Why are STOs of interest? The increasing number of wireless microwave applications that accompany our everyday life (GPS, mobile phones etc.) pushes suppliers to implement more and more frequency standards on a single device. This requires new concepts to cover a large frequency range and to achieve at the same time miniaturization while keeping costs low. With current technology, based on LC-tanks, miniaturization is difficult to achieve and defining alternative technologies is an active research field. Here spintronics based RF oscillators may provide solutions due to their extremely small dimensions of the magneto-resistive devices (100 nm) that will make it possible to replace one conventional Voltage Controlled Oscillator (VCO) by about several tens of spin torque oscillators (STOs). In this view SPINTEC in collaboration with LETI/DIHS and LETI/DACLE are studying such oscillators addressing basic issues such as the magnetization dynamics under spin transfer torque as well as the optimization of the corresponding microwave performances (power, linewidth etc) to achieve the required specifications of given microwave components and systems. 

Issues to address for STOs: One of the major advantages of STOs is their nanoscale size, so miniaturization is inherent to these devices. The other advantage is the possibility to tune the frequency by the injected spin polarized currents, with tuning ranges of 100% around the carrier (see below,What is an STO?).

However, to achieve compatibility with existing microwave circuits, a number of important questions need to be solved.

These are:

- operation in zero external field
- maximizing the oscillation amplitude
- enhancing the output power
- understanding the origin of the linewidth broadening and the associated phase noise and non-linear - magnetization dynamics properties.

Here SPINTEC, has made important contributions using respectively aperpendicular polarizer, magnetic tunnel junctions and by developing new time domain analysis techniques. The general concept and efforts of our research activity are summarized in the poster. For a short walk through this fascinating topic of spin torque oscillators check out our results (click to open):



PDF - 1 Mb



  • How to maximize the precession amplitude using a perpendicular polarizer ?
  • How to increase the output power : Spin valves vs. tunnel junctions !
  • How to increase the output power : High Impedance Amplification ?
  • How to reduce the linewidth : Synthetic antiferromagnets !
  • Origin of linewidth broadening - Frequency fluctuations !
  • Origin of linewidth broadening - Phase and amplitude noise !
  • How to reduce linewidth : Mode coupling !

What is an STO?


The operation of a spin transfer torque oscillator can be explained in analogy to a Voltage Controlled Oscillator (VCO) used in current microwave technologies, see Fig. 2.

JPEG - 32.9 kb
Figure 2

A VCO contains three major elements :


  • a resonator (LC circuit)
  • an ohmic resistance, meaning attenuation or energy losses
  • an external amplification, meaning energy feedback, also called negative resistance


The resonator in the case of STOs is the magnetic system as such. The magnetization, which is given by a spin angular momentum, responds to a perturbation with an oscillating motion of the magnetization vector around its equilibrium. This oscillation would keep on forever, in the absence of any energy losses. However, the coupling of the magnetization to its surrounding (electrons, lattice…) leads to adamped motion and thus to a gradual decrease of the precession amplitude (characterized by the damping constant α). In order to keep up the precessional motion, energy has to be fed back into the system. In the case of the STO this feedback is provided by the spin transfer torque. In this way very large amplitude magnetization oscillationscan be achieved that are unattainable using conventional means such as microwave pumping fields. The resulting precession trajectory (or voltage amplitude in the case of a VCO) is given by the balance between the spin transfer (energy feedback) and the damping torque (energy loss) and its stability is determined by the non-linear1contributions of these two terms.

The major difference between a VCO and a STO is the non-linear dependence of the precession frequency (resonator) on the precession amplitude, which is very weak for a VCO but very strong for the STO. This is why we call an STO a non linear1 oscillator. This non-linear dependence is quite unique and makes STOs of interest from a fundamental and technological point of view. In particular it provides the possibility to tune the precession frequency or to lock an oscillator to an external signal source, but it also plays a crucial role for linewidth broadening.

1non-linear = strong dependence of a given term on precession amplitude

JPEG - 26.2 kb
Figure 3

For further reading check out the following references:

[Slonczewski] J. Magn. Magn. Mater. 159, L1–L7 (1996), J. C. Slonczewski, et al;

[Slonczewski] J. Magn. Magn. Mater. 195, L261–L268 (1999), J. C. Slonczewski, et al;

[Berger] Phys. Rev. B 54, 9353–9358 (1996), L. Berger et al;

[Kim] Phys. Rev. Lett. 100, 017207 (2008), J.-V. Kim et al;

[Slavin] IEEE Trans. Magn. 45, 1875 (2009), A. N. Slavin et al;