Researchers at IBM San Jose have just reported the experimental demonstration of a magnetic bit made with only 12 atoms (press release). Sebastian Loth, the first author of the paper, just published in Science (link to the paper), will visit INL next week and give a seminar on Tuesday 31 January.
Non-volatile storage of digital information, as recorded in conventional hard-disks, makes use of the spin orientation of magnetic grains made of more than one million atoms. The magnetic grains are made of ferromagnetic materials, so that the magnetic moment of each atom tends to point parallel to their neighbors, in the so called ferromagnetic coupling, adding up to form a net magnetization. Magnetic grains have a so called easy axis, so that the collective magnetization points either parallel or anti-parallel to the easy axis. These two possible stable orientations define the two logical states of a classical bit, 0 or 1.
Increasing the storage density in magnetic recording involves reducing the size of the magnetic grains, which leads to a number of practical challenges involving the detection of their magnetization and the loss of stability: the time scale for unwanted flip of the magnetization decreases exponentially as the volume of the particle is reduced. This naturally leads to the question of how small can the magnetic particle be which leads a sufficiently large lifetime for the spin orientation.
The experiment of the IBM team addresses this question but changes the strategy in terms of the type of magnetic material. Instead of using ferromagnetic nanoparticle they use an atomically engineered antiferromagnetic system, in which the magnetic moments of a given atom likes to point anti-parallel to their neighbors. Taking advantage of a Scanning Tunneling Microscope (STM) they can move atoms, one by one, deposited on a surface, and design structures with arbitrary shapes. In this instance they have fabricated chains of a few (6,8) iron atoms, on top of a copper surface capped with a single monolayer of an insulating material, Copper Nitride.
The antiferromagnetic coupling between iron atoms result in a vanishing total magnetization for a given chain, which goes completely against the usual approach. They can get away with it taking advantage of a readout technique with atomic scale resolution: the spin polarized scanning tunneling microscope (STM). By reading the spin orientation of every atom in a given chain , the IBM team has been able to observe how the short chains of 4 atoms and less behave as quantum antiferroamgnets: the spin every atom is in a quantum superposition of being up and down, which results in an average null magnetization for each atom, and the impossibility of using such system so store digital information.
Interestingly, for chains of 6 atoms and higher the chains behave as expected in classical antiferromagnets and have two stable orientations: 1) Even atoms up, odd atoms down and 2) Even atoms down, odd atoms up. Taking advantage of the atomic scale readout the IBM team could trace the thermal stability of the different chains and found how the stability improved for larger chains. In particular, 2 chains with 6 atoms lying parallel, referred to as (2×6) have been shown to be suitable to store a bit of classical information.
In addition to the use of STM to fabricate the chains and to read their magnetization, the IBM team has also shown how to write the logical state in a bit using electrical pulses send with the STM. Combining the 3 capabilities together, the IBM group has fabricated a byte made with 8 (2×6) chains and has recorded on it the digital code for the letters S, P, I and N.
Whereas there is a long way between these experimental breakthroughs and massive commercial applications, mostly duet the low temperature and ultra-high vacuum experimental conditions at which the experiments have been done, the IBM team has made it real the vision of Feynman, back in 1959, in his inspiring talk “There is Plenty of Room at the bottom” where he prophetically said “When we get to the very, very small world – say circuits of seven atoms – we have a lot of new things that would happen that represent completely new opportunities for design”. The group of Theory of Nanostructures at INL is making progress to understand the working principles of all these striking experiments.
Joaquín Fernández-Rossier, Staff Researcher at INL