Nanoscale device inspired by the brain uses soft polymer material to process and store information
New device uses tiny mechanical movements to mimic how brain cells function, potentially leading to electronics that can perform multiple functions with less energy
Researchers funded by the U.S. National Science Foundation have developed tiny new devices that can simultaneously perform multiple computing functions, like processing and temporarily storing information. This technology uses the mechanical movements of a polymer material to remember past inputs and process information, similar to how neurons integrate and retain information in the brain.
Combining memory and processing in a single device could reduce the need for additional components, potentially leading to more compact and energy-efficient computing platforms. The research was supported by the NSF Center for the Chemistry of Molecularly Optimized Networks and the NSF National Nanotechnology Coordinated Infrastructure Network. The findings were published in Science Advances.
In mechanical computing, calculations are performed through physical transformations, such as movement and compression. However, achieving the reversible mechanical transformations needed for such computing is a fundamental challenge at the nanoscale. To address this, the researchers used the unique mechanical response of the polymer polydimethylsiloxane. They sandwiched a film of this polymer roughly 40,000 times thinner than a human hair between two tiny metal plates called electrodes. The polymer layer acts as a nanoscale spring, preventing the electrodes from permanently sticking together while enabling controlled, reversible movement.
This newly developed nanodevice mirrors how neurons in the brain communicate. Neurons accumulate electrical charge until a threshold is reached, triggering an electrical signal that rapidly transmits information to other neurons. Similarly, when voltage is applied to the device over time, the electrodes are drawn toward one another, gradually compressing the polymer layer and altering the electrical current flowing through it. Once a threshold is reached, the device produces a strong electrical response, mimicking the firing behavior of a neuron. The polymer then relaxes back to its original state, a process that takes time and gives the device a form of memory. As a result, the device's response to voltage depends on both past and current inputs. Using this behavior, researchers demonstrated an artificial neuron.
By incorporating temporary memory and computing capabilities into the same nanodevice, this brain-inspired technology could potentially enable smaller, more energy-efficient electronics. Future applications could include smart prosthetics that process touch signals, wearable patches that monitor health indicators in real time, and new environmental monitoring systems.
This article was originally published by the Massachusetts Institute of Technology. It has been edited for length and style.
Read the full story: Nanoscale mechanics could enable brain-inspired computing