Researchers Develop Groundbreaking Artificial Tongue Using Graphene

Researchers in Beijing have unveiled an innovative artificial tongue capable of learning and identifying various flavors with remarkable precision. This new graphene oxide “tongue” not only detects chemical compounds but also processes them, achieving nearly 99% accuracy in recognizing sour, salty, bitter, and sweet tastes. Developed by a team at the National Center for Nanoscience and Technology, this breakthrough could revolutionize how machines interact with the sensory world.

Transforming Taste Detection

While machines have excelled in mimicking sight and sound, digitizing the sense of taste has proven more complex. Previous artificial tongues focused on specific flavors, like sweetness or wine. In contrast, this new system represents a more generalist approach. The artificial gustatory system employs layered graphene oxide membranes, designed to mimic the interaction between biological taste buds and neurons.

One key challenge in taste detection lies in the need for functionality in liquid environments, since taste operates through ions rather than electrons. The researchers addressed this by creating a graphene oxide ionic sensory memristive device (GO-ISMD). Within this device, ions undergo interfacial adsorption and desorption, creating a memory-like electrical response. This unique response allows the system to detect chemicals while performing computation in wet environments, marking a first in the field.

In laboratory tests, the device demonstrated synapse-like behavior. It can adjust its response strength, exhibit memory effects, and even retain information from closely arriving signals. The memory duration varies, depending on the thickness of the membrane, lasting up to 140 seconds, exceeding expectations for simple ion movement.

Learning Flavors and Future Applications

The research team utilized reservoir computing to translate the electrical signals into recognizable digital patterns. “Inspired by the biological taste system, we developed a smart system using our devices to ‘recognize’ chemicals based on their flavors,” explained Yon Yang in a recent email. The system consists of three main components: a sensing input, a reservoir layer, and a single-layer fully connected neural network. This architecture enables the system to learn and recall different flavors, creating a functional memory.

In their proof-of-concept study, the researchers tested the device with four representative tastants: sour (acetic acid), salty (NaCl), bitter (MgSO4), and sweet (lead acetate). The neural network achieved approximately 98.5% accuracy in distinguishing these flavors, with test accuracies ranging between 75% and 90% depending on the sample. The device also performed well in classifying complex beverages such as coffee and Coke.

Despite these promising results, the researchers acknowledge that the current setup is still in the development stage. The device is bulky and requires significant energy to function. Future advancements will focus on miniaturizing the technology and integrating circuits to enhance practical applications outside laboratory settings.

“This technology perfectly bridges brain-inspired computing, chemical detection, and biologically inspired systems,” noted Yan, one of the lead researchers. The team anticipates that with ongoing enhancements in production scaling, power efficiency, and multi-sensor integration, the technology could lead to transformative applications in healthcare, robotics, and environmental monitoring within the next decade.

The study was published in the Proceedings of the National Academy of Sciences, representing a significant step forward in biomimetic gustation and neuromorphic engineering, potentially paving the way for tools that may extend or even reconstruct the sense of taste.