The concept of memory, whether in computers or human brains, is a fundamental operation that allows for the storage and accessibility of information. However, a key disparity lies in how this information is processed. While human brains perform computations directly on stored data, computers rely on shuttling data back and forth between a memory unit and a central processing unit (CPU). This separation, known as the von Neumann bottleneck, is highly inefficient and contributes to the growing energy cost of computers.
In recent years, researchers have been exploring the potential of memristors (memory resistors) as electronic components that can not only store data but also compute, similar to a synapse in the brain. However, Aleksandra Radenovic and her team at the Laboratory of Nanoscale Biology (LBEN) in EPFL’s School of Engineering have taken a step further by developing a nanofluidic memristive device that operates using ions instead of electrons and holes. This new approach aims to mimic the brain’s energy-efficient way of processing information.
The research conducted by LBEN has led to the creation of a nanofluidic device for memory applications that is more scalable and performant than previous iterations. The device, which has been recently published in Nature Electronics, can switch between two conductance states by manipulating the applied voltage. Unlike traditional electronic memristors that rely on electrons and holes, LBEN’s memristor can leverage a variety of ions for information processing.
The device fabricated by the researchers at EPFL’s Center of MicroNanoTechnology features a nanopore at the center of a silicon nitride membrane. By adding layers of palladium and graphite, nano-channels for ions are created. As a current flows through the chip, ions flow through the channels and converge at the pore, generating a blister that alters the device’s conductivity and memory state. This novel approach closely resembles the structural changes that occur in ion channels within the brain synapse.
The team’s achievement in connecting two artificial synapses using ion flow represents a significant milestone in the development of brain-inspired liquid hardware. Through collaboration with experts in nanoscale electronics and structures, the researchers have successfully demonstrated digital logic operations based on synapse-like ionic devices. Their next endeavor involves connecting a network of highly asymmetric channels (HACs) with water channels to create fully liquid circuits.
The utilization of water in nanofluidic devices not only provides an inherent cooling mechanism but also opens up possibilities for bio-compatible devices with potential applications in brain-computer interfaces and neuromedicine. By combining the principles of ion-based memory processing with innovative fabrication techniques, researchers are paving the way for a new era of memory technology that closely mimics biological processes.