Significant suppression of two-magnon scattering in ultrathin Co by controlling the surface magnetic anisotropy at the Co/nonmagnet interfaces

To enable suppression of two-magnon scattering (TMS) in nanometer-thick Co (ultrathin Co) layers and realize low-magnon damping in such layers, the magnon damping in ultrathin Co layers grown on various nonmagnetic seed layers with different surface magnetic anisotropy (SMA) energies are investigated. We verify the significantly weak magnon damping realized by varying the seeding layer species used. Although TMS is enhanced in ultrathin Co on Cu and Al seeding layers, the insertion of a Ti seeding layer below the ultrathin Co greatly suppresses the TMS, which is attributed to suppression of the SMA at the interface between Co and Ti. The Gilbert damping constant of the ultrathin Co layer on Ti (3 nm), 0.020, is comparable to the value for bulk Co, although the Co layer thickness here is only 2 nm. Realization of such weak magnon damping can open the door to tunable magnon excitation, thus enabling coupling of magnons with other quanta such as photons, given that the magnetization of ultrathin ferromagnets can be tuned using an external electric field.

Figure 1

Schematics of the sample structures shown in Table 1. The Co film thickness ($t_{\mathrm{Co}}$) was varied from 2 to 20 nm. The samples were capped with a 2-nm-thick MgO layer and a 10-nm-thick ${\mathrm{SiO}}_2$ layer for samples 1, 2, 3, and 4.

Figure 2

(a) Schematic of the experimental setting used in the CPW-FMR measurements. The external magnetic field was applied along the in-plane direction of the samples. (b) Frequency dependence of the FMR signals of Co (20 nm)/Cu, where the frequency was varied from 5 to 15 GHz. (c) Frequency dependence of the HWHM in the Co (10 nm)/${\mathrm{SiO}}_2$, Al, Cu, and Ti sample. (d) Co thickness dependence of the Gilbert damping constant α for the samples with the various NM seeding layers and the ${\mathrm{SiO}}_2$ underlayer (i.e., the reference sample).

Figure 3

(a) Schematic of the measurement setup used to determine the angular dependence of the FMR signals when using the ESR apparatus. (b) Angular dependence of the FMR signals of the Co (10 nm)/${\mathrm{SiO}}_2$ sample. (c) Angular dependence of the resonance fields of the Co (3, 5, and 10 nm)/${\mathrm{SiO}}_2$ samples. The broken red line represents the fitting line. (d) Co film thickness dependence of the uniaxial magnetic anisotropy field $h_{\mathrm{u}}$, where closed lines represent the results obtained using the fitting function $h_{\mathrm{u}}=h_{{\mathrm{V}}_{\mathrm{u}}}+h_{{\mathrm{S}}_{\mathrm{u}}}{t}_{\mathrm{FM}}^{-1}$.

Figure 4

(a) Correlation diagram between the TMS coefficient ${\beta }_{\mathrm{TMS}}$ and the square of the SMA field ${\left(h_{{\mathrm{S}}_{\mathrm{u}}}\right)}^2$. The broken line indicates the fitting result for the samples with the NM seeding layers. In the correlation between ${\beta }_{\mathrm{TMS}}$ and ${\left(h_{{\mathrm{S}}_{\mathrm{u}}}\right)}^2$, the blue, red, green, and black points represent the data points for Co/Ti, Co/Cu, Co/Al, and $\mathrm{Co}/\mathrm{Si}{\mathrm{O}}_2$, respectively. (b) The enlarged figure in the block broken square. Surface energy density dependence of the NM materials selected as the seeding layers on (c) ${\beta }_{\mathrm{TMS}}$ and (d) $h_{{\mathrm{S}}_{\mathrm{u}}}$, where the black broken lines in (c) and (d) show the values of the $\mathrm{Mg}\mathrm{O}/\mathrm{Co}/\mathrm{Si}{\mathrm{O}}_2$.

Figure 5

(a) Schematic of the measurement setup used to perform the FMR measurements using the rutile cavity. The external magnetic field was applied in-plane along the sample and the microwave signal was applied via the coil. (b) Frequency dependence of $|S_{11}|$ at a zero external magnetic field, where the red line shows the fitting result obtained using Eq. (5). (c) External magnetic field dependence of ${\kappa }_i$, where the red broken line represents the fitting result obtained using Eq. (6). (d) Color map of $|S_{11}|$ for the external magnetic fields and microwave frequencies.