We are, collectively, producing an astounding amount of bits and there is no ending in sight to the growth. Google, in 2016, is storing bits in 15 data centres (as of 2016), managing 40 million searches per second through over 30 million servers (if you think about it is means a server answering about one search per second, which seems to be very little, but there is much much more running behind the scene that these servers are doing…).
Overall the energy used is impressive and so is the cost. No wonder that scientists al over the world are looking at ways to increase the storage density going beyond present technology.
A way to make a significant step forward would be to use single molecules to store a bit and scientists have already proved the technical feasibility, as an example using a terbium atom as a single magnet. The problem when you are scaling down to molecular level is the ambient “noise” that can alter the data, and noise at atomic level equals temperature. Researchers have been able to store a bit on a single molecule but they had to cool it down to a few degrees above absolute zero. This, in turns, requires a tremendous amount of power and in the end it would not solve the problem.
This is why the research published on Nature by scientists at University of Manchester is so interesting.
They found a composite material, hexa-tert-butyldysprosocenium complex, based on dysprosium (atomic number 66), that can have a good magnetic hysteresis cycle at 60 Kelvin (-213C). Having a good magnetic hystereses cycle is great: basically that it takes quite a bit of effort to change its magnetic orientation, which in turns ensure that if we store a data that data will remain stored and will not be affected by noise.
Now, -213 C still seems pretty cold, way beyond freezing. However it is within the freezing power of liquid nitrogen, and freezing with liquid nitrogen is way more cheaper than using liquid helium, which is what you would need to use to operate at atomic level with other materials.
It brings storage on single molecules within the range of economic feasibility. There is still a lot of work, engineering work, that needs to be done to make this happen. The incentive is great. It may mean to increase storage density a hundred times decreasing cost at least ten times.
Additionally, this can have an impact on development of quantum computers where again the “noise” is affecting the stability of the computation.