Highlight Publication - New Physics Model "Non-pinched hysteresis in CrOx/TiOy-based memristive devices: Modeling and analysis"

Congratulations to our colleague, PhD Student. Msc. Pham Phu Quan, for his recent publication entitled "Non-pinched hysteresis in CrOx/TiOy-based memristive devices: Modeling and analysis" in the journal "Applied Physics Letters" (Q1, h-index = 486).

Appl. Phys. Lett. 128, 153502 (2026), DOI: https://doi.org/10.1063/5.0332014 

This research paper presents a comprehensive study on the fabrication, characterization, and simulation of transition-metal oxide memristors designed for neuromorphic computing. Published in Applied Physics Letters, the study addresses a critical gap in existing SPICE models, which typically fail to reproduce "non-pinched" hysteresis—a behavior in which the I-V characteristic curve does not intersect at (0,0).

The key aspects of the research are described below

• Improved Modeling: The authors developed an enhanced version of the Yakopcic model by incorporating new parameters (u1, u2, k1, k2) to account for internal electromotive forces and surface-trapped charges. This allows the model to accurately capture the non-zero crossing points observed in experimental I-V curves
• Synaptic Emulation: The device successfully emulates biological synaptic functions, including Long-Term Potentiation (LTP) and Long-Term Depression (LTD). The response is highly dependent on pulse width and amplitude, with the best agreement between simulation and experiment at pulse widths around 10 ms.
• Scalability: To demonstrate practical application, the model was extended to a 4x4 crossbar array. Experimental results from the array showed high uniformity and comparable performance to individual cells, highlighting the potential for large-scale integration in artificial neural networks.

While the model achieves high fidelity in reproducing I-V characteristics, the authors note that further improvements are needed to better account for pulse-driven Spike-Number-Dependent Plasticity (SNDP) switching kinetics. Overall, the work provides a vital framework for bridging experimental device physics with circuit-level neuromorphic design.