Storage is a very exciting thing these days: SSDs are increasing in capacity and becoming cheaper, hard drives are offering storage capacity that’s unprecedented at the consumer level, and recently, scientists have been able to store significant amounts of data using unusual mediums, such as strings of DNA and small groups of atoms. Now, scientists have managed to store data in individual molecules.
Using a new, still-experimental technology, researchers have managed to turn individual molecules into a storage medium. In theory, this molecular memory could increase current storage capacities by one thousand times over more conventional means.
Molecular memory isn’t an entirely new concept but there have always been significant hurdles, the first of which is no stranger to the computing world: cooling. Previously, molecular memory needed to be cooled to temperatures close to absolute zero — not exactly practical. However, a team led by Jagadeesh Moodera, a senior research scientist at MIT, has discovered a way to cool molecular memory using only temperatures approaching the freezing point of water — an easy-to-achieve temperature.
The team also overcame another significant hurdle standing in the way of molecular memory. Previously, the storage molecules — which take the form of a thin sheet of carbon and zinc atoms attached to each other, and are called “graphene sheets” – needed to be kept between two ferromagnetic (the common form of magnetism that we experience in everyday life) electrodes. Because of this, the molecular memory had to be shaped in a specific way. When the team measured the conductivity of the two electrodes, they expected to find a single change, as the two units were supposed to be working together, but rather they found two separate jumps in conductivity, realizing that the units were working separately of each other.
Considering the memory only requires one jump in conductivity to work, the team realized they could remove one of the ferromagnetic electrodes, and replaced it with a regular metal electrode, which didn’t disrupt the required conductivity. Due to this discovery, the shape of molecular memory theoretically does not have to be as constrained as it once was, making production easier and more malleable.
If the molecular memory came too close to the top ferromagnetic electrode, it could become disrupted; however, the regular metallic electrode won’t interfere, and thus the molecular memory has more room to be shaped and layered. Interestingly, if too many molecules are stuffed into a memory cell, that made control of the cell more difficult, so cutting down on the number of molecules actually increases a cell’s control.
Unfortunately, like most experimental technology that sounds so amazing that you want it right now, the molecular memory cell doesn’t provide enough power for a commercial device, and is currently only able to produce a 20% jump in conductivity. However, the field is promising, and having eliminated the need for near-absolute zero temperatures and removing some of the constraints of the shape and number of layers of the molecule sheets seems to mean that two of the biggest barriers have been removed.
Research paper: Interface-engineered templates for molecular spin memory devices
[Image credit: Wallsonline]