In quantum mechanics particles can escape from their confines, even if a barrier stands in their way, via a process known as tunneling. Tunneling is no mere quantum curiosity—tunneling electrons, for instance, are harnessed by scanning tunneling microscopes to observe on the smallest scales. Those probes can image a surface at the atomic level by detecting the tunneling of electrons from the surface across a small gap to the microscope's tiny scanning tip.

A paper in this week's Science adds new depth to tunneling by showing how readily electrons can tunnel out from multiple orbitals in a molecule. "Until just recently, everyone would have thought that only the most easily available electron could tunnel," says study co-author Paul Corkum, a physicist at the University of Ottawa and director of Attosecond Science at the National Research Council Canada. A series of research papers in the past few years has begun to revise that thinking, showing that lower-lying orbitals get into the act, as well.

In the new research, Corkum and his colleagues observed electrons tunneling out of hydrogen chloride (HCl) molecules subjected to laser pulses and traced the electrons back to their parent orbitals. "You would think that the highest one, which has to go underneath the classically allowed barrier by only a little bit, would have a huge advantage, and the [next] lower one, which has to go under the barrier by a lot, would be highly suppressed," Corkum says. But the team found that the second-highest orbital contributed a measurable amount to the total tunneling current.

Late last year two groups published papers in Science showing how intense laser pulses could be used to liberate electrons not only from the highest molecular orbital but also from the next orbital below. Markus Gühr, a Stanford University chemical physicist at the SLAC National Accelerator Laboratory in Menlo Park, Calif., co-authored one of those papers with a view toward examining molecular processes in real-time.

"The general vision that we have in the community is we want to look at chemistry," says Gühr, who did not participate in the new research. Probing the way electrons form and break bonds between atoms is critical to tracing the workings of chemistry at the ground level. "The sensitivity on electrons is really a new crucial step, I would say," Gühr adds.

In the 2008 work Gühr's group examined electrons tunneling in nitrogen. "The nitrogen molecule has the advantage that these two orbitals...are pretty close together," Gühr says. In the hydrogen chloride molecule probed by Corkum's group, he adds, the orbitals are much farther apart, making the tunneling contribution from the lower-lying one all the more notable.

The hydrogen chloride molecule made a handy test bed for such a tunneling experiment. When the electron is stripped from hydrogen chloride's highest orbital, an ion (a charged version of the molecule) survives. But the electron in the next orbital down accounts for the bond between the molecule's atoms, so when an electron tunnels from that orbital, the HCl molecule breaks apart. Such a fragmentation is one signature of the lower-level tunneling.

Aside from demonstrating lower-orbital tunneling in a molecule that is less amenable to it, Corkum's group was also able to show just how often it takes place. "I would say in the evidence that we presented [last year], and also that other groups presented, it is clear that [the lower-level orbital] definitely has a contribution in tunnel ionization," Gühr says. "But it is not clear quantitatively to which exact extent." Corkum's group, Gühr adds, has taken the next step of directly quantifying the contribution of the lower orbital—in this experiment, the lower-lying orbital contributed 0.2 percent of the total tunneling current.

Corkum notes that it is exponentially more difficult, but not theoretically prohibited, to get electrons from lower orbitals rather than from higher orbitals. So although the new results were at first glance surprising, they make sense from a physics perspective. "I would say our prejudice was wrong, not the theory itself," Corkum says.