Ultrathin ferroelectric capacitors; memristor ADC; molecular electronics.
Researchers from the Institute of Science Tokyo and Canon ANELVA Corporation built an ultrathin ferroelectric memory capacitor stack using scandium-substituted aluminum nitride ((Al,Sc)N) thin films with platinum electrodes. The total thickness is just 30nm: a 20nm ferroelectric layer sandwiched between 5nm platinum top and bottom electrodes.
“Previous research on downscaling ferroelectric memory has only focused on thinning the ferroelectric layers. What makes our research stand out is that we focused on downscaling the complete device stack, not just the ferroelectric film,” said Hiroshi Funakubo, a professor from the School of Materials and Chemical Technology, Institute of Science Tokyo, in a press release. “The results demonstrated that high ferroelectric performance can be sustained even when the entire capacitor stack thickness is drastically reduced, and this brings us one step closer to the practical implementation of ultrathin memory devices.”
Post-heat treatment of the bottom platinum electrode at 840 °C enhanced its crystal orientation and improved the polarization switching in thinner films. While the team says the capacitor is suitable for direct embedding in semiconductors and logic systems, they plan to explore alternative electrode materials with more suitable crystal orientations, which may reduce the thermal processing requirement and improve durability. [1]
Researchers from the University of Hong Kong, Xidian University, and Hong Kong University of Science and Technology created a memristor-based analog-to-digital converter (ADC) that is capable of automatically adjusting its settings based on the data it receives to fine-tune how signals are converted.
The design uses analog content-addressable memory cells with programmable overlapped boundaries to establish optimized quantization thresholds, an approach the researchers said provided a 15x improvement in energy efficiency alongside almost 13x circuit area reduction compared with state-of-the-art solutions. Additionally, the ADC maintained high accuracy when running neural network tasks across various types of AI models. When integrated into compute-in-memory systems, further reductions in overall energy consumption and chip size were demonstrated. [2]
Researchers from the Indian Institute of Science and University of Limerick created a ruthenium-based molecular device that can behave as a memory unit, a logic gate, a selector, an analog processor, or an electronic synapse, depending on how it is stimulated.
Key to creating the device was developing a framework capable of predicting function from molecular structure. The team synthesized 17 metal-organic ruthenium complexes by making small changes to the arrangement of ligands and ions. This enabled them to map how electrons traverse the molecular film, how individual molecules undergo oxidation and reduction, and how counterions rearrange within the molecular matrix, which combine to govern the switching and relaxation dynamics and the stability of each molecular state.
“What surprised me was how much versatility was hidden in the same system,” said Pallavi Gaur, first author and PhD student at IISc’s Centre for Nano Science and Engineering, in a press release. “With the right molecular chemistry and environment, a single device can store information, compute with it, or even learn and unlearn. That’s not something you expect from solid-state electronics.”
The researchers aim to integrate such adaptable materials into silicon chips to create neuromorphic hardware in which learning can be encoded into the material itself. [3]
[1] S. Doko, N. Matsui, T. Irisawa, et al. “Thickness Scaling of Integrated Pt/(Al0.9Sc0.1)N/Pt Capacitor Stacks to 30 nm.” Adv. Electron. Mater. (2025): e00451. https://doi.org/10.1002/aelm.202500451
[2] H. Hong, Z. Du, M. Jiang, et al. Memristor-based adaptive analog-to-digital conversion for efficient and accurate compute-in-memory. Nat Commun 16, 9749 (2025). https://doi.org/10.1038/s41467-025-65233-w
[3] P. Gaur, B. Kundu, P. Ghosh, et al. “Molecularly Engineered Memristors for Reconfigurable Neuromorphic Functionalities.” Adv. Mater. (2025): e09143. https://doi.org/10.1002/adma.202509143
 |
 |
 |
 |
 |  |
 |  |
 |
Leave a Reply