“Topological insulators” (TIs) are artificially constructed materials that are insulating in the bulk but can conduct electricity on their surface via special surface electronic states. “Most TIs made to date have not been completely insulating in the bulk, however, because of impurities (unintentionally introduced during material synthesis or processing) that doped the bulk and made it conducting,” explains corresponding author Yong Chen of Purdue University. “Our TI appears not to conduct at all in the bulk but does so only at its surface.”

The researchers measured how thin flakes of BiSbTeSe2 of various thicknesses conducted electricity. They found that the conductance of these materials was almost independent of their thickness. Such behaviour is completely different to that seen in normal 3D materials in which conductance is proportional to sample thickness.

Only conducting at the surface

“Our result is consistent with the picture that bulk BiSbTeSe2 is only conducting at its surface,” Chen tells nanotechweb.org. “It is like you were to keep cutting and reducing the material to ever smaller thicknesses and strangely never finding the conductance to change much. This is because every time you create a new surface, you get the same conduction.”

Indeed, the researchers observed this topological surface conduction even at room temperature for samples thinner than around 100 nm – something that will be important for practical applications.

“Half-integer” QHE

And that is not all: Chen and colleagues also found evidence for a well-defined “half-integer” quantum Hall effect (QHE), where the top and bottom surfaces of their thin-slab samples each contribute a half integer unit of quantum conductance (e2/h), where e is the charge on the electron and h is the Planck constant. These two half integer units make up the measured Hall conductance plateau – quantized at integer units of e2/h. Such half-integer QHE is another unique signature of topological surface state charge carriers, which are in fact, spin-polarized massless Dirac fermions.

As an aside, these massless Dirac fermions are analogous to the massless Dirac fermions that exist in graphene, explains Chen. It is these massless charge carriers that make graphene so unique, with its high conductivity among other exceptional physical properties. The difference is that in graphene, the charge carriers are not spin-polarized and come in four “degeneracies”. For a TI surface, on the other hand, there is only one species of Dirac fermion – the simplest, or “1/4 graphene”.

The researchers observed the QHE in BiSbTeSe2 flakes that were between tens to hundreds of nanometres thick, at cryogenic temperatures of below 30K and at high magnetic fields applied perpendicular to the top and bottom surfaces of the samples.

“For this part of our experiments, we used a powerful magnet (where we can get up to 33 Tesla) at the National High Magnetic Field Lab in Tallahassee, Florida ” says Chen.

A "perfect" TI

Our results point to a “perfect” TI – and one that behaves just how theory says it should. It is thus an excellent material platform in which to look for exotic physics predicted to exist in these structures.

For example, “quasiparticles” that are condensed matter analogues of the exotic Marjorana fermions (fermions that are their own antiparticles) could be made by combining an ordinary superconductor with such a TI. These fermions are predicted in high-energy physics but have yet to be observed in particle physics experiments. If they were to be found in such a solid-state setting, they might be used to build a so-called topological quantum bit, or qubit – which could help make a fault-tolerant quantum computer, for example.

Why fault-tolerant? Marjorana fermions – unlike more familiar Dirac fermions, such as electrons – could be made to obey "non-Abelian statistics" in solid-state structures and so should be robust to background environmental noise. They would therefore be able to store and transmit quantum information without being perturbed by the outside environment - one of the main challenges facing anyone trying to build a practical, fully functioning quantum computer today.

Spintronics and “topological magnetoelectronics”

Another promising application is in spintronics, says Chen, thanks to the fact that surface carriers on TIs are spin polarized. Spintronics is a relatively new technology in which devices exploit the spin of an electron instead of just its charge - as is the case in conventional electronics.

As well as these exotic physics and applications, we might also be able to realize phenomena like “topological magnetoelectronics” in the new 3D TI, he adds. These would allow us to create effective magnetic monopoles, for instance, by exploiting an unusual form of electromagnetism (different from that described by conventional Maxwell equations) predicted for such 3D TIs.

The team, which includes physicists from Princeton University and the University of Texas at Austin, reports its present work in Nature Physics.