Cycle-to-Cycle Intrinsic RESET Statistics in ${\rm HfO}_{2}$-Based Unipolar RRAM Devices

Abstract

The statistics of the RESET voltage (V RESET ) and the RESET current (I RESET ) of Pt/HfO 2 /Pt resistive random access memory (RRAM) devices operated under unipolar mode are analyzed. The experimental results show that both the distributions of I RESET and V RESET are strongly influenced by the distribution of initial resistance in the ON state (R ON ), which is related to the size of the conductive filament (CF) before RESET. By screening the statistical data into different resistance ranges, both the distributions of I RESET and V RESET are shown to be compatible with a Weibull model. Contrary to previous reports for NiO-based RRAM, the Weibull slopes of the I RESET and V RESET are demonstrated to be independent of R ON . This is an indication that the RESET point, defined in this letter as the point of maximum current, corresponds to the initial phase of CF dissolution. On the other hand, given that the scale factor of the V RESET distribution (V RESET63% ) is roughly independent of R ON , the scale factor of the I RESET (I RESET63% ) is inversely proportional to R ON . This is analogous to what was found in NiObased RRAM and it is consistent with the thermal dissolution model of RESET. Our results highlight the intrinsic link between the SET and RESET statistics and the need for controlling the variation of ON-state resistance to reduce the variability of the RESET voltage and current. Index Terms-RESET statistics, resistive random access memory (RRAM), resistive switching (RS).

Voltage and Power-Controlled Regimes in the Progressive Unipolar RESET Transition of HfO2-Based RRAM

“…RS is usually induced with a ramped or pulsed electrical stress of the same (unipolar switching) or opposite polarity (bipolar switching) and has been reported in a wide variety of materials, including binary oxides1011121314151617 and complex perovskite oxides1819, etc., among which HfO 2 is one of the most widely investigated and most competitive resistive switching functional materials. To successfully push RRAM into applications, it is imperative to solve some key issues, such as revealing the physics underlying the RS phenomena20212223242526 with special emphasis on the related statistics2728293031 and the effective control of the statistical variation of the switching parameters32. For the filament-type RRAM devices, the mechanism of the SET transition from HRS to LRS or the conductive filament (CF) formation is comparatively clear and is interpreted as a defect-induced soft-breakdown, involving the oxidation/reduction (redox) processes of metal cations or oxygen anions in the interfaces and bulk of the RS layer and their migration through this layer22031.…”

“…The switching occurs by the destruction and reformation of a CF due to redox processes involving the generation and movement of oxygen vacancies133031. In a typical RESET cycle with a voltage ramp, an abrupt current jump event (RESET1) usually occurs just after reaching the maximum current, followed by a progressive RESET phase consisting of several successive smaller current drops between well-defined intermediate states with the conductance falling between those of the LRS and the HRS.…”

“…The transition from the voltage-controlled to the power-controlled regime depends on the CF conductance and the series resistance. To further understand these RESET processes, the evolution of the statistical distribution of conductance is reported as a function of the voltage, and all of the experimental observations are reproduced with a Monte Carlo simulator based on the thermal dissolution model2930333435.…”

An in-depth study of thermal effects in reset transitions in HfO2 based RRAMs

“…RRAMs fabricated with metal oxides show different resistance states produced by alternating the formation and rupture of one or several conductive filaments (CFs) [3,[5][6][7][8]: a low resistance state (LRS) is achieved if a CF is formed, and it can be switched to a high resistance state (HRS) by the partial dissolution of the CF (reset). The HRS can be turned back again to the LRS through a set process.…”

Dual bipolar resistive switching in the sub-forming regime of HfO2 resistive switching devices

“…Here the Ti top electrode acts effectively as an inert electrode, similar to the bottom Pt electrode, since no thermal treatment is performed, no significant layer of TiO 2Àx is formed and the stoichiometry of the HfO 2 is not altered drastically [24]. Additionally, other researchers have studied resistive switching in Pt/HfO 2 /Pt devices that show evidence for oxygen vacancies based filaments [25][26][27][28]. In support of this claim we note that devices with Pt as top electrode measured under identical conditions show similar DBRS behavior (not shown), however their yield is lower possibly due to poor TE-oxide interface [29].…”