An international team of researchers led by Alberto Salleo at Stanford University has discovered a way to create an artificial synapses using layers of polymer (plastic) separated by a thin film of electrolyte. A living synapses has complex electrical characteristics: depending on its size and geometry at the contact point it can modulate the transmission of electricity. This changes over time as result of the communications taking place (in a way a synapses "remembers" what signals it sent before, it "learns") and therefore a singly synapses will send a variety of signals during its operation.
Mimicking this behaviour in an artificial synapses is therefore quite complex.
The team of researchers created a polymer structure that can change its conductivity in 500 different ways, hence it can send 500 different signals, within a 1 V (volt) range.
This is an amazing result considering that in general there is a tradeoff between the voltage required to change the state in a material and the stability of the material (you would like to have a very stable material that keeps its state and only changes it when you want it to change: this in general requires using a significant switching voltage to make sure that unwanted switching due to environmental noise occurs).
The polymer base synapses has the desired stability and yet it can switch its state with very little energy, in the order of a few aJ (1 billionth billionth of a Joule), that is using far less energy than a natural synapses.
The researchers hope that this result will give a boost to the creation of neuromorphic structures, computers that mimic the brain circuitry.