Predicted state of atomic collapse seen for first time

Atomic collapse, a phenomenon first predicted in the 1930s based on quantum mechanics and relativistic physics but never before observed, has now been seen for the first time in an ''artificial nucleus'' simulated on a sheet of graphene.

Scanning tunneling microscope image shows an artificial atomic nucleus on graphene, consisting of five pairs of calcium atoms (slightly darker circles at center), in an electron cloud that is on the verge of collapse.

The observation not only provides confirmation of long-held theoretical predictions, but could also pave the way for new kinds of graphene-based electronic devices, and for further research on basic physics.

The achievement, by a team of scientists from MIT, the University of California at Berkeley, and other institutions was reported online last week and will appear in a forthcoming issue of the journal Science.

Leonid Levitov, a professor of physics at MIT and a co-author of the paper, says this work follows up on an early success of quantum mechanics that showed why matter is stable: It detailed how the positive charge of an atomic nucleus, and the negative charge of its surrounding electrons, balance each other out, preventing the atom from collapsing or flying apart.

But those early calculations also showed that this balance should break down above a certain point - specifically, if the charge of the nucleus is more than 137. According to theory, in such super-charged atomic nuclei, the electrons should collapse into the nucleus, where they would then eject their antimatter opposites, positrons, which would spiral outward and away.

Later refinements of the theory raised the threshold number from 137 to 170, but the underlying principle remained. ''Such atoms were expected to collapse by grabbing an electron from vacuum, pulling it onto the nucleus and recharging,'' Levitov says.

Since the known natural and man-made elements only reach an atomic number of 118, the prediction has been hard to demonstrate experimentally. Physicists have tried to demonstrate atomic collapse in particle accelerators by taking two heavy nuclei, such as those of uranium atoms (atomic number 92) and smashing them together, Levitov explains. ''These experiments have been tried for decades,'' he says, but no clear-cut evidence of collapse has been found.

But last year, a team at Berkeley and at Lawrence Berkeley National Laboratory demonstrated a new technique that could simulate heavy atoms in a way that was much easier to manipulate and observe. The new findings are a culmination of that research.

What the new Science paper reports is that atoms sitting on a sheet of graphene - a two-dimensional structure composed of carbon atoms linked in a chicken-wire-like mesh of hexagonal bonds - exactly mimic the properties of atomic nuclei, and can be manipulated to recreate and observe complex atomic phenomena.

The key is that while electrons move through graphene as relativistic particles - as though they were massless, even though they actually do have mass - their motion is 300 times slower than that of true massless particles. As a result, the expected phenomenon of collapse should take place at one-three-hundredth the normal nuclear charge - putting it well within reach of experimental observations.

To simulate atomic nuclei, the researchers used pairs of calcium atoms on the graphene surface; they were able to manipulate these pairs (called dimers) on the surface using the probe tip of a scanning tunnelling microscope.

As soon as three dimers were pushed close together, the surrounding field of electrons showed a specific spectrum of resonances that precisely matched the decades-old predictions of atomic collapse. The observed resonances persisted in a four-dimer and five-dimer artificial nucleus.

These results provide ''a clear case for collapse,'' Levitov says. ''We are certain this is correct.''

Though the initial impetus for the work was the desire to prove a long-accepted theory about the quantum-mechanical behaviour of atoms, the work has more than just theoretical relevance. As researchers around the world race to create electronic devices on graphene - harnessing that material's unique strength, flexibility and electronic properties - these findings, and the techniques developed to produce them, could provide important insights into graphene's behavior, Levitov says.

What's more, the delicate sensitivity of artificial atoms on a graphene surface make them incredibly responsive to surrounding conditions, which could lead to new detectors for trace chemicals or biomarkers, he suggests. The technique will continue to be used to probe different configurations of artificial atoms - perhaps linking them together into artificial molecules - to probe factors important to basic physics and chemistry.

In addition to confirming a decades-old prediction, this study produced findings that remain to be explained; the researchers found that removing a charge from a nucleus produced a very different response than adding a charge to it - something that had not been expected. ''It means there is something going on that we don't understand,'' Levitov says.

Philip Kim, an associate professor of physics at Columbia University who was not involved in this research, says, ''This is extremely nice work. … This paper nicely demonstrates a long-sought relativistic quantum-mechanics analogy in graphene … (and) both provides clear experimental realisation as well as a deep and precise theoretical understanding behind this unique phenomenon.''

The paper was co-authored by 13 researchers from MIT, Berkeley, Lawrence Berkeley National Laboratory, the University of Exeter in the UK, and the University of California at Riverside. The work was supported by the Office of Naval Research, the U.S. Department of Energy, the National Science Foundation, and the Engineering and Physical Sciences Research Council in the UK.