Main Logo Main Logo Main Logo
 
HomeResearch Magnetic Semiconductors

Magnetic Semiconductors

Magnetic semiconductors are materials that exhibit both semiconducting and magnetic response. These materials are of great interest in the context of spintronics, as they allow the control of charge and spin current. We can classify the magnetic semiconductors according to their structure, composition and arrangement of the magnetic ions in the semiconducting host as: dilute magnetic semiconductors (DMS) and condensed magnetic semiconductors (CMS).

Dilute Magnetic Semiconductors (DMS)

In DMS, a fraction of the host atoms (or cations in the case of binary semiconductors) is randomly replaced by magnetic ions, like transition metals (TM) or rare earths (RE). In the presence of delocalized carriers, these substitutional magnetic impurities are magnetically coupled resulting in a macroscopic ferromagnetic order. Up to date an impressive variety of combinations of semiconductors and magnetic elements have been studied [1], however up to date a DMS operating at room temperature is still missing.

Condensed Magnetic Semiconductors (CMS)

The thermodynamical miscibility of many transition metals in semiconductors is in fact extremely low and even under non-equilibrium growth techniques the magnetic dopants tend to aggregate in: (i) embedded nanocrystals with a crystalline structure other than the host, and (ii) into volumes preserving the crystal structure of the host, but rich in the magnetic element [2].

Challenges and main key points in magnetic semiconductor study

  • control on the reproducibility of the epitaxial growth processes
  • accurate characterization of the material’s structural properties, and chemical composition, down to atomic scale
  • control on the incorporation of magnetic elements: dilution/segregation/precipitation
  • correlation of the structural, electrical, chemical properties of the system with the magnetic response from the different magnetic phases in the case of both:
  • study of the spin phenomena: spin polarization, spin transfer, etc
  • implementation into spintronic devices

Nitride based Magnetic Semiconductors

(Ga,Fe)N

Under our growth conditions, (Ga,Fe)N is a DMS below 0.4% of Fe and a CMS above this concentration. In the DMS regime, this material system has proven to be a model system for magneto-optical measurements, where the three excitons of GaN and a huge Zeeman splitting can be observed [3]. As a CMS, we have found ε-Fe3N ferromagnetic (FM) nanocrystals which give a layer TC of up to 573 K [4-6]. Temperature dependent studies have allowed us to identify - as function of the fabrication parameters - the onset of other Fe-rich phases (γ-Fe4N, ζ-Fe2N, α-Fe, γ-Fe), each with a characteristic magnetic response [7]. Moreover, by doping (Ga,Fe)N with donors (Si) we reduce the charge state from Fe3+ to Fe2+, affecting in this way the interaction between ions and quenching the tendency of the Fe-ions to aggregate [6,8].

Recently, we have achieved well-aligned 2D nanocrystal (NC) layers embedded in GaN [9], which open the perspectives to implement these structures into spintronic devices. By varying the fabrication conditions, the NCs composition and therefore, the magnetic properties can be tuned from FM – Fe3N and Fe4N- to antiferromagnetic (AF) – GaFe3N. This makes this CMS suitable for FM and AF spintronics [10].

TEM and XRD of the tunable GaxFe4-xN nanocrystals in GaN
Figure 1: TEM and XRD of the tunable GaxFe4-xN nanocrystals in GaN [9].

Our current goal is to optimize the 2D NCs layers in (Ga,Fe)N to obtain single-phase Fe-rich nanocrystals on demand and to understand their magneto-transport and magneto-optical properties. Moreover, we want to explore the spin phenomena that occur in these structures for their implementation into FM and AF spintronic devices.

(Ga,Mn)N

Our (Ga,Mn)N with a concentration of Mn ions below 1% is paramagnetic, and can be described as an ensemble of non-interacting Mn3+ ions substituting Ga[12]. Recently, we have proved that - as expected from many theoretical works, the interaction between Mn pairs also in samples with a Mn concentration up to 3,3% is definitely FM, but short-ranged, with no ferromagnetic percolation above 2 K[13]. Similar as for (Ga,Fe)N, by codoping (Ga,Mn)N with Si, the charge state of Mn is reduced from Mn3+ to Mn2+ leading to an AF interaction between Mn atoms and reducing the magnetic anisotropy observed in (Ga,Mn)N.

Recently, we have shown that the magnetic properties of (Ga,Mn)N can be modified by applying an electric field. The changes in magnetization generated by the electric field in the films are caused by the inverse piezoelectric effect that alters the out-of-plane lattice parameter c modifying the trigonal distortion and, thus, the uniaxial magnetic anisotropy [14].

Modification of magnetic anisotropy in (Ga,Mn)N with an external electric field
Figure 2: Modification of magnetic anisotropy in (Ga,Mn)N with an external electric field [14].

(Ga,Mn)N:Mg

According to the predictions of T. Dietl [15], GaN and ZnO with a sufficient concentration of both dilute magnetic ions and free holes mediating the magnetic interaction should have a Curie temperature Tc far above room temperature. In order to add holes to (Ga,Mn)N, we cooping with Mg was done. Instead of an increase in the TC, the formation of magnetic complexes, which composition and spin can be tuned entirely by the fabrication conditions, were achieved [16]. The Mn-Mgk complexes reduce strongly the observed magnetic anisotropy of (Ga,Mn)N and show striking optical properties with an emission at technological important IR region of 1.2 µm persisting up to RT. These remarkable characteristics open a variety of applications ranging from single photon generation to the possibility of achieving a GaN IR-emitting laser [17].

Our current research is devoted to the fabrication of distributed Bragg reflectors containing a (Ga,Mn)N:Mg cavity with a stop band at 1.2 µm [18-19].

Magnetic Mn-Mgk complexes in GaN with tunable spin and IR-emission
Figure 3: Magnetic Mn-Mgk complexes in GaN with tunable spin and IR-emission [16].

References

  1. T. Dietl and H. Ohno. Rev. Mod. Phys. 86 (2014)
  2. T. Dietl et al. Rev. Mod. Phys. 87(2015)
  3. W. Pacuski et al. PRL 100 (2008)
  4. A. Bonanni et al. PRB 75 (2007)
  5. T. Li et al. J. Of Cryst.Growth 310 (2008)
  6. A. Bonanni et al. PRL 101 (2008)
  7. A. Navarro-Quezada et al. PRB 81 (2010)
  8. M. Rovezzi et al. PRB 79 (2009)
  9. A. Navarro-Quezada et al. APL 110 (2012)
  10. T. Jungwirth et al. Nature Nanotechnology 11 (2016)
  11. T. Jungwirth et al., Rev. Mod. Phys. 78, 809 (2006)
  12. W. Stefanowicz et al. PRB 81 (2010)
  13. A. Bonanni et al. Phys. Rev. B 84 (2011)
  14. D. Stzenkiel et al. Nature Comm. 7 (2016)
  15. T. Dietl, Science 287 (2000)
  16. T. Devillers et al. Sci. Reports (2012)
  17. Austrian patent No. 512938
  18. G. Capuzzo et al. Sci. Reports (2017)
  19. D. Kysylychyn, Master Thesis (2015)
JKU Institute of Semiconductor and Solid State Physics, Altenbergerstr. 69, 4040 Linz, Austria, Tel. +43 732 2468 9639, Fax +43 732 2468 8650