Comparison of data retention measured by static laser testing and in PCRAM devices

Phase change materials are often evaluated for their thermal stability properties by measuring crystallization temperatures as function of heating rate or during isothermal annealing processes using, for example, resistivity versus temperature measurements, differential scanning calorimetry or time-resolved x-ray diffraction. These measurements are typically performed on as-deposited films. We have developed a method to measure data retention for amorphous, melt-quenched materials using a static laser tester combined with isothermal annealing in a furnace. The results show a much better data retention for as-deposited amorphous films compared to melt-quenched amorphous material. It has been known that as-deposited amorphous films can have much longer crystallization times compared to meltquenched amorphous material. This was attributed to the fact that in the latter case no nucleation is required because the melt-quenched material typically is surrounded by crystalline material. In addition, the melt-quenched material might contain tiny crystallites that will accelerate crystallization. We find that the data retention results on asdeposited, amorphous films greatly overestimate the performance of the same phase change material in actual phase change random access memory (PCRAM) devices. Laser testing on melt-quenched material leads to a better prediction of the performance of a given phase change material in PCRAM devices.