Digging in the Dirt Chemical and biological sensors could aid the search for hidden land mines

 Many people regard antipersonnel land mines as the worst form of pollution on the planet. An estimated 100 million of these small explosive packages, designed specifically to maim or kill when stepped on, lie buried around the world, remnants of past military maneuvers or terror campaigns. These devices have turned fields, forests, and villages into treacherous, unusable terrain.

 Nations with the severest land mine problems include Afghanistan, Angola, Bosnia, and Cambodia, where years of war have littered the landscape with the deadly devices. In many parts of these countries, the activities of everyday life threaten civilians with possible injury or death. In Cambodia, 1 of every 250 people has lost a limb or limbs because of an encounter with a mine. These deaths and permanent injuries among the population exact a crushing economic toll from nations that can ill afford it.

 With so many mines already in the ground, the problem of removing them seems almost insurmountable. Even if no more mines were laid, it would take 1,000 years to clear them all at current removal rates, says Ron Woodfin of Sandia National Laboratories in Albuquerque, N.M. A dramatic increase in the number of workers could speed up the demining process, but many people are hoping that new technology can quicken the pace.

 Some of the more promising approaches now in development employ chemical and biological sensors to detect traces of explosives emitted by mines. Such sensors could enable a demining team to survey a large area and identify hot spots quickly and safely.

 Perhaps more important, sensors can indicate where there aren't any mines. The ability to see that an area is safe would allow demining teams to focus their energies on the trouble zones, says Albert M. Bottoms, president of the Mine Warfare Association (MINWARA), an educational organization in Monterey, Calif. However, "even that is beyond our current technology," he says. Bosnia alone has an estimated 18,000 minefields. "To our utter dismay, we don't know how to do a survey of even one of them, let alone 18,000," he remarks.

 The various designs of antipersonnel mines form a gruesome gallery of weapons. According to the humanitarian organization CARE in Atlanta, there are more than 600 types. Explosive blast-effect mines tear off a person's lower leg and drive dirt and bone fragments into the wound. Fragmentation mines are stuffed with ball bearings and metal scraps to pepper their victims with shrapnel. Mines nicknamed "bouncing Betties" pop out of the ground and explode at waist height, shooting out fragments on all sides.

 The military can often use brute force to rid an area of antipersonnel mines. "They're not trying to remove the mines as much as defeat them," says Dick Davis, director for defense programs at Oak Ridge (Tenn.) National Laboratory. A heavily shielded, remote-controlled vehicle can quickly forge a safe, though limited, path for troops by turning up and detonating mines as it rolls along.

 Demining an entire area for humanitarian purposes -- so that children can play and families can move about free of fear -- is much trickier. Workers, usually trained civilians, slide a rod at an angle into the soil, probing gently for any suspicious objects, which include both unexploded artillery and mines. This process, required to clear an area as thoroughly as possible, is dangerous and frustratingly slow.

 Bottoms sees promise in new technologies that might make mine clearance safer and faster. He recalls a videotape he once viewed of "a line of peasants, walking across a field, poking the ground with sticks. Technology has to offer us something better than that."

 Finding effective detection methods is particularly difficult because most of the mines are made of plastic instead of metal, rendering standard metal detectors useless for humanitarian demining.

 Other types of imaging systems, such as infrared detectors or ground-penetrating radar, can alert deminers to objects buried in the ground, but they don't distinguish among mines, unexploded shells, or innocuous metal fragments. "In Afghanistan, they find 115 [harmless objects] for every one that has explosives in it," says Woodfin. Nevertheless, each hit has to be carefully probed, wasting valuable time. The large number of false positives can also cause a deminer's attention to flag, with potentially disastrous results. "False positives are bad, but false negatives are even worse," says Bottoms.

 In order to make detectors that don't cry wolf so often, several groups of researchers are focusing on the essential difference between benign objects and deadly mines: the presence of explosives. Since most mines leak a little bit of explosive into their surroundings, chemical and biological sensors can sniff them out. Deminers can then ignore other buried objects.

 These new sensing techniques are variations on methods developed to detect chemicals and hazardous waste in the environment.

 Most of the "cheap and dirty" mines contain the explosive trinitrotoluene, or TNT, says Woodfin. Some use plastic explosives such as RDX, a favorite of terrorists. 

 At Oak Ridge, scientists are developing sensors that can identify the chemical signatures of these compounds. "None are perfected at this point," says Davis, "but they will prove beneficial in 2 to 4 years."

 The sensors could be included in long-handled, handheld detectors or mounted on a robotic vehicle. Such devices could quickly indicate the presence of mines while the deminers maintain a safe distance.

 Some researchers are working to make a standard laboratory apparatus portable enough to be taken to a minefield. Those at Oak Ridge have trimmed a chemical-sensing machine called an ion trap mass spectrometer to the size of a suitcase, says Davis. The current version, including batteries, weighs 60 to 70 pounds, making it "luggable, not quite portable." Their goal is to get the spectrometer to the size of an attache case and run it off a car battery.

 At Sandia, researchers are scaling down another type of mass spectrometer, originally developed to check airline passengers for explosives.

 A different type of sensor under development depends on a tiny silicon cantilever just 1 micrometer wide that "looks like a small diving board," says Thomas G. Thundat, a scientist at Oak Ridge. A coating on the cantilever -- platinum, for example -- absorbs molecules of explosives. When the cantilever is heated, the miniature explosion occurs at a characteristic temperature, indicating the presence of the substance. Once the explosive burns off, the sensor can be used again within 1 second, says Thundat.

 Other cantilever-based sensors have also been designed to detect individual chemicals by measuring vibrations instead of burn temperatures. Thundat and his colleagues have coated the cantilevers with different materials that selectively absorb a range of chemicals given off by explosives.

 Absorbed molecules change the mass of the cantilever and, therefore, the frequency at which the cantilever vibrates. By focusing a laser on the end of the cantilever, researchers can detect the change in frequency. So far, Thundat says, they can detect environmental pollutants such as acetone, mercury, and toluene -- but not yet its derivative TNT -- using these devices.

 In 1 square centimeter of a silicon chip, researchers can carve out 648 cantilevers using standard circuit manufacturing techniques. Such an array could serve as an artificial nose, sensing a variety of substances without using a lot of power.

 Indeed, sensors can take lessons from some real noses. Dogs can smell explosives with "breathtaking" efficiency, says Bottoms, but they are expensive to train and handle. Sandia scientists are assisting a company called Nomadics in Stillwater, Okla., in developing a chemical sniffer. Trying to emulate what dogs do when they inhale and exhale near the ground, the sniffer stirs up dirt particles, sucks them in, and electrostatically traps them. Traces of explosives that stick to the particles can then be identified.

"By far the largest percentage [of explosives] is adsorbed to soil particles," says Woodfin. "Only minor fractions are found in vapor." Soil concentrates the substances so that they can be analyzed more easily.

 Organisms that can be added to soil also play a part in explosives detection. At Oak Ridge, researchers have genetically engineered bacteria to light up in the presence of TNT. 

 When certain bacteria ingest organic molecules, they turn on the production of regulatory proteins. By inserting a gene for a luminescent or fluorescent protein next to the gene for the regulatory protein, the researchers can induce the bacteria to produce both proteins whenever they come into contact with organic molecules. In this way, glowing bacteria signal the presence of the explosive in the environment.

 Robert S. Burlage and his colleagues at Oak Ridge have engineered several strains of the bacterium Pseudomonas putida to glow with visible or fluorescent light when they scavenge TNT and dinitrotoluene, a related chemical. Burlage is presenting the results of this project at the Third International Symposium on Technology and the Mine Problem to be held at the Naval Postgraduate School in Monterey, Calif., next week.

 Later this year, the group will test the bacteria on a small simulated minefield, spraying the plot of land with the bacteria and waiting 3 hours for them to produce the glowing proteins. Luminescent bacterial strains should be visible to the naked eye, whereas fluorescent ones will require ultraviolet light in order to be seen.

 Burlarge expects eventually to apply the technique to real minefields, using a crop duster to shower an area with the engineered bacteria. Where the bacteria contact explosives, they will give off light that can be mapped from the air or viewed on the ground.

 Burlage's team had previously engineered bacteria to report the presence of several environmental pollutants: toluene, naphthalene, and mercury. In 1996, the Environmental Protection Agency approved the use of these organisms for cleaning up polluted areas.

 A Mines lie buried in so many different environments that no single method can deal with them all. Researchers are therefore developing a variety of technologies. In some places, deminers may need to apply a combination of techniques, each based on a different physical principle, in order to reveal a minefield's secrets. Discovering how to use advanced data analysis to combine these disparate lines of information adds another layer of complexity to the challenge.

 Some land mines create a shorter-term hazard than others. So-called smart mines self-destruct after a set period of time. The United States makes an effort to use only smart mines, says Bottoms. However, a country like China, which has a stockpile of 50 million conventional mines, probably won't replace them in the near future.

 At the upcoming conference, a representative from the U.S. Army has been invited to speak on the Army's efforts to find alternatives to antipersonnel land mines, Bottoms says. "Offhand," he adds, "I can't think of what those would be."