The nose knows: Electric schnoz can smell when your food’s gone bad

Most of us have used the sniff test to decide whether a slightly expired bottle of milk or a week-old box of takeout is still good to eat. But while the human nose can be quite astute, it doesn’t always catch everything. Each year, millions of people in the U.S. are sickened by food-borne pathogens that thrive in undercooked or spoiled food.

Luckily for our collective stomachs, a new “electronic nose” developed at UC Berkeley can detect the scents associated with spoiled food much more accurately than the human nose. It can also sniff out the presence of common food allergens, like walnuts and peanuts, which can be deadly for those with sensitivities. The nose is described in a new study published today (June 17) in the journal Science Advances. 

“I think ‘smart’ fridges — which come with sensors that you can control on your phone — would be a great application for this kind of technology,” said study lead author Carla Bassil, a Ph.D. student in electrical engineering and computer sciences at Berkeley. “How great would it be if your fridge could tell you, ‘Hey, your broccoli's going to go bad soon, so you should probably eat that’? Or, ‘Your chicken is on its last day’?”

The new artificial schnoz nose is made up of an array of 16 tiny gas sensors, each of which is sensitive to a slightly different combination of gaseous compounds. 

“You can think of it like a set of digital taste buds, where each sensor on this chip responds uniquely to the various gas molecules presented to it,” Bassil said in a UC Grad Slam talk about her research. “Each of these 16 sensors has a different sensing film on it, and it works by converting chemical reactions between the sensor surface and the gas molecule into electrical signals.”

Using machine learning, Bassil trained a model to recognize the sensor response profiles associated with seven different foods: strawberry, blueberry, banana, walnut, hazelnut, cashew and peanut. She also trained it to recognize the scent of raw chicken, milk and eggs when they were fresh and when they had been left out at room temperature for 24 hours and 48 hours. 

Bassil found that the nose was sensitive enough to smell 0.05 grams of isolated walnut, which is about one hundredth of an average shelled walnut. However, she has yet to test the sensitivity of the device in environments where other gases are present, such as when walnuts are in a salad or a cake, or when spoiled food is in a refrigerator with other foods. 

“The idea is that we can use the relative selectivity of the gas sensors, paired with the pattern recognition abilities of machine learning, to sort out which gas fingerprint is associated with each food,” Bassil said. “The result is a sensor chip that is far more sensitive and far more objective than any human nose can be.”

While the concept of the electronic nose has been around since the 1980s, bringing the technology to life has been tricky. Single gas sensors, like those found in the carbon monoxide detectors in your home, are relatively straightforward to manufacture. But integrating an array of different sensing films onto a single chip is a lot more difficult. 

Bassil overcame many of these challenges by using carbon nanotubes as the conducting material, rather than metal oxides. Carbon nanotubes can form layers that are only around a few nanometers thick, which is the same as just a few atoms, or one one hundredth of a human hair. Their large surface area gives them many special qualities, including being highly sensitive at room temperature. 

Using a device structure that works at room temperature — rather than needing to be hot — gave Bassil the ability to choose a wider variety of different gas-sensitive materials, including those that might degrade at high temperatures such as polymers. It also allowed her to fabricate the sensing chip using a simple process called drop casting, rather than requiring more complicated technique.

“The truly scalable aspect of my electronic nose is that we can use all these different types of sensing materials while depositing them all in a single step,” Bassil said. 

Though not included in the new study, Bassil has now created a portable version of the e-nose that can be operated with an iPhone app. She plans to test the next generation of the device in a wider variety of environments while continuing to improve its sensitivity and reliability. 

Ali Javey, the Lam Research Distinguished Chair in Semiconductor Processing at Berkeley, is the senior author of the study. Additional authors include Kichul Lee, Xun Liao, Divya Krishnan, Yifei Zhan, Theodorus Jonathan Wijaya and Edward Hester of Berkeley; and Minhyun Kim, Il-Doo Kim and Inkyu Park of KAIST.