Global Sources
EE Times-Asia
Stay in touch with EE Times Asia
EE Times-Asia > Memory/Storage

IBM claims PCM non-volatility not necessary

Posted: 19 Dec 2014 ?? ?Print Version ?Bookmark and Share

Keywords:PCM? non-volatility? IEDM? SCM? DRAM?

Moving Ge atoms in TRAMs?

If there is to be a future for memory based on GeSbTe compositions as an SCM or as an incumbent memory replacement, then the topological-switching random access memory (TRAM) may have the answer with its GexT1-x/Sb2Te3 super-lattice structure. The TRAM is a derivative of the super-lattice based interface phase change memory (IPCM) and might be considered as an order-order transition-based device rather than the order-disorder of the conventional PCM.

Another IEDM2014 paper [Ref 3] from a team made up of members from the Low Power Electronics Association Project (LEAP), University Groups, and The National Institution of Advanced Industrial Science and Technology in Japan reports the progress they have made. They are demonstrating a TRAM with equal 55uA reset and reset currents and without any evidence of melting of the active material. The required RESET and SET pulse widths were 75ns and 120ns, respectively, in a manner similar to that of a conventional PCM.

While Sb2Te3 is known as one of a number of materials that can display topological insulator (TI) properties, it is a material with a conducting surface layer and an insulating body. The use of this effect for memory requires depositing films that are thin enough to represent just the conducting surface layer and then electrically modulate the TI effect.

In the year since IEDM 2013, the emphasis in the description of this super-lattice device has moved from a charge-injection PCM to topological-switching RAM.

One example of a TRAM structure is illustrated in Figure 4(a). Fabricated by the team in Japan, it consists of eight alternate layers with 1nm of GeTe, and 4nm of Sb2Te3. To stabilise the film during the film deposition using a multi-cathode PVD fabrication process, the layered structure is deposited on an initial layer of Sb2Te3.

It would appear the operation of the TRAM, supported by some analysis and experimental evidence, depends of moving Ge atoms in and out of the Te-Te interface region of the 1nm thick GeTe layer where they either allow or interrupt the flow of current. This is illustrated in Figure 4(b), a simplified two-dimensional view, in effect the blocking if the Te-Te conduction channel isolates the Sb2Te3 blocks.

Figure 4

The super-lattice in what is an artificial material and the switching effect is best observed when the thickness of Sb2Te3 blocks is kept to a minimum, less than 6nm.

It is claimed that the short-range movement of the Ge atoms is not as might be expected by Joule heating and/or electro-migration but by movement related to the density of injected charge, holes, or electrons. The justification for the injected charge conclusion was the lack of any difference in switching characteristics when the device was fabricated on substrates of different thermal conductivity. My calculation of current density of 2.8x106 A/sq-cm at a write current of 55uA would also mitigate against electro-migration, providing, that is, the switching operation does not occur in a narrow filament region.

?First Page?Previous Page 1???2???3???4?Next Page?Last Page

Article Comments - IBM claims PCM non-volatility not ne...
*? You can enter [0] more charecters.
*Verify code:


Visit Asia Webinars to learn about the latest in technology and get practical design tips.

Back to Top