Phase-Change Materials

The idea of using structure for storage dates back to the 1960’s with the seminal work of Stanford Ovshinsky who proposed a four component chalcogenide alloy as a form of nonvolatile electrical memory. The higher power requirements of the phase-change memory (PCM) limited its application until the 1980’s when the pioneering work at Matsushita lead to the development of Ge-Sb-Te PCM alloys and its application to re-writeable optical discs. This has now lead to optical disc storage capacities to exceed 50 GB. Moreover, the same Ge-Sb-Te alloys are now employed in the form of storage-class memory (as epitomized by Intel’s Optane memory subsystems), a form of nonvolatile electrical memory with latencies and write times within a factor of 10 of DRAM alleviating the bottlenecks associated with the larger latency of the denser nonvolatile FLASH memory. Even more recently a new variation of electrical memory called interfacial phase-change memory (iPCM) has appeared which uses highly oriented superlattices of GeTe and Sb2Te3 to achieve more than a factor of ten reduction in energy requirements by use of what is thought to be a crystalline-crystalline transition. The figure above represents a possible atomic level model for the switching process in iPCM devices.