are being studied, including fluorinated polymers and ionomers, for possible use on 1996-model cars.

Following Nissan and Ford is a new player in plastic gas tanks, Walbro Automotive Corp., whose new plant in Ossian, Ind., has had a Krupp Kautex continuous-coextrusion blow molder since last August and a second, larger one on order. Walbro is prototyping six-layer gas tanks with EVOH barrier.

As with coex sheet extruders, the new coex blow molding machines benefit from greatly improved computer control. "When you add together control of all six extruder speeds (in increments of 0.01 rpm), plus six heat zones in each extruder and multiple fiber-optic temperature sensors in the head, this process would not have been doable four years ago," says John Thorn, Ford supervisor of fuel-tank engineering in Detroit.