Magnetization reversal in YIG/GGG(111) nanoheterostructures grown by laser molecular beam epitaxy

Abstract

Thin (4–20 nm) yttrium iron garnet (Y3Fe5O12, YIG) layers have been grown on gadolinium gallium garnet (Gd3Ga5O12, GGG) 111-oriented substrates by laser molecular beam epitaxy in 700–1000 °C growth temperature range. The layers were found to have atomically flat step-and-terrace surface morphology with step height of 1.8 Å characteristic for YIG(111) surface. As the growth temperature is increased from 700 to 1000 °C the terraces become wider and the growth gradually changes from layer by layer to step-flow regime. Crystal structure studied by electron and X-ray diffraction showed that YIG lattice is co-oriented and laterally pseudomorphic to GGG with small rhombohedral distortion present perpendicular to the surface. Measurements of magnetic moment, magneto-optical polar and longitudinal Kerr effect (MOKE), and X-ray magnetic circular dichroism (XMCD) were used for study of magnetization reversal for different orientations of magnetic field. These methods and ferromagnetic resonance studies have shown that in zero magnetic field magnetization lies in the film plane due to both shape and induced anisotropies. Vectorial MOKE studies have revealed the presence of an in-plane easy magnetization axis. In-plane magnetization reversal was shown to occur through combination of reversible rotation and abrupt irreversible magnetization jump, the latter caused by domain wall nucleation and propagation. The field at which the flip takes place depends on the angle between the applied magnetic field and the easy magnetization axis and can be described by the modified Stoner–Wohlfarth model taking into account magnetic field dependence of the domain wall energy. Magnetization curves of individual tetrahedral and octahedral magnetic Fe3+ sublattices were studied by XMCD.

Direct detection of induced magnetic moment and efficient spin-to-charge conversion in graphene/ferromagnetic structures

“…The bottom of the figures present the XMCD asymmetries, which were evaluated by taking the averaged difference between the difference of the spectra taken with two different polarizations under 300 Oe (shown in the figure) and the difference of the spectra taken under -300 Oe (not shown), as explained in the methods section. The Fe L 3,2 XAS/XMCD spectrum is in agreement with published spectroscopic data of iron(III) coordinated as it is in YIG 66 . The most important result of the XAS experiments is the clear observation of a dichroic signal at the C K absorption edge.…”

Role of gallium diffusion in the formation of a magnetically dead layer at the Y3Fe5O12/Gd3Ga5Korovin1, Bursian2, Lutsev3 et al. 2018Phys. Rev. Materials442130View full textAdd to dashboardCite

“…The epitaxial YIG layers were grown at 700-1000°C by laser molecular beam epitaxy onto annealed GGG (111) substrates, following the approach addressed in our earlier works [13,14]. As described therein, growth results in high-quality YIG films with sharp x-ray diffraction Bragg peaks, highcontrast Laue oscillations, smooth atomically flat surface, ultranarrow magnetization loops, and a low spin-wave damping coefficient [6].…”

“…[6] that SW damping in a 10 nm epitaxial YIG layer can be as low as α = 3.6 × 10 −5 , approaching the bulk value obtained for YIG single crystals grown by the Czochralski method. Various deposition techniques including laser molecular beam epitaxy (LMBE) have been used [6][7][8][9][10][11][12][13][14] in recent years to grow high-quality YIG films onto gadolinium gallium garnet (Gd 3 Ga 5 O 12 , GGG) substrates. Despite the fact that GGG is very well lattice matched to YIG ( a/a = 6 × 10 −4 ), it was claimed in a number of studies that the crystal structure and magnetic properties of YIG nanolayers can be quite different * Correspondence: suturin@mail.ioffe.ru from the bulk.…”

“…Despite the fact that GGG is very well lattice matched to YIG ( a/a = 6 × 10 −4 ), it was claimed in a number of studies that the crystal structure and magnetic properties of YIG nanolayers can be quite different * Correspondence: suturin@mail.ioffe.ru from the bulk. Particularly, the (111) interlayer spacing in films is often significantly larger (by 1-1.5%) than in bulk YIG due to rhombohedral distortions [13,14]. This can be caused by stoichiometry deviations due to oxygen and iron vacancies, gallium or gadolinium diffusion from the substrate, etc.…”

Electrical generation and detection of spin waves in polycrystalline YIG/Pt grown on silicon wafers

“…Additionally, as a ferrimagnetic insulator, YIG is an ideal material of low-power consumption to generate a pure spin current in magnon-based devices without introducing extra charge current. At present, high-quality single-crystal YIG films (thickness<100 nm) have been successfully achieved on a costly single crystal substrate, e.g., gadolinium gallium garnet (GGG) via epitaxial growth techniques such as liquid phase epitaxy (LPE), molecular beam epitaxy (MBE), pulsed laser deposition (PLD) and magnetron sputtering [3][4][5][6], which have been intensively studied in various spintronic researches, such as the spin pumping [7][8][9][10][11][12], the spin Hall magnetoresistance (SMR) [13][14][15][16], and the electrical excitation and detection of spin waves [1,17].…”