Hydration of gadolinium oxide ($\mathrm{Gd}{\mathrm{O}}_x$) and its effect on voltage-induced Co oxidation in a $\mathrm{Pt}/\mathrm{Co}/\mathrm{Gd}{\mathrm{O}}_x/\mathrm{Au}$ heterostructure

Magneto-ionic control of magnetism has garnered great interest in recent years due to the large magnetic changes that can be induced using a relatively small voltage. One model structure for this is $\mathrm{Pt}/\mathrm{Co}/\mathrm{Gd}{\mathrm{O}}_x/\mathrm{Au}$, where Co is the magnetic layer and $\mathrm{Gd}{\mathrm{O}}_x$ is the ionic conductor, with the magnetic properties dependent on the oxidation state of Co. While this structure is commonly used, there is limited understanding of the effect of $\mathrm{Gd}{\mathrm{O}}_x$ properties on voltage-induced magnetic changes. In this work, we show that hydration of $\mathrm{G}{\mathrm{d}}_2{\mathrm{O}}_3$ to form $\mathrm{Gd}{\left(\mathrm{OH}\right)}_3$ is crucial for voltage-induced Co oxidation in a $\mathrm{Pt}/\mathrm{Co}/\mathrm{Gd}{\mathrm{O}}_x/\mathrm{Au}$ device. By examining the rate of Co oxidation in nonhydrated and hydrated devices, we conclude that ${\mathrm{H}}_2\mathrm{O}$ in the $\mathrm{Gd}{\mathrm{O}}_x$ layer acts as an oxidant during the voltage-induced Co oxidation process. Co oxidation through this interfacial reaction process is confirmed by in situ x-ray absorption spectroscopy.

Figure 1

(a) Schematic of a nonhydrated, partially hydrated, and fully hydrated $\mathrm{Gd}{\mathrm{O}}_x$ thin film on a $\mathrm{Si}{\mathrm{O}}_2/\mathrm{Si}$ substrate. (b) SLD of the $\mathrm{Gd}{\mathrm{O}}_x$ thin film as a function of hydration time. The fitted mass densities of $\mathrm{G}{\mathrm{d}}_2{\mathrm{O}}_3$ and $\mathrm{Gd}{\left(\mathrm{OH}\right)}_3$ are 8.3and 6.0 g/cc, respectively. (c) Fitted thickness of $\mathrm{Gd}{\left(\mathrm{OH}\right)}_3$ as a function of hydration time. (d) XRR spectra of a nonhydrated (0 h) and hydrated $\mathrm{Gd}{\mathrm{O}}_x$ thin film (144 h of hydration). The solid lines are the fitted models to the measured data (hollow circles). (e) XPS data of a 3-nm $\mathrm{Gd}{\mathrm{O}}_x$ thin film surface that has been exposed to ambient pressure for >2 weeks.

Figure 2

(a)–(e) MOKE hysteresis loops of nonhydrated Pt(3 nm)/Co(0.9 nm)/$\mathrm{Gd}{\mathrm{O}}_x$(10 nm)/Au(3 nm) device in virgin state (a) and after $V_G=-3\mathrm{V}$ was applied for 600 s in ambient (b), vacuum (c), wet ${\mathrm{N}}_2$ (d), and dry ${\mathrm{O}}_2$ (e). (f)–(j) MOKE hysteresis loops of hydrated Pt(3 nm)/Co(0.9 nm)/$\mathrm{Gd}{\mathrm{O}}_x$(10 nm)/Au(3 nm) device in virgin state (f) and after $V_G=-3\mathrm{V}$ was applied for 600 s in ambient (g), vacuum (h), wet ${\mathrm{N}}_2$ (i), and dry ${\mathrm{O}}_2$ (j). (k)–(l) Schematic of voltage-induced reaction in a nonhydrated (k) and hydrated (l) device. The color gradient in the oxide layer in (l) represents a gradient in the incorporated water content.

Figure 3

(a)–(c) MOKE hysteresis loops of hydrated Pt(3 nm)/Co(0.9 nm)/$\mathrm{Gd}{\mathrm{O}}_x$(10 nm)/Au(15 nm) device in virgin state (a) and after $V_G=-3\mathrm{V}$ was applied for 600 s in ambient (b) and vacuum (c). (d), (e) Optical micrographs of Pt(3 nm)/Co(0.9 nm)/$\mathrm{Gd}{\mathrm{O}}_x$(10 nm)/Au (15 nm) devices before (d) and after (e) bias voltage application ($V_G=-3\mathrm{V}$ for 600 s) showing generation of hydrogen bubbles under the electrode. (f) Optical micrograph of Pt(3 nm)/Co(0.9 nm)/$\mathrm{Gd}{\mathrm{O}}_x$(10 nm)/Au (3 nm) after applying $V_G=-3\mathrm{V}$ for 600 s. The scratch marks on the side of the Au electrodes are due to the CuBe probes.

Figure 4

(a) MOKE hysteresis loop of hydrated Pt(3 nm)/Co(0.9 nm)/$\mathrm{Gd}{\mathrm{O}}_x$(10 nm)/Au(3 nm) after $V_G=-3\mathrm{V}$ was first applied for 300 s in ambient to completely oxidize the metallic Co layer. (b), (c) MOKE hysteresis loops after $V_G=+3\mathrm{V}$ was applied to device in (a) under vacuum (b) and ambient (c) conditions in order to reduce the oxidized Co. (d) MOKE hysteresis loop of hydrated Pt(3 nm)/CoO(0.9 nm)/$\mathrm{Gd}{\mathrm{O}}_x$(10 nm)/Au(3 nm) in virgin state. (e), (f) MOKE hysteresis loops after $V_G=+2\mathrm{V}$ was applied to device in (d) in vacuum (e) and in ambient atmosphere (f).

Figure 5

(a)–(c) XAS spectra of Mg(30nm)/Pd(10 nm)/Co(0.9 nm)/$\mathrm{Gd}{\mathrm{O}}_x$ (30 nm)/Au(3 nm) device in virgin state (a) and after $V_G=+3\mathrm{V}$ was applied for 600 s in vacuum (b) and in 10 Torr of ${\mathrm{P}}_{\mathrm{H}2\mathrm{O}}$ (c), respectively. The experiments from (a) to (c) were done sequentially. (d), (e) XAS spectra of the device in (c) after $V_G=-3\mathrm{V}$ was applied for 600 s in vacuum and in 10 Torr of ${\mathrm{P}}_{\mathrm{H}2\mathrm{O}}$, respectively. Two different devices in (c) were used for experiments in (d) and (e).