A startling discovery at the University of Limerick in Ireland has revealed for the first time that unconventional computing that resembles a brain at the smallest scale of atoms and molecules is possible.
Researchers at the Bernal Institute at the University of Limerick have worked with an international team of scientists to create a new type of organic matter that learns from its past behaviour.
The discovery of a “dynamic molecular switch” that mimics synaptic behavior is reported in a new study published in the International Journal of Neuroscience nature materials.
The study was led by Damien Thompson, Professor of Molecular Modeling in the Department of Physics at UL and Director of SSPC, the Ireland Research Center for Science hosted by UL, along with Christian Nigues in the Center for Brain-Inspired Molecular and Nanosystems at the University of Twente and Enrique Del Barco of the University of Central Florida.
Working during the closures, the team developed a layer of molecules 2 nanometers thick, which is 50,000 times thinner than a strand of hair and remembers its history as electrons pass through it.
“The switching probability and values of the on/off states are constantly changing in the molecular material, providing a disruptive new alternative to traditional silicon-based digital switches that cannot be turned on or off,” explained Professor Thompson.
The newly discovered dynamic organic switch exhibits all the mathematical logic functions necessary for deep learning, and successfully simulates Pavlovian “call and response” interlocking behaviour.
The researchers demonstrated the properties of the new materials using comprehensive experimental characterization and electrical measurements supported by multiscale modeling that extends from predictive modeling of molecular structures at the quantum level to analytical mathematical modeling of electrical data.
To simulate the dynamic behavior of synapses at the molecular level, the researchers combined rapid electron transfer (similar to action potentials and rapid depolarization processes in biology) with slow, diffusion-limited proton coupling (similar to the role of biological calcium ions or neurotransmitters).
Because the electron transfer and conjugation steps of protons within the material occur on vastly different time scales, the transduction can simulate the plastic behavior of synaptic connections, Pavlovian learning, and all logic gates of digital circuits, simply by changing the applied voltage and the duration of the voltage pulses during synthesis, they explained.
“This was a great closing project, as Chris, Enrique, and I pushed each other through zoom meetings and massive email threads to bring our teams’ combined skills in materials modeling, synthesis, and characterization to the point where we could demonstrate these brain-like computing properties,” explained Professor Thompson. .
“Society has long known that silicon technology works very differently from how our brains work, and so we’ve used new types of electronic materials based on soft particles to simulate brain-like computing networks.”
The researchers show that the method can be applied in the future to dynamic molecular systems that are driven by other stimuli such as light and coupled with different types of dynamic covalent bond formation.
This breakthrough opens up a whole new range of adaptive and reconfigurable systems, creating new opportunities in sustainable and green chemistry, from producing more efficient flow chemistry for pharmaceutical products and other value-added chemicals to developing new organic materials for high-density computing, memory and storage in massive data centers. .
“This is just the beginning. We are already busy scaling this next generation of smart molecular materials, which enable the development of sustainable alternative technologies to address major challenges in energy, environment and health,” explained Professor Thompson.
Professor Norelli Kennedy, Vice President of Research at UL, said: “Our researchers are constantly searching for new ways to make more efficient and sustainable materials. This latest discovery is very exciting, demonstrating the reach and ambition of our international collaboration and showcasing our world – leading at UL to encode beneficial properties into materials membership.”