MTBE: Properties, Production, and Environmental Impact

Methyl tert-butyl ether, known universally as MTBE, is a synthetic chemical that was once blended into gasoline at enormous scale to reduce air pollution from vehicle exhaust. It did that job well. But its other defining property, an extreme affinity for water and resistance to breaking down underground, turned it into one of the most widespread groundwater contaminants in the United States and parts of Europe. The tension between those two facts shaped fuel policy for decades and continues to influence how contaminated sites are cleaned up today.

What MTBE Is and Why It Ended Up in Gasoline

MTBE is a simple ether, a small molecule made of carbon, hydrogen, and oxygen. It is a colorless liquid at room temperature, mixes easily with gasoline, and has an oxygen content of about 18% by weight. That oxygen content is the whole reason it mattered to the fuel industry. When gasoline burns in an engine, oxygen-rich additives help the fuel combust more completely, which lowers carbon monoxide and certain smog-forming emissions from tailpipes. MTBE became the dominant “oxygenate” in reformulated gasoline programs, particularly in the United States after the 1990 Clean Air Act Amendments required cleaner-burning fuels in cities with the worst smog.

By the mid-1990s, MTBE was one of the highest-volume chemicals produced in the United States. It was made by reacting isobutylene (a byproduct of petroleum refining) with methanol over an acid catalyst. The process was relatively cheap, the feedstocks were abundant at refineries, and the product blended seamlessly into gasoline without requiring major changes to engines or fuel infrastructure. At peak use, reformulated gasoline in some regions contained roughly 11% MTBE by volume.

The air quality side of the bargain largely worked. Analyses of oxygenated fuel programs found that MTBE-containing blends reduced carbon monoxide and certain volatile organic compound emissions from vehicles, contributing to measurable improvements in urban air quality.¹ The problem was what happened to the MTBE that leaked, dripped, and spilled before it ever reached an engine.

How MTBE Behaves in the Environment

MTBE has a set of physical properties that make it a particularly stubborn groundwater pollutant. It is highly soluble in water, far more so than the other components of gasoline like benzene or toluene. While those hydrocarbons tend to stick to soil particles and biodegrade relatively quickly, MTBE dissolves into water and travels with it. It does not adsorb strongly to soil or organic matter, so a plume of MTBE in groundwater can spread much farther and faster than plumes of other gasoline constituents from the same spill.

The main sources of contamination were leaking underground storage tanks at gas stations, pipeline breaks, and fuel spills. Because MTBE was present in such large volumes in the fuel itself, even small leaks released significant amounts into shallow aquifers. In the late 1990s, monitoring programs began detecting MTBE in drinking water wells and surface water supplies across the country, sometimes miles from the nearest known gasoline release point.

Once in groundwater, MTBE is resistant to natural breakdown compared to other gasoline chemicals. At two petroleum spill sites investigated in detail, researchers found that natural attenuation (the combination of dilution, adsorption, and microbial activity) was occurring, but biodegradation was the dominant process, responsible for roughly 60 to 80% of total MTBE mass reduction depending on the site.² A separate study estimated the half-life of MTBE in groundwater undergoing natural attenuation at about 23 days under favorable conditions.³ That sounds quick, but “favorable conditions” is doing a lot of work in that sentence. Many contaminated aquifers are oxygen-depleted, cold, or lack the microbial communities needed to break MTBE down efficiently, which means plumes can persist for years or decades.

What MTBE Does to Drinking Water

The most immediate concern for the public was never toxicity at the levels found in drinking water. It was taste and smell. MTBE gives water a distinctive turpentine-like or medicinal flavor at very low concentrations. People can detect it in their tap water long before it reaches levels that would pose a health risk, which created alarm in communities where wells had been contaminated.

Pinning down the exact detection threshold has been surprisingly contentious. A widely cited 1993 study suggested very low thresholds, but a reassessment in 2004 using trained panels could not reproduce those results. The later analysis concluded that the original 1993 study had flaws in its test solutions and should not be used to set cleanup standards, finding that the odor threshold across valid studies was above 15 micrograms per liter.⁴ A Dutch study looking at several gasoline oxygenates found MTBE odor and flavor thresholds in the range of 7 to 16 micrograms per liter, consistent with that reassessment.⁵

Those numbers matter because many U.S. states set advisory or cleanup levels for MTBE in drinking water somewhere in the range of 20 to 70 micrograms per liter, driven more by aesthetic concerns than by health-based risk assessments. The U.S. EPA issued a drinking water advisory of 20 to 40 micrograms per liter based on taste and odor but never established a federal maximum contaminant level for MTBE.

Cancer Risk and Human Health

Whether MTBE causes cancer in humans has been debated since the 1990s. The evidence that triggered concern came from animal studies. High-dose inhalation experiments in rats found increases in testicular tumors and blood cancers in female rats exposed to MTBE.⁶ Those results understandably raised red flags, particularly given how many people were exposed to trace amounts in their water and fuel vapors every day.

But the picture is more complicated than the animal data alone suggest. MTBE is not genotoxic, meaning it does not directly damage DNA the way many established carcinogens do. A detailed integration of the rodent tumor data with information about how MTBE is processed in the body concluded that the tumors seen in rats and mice occurred only at extremely high doses, well beyond what any human would encounter through fuel use, and that the underlying mechanisms were specific to those rodent species rather than relevant to humans.⁷ For example, the male rat kidney tumors and the female mouse liver tumors arose at inhalation exposures of 3,000 to 8,000 parts per million, doses that far exceed anything achievable in a real fuel-use scenario.

The International Agency for Research on Cancer weighed in back in 1999, classifying MTBE as “not classifiable” regarding human carcinogenicity, meaning the evidence was inadequate to draw a conclusion either way.⁸ A more recent systematic evaluation of all the available animal and mechanistic evidence reinforced the conclusion that MTBE does not pose a realistic carcinogenic hazard to people. The scientific consensus, to the extent one exists, is that MTBE’s practical threat to human health at environmental exposure levels is more about rendering water undrinkable due to its taste and odor than about causing disease.

Effects on Aquatic Life

Given how much MTBE ended up in water, its toxicity to fish, invertebrates, and other aquatic organisms matters. The good news, relatively speaking, is that MTBE is not particularly toxic to most aquatic species at the concentrations actually found in the environment. A comprehensive review of freshwater toxicity data found that MTBE caused harmful effects in invertebrates at concentrations starting around 57 milligrams per liter and in fish at concentrations above 388 milligrams per liter. Bacterial assays were the most sensitive, showing effects at about 7 milligrams per liter. Surface water concentrations of MTBE at contaminated sites were typically below 0.1 milligrams per liter, several orders of magnitude below any observed toxic threshold.⁹

Testing in marine species showed a similar pattern. Acute effects ranged from about 166 milligrams per liter for grass shrimp up to 1,950 milligrams per liter for mussels, and reported MTBE concentrations in coastal waters were several orders of magnitude below those levels.¹⁰ MTBE also does not appear to bioaccumulate in fish tissue the way persistent organic pollutants do.

That said, laboratory studies using more sensitive endpoints paint a more nuanced picture. Experiments exposing Nile tilapia to MTBE found reduced growth performance along with signs of liver stress, disrupted blood chemistry, and elevated markers of oxidative damage, effects that would not show up as acute mortality but could affect fish health over time. The study also found that MTBE could accumulate in fish muscle tissue and that its presence amplified the toxicity of other contaminants like tire rubber particles, suggesting a synergistic effect that field-relevant mixtures could make worse than single-chemical tests imply.¹¹

How MTBE Breaks Down in the Atmosphere

Not all MTBE ends up in groundwater. A substantial fraction evaporates during fueling and from vehicle emissions and enters the atmosphere, where it undergoes a different set of reactions. In air, MTBE reacts primarily with hydroxyl radicals and has an atmospheric half-life of a few days under typical conditions. The degradation products include tert-butyl formate, methyl acetate, tert-butyl alcohol, acetone, and formaldehyde. Researchers studying the breakdown pathway in simulated atmospheric water droplets identified more than 15 intermediate products and found that acetone was about 15 times more stable than MTBE under those conditions, meaning it lingers much longer once formed.¹² The atmospheric route also explains how MTBE can appear in rainwater and snowpack, contributing to background contamination far from fuel sources.

Cleaning Up MTBE Contamination

Remediating MTBE in groundwater has proven to be an expensive headache, partly because the same properties that make it spread easily also make it hard to remove. The main engineering approaches include air stripping, granular activated carbon adsorption, and advanced oxidation processes, each with distinct trade-offs.

Air stripping forces contaminated water through a packed tower where air is blown upward through falling water, transferring MTBE from the water into the air stream. It works, and a comparative study of five different groundwaters found it had the lowest treatment cost at higher flow rates, though tall towers were needed for heavily contaminated water. Activated carbon was effective across most conditions but was consistently the most expensive option.¹³ A practical complication with air stripping is that it does not destroy MTBE; it moves the chemical from water into air, so off-gas treatment is usually required. Pilot-scale work achieved greater than 99% removal efficiency from water using air stripping, with the off-gas then passed through activated carbon beds or synthetic resins to capture the MTBE before it entered the atmosphere.¹⁴

Advanced oxidation processes take a fundamentally different approach by chemically destroying MTBE rather than just moving it. Combinations of ozone and hydrogen peroxide generate hydroxyl radicals, which are extremely reactive and break MTBE apart rapidly.¹⁵ One concern with ozone-based treatment is the potential formation of bromate, a regulated byproduct, in water that contains natural bromide.¹⁶ Newer approaches combining ultraviolet light with iron and hydrogen peroxide (a photo-Fenton process) have shown promising results, achieving complete MTBE destruction and nearly 88% mineralization at near-neutral pH, which makes the technique more practical for real-world water treatment where highly acidic conditions are undesirable.¹⁷

Can Microbes Do the Work Instead?

Bioremediation, using microorganisms to eat the contaminant, is often the cheapest and least disruptive cleanup strategy for organic pollutants in groundwater. For MTBE, it is possible but slow and finicky compared to the biodegradation of simpler fuel compounds.

Under aerobic conditions (where oxygen is available), several bacterial strains and mixed microbial communities can break down MTBE. The degradation pathway starts with a monooxygenase enzyme that clips the ether bond, producing tert-butyl alcohol (TBA) as a key intermediate along with formaldehyde. Some microbes can then further degrade TBA, while others treat it as a dead-end product that accumulates in the water.¹⁸ That accumulation of TBA is a practical concern, because TBA itself is a regulated contaminant in some jurisdictions and can linger in groundwater long after MTBE concentrations have dropped.¹⁹

For a long time, researchers believed MTBE could not be degraded under anaerobic conditions, which was discouraging because many contaminated aquifers are oxygen-depleted precisely because other fuel components have already consumed the available oxygen. That assumption turned out to be wrong, though the anaerobic process is much slower. Laboratory experiments demonstrated that aquifer sediments could break down MTBE anaerobically when iron was available as an alternative electron acceptor and humic substances were present to shuttle electrons. An adaptation period of roughly 250 to 300 days was required before rapid anaerobic degradation began, and the end products included carbon dioxide and methane.²⁰ Field evidence supports these lab findings. At a gasoline spill site, MTBE concentrations dropped from over 1,460 micrograms per liter to below 10 micrograms per liter within about 200 days under anaerobic conditions, with TBA rising as MTBE fell, confirming active biodegradation rather than just dilution.²¹

One clever way researchers have confirmed that biodegradation is actually happening underground, rather than MTBE simply dispersing, involves tracking stable carbon isotope ratios. As microbes preferentially consume lighter carbon-12 over heavier carbon-13, the remaining MTBE becomes progressively enriched in carbon-13. At one gasoline release site, the carbon isotope signature of MTBE shifted dramatically along the plume, confirming that anaerobic biodegradation was the dominant attenuation mechanism.²²

MTBE Versus Ethanol as a Fuel Oxygenate

The political and regulatory response to MTBE contamination was, in many U.S. states, to ban MTBE and replace it with ethanol as the fuel oxygenate. California, New York, and about two dozen other states phased MTBE out between 2003 and 2006. Ethanol now serves as the primary oxygenate in American gasoline, but it is not a straightforward upgrade.

A risk comparison found that both MTBE and ethanol fuel blends offer substantial air quality benefits over conventional gasoline. However, ethanol’s advantages are partially undercut by its tendency to increase emissions of certain air contaminants, including acetaldehyde and potentially ozone precursors. On the water side, the trade-off is different but still real. MTBE’s impact on drinking water is mainly aesthetic, making water taste and smell bad at levels well below any health threshold. Ethanol, by contrast, does not persist in groundwater the way MTBE does, but if a gasoline-ethanol blend spills, the ethanol can accelerate the movement of genuinely toxic gasoline components like benzene through the ground, potentially increasing human health exposure to those compounds.¹

That last point is underappreciated. MTBE itself is relatively harmless in drinking water at realistic concentrations, but it signals the presence of a gasoline plume that may also contain benzene and other carcinogens. Ethanol biodegrades quickly and does not create the same taste and odor alarm system, which could paradoxically mean that gasoline contamination from ethanol-blend spills goes unnoticed longer.

Detecting MTBE at Trace Levels

Monitoring for MTBE in groundwater requires sensitive analytical methods because the concentrations that matter, both for taste and odor and for regulatory compliance, are very low. The standard approach uses solid-phase microextraction coupled with gas chromatography and mass spectrometry. This technique can detect MTBE in water down to 0.008 micrograms per liter, far below any taste threshold or regulatory level. The method also picks up related oxygenates and the TBA breakdown product, which is useful for tracking the progress of biodegradation at contaminated sites.²³ The ability to measure such low concentrations is what made it possible to discover how widespread MTBE contamination had become in the first place. It is also what keeps MTBE showing up in monitoring data years after it was phased out of gasoline in most U.S. markets, because legacy contamination plumes are still slowly migrating through aquifers and the instruments are sensitive enough to pick up what remains.