Molybdenum Disulphide: Wonder Material for Memory Chips

With Two-Dimensional Materials, Sometimes Less is More

Scientists in the Department of Electrical and Computer Engineering at the University of Minnesota recently demonstrated that the unique properties of molybdenum disulphide (MoS2 ) make it an excellent candidate for use in memory chips.

Not too long ago the best place to find molybdenum disulphide was in hobby shops as a solid lubricant for model car wheels. Leaving its hobby shop days behind, today MoS2 has become one of the most widely studied semiconductor materials, due to the discovery that it can be made two-dimensional (2D). 2D MoS2 is atomically thin and displays a wide range of remarkable electronic and optical properties that have led numerous researchers to hail it as a replacement for silicon in modern computer chips. However, MoS2 has one problem: the electrons are too heavy. When given an electrical push, they  do not speed up enough to provide sufficient electrical current for most  electronic circuits.

Prof. Steven Koester and his research team have discovered that while the heavy electrons in MoS2 are a disadvantage for high-speed applications, they can actually provide enormous power savings in memory chips. In a paper titled “Dynamic Memory Cells Using MoS2 Field-Effect Transistors Demonstrating Femtoampere Leakage Currents” (published in ACS Nano), the team follow the operation of one and two transistor dynamic random access memory (DRAM) circuits fabricated using MoS2 transistors to illustrate the power savings enabled by this 2D material.

Prof. Chris Kim, ECE faculty and collaborator says, “One of the things most people don’t realize about the DRAM in their laptops is that it needs to constantly refresh in order to retain the stored information in memory. MoS2 transistors allow the DRAM to be refreshed about 100 times less often, which can save enormous amounts of power in certain systems.”

Low Leakage Leads to Higher Power Savings

Prof. Koester’s group demonstrated that the MoS2 devices have leakage on the order of 1 femtoampere, which is one millionth-billionth of an ampere (1 ampere is approximately the current that is drawn by a 100W light bulb). This value is roughly 100x lower than can be obtained in silicon devices of comparable size. In a DRAM, the lower leakage translates directly to lower power consumption because the memory does not have to be refreshed as often. The improved power savings in MoS2 transistors comes from the larger energy gap in MoS2 as well as the higher effective mass of electrons which allow the transistors to have a lower leakage current compared to standard silicon devices. The effective mass refers to how heavy electrons appear in a semiconductor, and materials with heavier effective mass can have much lower leakage currents.

Dr. Koester explains, “A transistor is an electronic switch, but unlike a light switch, when a transistor is turned off, it still “leaks” a little bit of electrical current. MoS2 transistors leak much less current than silicon and this is what allows us to save power.”

Future Implications 

Dr. Koester’s team has shown that the leakage current in MoS2 transistors is in fact so low that it could not be measured using normal direct electrical current measurement techniques. Instead, the team utilized a so-called gain cell memory, a design technique pioneered by Prof. Kim, which uses two MoS2 transistors. In this design, one of the transistors holds stored electrical charge on the gate of a second transistor which is used to read out the value of the stored charge. In this way, the  inherent gain of the transistor is used to enable the femtoampere-level leakage currents in one device to be deduced by analyzing the much larger microampere-level currents through the other transistor.

The breakthrough could have wide ranging implications for reducing power consumption, thereby improving battery life in laptops and handheld devices, while also increasing storage capacity in semiconductor memories. The low leakage could also enable higher-capacity cache memories which are important in multi-core microprocessors. MoS2 MOSFETs also have enormous potential for use in solid-state hard drives, where the extremely low leakage combined with the atomic-scale thickness could result in order-of-magnitude increases in storage capacity.

Given the seemingly endless hunger for more memory, the ultra-low leakage, super-thin, 2D wonder material MoS2 is proving that in the world of semiconductors, less  is truly more.

Learn more about research conducted by Prof. Steven Koester and Prof. Chris Kim