How Things Work: Nanoparticle-based memory
Modern hard disk drives last a few years before they have to be discarded — they are plagued by mechanical failures that can occur anywhere between two and six years into their lives. Solid state drives offer some consolation in that the amount of time they can store data is approximately 2400 years if handled carefully, according to a paper published by Imation Corp. However, for those among us who would need to store data for over a billion years, we would probably have to resort to nanoparticle-based non-volatile memory.
Of the many nanomaterials, carbon nanotubes have attracted the most attention, and not without cause. Carbon nanotubes are tubes made of interconnected carbon atoms.
They may be single-walled, made of a single layer of carbon atoms, or they could be multi-walled, with multiple layers forming the tube. These nanotubes possess many properties that set them apart from most materials on the nano, micro, and macro scales.
Carbon nanotubes are the strongest and stiffest materials known today, and certain forms of these nanotubes are capable of conducting electricity. It is mainly on account of these two properties that carbon nanotubes are used in nanoparticle-based memory.
Digital data is stored — and processed — in binary digits, known as bits. These bits can be thought of as the states of a switch; they can be “on” or “off.” In nanoparticle-based memory, the basic idea is to shuttle a nanoparticle (made of a metal like silver or iron) between the two ends of a multi-walled carbon nanotube. Usually, one end of the tube is considered to be at a certain logic level zero while the other end is the logic level one. The position of the shuttle within the nanotube, therefore, simulates the same states of “on” and “off,” and if the shuttle position can be frozen, the system will function as non-volatile memory. Non-volatile memory is capable of storing data even when the power supply is removed.
“Writing” data into this sort of memory is to simply position the shuttle in one of the different predefined regions. Particle motion within the nanotube is brought about by flowing a current through the device. The direction of motion can be reversed by reversing the direction of supplied current.
The speed of the particle as it travels along the tube is determined by the magnitude of the voltage applied across the ends of the tube. It follows that by reducing the supplied current to zero, the position of the shuttle can be fixed at some point within the tube.
The solid state physics group at the University of California at Berkeley, led by Alex Zettl, noticed that the electrical resistance offered by the nanotubes varies with the position of the particle, making it possible to locate the particle — with reasonable accuracy — by simply measuring the resistance along the tube. This is how the written data is then “read.”
The nanotubes are sealed at both ends with electrodes, and these electrodes provide the shuttle with the current necessary for shuttle motion or positioning.
At the nano scale, friction is negligible and the friction offered by the nanotube to the shuttle will be so trivial that no mechanical damage will be caused during particle movement.
Unlike hard disk drives, the mechanical components in this memory will not, in theory, fail. Furthermore, this lack of friction allows for an unlimited number of write/rewrite processes, unlike solid state drives that become unreliable after a certain number of write processes.
Zettl’s group has shown that data can be stored for longer than a billion years using this method of storage.
The nanotubes are protected, as they are hermetically sealed by the two electrodes, and this prevents contamination and possible loss of data.
They also claim that compared to modern data storage systems, more data can be stored per area of storage material used.
While today’s state-of-the-art hard disk drives can store approximately 200 gigabits (200 billion bits) per square inch, it is believed that this proposed method can increase data densities to around 1000 gigabits per square inch.
So far, only small numbers of such carbon nanotubes have been assembled in order to test these data storage capabilities. These numbers have not been large enough to warrant mass production.
However, they do prove that this model works, and can be implemented with good results. It could well be quite a few years before this goes into production, quite possibly more than a decade before this system is perfected. Until then, we will have to wait, and while doing so, back up our digital data on solid state drives.