In With The New

For 17 years, chemists and chemical engineers at Lucite International worked to reinvent the process for making methyl methacrylate (MMA) that Lucite's predecessor company, ICI, pioneered in 1933. They succeeded, and only a year after Lucite opened the world's first facility to use the new process, the firm is already planning an even larger plant.

For Lucite, which this year was acquired by Mitsubishi Rayon, the new technology frees the production of MMA from some of the fetters of the old process. The new route, dubbed Alpha, uses feedstocks that are more readily available, and it doesn't yield unwieldy by-products.

Most of the world's production of MMA, used to make polymethyl methacrylate polymers, is based on the old ICI route. Known as the ACH process, it begins with hydrogen cyanide and acetone, which are combined to make acetone cyanohydrin (ACH). This intermediate is reacted with sulfuric acid and then methanol to make methyl methacrylate and ammonium bisulfate.

David Johnson, the senior chemist at Lucite who has led the Alpha effort since the beginning, says the ammonium bisulfate by-product is either reacted with ammonia to make ammonium sulfate fertilizer or converted into sulfuric acid and recycled back into the process. "Neither way is attractive, and that really is the biggest pitfall to the ACH process," he says.

The use of ammonium sulfate as a fertilizer is confined to geographical niches such as rice-growing regions or places with a sulfur deficiency in the soil, Johnson notes. Recycling the ammonium bisulfate is an expensive alternative; when the process is deployed, it accounts for half of the capital cost of an ACH plant.

Not surprisingly, chemical makers have been looking for a way around the ammonium bisulfate problem for some time, notes Mark Morgan, principal for energy and chemicals in Europe, Africa, and the Middle East at the consulting firm Nexant. For instance, Evonik Industries' new Aveneer MMA process starts with the same basic raw materials as the ACH process but does not produce sulfate by-products.

Another drawback of the ACH process is the need for hydrogen cyanide. HCN can be procured by reacting ammonia with methane, but synthesis is capital intensive and offers only about 60% selectivity, Johnson says.

It can also be purchased from an acrylonitrile plant, which makes HCN as an unwanted by-product. HCN can be about 15% of the output of an acrylonitrile plant, Johnson says. But the use of acrylonitrile in acrylic fibers is in decline, he notes, crimping new HCN supplies. Acetonitrile, an acrylonitrile coproduct used as a solvent in high-performance liquid chromatography, has been in short supply for the same reason.

Johnson observes that the world's last major acrylonitrile plant opened in 2004 in Shanghai as part of the SECCO petrochemical joint venture between BP and Sinopec. Lucite snapped up its HCN output a year later when it completed an MMA plant in Shanghai.

Nexant's Morgan agrees that sourcing cyanide is a problem. "The growth in HCN supply from the acrylonitrile source cannot keep up with MMA demand growth," he says. "To maintain supply to meet MMA industry demand growth in the future, investment in alternative technologies to the ACH process is needed."

In 1992, Lucite, which was then part of ICI, decided to find a way around the ACH process. Charged with researching new routes, Johnson found many potential ways to make MMA that, as he puts it, "never saw the light of day." Eventually, the Alpha process emerged as the leading contender.

The process starts with ethylene, methanol, and carbon monoxide, which are reacted together in one step to make methyl propionate. It, in turn, is combined with formaldehyde to make MMA and water. Some heavy esters—about 5% of MMA yield—are produced in the second step of the process and are burned as fuel.

The three starting raw materials are as simple and abundant as petrochemicals get. "We did a study to look at possible sites for the process," Johnson says. "What we found is that there were so many that it was a waste to do a study. You can essentially build it anywhere."

And there is virtually no limit to how large the plant can be. ACH plants are capped at about 100,000 tons per year by the size of available HCN coproduct streams. Another alternative MMA process, based on isobutylene and practiced by Sumitomo Chemical and Mitsubishi Rayon, has a similar limitation because of the amount of isobutylene that can be secured from ethylene cracker sources.

In 2001, Lucite started up a 10-kg-per-day Alpha pilot plant at its laboratory in Wilton, England. A dozen chemists and eight chemical engineers, representing more than half of the relatively small firm's research department, were employed in the effort.

The group worked out some kinks along the way. The ligands of the palladium phosphine catalysts the company was working with were being used up at a rate of one for every 4,000 methyl propionate molecules produced. Making a more stable catalyst and adjusting operating conditions boosted that number to 10 million. The cesium silica catalyst used in the next step couldn't stand the high process temperatures and lasted only two to three weeks before needing to be replenished. Johnson says the company added zirconium, improving the catalyst's life to three years.

Overall, the pilot program was a big enough success that Johnson was confident Lucite's engineers could design a commercial-scale facility without the intermediate "semiworks-scale" plant that is often tried with new chemical technologies. This decision saved the company about $60 million, Johnson estimates.

Lucite completed the 120,000-metric-ton-per-year plant in Singapore last November. It is no coincidence that Mitsubishi Rayon unveiled its $1.6 billion bid for Lucite on the same day the start-up was announced. Mitsubishi Rayon noted that the acquisition would make it the only MMA producer to use all three major commercial processes.

Since then, Mitsubishi Rayon and Saudi Basic Industries Corp. (SABIC) have signed a letter of intent to build a 250,000-metric-ton-per-year Alpha process MMA plant in Saudi Arabia. Alpha's feedstock flexibility, Johnson notes, is particularly advantageous in the Middle East. He estimates that the location offers costs of only 40% of a similarly sized ACH plant elsewhere. The Singapore unit can make MMA at about 80% of the cost.

A $1.6 billion acquisition and huge breakthrough plants in Singapore and Saudi Arabia—these accomplishments are good indications that Johnson and his band of British chemists did their job well.