4
P. K. Muduli et al.
Fig. 1 a Schematic of the operation of a spin-torque nano-oscillator in a system consisting of a
ferromagnet/spacer/ferromagnet trilayer. b Schematic presentation of damping torque, in-plane spin
torque, T IP , and out-of-plane torque, T OPP on the local magnetization. c Schematic presentation of
resistance oscillations arising due to the spin-torque-induced magnetization precession, as a result
of the GMR or TMR effect
1 Introduction
The theoretical prediction of the spin-transfer torque (STT) effect in the year 1996
constituted a remarkable breakthrough in the development of spintronics. This prediction made independently by Slonczewski [1] and Berger [2] after the 1988 discovery
of giant magnetoresistance (GMR) [3, 4] and the first report of high tunneling magnetoresistance (TMR) in room-temperature magnetic tunnel junctions (MTJ) in 1995 [5,
6]. STT made possible new methods of manipulating magnetization dynamics using
spin-polarized currents instead of magnetic fields, opening several opportunities for
the development of novel spintronics devices. The two most important types of STT
devices are (i) magnetic memories, such as magnetoresistive random-access memories (MRAM) [7, 8], and (ii) nanoscale microwave signal generators/detectors. In
this chapter, we will focus on the second application of STT. The reader may also
refer to reviews of STT-based MRAM and STNOs in Ref. [9–15].
The basic principle of operation of an STNO is illustrated in Fig. 1. When electrons
pass through the thick fixed magnetic layer, their spins become aligned with the
magnetization of this layer. When these spin-polarized electrons enter the free layer,
after passing through the non-magnetic/insulating barrier, they exert a torque on
the local magnetization of the free layer if current is sufficiently high, i.e. more
than some threshold value. It can lead to magnetization switching if the torque is
sufficiently large. Smaller torque values result in magnetization precession around
the effective magnetic field. In the general case, an STNO consists of a nanopillar
with a giant magnetoresistance (GMR) or a magnetic tunnel junction (MTJ) structure.
The oscillation of the magnetization is detected as a microwave or radio frequency
(RF) voltage signal by virtue of either the GMR or TMR effect (Fig. 1c).
STNOs are particularly useful in wireless telecommunications, where they offer
several advantages over commercially used oscillators, such as yttrium iron garnet (YIG) tuned oscillators (YTOs), dielectric resonator oscillators (DROs) based
on YIG, and semiconductor-based voltage-controlled oscillators (VCOs). These
advantages include a very wide frequency tuning range [16–19], high modulation
rates [20–26], submicron footprints [27], and compatibility with standard complementary metal oxide semiconductor processes [7, 8]. However, low output power and
P. K. Muduli et al.
Fig. 1 a Schematic of the operation of a spin-torque nano-oscillator in a system consisting of a
ferromagnet/spacer/ferromagnet trilayer. b Schematic presentation of damping torque, in-plane spin
torque, T IP , and out-of-plane torque, T OPP on the local magnetization. c Schematic presentation of
resistance oscillations arising due to the spin-torque-induced magnetization precession, as a result
of the GMR or TMR effect
1 Introduction
The theoretical prediction of the spin-transfer torque (STT) effect in the year 1996
constituted a remarkable breakthrough in the development of spintronics. This prediction made independently by Slonczewski [1] and Berger [2] after the 1988 discovery
of giant magnetoresistance (GMR) [3, 4] and the first report of high tunneling magnetoresistance (TMR) in room-temperature magnetic tunnel junctions (MTJ) in 1995 [5,
6]. STT made possible new methods of manipulating magnetization dynamics using
spin-polarized currents instead of magnetic fields, opening several opportunities for
the development of novel spintronics devices. The two most important types of STT
devices are (i) magnetic memories, such as magnetoresistive random-access memories (MRAM) [7, 8], and (ii) nanoscale microwave signal generators/detectors. In
this chapter, we will focus on the second application of STT. The reader may also
refer to reviews of STT-based MRAM and STNOs in Ref. [9–15].
The basic principle of operation of an STNO is illustrated in Fig. 1. When electrons
pass through the thick fixed magnetic layer, their spins become aligned with the
magnetization of this layer. When these spin-polarized electrons enter the free layer,
after passing through the non-magnetic/insulating barrier, they exert a torque on
the local magnetization of the free layer if current is sufficiently high, i.e. more
than some threshold value. It can lead to magnetization switching if the torque is
sufficiently large. Smaller torque values result in magnetization precession around
the effective magnetic field. In the general case, an STNO consists of a nanopillar
with a giant magnetoresistance (GMR) or a magnetic tunnel junction (MTJ) structure.
The oscillation of the magnetization is detected as a microwave or radio frequency
(RF) voltage signal by virtue of either the GMR or TMR effect (Fig. 1c).
STNOs are particularly useful in wireless telecommunications, where they offer
several advantages over commercially used oscillators, such as yttrium iron garnet (YIG) tuned oscillators (YTOs), dielectric resonator oscillators (DROs) based
on YIG, and semiconductor-based voltage-controlled oscillators (VCOs). These
advantages include a very wide frequency tuning range [16–19], high modulation
rates [20–26], submicron footprints [27], and compatibility with standard complementary metal oxide semiconductor processes [7, 8]. However, low output power and