On March 22, 2000, the U.S. Environmental Protection Agency (EPA) announced that it will ban the gasoline additive MTBE as a risk to public health and the environment. The agency expects the ban to take up to three years to implement.

Maybe you first read about it in the newspapers. Maybe you saw it on "60 Minutes."

Or maybe, worse still, you've encountered it personally.

You turn on your tap and the water coming out smells like turpentine. Awful to drink, no good for cooking, appalling in the shower.

If you're experiencing this now, it's probably cold comfort to know that what has contaminated the water you drink is the result of a perfectly laudable attempt to clean up the air you breathe.

And small consolation to discover that you're not alone. The problem is widespread and getting worse. Some believe it will be the coming decade's biggest environmental problem.

What's going on here? Let's go back to the beginning.

RFG -> MTBE
In 1990, the United States Congress passed the Clean Air Act, which mandated the use of reformulated gasoline (RFG). Petroleum refiners were directed to add a compound to their gasoline formulations that would enable more complete combustion and thus reduce emissions. Most refiners chose a compound called MTBE, methyl tertiary-butyl ether, because it was plentiful, cheap, and efficient.

And it worked. Our smog-saturated cities are actually beginning to breathe easier.

Unfortunately, RFG is stored in tens of thousands of leaking underground storage tanks (LUSTs). Worse, MTBE is more soluble in water than gasoline is, so it tends to migrate into water faster than the gasoline does, seeping into the ground and leaching its way into our wells and reservoirs, our water glasses and shower stalls.

All over North America, people are gagging on MTBE. A minimum estimate of impacted sites in California alone exceeds 10,000 and virtually every state in the Union is suffering from MTBE in some measure. An EPA advisory places the acceptable limits of MTBE in water at 20-40 parts per billion (ppb). In California, concentrations have been found as high as 100,000 ppb. And while there are no conclusive data as yet on the effects on humans of drinking MTBE-contaminated water, in laboratory tests, animals directly injected with MTBE developed cancer and non-cancer effects at high levels of exposure.

Demand for MTBE has reached 265,000 barrels/day. Yet a small cup of the stuff is enough to contaminate a substantial city reservoir. Clearly, even with the new EPA ban in effect, this is one problem that's going to get worse before it gets better.

A blue ribbon panel convened to consider the problem estimated that 20,000 new releases from LUSTs were taking place annually. Anecdotal reports from California, Maine and Delaware indicated that even upgraded LUSTs continue to have releases.

What to do about it?

Good news and bad news
As always, the first step in remediation of a pollution problem is detection. As it happens, there is no EPA method directly connected with MTBE detection. There is, however, EPA Method 8260 for detecting a whole laundry list of volatile organic compounds (VOCs). The method uses purge and trap (P&T) with gas chromatography and mass spectrometry (GC/MS).

The good news is, even though MTBE isn't on the EPA Method 8260 list, the method works for MTBE. The bad news is, it costs about $150 and takes about 20-30 minutes per sample.

There are other problems with purge and trap, as well, which Agilent's Mike Szelewski and Bruce Quimby have addressed and resolved. They have written a paper on the subject, which states in part:

While P&T allows analysis at very low levels, problems arise with samples containing unexpectedly high levels of volatiles. Instrument contamination and subsequent carryover result in reduced productivity and higher cost.

"We were sitting around the coffee pot," says Mike Szelewski, "trying to think of ways to do the EPA method simpler, faster and cheaper. And we thought about headspace analysis."

The Szelewski-Quimby paper continues:

Prescreening using headspace analysis can prevent instrument contamination problems. Lab productivity is also increased with prescreening, because the approximate concentration range of analytes is known before P&T. Re-work of samples outside the P&T calibration range is eliminated. Ambient headspace is a fast, low-cost technique for analyzing non-polar volatiles in water. It can be used instead of [the more expensive] normal heated headspace for prescreening.

"Ambient headspace chemistry has been around for years, but nobody's really been promoting it," says Mike. "We had satisfied ourselves that it would work on the EPA's volatiles list. And on the second round of the coffee pot, Bruce said, "Hey, I wonder if we can see MTBE with it."

Sure enough, there it was. Not only that, but the detection limits were actually below the EPA's recommended action limit of 20 ppb.

Salting out
So what exactly is "ambient headspace analysis"? "If you have a bottle partially filled with water," says Mike, "the gaseous area above the water is called the head space. And all volatile compounds - the 8260 list and MTBE as well - just sitting at room temperature establish an equilibrium between the water and the head space."

To concentrate the volatiles in the vial head space, Szelewski and Quimby use a technique called "salting out". They take a standard two-milliliter autosampler vial and half fill it with the sample in question. They then saturate the sample with an inorganic salt - in this case, sodium sulfate. This forces more of the non-polar organics into the head space of the vial. No heat required, hence the term "ambient".

Conveniently, the Agilent 7683 automatic liquid sampler (ALS) accepts gas-tight syringes up to 100 uL. That permits the whole process to be automated. Once the samples have been salted, they go into a regular autosampler routine, the syringe goes in, sucks out 50 uL of gas and injects it directly into the GC/MSD.

Just eight minutes later, not only MTBE but all 65 volatiles on the EPA 8620 list separate out.

The technique actually works better than expected. Linearity and repeatability are good enough that the method can be used for quantitation, not just pre-screening

"The really neat thing is, we have built a library of compounds that, based solely on retention time, lets you identify the compounds," says Szelewski. "So you can either do this on just GC, the inexpensive solution, or you can do it on the more expensive GC/ mass-spec solution if, in addition to retention time confirmation, you want mass spectral confirmation."

The bottom line: MTBE isn't going away any time soon. But the techniques for its rapid and inexpensive detection, which will lead the way to faster remediation, are now in place.

###