Multilayered metallic nanostructures are promising for the fabrication of spin valves based on the giant magnetoresistive effect and for studies of the nature of topological magnetism, aimed at the development of new nanoscale data storage and transfer devices, e.g. those based on magnetic skyrmions. It is still an important task to develop methods of synthesis and configuration of thin-film nanostructures and control of spin textures in those nanostructures under electric and spin currents generated as a result of the spin Hall effect in external electric fields. Thin-film polycrystalline ferromagnetic / heavy metal Ru(10nm)/Co(0,8)/Ru(2), Ru(10)/Co(0,8)/Ru(2)/W(4), Pt(5)/Co(0,8)/MgO(2)/Pt(2) and Pt(15)/Co(0,8)/MgO(2)/Pt(2) nanostructures have been synthesized using magnetron sputtering. Electric contacts and Hall structures with different conductive bridge thicknesses have been synthesized on the specimens using electron beam photolithography. Experimental vibration magnetometric data have been utilized to calculate magnetic parameters of the specimens, i.e., saturation magnetization, magnetic anisotropy energy and field and coercive force as functions of ferromagnetic and heavy metal layer types. The domain structure of the specimens has been studied using Kerr microscopy. The electrical resistivity has been simulated and the critical current and current density of the nanostructures have been assessed. We show that all the film specimens exhibit perpendicular magnetic anisotropy and can be used in the studies of current-induced phenomena and spin moment transfer processes in nanostructures.

magnetron sputtering, nanostructures, photolithography, Kerr microscopy, spintronics, perpendicular magnetic anisotropy, metallic films

Multilayered films with ultrathin ferromagnetic metal layers exhibit unique magnetic and transport properties and have for a long time been attracting the attention of researchers due to the possibility of developing magnetic field sensing devices on their basis [1–3]. The synthesis of magnetic nanostructures requires understanding the processes occurring therein and the fundamental origins of their magnetic behavior, as well as high engineering capabilities and well-developed methods of synthesizing specimens with required functional properties. For example, varying the nanostructure growth process and geometry, number and thickness of layers and substrate parameters, one can control the magnetostatic energy and effective magnetic anisotropy of the growing nanostructures and change the orientation of their easy and hard magnetization axes [4–10], finally obtaining the required mciromagnetic configuration and general magnetic behavior. Study of the domain structure and remagnetization mechanisms in magnetic nanostructures is also of practical interest. For example, the effect of domain boundary movement in nanomaterials is widely used for the development of new types of non-volatile magnetic memory and logical devices [11, 12]. Studies at the nexus of topological magnetism and spintronics (spin-orbitronics and skyrmionics) open new technological opportunities for the development of UHF and THz ultrafast nanoscale data transfer and processing devices with controlled losses and noise [13–15]. Thus, a fundamental task of spintronics and its new branch spin-orbitronics is to study the nature of topological magnetism in collinear and non-collinear spin systems with strong spin-orbit coupling, over the discovery of magnetic skyrmions that can be stabilized in ferromagnetic media at room temperature. The most interesting results were obtained for heavy metal (HM = Pt, W, Ta etc.) / ferromagnetic (FM = Co, CoFe etc.) metallic nanostructures having perpendicular anisotropy and strong Dzyaloshinskii–Moriya interaction [16–21]. An important task of spintronics is to develop methods of controlling the spin textures in thin-film HM/FM nanostructures affected by polarized and pure spin currents caused by the spin Hall effect in the HM layer and by external magnetic fields (e.g. UHF) for the goal-oriented generation of individual static and dynamic magnetic skyrmions [19–24]. However, despite the large number of publications on the topic it is still not completely clear which materials or structures are optimum for the practical applications in this field. Furthermore, studies of the transport properties and spin moment transfer processes in metallic nanostructures require complex-geometry specimens with various types of current contacts, e.g. Hall bridges, thus imposing limitations on the research capabilities.

Presented below are our new synthesis method and experimental results for synthesized and calibrated thin-film magnetic HM (Ru, Pt, W) / FM (Co) Hall contact nanostructures. The saturation magnetization, perpendicular magnetic anisotropy, domain structure and critical currents of the films were assessed based on the experimental Kerr microscopy, magnetic and electrical data. We show that the specimens can be used for studying spin transport properties.

Polycrystalline films were synthesized on one side of polished thermally oxidized silicon substrates Si(675mm)/SiO₂(500nm) (Si-mat Co., Korea) in an Omicron ultrahigh-vacuum plant (Fig. 1a) equipped with four magnetron sputtering sources. The film compositions and layer thicknesses are presented below.

Specimen were placed on a rotating HM holder for the deposition of films having uniform and isotropic thicknesses and compositions. The base pressure in the chamber was 1.33·10^-6 Pa. Metallic layers (Pt, W, Co, Ru) were sputtered in direct current (DC) mode at an Ar gas pressure of 0.4 Pa and a 22 W power. The deposition rates were monitored for different layers with a quartz thickness gage: V_Pt = 0.05 nm/s, V_MgO = 0.02 nm/s, V_W = 0.015 nm/s and V_Ru = 0.018 nm/s. Then a photoresist layer was applied for further photolithographic stripping to form the protective template of a Hall contact structure (see below). The sputtering rate was calibrated with a NTEGRA Aura atomic force microscope (AFM). To this end, single-layered films of each material used with thicknesses of approx. 50 nm were synthesized. Some sections of the films were deposited on weakly adhesive domestic alcohol marker substrates and subsequently mechanically removed from the substrates so as to produce a step for further high-accuracy film thicknesses measurements. Furthermore, for each composition the mean-square surface roughness was assessed to be approx. 0.3 nm.

The structure of the specimens was studied using X-ray diffraction on a D8 Advance diffractometer in CuK_α radiation. The studies showed that all the experimental specimens were polycrystalline and contained small fractions of amorphous structure.

Measurements of current-induced spin transport required Hall bridge structures with variable width (with narrowed sections) be formed on the specimens. The 20.5 and 0.7 mm narrow sections were synthesized using electron beam lithography on the basis of a Scios 2Dual Beam electron microscope with a Raith attachment (10 kV accelerating voltage, 0.4 nA beam current and 30 nm beam diameter). The printing was accomplished at a 180^× magnification, i.e., the exposed field area was 1×1 mm², the exposure step being 16 nm. The photoresist was Micro Chem PMMAA2, the irradiation dose being 125 mC/cm². A photoresist film was applied at a 3000 rpm speed and baked at 453.15 K for 1 min. The photoresist film thickness was approx. 300 nm. Ion beam etching was conducted with an Oxford PlasmaLab 80Plus system. Unprocessed film sections were stripped off by plasma etching from the entire specimen surfaces except the areas protected by the photoresist template. The base pressure in the chamber was 1·10^-3 Pa, the Ar gas pressure was 0.7 Pa, the plasma generator power was 400 W and the etching time was 3 min.

Electric contact templates were printed on a Suss Microtec MJB4 contact photolithographic system (Fig. 1b). FP-2514 positive photoresist from FRAST-M Co. was used for exposure. The photoresist was applied at a 3000 rpm speed and baked at 368.15 K for 1 min. The photoresist film thickness was ~1.5 mm. Following photoresist stripping in acetone, Hall structures with 200 long and 20 mm wide conductors were formed on the specimen surface. Electric contacts were formed by coating the specimens with another photoresist layer in which electric contact pads were formed by photolithography. The surface adhesion of the electric microcontacts was improved by applying a Ti/Au layer in a custom-made thermal deposition plant. A current of ~30 A was sequentially passed through heating coils inside which a titanium and gold weighed dose was placed. Titanium was initially deposited from solid state for 20 s to form a 5–15 nm thick film for gold adhesion improvement. The Au weighed dose was sputtered completely: sputtering of a 70 mg gold dose produced a ~60 nm thick film on the specimen surface. The sputtering chamber had an exhaust cart with a turbomolecular vacuum pump developing a pressure of ~0,1·10^-3 Pa.

As a result, single-layered films and Hall-shaped contact film structures were synthesized. The final film structure specimens formed on a silicon chip (substrate) had 4×4 mm² dimensions and were then separated using ultrasonic welding with ~20 mm diam. aluminum wires to form a wedge-to-wedge arrangement. The welded contact diameter was ~60 mm. Thus, each chip specimen contained nine Hall-type structures having different conductive contact thicknesses (Type 1: 20 mm, Type 2: 5 mm and Type 3: 2 mm, see Fig. 2). Each Hall-type structure had six contacts: two at the edges of the current conductor and four transverse contacts for measuring the longitudinal and lateral potential drops under bias. Thin copper wires were soldered to the chip leads for easy connection.

Figure 1.

Nanostructure synthesis equipment: (a) Omicron ultrahigh-vacuum plant with four magnetron sputtering sources; (b) Suss Microtec MJB4 contact photolithography system

Figure 2.

Schematic of final on-chip specimen with nine Hall contact film structures

The magnetic properties (magnetization state, coercive force, magnetic anisotropy energy etc.) were studied for continuous films using vibration magnetometry (Lake Shore 7401VSM) in up to 3 T fields at 78–800 K. The magnetometer holder allowed measuring the specimen orientation relative to the magnetic field direction. Hysteresis loops were taken for each film in the easy magnetization axis (EMA) direction which for all the test specimens was perpendicular to the film plane, and along the hard magnetization direction, i.e., within the film plane. The former curves were used for calculating the saturation magnetization and the coercive force, and the latter ones, for determining the magnetic anisotropy energy.

Figure 3a shows a typical magnetic hysteresis loop for a continuous film with field applied along the easy magnetization axis, i.e., perpendicularly to the film. The loop exhibits four typical points: the positive and negative magnetic inductions B_с at which the magnetization is zero and the positive and negative magnetizations M. The mean magnetic induction В_с characterizes the coercive force and the mean magnetic moment M₀ shows the saturation magnetization. The saturation magnetization M_s, i.e., the specific magnetic moment per unit volume, is calculated based on the accurate measurement of the film surface area S on a specimen photograph superimposed with millimeter grid paper. Knowing the magnetic layer thickness t_F one can calculate the saturation magnetization M_s using the following formula:

$M_{\mathrm{s}}=\frac{M_0\bullet {10}^{-3}}{S{t}_F}$(1)

Figure 3b shows a typical magnetization curve for a continuous film measured perpendicular to the EMA, i.e., within the film plane. Since the magnetic moment can be measured with a higher accuracy upon remagnetization along the EMA, the curve was reduced to the units of M_n, and the saturation magnetization M_s for the calculation was taken from the previous calculation. The area between the coordination axis and the magnetization curve is proportional to the magnetic anisotropy energy which can be thus calculated using the following formula [25]:

$K_u=M_{\mathrm{s}}{\int }_0^1{B}_x\mathrm{d}{M}_n$(2)

The field at which the magnetization is saturated in the plane was accepted as the anisotropy field B_a. The latter parameter was subsequently used for analysis of magnetic anisotropy in the film.

The EMA direction for all the test specimens was perpendicular to the film plane, i.e., the specimens had perpendicular magnetic anisotropy which meets the requirements. The hysteresis loop width, the coercive force and the anisotropy and magnetization fields are controlled by the parameters of the structures, primarily, by the thicknesses and combination of the layers. The minimum hysteresis loop width and hence coercive force were obtained for the Ru/Co specimens (Specimens 1 and 2, see Table 1). For the Pt/Co structures (Specimens 3 and 4) the above parameters were almost one order of magnitude higher, the hysteresis loop being almost rectangular. Additional deposition of a top nonmagnetic metallic W layer onto the Ru/Co film did not affect the saturation magnetization and the coercive force of Specimen 2 but substantially reduced the anisotropy field and magnetic anisotropy energy (Fig. 4).

The remagnetization processes were visualized and the domain sizes in the continuous film specimens and the contact structure specimens were assessed from Kerr microscopy data (Evico Magnetics). The microscope allowed visualization of magnetization switching in the test objects (domains) sized up to 800 nm. The Kerr microscope specimen holder had two electric magnets for simultaneous generation of two magnetic fields, i.e., within (±400 mT) and perpendicularly (±120 mT) to the film plane. Figure 5 illustrates the magnetic structure of the film in a demagnetized state and the Hall structure-shaped films with and without an external magnetic field.

The specific electrical resistivity of the test contact nanostructures was ~10^-7 Ohm·m (Table 1). The critical current in the structures was 0.5 to 60 mA. The differences in the electrical resistivities and in the critical currents (above which the Hall bridge is destroyed) for the Hall contact films originates, along with the different compositions and structure configuration, from the fact that currents in those systems pass not only through the HM layer but also through the FM one and other auxiliary films. One should therefore take into account shunting in the design of the test structures for studying various current-induced processes. Nevertheless, simple Kirchhoff rule assessments yield a maximum current density of ~10¹¹ A/m² for the narrow structure section. This suggests that the required magnitude of current-induced effects (Oersted fields) on the magnetic structure of ferromagnetic materials is achievable [26, 27] and one can therefore use the test specimens for both the spintronics and spin-orbitronics applications.

Figure 3.

(a) Magnetic hysteresis loop for continuous film with magnetic field perpendicular to the film plane and (b) half of hysteresis loop for continuous film with magnetic field within the film plane

Figure 4.

Magnetic hysteresis loop for magnetic field orientation (a–d) perpendicular and (e–h) along the film plane: (a, e) Ru/Co/Ru; (b, f) Ru/Co/Ru/W; (c, g) Pt(5)/Co/MgO/Pt; (d, h) Pt(15)/Co/MgO/Pt

Figure 5.

Magnetooptic Kerr microscopy of test films: (a, b) Ru/Co/Ru; (c, d) Ru/Co/Ru/W (the typical scale of the structure in the drawings is indicated by red numbers); (e, f) Pt(5)/Co/MgO/Pt; (g, h) Pt(15)/Co/MgO/Pt. (a, c, e, g) continuous films in demagnetized state; (b, d, f, h) Hall-shaped films with and without field

Multilayered polycrystalline HM/FM Ru(10)/Co(0,8)/Ru(2), Ru(10)/Co(0,8)/Ru(2)/W(4), Pt(5)/Co(0,8)/MgO(2)/Pt(2) and Pt(15)/Co(0,8)/MgO(2)/Pt(2) metallic structures were synthesized using magnetron sputtering. Analysis of the magnetic properties of the structures showed that the specimens exhibit perpendicular magnetic anisotropy and low-field domain structure switching. Different configurations of Hall contact structures were synthesized using lithographic techniques with Hall bridge thicknesses of 0.7–2 mm. The critical current and current density of the structures were assessed. The synthesized specimens can be used for studies of current-induced effects on the magnetic structure of ferromagnetic materials and the formation of different topological spin textures, as well as for the electric control of spins in HM/FM structures. Results of such experiments are indispensable for the improvement of strong spin-orbital interaction structure technologies.

The work was supported by the Russian Science Foundation grant No. 21-72-20160 (https://rscf.ru/en/project/21-72-20160). The Authors are also grateful to the Joint Use Center of the Far East Federal University.