POTENTIAL APPLICATIONS OF SPINTRONICS Dept. of ECECS, Univ.of Cincinnati, Cincinnati, Ohio 45221...

POTENTIAL APPLICATIONS OF SPINTRONICS

Dept. of ECECS, Univ.of Cincinnati,

Cincinnati, Ohio 45221

http://www.ececs.uc.edu/~mcahay

M.CahayM.Cahay

February 4, 2005February 4, 2005

Outline

A Little quBit of HistorySuccess Story: Giant MagnetoresistanceSpin ValveRequirements for spintronicsZeeman, Spin-Orbit EffectsInjection, Manipulation, DetectionMagnetoresistive biosensorsConclusions

Brief HistorySpintronics-Magnetoelectronics

• Stern-Gerlach Experiment (Early 1920s)• spin concept – 1920s• Pauli-Dirac Equation (Late 1920s)• 1980s… : Study of Mesoscopic systems – Landauer-

Buttiker Formalism • Breakdown of Moore’s Law?• 1988: Giant Magnetoresistance in Magnetic multilayers,

magnetic read heads, magnetic sensors, spin valves.

Brief HistorySpintronics-Magnetoelectronics

• 1990: SPINFET proposal by Datta and Das

• 1990s: Lots of work on Ferromagnet/ (metal, semiconductor, superconductor) interfaces.

• DMS – diluted magnetic semiconductors (ZnMnSe, GaMnAs,GaMnN,…)

• 1985: David Deutsch:

Quantum Mechanical Turing Machine

Brief HistorySpintronics-Magnetoelectronics

• 1995: P.Schor’s algorithm for fast factorization of large integers (cryptography)

• 1997: L.K.Grover’s search algorithm for efficient search of large database

• 1990s: Lots of proposal for implementation of qubits and quantum computers (NMR, Ion trap, quantum dot)

• Search in Spintronics and Quantum Computing will continue to feed on each other

SIA ROADMAP - Moore’s Law

Requirements for Spintronic Integrated Circuits

• Simple device structure for high degree

of integration and high process yield.

• Large magnetocurrent for high speed operation

• High transconductance for high speed operation

• High amplification capability (V, I, and/or power)

• Small power delay product and small off-current

for low power dissipation

Preliminaries: Stoner Model

Exchange Energy

E(k)

k

Ferromagnetic Contact

GMR Read Head

RAM with GMR Elements

Preliminiaries: Zeeman Preliminiaries: Zeeman EffectEffect

Preliminaries:Spin-Orbit Preliminaries:Spin-Orbit InteractionInteraction

Ferromagnet/Sm Interface

Contact Selection

• Ferromagnetic contact (Fe, Ni, Co)• HMF candidates: Heusler Materials• Dilute Magnetic Semiconductors (GaMnAs,

ZnMnSe, ZnMnTe,GaMnN…)• More recently, wide band gap ferromagnetic

semiconductors and oxides S.J. Pearson et al., “Wide band gap ferromagnetic

Semiconductors and oxides”, Journal of Applied Physics, Vol.93, pp.1-13 (2003)

Ferromagnetic contact/semiconductor interfacesHow good are they?

Why Ferromagnetic Contacts (Fe, Ni, Co)? Because Curie Temperature Is Above Room Temperature!

Hence, devices could work at 300k. FM are good source of spin polarized electron sources

(Stoner model)

Theoretical PredictionsTheoretical Predictions Classical diffusion eq. predicts very small spin Classical diffusion eq. predicts very small spin injection efficiency across Fe/Sm interface (G. Schmidt et injection efficiency across Fe/Sm interface (G. Schmidt et al. PRB 62,R4790 (2000). Main reason: Large conductivity al. PRB 62,R4790 (2000). Main reason: Large conductivity mismatch between the two materials.mismatch between the two materials. Not so fast! E.I.Rashba (Phys.Rev.B 62 R16267 (2000)).Not so fast! E.I.Rashba (Phys.Rev.B 62 R16267 (2000)).““If you can adjust interface resistance by using a If you can adjust interface resistance by using a tunneling barrier, the situation can improved tunneling barrier, the situation can improved drastically!”.drastically!”.

Ferromagnetic contact/semiconductor interfacesHow good are they?

Rashba's prediction was confirmed

experimentally using

(a) Schottky barriersH.J.Zhu et al., PRL 87, 016601 (2001) (Fe/GaAs), 2% efficiency

A.T.Hanbicki et al, APL 80, 1240 (2002)

A.T.Hanbicki et al, APL 82, 4092 (2003), (Fe/AlGaAs), 33% efficiency

(b) Thin Metal Oxides V.F. Motsynyi et al, APL 81, 265 (2002)

T. Manago and H. Akinaga, APL 81, 694 (2002)

(c) AlAs barriers S.H.Chun et al, PRB 66, R100408 (2002).

Spin Relaxation Mechanisms

The Elliot-Yafet Scattering Mechanism

As a result of the SO-contribution to the crystalHamiltonian, conduction-band states of somesemiconductors are not spin eigenstates. This leads tothe possibility for spin-flip scattering even for spinindependent impurity scattering (due to Coulombicscattering for instance).

For the same reason, spin-independent electron-electronscattering can also cause spin-flip transitions

DYAKONOV-PERELSPIN RELAXATION IN A QUANTUM WIRE

• DRESSELHAUS HAMILTONIAN

• RASHBA HAMILTONIAN

y

z

x

x

RAMSAUER (FABRY-PERROT) RESONANCES

ONE REPULSIVE IMP. 300 angs from left contact

INFLUENCE OF SCATTERING STRENGTH

Magnetoresistive-Based BiosensorsD.L.Graham et al, Trends in biotechnology vol.22, 455 (2004)

Conclusions

Spintronics has already some success stories!

(giant magnetoresistance/spin valve)

Quantum Computing: Too early to tell!

Other potential: Spintronics & organics,

Spintronics & Biosensors, Magnetic Sensors.

Want to know more about it? Buy the book:

“Introduction to Spintronics”….in 2006.