Self-compliance and high-performance GeTe-based CBRAM with Cu electrode
Introduction
As an indispensable part of the information age, non-volatile memory occupied a pivotal position in global semiconductor market due to its wide application in the fields of big data processing, smart medical health, and personal intelligent wear. However, traditional flash memory was facing various challenges like physical limits with thickness reduction of the oxide layer. To overcome its defect, new nonvolatile memory technologies were developed, such as phase change random access memory (PCRAM), ferroelectric random access memory (FeRAM), magnetic random access memory (MRAM), and resistive random access memory (RRAM). As a new type of nonvolatile memory device, RRAM became a competitive alternative on account of its advantages of structure, forming voltage, scalability, integration, stability and power consumption [[1], [2], [3], [4], [5]]. Therefore, it was widely considered as the neuromorphic computing and future memories [6].
RRAM had a simple “MIM” structure, where “I” was a functional layer made of insulating or resistive material, and “M” was the electrode at both interfaces of the functional layer. It was a simple structure to store “0” and “1” through two different resistance states. But large switching voltage and current were still the key issues for realizing power efficiency and high-density applications in RRAM arrays [[7], [8], [9]]. The physical mechanisms of forming process had been studied for many years [10,11]. The electroforming process and polarity resulted in the properties of defects and resistive switching characteristics [[12], [13], [14]]. Besides, it was reported that the forming voltage of RRAM increased when the size of device was reduced, which was quite pivotal with the devices scaling down [[15], [16], [17]]. Many materials had been utilized as either insulator or electrode in RRAM to lower the forming voltage and optimize the RS characteristics. However, for GeTe-based RRAM, uncontrollable current and large voltage of low resistance state (LRS) were barriers for scalability and high-density integration.
In this work, the prepared Cu/GeTe/TiN RRAM exhibited low forming voltage, which was able to be interpreted as the movement of Cu ions in GeTe electrolyte. Besides, Cu ions were able to migrate to the functional layer to form more stable CuTe which had a relatively larger resistance, inducing the self-compliance phenomenon. Finally, the mechanisms of the device were further studied and a physical model was proposed to figure out the RS characteristics.
Section snippets
Experimental
The structure of Cu/GeTe/TiN device was formed in a via hole on the substrate of TiN/Ti/SiO2/Si for the strict control of device size and the uniformity of electrical properties [18]. To acquire the TiN/Ti/SiO2/Si substrate, firstly, photolithography was adopted to mark the position of the bottom electrode on the 300-nm-thick SiO2 oxide layer, and then the 200-nm-thick TiN film was deposited by atomic layer deposition (ALD). Subsequently, the prepared TiN substrate was dephotoresisted, and the
Results and discussion
As shown in Fig. 1(a), a typical metal-insulator-metal sandwich structure of Cu/GeTe/TiN device was prepared. After a series of mask aligner fabrication processes, the TiN bottom electrode and the via hole with side walls of SiO2 on the TiN/Ti/SiO2/Si substrate were patterned. Then, GeTe thin film and Cu top electrode were deposited sequentially. To confirm the structure of the device, a cross-sectional view of SEM image was performed in Fig. 1(b), where a via hole structure was found clearly
Conclusions
In summary, the device based on GeTe electrolyte exhibited outstanding performance of low forming voltage, fast switching speed and self-compliance properties, demonstrating that the device had impressive scalability and great potential in high-density integration. According to the experiment results, the physical model was established to understand the mechanism. The device acquired a low forming voltage, low switching voltage and fast switching speed because copper ions migrated easily in the
Author statement
Zhao Jiayi: Investigation, Writing – review & editing. Chen Qin: Formal analysis, Writing – original draft, Investigation. Zhao Xiaohu: Formal analysis. Yang Gaoqi: Investigation, Writing – review & editing. Ma Guokun: Project administration, Supervision, Funding acquisition. Wang Hao: Supervision
Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgements
Supported by the National Natural Science Foundation of China (NSFC) under Grant 61904050, Science and Technology Major Project of Hubei under Grant 2020AAA005 and Grant 2020AEA017 and Hubei Key Laboratory of Advanced Memories.
References (49)
- et al.
A remarkable performance of CH3NH3PbI3 perovskite memory based on passivated method
Org. Electron.
(2018) - et al.
Revealing conducting filament evolution in low power and high reliability Fe3O4/Ta2O5 bilayer RRAM
Nano Energy
(2018) - et al.
Good endurance and memory window for Ti/HfOx pillar RRAM at 50-nm scale by optimal encapsulation layer
IEEE Electron. Device Lett.
(2011) - et al.
A novel high-performance and energy-efficient RRAM device with multi-functional conducting nanofilaments
Nano Energy
(2021) - et al.
Bipolar switching properties and electrical conduction mechanism of manganese oxide RRAM devices
Ceram. Int.
(2017) - et al.
Resistive random access memory (RRAM): an overview of materials, switching mechanism, performance, multilevel cell (MLC) storage, modeling, and applications
Nanoscale Res. Lett.
(2020) - et al.
Advances of RRAM devices: resistive switching mechanisms, materials and bionic synaptic application
Nanomaterials
(2020) - et al.
Multilevel HfO2-based RRAM devices for low-power neuromorphic networks
Apl. Mater.
(2019) - et al.
Self-controlled multilevel writing architecture for fast training in neuromorphic RRAM applications
Nanotechnology
(2018) - et al.
A robust memristor based on epitaxial vertically aligned nanostructured BaTiO3−CeO2 films on silicon
Adv. Mater.
(2022)