
The future of computing could be shaped by a new kind of memory called resistive random access memory, or ReRAM.
Unlike traditional memory chips, ReRAM is based on oxide materials and promises faster speeds, better data storage, and simpler designs.
It could even play a central role in neuromorphic computing, where devices are designed to work more like the human brain.
But until now, scientists didn’t fully understand how ReRAM actually works at the microscopic level.
A team of researchers at KAIST in South Korea has now provided the clearest explanation yet. Led by Professor Seungbum Hong, in collaboration with Professor Sang-Hee Ko Park, the team has revealed exactly how oxygen defects—tiny “missing atoms” inside oxide materials—control the way ReRAM stores and erases information.
Their findings were published in ACS Applied Materials & Interfaces.
To uncover this mystery, the researchers used a powerful tool called a multi-modal scanning probe microscope (Multi-modal SPM).
This device combines several imaging techniques, allowing them to watch, in real time, how electricity moves through an oxide thin film.
They could see the channels where electrons flowed, track the movement of oxygen ions, and measure how the surface charge changed as the material switched between different states.
The team focused on a thin layer of titanium dioxide (TiO₂), a common oxide material. By applying electrical signals, they simulated the writing and erasing of digital information inside the memory. What they discovered was that oxygen defects play a critical role.
When there are many oxygen defects clustered together, they form wide pathways that let current flow easily—this represents the “on” state of memory. But when the defects scatter or shift position, the pathways close off, blocking the current and turning the memory “off.”
This process explains how ReRAM can store binary information, the ones and zeros at the core of all digital devices.
But the team’s discovery went further. They found that the stability of the “off” state depends on injecting oxygen ions during the reset process. This keeps the material in a high-resistance state for a long time, which is essential for reliable, long-term data storage.
Importantly, their work showed that oxygen defects don’t act alone. The movement of electrons is tightly linked with how oxygen ions shift inside the material. This interaction between ions and electrons ultimately determines how well ReRAM performs.
Professor Hong explained that this was the first time researchers could directly observe the spatial relationship between oxygen defects, ions, and electrons in such detail. He believes the technique could be applied beyond memory, helping to design future metal oxide devices for semiconductors.
This discovery not only clarifies how ReRAM works but also provides a roadmap for building faster, more reliable, and energy-efficient memory devices.
With the growing need for high-performance computing in areas like artificial intelligence and renewable energy, the ability to harness oxygen defects may help unlock a new generation of technology.