Characteristics of hafnium oxide resistance random access memory with different setting compliance current

Su, Yi-Sheng; Chang, Kuan‐Chang; Chang, Ting‐Chang; Tsai, Tsung‐Ming; Zhang, Rui; Lou, Jen‐Chung; Chen, Jung‐Hui; Young, Tai‐Fa; Chen, Kai-Huang; Tseng, Bae–Heng; Shih, Chih‐Cheng; Yang, Ya‐Liang; Chen, Min-Chen; Chu, Tian-Jian; Pan, Chunxu; Syu, Yong-En; Sze, Simon M.

doi:10.1063/1.4825104

Cited by 47 publications

(21 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Consequently, stronger conductive filament is formed and the LRS resistance is reduced. Similar mechanism understanding has been reported in literature discussing about the compliance current effect [10,11]. The remarkable dependence of the LRS resistance upon the current endpoint suggests a reliable approach to multilevel programming.…”

Section: Resultssupporting

confidence: 63%

Self-compliance SET switching and multilevel TaO<inf>x</inf> resistive memory by current-sweep operation

Jiang

Zhuo

et al. 2014

2014 14th Annual Non-Volatile Memory Technology Symposium (NVMTS)

View full text Add to dashboard Cite

Section: Resultssupporting

confidence: 63%

Self-compliance SET switching and multilevel TaO<inf>x</inf> resistive memory by current-sweep operation

Jiang

Zhuo

et al. 2014

2014 14th Annual Non-Volatile Memory Technology Symposium (NVMTS)

View full text Add to dashboard Cite

mentioning

confidence: 99%

“…4,5 The resistance switching behavior due to formation and rupture of conductive paths within HfO 2 deposited by atomic layer deposition was exploited for random access memory fabrication. 6 Since these are only a few examples of most recent applications of HfO 2 , still the predominant effort is put into using it as high-k material for gate oxides for replacing SiO 2 gates in MOS-FET structures. 7 For a typical operation voltage of 1-1.5 V, the leakage current through HfO 2 dielectric films was found to be several orders of magnitude lower than that of SiO 2 for the same equivalent oxide thickness of 0.9-2 nm.…”

mentioning

confidence: 99%

Anodization Behavior of Glassy Metallic Hafnium Thin Films

Mardare

Siket

Gavrilović-Wohlmuther

et al. 2015

J. Electrochem. Soc.

View full text Add to dashboard Cite

Glassy metallic Hf thin films were obtained using electron beam deposition at room temperature due to the low energy received by Hf atoms during the film formation process. The amorphous nature of the Hf films suggested by XRD was confirmed by low temperature electrical conductivity measurements where a negative temperature coefficient of resistivity was identified. Anodic oxidations using a scanning droplet cell microscopy were performed on a typical crystalline (hexagonal) Hf film obtained by sputtering and on the glassy Hf. A decrease of the oxide formation factor by 30% (from 2.4 to 1.7 nm V −1 ) was evidenced on the amorphous Hf. Electrochemical impedance spectroscopy on both samples revealed almost identical capacitances while the electrical permittivities were found to differ by 40%. The dielectric constant of the HfO 2 decreased from 33.5 on the crystalline parent metal sample to 19.8 on the amorphous one. The unexpected 10% deviation was attributed to a smoother defect-free oxide/metal interface in the case of Hafnium dioxide has a high dielectric constant, a relatively large bandgap, larger heat of formation as compared to SiO 2 , good chemical and thermal stability on Si and large barrier heights at interfaces with Si. Its modern applications are diverse, most recent ones focused on electronic devices. The chosen methods for obtaining hafnia thin films are also varied, most common one being reactive sputtering.1 Doping of sputtered hafnia with elements such as Ce has led to an increase in the dielectric constant and a decrease of the leakage currents, while Si doping revealed intrinsic ferroelectricity due to a non-centrosymmetric orthorhombic phase with direct applications in ferroelectric memories.2,3 Cathodoluminescence due to oxygen vacancies was also recently found in ion beam sputtered HfO 2 and first steps toward fabrications of hafnia micro-light emitting devices were already taken. 4,5 The resistance switching behavior due to formation and rupture of conductive paths within HfO 2 deposited by atomic layer deposition was exploited for random access memory fabrication. 6 Since these are only a few examples of most recent applications of HfO 2 , still the predominant effort is put into using it as high-k material for gate oxides for replacing SiO 2 gates in MOS-FET structures.7 For a typical operation voltage of 1-1.5 V, the leakage current through HfO 2 dielectric films was found to be several orders of magnitude lower than that of SiO 2 for the same equivalent oxide thickness of 0.9-2 nm.8-10 However, clearly a weak point in using thermally grown or vapour phase deposited HfO 2 as gate dielectrics arises from its poor interfacing with semiconductors which has led to a fervent investigation of hafnium-based complex oxides in recent years. 7Nowadays it is common knowledge that the anodic oxides on valve metals (e.g. Al, Hf, Ta, Nb etc.) are some of the most compact oxides available due to their growth mechanism obeying a high field regime.11 However, their use for electronic devices is weakly ex...

show abstract

“…1 [21], while in the case of CBRAM it consists of reduced metal ions sourced from a thin sacrificial metal anode [22]. For both CBRAM and OXRAM devices, the SET-state resistance (LRS) values can be well defined by controlling the dimensions of the conductive filament [8], [21], or the conduction mechanism (specific to some OXRAM devices) [23], which depends on the amount of current flowing through the active layer. Current flowing through the active layer is controlled either by externally imposed current compliance or by using an optional selector device that is used to drive the RRAM device (i.e.…”

Section: A Basic Principle and Workingmentioning

confidence: 99%

Exploiting Intrinsic Variability of Filamentary Resistive Memory for Extreme Learning Machine Architectures

Parmar

2015

IEEE Trans. Nanotechnology

View full text Add to dashboard Cite

In this paper, we show for the first time how unavoidable device variability of emerging non-volatile resistive memory devices can be exploited to design efficient low-power, low-footprint Extreme Learning Machine (ELM) architectures. In particular we utilize the uncontrollable Off-state resistance (Roff/HRS) spreads, of nanoscale filamentary-resistive memory devices, to realize random input weights and random hidden neuron biases; a characteristic requirement of ELM. We propose a novel RRAM-ELM architecture. To validate our approach, experimental data from different filamentary-resistive switching devices (CBRAM, OXRAM) is used for full network simulations. Learning capability of our RRAM-ELM architecture is illustrated with the help of two real world applications-(i) diabetes diagnosis test (classification) and (ii) SinC curve fitting (regression).

show abstract

Characteristics of hafnium oxide resistance random access memory with different setting compliance current

Cited by 47 publications

References 24 publications

Self-compliance SET switching and multilevel TaO<inf>x</inf> resistive memory by current-sweep operation

Self-compliance SET switching and multilevel TaO<inf>x</inf> resistive memory by current-sweep operation

Anodization Behavior of Glassy Metallic Hafnium Thin Films

Exploiting Intrinsic Variability of Filamentary Resistive Memory for Extreme Learning Machine Architectures

Contact Info

Product

Resources

About