Abstract

INTRODUCTION TO SPINTRONICS

Pavel Streda

Institute of Physics ASCR, Praha

Communication systems demand the miniaturization and integration of low-power electronic devices. At the same time, faster devices are needed to process information. In all devices the properties of electrons carrying electric charge and spin are decisive. Today, the charge of the electron serves for information processing, while its spin, internal momentum, is used for information storing (magnetic memories). Spin electronics, spintronics, is supposed to develop devices in which spin of electrons will be used in information processing as well. The main goal of spintronics is to gain knowledge on spin-dependent phenomena of electron systems, and to exploit them for new functionalities - new devices.

Some of the problems to be solved by spintronics are the following ones:

• The increase density of memory elements requires a reduction in size of the sensors. The new generation of sensors are exploiting a giant magnetoresistance effect (GMR) of multiple layer structures made of alternating ferromagnetic, antiferromagnetic, and paramagnetic metals. In contrast to the traditional Hall probe, the operation of the GMR sensors depends on the electron spin and not on the charge. The most recent research on the field sensors focused on spin-dependent electron tunneling between ferromagnetic layers through an insulator. It is expected, that the parameters of tunneling magnetoresistance (TMR) sensors will make them suitable not just for reading devices, but also as position detectors, where Hall-effect sensors dominate today.
   

• A much more ambitious spintronic objective is to develop magnetic random access memories (MRAM). For this purpose, it is necessary to find means of writing and reading the direction of magnetization in given cells without employing any moving parts. An important step would be the elaboration of methods of controlling magnetization by light or electric field. Magnetic random access memories (MRAM) are non-vocative opposite to memories based on the accumulated electric charge (Dynamic Random Access Memory - DRAM) that require a frequent refreshing.
   

• Spin-transistor consists of two ferromagnetic metals separated by non-magnetic conductor which serves as the base. If spin-polarized carriers injected to non-magnetic layer conserve spin orientation, the spin accumulation effect will influence the current between source and base in the dependence on the relative directions of magnetization in the two ferromagnetic layers. If the non-magnetic layer is a semiconductor quantum well structure with a metallic gate, strength of the spin-orbit interaction can be modified by an external voltage. If the spin-coherence is preserved, the device exhibits the current modulation due to the spin precession in analogy with the electro-optic modulator.
   

• The fundamental quantum mechanical nature of spin places out of reach many of the forces in a host material and the orientation of electron spin can be very long-lived. It means that the phase coherence of the spin degree of freedom can be preserved for much longer time than the orbital degrees of freedom. Spin quantum devices are thus much more promising for quantum computation, the idea which is based on the quantum coherence.

The aim of the talk is to review the fundamental physics of the electron spin, coherent transport and elementary spin devices in the following blocks:

• The electron spin as the natural feature of the Dirac relativistic wave equation.
   

• Basic spin interactions: with magnetic fields (Zeeman energy splitting), with electric field (spin-orbit interaction) and with other electrons (exchange coupling).
   

• Electron transport and its description for coherent and non-coherent systems.
   

• Giant magnetoresistance and spin transistor.
   

• Coherent spin-polarized transport, spin precession and field-effect spin transistor.
   

• Semiconductor spintronic - promising but risky investment.