A New Scheme to Determine the Spin Hall Coefficient

The recent interest in the spin Hall effect has incited a large number of theoretical studies. Experimentally, it is still very difficult to determine the spin Hall coefficient. We proposed a new scheme to determine the  spin Hall coefficient of a semiconductor sample.  The scheme is simple,  accessible to most of experimental groups in the field.

Our scheme employs a continuous circularly-polarized laser light beam that induces an inhomogeneous spin density in the sample.  We then measure the induced charge current along the direction transverse to the gradient of the spin density.  By applying the Einstein relation and Onsager relation derived from the general principles of non-equilibrium thermodynamics, we can determine the spin Hall coefficient from the measurement.

The scheme was applied in a recent experimental study on InGaAs quantum well.  The spin Hall coefficient was estimated to be ~e/8p.  The value is surprisingly close to the theoretical prediction of the intrinsic spin Hall effect.

See the preprint for details.

Proper Definition of Spin Current in Spin-Orbit Coupled Systems

The study of spin-Hall effect and spin transport has attracted much attention recently.  One of the fundamental issues is the definition of spin current.  An intuitive choice of the spin current is to replace the electron charge -e in the usual charge current with the spin operator.  Unfortunately, this approach neglects one of the most important aspects of spins: unlike the charge, spins are not conserved in general.  This has caused some apparent difficulties already observed by a number of authors.

Our work surveyed this issue in depth.  We found that the conventional spin current includes the extraneous contribution from the local spin torque that must be deducted before the resulting conserved spin current can be considered as the true transport current.  It turns out that the proper definition of the spin current should be the total time derative of the spin displacement operator rsz.  The spin Hall coefficients of Rashba models, both linear and cubed k, were recently re-calculated using the proper definition. It was found that in the presence of the non-magnetic impurities, there is no intrinsic spin Hall effect.  This contradicts with the most of the earlier conclusions, and suggests that it may be necessary to re-survey the whole subject thoroughly.

See this PRL paper and these  slides  for the theory.

Electron Dynamics in Exotic Materials

For certain class of solids such as ferromagnetic metals, it had been found that the (Bloch) electron follows different dynamics from the one found in textbooks.  Most notably, there exists an extra "magnetic field" in reciprocal (k-) space that could drive the motion of electrons.  In theory, all solids that lack either time-reversal or spatial inversion symmetry could possess such a field.   So, it is foreseeable that a whole class of new materials could be discovered/synthesized in laboratory in the near future.

Our research try to laid down the theoretical ground for this new class of exotic materials.  Based on the previous works on the effective dynamics of Bloch electron, we are able to determine the complete form of the new quantum mechanics of Bloch electron in these exotic materials.  One of its most notable features is that the phase space of the electron acquires a measure which is non-homogeneous in general.  This will have profound effects in equilibrium and transport properties.  Its could also have important implications to some fundamental aspects of the condensed matter physics, such as the Fermi liquid theory.

See this PRL paper and these slides for a detailed survey of the theory.

Direct Extraction of Eliashberg Function From Photoemission

It turns out the photoemission technique can reveal more in-depth information about electron correlation in solids than that we had expected! The development of a systematic technique to extract the Eliashberg function directly from the high resolution angle resolved photoemission data made this possible.  The technique had been successfully applied to Be(1010) surface.  Recently this technique was also applied to high-Tc materials and the fine structures of the electron-boson coupling were identified in the LSCO compound for the first time. A number of experimental groups have employed the technique to study the properties of other materials.

Look at these slides to have a rough idea of the technique.  The computer codes doing the analysis can be downloaded in the Files section.

Surface Structural Phase Transitions Induced by Electron-Mediated Adatom-Adatom Interaction

The electron-phonon coupling is enhanced on many surfaces.  In metallic adlayers, it may lead to very strong adatom-adatom interaction  mediated by electrons.  This is responsible for the surface structural phase transition observed in Sn/Ge(111).   Our paper in Physical Review Letters explained many puzzling behaviors of the transition based on this model.  The prediction of the paper, i.e., the system may enter into a glassy phase at low temperature, has been confirmed by a recent experiment in iso-electronic Pb/Ge(111) system.

Look at these slides for a brief introduction.

"Zero Resistance States" in Microwave Radiated  2DEG

Recent experimental discoveries of "zero resistance state" in a microwave radiated 2DEG revealed that intriguing and unexpected transport behaviors may  arise in a far-from-equilibrium system. Our paper published in Physical Review Letters provided a simple theoretical rationale for the occurrence of the "zero resistance state".  From the physical view, this is the typical behavior of a far-from-equilibrium driven system, albeit in a quantum system.  It is still not clear  whether or not  the quantum nature can play essential role in the far-from-equilibrium transport.

See these slides for a brief introduction.

Two-Dimensional Metal-Insulator Transition

It had been hoped that a series of experiments could have provided evidences that a metallic phase may exist in two-dimensions. This contradicts with the well known weak localization theorem. Our Studies suggested that the observed transport anomaly could be well understood in a semi-classical picture in which the incoherent transport dominates.  In general, the transition is the Mott-Hubbard type, and the weak localization theorem, upon which many earlier arguments were based, is not relevant.

visitors since December 09, 2004