Scientists create single-atom bit, smallest memory in the world
November 17, 2013
The scanning tunneling microscope makes single holmium atoms on a platinum surface visible (credit: T. Miyamachi/KIT)
Karlsruhe Institute of Technology (KIT) researchers have taken a big step towards miniaturizing magnetic data memory down to a single-atom bit: they fixed a single atom on a surface so the magnetic spin remained stable for ten minutes.
“A single atom fixed to a substrate is [typically] so sensitive that its magnetic orientation is stable only for less than a microsecond,” said Wulf Wulfhekel of KIT.
A compound of several million atoms has been needed to stabilize a magnetic bit longer than that. That’s because the magnetic moments of these atoms are normally easily destabilized by interactions with electrons, nuclear spins, and lattice vibrations of the substrate.
The finding opens up the possibility of designing more compact computer memories and could also be the basis for quantum computers, Wulfhekel said.
How to create a single-atom bit
In their experiment, the researchers placed a single holmium atom onto a platinum substrate. At temperatures close to absolute zero (about 1 degree Kelvin), the atom was nearly vibration-free. They measured the magnetic orientation of the atom using the fine tip of a scanning tunneling microscope. The magnetic spin changed after about 10 minutes — “about a billion times longer than that of comparable atomic systems,” Wulfhekel said.
Abstract of Nature paper
Single magnetic atoms, and assemblies of such atoms, on non-magnetic surfaces have recently attracted attention owing to their potential use in high-density magnetic data storage and as a platform for quantum computing1, 2, 3, 4, 5, 6, 7, 8. A fundamental problem resulting from their quantum mechanical nature is that the localized magnetic moments of these atoms are easily destabilized by interactions with electrons, nuclear spins and lattice vibrations of the substrate3, 4, 5. Even when large magnetic fields are applied to stabilize the magnetic moment, the observed lifetimes remain rather short5, 6 (less than a microsecond). Several routes for stabilizing the magnetic moment against fluctuations have been suggested, such as using thin insulating layers between the magnetic atom and the substrate to suppress the interactions with the substrate’s conduction electrons2, 3, 5, or coupling several magnetic moments together to reduce their quantum mechanical fluctuations7, 8. Here we show that the magnetic moments of single holmium atoms on a highly conductive metallic substrate can reach lifetimes of the order of minutes. The necessary decoupling from the thermal bath of electrons, nuclear spins and lattice vibrations is achieved by a remarkable combination of several symmetries intrinsic to the system: time reversal symmetry, the internal symmetries of the total angular momentum and the point symmetry of the local environment of the magnetic atom.
comments 13
by silicaroach
Totally irrelevant but: I read the title and thought, “Like who else but scientists?” I suppose it would have been a _real_ attention grabber with something like “Sea slug creates … ” or “Hillbilly granny creates … “, but scientiests, aren’t they always doing stuff like that? Time for a coffee.
by Trance_Ender
It is always time for coffee.
In the picture, don’t the holmium atoms lookjust like those cored green olives with the pimento stuffing? I also like the ripples in the blue plasmon surface of the platinum substrate.
by erichlof
Should the title of this article instead read “…smallest memory in the universe” ?
Is an atom-size bit as low as we can go? Or can single protons and electrons (quarks?) be turned into computational memory as well?
Regardless, this is a great leap forward – thanks for the article. :)
by Editor
I don’t have information on memory technologies outside this planet, and I don’t know of any studies that define the lower bound for memory bits, but here’s one speculation:
http://www.kurzweilai.net/femtotech-computing-at-the-femtometer-scale-using-quarks-and-gluons
by Transhuman
Soon enough, we will engineer on that level as well, as long as the known laws of physics do not forbid it.
Since we have continued to refine our known laws as time has progressed, it is conceivably feasible that we may one day reduce our computronium to the femtometer, and possibly smaller, presuming the fundamental substrate of the universe is not infinitely granular.
Of course, once we do get to that level, manipulating matter and charge will seem like childs play.
Realm of the Gods, indeed.
by Trance_Ender
Manipulating mass and charge even at this gross nano level is indeed child’s play, to the agelessly young children of all ages who are involved directly with doing it, and to those of us who know a little math and physics and understand a few simple equations but mostly have our fun just looking at the pictures.
Does current quantum doctrine assert that cosmic substrate granularity actually is finite, reaching its ultimate granularity at “Planck’s distance”, on the order of 10^-43 m ?
by Ryan
I think it would be easiest to say, and most accurate, that it would be the smallest we can currently achieve with our knowledge and technology. I know there are efforts to explore femtotech but right now its pretty much all theory, somewhat like nanotech was a few decades back. Femtotech would manipulate subatomic particles such as neutrons, protons and electrons. Again, we’re still just getting a handle of the rules of things at the nanotech level, so unless we do reach an intelligence explosion with advanced AI it could be a while before we see anything real arising in the femtotech arena. After that we’re aware of things at smaller scales but whether we can harnes and shape those things is up in the air. Personally I would say it is possible but we simply lack the tools to do so.
by Trance_Ender
To go, much further there is.
(Hint: according to Pauli’s Exclusion Principle, no more than 2 electrons may occupy an orbital.)
by nanotech@republika.pl
“(…) to absolute zero (about 1 degree Kelvin)” should read “(…) to absolute zero (about 1 Kelvin)”. The unit is Kelvin, not degree Kelvin. That’s a common mistake.
by victor
What was once a dream, is becoming a reality. Someone once said that in a grain of sand, theoretically, one can put as much computing power as in all of the world (at the time of saying). First step is done, then. And this will lead extremely cheap computers…
Next they may find a kind of cage to isolate the atom from different interference and make possible single atom bit at higher temperature.
by Trance_Ender
I could swear I recently came across an article about an experiment involving trapping one atom inside a cage of light beams. Yes – it is in TIME magazine, issue dated Monday 2013.11.25: a brief article about a new kind of atomic clock said to be accurate beyond the time limit for the existence of this universe. The atom in this case happens to be ytterbium, which happens to be a lanthanide rare earth element like holnium.
In “The Singularity Is Near”, Ray Kurzweil discusses an hypothetical material he calls “computronium”, which would have the attributes of your grain of supercomputing sand.
by Editor
http://www.kurzweilai.net/a-step-toward-the-quantum-internet
http://www.kurzweilai.net/worlds-most-precise-clock
by Trance_Ender
Thanks, Amara!