The State of the Art of Electronic Noses

Three new e-noses use three different methods to sniff out everything from freon to fatty acids

NASA's E-Nose

courtesy JPL

A rose by any other name would smell as sweet; we all know that. But what about a rose smelled by a non-human nose? What would it smell like?

Well, an electronic nose is no Shakespeare, so you'd lose some of the poetry. But a new generation of e-noses is is poised to give a whole new meaning to the sense of smell.

Electronic noses and sniffers keep airports and the space station safe by noticing the tiniest amounts of dangerous chemicals. They can tell the difference between things like Coke and Pepsi; sick trees and healthy ones; cancer cells and normal cells; even different human organs. Some use polymer-based sensors, which expand or contract based on the substance they're "smelling." Some break up smells into their molecular constituents. Others use nanomaterials; still others act like mood rings, changing color in the presence of certain compounds. In most cases, software analyzes the patterns of each smell to determine what it is.

The field requires a harmony among chemists, biologists, engineers, geneticists and other assorted scientists who frequently don't speak the same language.

Three new e-noses use three different methods to sniff out everything from freon to fatty acids.

One e-nose just came back from six months in space, where it sniffed the air on the International Space Station every few seconds. Various gases that provide lighting and help cool the station, like mercury, freon and ammonia, can be harmful to astronauts' health, so NASA wants to monitor for leaks. The shoebox-sized e-nose has 32 sensors that can detect a wide variety of chemicals; in the latest experiment, it sniffed for 10 contaminants, said Amy Ryan, the project's principal investigator at NASA's Jet Propulsion Laboratory.

The e-nose consists of several polymers, which react differently to various substances. When exposed to a chemical, the polymers change size, which affects the resistance of an electrical current running through them. The changes become a pattern interpreted by a complex algorithm -- our brains use similar pattern-recognition to decipher smells.

"You'd use the array of sensors and look not at the response of one sensor for one chemical species, but the response of the array as a whole, and try to de-convolute what is there," Ryan said. "The advantage is that you can sense more; the disadvantage is that it becomes a computational issue. You need to do a lot more work to be sure you understand the signals."

The e-nose can detect contaminant concentrations from one to 10,000 parts per million. Such a sensitive instrument can even detect the difference between cancerous cells and normal ones. That led to a study this spring involving neuroscientists who used NASA's nose to sniff brain cancer cells.

In especially complex procedures like brain surgery, the line between healthy cells and cancerous ones is not always clear. But the e-nose was able to distinguish normal cells from cancerous ones, a finding that could pave the way for more detailed experiments.

It's just one of a growing number of noses developed for human health. Last month, researchers at the Technion Department of Chemical Engineering and the Russell Berrie Nanotechnology Institute in Haifa, Israel, announced an e-nose using gold nanoparticles that can [ between the breath of healthy people and that of lung cancer patients.

Yet another e-nose developed at the University of Massachusetts-Amherst combines nanoparticles with polymers to detect cancer. That nose uses an array of sensors to signal the presence of any abnormal cell -- one member of the team compared it to a "check engine" light in a car.

And a tree sniffer being developed in California could be modified to look for any number of ailments, in plants or in people.

Scientists at the University of California-Davis are trying to sniff out a tree disease that could wreak havoc on California's $1.3 billion citrus industry. But it could also help doctors looking for new, fast ways to diagnose disease by monitoring for certain metabolic changes.

The Asian citrus psyllid, an aphid-sized insect, can carry bacteria that cause greening disease, which is fatal to citrus trees. Finding infected trees and removing them quickly is key to stopping the disease's spread. To that end, Cristina Davis and a team of researchers at UC Davis are working on a sniffer that can tell the difference between a sick tree and a healthy one.

Breathalyzer for Cancer

Not a breathalyzer test you want to take.Nature Nanotechnology

Living organisms emit a variety of gases and metabolites, which change depending on the circumstances. Changes in those compounds indicate something is happening to the organism. Like a doctor taking your temperature, the sniffer can look for those changes -- but at a molecular level.

"Rather than detecting the infectious agent or bacterium itself, you can monitor for what it is doing to the living system," Davis said. "We are working on developing bio-markers of the symptoms, and simultaneously working on the hardware and software."

The sniffer uses a differential mobility spectrometer to measure ions of certain emitted compounds. Different ions, even very similar ones, move differently throughout an electrical field. If you know how they move, you can tune the field to only let certain ions pass through it, Davis said.

"It's like a tunable ion filter. You can let through the stuff you want to see, or you can have the filter let you see everything that is present," she said.

The result is a complicated data set that tells scientists what's being emitted. In the case of a citrus tree, the data would provide evidence of all the metabolic processes happening in that tree -- you'd just have to write software to interpret it. That could lead to an all-purpose sniffer that can be trained to recognize just about anything.

Such devices could do wonders for medicine, security and even biometrics.

Scientists have long wondered how dogs' superior sense of smell helps them identify their owners. We know dogs can smell a lot more than we can -- a dog's nose has more than 220 million receptors, compared with a human's roughly five million -- but we don't always know what the dogs are smelling.

Yale scientist Juan Fernandez de la Mora and a team from Boecillo Technology Park in Valladolid, Spain, may have an answer.

"Human skin is very special. We transpire more than any other mammal, which provides an effective cooling system to remove heat generated by intense exercise," Fernandez de la Mora said. "Our ancestors were able to hunt many faster but better-insulated animals by trotting after them until they became too hot to go any further."

Human skin constantly emits several vapors, including large quantities of lactic acid, best known as a byproduct of exercise, along with smaller quantities of fatty acids released by bacterial activity.

Every person's emissions seem sufficiently unique to enable recognition, like an olfactory fingerprint. Fernandez de la Mora, a professor of mechanical engineering at Yale and colleagues of the Spanish company SEADM, developed a sniffer that can distinguish those fingerprints. He believes the vapors emitted by human skin and breath allow dogs to recognize their owners and even help mosquitoes choose their human prey.

The e-nose works by charging the vapors with a spray of charged drops and analyzing the resulting ions with a mass spectrometer. The e-nose can also detect the tiniest amounts of explosives, below a few parts per trillion.

Like so many other breakthroughs, the skin vapor discovery was an accident.

"We were studying negatively charged breath vapors originally, and then we saw a lot of contaminants in the lab. It took us a while to realize these contaminants were produced by people, and they came from the skin of people," Fernandez de la Mora said. "The fatty acids in both breath and skin stood out over most other vapors."

With new findings announced every month, the field of e-noses holds plenty of promise.

Davis, whose background includes work in aeronautical engineering, said collaboration is key.

"The people working in this area span a broad range. There are engineers, chemists, medical doctors, plant physiologists -- it's a big big spectrum. It takes all of those people with their unique perspective to help to solve these problems," she said.