Printer Friendly

How they make PEs.

Sophisticated processors are asking more questions about their polyethylene resins than ever before. One key reason is that today's high-speed machinery is far more sensitive to minute variations in resin rheology--variations that cannot be detected by conventional "data-sheet" properties like melt index and density. Resins that appear to be perfectly in spec can cause production upsets or product quality problems that sometimes baffle the processor and resin supplier both.

"The molder is using more and more sophisticated equipment, so resin subtleties show up that wouldn't have 10 years ago. That's why this new interest is happening," says polymer chemist and consultant John DeManuelle in Houston.

Even without the increasing demands imposed by their own machinery, molders face materials choices that include an extraordinary proliferation of polyethylene products, all claiming distinctive and superior properties. This proliferation reflects the ferment in polyolefins R&D, which is spawning new polymerization and catalyst technologies.

Traditionally, resin companies told processors little about how their polymers were made, but now suppliers say they're letting molders into their tent and exchanging information more openly. That's partly because processors are also more involved in resin selection these days. Processors with proprietary products work with resin suppliers to develop unique grades just for their applications. Custom processors, too, are far more likely today to have a hand in material selection.

As a result of these concerns, one of the country's more technically sophisticated custom molders asked PLASTICS TECHNOLOGY if it would be possible to prepare a guide to "family distinctions" between resins produced by different polymerization technologies. One purpose would be to enable readers to anticipate some of the basic processing or performance characteristics of resins with which they were unfamiliar. And, because of the widespread licensing of polymer technology, such a guide could suggest appropriate alternative suppliers of resins suitable to a particular processor's needs. This article is a response to that request. Although necessarily a somewhat simplified overview of a very large subject, it briefly summarizes the different ways polyethylenes are made today and attempts to identify traits the resins "inherit" from the type of reactor and catalyst that generated them.

Resin companies say they've spent years of advanced R&D to obliterate exactly the distinctions we were looking to identify, so that resins made by different reactor processes will "drop into" the same molding machine or extruder interchangeably. The question remains whether the subtle differences in chain branching, comonomer contents, and molecular-weight distribution have detectable consequences for processors and end-users. "The processor that knows the most about the resins it uses has a tremendous competitive advantage," says Dr. Saleh Jabarin, director of the Polymer Institute of the University of Toledo, Ohio.


All polyethylene is made in one of four basic reactor types: high-pressure, solution, slurry and gas-phase. Each, together with specific catalyst groups, makes certain density ranges and achieves certain performance advantages. Beyond reactor type, each individual plant's reactor unit imparts its own "fingerprint" in terms of quality differences. One resin company has a customer who specifies the exact reactor unit on which his material is to be run. Agrees polymer consultant Kenneth Sinclair of SRI International in Menlo Park, Calif., "It gets down to specific plant configurations, a valve here or there, or something done on a product change-out. One plant makes consistent quality resin, another doesn't."

* High-pressure reactors are the oldest method of making PE, dating back to the early 1940s. They make what's known today as conventional or "high-pressure" LDPE, which was the first form of commercial PE resin. High-pressure reactors are either autoclave cylinders or long-tubes. Autoclaves are essentially heated pressure chambers--usually two to four in sequence. Reaction time is very short--around 30 sec.

Tube reactors are long pipes laid back and forth horizontally in large, walled enclosures or cells. The pressurized tubes may be a quarter of a mile long with thick, cannon-like walls built to contain a series of explosions that travel from one end to the other in a minute or two, creating pressures of up to 40,000 psi and temperatures up to 600 F. Originally, all monomers were put in at one end. Now monomers, comonomers, and fresh catalyst are fed at intervals all along the tube, allowing manufacture of broad-MWD resins.

Autoclaves use various peroxide catalysts; tube reactors use peroxide or air (which generates peroxides at high temperatures) or both. Peroxide generates a free radical, which initiates polymerization by forming a chain. The active end of the chain can "randomly either add another ethylene molecule, transfer to another molecule to start a long chain branch, or |bite back' on itself to form a short chain branch," explains SRI's Sinclair. "All three happen at the same time and their relative rates determine molecular weight and the degrees of long-and short-chain branching."

High-pressure autoclaves are licensed by DuPont Co., Wilmington, Del., and Quantum Chemical Corp. in Cincinnati. High-pressure tubes are licensed by Union Carbide Chemicals & Plastics Co. in Danbury, Conn., and BASF AG in Germany.

* Solution reactors are pots that cook batches of monomer and solvent at high temperatures. They operate over the melting point of the resin in order for the polymer to form in a solution. The diluent or transport medium is typically cyclohexane or isoparaffin. Residence time is short--one to 10 min.

These reactors are something of an anachronism. Phillips 66 Co., Bartlesville, Okla., first developed them, but production rates were relatively low. And there's the high cost of filtering to remove the resin once it's solidified. So most solution reactors were dismantled 20 years ago.

However, they're still used by DuPont Canada, Mississauga, Ont., and DSM NV in the Netherlands. DuPont's multi-zone reactor is the most widely used solution process internationally and is still actively licensed. This flexible system makes resins with the broadest range of melt index (0.2 to 120 MI). The only DuPont-type reactors in North America are at DuPont Canada and at Occidental Chemical Corp., Dallas, whose PE operations formerly belonged to DuPont.

The only other solution reactor in North America is a proprietary and closely guarded process of Dow Chemical Co., Midland, Mich. It differs from the other, "adiabatic," types in being cooled to remove some heat of reaction. Dow's solution process is used to make specialized copolymers with higher-molecular-weight comonomers. Solution reactors are also of current interest because they can be used with new metallocene "single-site" catalysts.

* Slurry reactors are either loops or stirred tanks in series. They're less costly to build and run than either high-pressure or solution reactors, so slurry-made material has replaced high-pressure and solution material in many commodity applications. Loops predominate in the U.S., stirred tanks in Europe and Japan. Loop reactors use isobutane as a diluent; tanks use hexane. Residence time can be less than 45 min or up to 3 hr.

Phillips licensed the original loops in the 1950s; they look like huge, upright paperclips and are used with Phillips' chromium-based or titanium/aluminum catalysts. The catalyst is fed continuously into the slurry in the loop. Polymer is taken out frequently in small amounts. Since Phillips' loop has been licensed for 30 years, lots of variations exist among licensees. Quantum and Amoco Chemical Co., Chicago, turned the loop on its side, though horizontal loops do essentially the same thing as vertical ones. They just require less mechanical sup port structure.

Other slurry variations include that of Showa Denko KK in Japan, which uses loops in series to make bimodal material. And Neste Oy in Finland has developed a "supercritical slurry-in-loop" process using propane as the diluent instead of isobutane. Neste's process, still in the pilot stage, works like a conventional liquid slurry, but is really a gas compressed to a liquid at high pressure. This process will make medium-density PE.

Multiple loop or stirred-tank configurations are more flexible in controlling product properties than a single loop because each reactor in the series can be adjusted separately. Mitsui's liquid-pool reactor series uses a new high-activity titanium-based catalyst to make high-performance resins. Exxon uses Mitsui's technology, and Occidental Chemical Corp., Dallas, uses Nissan Chemical Industries' technology (see PT, Nov. '90, p. 25). Sumitomo Chemical in Japan also has a multi-reactor series.

* Gas-phase reactors look like large inverted onions. In the narrow vertical neck, gas bubbles up through growing polymer particles formed by Ziegler-type titanium catalyst. Licensed by Carbide, Unipol reactors use low pressure (145-510 psi), low temperature (as low as 122 F), and the longest residence time of any process (up to 4 hr).

A new variation is Carbide's two-reactor Unipol II series of fluid-bed reactors, which is similar to a stirred-tank slurry series. The newest PE process, Spherilene from Himont Inc., Wilmington, Del., is a cascade fluid bed gas-phase process. It will be used first to make LLDPE.

Another gas-phase fluid-bed process is licensed by BP Chemicals Ltd. of the U.K. Eastman Chemical Co., Kingsport, Tenn., recently brought on-stream a plant that uses BP's process to make LLDPE film resins and HDPE injection molding grades with high ESCR for crates and pails (PT, May '93, p. 86).


All commercial LDPE of the "conventional" or branched type is made in high-pressure reactors of some kind. LDPE comes in densities of 0.910-0.930 g/cc and is characterized by long- and short-chain branching. Long chain entanglements raise molecular weight and viscosity and give strength in both melt and solid state. Short chains act like little waxy lubricants.

Typically, autoclave resins have much less long-chain branching and narrower MWD than tubular resins. Autoclaves excel at making PE copolymers and terpolymers with polar vinyl monomers, such as vinyl acetate. "No one has been able to synthesize coating grades with success in tubes," says Quantum research scientist Lloyd Pebsworth. Eastman uses autoclaves to make its LDPE extrusion coating grades. DuPont uses autoclaves to make its EVA copolymers. And Quantum uses autoclaves to make high-clarity Vynathene EVA film grades and EnBA copolymers.

Tube reactors excel in production of broad-MWD LDPEs for film. These are typically extra-tough, high-elasticity grades used in garbage bags. Rexene Products Co. (Dallas); Lyondell Polymers Corp. (Houston); Dow; Exxon; Quantum; Union Carbide; Chevron Chemical Co. (Houston); Mobil Chemical Co. (Houston); Novacor Chemicals Ltd. (Calgary, Alberta); and Westlake Polymer Corp. (Houston) all use tubes.

For now, LDPE resins made in high pressure reactors are unique among PEs in terms of clarity, heat-sealability, and suitability for cast film and extrusion coating processes. But Dow intends to develop competing resins with its solution process and brand-new metallocene catalysts. Carbide intends to do the same with its two-reactor Unipol II gas-phase process and metallocene catalysts.


Tough, stress-crack-resistant LLDPE resins took processors by storm 15 years ago when Carbide introduced them. They have 0.915-0.930 density and are characterized by narrow MWD with short, regular side-chain branches, determined by the comonomers used (butene, hexene, methylpentene, or octene).

LLDPE can be made in all four reactor types, with gas-phase the most prevalent. High-pressure autoclave reactors can make LLDPE only by switching from their usual peroxide catalyst. For example, Exxon is using metallocene catalyst in its high-pressure reactor in Baton Rouge, La., to make LLDPEs with special toughness properties. Repsol Quimica SA in Spain, Elf Atochem SA in France, and Sumitomo Corp. in Japan also make LLDPE in high-pressure autoclaves using conventional Ziegler titanium catalyst.

Solution reactors are the most convenient way to make higher-melt-index LLDPE. Solution reactors can efficiently incorporate higher-MW comonomers like octene. Dow's solution plant uses octene to make Dowlex LLDPE. DuPont Canada's solution plant uses octene to make Sclair LLDPE injection molding and film grades, including ones for specialty films like high-strength milk pouches and low-gel lamination grades. DSM's single-vessel solution plant also makes octene LLDPEs for film and injection molding. Octene LLDPEs are known for superior physical properties that have only recently begun to be challenged by gas-phase "super-hexene" LLDPEs from such firms as Quantum, Carbide, Novacor, and Mobil.

Long ago, LLDPEs were made in loops, but inefficiently. Recently, Phillips announced new LLDPE slurry-loop materials (PT, Nov. '92, p. 63;June '93, p. 70). These are for film, injection and blow molding. They have medium MW and broader MWD than other LLDPEs for greater toughness, Phillips says. These resins are also said to have a broad processing window. Himont's new Spherilene process will make typical butene LLDPE film resins, as well as higher-alpha-olefin film grades that reportedly process like LDPE (PT, April '93, p. 105). Other Spherilene LLDPEs will include HMW, high-ESCR, and injection molding grades.

The low-pressure gas-phase method is what put LLDPE on the map. (LLDPE was initially known as "low-pressure" resin.) Gas-phase processes make standard film and injection molding grades of butene- and hexene-type LLDPE. BP's gas-phase process also makes a lamination grade that's especially known for low gel count and low odor because of its very uniform, narrow MWD. Gas-phase reaction is limited in molecular weight by the comonomers it can vaporize economically, like butene and hexene rather than octene. These form short side-chain branches and lower resin density.


"Very-low" or "ultra-low" density LLDPE, known as VLDPE or ULDPE, has densities of 0.900-0.915. Dow and Union Carbide are the only domestic makers of VLDPE, though Exxon's new 30-million-lb/yr pilot plant in Baton Rouge, La., makes developmental VLDPE and PE "plastomers" down to 0.870 density (PT, Nov. '91, p. 17). That plant uses high-pressure autoclave technology from Mitsubishi Petrochemical Co. of Japan and Exxon's own metallocene catalyst to make the brand-new Exact series. Dow makes its Attane ULDPE resins in solution. DSM, Mitsui, Sumitomo and Elf Atochem make VLDPE abroad. Like Exxon, Sumitomo makes VLDPE in a high-pressure reactor. Like Dow, DSM in Holland and Mitsui in Japan make VLDPE using Ziegler catalysts in solution.

Carbide's Flexomer VLDPE film resins are made in a Unipol gas-phase reactor with a proprietary catalyst (which Carbide doesn't license) that makes a broad-MWD material. Carbide also uses Ziegler catalyst to make a narrow-MWD VLDPE similar to Dow's and Mitsui's.


HDPE with density over 0.940 was the first "linear" PE. Homopolymers are characterized by very long, unbranched carbon chains; copolymers have long chains with a few (1-1.5%) side chains of hexene or butene comonomer. High-pressure reactors can't make HDPE at all because they add too many branches. HDPE is made by solution, slurry and gas-phase processes.

The slurry reactor is the principal method used to make commodity blow molding grades of HDPE in the U.S. Phillips' material has a medium to high MW and relatively broad MWD, conferring both toughness and processability. Large U.S. blow molders with high-speed wheel machines use mainly Phillips-process material. Phillips catalyst, however, isn't used to make resin with MI over 2 g/10 min, though it is used to make HMW-HDPE.

Substituting Ziegler titanium catalyst in a conventional Phillips slurry loop (as is done by producers like Quantum and Solvay Polymers, Houston) makes narrower-MWD HDPE for injection molding and film. These resins are said to be inherently glossier because they have smaller crystals.

Paxon Polymer Co. in Baton Rouge, La., another Phillips loop-reactor licensee, also uses a unique BASF catalyst that adds an extremely high-MW "tail" onto the polymer molecule, providing higher-HLMI resins for blow molding 55-gal drums. High-MW tails also broaden MWD and impart greater elasticity. Gas-tank blow molding grades are all made by Phillips-loop licensees (BASF, Solvay, Paxon and Showa Denko) because of this extra elasticity.

European and Japanese slurry processes, which use a cascade of tanks, can also make HMW-HDPEs with broader unimodal, bimodal, or multimodal MWD. Hoechst AG in Germany was first in the world to make bimodal HDPE commercially in the early '70s. Japanese producers followed suit because they already used a tank-reactor cascade to increase residence time. Asahi Chemical Industry KK and Mitsui Petrochemical Co. Ltd. use such tank-slurry processes.

A disadvantage of some bimodal materials that use zinc stearate as a processing aid and catalyst scavenger is that they leave trace metals in the resin. These resins may not be suitable for pharmaceutical or other applications with tight extractable specifications, says Martin Germak, materials lab supervisor of drum maker Smurfit Plastic Packaging Inc., Wilmington, Del.

Narrow-MWD resins made with Ziegler catalyst "have a tendency to fail ESCR tests for bottles and have a greater tendency toward melt fracture under blow molding conditions," says James Fargher, head of materials and development at Owens-Brockway in Toledo, Ohio. Melt fracturing makes "alligator skin" on a bottle surface.

Low-pressure gas-phase processes make medium-MW HDPE for food wrap and HMW for film, blow molding, pipe, and sheet. Gas-phase materials seem to be more consistent in shear properties than materials from other processes because they have fewer low-MW ends, says Smurfit's Germak. The first Unipol gas-phase blow molding HDPEs weren't accepted in the U.S. because blow molding machines were already designed to run Phillips-type material, which has different die-swell and flair characteristics. So Carbide used a chromium catalyst to come up with HDPE blow molding grades that mimic Phillips material, but "our ESCR is better," Carbide says. These Unipol-process HDPEs now have 15% of the milk-jug market.

BP's gas-phase process uses Ziegler catalysts to make LLDPE film and HDPE injection molding grades. It isn't used much commercially to make pipe or HMW resins, although BP does license chromium catalysts to make these grades.

Among solution processes, DSM's makes HDPE resins for injection molding using butene and octene comonomers. DSM can't go below 0.8 MI, so it doesn't make HMW-HDPE film or pipe grades. DSM licenses its process to Idemitsu Petrochemical Co. in Japan and to producers in Korea and Taiwan.
COPYRIGHT 1993 Gardner Publications, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1993, Gale Group. All rights reserved. Gale Group is a Thomson Corporation Company.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:polyethylenes
Author:Schut, Jan H.
Publication:Plastics Technology
Date:Sep 1, 1993
Previous Article:Low-pressure alternatives for molding large automotive parts.
Next Article:Twin-screw machines explore solid-state extrusion.

Related Articles
PTT Petrochemical Co. Ltd.
Two Tough Pes For Rotomolding.
Chevron changes PE brand names. (In Brief).
Vegetarianism and male fertility in the presence of environmental pollutants. (Scientific update: a review of recent scientific papers related to...
Modified time-based servo enables increased track density on tape.
Modifier boosts flow of HDPE & LLDPE.

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters