A new artificial-nose-like sensor can sniff out organic compounds

Most of us carry around a phone with a microphone and camera that can replicate our senses of vision and hearing. What our phones can’t do, and what no other device can do, is replicate our sense of smell. But now a team of researchers in South Korea has developed an organic electronic device that can, at least partially, detect odor molecules in the air and determine their identity (Sci. Adv. 2024, DOI: 10.1126/sciadv.adl2882).

The critical components of this artificial-nose-like sensor are actually three different human olfactory receptor proteins engineered into and purified from Escherichia coli. The sensor contains sets of those three olfactory receptors bound to a thin graphene sheet. The receptors are then connected to artificial neurons that make up a neural network. When gaseous odorant molecules bind to those receptors, they activate and create an electrical signal across the graphene sheet unique to that odorant.

The resulting activation pattern is what allows the device to “distinguish what the smell is and also the concentration,” according to Joon Hak Oh, a polymer engineer at Seoul National University who led the research team. Today’s commercialized sensors can detect only the presence and concentration of organic compounds in the air, so the ability of Oh’s sensor to determine what compound is in the air makes it “kind of a breakthrough in chemical sensing.”

Jianguo Mei, a polymer chemist at Purdue University who wasn’t involved in the research, agrees and calls the work “a significant leap towards the long-pursued goal of creating an electronic nose.”

In the study, Oh and his team trained the device’s neural network by feeding in four different short-chain fatty acids, ranging from three-to-six carbon molecules. From there, the device could distinguish between those four fatty acids with over 90% accuracy and determine their concentration in an air sample. “Although the chemical structures [of the fatty acids] are really similar, this sensor can distinguish the differences,” Oh says.

Those fatty-acids were selected because their presence is implicated in diseases like gastric cancer, which is why Oh says a sensor like this could eventually aid in medical diagnoses. He also imagines placing sensors on drones to monitor the content of emissions from industrial manufacturing.

For now though, the device hasn’t been thoroughly tested on anything other than these fatty acids. The study’s approach was just a “proof-of-concept starting point,” Oh says. His team is now testing over 20 different receptors to create more advanced smell sensors that could detect other kinds of molecules.