Roughly five million liters of dispersants have now been used to break up the oil spilling into the Gulf of Mexico, making this the largest use of such chemicals in U.S. history. If it continues for 10 months, as long as Mexico's Ixtoc 1 blowout in 1979 in the same region, the Macondo well disaster has a good chance of achieving the largest global use of these chemicals, surpassing 10 million liters.

And there is no doubt that dispersants are toxic: Both types of the dispersal compound COREXIT used in the Gulf so far are capable of killing or depressing the growth of a wide range of aquatic species, ranging from phytoplankton to fish. "It's a trade-off decision to lessen the overall environmental impact," explained marine biologist Jane Lubchenco, director of the National Oceanic and Atmospheric Administration (NOAA), at a press conference on May 12. "When an oil spill occurs, there are no good outcomes."

The trade-off in this case is the addition of toxic chemicals in a bid to protect the marshes of Louisiana and the beaches of Florida. But the U.S. Environmental Protection Agency (EPA), for one, has become concerned about the toxicity of the most-used dispersant at the Gulf of Mexico spill—COREXIT 9500—and ordered BP to look at alternatives. (COREXIT 9527 was used earlier during the spill, but it was discontinued because it was considered too toxic.)

The problem? The EPA's industry-generated data is unclear as to the relative toxicity of various dispersants. "If you think the data on COREXIT is bad, try to find any decent toxicology data on the alternatives," says toxicologist Carys Mitchelmore of the University of Maryland's Chesapeake Biological Laboratory, who helped write a 2005 National Research Council (NRC) report on dispersants. "I couldn't compare and contrast which one was more toxic than the other based on that."

Dispersed oil
Both COREXIT 9500 and 9527 are produced by Naperville, Ill.–based Nalco, a company better known for its water purification work with the oil industry. "For every barrel of oil produced, 3.5 barrels of water are produced," explains chemist Mani Ramesh, chief technology officer for Nalco. "That needs to be treated before it can be released. That water treatment has been a core area for us."

But at the same time Nalco keeps busy cleaning the oil industry's water, it also provides COREXIT, a product to minimize the impact of any oil that spills into the water. Developed in a joint venture with ExxonMobil, the compound is largely made at facilities in Sugarland, Tex., and Garyville, La. The company expects to sell some $40-million worth of COREXIT as a result of the latest spill. "What the dispersant process enables is to prevent the oil from reaching the shore and converts that oil to easy food for naturally occurring microbes," Ramesh says. "If the oil reaches the shore the decomposition rate of oil is so low it would remain on the shore for probably 100 years."

By last week, the EPA and Nalco had both released the ingredient list for COREXIT 9500 in response to widespread public concern. Its constituents include butanedioic acid (a wetting agent in cosmetics), sorbitan (found in everything from baby bath to food), and petroleum distillates in varying proportions—and it decomposes almost entirely in 28 days. "All six [ingredients] are used in day-to-day life—in mouthwash, toothpaste, ice cream, pickles," Ramesh argues. "We believe COREXIT 9500 is very safe."

The U.S. Centers for Disease Control and Prevention agrees, noting in a document for health professionals that "the dispersants contain proven, biodegradable and low-toxicity surfactants," which are "detergentlike" and "in low toxicity solvents."

However, those solvents—petroleum distillates—are also known animal carcinogens, according to toxicology data, and make up 10 to 30 percent of a given volume of COREXIT. And those same everyday products can be deadly to wildlife. "It's the same products in Dawn dishwasher soap," Mitchelmore notes, which is being used widely to clean up oiled birds and other animals. "I wouldn't want to put a fish in Dawn dishwashing soap either. That would kill it."

As a result, the EPA ordered BP to stop spraying dispersants on the oil slick on May 26. The EPA also ordered BP to look for less toxic alternatives on May 20, and the company responded in a letter dated that same day that "BP continues to believe that COREXIT EC9500A is the best alternative." The dispersant continues to be sprayed onto the ongoing oil spill.

No alternative
One reason BP can make such claims is due to a lack of clear data on any of the alternative dispersants. As part of the National Contingency Plan required for offshore drilling, one of 18 EPA-approved dispersants must be on hand to handle spilled oil. Each of those dispersants has been preapproved for use, and each of those dispersants has been tested—by the companies that make them—for toxicity using representative species of estuarine shrimp (Mysidopsis bahia) and fish (Menidia beryllina). Specifically, these animals are exposed to a mix of one liter of dispersant for every 10 liters of heavy fuel oil in water.

Yet, the results of those tests vary wildly, from toxic impacts occurring at levels of just 2.6 parts per million for COREXIT to 100 ppm for another dispersant, NOKOMIS 3-F4. That suggests to experts that the tests which showed lower toxicity may have employed heavy fuel oil that had lost its potency. After all, volatile organic compounds in oil evaporate quickly when exposed to air and can even wash off in water. "These are order of magnitude differences," Mitchelmore notes. "A lot of that can relate to how those tests were set up."

Adds Nalco toxicologist Sergio Alex Villalobos, "If the oil is aged, then the oil loses its toxicity. Using an oil that is not very toxic, if you disperse that oil you are going to get very favorable numbers. Do those numbers really exist?"

EPA, for its part, did not show the best understanding of toxicological data in making its recommendations, urging BP to use dispersants with less than a certain cutoff of toxicity (pdf). Of course, in toxicology the lower the concentration the more toxic a given substance is. "They completely got that wrong," Mitchelmore says. EPA is now undertaking its own toxicology testing of COREXIT and Louisiana crude oil, but results are pending.

Nevertheless, just 20 ppm of COREXIT 9500—or one drop in 2.5 liters of water—inhibits growth of Skeletonema costatum, a Gulf of Mexico diatom, according to toxicology test data presented in the 2005 NRC report. It appears to inhibit the phytoplankton's ability to perform photosynthesis, specifically blocking part of the biochemistry that enables the photosystem II complex, Villalobos says. "Skeletonema seems to fall among the most sensitive ones," he says. "Like many aquatic plants, these are organisms that are resilient, that tend to come back even though you wipe them out in some cases chemically."

COREXIT is also not approved for use in U.K. waters because it fails the so-called "limpet test". That test involves spraying the dispersant and oil on rocks and seeing if limpets (a type of small mollusk) can still cling to them, a test which COREXIT and many other dispersants with slippery surfactants fail. "This is not a product for rocky shores," Villalobos says. "These are only for open sea waters."

Novel use
Of course, in the case of the oil spewing from BP's Macondo well in the Gulf of Mexico, COREXIT is being used in another unapproved way. A wand from one of the remote-operated robots has sprayed more than 1.5 million liters of dispersants directly onto the escaping oil and natural gas roughly 1,500 meters beneath the ocean's surface. "I don't think anybody knows what would happen by applying the dispersants at depth," Ramesh says. "We do not have any knowledge that would allow us to predict what would happen."

In addition to creating subsurface plumes (and providing a rich feast for oil-eating microbes), it remains unclear what kind of dosage of dispersed oil sea life throughout the water column is facing. NOAA measurements show that levels reach 100 ppm of dispersed oil in the first half-meter of water, dropping to 12.5 ppm at 10 meters and unknown levels even deeper. "There isn't any information on what is the environmentally relevant level of dispersant," Mitchelmore notes. "Dispersed oils are going to be toxic, particularly in the top 10 meters that contains all the sensitive life stages. Anything that has sensitive membranes can be affected by dispersants and dispersed oil."

Sunlight falling on the dispersed oil may make the problem worse through a phenomenon known as phototoxicity. Compounds in the oil act as a catalyst to transfer some of the sun's energy into oxygen, converting the latter to a more reactive state that can literally burn up cells. And as fish and other sea life ingest the dispersed oil, it can be broken down into more toxic by-products. "What do these things break down into?" Mitchelmore says. "In toxicology it's quite often not the original compound that's the toxic entity."

Ultimately, the problem is that too little is known about the dispersants and the dispersed oil. "Given that this is a billion-dollar industry, why were those data gaps not filled?" Mitchelmore asks. "The whole issue regarding limited toxicity data—that's not just common to dispersants, that's common to tens of thousands of chemicals we're putting out into the environment daily."

After all, it was only after decades of using bisphenol A, polybrominated flame retardants and other chemicals that significant concerns began to manifest. In effect, usage replaced safety testing—and that's exactly what is happening with dispersants and the massive spill in the Gulf. Different regulation of chemicals and the chemical industry might forestall toxicological mysteries like those surrounding dispersants—and their thousands of chemical cousins—in the future.

"We're using an awful lot of dispersants," said EPA administrator Lisa Jackson during the same May 12 press briefing on the chemical's use at which NOAA's Lubchenco spoke. "This is going on longer than one might have known on day three or four. We're still dealing with a constant release of fresh oil and we need to continue to disperse."