240 likes | 363 Views
Spin Electronics. Peng Xiong. Department of Physics and MARTECH Florida State University. QuarkNet, June 28, 2002. SOURCE. GATE. DRAIN. MOSFET. Moore’s Law… is the end in sight?. Speed: 10 0 Hz Size: 10 -2 m Cost: $10 6 /transistor. Speed: 10 9 Hz Size: 10 -7 m
E N D
Spin Electronics Peng Xiong Department of Physics and MARTECH Florida State University QuarkNet, June 28, 2002
SOURCE GATE DRAIN MOSFET Moore’s Law… is the end in sight? Speed: 100 Hz Size: 10-2 m Cost: $106/transistor • Speed: 109 Hz • Size: 10-7 m • Cost: $10-5/transistor
Magnetic Information Storage: superparamagnetic limit • Density: 20 Gb/in2 • Speed: 200 Mb/s • Size: f2.5” x 2 • Capacity: 50 Gb • Density: 2 kb/in2 • Speed: 70 kb/s • Size: f24” x 50 • Capacity: 5 Mb
Superparamagnetic Limit: thermal stability of magnetic media
Semiconductor Random Access Memory: alternatives? M O S • High speed • Low density • High power consumption • Volatile
H R H E E E E H M EF EF N(E) N(E) Metal-based Spintronics: Spin valve and magnetic tunnel junction Applications: magnetic sensors, MRAM, NV-logic
GATE Spintronics in Semiconductor: spin transistor • Dreams • High performance • opto-electronics • Single-chip computer • (instant on; low power) • Quantum computation Datta and Das, APL, 1990 H SOURCE DRAIN GaAs • Issues • Spin polarized material • Spin injection • Spin coherence • Spin detection H
Solutions: • Use injector with 100% • spin polarization • Non-diffusive injection • Conductivity matching Spin Injection: the conductivity mismatch I Schmidt et.al., PRB, 2000 I RN RF I¯ SC mF RN¯ RF¯ mN mF¯ mN¯ FM
E E CrO2: a half metal Tc = 400 K m = 2mB/Cr p = 100% Uex E 4s Schwarz, J. Phys. F, 1986 normal metal half-metallic ferromagnet 3d metallic ferromagnet Measurement of spin polarization: using a superconductor
Question: • What could happen to an electron with energy eV < D when it hits S from N? • bounce back; • go into S as an electron; • C. go into S in a Cooper pair. • A and B • B and C • C and A • A and B and C Andreev reflection: normal metal/superconductor E S N D eV EF -D N(E) N S
Andreev reflection: normal metal/superconductor p = 0 Z = 0 clean metallic contact Z ~ 1 in-between Z >> 1 tunnel junction Blonder, Tinkham, and Klapwijk, PRB, 1982
Andreev reflection: ferromagnet/superconductor p = 75% E F S Z = 0 metallic contact D eV EF -D Z ~ 1 in-between DOS Z >> 1 tunnel junction V
Comparison: normal metal and ferromagnet p = 75% p = 0 Z = 0 metallic contact Z = 0 metallic contact Z ~ 1 in-between Z ~ 1 in-between Z >> 1 tunnel junction Z >> 1 tunnel junction V V
Spin Polarization of CrO2: our approach Planar junction real device structure Artificial barrier controlled interface Preservation of spin polarization at and across barrier Key step: controlled surface modification of CrO2 via Br etch
CrO2 Film Growth: Chemical Vapor Deposition Furnace, T=280° C O2 flow Heater block, T=400°C substrate Cr8O21 precursor Ivanov, Watts, and Lind, JAP, 2001
~ V Lock-in dV/dI vs V in He4 (1K) or He3 (0.3K) cryostats Junction Fabrication and Measurement • Grow CrO2 film • Pattern CrO2 stripe • Surface modification: Br etch • Deposit S cross stripes Pb or Al Pb or Al I CrO2 CrO2 TiO2
Results: CrO2/(I)/Pb junctions Metallic contact Z = 0 p = 97% T = 1.2 K • = 1.44 meV Tunnel junction T = 400 mK High quality barrier w/o inelastic scattering
mH H Measurement of spin polarization in high-Z junctions: using Zeeman splitting E D eV EF -D eV/D N(E) Meservey and Tedrow, Phys. Rep., 1994 S F
Zeeman splitting in an F/I/S junction CrO2 In order to get high Hc: Ultrathin S film Parallel field Negligible s-o interaction H Al Al CrO2
Results: Zeeman splitting +2.5T -2.5T T =400 mK
Summary (CrO2) Verified half-metallicity of CrO2 Engineered an artificial barrier on CrO2 surface Preserved complete spin polarization at interface Achieved full spin injection from a half metal Future Apply the technique to other systems Magnetic tunnel junction
CrO2/I/Co magnetic “tunnel” junction H Co CrO2 AlOx
The People Jeff Parker Jazcek Braden Steve Watts Pavel Ivanov Stephan von Molnár Pedro Schlottmann David Lind
Let’s build “computers with wires no wider than 100 atoms, a microscope that could view individual atoms, machines that could manipulate atoms 1 by 1, and circuits involving quantized energy levels or the interactions of quantized spins.” Richard Feynman – “There’s Plenty of Room at the Bottom” 1959 APS Meeting