Material Modeling of Resistive Switching for Non-Volatile Memories
Using ATLAS C-Interpreter/Giga


1. Introduction

Recently, a variety of materials having large non-volatile resistance change have been studied as potential candidates for next generation non-volatile memory devices. For example, chalcogenides for PCM (Phase-Change Memory) [1] and perovskite oxides or transition metal oxides for RRAM (Resistive Random Access Memory) [2] etc.

The basic operation of these devices is as follows: there are two states, RESET and SET. The RESET state is a high resistance state obtained by applying a sufficiently high electrical pulse to change crystal phase to amorphous phase for PCM, or to break the conduction path for RRAM. The SET state is a low resistance state obtained by applying a lower and longer pulse to change amorphous phase to crystal phase or to re-form conduction path.


2. A Simple Material Model for Resistive Switching Operation

The detailed mechanisms of the resistive switching especially for RRAM materials are still under investigation, so developing better models which can account for experimental I-V curves of these devices are useful for comprehending the operation and optimizing both the operation and structure of the device.

For that purpose, the C-Interpreter is very helpful. It enables the user to create their own models in order to investigate material and device behavior.

For example, assuming that the phase change or conduction path destruction/re-formation is dependent on the material’s temperature and its resistivity change can be expressed as a mobility change, a user-definable temperature dependent C-Interpreter mobility model can be used. The Giga module is used to account for self-treating effects.

Figure 1 is an example in which the mobility is described as a function of temperature, depending on the range and the direction of the temperature and mobility change.

Figure 1. A description of a user defined mobility model as a function of temperature using C-Interpreter.


3. Simulation Results

The device structure simulated is very simple as shown in Figure 2. A resistive switching material with the user defined mobility model is sandwiched between two electrodes. A bipolar triangular voltage sweep of 200ns shown is applied as shown at the top of Figure 3.

Figure 2. Mobility change at initial, 50ns, 150ns and 200ns.


Figure 3. (Top) Triangular voltage sweep applied. (Middle) A temperature change at a point near the center of the device. (Bottom) Mobility change at that point.


Figure 2 shows the mobility change in an applied voltage cycle. The top left picture is the initial SET state of high mobility. The top right picture shows the mobility and temperature contours at 50ns, the mobility decrease depends on the temperature distribution, and corresponds to a phase change from crystal to amorphous or to the breaking of conduction path. The bottom left picture at 150ns shows that the mobility increases again; the increase corresponds to re-crystallization or re-formation of the conduction path. The bottom-right is the state at 200ns in which mobility is kept at the SET state once reached at 150ns.

The mobility and temperature near the device center at X=0.6um Y=0.5um traced for one cycle are shown in Figure 3. The I-V hysteresis curve of the device is shown in Figure 4. A typical hysteresis curve can be obtained by the simple temperature dependent mobility functions defined in Figure 1.

Figure 4. Simulated I-V hysteresis loop for one cycle.


4. Conclusion

When conventional models are not applicable for new material devices, it is useful to develop a user defined model using the C-Interpreter. It provides users with a flexible method to comprehend the device behavior and to optimize its operation and structure.



  1. S.Lai, T. Lowrey, ”OUM-A 180nm nonvolatile memory cell element technology for standalone and embedded applications”, IEDM Tech., Dig., 2001, pp. 36.5.1 - 36.5.4
  2. Y.Hosoi, et al., “High Speed Unipolar Switching Resistance RAM(RRAM) Technology”, IEDM Tech., Dig., 2006, pp.30.7.1 - 30.7.4.


Download pdf version of this article