LBNL Masthead A-Z IndexBerkeley Lab mastheadU.S. Department of Energy logoPhone BookJobsSearch
A Spintronic Semiconductor with Selectable Charge Carriers Print
Wednesday, 28 August 2013 00:00

Accentuating the Positive
(or the Negative)

Spintronics—a type of electronics that makes use of electron spin as well as charge—is already here to a certain extent. The discovery of giant magnetoresistance, a spin-based effect, has revolutionized the information storage industry. Beyond this, however, scientists envision the possibility of combining storage and processing functions in one integrated system. In electronics, processing is done using semiconductor materials like silicon and germanium that have the requisite properties to perform logical operations with both electrons (negative n-type charge carriers) and holes (positive p-type charge carriers). Thus, a spintronically desirable semiconductor would simultaneously have discrete spin-up and spin-down states as well as both positive and negative charge carriers.

Strategies for developing spintronic semiconductors have been based on surface doping or on alloying, both of which have drawbacks such as chemical instability or reduced mobility. In BiTeI, however, electron and hole conduction is achieved without modifying the ideal crystal structure. One of the things discovered by Crepaldi et al. was that the electronic band structure of BiTeI bends in different ways near the surface depending on which layer is on top. That, in turn, means that the Fermi level (which determines a material's conductivity) can be located in either the valence band (for positive charge carriers) or the conduction band (for negative charge carriers). With techniques such as molecular-beam epitaxy and chemical vapor deposition, it is realistic to consider that regions with opposite band bending could be patterned on a substrate, opening new possibilities for the manipulation of spin-polarized states.

The ultimate goal of spintronics is to utilize electron spin—in addition to charge—for the storage and processing of data. However, the manipulation of spin typically requires magnetic materials. While commonly found in components used for data storage, magnetic materials are not generally suitable for semiconductor devices used in data processing. Now, with angle-resolved photoemission spectroscopy (ARPES) at the ALS, researchers have shown that two properties crucial for enabling semiconductor-based spintronics—a large Rashba effect (splitting of degenerate spin states) and ambipolarity (conduction via both electrons and holes)—are present in the semiconductor BiTeI. Furthermore, they found that it is possible to control whether the charge carriers are electrons or holes by engineering the surface layer.

A fundamental law of quantum physics states that, in a crystal lattice possessing space-inversion symmetry, degenerate spin states cannot be separated in the absence of a magnetic field. However, the key property of most Rashba systems is that they are two dimensional—typically surfaces or interfaces—so that space symmetry along the normal direction is broken by definition and spin separation can in fact happen. Unfortunately, this means that the spin-split states available for conduction are only those confined to the surface of the material. However, there is another class of nonmagnetic materials in which spin separation is possible because space-inversion symmetry is broken in the bulk.

BiTeI has a layered structure, with Bi, Te, and I planes alternating along the c‑axis. The Bi and Te planes are covalently bonded to form a positively charged (BiTe)+ bilayer. The ionic coupling between the bilayer and the adjacent I plane defines the natural cleavage plane.

BiTeX compounds consist of alternating layers of bismuth, tellurium, and a halogen atom, like chlorine, bromine, or iodine. The layer sequence is such that the axis normal to the layers is chiral, meaning that the crystal is not symmetric with respect to an upside-down flip. In addition, they contain Bi and Te as heavy elements, which induce a strong spin-orbit interaction, an essential ingredient for obtaining a sizeable spin separation. Previous ARPES experiments had revealed that the electronic structure at the Fermi level consists of two separate bands, with spin-up and spin-down states clearly split in energy. Using ARPES on Beamline 7.0.1, a collaboration between groups from the Swiss Federal Institute in Lausanne (EPFL) and the ALS has now found that BiTeI presents another crucial quality: the material is ambipolar, meaning that it can support conduction by both negative (electron) and positive (hole) charges.

Top panels: ARPES data showing electron-like (left) and hole-like (right) split bands for Te and I, respectively.
Bottom panels: Band structure calculated from first principles. The continuum of bulk states is shown in blue.

While the original purpose of the study was to disentangle the surface versus bulk origin of the different electronic states, in measuring a series of samples, something even more interesting was noticed: there seemed to be two different sets of data, with some resemblance but overall distinct structures. In one case, the low-energy states were electron-like, while in the other they were hole-like. A close look at the core levels helped to clarify that the two data sets were associated with the two surface layers found on the samples, either Te or I. (Strong covalent bonding between the Bi and Te layers prevented exposure of the Bi layer when the material was cleaved.)

BiTeI has a chiral axis and a preferential cleaving plane between the I and Te layers. The two surface terminations can be obtained either by flipping the sample upside down or as a result of stacking faults, as in the figure. The bottom panels show the band bending for the two surface terminations, with the Te (or I) termination behaving as a p (or n) semiconductor.

In other words, the orientation of the sample determined the sign of the charges at the Fermi level, similar to p or n doping in conventional electronic semiconductors. Therefore, not only did the surface-related states host charges of opposite sign, but thanks to the semiconducting nature of the material, the team showed that they could move the chemical potential (Fermi level) from the conduction band to the valence band by controlling the surface termination, which is equivalent to selecting either spin-polarized electron or hole states as the conducting states. The observation is even more interesting because such layered samples are typically suitable for making epitaxial films. In that case, the surface termination could be controlled at the growth stage and, in principle, a device could be patterned with different paths for the two terminations.

 


 

Research conducted by: A. Crepaldi, G. Autès, C. Tournier-Colletta, S. Moser, N. Virk, H. Berger, Ph. Bugnon, K. Kern, O.V. Yazyev, and M. Grioni (Ecole Polytechnique Fédérale de Lausanne, Switzerland) and L. Moreschini, Y.J. Chang, A. Bostwick, and E. Rotenberg (ALS).

Research funding: Swiss National Science Foundation. Operation of the ALS is supported by the U.S. Department of Energy, Office of Basic Energy Sciences.

Publication about this research: A. Crepaldi, L. Moreschini, G. Autès, C. Tournier-Colletta, S. Moser, N. Virk, H. Berger, Ph. Bugnon, Y.J. Chang, K. Kern, A. Bostwick, E. Rotenberg, O.V. Yazyev, and M. Grioni, "Giant Ambipolar Rashba Effect in the Semiconductor BiTeI," Phys. Rev. Lett. 109, 096803 (2012).

ALS Science Highlight #278

 

ALSNews Vol. 345