The intensity map (black circle) represents the electron distribution of the surface state close to the Fermi level. The green (blue) arrows indicate the spin polarization in the y (x) direction; the red arrows the total spin polarization. The lengths of the arrows indicate the magnitudes of the spin polarizations. The surface spin polarizations (red arrows) measured at momentum locations A and B are tangential to the circle (helical spin structure) and are very large.
With an eye on the developing and exciting fields of spintronics (electronics based on electron spin rather than charge) and quantum computing, scientists working in part at the National Synchrotron Light Source have learned important information about a new state of matter known as a topological insulator. These materials may eventually play an important role in these technologies due to their unusual electronic and spin structures.
Topological insulators (TIs) are unique in that their interior, or bulk, acts like an insulator, but the surface of the material conducts. A regular insulator is characterized by a large “band gap,” the difference between the energies of the material’s valence electrons and the energies they need to become mobile charge carriers. In a TI, this gap disappears at the surface.
The research group includes scientists from Brookhaven National Laboratory, Lawrence Berkeley National Laboratory (LBNL), and the Massachusetts Institute of Technology. They studied the model topological insulator Bi2Se3, composed of the elements bismuth and selenium. Bi2Se3 is of particular interest to researchers because its bulk band gap is relatively large and its surface structure is relatively simple, both key traits in promising TIs.
“TIs have quickly become one of the most interesting materials around. Not only do they show a huge potential to serve as a basis for technological applications (spintronics, quantum computing), but also they display a wide spectrum of extremely exotic phenomena that no other material can match,” said Brookhaven physicist Tonica Valla, who participated in the research.
The researcher’s results refute those of past experiments, which concluded that the electron spin polarization on the Bi2Se3 surface – the degree to which the spins line up in a certain direction – is too low to make the material suitable for quantum computing or spintronics. Also troubling is that theory predicts full spin polarization, pointing to the possibility that the theory is wrong and, by extension, that scientists don't have a basic understanding of TIs.
But via studies at NSLS and LBNL’s Advanced Light Source (ALS), the researchers measured almost full spin polarization– meaning that the electron spins were nearly perfectly aligned – much higher than previously reported, just as theories predicted.
“The exact amount of spin-polarization depends somewhat on the precision of the experimental apparatus and on details of the analysis, but our results show large, nearly full spin polarization of these states,” said the paper’s lead author, Brookhaven scientist Zhihui Pan.
Added Valla, “However, there are still some significant obstacles that may prevent these materials from showing their full possibilities. For example, due to the crystal imperfections, none of the materials synthesized so far are truly insulating, causing the exotic elusive effects on their surfaces to be hidden by the conventional bulk conduction.”
At NSLS, the group employed a technique called spin- and angle-resolved photoemission spectroscopy (SARPES) to study the Bi2Se3 surface. In SARPES, a focused beam of ultraviolet light is aimed at a sample, which absorbs it and emits the electrons. These electrons are collected by detectors and the data analyzed, allowing scientists to determine the electrons' binding energies, momenta, and spins.
The NSLS work was complemented at ALS by conventional angle-resolved photoemission spectroscopy (ARPES), the technique first used to confirm the existence of TIs.
The paper corresponding to this work was published in the September 28, 2011, edition of Physical Review Letters.
Summary Slide (.pdf)