This article also appears in
Subscribe now »

The ORNL/RMX plasma oxidation technology is projected to reduce the costs of making carbon-fiber automotive parts.

Plasma process could cut carbon fiber costs

Carbon fiber is to composites as steel rebar is to concrete. When a support skeleton of either is embedded inside a solidifying mass of resin or cement, the strong sinews bind everything together into a resilient, ultra-functional whole.

But just as building contractors have to charge more if rebar prices rise, the same goes for the makers of carbon-fiber composites. Trouble is, carbon fiber always has been too pricey. Although automotive OEMs expect to get carbon fiber at somewhere around $5 per pound, reported Sanjay Mazumdar, CEO at market consultants Lucintel in January, the market price ranges from $10 to $15 per lb.

Carbon fiber’s premium pricetag derives mainly from three factors: high precursor costs, high energy consumption, and long processing times during the fiber-conversion process.

But that could change if a new, lower-cost manufacturing method achieves full industrial commercialization: a novel cold plasma-based processing technology, developed by researchers at Oak Ridge National Laboratory (ORNL) and RMX Technologies of Knoxville, Tenn., can dramatically cut the time and energy that’s expended on the carbon fiber production line, said co-inventor and ORNL principal investigator Felix Paulauskas.

Paulauskas developed the basic concept for the method eight years ago and worked with RMX to develop prototypes and demonstrate the technology at the laboratory scale. In 2014, RMX constructed a 1-metric ton plasma oxidation oven (see

Plasma magic

The ORNL/RMX plasma-processing technology aims to improve what’s called the oxidation stage of the conversion process, said Truman Bonds, the co-developer and RMX plasma engineer, who’s been Paulauskas’ partner over the last decade. The oxidation line is where the precursor rows of polyacrylonitride (PAN) polymers are unrolled into oven chambers to be oxidized (or stabilized), he explained. Oxidization is needed for the fiber to survive the next step in the sequence in which it is cooked down to pure carbon, first at low, then at high temperatures from 1000 to 1700°C.

“During oxidation, the thermoplastic PAN precursor is essentially converted to a thermoset material that can no longer be melted,” Bonds said. But it’s a slow reaction. Oxidation is the most time-consuming and costly phase of production. “It’s also the number one source of problems in the conversion process.”

“It takes a conventional system somewhere between 75 and 120 min to full oxidization,” Paulauska noted. “Our plasma process cuts the time by a factor of 2.5 to 3 times, so we can process fiber in 25 to 35 min. And that’s with equal or better mechanical properties.

Reactive ionic wind

The patented non-thermal, or cold plasma, technology creates a plasma—“a fourth state” of weakly bonded matter, Bonds said. “It’s a collection, or cloud, of charged ions and electrons that we create using applied electric fields.

“The trick that we invented is a way to generate the specific type of plasma chemistry we need directly from the air,” he continued. “We can make a special reactive, oxidative bath that can accelerate the oxidation process at atmospheric pressure, tripling throughput.”

The plasmas help herd clouds of reactive—and so short-lived—chemical species to the fiber surfaces where they can accelerate the diffusion of oxygen and push the oxidation reaction toward the fibers’ centers. “Using fields we can create thermal flows, or use convective forces to induce the gases to move where we want and then keep them in place,” he said. “The new technology could mean using 75% less energy per pound of fiber processed, saving 20% in costs.”

The new process had been proven in the pilot-plant by five unidentified fiber manufacturers which then requested further demonstration at a larger scale, they reported. RMX has licensed the technology and plans to build a plant with a 15,000-lb capacity to show feasibility for mass production. Meanwhile, an RMX subsidiary called 4M Industrial Oxidation has partnered with C.A. Litzler, a Cleveland-based industrial dryer and oven maker, to design and build 175-metric ton oven design by 2017; the subsidiary plans to sell the oven to at least eight prospective industrial customers.

Low-cost PAN precursor

“What the auto industry needs to grow while growing greener is industrial-grade fiber at an acceptable price,” Bonds said. “Our oxidation process gets you a 20% cost reduction, but more is needed. That’s why we’re working with acrylic textile maker Dralon of Lingen, Germany, who’s got a cheaper textile-grade PAN precursor that can be optimized for plasma oxidation.”

The resulting industrial-grade fiber is to provide 600-ksi tensile strength—what’s needed for automotive applications. The textile precursor will bring another 20% cost reduction, he predicted.

The combination of the two technologies could mean big value on any investments by the industry, said Rodney Grubb, RMX’s president. “A conventional plant produces a couple of million pounds of carbon fiber a year. The combination looks like it could save $2 to $3 per lb so that’s worth multiple millions of dollars per year per line.”

Bonds added that RMX has initiated development of new plasma-based and related technology that could reduce the production costs accrued in the other two main stages of the fiber-conversion process—carbonization and surface treatment (that is, the use of sizing powders as a matrix-bonding interface and materials-handling aide). Both efforts are at the early stage.

For surface-treatment work, researchers are studying the use of atmospheric-pressure plasmas that help low-cost sizings and resins to better bond with the carbon fiber's outer skin, which has been rendered almost entirely chemically inert by carbonization. Although it’s still early days, Bonds said, RMX has identified some performance and efficiency advantages over the standard wet-chemical-dip (“wash and dry”), or thermochemical processing techniques.

Continue reading »