Sensing devices are becoming more adept at detecting toxins and pollutants

Local boozers were breathing a bit easier yesterday following news reports that an audit by the Washington Metropolitan Police Department found some 80 percent of its "breathalyzer" alcohol monitors may have been giving faulty readings.

The impact on current and past DUI cases is uncertain. But one thing is for sure: Breath analysis devices are society's front-line weapon against drunken driving, and they need to be unassailably accurate. Many well-calibrated units are, and the technology has improved dramatically since the original "drunkometer" made its highway debut in 1938.

But even today's most impressive devices may soon seem downright primitive, thanks to relentless progress in microsensor science. In the fast-growing, fast-shrinking field of single-molecule detection, some gizmos are now able to spot substances in the parts-per-trillion range. Sensors known as "e-noses" function as artificial snouts that can identify the barest trace of compounds in the air, while microfluidic "lab on a chip" sensors can flag individual DNA strands and other entities in liquids.

Few of them are commercialized as yet. But the trend portends a revolution in public safety , according to Stephen Semancik, who heads the Chemical Microsensor Program at the National Institute of Standards and Technology. "What we can't smell can hurt us," he says, citing dangers from carbon monoxide to spoiled food, low-level industrial toxins, water contaminants and building fires where "five minutes can be life or death." The smaller the amount you can detect, the earlier your warning.

For example, to be perfectly morbid, in the event that terrorists use chemical weapons, you really don't want to wait until concentrations reach levels at which your Metro seatmate says, "Hey, what's that funny odor?" At that point, even if you just got on the train, you're probably at the end of the line.

In medical applications, Semancik says, "smelling other kinds of molecules might be able to help us in early disease detection." He foresees breath-type detectors deployed to detect a host of maladies, including skin cancer (recognizable now by trained dogs), diabetes (whose sufferers exhale a telltale scent of acetone, the working ingredient in nail-polish remover) and cystic fibrosis (where breath is abnormally acidic and rich in sulfur compounds). If sniffer diagnostics could be used for initial screening in lieu of a battery of expensive and invasive tests, both patients and the health-care system would benefit.

Whatever their uses, the next of microsensors will be small, highly automated and sufficiently sensitive to operate in the chemical cacophony of the real world. There are numerous designs.

In a typical mechanical system, airborne molecules stick to the end of a microscopic cantilever. Just as a diving board vibrates differently if it's occupied by your 6-year-old or your double-wide neighbor, the vibrational frequency of the tiny cantilever changes in response to the mass of the molecule trapped on it and sounds the alarm when its target compound shows up. Other mechanical systems use nanoscale tubes (that is, with dimensions in the range of billionths of a meter) to snag pathogens by size. For example, University of Missouri scientists recently created a glass "nanopore" system custom-tailored to trap single molecules of ricin, a nasty little poison extracted from castor beans.

Optical systems sense changes in the properties of transmitted or reflected light that occur when the molecule of interest binds to another molecule or undergoes a chemical reaction. David Walt of Tufts University devised a novel variation: "My lab recently demonstrated that we could isolate single molecules into very tiny wells etched into the end of a fiber optic bundle," Walt said in an e-mail. The bundle had tens of thousands of such wells, each just big enough for one target DNA molecule coated with special compounds that would react with the DNA to produce fluorescence. When a solution containing the molecules flowed across the bundle, wells that trapped the molecules lit up. Counting the ratio of dark to light fibers in the bundle, Walt's team was able to determine the concentration of molecules in the solution to unprecedented accuracy.

But the ultimate goal is to get out of the lab. "I believe that the focus of this field is moving from 'sensors' to 'point of care technologies' (POCT) that bring bioanalytic methods from the laboratory to the point of need, whether it is the bedside, emergency room or for preventative medicine such as cholesterol monitoring," Jerome Schultz, chair of the Department of Bioengineering at the University of California at Riverside, said in an e-mail. "This type of device would have wide applications for water quality, especially for third world countries, food contamination and bioterrorism."

If such sensors can be made sufficiently small, sensitive and cheap, the public could become a giant roving detection system. "We all have cellphones now," Semancik says. "Suppose we put a chemical sensor in every phone?" If it encountered a threat, it would send out a signal. If the monitoring system started getting lots of the same signals from one place (as determined by the phones' GPS locators), responders could home in on the threat.

Boozers, beware. Before you know it, the old cliche will become the new reality: They can smell you coming a mile away.