Interfacial exchange coupling in Fe-Tb/[Co/Pt] heterostructures

The exchange interaction and magnetization reversal processes in exchange-biased heterostructures consisting of amorphous ferrimagnetic Fe-Tb alloy films and ferromagnetic Co/Pt multilayers were investigated. Both film systems exhibit an easy axis magnetization perpendicular to the film plane. The dependence of the interfacial exchange coupling on the stoichiometry of the Fe-Tb layer as well as the number of repetitions in the Co/Pt multilayers were analyzed. The net magnetization and effective magnetic anisotropy of the Fe-Tb alloy films have an influence on the exchange energy per unit area. A large exchange-bias field up to 8 kOe is found to be accompanied by an interfacial domain wall as probed by element specific x-ray magnetic circular dichroism absorption measurements. This domain wall exhibits a total thickness between 3 and 4 nm and affects strongly the magnetization reversal in the heterostructure. Furthermore, a linear relation between the thickness of the ferromagnetic Co/Pt multilayers and the exchange-bias field is observed and can be attributed to a change of the location where the interfacial domain wall nucleates in the heterostructure.

Figure 1

(a) General schematic view of the magnetic configuration in heterostructures with PMA consisting of FI amorphous Fe-Tb alloy films and FM Co/Pt multilayers. The black and the blue arrows represent the collinear alignment of the Fe and Co sublattices, respectively. In addition a drawing of a sample layer stack is given on the right. The red arrows with different length and opening angles θ illustrate the magnetic moment of the Tb sublattices and their average orientational distribution depending on the composition. (b)–(d) Average magnetic configuration illustrated by macrospins depending on the composition of the Fe-Tb alloy film. The opening angles ${\theta }_{1...3}$ of the Tb macro spins refer to the different fanning cone structures of the Tb sublattice. (b) The net magnetization of the Fe-Tb alloy film is dominated by the Tb sublattice. (c) The magnetic sublattices of the Fe and the Tb compensate each other and the net magnetization is zero. (d) The net magnetization of the Fe-Tb alloy film is dominated by the Fe sublattice. Please note that the different states depend on the Tb content and temperature.

Figure 2

TEM images with two different magnifications obtained in cross section geometry of a Fe${}_{81}$Tb${}_{19}$(20 nm)/[Co(0.4 nm)/Pt(0.8 nm)]${}_{10}$ heterostructure embedded in Pt seed and capping layers revealing some roughness and intermixing at the Fe-Tb/[Co/Pt] interface.

Figure 3

Out-of-plane remanence net magnetization as a function of the temperature and the composition obtained from 20-nm-thick Fe${}_{100-x}$Tb${}_x$ single layers used as reference samples for the heterostructures. The remanence magnetization was measured from 4 to 400 K after saturating the samples in an external magnetic field of 70 kOe at RT. The superconducting magnet of the SQUID VSM was reset by heating over its transition temperature before the measurement to avoid the influence of a small magnetic field coming from the remanence of the magnet. The positive and negative region of the net magnetization refers to a dominant Fe and Tb sublattice, respectively.

Figure 4

(a) The effective magnetic anisotropy and (b) the saturation magnetization of 20-nm-thick Fe${}_{100-x}$Tb${}_x$ alloy films as a function of the Tb content for various temperatures. These single layers serve as reference samples. The dashed curves act as a guide to the eye. The inset reveals the composition dependence of the compensation temperature.

Figure 5

(a) Hysteresis loops of Fe${}_{81}$Tb${}_{19}$(20 nm)/[Co(0.4 nm)/Pt(0.8 nm)]${}_{10}$ heterostructures measured in out-of-plane geometry at different temperatures using a SQUID-VSM and (b) switching field of the Fe-Tb layer ($H_{\mathrm{S},\mathrm{Fe}-\mathrm{Tb}}$) and the Co/Pt multilayers ($H_{\mathrm{S},\mathrm{Co}/\mathrm{Pt}}$) taken from the right branch of the hysteresis loop as a function of the temperature compared to the temperature dependence of $K_{\mathrm{eff}}$ of the single reference layers (inset).

Figure 6

Minor loop reversal taken from out-of-plane SQUID-VSM measurements and a schematic drawing of the assumed domain configuration during the magnetization reversal in Fe${}_{81}$Tb${}_{19}$(20 nm)/[Co(0.4 nm)/Pt(0.8 nm)]${}_{10}$ heterostructures.

Figure 7

Element specific full and minor hysteresis loops of Fe${}_{83}$Tb${}_{17}$(20 nm)/[Co(0.4 nm)/Pt(0.8 nm)]${}_{10}$ heterostructures obtained from XMCD absorption measurements at the $L_3$ edge of Fe (708 eV) and Co (778 eV) and the $M_5$ edge of Tb (1241 eV) of 160 K. The percentages indicate the portion of the XMCD signal related to the IDW.

Figure 8

(a) Hysteresis loops of Fe${}_{75.5}$Tb${}_{24.5}$(20 nm)/[Co(0.4 nm)/Pt(0.8 nm)]${}_{10}$ heterostructures measured in out-of-plane geometry at different temperatures using a SQUID VSM and (b) switching field of the Fe-Tb layer ($H_{\mathrm{S},\mathrm{Fe}-\mathrm{Tb}}$) and the Co/Pt multilayers ($H_{\mathrm{S},\mathrm{Co}/\mathrm{Pt}}$) as a function of temperature compared to the temperature dependence of $K_{\mathrm{eff}}$ of the single reference layers (inset). Please note that the increase in the switching field towards higher temperatures is related to the compensation point of the Fe-Tb layer, which is close to the Curie temperature.

Figure 9

Minor loop reversal taken from out-of-plane SQUID-VSM measurements and a schematic drawing of the assumed domain configuration during the magnetization reversal in Fe${}_{75.5}$Tb${}_{24.5}$(20 nm)/[Co(0.4 nm)/Pt(0.8 nm)]${}_{10}$ heterostructures.

Figure 10

(a) The interfacial exchange energy density and (b) the magnitude of the EB field as a function of the Tb content obtained from Fe${}_{100-x}$Tb${}_x$/[Co/Pt]${}_{10}$ heterostructures at 10 and 120 K. The dashed curves act as a guide to the eye. The error for the EB field lies within the size of the symbols. And the scattering of data points is most likely related to the interface roughness induced by changes in growth conditions depending on the Tb content of the alloy. Please note that the sign of the EB field changes at a composition of about 22 at.$\%$ due to the transition of the dominant magnetic moment from Fe to Tb in the Fe-Tb alloy film as shown in the inset.

Figure 11

(a) EB field $H_{\mathrm{EB},\mathrm{Co}/\mathrm{Pt}}$ of the Co/Pt multilayers at 10, 40, 120, and 160 K and (b) the total as well as partial IDW thickness ($t_{\mathrm{IDW}}$ $=$ $t_{\mathrm{IDW},\mathrm{Co}/\mathrm{Pt}}$ $+$ $t_{\mathrm{IDW},\mathrm{Fe}-\mathrm{Tb}}$) related to both layers of the heterostructure at 160 K as a function of the nominal thickness of the multilayer stack obtained from SQUID-VSM and XMCD absorption measurements. The dashed lines act as a guide to the eye.