Graphene "Tattoos" for Plants Could Form Neural Networks

Overview

Researchers at the University of Texas at Austin developed a graphene 'tattoo' that sticks directly onto a plant leaf to provide real-time moisture readings. Beyond sensing hydration, the team believes the patches could one day act as artificial synapses and be linked into a neural network that computes on the plants themselves. The work, led by associate professor Jean Anne Incorvia and colleagues, was published in Nano Letters in February.

Key Takeaways

The graphene 'tattoo' sticks onto a leaf and gives real-time moisture readings without cutting the leaf from the plant.
The sensor is functionally a three-terminal transistor using a graphene channel, gold electrodes, and the leaf as its dielectric insulator.
It senses hydration by sending an electric pulse into the leaf, moving ions and changing the graphene's conductance.
The sensor can act like a synapse, with pulses adjusting its conductance and a slow return to baseline acting as short-term memory.
Graphene is nearly transparent so it does not block light, and it can stretch as the leaf grows or twists.
Researchers envision networked sensors across a grove of plants to gauge fire or drought risk in real time.

Stats & Key Facts

#Published in Nano Letters in February
#Sensor returns to original conductance over about 90 seconds

Graphene "Tattoos" for Plants Could Form Neural Networks

The problem with monitoring leaf hydration

Traditional methods are slow and destructive.

›A hydrated leaf is a healthy leaf for both crops and trees in fire-prone areas.
›Traditional techniques require cutting leaves from plants.
›Cutting is time-consuming and cannot give live measurements.

Because traditional methods cannot provide live readings, many researchers are building sensors that measure a plant's health in real time.

How the graphene tattoo works

›A graphene patch is pasted onto a leaf like a stick-on tattoo.
›It is functionally a three-terminal transistor with a graphene channel and gold strip electrodes.
›The leaf itself serves as the dielectric insulator.

The sensor gauges hydration by sending an electric pulse into the leaf, which moves ions within the leaf and changes the graphene's conductance. Because the magnitude of those conductance changes depends on moisture, the researchers read out hydration without external processing. The team used Monstera plants in their tests.

Why graphene

›Graphene is nearly transparent, so it will not block light and disrupt photosynthesis.
›It can stretch and squeeze as the leaf grows, shrinks or twists.
›Real-time hydration sensors are not common in the field.

The researchers hope this sensor can make real-time hydration monitoring more common, partly because it can also fit into a neural network.

Acting like a synapse

›Electric pulses can slightly adjust the sensor's conductance up or down, like strengthening or weakening a synapse.
›After a pulse ends, the sensor returns to its original conductance slowly, over about 90 seconds.
›During that time it can act as a sort of short-term memory.

The researchers imagine using these artificial synaptic qualities to tune and store a neural network's weights.

Neuromorphic plant computing

›Incorvia's group has designed non-leaf-based neuromorphic devices for several years.
›They typically used graphene with Nafion, a polymer that conducts protons well.
›A leaf can conduct protons too, prompting a part-leaf device.

Undergraduate Maya Borowicz suggested swapping out the Nafion to make a part-leaf device. Incorvia said they tried it a few years ago and it worked, but they tabled the idea until a later encounter with geologist Ashley Matheny, who was working on better ways to monitor moisture.

A possible future forest

›A future forest might hold a grove of networked sensors.
›The network could gauge fire or drought risk in real time.
›Each sensor could double as an artificial synapse in a larger network.

Incorvia and colleagues think networked leaf sensors could turn plants into both a monitoring system and a computing substrate.

Frequently Asked Questions

What is the graphene 'tattoo' for plants?

It is a graphene patch pasted onto a leaf like a stick-on tattoo that provides real-time moisture readings.

How does the sensor measure hydration?

It sends an electric pulse into the leaf, moving ions and changing the graphene's conductance, with the magnitude of the change depending on moisture.

How can the sensor act like a synapse?

Electric pulses can adjust its conductance up or down, and after a pulse ends it returns to its original conductance slowly over about 90 seconds, acting as short-term memory.

Why is graphene a good material for this?

It is nearly transparent so it does not block light and disrupt photosynthesis, and it can stretch and squeeze as the leaf grows, shrinks or twists.

Who did the research and where was it published?

The work was led by University of Texas at Austin associate professor Jean Anne Incorvia and colleagues, published in Nano Letters in February.

The team sees networked graphene leaf sensors as a path to real-time plant monitoring and computing directly on plants.