in ethylene, fourth in polyethylene, and third in ^-xylene capacity. The company will also be the fourth-largest polyethylene player in the U.S., with a market share of around 12%, according to the Houston-based consultancy TownsendTarnell. The deal also marries the largest domestic ethylene buyer, Chevron, with the biggest U.S. ethylene seller. In addition, Chevron will contribute one of the world's largest styrene and polystyrene businesses. Phillips will add its 50% interest in a polypropylene plant in Pasadena, Texas, as well as strong positions in numerous specialty petrochemical products. Chevron's Oronite lubricant and fuel-additives subsidiary will not be included in the joint venture. Although there is some overlap in major products, the companies do not foresee any problems securing regulatory approvals and expect to complete the deal by mid-2000. Phillips and Chevron say they will most likely exceed their target of achieving $150 million in annual pretax cost savings by the end of 2001. About $85 million of this savings will come from personnel cutbacks; the companies intend to reduce the combined workforce of 6,000 people by 10%. Alexander Tullo
Physicists move closer to molecular lasers Using beams of light to synchronize the behavior of atoms, researchers in Texas have prepared a collection of nearly motionless molecules in a single rotationalvibrational state [Science, 2 8 7 , 1016 (2000)]. The study not only offers insights into fundamental science topics—like properties of quantum fluids— but it may also lead to applications in molecular lasers and other areas. The investigation, which advances understanding of Bose-Einstein condensates (collections of ultracold particles in which all members reside in a single quantum state), was conducted by University of Texas, Austin, associate professor of physics Daniel J. Heinzen and graduate students Roahn Wynar, Riley S. Freeland, Dian-Juin Han, and Changhyun Ryu. Scientists who study low-temperature systems often describe a molecule's energy in terms of its translational or vibrational temperature. A few research groups have thus far prepared
Using a coherent, two-color laser technique (depicted with red and blue arrows), University of Texas physicists stimulate closely spaced pairs of Rb atoms in a Bose-Einstein condensate (yellow, top) to respond to the laser light in concert. Each pair absorbs a photon of one color while simultaneously emitting a photon of the other color. The laser-driven process causes pairs of atoms to undergo a transition to a bound molecular state, creating Rb2 molecules (orange, bottom) in a Rb-atom BoseEinstein condensate. molecules with translational temperatures in the millikelvin range. What's unique in the present study, points out William C. Stwalley, professor of physics and chemistry at the University of Connecticut, Storrs, is that the Texas physicists synthesized ultracold, stationary molecules (rubidium dimers)
in a state-selected manner—that is, the product molecules occupy a particular rotational-vibrational state, not a distribution of states. That accomplishment is a key step toward preparing a BoseEinstein condensate of molecules and ultimately a coherent beam of stateselected molecules—a molecular laser. Starting with established lasercooling and magnetic-trapping techniques to create a Bose-Einstein condensate of rubidium atoms, the researchers use a coherent, two-color laser procedure to photoassociate some of the atoms into Rb2 molecules. The procedure creates molecules with nearly zero kinetic energy in a weakly bound vibrational state. The combination of motionless atoms and molecules permits spectroscopic measurements to be made with very high precision. "This is an extremely clever experiment," remarks Daniel Kleppner, professor of physics at Massachusetts Institute of Technology. "It provides a new way to measure energies of molecular states with unprecedented precision." Kleppner notes that the experimental scheme devised by the Texas group also provides "new opportunities to measure molecule-molecule and molecule-atom interactions." The present advance may make a significant mark in chemistry. As Chris H. Greene, professor of physics at the University of Colorado, Boulder, points out, "If molecular condensates can be created and controlled to an extent comparable to their atomic brethren, it might have a revolutionary impact in chemistry, commensurate with its huge ongoing impact in physics." Mitch Jacoby
Alliances are the rule for making acrylic acid Production alliances are on the rise in the I chemical industry, and nowhere is this more the case than in acrylic acid, where three manufacturing ventures are either starting up or getting ready to do so. Following receipt of government approvals, Rohm and Haas and DegussaHiils' Stockhausen subsidiary say they have formally established StoHaas Monomer, a joint venture for the production of acrylic acid. The deal was first announced in June 1999. By the time the venture is fully operational by year's end, each company will have contributed 365 million lb per year of existing acrylic acid capacity—Rohm I
and Haas in the U.S. and Stockhausen in Germany. The venture will add 220 million lb of capacity to the German plant in 2003. StoHaas is also planning an acrylic acid plant in Brazil, where the partners are pursuing access to the necessary raw material, propylene. Acrylic acid, a fast-growing monomer used to make superabsorbent polymers for hygiene products and acrylic esters for adhesives and paints, is well suited for such ventures because plant construction is very capital intensive. In addition, because the partners in acrylic acid ventures tend to occupy different downstream niches, they aren't competing for customers. FEBRUARY 14, 2000 C&EN
19
n***A/c
Λ * +ΙΊΑ %A#AAI4-
This is the case for a 175 million-lb-peryear acrylic acid manufacturing venture between Dow Chemical and Celanese that just started up at the BSL Bohlen chemi cal complex in the former East Germany. Owned by Dow and managed by Celanese, the new facility is producing acrylic acid and butyl acrylate with tech nology developed at Celanese's Clear Lake, Texas, site. Dow will take about half the plant's acrylic acid output for de rivatives such as superabsorbents and emulsion polymers, and Celanese will be the sole external marketer. Meanwhile, a third acrylic acid man ufacturing alliance is advancing in
Houston. American Acryl, a joint ven ture between Elf Atochem North Ameri ca and Japan's Nippon Shokubai, is pre paring a site for a 240 million-lb-per-year plant expected to start up in late 2001. The project was first announced in 1997 for completion this year, but it has been delayed because of permit problems. Once the facility is onstream, Nippon Shokubai will use its share of acrylic acid in superabsorbent production. Elf will consume its portion in a butyl acry late plant that it is also constructing on the Houston site and in plastics addi tives manufacture. Michael McCoy
Polyvalent ligand trounces Shiga toxin A multivalent carbohydrate ligand fash ioned by researchers in Canada and the U.K. neutralizes Shiga toxin in vitro up to 10 million times more potently than univalent ligands do [Nature, 4 0 3 , 669 (2000)]. David R. Bundle, a professor of chemistry at the University of Alberta, led the eight-member team that synthe sized the inhibitor. Shiga toxin, produced by the bacteri um Shigella dysenteriae, belongs to a family of toxins, including cholera, that are made up of two subunits. The socalled Β subunit binds to susceptible mammalian cells, allowing the A subunit to gain entry and do the toxin's dirty work. A strain of Escherichia colt has picked up the gene for Shiga toxin, Bundle notes, expressing it as the "hamburger toxin" that's proven deadly in under cooked meat. Shiga toxin's doughnutshaped Β subunit is made up of five identical mono mers, each of which has three binding sites with varying affinities for cellsurface carbohydrates. The pentamer locks onto cells by gripping five or more carbohydrate li gands simultaneously. Bundle and his colleagues sought an inhibitor that would mimic this binding approach and latch tena ciously onto the toxin. In designing their molecule, they exploited the crystal structure of the E. coli toxin's Β subunit corn20
FEBRUARY 14. 2000 C&EN
plexed to an analog of its carbohydrate receptor. Initially, Bundle's team designed a li gand they anticipated might have a high affinity for binding sites 11 A apart on the subunit's monomers. That ligand consisted of two trisaccharide units con nected by an 11-A-long tether. But the results were disappointing—"almost useless actually," Bundle says. "Then my postdoc, Pavel I. Kitov, came up with the idea of attaching the trisaccharide dimer on each of five spokes on a core molecule [glucose] in the hopes of 'hit ting' 10 binding sites on the five mono mers." Because of its configuration, the new ligand was dubbed Starfish.
Starfish sticks like Velcro to each of the five Β subunit monomers—but not as expected. Instead of binding to two sites per monomer, it binds only to one. Only five of Starfish's trisaccharide arms are engaged, leaving the other five free to bind another Β subunit. The up shot is that Starfish is sandwiched be tween two subunits. 'That mode of ac tion differs from the one envisioned by rational design," Bundle notes. Nonetheless, "the generation of a high-affinity inhibitor for Shiga toxin by rational design is a landmark achieve ment," says James Paulson, a molecular biologist at Scripps Research Institute, La Jolla, Calif. "Like Shiga toxin, microbial and eukaryotic carbohydrate-binding proteins Oectins) in general have low in trinsic affinities for their ligands, and many laboratories have pursued the de sign of multivalent ligands to increase molar affinities with notable lack of success." George M. Whitesides, a professor of chemistry at Harvard University, calls the work "a spectacular example of polyvalency. It is also another signpost pointing to new classes of medicinal agents that act with targets on cell sur faces (and with oligomeric proteins)" via polyvalent interactions. "It's a beautiful study," comments Laura L. Kiessling, a professor of chemis try at the University of Wisconsin, Madi son. 'The structural studies reveal that these potent ligands function by two mechanisms: They act by the chelate effect, and they cluster two receptor com plexes. These results are highly significant." More work is on Bun dle's agenda. Antidotes to treat Shiga toxin in the gut are in clinical trials, but no compounds are on the hori zon for treating the toxin af ter it enters the blood stream, an event that can trigger kidney toxicity and death. "The Starfish mole cule offers the potential for the design of a potent [in jectable] drug" that can pre vent such toxicity, Paulson suggests. Bundle's group is gearing up to produce Star fish in sufficient quantities to explore that possibility in animal models for "ham burger disease." Mairin Brennan
